IPP - XLsoft.com

® 

 

 

(IPP) 

 

1 : 

 

© 2000-2001 Intel Corporation. 

 

: A24968J-1101 

Web: http://www.intel.co.jp/jp/developer/ ( ) 

http://developer.intel.com ( )

-1001 2000 9 

-1002 IPP1.0 NormDiff 2001 2 

AutoCorr ZeroMean 

-1101 IPP1.1 2001 4 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IntelPentiumMMXItanium Intel Corporation 

 

* 

© 2000-2001 Intel Corporation. 

ii

1 

2 

 

IPP ................................................................................... 1-1 

............................................................ 1-1 

............................................................... 1-2 

..................................................................................................... 1-2 

................................................................................................... 1-2 

................................................................................................... 1-3 

............................................................................................ 1-3 

............................................................................................ 1-3 

....................................................................................................... 1-3 

......................................................................................................... 1-4 

................................................................................................... 1-4 

......................................................................................... 1-4 

....................................................................................................... 1-5 

 

IPP ......................................................................................................... 2-1 

IPP PLS .................................................................................. 2-1 

......................................................................................................... 2-2 

.............................................................................................................. 2-2 

.............................................................................................................. 2-3 

............................................................................................ 2-3 

............................................................................................ 2-5 

.............................................................................................................. 2-5 

................................................................................................. 2-6 

................................................................... 2-6 

..................................................................................... 2-6 

iii

IPP : 1 

 

.............................................................................. 2-6 

............................................................................................ 2-7 

.................................... 2-8 

.......................................................................................... 2-8 

..................................................................................................... 2-9 

3 

4 

5 

 

............................................................................................................ 3-1 

GetLibVersion ............................................................................................... 3-1 

GetStatusString ............................................................................................ 3-3 

.......................................................................................... 3-4 

Malloc ........................................................................................................... 3-4 

Free .............................................................................................................. 3-5 

 

.......................................................................................... 4-1 

Copy ............................................................................................................. 4-1 

Set ................................................................................................................ 4-2 

Zero .............................................................................................................. 4-4 

.............................................................................................. 4-5 

............................................................................................ 4-5 

Tone ............................................................................................................. 4-5 

.............................................................................. 4-7 

Triangle ........................................................................................................ 4-9 

.............................................................................................. 4-11 

RandUniform .............................................................................................. 4-11 

.......................................................................................... 4-12 

RandGauss ................................................................................................ 4-12 

Jaehne ................................................................................ 4-13 

VectorJaehne ............................................................................................. 4-13 

 

....................................................................................... 5-1 

AndC ............................................................................................................ 5-1 

And ............................................................................................................... 5-2 

OrC ............................................................................................................... 5-3 

iv


 

Or ................................................................................................................. 5-4 

XorC ............................................................................................................. 5-5 

Xor ................................................................................................................ 5-6 

Not ................................................................................................................ 5-7 

LShiftC .......................................................................................................... 5-8 

RShiftC ......................................................................................................... 5-9 

.......................................................................................................... 5-11 

AddC .......................................................................................................... 5-12 

Add ............................................................................................................. 5-13 

MulC ........................................................................................................... 5-15 

Mul .............................................................................................................. 5-16 

SubC .......................................................................................................... 5-18 

SubCRev .................................................................................................... 5-20 

Sub ............................................................................................................. 5-21 

DivC ........................................................................................................... 5-23 

Div ............................................................................................................. 5-24 

Abs ............................................................................................................. 5-28 

Sqr .............................................................................................................. 5-29 

Sqrt ............................................................................................................. 5-31 

Exp ............................................................................................................. 5-34 

Ln ............................................................................................................... 5-36 

Normalize ................................................................................................... 5-38 

.......................................................................................................... 5-40 

Convert ....................................................................................................... 5-40 

Conj ............................................................................................................ 5-42 

ConjFlip ...................................................................................................... 5-43 

Magnitude ................................................................................................... 5-44 

Phase ......................................................................................................... 5-46 

Phase ......................................................................................................... 5-47 

PowerSpectr ............................................................................................... 5-48 

PowerSpectr ............................................................................................... 5-49 

Real ............................................................................................................ 5-51 

Imag ........................................................................................................... 5-51 

RealToCplx ................................................................................................. 5-52 

CplxToReal ................................................................................................. 5-53 

v


 

Threshold ................................................................................................... 5-54 

Threshold_LTThreshold_GT ................................................................... 5-58 

Threshold_LTValThreshold_GTVal ........................................................ 5-62 

Threshold_LTInv ........................................................................................ 5-65 

CartToPolar ................................................................................................ 5-68 

CartToPolar ................................................................................................ 5-69 

PolarToCart ................................................................................................ 5-70 

PolarToCart ................................................................................................ 5-71 

MaxOrder ................................................................................................... 5-72 

Preemphasize ............................................................................................ 5-73 

.......................................................................................................... 5-74 

MuLawToLin ............................................................................................... 5-75 

LinToMuLaw ............................................................................................... 5-76 

ALawToLin ................................................................................................. 5-77 

LinToALaw ................................................................................................. 5-78 

MuLawToALaw ........................................................................................... 5-79 

ALawToMuLaw ........................................................................................... 5-80 

(Window) ............................................................................................. 5-81 

(Window) ............................................................................. 5-81 

WinBartlett .................................................................................................. 5-83 

WinBlackman ............................................................................................. 5-85 

WinHamming .............................................................................................. 5-88 

WinHann ..................................................................................................... 5-90 

WinKaiser ................................................................................................... 5-92 

.......................................................................................................... 5-95 

Sum ............................................................................................................ 5-95 

Max ............................................................................................................. 5-97 

MaxIndx ...................................................................................................... 5-97 

Min .............................................................................................................. 5-98 

MinIndx ....................................................................................................... 5-99 

Mean ........................................................................................................ 5-101 

StdDev ...................................................................................................... 5-102 

Norm ......................................................................................................... 5-104 

NormDiff ................................................................................................... 5-106 

DotProd .................................................................................................... 5-108 

vi


 

.......................................................................................... 5-111 

SampleUp ................................................................................................. 5-111 

SampleDown ............................................................................................ 5-113 

6 

 

............................................................................... 6-1 

Conv ............................................................................................................. 6-1 

ConvCyclic ................................................................................................... 6-3 

AutoCorr ....................................................................................................... 6-4 

CrossCorr ..................................................................................................... 6-6 

.......................................................................................... 6-9 

FIR ......................................................................................... 6-9 

FIRInitAllocFIRMRInitAlloc ..................................................................... 6-10 

FIRFree ...................................................................................................... 6-14 

FIROne ....................................................................................................... 6-15 

FIR .............................................................................................................. 6-17 

FIRGetTaps ................................................................................................ 6-22 

FIRGetDlyLine, FIRSetDlyLine ................................................................... 6-23 

FIR LMS ............................................................................... 6-26 

FIRLMSInitAlloc .......................................................................................... 6-27 

FIRLMSFree ............................................................................................... 6-28 

FIRLMSOne ............................................................................................... 6-28 

FIRLMS ...................................................................................................... 6-31 

FIRLMSGetTaps ........................................................................................ 6-33 

FIRLMSGetDlyLineFIRLMSSetDlyLine .................................................. 6-34 

IIR ........................................................................................ 6-35 

IIRInitAllocIIRInitAlloc_BiQuad ............................................................... 6-35 

IIRFree ....................................................................................................... 6-39 

IIROne ........................................................................................................ 6-40 

IIR ............................................................................................................... 6-42 

IIRGetDlyLineIIRSetDlyLine ................................................................... 6-48 

........................................................................ 6-51 

FilterMedian ................................................................................................ 6-51 

vii


 

7 

 

.............................................................................................. 7-1 

......................................................................................... 7-1 

flag hint ................................................................................ 7-1 

Pack ................................................................................................ 7-2 

Perm ................................................................................................ 7-3 

CCS ................................................................................................. 7-3 

................................................................... 7-3 

ConjPerm ..................................................................................................... 7-4 

ConjPack ...................................................................................................... 7-6 

ConjCCS ...................................................................................................... 7-8 

............................................................................ 7-10 

MulPackMulPerm .................................................................................... 7-10 

............................................................................... 7-13 

FFTInitAlloc_CFFTInitAlloc_R ................................................................ 7-13 

FFTFree_CFFTFree_R ........................................................................... 7-15 

FFTGetBufSize_CFFTGetBufSize_R ..................................................... 7-16 

FFTFwd_CToCFFTInv_CToC ................................................................ 7-17 

FFTFwd_RToPermFFTInv_PermToRFFTFwd_RToPack 

FFTInv_PackToRFFTFwd_RToCCSFFTInv_CCSToR .................... 7-20 

............................................................................... 7-24 

DFTInitAlloc_CDFTInitAlloc_R ............................................................... 7-24 

DFTFree_CDFTFree_R .......................................................................... 7-26 

DFTGetBufSize_CDFTGetBufSize_R .................................................... 7-27 

DFTFwd_CToCDFTInv_CToC ............................................................... 7-28 

DFTFwd_RToPermDFTInv_PermToRDFTFwd_RToPack 

DFTInv_PackToRDFTFwd_RToCCSDFTInv_CCSToR ................... 7-31 

DFT(Goertzel) .............................................................. 7-35 

Goertz ......................................................................................................... 7-35 

GoertzTwo .................................................................................................. 7-38 

............................................................................... 7-39 

DCTFwdInitAllocDCTInvInitAlloc ............................................................ 7-39 

DCTFwdFreeDCTInvFree ....................................................................... 7-40 

DCTFwdGetBufSizeDCTInvGetBufSize ................................................. 7-41 

DCTFwdDCTInv ..................................................................................... 7-43 

viii


 

................................................................................. 7-46 

.......................................................... 7-49 

WTHaarFwdWTHaarInv ......................................................................... 7-49 

................................................... 7-54 

WTFwdInitAllocWTInvInitAlloc ................................................................ 7-54 

WTFwdFreeWTInvFree .......................................................................... 7-57 

WTFwd ....................................................................................................... 7-58 

WTFwdSetDlyLineWTFwdGetDlyLine .................................................... 7-61 

WTInv ......................................................................................................... 7-64 

WTInvSetDlyLineWTInvGetDlyLine ........................................................ 7-67 

8 

 

.......................................................................................... 8-1 

....................................................................................................... 8-1 

AddAllRowSum ........................................................................................... 8-1 

AddMulCol .................................................................................................... 8-3 

AddMulRow .................................................................................................. 8-4 

DotProdCol ................................................................................................... 8-5 

MulCol .......................................................................................................... 8-6 

QRTransCol ................................................................................................. 8-7 

SumCol ......................................................................................................... 8-8 

SumColAbs ................................................................................................ 8-10 

SumColSqr ................................................................................................. 8-11 

RecSqrt ...................................................................................................... 8-12 

SubRow ...................................................................................................... 8-13 

SumRow ..................................................................................................... 8-14 

SumRowAbs ............................................................................................... 8-15 

SumRowSqr ............................................................................................... 8-16 

AddNRows ................................................................................................. 8-17 

ZeroMean ................................................................................................... 8-19 

LMThreshold .............................................................................................. 8-20 

..................................................................................................... 8-21 

MeanCol ..................................................................................................... 8-21 

VarCol ........................................................................................................ 8-22 

MeanVarCol ............................................................................................... 8-24 

ix


 

MeanVarAcc ............................................................................................... 8-25 

NormalizeCol .............................................................................................. 8-26 

................................................................................................... 8-27 

..................................................................................................... 8-27 

LogAdd ....................................................................................................... 8-27 

LogSub ....................................................................................................... 8-28 

......................................................................................................... 8-29 

CopyCol ...................................................................................................... 8-31 

EvalDelta .................................................................................................... 8-32 

Delta ........................................................................................................... 8-33 

DeltaDelta ................................................................................................... 8-35 

............................................................................................... 8-37 

..................................................................................................... 8-37 

LogGaussSingle ......................................................................................... 8-37 

MahDistSingle ............................................................................................ 8-38 

.............................................................................................. 8-39 

LogGauss1 ................................................................................................. 8-39 

LogGauss2 ................................................................................................. 8-41 

MahDist1 .................................................................................................... 8-42 

MahDist2 .................................................................................................... 8-44 

LogGaussMax1 ........................................................................................ 8-45 

LogGaussAdd1 ......................................................................................... 8-47 

.......................................................................................... 8-48 

GaussianDist .............................................................................................. 8-48 

GaussianMerge .......................................................................................... 8-50 

GaussianSplit ............................................................................................. 8-51 

 

 

 

 

2-1 IPP ............................... 2-4 

2-2 ................................... 2-9 

x


 

7-1 flag ............................................................ 7-2 

7-2 hint ............................................................ 7-2 

7-3 PackPerm CCS FFT ........ 7-3 

7-4 FFT ................................... 7-5 

7-5 Pack ....................................................... 7-7 

7-6 CCS ........................................................ 7-9 

7-7 PackPerm CCS FFT ...... 7-22 

 

 

7-1 ...................... 7-47 

7-2 3 ................................................ 7-48 

7-3 3 ............................................. 7-48 

7-4 1 ............................................. 7-61 

7-5 1 ............................................. 7-67 

3-1 ippsGetLibVersion ....................................................... 3-2 

3-2 ippGetStatusString ....................................................... 3-3 

4-1 ippsCopy ...................................................................... 4-2 

4-2 ippsSet ......................................................................... 4-3 

4-3 ippsZero ....................................................................... 4-5 

4-4 ippsVectorJaehne ...................................................... 4-14 

5-1 ....................................................... 5-11 

5-2 ippsAdd ...................................................................... 5-15 

5-3 ippsDiv_16s_ISfs ....................................................... 5-27 

5-4 ippsDiv_32f_I ............................................................. 5-27 

5-5 ippsAbs ...................................................................... 5-29 

5-6 ippsSqr ....................................................................... 5-31 

5-7 ippsSqrt ...................................................................... 5-33 

5-8 ippsExp_16s_ISfs ...................................................... 5-36 

5-9 ippsExp_64f_I ............................................................ 5-36 

5-10 ippsLn ........................................................................ 5-38 

5-11 ippsMagnitude ........................................................... 5-46 

5-12 ippsThreshold ............................................. 5-57 

5-13 ippsThreshold ......................................... 5-58 

xi


 

5-14 ippsThreshold_LTInv ................................................. 5-67 

5-15 ippsCartToPolar ......................................................... 5-69 

5-16 ............................................................................. 5-81 

5-17 Window FFT ... 5-82 

5-18 ippsWinBartlett ........................................................... 5-84 

5-19 ippsWinBlackmanStd ................................................. 5-88 

5-20 ippsWinHamming ....................................................... 5-90 

5-21 ippsWinHann ............................................................. 5-92 

5-22 ippsWinKaiser ............................................................ 5-94 

5-23 ippsSum ..................................................................... 5-96 

5-24 ippsMin ippsMax .......................................... 5-100 

5-25 ippsMean ................................................................. 5-102 

5-26 ippsStdDev .............................................................. 5-104 

5-27 ippsNorm ................................................................. 5-108 

5-28 ippsDotProd ... 5-110 

5-29 ippsSampleDown ..................................................... 5-115 

6-1 ippsConv 2 .......................... 6-2 

6-2 ippsCrossCorr .............................................................. 6-8 

6-3 ippsFIR .................... 6-21 

6-4 ippsLMSOne .............................................................. 6-30 

6-5 ippsFIRLMS ......................................... 6-32 

6-6 ippsIIR_32f 60Hz ........................................ 6-46 

6-7 ippsIIR ................................ 6-47 

6-8 ippsFilterMedian ...... 6-53 

7-1 ippsConjPerm .............................................................. 7-5 

7-2 ippsConjPack ............................................................... 7-7 

7-3 ippsConjCCS ............................................................... 7-9 

7-4 ippsMulPack .............................................................. 7-12 

7-5 ippsFFTFwd_RToCCS .............................................. 7-23 

7-6 ippsDFTFwd_RToCCS .............................................. 7-34 

7-7 Goertzel .......................... 7-37 

7-8 ippsDCTFwd ippsDCTInv .................... 7-45 

7-9 ippsWTHaarFwd ........................................................ 7-53 

xii

1 

1 ® IPP(Integrated Performance 

Primitives) IPP 

1 IPP 2 IPP 

 

(IA)MMX ® SIMD 

 

IPP 1 (1D) 

 

IPP 


 



IPP MMX SSE SIMD 

(single-instruction, multiple-data) MMX SSE 

 

 

 

IPP Microsoft* Windows* 95Windows 98Windows 2000Windows NT* 

Windows XP 

IPP C C++ 

 

1-1

IPP : 1 1 

 

IPP Windows 

ANSI C 


 

 

 

IPP IPP 


( 3 8 ) 

 

1 IPP 

 

2 

 

3 copy set 

 

4 

5 

6 

 

7 ( ) 

 

8 

 

1-2


 

 

 

 

3 8 ( ipps 

) 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/ 

C 

PDF Adobe* Acrobat* 

(Adobe Acrobat 

) 

 

 

1-3


 

 

 

• ( ) 

• 

• 

 

THIS TYPE STYLE 

This type style 

This type style 


: IPP_MAX_64S 

IppLibraryVersion 

 

 

: void ippsFree() 

 

: val, srcLen 

 

1D x(n) 

x[n] 

 

x[n], 0 ≤ n < len 

len 

n 

pDst[n] = pSrc[n] + val pDst pDst[n] n (0 

len-1) 

 

1-4


 


• 

: IPP_MIN_64S 

• Ipps 

IPP Ipp 

 

: IppsROIIppLibraryVersion 

• ipps 


 

ipps 

 

 

• 

: ippsAddAllRowSum 

IPP 2-2 

1-5


1-6

2 





 


MMX 

 



 

IPP PLS 



 

• IPP 

 


 

 


 

• OS 

2-1


 


 


 

IPP IA-32 

Itanium ® 

StrongARM* IPP 

ANSI C 

 

 


ipps ippi 

 

 

ipp_[_descriptor](); 

 

 


 

s ( ) 

i ( ) 

ipps 

2-2


 

SetAddCopy 

 

ippGetStatusString 

 

 

= [_modifier] 

 

ippsFFTInv_CToC_32fc CToC 

 

 

 

[c] 

u “ ”s “ ”f “ ” 

c “ ” 

IPP 2-1 

 

Ipp 

8 Ipp8s 

 

2-3


2-1 


 

 


8 unsigned char Ipp8u 

8 signed char Ipp8s 

16 unsigned short Ipp16u 

16 signed short Ipp16s 

16 complex short Ipp16sc 

32 unsigned int Ipp32u 

32 signed int Ipp32s 

32 float Ipp32f 

32 complex float Ipp32fc 

64 __int64 (Windows ) Ipp64s 

long long (Linux ) 

64 double Ipp64f 

64 complex double Ipp64fc 

1 

 

 

 

 

= [src2Depth][dstDepth] 

ippsDotProd_16s16sc_Sfs 16 short 16 

complex short 16 

complex short 

2 

 

2-4


 

 

 

 

 

 

 

( : ippsThreshold) 

 

I 

 

Sfs 

 

 

2 copy 

 

 

 

 

1. 

2. 

3. 

 

• p ( : pPhasepSrcpSeed) 

( : valsrcsrcLen) 

• ( : pSrc 

lenSrcpDlyLine) 

• pSrc src 

pSrc2 srcLen 

pDst dst ( : 

pDstdstLen) / pSrcDst 

: pPhaselengthdownPhase 

pSize 

2-5


 


 

IppLibraryVersion IPP 

 

• major minor 

• NameIPP : “ippsa6” 

• Version : “v0.0 Alpha 0.0.3.3” 

 

IPP 2 ( ) 

Ipp32fc 

 

typedef struct { 

Ipp32f re; 

Ipp32f im; 

} Ipp32fc; 

:Ipp16scIpp32fcIpp64fc 

 

IPP 

 

IppsFFTSpec 

 

 

 

 

 

2-6


 

IppStatus IPP 

 

2-9 

IppCmpOp 5-54 

 

typedef struct { 

ippCmpLess, 

ippCmpLessEq, 

ippCmpEq, 

ippCmpGreaterEq, 

ippCmpGreater 

} IppCmpOp; 

IppCnvrtRnd 5-40 

typedef enum { 

ippRndZero, 

ippRndNear 

} IppCnvrtRnd; 

IppHintAlgorithm 

( ) 

7-1 flag hint 

typedef enum { 

ippAlgHintNone, 

ippAlgHintFast, 

ippAlgHintAccurate 

} IppHintAlgorithm; 

2-7


 

IPP /Zp8 /Zp8 

8 ( 8 ) 

 

IPP ippsMalloc 32 

 

 

 

scaleFactor (scaleFactor 

)Sfs 

 

 

 

 

 

 

2 -scaleFactor 

 

 

200 2 ippsSqr Ipp16s 40000 

32767 

scaleFactor = 1 

20000 20000*2 

 

 

2 ippsSqrt ( ) 1.414 

1 scaleFactor = -3 

11 

11*2 -3 = 1.375 

2-8


 

2-2 


 

 

 

IppStatus 

2-2 IPP 

 

 

 

ippStsToneMagnErr -38 Incorrect magnitude for tone generation 

 

ippStsToneFreqErr -37 Incorrect frequency for tone generating 

 

ippStsTonePhaseErr -36 Incorrect phase for tone generation 

 

ippStsTrnglMagnErr -35 Incorrect magnitude for triangle generation 

 

ippStsTrnglFreqErr -34 Incorrect frequency for triangle generation 

 

ippStsTrnglPhaseErr -33 Incorrect phase for triangle generation 

 

ippStsTrnglAsymErr -32 Incorrect asymmetry for triangle generation 

 

ippStsHugeWinErr -31 Incorrect size of the Kaiser window 

Kaiser window 

ippStsJaehneErr -30 Incorrect value of the magnitude 

 

ippStsStepErr -29 Step value is less or equal zero 

 

ippStsDlyLineIndexErr -28 Incorrect value of the delay line sample index 

 

ippStsStrideErr -27 Stride value is less than the row length 

 

2-9


2-2 

( ) 

ippStsEpsValErr -26 Negative epsilon value error 

 

ippStsScaleRangeErr -25 Scale bounds are out of range 

 

ippStsThresholdErr -24 Incorrect threshold bounds 

 

ippStsWtOffsetErr -23 Invalid offset value of the wavelet filter 

 

ippStsAnchorErr -22 Anchor point is outside the mask 

 

ippStsMaskSizeErr -21 Invalid mask size 

 

ippStsShiftErr -20 Shift value is negative 

 

ippStsSampleFactorErr -19 Sampling factor is less than or equal to zero 

 

ippStsSamplePhaseErr -18 Phase value is out of range, 0 ≤ phase < factor 

0 ≤ < 

ippStsFIRMRFactorErr -17 Incorrect value of the MR FIR sampling factor 

MR FIR 

ippStsFIRMRPhaseErr -16 Incorrect value of the MR FIR sampling phase 

MR FIR 

ippStsRelFreqErr -15 Relative frequency value is out of range 

 

ippStsFIRLenErr -14 Length of a FIR filter is less or equal to zero 

FIR 

ippStsIIROrderErr -13 Order of an IIR filter is less or equal to zero 

IIR 

ippStsResizeFactorErr -12 Resize factor(s) is less or equal to zero 

 

ippStsDivByZeroErr -11 Attempt to divide by zero 

 

ippStsInterpolationErr -10 Invalid interpolation mode 

 

ippStsMirrorFlipErr -9 Invalid flip mode 

 

2-10


2-2 

( ) 

ippStsMoment00ZeroErr -8 Moment value M(0,0) is too small for further calculations 

M(0,0) 

 

ippStsThreshNegLevelErr -7 Negative value of the level in the threshold operation 

 

ippStsContextMatchErr -6 Context parameter does not match the operation 

 

ippStsFftFlagErr -5 Incorrect value of the FFT flag parameter 

FFT 

ippStsFftOrderErr -4 Incorrect value of the FFT order parameter 

FFT 

ippStsMemAllocErr -3 Not enough memory allocated for the operation 

 

ippStsNullPtrErr -2 Null pointer error 

 

ippStsSizeErr -1 Incorrect value of the data size 

 

ippStsNoErr 0 No error 

 

ippStsNoOperation 1 No operation executed 

 

ippStsMisalignedBuf 2 Misaligned pointer encountered in the operation where it 

must be aligned 

 

 

ippStsSqrtNegArg 3 Negative value(s) of the argument in the function Sqrt 

Sqrt 

ippStsInvByZero 4 Inf result. Zero value was encountered by the function 

InvThresh called with zero level 

inf 

InvThresh 

ippStsEvenMedianMaskSize 5 Even size of the Median Filter mask was replaced by an 

odd number 

 

 

ippStsDivByZero 6 Zero value(s) of the divisor in the function Div 

Div 

2-11


2-2 

( ) 

ippStsLnZeroArg 7 Zero value(s) of the argument in the function Ln 

Ln 

ippStsLnNegArg 8 Negative value(s) of the argument in the function Ln 

Ln 

ippStsNanArg 9 Not-a-number (NaN) met in the input 

NaN(Not-a-number) 

Err (ippStsNoErr ) 

 

 

 

 

ippsDiv_8u 

 

ippStsDivByZero 

2-12

3 


 

 

ippsGetLibVersion IPP 


 

ippGetStatusString IPP 

IppStatus 

 

GetLibVersion 

 

 

const IppLibraryVersion* ippsGetLibVersion(void); 

 

ippsGetLibVersion IPP 


 

3-1



major ( ) 

minor ( ) 

Name 

 

Version 

BuildDate 

“v1.2 Beta” “ippsm6” 

“Jul 20 99” 

major = 1, minor = 2, Name = “ippsm6”, 

Version = “v1.2 Beta”, BuildDate = “Jul 20 99” 

3-1 ippsGetLibVersion 

3-1 

ippsGetLibVersion 

void libinfo(void) { 

const IppLibraryVersion* lib = ippsGetLibVersion(); 

printf(“%s %s %d.%d.%d.%d\n”, lib->Name, lib->Version, 

lib->major, lib->minor, lib->majorBuild, lib->build); 

} 

Output: 

ippsa6 v0.0 Alpha 0.0.5.5 

3-2


GetStatusString 

 

 

const char* ippGetStatusString(IppStatus StsCode); 

 

stsCode 

IppStatus 

 

ippGetStatusString IppStatus 

 

 

3-2 ippGetStatusString 

ippsAdd_16s_I NULL - 2 

 

3-2 

ippGetStatusString 

void statusinfo(void) { 

IppStatus st = ippsAddC_16s_I (3, 0, 0); 

printf(“%d : %s\n”, st, ippGetStatusString(st)); 

} 

Output: 

-2, Null Pointer Error 

3-3


 

 


(len) 

 

ippsFree() 

Malloc 

32 

Ipp8u* ippsMalloc_8u(int len); 



Ipp8s* ippsMalloc_8s(int len); 




Ipp32f* ippsMalloc_32f(int len); 

Ipp64f* ippsMalloc_64f(int len); 

Ipp8sc* ippsMalloc_8sc(int len); 




Ipp32fc* ippsMalloc_32fc(int len); 

Ipp64fc* ippsMalloc_64fc(int len); 

3-4


 

len 

 

 

ippsMalloc 32 

 

 

ippsMalloc 

NULL 

ippsFree 

Free 

ippsMalloc 

 

void ippsFree(void* ptr); 

 

ptr 

ptr 

ippsMalloc 

 

ippsFree ippsMalloc 

 

malloc calloc 

ippsFree ippsMalloc 

free 

3-5


3-6

4 

 

 

 

 

2 

 

Copy 

 

IppStatus ippsCopy_8u(const Ipp8u* pSrc, Ipp8u* pDst, int len); 

IppStatus ippsCopy_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len); 

IppStatus ippsCopy_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len); 

IppStatus ippsCopy_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsCopy_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len); 

IppStatus ippsCopy_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len ); 

IppStatus ippsCopy_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len); 

4-1


 

pSrc 

pDst 

len 

pDst 

 

 

 

ippsCopy pSrc pDst 

len 

4-1 ippsCopy 

 

ippStsNoErr 

ippStsNullPtrErr 

ippStsSizeErr 

 

pDst pSrc NULL 

len 

4-1 

ippsCopy 

IppStatus copy(void) { 

char src[] = “to be copied\0”; 

char dst[256]; 

return ippsCopy_8u(src, dst, strlen(src)+1); 

} 

Set 

 

IppStatus ippsSet_8u(Ipp8u val, Ipp8u* pDst, int len); 

IppStatus ippsSet_16s(Ipp16s val, Ipp16s* pDst, int len); 

IppStatus ippsSet_16sc(Ipp16sc val, Ipp16sc* pDst, int len); 

IppStatus ippsSet_32f(Ipp32f val, Ipp32f* pDst, int len); 

4-2


IppStatus ippsSet_32fc(Ipp32fc val, Ipp32fc* pDst, int len); 

IppStatus ippsSet_64f(Ipp64f val, Ipp64f* pDst, int len); 

IppStatus ippsSet_64fc(Ipp64fc val, Ipp64fc* pDst, int len); 





 

pDst 

len 

val 

 

 

pDst 

 

ippsSet pDst len val 

(pDst ) 

4-2 ippsSet 

 

ippStsNoErr 


ippStsSizeErr 

 

pDst NULL 

len 

4-2 

ippsSet 

IppStatus set(void) { 

char src[] = “set”; 

return ippsSet_8u(‘0’, src, strlen(src)); 

} 

4-3


Zero 

 

IppStatus ippsZero_8u(Ipp8u* pDst, int len); 

IppStatus ippsZero_16s(Ipp16s* pDst, int len); 

IppStatus ippsZero_16sc(Ipp16sc* pDst, int len); 

IppStatus ippsZero_32f(Ipp32f* pDst, int len); 

IppStatus ippsZero_32fc(Ipp32fc* pDst, int len); 

IppStatus ippsZero_64f(Ipp64f* pDst, int len); 

IppStatus ippsZero_64fc(Ipp64fc* pDst, int len); 

 

pDst 

len 

 

 

 

ippsZero pDst len pDst 

 

4-3 ippsZero 

 

ippStsNoErr 


ippStsSizeErr 

 

pDst NULL 

len 

4-4


4-3 

ippsZero 

IppStatus zero(void) { 

char src[] = “zero”; 

return ippsZero_8u(src, strlen(src)); 

} 

 

 

 

Jaehne 


( ) 

 

 

 

C sin() 

sin() cos() 

 

 

Tone 

 

 

IppStatus ippsTone_Direct_32f(Ipp32f* pDst, int len, float magn, 

float rfreq, float* pPhase, IppHintAlgorithm hint); 

IppStatus ippsTone_Direct_32fc(Ipp32fc* pDst, int len, float magn, 


4-5


IppStatus ippsTone_Direct_64f(Ipp64f* pDst, int len, float magn, 


IppStatus ippsTone_Direct_64fc(Ipp64fc* pDst, int len, float magn, 


IppStatus ippsTone_Direct_16s(Ipp16s* pDst, int len, Ipp16s magn, 


IppStatus ippsTone_Direct_16sc(Ipp16sc* pDst, int len, Ipp16s magn, 


 

magn 

pPhase 

rfreq 

pDst 

len 

hint 

 

0.0~2π 

 

0.0 0.5 

 

 

 

hint flag 

hint 

 

ippsTone rfreq pPhase magn 

len 

pDst x(n) 

x(n) = magn ⋅ cos(2πn ⋅ rfreq + phase) 

x(n) 

x(n) = magn ⋅ (cos(2πn ⋅ rfreq + phase)+j ⋅ sin(2πn ⋅ rfreq + phase)) 

hint ( 

) hint 7-2 

4-6


 

ippStsNoErr 


ippStsSizeErr 

ippStsToneMagnErr 

ippStsToneFreqErr 

ippStsTonePhaseErr 

 

pDst pPhase NULL 

len 

magn 

rfreq 0.5 

pPhase IPP_2PI 

 

 

( ) 

 

rfreq phase magn h 

x[n] ( ) 

xn [ ] = magn ⋅ ct h 

( 2π ⋅ rfreq ⋅ n + phase) , n = 012… , , , 

rfreq phase mag h 

x[n]( ) 

xn [ ] = magn ⋅ ( ct h ( 2π ⋅ rfreq ⋅ n + phase) 

+ 

j ⋅ st h ( 2π ⋅ rfreq ⋅ n + phase) 

), 

n= 

012… , , , 

ct h () 

H = π + h 

⎧ 2 

⎪– --- ⋅ ⎛α H 

– --- H ⎞, 

0 ≤α ≤H 

⎪ ⎝ 2⎠ 

ct h 

( α) 

= ⎨ 

⎪ 2 

--------------- ⋅ ⎛α – 

2π ---------------- 

+ H ⎞ 

⎪2π 

– H ⎝ 2 ⎠ 

, H ≤α ≤2π 

⎩ 

ct h 

( α + k ⋅ 2π) = ct h 

( α) , k= 

0, 

± 1, 

± 2 , … 

4-7


H = π h = 0 ct h () cos() 

 

ct h 

( H ⁄ 2 + k ⋅ π) = 0, k= 

0, 

± 1, 

± 2 , … 

ct h 

( k ⋅ 2π) = 1, k= 

0, 

± 1, 

± 2 , … 

ct h 

( H + k ⋅ 2π) = – 1, 

k= 

0, 

± 1, 

± 2 , … 

st h () 

⎧ 

2 

--------------- ⋅ α, 

0≤α ≤ 2π --------------- 

– H 

⎪2π 

– H 

2 

⎪ 

⎪ 2 2π – H 

st h ( α) 

= –--- 

⋅ ( α – π) 

, --------------- ≤ α ≤ ---------------- 

2π + H 

⎨ H 

2 

2 

⎪ 

⎪ 2 

2π + H 

⎪--------------- ⋅ ( α – 2π) 

, ---------------- ≤α ≤2π 

⎩2π 

– H 

2 

st h 

( α + k ⋅ 2π) = st h 

( α) , k= 

0, 

± 1, 

± 2 , … 

H = π h = 0 st h () 

 

st h 

( α) = ct h 

( α + ( 3π + h) ⁄ 2) 

st h 

( πk) = 0, k= 

0, 

± 1, 

± 2 , … 

st h 

(( π – h) ⁄ 2+ 

2πk) = 1, k= 

0, 

± 1, 

± 2 , … 

st h 

(( 3π + h) ⁄ 2 + 2πk) = – 1, 

k= 

0, 

± 1, 

± 2 , … 

4-8


Triangle 

 

 

IppStatus ippsTriangle_Direct_64f(Ipp64f* pDst, int len, float magn, 

float rfreq, float asym, float* pPhase); 

IppStatus ippsTriangle_Direct_64fc(Ipp64fc* pDst, int len, float magn, 


IppStatus ippsTriangle_Direct_32f(Ipp32f* pDst, int len, float magn, 


IppStatus ippsTriangle_Direct_32fc(Ipp32fc* pDst, int len, float magn, 


IppStatus ippsTriangle_Direct_16s(Ipp16s* pDst, int len, Ipp16s magn, 


IppStatus ippsTriangle_Direct_16sc(Ipp16sc* pDst, int len, Ipp16s magn, 


 

rfreq 

pPhase 

magn 

asym 

pDst 

len 

0.0 0.5 

 

0.0 2π 

 

 

 

h- π π h=0 

 

 

 

4-9


 

ippsTriangle rfreqpPhase 

magn len 

pDst 

x(n) 

xn [ ] = magn ⋅ ct h ( 2π ⋅ rfreq ⋅ n + phase), n = 012… , , , 

x(n) 

xn [ ] = magn ⋅ ( ct h ( 2π ⋅ rfreq ⋅ n + phase) + j ⋅ st h ( 2π ⋅ rfreq ⋅ n + phase)), n = 012… , , , 

ct h st h 4-7 

 

 

ippStsNoErr 


ippStsSizeErr 

ippStsTrnglMagnErr 

ippStsTrnglFreqErr 

ippStsTrnglPhaseErr 

ippStsTrnglAsymErr 

 

pDst pPhase NULL 

len 

magn 

rfreq 0.5 

pPhase IPP_2PI 

asym -IPP_PI IPP_PI 

4-10


 

 

RandUniform 

 

 

IppStatus ippsRandUniform_Direct_16s(Ipp16s* pDst, int len, Ipp16s low, 

Ipp16s high, unsigned int* pSeed); 

IppStatus ippsRandUniform_Direct_32f(Ipp32f* pDst, int len, Ipp32f low, 

Ipp32f high, unsigned int* pSeed); 

IppStatus ippsRandUniform_Direct_64f(Ipp64f* pDst, int len, Ipp64f low, 

Ipp64f high, unsigned int* pSeed); 

 

pSeed 

low 

high 

pDst 

len 

 

 

 

 

 

 

ippsRandUniform len 

pDst 

 

ippStsNoErr 


ippStsSizeErr 

 

pDst pSeed NULL 

len 

4-11


 

 

 

RandGauss 

 

 

IppStatus ippsRandGauss_Direct_16s(Ipp16s* pDst, int len, Ipp16s mean, 

Ipp16s stdev, unsigned int* pSeed); 

IppStatus ippsRandGauss_Direct_32f(Ipp32f* pDst, int len, Ipp32f mean, 

Ipp32f stdev, unsigned int* pSeed); 

IppStatus ippsRandGauss_Direct_64f(Ipp64f* pDst, int len, Ipp64f mean, 

Ipp64f stdev, unsigned int* pSeed); 

 

pSeed 

mean 

stdev 

pDst 

len 

 

 

 

 

 

 

ippsRandGauss len 

pDst 

 

ippStsNoErr 


ippStsSizeErr 

 

pDst pSeed NULL 

len 

4-12


Jaehne 

Jaehne 

 

VectorJaehne 

Jaehne 

IppStatus ippsVectorJaehne_8u(Ipp8u* pDst, int len, Ipp8u magn); 

IppStatus ippsVectorJaehne_8s(Ipp8s* pDst, int len, Ipp8s magn); 





IppStatus ippsVectorJaehne_32f(Ipp32f* pDst, int len, Ipp32f magn); 

IppStatus ippsVectorJaehne_64f(Ipp64f* pDst, int len, Ipp64f magn); 

 

pDst 

len 

magn 

 

 

 

 

ippsVectorJaehne Jaehne pDst 

magn 

 

0.5πn 2 

pDst[ n] = magn ⋅ sin -------- 

⎝ 

⎛ , 0≤n


 

ippStsNoErr 


ippStsSizeErr 

ippStsJaehneErr 

 

pSrcDst NULL 

len 

magn 

4-4 ippsVectorJaehne 

IppStatus Jaehne (void) { 

Ipp16s buf[100] ; 

return ippsVectorJaehne_16s ( buf, 100, 255 ); 

} 

4-14

5 

(window) 

 

 

IPP( ) 

 

ANDOR XOR 

- : AndCOrCXorC 

- : AndOrXor 

AndC 

AND 

 

IppStatus ippsAndC_8u_I(Ipp8u val, Ipp8u* pSrcDst, int len); 

IppStatus ippsAndC_8u(const Ipp8u* pSrc, Ipp8u val, Ipp8u* pDst, int len); 

IppStatus ippsAndC_16u_I(Ipp16u val, Ipp16u* pSrcDst, int len); 

IppStatus ippsAndC_16u(const Ipp16u* pSrc, Ipp16u val, Ipp16u* pDst, int len); 

5-1


 

val 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

 

ippsAndC val pSrc AND 

pDst 

ippsAndC val pSrcDst 

AND pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

And 

2 AND 

IppStatus ippsAnd_8u_I(const Ipp8u* pSrc, Ipp8u* pSrcDst, int len); 

IppStatus ippsAnd_8u(const Ipp8u* pSrc1, const Ipp8u* pSrc2, 

Ipp8u* pDst, int len); 

IppStatus ippsAnd_16u_I(const Ipp16u* pSrc, Ipp16u* pSrcDst, int len); 

IppStatus ippsAnd_16u(const Ipp16u* pSrc1, const Ipp16u* pSrc2, 


5-2


 

pSrc1pSrc2 

pDst 

pSrc 

pSrcDst 

len 

2 

 

 

 

 

 

 

ippsAnd pSrc1 pSrc2 

AND pDst 

ippsAnd pSrc pSrcDst 

AND pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

OrC 

OR 

 

IppStatus ippsOrC_8u_I(Ipp8u val, Ipp8u* pSrcDst, int len); 

IppStatus ippsOrC_8u(const Ipp8u* pSrc, Ipp8u val, Ipp8u* pDst, int len); 

IppStatus ippsOrC_16u_I(Ipp16u val, Ipp16u* pSrcDst, int len); 

IppStatus ippsOrC_16u(const Ipp16u* pSrc, Ipp16u val, Ipp16u* pDst, int len); 

5-3


 

val 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

 

ippsOrC val pSrc OR 

pDst 

ippsOrC val pSrcDst 

OR pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Or 

2 OR 

IppStatus ippsOr_8u_I(const Ipp8u* pSrc, Ipp8u* pSrcDst, int len); 

IppStatus ippsOr_8u(const Ipp8u* pSrc1, const Ipp8u* pSrc2, 


IppStatus ippsOr_16u_I(const Ipp16u* pSrc, Ipp16u* pSrcDst, int len); 

IppStatus ippsOr_16u(const Ipp16u* pSrc1, const Ipp16u* pSrc2, 


5-4


 

pSrc1pSrc2 

pDst 

pSrc 

pSrcDst 

len 

2 

 

 

 

 

 

 

ippsOr pSrc1 pSrc2 

OR pDst 

ippsOr pSrc pSrcDst 

OR pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

XorC 

XOR 

 

IppStatus ippsXorC_8u_I(Ipp8u val, Ipp8u* pSrcDst, int len); 

IppStatus ippsXorC_8u(const Ipp8u* pSrc, Ipp8u val, Ipp8u* pDst, int len); 

IppStatus ippsXorC_16u_I(Ipp16u val, Ipp16u* pSrcDst, int len); 

IppStatus ippsXorC_16u(const Ipp16u* pSrc, Ipp16u val, Ipp16u* pDst, int len); 

5-5


 

val 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

 

ippsXorC val pSrc XOR 

pDst 

ippsXorC val pSrcDst 

XOR pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Xor 

2 XOR 

IppStatus ippsXor_8u_I(const Ipp8u* pSrc, Ipp8u* pSrcDst, int len); 

IppStatus ippsXor_8u(const Ipp8u* pSrc1, const Ipp8u* pSrc2, 


IppStatus ippsXor_16u_I(const Ipp16u* pSrc, Ipp16u* pSrcDst, int len); 

IppStatus ippsXor_16u(const Ipp16u* pSrc1, const Ipp16u* pSrc2, 


5-6


 

pSrc1pSrc2 

pDst 

pSrc 

pSrcDst 

len 

2 

 

 

 

 

 

 

ippsXor pSrc1 pSrc2 

XOR pDst 

ippsXor pSrc pSrcDst 

XOR pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Not 

NOT 

IppStatus ippsNot_8u_I(Ipp8u* pSrcDst, int len); 

IppStatus ippsNot_8u(const Ipp8u* pSrc, Ipp8u* pDst, int len); 

IppStatus ippsNot_16u_I(Ipp16u* pSrcDst, int len); 

IppStatus ippsNot_16u(const Ipp16u* pSrc, Ipp16u* pDst, int len); 

 

pSrc 

pDst 

 

 

5-7


pSrcDst 

len 

 

 

 

 

ippsNot pSrc NOT 

pDst 

ippsNot pSrcDst 

NOT pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

LShiftC 

 

IppStatus ippsLShiftC_8u_I(int val, Ipp8u* pSrcDst, int len); 

IppStatus ippsLShiftC_8u(const Ipp8u* pSrc, int val, Ipp8u* pDst, int len); 

IppStatus ippsLShiftC_16u_I(int val, Ipp16u* pSrcDst, int len); 

IppStatus ippsLShiftC_16u(const Ipp16u* pSrc, int val, Ipp16u* pDst, int len); 

IppStatus ippsLShiftC_16s_I(int val, Ipp16s* pSrcDst, int len); 

IppStatus ippsLShiftC_16s(const Ipp16s* pSrc, int val, Ipp16s* pDst, int len); 

IppStatus ippsLShiftC_32s_I(int val, Ipp32s* pSrcDst, int len); 

IppStatus ippsLShiftC_32s(const Ipp32s* pSrc, int val, Ipp32s* pDst, int len); 

 

val 

pSrc 

pSrc pSrcDst 

 

 

5-8


pDst 

pSrcDst 

len 

 

 

 

 

 

ippsLShiftC pSrc val 

pDst 

ippsLShiftC pSrcDst val 

pSrcDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

RShiftC 

 

IppStatus ippsRShiftC_8u_I(int val, Ipp8u* pSrcDst, int len); 

IppStatus ippsRShiftC_8u(const Ipp8u* pSrc, int val, Ipp8u* pDst, int len); 

IppStatus ippsRShiftC_16u_I(int val, Ipp16u* pSrcDst, int len); 

IppStatus ippsRShiftC_16u(const Ipp16u* pSrc, int val, Ipp16u* pDst, int len); 

IppStatus ippsRShiftC_16s_I(int val, Ipp16s* pSrcDst, int len); 

IppStatus ippsRShiftC_16s(const Ipp16s* pSrc, int val, Ipp16s* pDst, int len); 

IppStatus ippsRShiftC_32s_I(int val, Ipp32s* pSrcDst, int len); 

IppStatus ippsRShiftC_32s(const Ipp32s* pSrc, int val, Ipp32s* pDst, int len); 

5-9


 

val 

pSrc 

pDst 

pSrcDst 

len 

pSrc pSrcDst 

 

 

 

 

 

 

 

ippsRShiftC pSrc val 

pDst 

ippsRShiftC pSrcDst val 

pSrcDst 

 

 

5-1 

[0...255] 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-10


5-1 

 

void saturate(void) { 

Ipp16s x[8] = {1000, -257, 127, 4, 5, 0, 7, 8}, lo[8], hi[8]; 

IppStatus status = ippsNot_16u((Ipp16u*)x, (Ipp16u*)lo, 8); 

ippsRShiftC_16s_I(15, lo, 8); 

ippsCopy_16s(x, hi, 8); 

ippsSubCRev_16s_ISfs(255, hi, 8, 0); 

ippsRShiftC_16s_I(15, hi, 8); 

ippsAnd_16u_I((Ipp16u*)lo, (Ipp16u*)x, 8); 

ippsOr_16u_I((Ipp16u*)hi, (Ipp16u*)x, 8); 

ippsAndC_16u_I(255, (Ipp16u*)x, 8); 

printf_16s(“saturate =”, x, 8, status); 

} 

Output: 

saturate = 255 0 127 4 5 0 7 8 

 

IPP( ) 

 

2 

 

2 1 

“ ” 1 

 

5-11


AddC 

 

IppStatus ippsAddC_16s_I(Ipp16s val, Ipp16s* pSrcDst, int len); 

IppStatus ippsAddC_32f_I(Ipp32f val, Ipp32f* pSrcDst, int len); 

IppStatus ippsAddC_64f_I(Ipp64f val, Ipp64f* pSrcDst, int len); 

IppStatus ippsAddC_16s_ISfs(Ipp16s val, Ipp16s* pSrcDst, int len, 

int scaleFactor); 

IppStatus ippsAddC_32fc_I(Ipp32fc val, Ipp32fc* pSrcDst, int len); 

IppStatus ippsAddC_64fc_I(Ipp64fc val, Ipp64fc* pSrcDst, int len); 

IppStatus ippsAddC_16sc_ISfs(Ipp16sc val, Ipp16sc* pSrcDst, int len, 


 

val 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

pSrc pSrcDst 

 

 

pSrc[n] + val 

 

 

 

 

2 

 

ippsAddC pSrc val pDst 

 

ippsAddC pSrcDst val 

pSrcDst 

Sfs scaleFactor 

 

5-12


 

ippStsNoErr 


ippStsSizeErr 

 


len 

Add 

2 

IppStatus ippsAdd_16s_I(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len); 

IppStatus ippsAdd_32f_I(const Ipp32f* pSrc, Ipp32f* pSrcDst, int len); 

IppStatus ippsAdd_64f_I(const Ipp64f* pSrc, Ipp64f* pSrcDst, int len); 

IppStatus ippsAdd_16s_ISfs(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len, 


IppStatus ippsAdd_16s(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

Ipp16s* pDst, int len); 

IppStatus ippsAdd_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 

Ipp32f* pDst, int len); 

IppStatus ippsAdd_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


IppStatus ippsAdd_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

Ipp16s* pDst, int len, int scaleFactor); 

IppStatus ippsAdd_32fc_I(const Ipp32fc* pSrc, Ipp32fc* pSrcDst, int len); 

IppStatus ippsAdd_64fc_I(const Ipp64fc* pSrc, Ipp64fc* pSrcDst, int len); 

IppStatus ippsAdd_16sc_ISfs(const Ipp16sc* pSrc, Ipp16sc* pSrcDst, int len, 


IppStatus ippsAdd_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsAdd_32fc(const Ipp32fc* pSrc1, const Ipp32fc* pSrc2, 

Ipp32fc* pDst, int len); 

IppStatus ippsAdd_64fc(const Ipp64fc* pSrc1, const Ipp64fc* pSrc2, 


5-13


IppStatus ippsAdd_16sc_Sfs(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 

Ipp16sc* pDst, int len, int scaleFactor); 

 

pSrc1pSrc2 

pDst 

pSrc 

pSrcDst 

len 

scaleFactor 

 

pSrc2[n] + pSrc1[n] 

 

pSrcDst 

 

 

 

 

2 

 

ippsAdd pSrc1 pSrc2 

pDst 

ippsAdd pSrc pSrcDst 

pSrcDst 


 

5-2 

 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-14


5-2 

ippsAdd 

IppStatus add(void) { 

Ipp16s x[4] = {-1, 32767, 2, 30000}; 

IppStatus st = ippsAdd_16s_ISfs(x, x, 4, 1); 

printf_16s(“add =”, x, 4, st); 

return st; 

} 

Output: 

add = -1 32767 2 30000 

MulC 

 

IppStatus ippsMulC_16s_I(Ipp16s val, Ipp16s* pSrcDst, int len); 

IppStatus ippsMulC_32f_I(Ipp32f val, Ipp32f* pSrcDst, int len); 

IppStatus ippsMulC_64f_I(Ipp64f val, Ipp64f* pSrcDst, int len); 

IppStatus ippsMulC_16s_ISfs(Ipp16s val, Ipp16s* pSrcDst, int len, 


IppStatus ippsMulC_32f(const Ipp32f* pSrc, Ipp32f val, Ipp32f* pDst, int len); 

IppStatus ippsMulC_64f(const Ipp64f* pSrc, Ipp64f val, Ipp64f* pDst, int len); 

IppStatus ippsMulC_32fc_I(Ipp32fc val, Ipp32fc* pSrcDst, int len); 

IppStatus ippsMulC_64fc_I(Ipp64fc val, Ipp64fc* pSrcDst, int len); 

IppStatus ippsMulC_16sc_ISfs(Ipp16sc val, Ipp16sc* pSrcDst, int len, 


5-15


 

val 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

pSrc pSrcDst 

 

 

pSrc[n] * val 

 

 

 

 

2 

 

ippsMulC pSrc val pDst 

 

ippsMulC pSrcDst val 

pSrcDst 


 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Mul 

2 

IppStatus ippsMul_16s_I(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len); 

IppStatus ippsMul_32f_I(const Ipp32f* pSrc, Ipp32f* pSrcDst, int len); 

IppStatus ippsMul_64f_I(const Ipp64f* pSrc, Ipp64f* pSrcDst, int len); 

5-16


IppStatus ippsMul_16s_ISfs(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len, 


IppStatus ippsMul_16s(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsMul_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 


IppStatus ippsMul_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


IppStatus ippsMul_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsMul_32fc_I(const Ipp32fc* pSrc, Ipp32fc* pSrcDst, int len); 

IppStatus ippsMul_64fc_I(const Ipp64fc* pSrc, Ipp64fc* pSrcDst, int len); 

IppStatus ippsMul_16sc_ISfs(const Ipp16sc* pSrc, Ipp16sc* pSrcDst, int len, 


IppStatus ippsMul_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsMul_32fc(const Ipp32fc* pSrc1, const Ipp32fc* pSrc2, 


IppStatus ippsMul_64fc(const Ipp64fc* pSrc1, const Ipp64fc* pSrc2, 


IppStatus ippsMul_16sc_Sfs(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 


 

pSrc1pSrc2 

pDst 

pSrc 

pSrcDst 

len 

scaleFactor 

 

pSrc1[n] * pSrc2[n] 

 

pSrcDst 

 

 

 

 

2 

5-17


 

ippsMul pSrc1 pSrc2 

pDst 

ippsMul pSrc pSrcDst 

pSrcDst 


 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

SubC 

 

IppStatus ippsSubC_16s_I(Ipp16s val, Ipp16s* pSrcDst, int len); 

IppStatus ippsSubC_32f_I(Ipp32f val, Ipp32f* pSrcDst, int len); 

IppStatus ippsSubC_64f_I(Ipp64f val, Ipp64f* pSrcDst, int len); 

IppStatus ippsSubC_16s_ISfs(Ipp16s val, Ipp16s* pSrcDst, int len, 


IppStatus ippsSubC_32fc_I(Ipp32fc val, Ipp32fc* pSrcDst, int len); 

IppStatus ippsSubC_64fc_I(Ipp64fc val, Ipp64fc* pSrcDst, int len); 

IppStatus ippsSubC_16sc_ISfs(Ipp16sc val, Ipp16sc* pSrcDst, int len, 


5-18


 

val 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

pSrc pSrcDst 

 

 

pSrc[n] - val 

 

val pSrcDst 

 

 

2 

 

ippsSubC pSrc val pDst 

 

ippsSubC pSrcDst val 

pSrcDst 


 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-19


SubCRev 

 

( ) 

IppStatus ippsSubCRev_32f_I(Ipp32f val, Ipp32f* pSrcDst, int len); 

IppStatus ippsSubCRev_32fc_I(Ipp32fc val, Ipp32fc* pSrcDst, int len); 

IppStatus ippsSubCRev_64f_I(Ipp64f val, Ipp64f* pSrcDst, int len); 

IppStatus ippsSubCRev_64fc_I(Ipp64fc val, Ipp64fc* pSrcDst, int len); 

IppStatus ippsSubCRev_16s_ISfs(Ipp16s val, Ipp16s* pSrcDst, int len, 


IppStatus ippsSubCRev_16sc_ISfs(Ipp16sc val, Ipp16sc* pSrcDst, int len, 


 

val 

pSrcDst 

len 

scaleFactor 

pSrcDst 

val 

val - pSrcDst[n] 

 

 

2 

 

ippsSubCRev val pSrcDst 

pSrcDst 


 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-20


Sub 

2 

IppStatus ippsSub_16s_I(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len); 

IppStatus ippsSub_32f_I(const Ipp32f* pSrc, Ipp32f* pSrcDst, int len); 

IppStatus ippsSub_64f_I(const Ipp64f* pSrc, Ipp64f* pSrcDst, int len); 

IppStatus ippsSub_16s_ISfs(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len, 


IppStatus ippsSub_16s(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsSub_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 


IppStatus ippsSub_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


IppStatus ippsSub_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsSub_32fc_I(const Ipp32fc* pSrc, Ipp32fc* pSrcDst, int len); 

IppStatus ippsSub_64fc_I(const Ipp64fc* pSrc, Ipp64fc* pSrcDst, int len); 

IppStatus ippsSub_16sc_ISfs(const Ipp16sc* pSrc, Ipp16sc* pSrcDst, int len, 


IppStatus ippsSub_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsSub_32fc(const Ipp32fc* pSrc1, const Ipp32fc* pSrc2, 


IppStatus ippsSub_64fc(const Ipp64fc* pSrc1, const Ipp64fc* pSrc2, 


IppStatus ippsSub_16sc_Sfs(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 


5-21


 

pSrc1 

pSrc2 

pDst 

pSrc 

pSrcDst 

len 

scaleFactor 

pSrc2 

pSrc1 

pSrc2[n] - pSrc1[n] 

 

pSrcDst 

 

pSrc pSrcDst 

 

 

2 

 

ippsSub pSrc2 pSrc1 

pDst 

ippsSub pSrcDst pSrc 

pSrcDst 


 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-22


DivC 

 

IppStatus ippsDivC_32f(const Ipp32f* pSrc, Ipp32f val, Ipp32f* pDst, int len); 

IppStatus ippsDivC_32fc(const Ipp32fc* pSrc, Ipp32fc val, 


IppStatus ippsDivC_64f(const Ipp64f* pSrc, Ipp64f val, Ipp64f* pDst, int len); 

IppStatus ippsDivC_64fc(const Ipp64fc* pSrc, Ipp64fc val, 


IppStatus ippsDivC_16s_Sfs(const Ipp16s* pSrc, Ipp16s val, Ipp16s* pDst, 

int len, int ScaleFactor); 

IppStatus ippsDivC_8u_Sfs(const Ipp8u* pSrc, Ipp8u val, Ipp8u* pDst, 

int len, int ScaleFactor); 

IppStatus ippsDivC_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc val, 

Ipp16sc* pDst, int len, int ScaleFactor); 

IppStatus ippsDivC_32f_I(Ipp32f val, Ipp32f* pSrcDst, int len); 

IppStatus ippsDivC_32fc_I(Ipp32fc val, Ipp32fc* pSrcDst, int len); 

IppStatus ippsDivC_64f_I(Ipp64f val, Ipp64f* pSrcDst, int len); 

IppStatus ippsDivC_64fc_I(Ipp64fc val, Ipp64fc* pSrcDst, int len); 

IppStatus ippsDivC_16s_ISfs(Ipp16s val, Ipp16s* pSrcDst, int len, 

int ScaleFactor); 

IppStatus ippsDivC_8u_ISfs(Ipp8u val, Ipp8u* pSrcDst, int len, 


IppStatus ippsDivC_16sc_ISfs(Ipp16sc val, Ipp16sc* pSrcDst, int len, 


 

val 

pSrc 

pDst 

pSrc pSrcDst 

 

 

pSrc[n] / val 

 

5-23


pSrcDst 

len 

scaleFactor 

val pSrcDst 

 

 

2 

 

ippsDivC pSrc val pDst 

 

ippsDivC pSrcDst val 

pSrcDst 


 

 

ippStsNoErr 


ippStsSizeErr 

ippStsDivByZeroErr 

 


len 

val 

Div 

2 

IppStatus ippsDiv_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 


IppStatus ippsDiv_32fc(const Ipp32fc* pSrc1, const Ipp32fc* pSrc2, 


IppStatus ippsDiv_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


IppStatus ippsDiv_64fc(const Ipp64fc* pSrc1, const Ipp64fc* pSrc2, 


IppStatus ippsDiv_32f_I(const Ipp32f* pSrc, Ipp32f* pSrcDst, int len); 

5-24


IppStatus ippsDiv_32fc_I(const Ipp32fc* pSrc, Ipp32fc* pSrcDst, int len); 

IppStatus ippsDiv_64f_I(const Ipp64f* pSrc, Ipp64f* pSrcDst, int len); 

IppStatus ippsDiv_64fc_I(const Ipp64fc* pSrc, Ipp64fc* pSrcDst, int len); 

IppStatus ippsDiv_16s_ISfs(const Ipp16s* pSrc, Ipp16s* pSrcDst, int len, 


IppStatus ippsDiv_8u_ISfs(const Ipp8u* pSrc, Ipp8u* pSrcDst, int len, 


IppStatus ippsDiv_16sc_ISfs(const Ipp16sc* pSrc, Ipp16sc* pSrcDst, int len, 


IppStatus ippsDiv_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsDiv_8u_Sfs(const Ipp8u* pSrc1, const Ipp8u* pSrc2, Ipp8u* pDst, 

int len, int scaleFactor); 

IppStatus ippsDiv_16sc_Sfs(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 


 

pSrc1 

pSrc2 

pDst 

pSrc 

pSrcDst 

len 

scaleFactor 

pSrc2 

pSrc1 

pSrc2[n] / pSrc1[n] 

pDst 

pSrcDst 

 

pSrc pSrcDst 

 

 

2 

5-25


 

ippsDiv pSrc2 pSrc1 

pDst 

ippsDiv pSrcDst pSrc 

pSrcDst 


 

5-3 

5-4 (x / 0, 0 / 0) 

 

ippStsNoErr 


ippStsSizeErr 

ippStsDivByZero 

 


len 

 

 

 

NaN 

 

+Inf 

-Inf 

5-26


5-3 

ippsDiv_16s_ISfs 

IppStatus div16s( void ) { 

Ipp16s x[4] = { -3, 2, 0, 300 }; 

Ipp16s y[4] = { -2, 2, 0, 0 }; 

IppStatus st = ippsDiv_16s_ISfs( y, x, 4, -1 ); 

printf_16s(“div16s =”, x, 4, st ); 

return st; 

} 

Output: 

-- warning 6, Zero value(s) in the divisor of the function Div 

div16s = 3 2 0 32767 

5-4 

ippsDiv_32f_I 

IppStatus div32f( void ) { 

Ipp32f x[4] = { -3, 2, 0, 300 }; 

Ipp32f y[4] = { -2, 2, 0, 0 }; 

IppStatus st = ippsDiv_32f_I( y, x, 4 ); 

printf_32f( "div32f =", x, 4, st ); 

return st; 

} 

Output: 

-- warning 6, Zero value(s) in the divisor of the function Div 

div32f = 1.500000 1.000000 1.#IND00 1.#INF00 

5-27


Abs 

 

IppStatus ippsAbs_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsAbs_64f_I(Ipp64f* pSrcDst, int len); 

IppStatus ippsAbs_16s_I(Ipp16s* pSrcDst, int len); 

IppStatus ippsAbs_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsAbs_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsAbs_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len); 

 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

ippsAbs pSrc pDst 

 

ippsAbs pSrcDst 

pSrcDst 

ippsMagnitude 

5-5 ippsAbs_32f_I 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-28


5-5 

ippsAbs 

void abs32f(void) { 

Ipp32f x[4] = {-1, 1, 0, 0}; 

x[3] *= (-1); 

ippsAbs_32f_I(x, 4); 

printf_32f(“abs =”, x, 4, ippStsNoErr); 

} 

Output: 

abs = 1.000000 1.000000 0.000000 0.000000 

Sqr 

2 

IppStatus ippsSqr_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsSqr_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsSqr_64f_I(Ipp64f* pSrcDst, int len); 

IppStatus ippsSqr_64fc_I(Ipp64fc* pSrcDst, int len); 

IppStatus ippsSqr_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsSqr_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len); 

IppStatus ippsSqr_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsSqr_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len); 

IppStatus ippsSqr_16s_ISfs(Ipp16s* pSrcDst, int len, int scaleFactor); 

IppStatus ippsSqr_16sc_ISfs(Ipp16sc* pSrcDst, int len, int scaleFactor); 

IppStatus ippsSqr_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


IppStatus ippsSqr_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


5-29


 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

 

 

 

 

 

2 

 

ippsSqr pSrc 2 pDst 

 

pDst[n] = pSrc[n] 2 

ippsSqr pSrcDst 2 

pSrcDst 

pSrcDst[n] = pSrcDst[n] 2 

2 

5-6 200 2 

32767 

 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-30


5-6 

ippsSqr 

IppStatus sqr(void) { 

Ipp16s x[4] = {-3, 2, 30, 200}; 

IppStatus st = ippsSqr_16s_ISfs(x, 4, 1); 

printf_16s(“sqr =”, x, 4, st); 

return st; 

} 

Output: 

sqr = 4 2 450 20000 

Sqrt 

 

IppStatus ippsSqrt_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsSqrt_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsSqrt_64f_I(Ipp64f* pSrcDst, int len); 

IppStatus ippsSqrt_64fc_I(Ipp64fc* pSrcDst, int len); 

IppStatus ippsSqrt_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsSqrt_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len); 

IppStatus ippsSqrt_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsSqrt_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len); 

IppStatus ippsSqrt_16s_ISfs(Ipp16s* pSrcDst, int len, int scaleFactor); 

IppStatus ippsSqrt_16sc_ISfs(Ipp16sc* pSrcDst, int len, int scaleFactor); 

IppStatus ippsSqrt_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsSqrt_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 


5-31


IppStatus ippsSqrt_64s_ISfs(Ipp64s* pSrcDst, int len, int scaleFactor); 

IppStatus ippsSqrt_64s_Sfs(const Ipp64s* pSrc, Ipp64s* pDst, int len, 


 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

 

 

 

 

 

2 

 

ippsSqrt pSrc pDst 

 

pDst[ n] = pSrc[ n] 

ippsSqrt pSrcDst 

pSrcDst 

pSrcDst[ n] = pSrcDst[ n] 

 

a+ 

j ⋅ b 

= 

a 2 + b 2 + a 

-------------------------------- + j ⋅sign( b) 

⋅ 

2 

a 2 + b 2 – a 

-------------------------------- 

2 

 

ippsSqrt 

 

5-7 ippsSqrt_16s_ISfs 

5-32


 

ippStsNoErr 


ippStsSizeErr 

ippStsSqrtNegArg 

 


len 

 

 

NaN 

 

 

0 

5-7 

ippsSqrt 

IppStatus sqrt(void) { 

Ipp16s x[4] = {-3, 2, 30, 300}; 

IppStatus st = ippsSqrt_16s_ISfs(x, 4, -1); 

printf_16s(“sqrt =”, x, 4, st); 

return st; 

} 

Output: 

-- warning 3, Negative value(s) in the argument of the function Sqrt 

sqrt = 0 3 11 35 

5-33


Exp 

e 

IppStatus ippsExp_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsExp_64f_I(Ipp64f* pSrcDst, int len); 

IppStatus ippsExp_16s_ISfs(Ipp16s* pSrcDst, int len, int scaleFactor); 



IppStatus ippsExp_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsExp_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsExp_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, int len, 






IppStatus ippsExp_32f64f(const Ipp32f* pSrc, Ipp64f* pDst, int len); 

 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

 

 

 

pSrcDst 

 

2 

5-34


 

ippsExp pSrc pDst 

 

 

pDst[n] = e pSrc[n] 

ippsExp pSrcDst 

pSrcDst 

 

pSrcDst[n] = e pSrcDst[n] 

 

 

ippsExp_32f64f 

5-8 ippsExp_16s_ISfs 5-9 ippsExp_64f_I 

 

 

ippsExp ippsLn 

 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-35


5-8 

ippsExp_16s_ISfs 

IppStatus exp16s(void) { 

Ipp16s x[4] = {-1, 2, 30, 0}; 

IppStatus st = ippsExp_16s_ISfs(x, 4, -1); 

printf_16s(“exp16s =”, x, 4, st); 

return st; 

} 

Output: 

exp16s = 1 15 32767 2 

5-9 

ippsExp_64f_I 

IppStatus exp64f(void) { 

Ipp64f x[4] = {-1, 2, 1, log(1.234567)}; 

IppStatus st = ippsExp_64f_I(x, 4); 

printf_64f(“exp64f =”, x, 4, st); 

return st; 

} 

Output: 

exp64f = 0.367879 7.389056 2.718282 1.234567 

Ln 

 

IppStatus ippsLn_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsLn_64f_I(Ipp64f* pSrcDst, int len); 

IppStatus ippsLn_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

5-36


IppStatus ippsLn_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsLn_64f32f(const Ipp64f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsLn_16s_ISfs(Ipp16s* pSrcDst, int len, int scaleFactor); 

IppStatus ippsLn_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

 

 

 

 

 

2 

 

ippsLn pSrc pDst 

 

pDst[n] = log e (pSrc[n]) 

ippsLn pSrcDst 

pSrcDst 

pSrcDst[n] = log e (pSrcDst[n]) 

5-10 ippsLn_32f_I 

 

ippStsNoErr 


ippStsSizeErr 

ippStsLnZeroArg 

ippStsLnNegArg 

 


len 

 

 

-Inf 

 

 

NaN 

5-37


5-10 

ippsLn 

IppStatus ln32f(void) { 

Ipp32f x[4] = {-1, (float)IPP_E, 0, (float)(exp(1.234567))}; 

IppStatus st = ippsLn_32f_I(x, 4); 

printf_32f(“Ln =”, x, 4, st); 

return st; 

} 

Output: 

-- warning 8, Negative value(s) of argument in the Ln function 

Ln = -1.#IND00 1.000000 -1.#INF00 1.234567 

Normalize 

 

 

 

IppStatus ippsNormalize_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 

int len, Ipp64fc vsub, Ipp64f vdiv); 

IppStatus ippsNormalize_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len, Ipp32fc vsub, Ipp32f vdiv); 

IppStatus ippsNormalize_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len, Ipp16sc vsub, int vdiv, int scaleFactor); 

IppStatus ippsNormalize_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 

Ipp64f vsub, Ipp64f vdiv); 

IppStatus ippsNormalize_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 

Ipp32f vsub, Ipp32f vdiv); 

IppStatus ippsNormalize_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 

int len, Ipp16s vsub, int vdiv, int scaleFactor); 

5-38


 

pSrc 

vsub 

vdiv 

pDst 

len 

scaleFactor 

 

 

 

 

 

2 

 

ippsNormalize pSrc vsub vdiv 

pDst 

pSrc[ n] – vsub 

pDst[ n] 

= -------------------------------------------- 

vdiv 

 

ippStsNoErr 


ippStsSizeErr 


 

pSrc pDst NULL 

len 

vdiv 

5-39


 

 

• 

• 

• 

• 



ippsReal ippsImag 

 

ippsRealToCplx 2 

1 

ippsCplxToReal 2 

 


Convert 

 

 

2 

 

IppStatus ippsConvert_8s32f(const Ipp8s* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsConvert_8u32f(const Ipp8u* pSrc, Ipp32f* pDst, int len); 


IppStatus ippsConvert_16u32f(const Ipp16u* pSrc, Ipp32f* pDst, int len); 


IppStatus ippsConvert_32f8s_Sfs(const Ipp32f* pSrc, Ipp8s* pDst, int len, 

IppCnvrtRnd rndmode, int scaleFactor); 

5-40


IppStatus ippsConvert_32f8u_Sfs(const Ipp32f* pSrc, Ipp8u* pDst, int len, 




IppStatus ippsConvert_32f16u_Sfs(const Ipp32f* pSrc, Ipp16u* pDst, int len, 




 

pSrc 

pDst 

rndmode 

len 

scaleFactor 

 

 

ippRndZero ippRndNear 

ippRndZero 

 

ippRndNear 

 

 

2 

 

ippsConvert pSrc 

pDst 

Sfs ippsConvert pSrc 

pDst 

 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-41


Conj 

2 

( ) 

IppStatus ippsConj_64fc_I(Ipp64fc* pSrcDst, int len); 

IppStatus ippsConj_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsConj_16sc_I(Ipp16sc* pSrcDst, int len); 

IppStatus ippsConj_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len); 

IppStatus ippsConj_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len); 

IppStatus ippsConj_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len); 

 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

ippsConj pSrc pDst 

 

pDst[n].re = pSrc[n].re 

pDst[n].im = - pSrc[n].im 

ippsConj( ) pSrcDst 

pSrcDst 

 

pSrcDst[n].re = pSrcDst[n].re 

pSrcDst[n].im = - pSrcDst[n].im 

5-42


 

ippStsNoErr 


ippStsSizeErr 

 


len 

ConjFlip 

 

 

IppStatus ippsConjFlip_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 

int len); 

IppStatus ippsConjFlip_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len); 

IppStatus ippsConjFlip_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len); 

 

pSrc 

pDst 

pSrcDst 

len 

pSrc 

 

 

 

 

 

ippsConjFlip pSrc pDst 

 

 

pDst[n] = conj(pSrc[len - n - 1]) 

5-43


ippsConjFlip pSrcDst 

pSrcDst 

 

pSrcDst[n] = conj(pSrcDst[len - n - 1]) 

pSrc pDst 

 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Magnitude 

 

IppStatus ippsMagnitude_32fc(const Ipp32fc* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsMagnitude_64fc(const Ipp64fc* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsMagnitude_16sc32f(const Ipp16sc* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsMagnitude_16sc_Sfs(const Ipp16sc* pSrc, Ipp16s* pDst, 


IppStatus ippsMagnitude_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 


IppStatus ippsMagnitude_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 


IppStatus ippsMagnitude_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 


5-44


 

pSrc 

pSrcRe 

pSrcIm 

pDst 

len 

scaleFactor 

 

 

 

 

 

2 

 

“ ” ippsMagnitude pSrc 

pDst 

magn[ n] = pSrc[ n].re 2 + pSrc[ n].im 2 

“ ” ippsMagnitude pSrcRe 

pSrcIm 

pDst 

magn[ n] = pSrcRe[ n] 2 + pSrcIm[ n] 2 

5-11 sin 2 x +cos 2 x = 1 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-45


5-11 


void magn(void) { 

Ipp64f x[6], magn[4]; 

int n; 

for (n = 0; n> n = 0:9; x = sin(2*pi*n/8); z = [x(1:8)+j*x(3:10)]; abs(z(1:4)) 

Phase 

2 

 

IppStatus ippsPhase_64fc(const Ipp64fc* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsPhase_32fc(const Ipp32fc* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsPhase_16sc32f(const Ipp16sc* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsPhase_16sc_Sfs(const Ipp16sc* pSrc, Ipp16s* pDst, int len, 


 

pSrc 

pDst 

 

( ) 

(- π, π] 

5-46


len 

scaleFactor 

 

2 

 

ippsPhase pSrc pDst 

(- π, π] 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Phase 

2 

 

3 

 

IppStatus ippsPhase_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 


IppStatus ippsPhase_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 


IppStatus ippsPhase_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 


IppStatus ippsPhase_16s32f(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 


 

pSrcRe 

pSrcIm 

pDst 

 

 

( ) 

(- π, π] 

5-47


len 

scaleFactor 

 

2 

 

ippsPhase pSrcRe 

pSrcIm 

pDst (- π, π] 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcRepSrcIm pDst NULL 

len 

PowerSpectr 

2 

 

IppStatus ippsPowerSpectr_64fc(const Ipp64fc* pSrc, Ipp64f* pDst, int len); 

IppStatus ippsPowerSpectr_32fc(const Ipp32fc* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsPowerSpectr_16sc_Sfs(const Ipp16sc* pSrc, Ipp16s* pDst, 


IppStatus ippsPowerSpectr_16sc32f(const Ipp16sc* pSrc, Ipp32f* pDst, int len); 

 

pSrc 

pDst 

len 

scaleFactor 

 

 

 

2 

5-48


 

ippsPowerSpectr pSrc 

pDst 2 

 

pDst[n] = (pSrc[n].re) 2 + (pSrc[n].im) 2 

5-44 ippsMagnitude 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

PowerSpectr 

2 

 

3 

 

IppStatus ippsPowerSpectr_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 


IppStatus ippsPowerSpectr_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 


IppStatus ippsPowerSpectr_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 


IppStatus ippsPowerSpectr_16s32f(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 


5-49


 

pSrcRe 

pSrcIm 

pDst 

len 

scaleFactor 

 

 

 

 

2 

 

ippsPowerSpectr pSrcRe 

pSrcIm 

pDst 

2 

pDst[n] = (pSrcRe[n]) 2 + (pSrcIm[n]) 2 

5-44 ippsMagnitude 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcRepSrcIm pDst NULL 

len 

5-50


Real 

2 

 

IppStatus ippsReal_64fc(const Ipp64fc* pSrc, Ipp64f* pDstRe, int len); 

IppStatus ippsReal_32fc(const Ipp32fc* pSrc, Ipp32f* pDstRe, int len); 

IppStatus ippsReal_16sc(const Ipp16sc* pSrc, Ipp16s* pDstRe, int len); 

 

pSrc 

pDstRe 

len 

 

 

 

 

ippsReal pSrc pDstRe 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

Imag 

2 

 

IppStatus ippsImag_64fc(const Ipp64fc* pSrc, Ipp64f* pDstIm, int len); 

IppStatus ippsImag_32fc(const Ipp32fc* pSrc, Ipp32f* pDstIm, int len); 

IppStatus ippsImag_16sc(const Ipp16sc* pSrc, Ipp16s* pDstIm, int len); 

5-51


 

pSrc 

pDstIm 

len 

 

 

 

 

ippsImag pSrc pDstIm 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

RealToCplx 

2 

 

IppStatus ippsRealToCplx_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 


IppStatus ippsRealToCplx_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 


IppStatus ippsRealToCplx_16s(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 

Ipp16sc* pDst, int len); 

 

pSrcRe 

pSrcIm 

pDst 

len 

 

 

 

 

5-52


 

ippsRealToCplx 2 pSrcRe pSrcIm 

pDst 

pSrcRe NULL 

pSrcIm NULL 

NULL 

 

ippStsNoErr 


ippStsSizeErr 

 

pDst NULL pSrcRe 

pSrcIm NULL 

len 

CplxToReal 

2 

 

IppStatus ippsCplxToReal_64fc(const Ipp64fc* pSrc, Ipp64f* pDstRe, 

Ipp64f* pDstIm, int len); 

IppStatus ippsCplxToReal_32fc(const Ipp32fc* pSrc, Ipp32f* pDstRe, 

Ipp32f* pDstIm, int len); 

IppStatus ippsCplxToReal_16sc(const Ipp16sc* pSrc, Ipp16s* pDstRe, 

Ipp16s* pDstIm, int len); 

 

pSrc 

pDstRe 

pDstIm 

len 

pSrc 

 

 

 

5-53


 

ippsCplxToReal pSrc 2 

pDstRe pDstIm 

 

ippStsNoErr 


ippStsSizeErr 

 

NULL 

len 

Threshold 

level 

 

IppStatus ippsThreshold_32f_I(Ipp32f* pSrcDst, int len, 

Ipp32f level, IppThreshCmp relOp); 

IppStatus ippsThreshold_32fc_I(Ipp32fc* pSrcDst, int len, 


IppStatus ippsThreshold_64f_I(Ipp64f* pSrcDst, int len, 


IppStatus ippsThreshold_64fc_I(Ipp64fc* pSrcDst, int len, 


IppStatus ippsThreshold_16s_I(Ipp16s* pSrcDst, int len, 

Ipp16s level, IppThreshCmp relOp); 

IppStatus ippsThreshold_16sc_I(Ipp16sc* pSrcDst, int len, 


IppStatus ippsThreshold_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 


IppStatus ippsThreshold_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len, 


IppStatus ippsThreshold_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 


5-54


IppStatus ippsThreshold_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len, 


IppStatus ippsThreshold_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


IppStatus ippsThreshold_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


 

pSrc 

pDst 

pSrcDst 

len 

level 

relOp 

 

 

 

 

 

pSrc pSrcDst 

 

 

level 

relOp 

 

ippCmpLess level 

 

ippCmpGreater level 

 

5-55


 

ippsThreshold level pSrc 

 

ippsThreshold level 

pSrcDst 

relOp 

level ippCmpLess 

ippsThreshold 

pDst[ n] 

= 

level,pSrc[ n] < level 

⎧ 

⎨pSrc[ n] , otherwise 

⎩ 

ippsThreshold level ippCmpLess 


pDst[ n] 

⎧pSrc -------------------------, [ n] ⋅ level 

abs( pSrc[ n] 

) 

abs ( pSrc [ n ]) < level 

⎪ 

= ⎨ 

⎪pSrc[ n] , otherwise 

⎩ 

 

level 

 

 


 

(ippCmpLess 

) ( +Inf 

-Inf) (ippCmpGreater ) 

“ ” ippsThreshold_16sc_I 5-13 

 

5-12 “ ” ippsTreshold_16s_I 

5-56


 

ippStsNoErr 


ippStsSizeErr 

 

pSrcpDst pSrcDst NULL 

len 

ippStsThreshNegLevelErr 

level 

5-12 


IppStatus threshold( void ) { 

Ipp16s x[4] = { -1, 0, 2, 3 }; 

IppStatus st = ippsThreshold_16s_I( x, 4, 2, ippCmpLess ); 

printf_16s("threshold result =", x, 4, st ); 

return st; 

} 

Output: 

threshold result = 2 2 2 3 

5-57


5-13 


IppStatus cmplx_threshold(void) { 

Ipp16sc x[4] = {{2,3}, {3,3}, {4,3}, {4,2}}; 

/// level is near to the point {2,3} = 3.6 

/// the point {2,3} is to be replaced 

/// the computed coordinates are {2.2188,3.3282} 

/// the point used is {3,4}; 

/// notice that it is the point with the phase, 

/// nearest to the source 

IppStatus st = ippsThreshold_16sc_I(x, 4, 4, ippCmpLess); 

printf_16sc("complex threshold result =", x, 4, st); 

return st; 

} 

Output: 

complex threshold result = {3, 4} {3, 3} {4, 3} {4, 2} 

Threshold_LTThreshold_GT 

level 

 

IppStatus ippsThreshold_LT_32f_I(Ipp32f* pSrcDst, int len, Ipp32f level); 

IppStatus ippsThreshold_LT_32fc_I(Ipp32fc* pSrcDst, int len, Ipp32f level); 

IppStatus ippsThreshold_LT_64f_I(Ipp64f* pSrcDst, int len, Ipp64f level); 

IppStatus ippsThreshold_LT_64fc_I(Ipp64fc* pSrcDst, int len, Ipp64f level); 

IppStatus ippsThreshold_LT_16s_I(Ipp16s* pSrcDst, int len, Ipp16s level); 

IppStatus ippsThreshold_LT_16sc_I(Ipp16sc* pSrcDst, int len, Ipp16s level); 

IppStatus ippsThreshold_LT_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 

Ipp32f level); 

5-58


IppStatus ippsThreshold_LT_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len, 


IppStatus ippsThreshold_LT_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 


IppStatus ippsThreshold_LT_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len, 


IppStatus ippsThreshold_LT_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len, 

Ipp16s level); 

IppStatus ippsThreshold_LT_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


IppStatus ippsThreshold_GT_32f_I(Ipp32f* pSrcDst, int len, Ipp32f level); 

IppStatus ippsThreshold_GT_32fc_I(Ipp32fc* pSrcDst, int len, Ipp32f level); 

IppStatus ippsThreshold_GT_64f_I(Ipp64f* pSrcDst, int len, Ipp64f level); 

IppStatus ippsThreshold_GT_64fc_I(Ipp64fc* pSrcDst, int len, Ipp64f level); 

IppStatus ippsThreshold_GT_16s_I(Ipp16s* pSrcDst, int len, Ipp16s level); 

IppStatus ippsThreshold_GT_16sc_I(Ipp16sc* pSrcDst, int len, Ipp16s level); 

IppStatus ippsThreshold_GT_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 


IppStatus ippsThreshold_GT_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len, 


IppStatus ippsThreshold_GT_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 


IppStatus ippsThreshold_GT_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len, 


IppStatus ippsThreshold_GT_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


IppStatus ippsThreshold_GT_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


5-59


 

pSrc 

pDst 

pSrcDst 

len 

level 

 

 

 

 

 

pSrc pSrcDst 

 

 

 

level pSrc 

 

level pSrcDst 

 

ippsThreshold_LTippsThreshold_LT 

level ippsThresholdLT 

 

pDst[ n] 

= 

level,pSrc[ n] < level 

⎧ 


⎩ 

ippsThreshold_LT level 

ippsThreshold_LT 

pDst[ n] 

⎧pSrc -------------------------, [ n] ⋅ level 

abs( pSrc[ n] 

) 

abs ( pSrc [ n ]) < level 

⎪ 

= ⎨ 


⎩ 

ippsThreshold_GT ippsThreshold_GT 

level 

ippsThreshold_GT 

pDst[ n] 

= 

level,pSrc[ n] > level 

⎧ 


⎩ 

ippsThreshold_GT level 

5-60


ippsThreshold_GT 

pDst[ n] 

 

level 

 

 

 

 

(ippsThreshold_LT ) 

( +Inf -Inf) 

(ippsThreshold_GT ) 

 

ippStsNoErr 


ippStsSizeErr 

⎧pSrc[ n] ⋅ level 

⎪ 

-------------------------, abs( pSrc[ n] 

) 

abs ( pSrc [ n ]) > level 

= ⎨ 


⎩ 

 


len 


level 

5-61


Threshold_LTValThreshold_GTVal 

level 

 

IppStatus ippsThreshold_LTVal_32f_I(Ipp32f* pSrcDst, int len, Ipp32f level, 

Ipp32f value); 

IppStatus ippsThreshold_LTVal_32fc_I(Ipp32fc* pSrcDst, int len, Ipp32f level, 

Ipp32fc value); 

IppStatus ippsThreshold_LTVal_64f_I(Ipp64f* pSrcDst, int len, Ipp64f level, 


IppStatus ippsThreshold_LTVal_64fc_I(Ipp64fc* pSrcDst, int len, Ipp64f level, 


IppStatus ippsThreshold_LTVal_16s_I(Ipp16s* pSrcDst, int len, Ipp16s level, 

Ipp16s value); 

IppStatus ippsThreshold_LTVal_16sc_I(Ipp16sc* pSrcDst, int len, Ipp16s level, 

Ipp16sc value); 

IppStatus ippsThreshold_LTVal_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

int len, Ipp32f level, Ipp32f value); 

IppStatus ippsThreshold_LTVal_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len, Ipp32f level, Ipp32fc value); 

IppStatus ippsThreshold_LTVal_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsThreshold_LTVal_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 


IppStatus ippsThreshold_LTVal_16s(const Ipp16s* pSrc, Ipp16s* pDst, 

int len, Ipp16s level, Ipp16s value); 

IppStatus ippsThreshold_LTVal_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len, Ipp16s level, Ipp16sc value); 

IppStatus ippsThreshold_GTVal_32f_I(Ipp32f* pSrcDst, int len, Ipp32f level, 


IppStatus ippsThreshold_GTVal_32fc_I(Ipp32fc* pSrcDst, int len, Ipp32f level, 


IppStatus ippsThreshold_GTVal_64f_I(Ipp64f* pSrcDst, int len, Ipp64f level, 


5-62


IppStatus ippsThreshold_GTVal_64fc_I(Ipp64fc* pSrcDst, int len, Ipp64f level, 


IppStatus ippsThreshold_GTVal_16s_I(Ipp16s* pSrcDst, int len, Ipp16s level, 

Ipp16s value); 

IppStatus ippsThreshold_GTVal_16sc_I(Ipp16sc* pSrcDst, int len, Ipp16s level, 

Ipp16sc value); 

IppStatus ippsThreshold_GTVal_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 

Ipp32f level, Ipp32f value); 

IppStatus ippsThreshold_GTVal_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 


IppStatus ippsThreshold_GTVal_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsThreshold_GTVal_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 


IppStatus ippsThreshold_GTVal_16s(const Ipp16s* pSrc, Ipp16s* pDst, 

int len, Ipp16s level, Ipp16s value); 

IppStatus ippsThreshold_GTVal_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len, Ipp16s level, Ipp16sc value); 

 

pSrc 

pDst 

pSrcDst 

len 

level 

value 

 

 

 

 

 

pSrc pSrcDst 

 

 

level 

5-63


 

level pSrc 

 

level pSrcDst 

 

ippsThreshold_LTValippsThreshold_LTVal 

level level value 

 

ippsThreshold_LTVal 

pDst[ n] 

= 

value,pSrc[ n] < level 

⎧ 


⎩ 

ippsThreshold_LTVal level 

ippsThreshold_LTVal 

pDst[ n] 

= 

value, abs( pSrc[ n] 

) < level 

⎧ 


⎩ 

ippsThreshold_GTValippsThreshold_GTVal 

level level value 

 

ippsThreshold_GtVal 

pDst[ n] 

= 

value,pSrc[ n] > level 

⎧ 


⎩ 

ippsThreshold_GTVal level 

ippsThreshold_GTVal 

pDst[ n] 

= 

value, abs( pSrc[ n] 

) > level 

⎧ 


⎩ 

5-64


 

level 

 

ippStsNoErr 


ippStsSizeErr 

 


len 


level 

Threshold_LTInv 

level 

 

 

IppStatus ippsThreshold_LTInv_32f_I(Ipp32f* pSrcDst, int len, Ipp32f level); 

IppStatus ippsThreshold_LTInv_32fc_I(Ipp32fc* pSrcDst, int len, Ipp32f level); 

IppStatus ippsThreshold_LTInv_64f_I(Ipp64f* pSrcDst, int len, Ipp64f level); 

IppStatus ippsThreshold_LTInv_64fc_I(Ipp64fc* pSrcDst, int len, Ipp64f level); 

IppStatus ippsThreshold_LTInv_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 


IppStatus ippsThreshold_LTInv_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len, Ipp32f level); 

IppStatus ippsThreshold_LTInv_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 


IppStatus ippsThreshold_LTInv_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 

int len, Ipp64f level); 

5-65


 

pSrc 

pDst 

pSrcDst 

len 

level 

 

 

 

 

 

pSrc pSrcDst 

 

 

ippsThreshold_LTInv pSrc pDst 

level 

 

ippsThreshold_LTInv pSrcDst 

pSrcDst level 

 

level 

ippsThreshold_LTInv 

 

pDst[ n] 

= 

1 

⎧----------, abs( pSrc[ n] 

) = 0 

⎪level 

⎪ 

⎪ abs( pSrc[ n] 

) 

-------------------------, 0 < abs( pSrc[ n] 

) < level 

⎨pSrc[ n] ⋅ level 

⎪ 

⎪ 1 

⎪------------------------ , otherwise 

pSrc[ n] 

⎩ 

level Inf 

( ) 

pDst[ n] 

⎧Inf,pSrc[ n] = 0 

⎪ 

= ⎨ 

1 

------------------------ , otherwise 

⎪pSrc[ n] 

⎩ 

5-14 ippsThreshold_LTInv_32f_I 

5-66


 

ippStsNoErr 


ippStsSizeErr 

 


len 


level 

ippStsInvZero 

level 

 

Inf 

5-14 

ippsThreshold_LTInv 

IppStatus invThreshold(void) { 

Ipp32f x[4] = {-1, 0, 2, 3}; 

IppStatus st = ippsThreshold_LTInv_32f_I(x, 4, 0); 

printf_32f("inv threshold =", x, 4, st); 

return st; 

} 

Output: 

-- warning 4, INF result. Zero value met by invThreshold with zero level 

inv threshold = -1.000000 1.#INF00 0.500000 0.333333 

5-67


CartToPolar 

 

IppStatus ippsCartToPolar_32fc(const Ipp32fc* pSrc, Ipp32f* pDstMagn, 

Ipp32f* pDstPhase, int len); 

IppStatus ippsCartToPolar_64fc(const Ipp64fc* pSrc, Ipp64f* pDstMagn, 

Ipp64f* pDstPhase, int len); 

 

pSrc 

pDstMagn 

pDstPhase 

len 

 

pSrc () 

 

pSrc ( ) 

(- π, π] 

 

 

ippsCartToPolar pSrc 

( ) pDstMagn ( ) 

pDstPhase 

5-15 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcpDstMagn pDstPhase 

NULL 

len 

5-68


5-15 ippsCartToPolar 

IppStatus cart2polar( void ) { 

Ipp64f cart[6], magn[4], phase[4]; 

int n; 

for (n=0; n


 

pSrcRe 

pSrcIm 

pDstMagn 

pDstPhase 

len 

X/Y 

 

X/Y 

 

( ) 

 

() 

(- π, π] 

 

 

ippsCartToPolar pSrcRe pSrcIm 

/ ( X/Y) ( ) 

pDstMagn ( ) pDstPhase 

 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcRepSrcImpDstMagn 

pDstPhase NULL 

len 

PolarToCart 

 

/ 

 

IppStatus ippsPolarToCart_32fc(const Ipp32f* pSrcMagn, 

const Ipp32f* pSrcPhase, Ipp32fc* pDst, int len); 

IppStatus ippsPolarToCart_64fc(const Ipp64f* pSrcMagn, 

const Ipp64f* pSrcPhase, Ipp64fc* pDst, int len); 

5-70


 

pSrcMagn 

pSrcPhase 

pDst 

len 

( ) 

( ) 

 

()() 

 

 

 

ippsPolarToCart pSrcMagn pSrcPhase 

/ pDst 

 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcMagnpSrcPhase pDst 

NULL 

len 

PolarToCart 

 

/ 

IppStatus ippsPolarToCart_32f(const Ipp32f* pSrcMagn, const Ipp32f* pSrcPhase, 

Ipp32f* pDstRe, Ipp32f* pDstIm, int len); 

IppStatus ippsPolarToCart_64f(const Ipp64f* pSrcMagn, const Ipp64f* pSrcPhase, 

Ipp64f* pDstRe, Ipp64f* pDstIm, int len); 

5-71


 

pSrcMagn 

pSrcPhase 

pDstRe 

pDstIm 

len 

( ) 

 

( ) 

(- π, π] 

X/Y 

 

X/Y 

 

 

 

ippsPolarToCart pSrcMagn pSrcPhase 

/ 

pDstRe pDstIm 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcMagnpSrcPhasepDstRe 

pDstIm NULL 

len 

MaxOrder 

 

IppStatus ippsMaxOrder_64f(const Ipp64f* pSrc, int len, int* pOrder); 

IppStatus ippsMaxOrder_32f(const Ipp32f* pSrc, int len, int* pOrder); 

IppStatus ippsMaxOrder_32s(const Ipp32s* pSrc, int len, int* pOrder); 

IppStatus ippsMaxOrder_16s(const Ipp16s* pSrc, int len, int* pOrder); 

5-72


 

pSrc 

len 

pOrder 

 

 

 

 

ippsMaxOrder pSrc 

pOrder 

 

ippStsNoErr 


ippStsSizeErr 

ippStsNanArg 

 

pSrc pOrder NULL 

len 

NaN 

Preemphasize 

 

( ) 

IppStatus ippsPreemphasize_32f(Ipp32f* pSrcDst, int len, Ipp32f val); 

 

pSrcDst 

len 

val 

 

 

 

 

5-73


 

ippsPreemphasize( ) pSrcDst 

 

y(n) = x(n) - val ⋅ x(n - 1), 

y(n) x(n) val 

val = 0.95 0 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcDst NULL 

len 

 

( ) 

[Rab78] 

 


• 8 µ-lawPCM( 

) 

• 8Α-lawPCM( 

) 

• 8 µ-lawΑ-law 

( ) 

µ-law A-law 

S/N 

 

 

[-1; +1] 

CCITT G.711([CCITT]) µ-law Α-law 

[CCITT] 

5-16 

5-74


MuLawToLin 

8 µ-law 

 

IppStatus ippsMuLawToLin_8u32f(const Ipp8u* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsMuLawToLin_8u16s(const Ipp8u* pSrc, Ipp16s* pDst, int len); 

pSrc 

pDst 

len 

8 µ-law 

 

 

 

 

 

ippsMuLawToLin pSrc 8 µ-law 

PCM pDst 

µ-law 

Cµ ( x ) ln( 1 + 255 ⋅ x ) 

= -------------------------------------- ⋅ 

ln( 256) 

128 , – 1 ≤x ≤1 

x C µ (x) µ-law 

 

 

 

5-75


 

 

µ-law CCITT G.711 

2a/G.711 2b/G.711 

G.711 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

LinToMuLaw 

8 µ-law 

 

IppStatus ippsLinToMuLaw_32f8u(const Ipp32f* pSrc, Ipp8u* pDst, int len); 

IppStatus ippsLinToMuLaw_16s8u(const Ipp16s* pSrc, Ipp8u* pDst, int len); 

pSrc 

pDst 

len 

(1.0 ) 

 

ippsLinToMuLaw 

 

 

ippsLinToMuLaw pSrc PCM 8 µ-law 

pDst 

5-16 ippsLinToMuLaw_32f8u 

5-76


 

ippStsNoErr 


ippStsSizeErr 

 


len 

ALawToLin 

8 Α-law 

 

IppStatus ippsALawToLin_8u32f(const Ipp8u* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsALawToLin_8u16s(const Ipp8u* pSrc, Ipp16s* pDst, int len); 

pSrc 

pDst 

len 

 

ippsALawToLin 

 

 

ippsALawToLin pSrc 8 Α-law 

PCM pDst 

Α-law 

C A 

( x) 

⎧ 87.56 x 

1 --------------------------- + ln87.56 

⋅ 128 

1 

⎪ 

, 0 ≤ x ≤ ------------ 

⎪ 

87.56 

= ⎨ 

⎪ --------------------------------------- 1+ 

ln( 87.56 x ) 

, 

1+ 

ln87.56 

⋅ 128 

1 

⎪ 

, 87.56 

------------ < x ≤ 1 

⎩ 

x C Α (x) Α-law 

 

 

 

5-77


 

 

Α-law CCITT G.711 

1a/G.711 1b/G.711 

G.711 

5-16 ippsALawToLin_8u32f 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

LinToALaw 

8 A-law 

 

IppStatus ippsLinToALaw_32f8u(const Ipp32f* pSrc, Ipp8u* pDst, int len); 

IppStatus ippsLinToALaw_16s8u(const Ipp16s* pSrc, Ipp8u* pDst, int len); 

pSrc 

pDst 

len 

 

ippsLinToALaw 

 

 

ippsLinToALaw pSrc PCM 8 Α-law 

pDst 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-78


MuLawToALaw 

8 µ-law 

8 A-law 

 

IppStatus ippsMuLawToALaw_8u(const Ipp8u* pSrc, Ipp8u* pDst, int len); 

pSrc 

pDst 

len 

8 µ-law 

 

8 Α-law 

 

 

 

ippsMuLawToALaw pSrc 8 µ-law 

8 A-law 

pDst 

 

µ-law A-law CCITT G.711 

3/G.711 G.711 

5-16 ippsMuLawToALaw_8u 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-79


ALawToMuLaw 

8 A-law 

8 µ-law 

 

IppStatus ippsALawToMuLaw_8u(const Ipp8u* pSrc, Ipp8u* pDst, int len); 

pSrc 

pDst 

len 

8 Α-law 

 

8 µ-law 

 

 

 

ippsMuLawToALaw pSrc 8 Α-law 

8 µ-law 

pDst 

 

Α-law µ-law CCITT G.711 

4/G.711 G.711 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

5-80


5-16 

 

void compand( void ) { 

Ipp32f x[4] = { 0.1f, 0.2f, 0.3f, 0.4f }; 

Ipp8u m[4], a[4]; 

ippsLinToMuLaw_32f8u( x, m, 4 ); 

ippsMuLawToALaw_8u( m, a, 4 ); 

ippsALawToLin_8u32f( a, x, 4 ); 

// now x must be close to original 

printf_32f("x =", x, 4, ippStsNoErr); 

} 

Output: 

x = 0.099609 0.207031 0.304688 0.398438 

(Window) 

(Window) (Window) 

 

(Window) FFT 

(Window) 

IPP (Window) 

• Bartlett (Window) 

• Blackman( ) (Window) 

• Hamming( ) (Window) 

• Hann( ) (Window) 

• Kaiser (Window) 

(Window) 


 

5-81


(Window) 

1.0 (Window) 1 ( ippsWinHann) 

(Window) 

1 ( ippsMul) 


5-17 

5-17 

(Window) FFT 

void multiFrameWin( void ) { 

Ipp32f win[LEN], x[LEN], X[LEN]; 

IppsFFTSpec_R_32f* ctx; 

ippsSet_32f( 1, win, LEN ); 

ippsWinHann_32f_I( win, LEN ); 

/// ... initialize FFT context 

while(1 ){ 

/// ... get x signal 

/// 

ippsMul_32f_I( win, x, LEN ); 

ippsFFTFwd_RToPack_32f( x, X, ctx, 0 ); 

} 

} 

 

(Window) [Jac89] 7.3 Windows in Spectrum Analysis[Jac89] 9.1 

Window-Function Technique [Mit93] 16-2 Fourier Analysis of Finite-Time 

Signals 

 

5-82


WinBartlett 

Bartlett (Window) 

IppStatus ippsWinBartlett_16s_I(Ipp16s* pSrcDst, int len); 

IppStatus ippsWinBartlett_16sc_I(Ipp16sc* pSrcDst, int len); 

IppStatus ippsWinBartlett_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsWinBartlett_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsWinBartlett_16s(const Ipp16s* pSrc, Ipp16s* pDst, 

int len); 

IppStatus ippsWinBartlett_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len); 

IppStatus ippsWinBartlett_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

int len); 

IppStatus ippsWinBartlett_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len); 

 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

ippsWinBartlett pSrc Bartlett( ) (Window) 

pDst 

ippsWinBartlett pSrcDst Bartlett( ) 

(Window) pSrcDst 

(Window) 

5-83


Bartlett (Window) 

w bartlett ( n) 

2n 

⎧ -------------------- , 

len – 1 

0 ≤ n ≤ -------------------- 

len – 1 

⎪ 

2 

= ⎨ 

⎪ 2n len – 1 

2 – -------------------- -------------------- 

⎩ 

, 

< n≤ 

len– 

1 

len – 1 2 

5-18 ippsWinBartlett_32f_I 

 

ippStsNoErr 


ippStsSizeErr 

 


len 3 

5-18 

ippsWinBartlett 

void bartlett(void) { 

Ipp32f x[8]; 

ippsSet_32f(1, x, 8); 

ippsWinBartlett_32f_I(x, 8); 

printf_32f(“bartlett (half) =”, x, 4, ippStsNoErr); 

} 

Output: 

bartlett (half) = 0.000000 0.285714 0.571429 0.857143 

Matlab* Analog: 

>> b = bartlett(8); b(1:4)' 

5-84


WinBlackman 

Blackman( ) (Window) 

 

IppStatus ippsWinBlackmanQ15_16s_I(Ipp16s* pSrcDst, int len, 

int alphaQ15); 

IppStatus ippsWinBlackmanQ15_16sc_I(Ipp16sc* pSrcDst, int len, 


IppStatus ippsWinBlackman_16s_I(Ipp16s* pSrcDst, int len, 

float alpha); 

IppStatus ippsWinBlackman_16sc_I(Ipp16sc* pSrcDst, int len, 

float alpha); 

IppStatus ippsWinBlackman_32f_I(Ipp32f* pSrcDst, int len, 

float alpha); 

IppStatus ippsWinBlackman_32fc_I(Ipp32fc* pSrcDst, int len, 

float alpha); 

IppStatus ippsWinBlackmanQ15_16s(const Ipp16s* pSrc, Ipp16s* pDst, 

int len, int alphaQ15); 

IppStatus ippsWinBlackmanQ15_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len, int alphaQ15); 

IppStatus ippsWinBlackman_16s(const Ipp16s* pSrc, Ipp16s* pDst, 

int len, float alpha); 

IppStatus ippsWinBlackman_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 


IppStatus ippsWinBlackman_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsWinBlackman_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 


IppStatus ippsWinBlackmanStd_16s_I(Ipp16s* pSrcDst, int len); 

IppStatus ippsWinBlackmanStd_16sc_I(Ipp16sc* pSrcDst, int len); 

IppStatus ippsWinBlackmanStd_32f_I(Ipp32f* pSrcDst, int len); 

5-85


IppStatus ippsWinBlackmanStd_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsWinBlackmanStd_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len); 

IppStatus ippsWinBlackmanStd_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len); 

IppStatus ippsWinBlackmanStd_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

int len); 

IppStatus ippsWinBlackmanStd_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len); 

IppStatus ippsWinBlackmanOpt_16s_I(Ipp16s* pSrcDst, int len); 

IppStatus ippsWinBlackmanOpt_16sc_I(Ipp16sc* pSrcDst, int len); 

IppStatus ippsWinBlackmanOpt_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsWinBlackmanOpt_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsWinBlackmanOpt_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len); 

IppStatus ippsWinBlackmanOpt_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int len); 

IppStatus ippsWinBlackmanOpt_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

int len); 

IppStatus ippsWinBlackmanOpt_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

int len); 

 

pSrc 

pDst 

pSrcDst 

alpha 

alphaQ15 

len 

 

 

 

 

Blackman window 

alpha scaleFactor 15 

 

5-86


 

ippsWinBlackman pSrc Blackman( ) 

(Window) pDst 

ippsWinBlackman pSrcDst 

Blackman( ) (Window) pSrcDst 

(Window) 

Blackman( ) (Window) 

ippsWinBlackman ippsWinBlackman alpha 


w blackman ( n) 

= 

alpha + 1 

2πn 

----------------------------- 0.5 

2 

cos -------------------- 

⎝ 

⎛ len – 1⎠ 

⎞ alpha 4πn 

– 

– ------------------- 

2 

cos -------------------- 

⎝ 

⎛ len – 1⎠ 

⎞ 

ippsWinBlackmanQ15 ippsWinBlackmanQ15 alphaQ15( 

15) Blackman( ) (Window) 

ippsWinBlackmanStd Blackman( ) (Window) 

ippsWinBlackmanStd alpha 


alpha = -0.16 

ippsWinBlackmanOpt ippsWinBlackmanOpt alpha 

Blackman( ) (Window) 30dB/ 

Window 

0.5 

alpha = –--------------------------------------- 

2π 

1 + cos -------------------- 

len – 1 

len 4 len alpha alpha 

alpha 

alpha = -0.25 

5-19 ippsWinBlackmanStd_32f_I 

5-87


 

ippStsNoErr 


ippStsSizeErr 

 


len ippsWinBalckmanOpt 4 

3 

 

5-19 

ippsWinBlackmanStd 

void blackman(void) { 

Ipp32f x[8]; 


ippsWinBlackmanStd_32f_I(x, 8); 

printf_32f(“blackman (half) =”, x, 4, ippStsNoErr); 

} 

Output: 

blackman(half) = 0.000000 0.090453 0.459183 0.920364 


>> b = blackman(8)'; b(1:4) 

WinHamming 

Hamming( ) (Window) 

 

IppStatus ippsWinHamming_16s_I(Ipp16s* pSrcDst, int len); 

IppStatus ippsWinHamming_16sc_I(Ipp16sc* pSrcDst, int len); 

IppStatus ippsWinHamming_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsWinHamming_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsWinHamming_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len); 

IppStatus ippsWinHamming_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len); 

5-88


IppStatus ippsWinHamming_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsWinHamming_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len); 

 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

ippsWinHamming pSrc Hamming( ) (Window) 

pDst 

ippsWinHamming pSrcDst Hamming( ) 


(Window) 

Hamming( ) (Window) 

2πn 

w hamming ( n) = 0.54 – 0.46 cos -------------------- 

⎝ 

⎛ len – 1⎠ 

⎞ 

5-20 ippsWinHamming_32f_I 

 

ippStsNoErr 


ippStsSizeErr 

 


len 3 

5-89


5-20 

ippsWinHamming 

void hamming(void) { 

Ipp32f x[8]; 


ippsWinHamming_32f_I(x, 8); 

printf_32f(“hamming(half) =”, x, 4, ippStsNoErr); 

} 

Output: 

hamming(half) = 0.080000 0.253195 0.642360 0.954446 


>> b = hamming(8); b(1:4)' 

WinHann 

Hann( ) (Window) 

 

IppStatus ippsWinHann_16s_I(Ipp16s* pSrcDst, int len); 

IppStatus ippsWinHann_16sc_I(Ipp16sc* pSrcDst, int len); 

IppStatus ippsWinHann_32f_I(Ipp32f* pSrcDst, int len); 

IppStatus ippsWinHann_32fc_I(Ipp32fc* pSrcDst, int len); 

IppStatus ippsWinHann_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len); 

IppStatus ippsWinHann_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len); 

IppStatus ippsWinHann_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsWinHann_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len); 

5-90


 

pSrc 

pDst 

pSrcDst 

len 

 

 

 

 

 

 

ippsWinHann pSrc Hann( ) (Window) 

pDst 

ippsWinHann( ) pSrcDst Hann( ) 


(Window) 

Hann( ) (Window) 

2πn 

w hann ( n) = 0.5 – 0.5 cos -------------------- 

⎝ 

⎛ len – 1⎠ 

⎞ 

5-21 ippsWinHann_32f_I 

 

ippStsNoErr 


ippStsSizeErr 

 


len 3 

5-91


5-21 

ippsWinHann 

void hann(void) { 

Ipp32f x[8]; 


ippsWinHann_32f_I(x, 8); 

printf_32f(“hann(half) =”, x, 4, ippStsNoErr); 

} 

Output: 

hann(half) = 0.000000 0.188255 0.611260 0.950484 


>> N = 8; n = 0:N-1; 0.5*(1-cos(2*pi*n/(N-1))) 

WinKaiser 

Kaiser (Window) 

IppStatus ippsWinKaiser_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len, 

float alpha); 

IppStatus ippsWinKaiser_16s_I(Ipp16s* pSrcDst, int len, float alpha); 

IppStatus ippsWinKaiserQ15_16s(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


IppStatus ippsWinKaiserQ15_16s_I(Ipp16s* pSrcDst, int len, int alphaQ15); 

IppStatus ippsWinKaiser_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 

float alpha); 

IppStatus ippsWinKaiser_16sc_I(Ipp16sc* pSrcDst, int len, float alpha); 

IppStatus ippsWinKaiserQ15_16sc(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


IppStatus ippsWinKaiserQ15_16sc_I(Ipp16sc* pSrcDst, int len, int alphaQ15); 

IppStatus ippsWinKaiser_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 

float alpha); 

5-92


IppStatus ippsWinKaiser_32f_I(Ipp32f* pSrcDst, int len, float alpha); 

IppStatus ippsWinKaiser_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len, 

float alpha); 

IppStatus ippsWinKaiser_32fc_I(Ipp32fc* pSrcDst, int len, float alpha); 

IppStatus ippsWinKaiser_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 

float alpha); 

IppStatus ippsWinKaiser_64f_I(Ipp64f* pSrcDst, int len, float alpha); 

 

pSrc 

pDst 

pSrcDst 

alpha 

alphaQ15 

len 

 

 

 

 

Kaiser window 

alpha scaleFactor 15 

 

 

ippsWinKaiser pSrc Kaiser (Window) 

pDst 

ippsWinKaiser pSrcDst Kaiser (Window) 

pSrcDst 

ippsWinKaiser ippsWinKaiser alpha 

(Window) 


w kaiser ( n) 

I ⎛ len – 1 

0 alpha ⎛-------------------- 

⎞ 2 ⎛ len – 1 

n -------------------- 

⎝ 2 ⎠ ⎝ 

– ⎝ 

⎛ 2 ⎠ 

⎞ ⎞ 2 

– 

⎞ 

⎝ 

⎠ ⎠ 

= ------------------------------------------------------------------------------------------------------------------- 

I ⎛ 

0 alpha⎛-------------------- 

len – 1 ⎞⎞ 

⎝ ⎝ 2 ⎠⎠ 

I 0 () 1 Bessel 

ippsWinKaiserQ15 ippsWinKaiserQ15 alphaQ15( 15) 


5-22 ippsWinKaiser_32f_I 

5-93


 

ippStsNoErr 


ippStsSizeErr 

ippStsHugeWinErr 

 

pDstpSrc pSrcDst NULL 

len 1 

Kaiser Window 

5-22 

ippsWinKaiser 

void kaiser(void) { 

Ipp32f x[8]; 

IppStatus st; 


st = ippsWinKaiser_32f_I( x, 8, 1.0f ); 

printf_32f("kaiser(half) =", x, 4, ippStsNoErr); 

} 

Output: 

kaiser(half) = 0.135534 0.429046 0.755146 0.970290 


>> kaiser(8,7/2)' 

5-94


 

( ) IPP 

 

Sum 

 

IppStatus ippsSum_32f(const Ipp32f* pSrc, int len, Ipp32f* pSum); 

IppStatus ippsSum_64f(const Ipp64f* pSrc, int len, Ipp64f* pSum); 

IppStatus ippsSum_32fc(const Ipp32fc* pSrc, int len, Ipp32fc* pSum); 

IppStatus ippsSum_16s32s_Sfs(const Ipp16s* pSrc, int len, Ipp32s* pSum, 


IppStatus ippsSum_16sc32sc_Sfs(const Ipp16sc* pSrc, int len, Ipp32sc* pSum, 


IppStatus ippsSum_16s_Sfs(const Ipp16s* pSrc, int len, Ipp16s* pSum, 


IppStatus ippsSum_16sc_Sfs(const Ipp16sc* pSrc, int len, Ipp16sc* pSum, 


 

pSrc 

pSum 

len 

scaleFactor 

 

 

 

2 

 

ippsSum pSrc pSum 

5-95


pSrc 

sum = 

len – 1 

∑ 

n = 0 

pSrc[ n] 

 

scaleFactor 

 

5-23 ippsSum 

 

ippStsNoErr 


ippStsSizeErr 

 

pSum pSrc NULL 

len 

5-23 

ippsSum 

void sum(void) { 

Ipp16s x[4] = {-32768, 32767, 32767, 32767}, sm; 

ippsSum_16s_Sfs(x, 4, &sm, 1); 

printf_16s(“sum =”, &sm, 1, ippStsNoErr); 

} 

Output: 

sum = 32766 


>> x = [-32768, 32767, 32767, 32767]; sum(x)/2 

5-96


Max 

 

IppStatus ippsMax_32f(const Ipp32f* pSrc, int len, Ipp32f* pMax); 

IppStatus ippsMax_64f(const Ipp64f* pSrc, int len, Ipp64f* pMax); 

IppStatus ippsMax_16s(const Ipp16s* pSrc, int len, Ipp16s* pMax); 

 

pSrc 

pMax 

len 

 

 

 

 

ippsMax pSrc pMax 

5-24 ippsMax_32f 

 

ippStsNoErr 


ippStsSizeErr 

 

pMax pSrc NULL 

len 

MaxIndx 

 

 

IppStatus ippsMaxIndx_16s(const Ipp16s* pSrc, int len, Ipp16s* pMax, 

int* pIndx); 

IppStatus ippsMaxIndx_32f(const Ipp32f* pSrc, int len, Ipp32f* pMax, 

int* pIndx); 

5-97


IppStatus ippsMaxIndx_64f(const Ipp64f* pSrc, int len, Ipp64f* pMax, 

int* pIndx); 

 

pSrc 

pMax 

len 

pIndx 

 

 

 

 

 

ippsMaxIndx pSrc pMax 

pIndx NULL 

pIndx 

 

 

ippStsNoErr 


ippStsSizeErr 

 

pMax pSrc NULL 

len 

Min 

 

IppStatus ippsMin_32f(const Ipp32f* pSrc, int len, Ipp32f* pMin); 

IppStatus ippsMin_64f(const Ipp64f* pSrc, int len, Ipp64f* pMin); 

IppStatus ippsMin_16s(const Ipp16s* pSrc, int len, Ipp16s* pMin); 

 

pSrc 

pMin 

len 

 

 

 

5-98


 

ippsMin pSrc pMin 

5-24 ippsMin_32f 

 

ippStsNoErr 


ippStsSizeErr 

 

pMin pSrc NULL 

len 

MinIndx 

 

 

IppStatus ippsMinIndx_16s(const Ipp16s* pSrc, int len, Ipp16s* pMin, 

int* pIndx); 

IppStatus ippsMinIndx_32f(const Ipp32f* pSrc, int len, Ipp32f* pMin, 

int* pIndx); 

IppStatus ippsMinIndx_64f(const Ipp64f* pSrc, int len, Ipp64f* pMin, 

int* pIndx); 

 

pSrc 

pMin 

len 

pIndx 

 

 

 

 

5-99


 

ippsMinIndx pSrc pMin 

pIndx NULL 

pIndx 

 

 

ippStsNoErr 


ippStsSizeErr 

 

pMin pSrc NULL 

len 

5-24 

ippsMin ippsMax 

void minmax(void) { 

Ipp32f *x = ippsMalloc_32f(1000), minmax[2]; 

int i; 

for (i = 0; i> x = rand(1,1000); min(x), max(x) 

5-100


Mean 

 

IppStatus ippsMean_32f(const Ipp32f* pSrc, int len, Ipp32f* pMean); 

IppStatus ippsMean_32fc(const Ipp32fc* pSrc, int len, Ipp32fc* pMean); 

IppStatus ippsMean_64f(const Ipp64f* pSrc, int len, Ipp64f* pMean); 

IppStatus ippsMean_16s(const Ipp16s* pSrc, int len, Ipp16s* pMean, 


IppStatus ippsMean_16sc(const Ipp16sc* pSrc, int len, Ipp16sc* pMean, 


 

pSrc 

pMean 

len 

scaleFactor 

 

 

 

2 

 

ippsMean pSrc pMean 

pSrc 

len – 1 

1 

mean = ----- pSrc[ n] 

len 

∑ 

n = 0 

5-25 ippsMean_32f 

 

ippStsNoErr 


ippStsSizeErr 

 

pMean pSrc NULL 

len 

5-101


5-25 

ippsMean 

void mean(void) { 

Ipp32f *x = ippsMalloc_32f(1000), mean; 

int i; 

for(i = 0; i> x = rand(1,1000); mean(x) 

StdDev 

 

IppStatus ippsStdDev_32f(const Ipp32f* pSrc, int len, Ipp32f* pStdDev); 

IppStatus ippsStdDev_64f(const Ipp64f* pSrc, int len, Ipp64f* pStdDev); 

IppStatus ippsStdDev_16s32s_Sfs(const Ipp16s* pSrc, int len, 

Ipp32s* pStdDev, int scaleFactor); 

IppStatus ippsStdDev_16s_Sfs(const Ipp16s* pSrc, int len, 

Ipp16s* pStdDev, int scaleFactor); 

5-102


 

pSrc 

pStdDev 

len 

scaleFactor 

 

 

 

2 

 

ippsStdDev pSrc pStdDev 

2 pSrc 

 

stdev = 

len – 1 

∑ 

n = 0 

( pSrc[ n] – mean( pSrc) 

) 2 

------------------------------------------------- 

len – 1 

5-26 ippsStdDev_32f 

 

ippStsNoErr 


ippStsSizeErr 

 

pStdDev pSrc NULL 

len 1 

5-103


5-26 

ippsStdDev 

void stdev(void) { 

Ipp32f *x = ippsMalloc_32f(1000), stdev; 

int i; 

for (i = 0; i> x = rand(1,1000); std(x) 

Norm 

CL1 L2 

IppStatus ippsNorm_Inf_16s32f(const Ipp16s* pSrc, int len, Ipp32f* pNorm); 

IppStatus ippsNorm_Inf_16s32s_Sfs(const Ipp16s* pSrc, int len, Ipp32s* pNorm, 


IppStatus ippsNorm_Inf_32f(const Ipp32f* pSrc, int len, Ipp32f* pNorm); 

IppStatus ippsNorm_Inf_64f(const Ipp64f* pSrc, int len, Ipp64f* pNorm); 

IppStatus ippsNorm_L1_16s32f(const Ipp16s* pSrc, int len, Ipp32f* pNorm); 

IppStatus ippsNorm_L1_16s32s_Sfs(const Ipp16s* pSrc, int len, Ipp32s* pNorm, 


IppStatus ippsNorm_L1_32f(const Ipp32f* pSrc, int len, Ipp32f* pNorm); 


IppStatus ippsNorm_L2_16s32f(const Ipp16s* pSrc, int len, Ipp32f* pNorm); 

5-104


IppStatus ippsNorm_L2_16s32s_Sfs(const Ipp16s* pSrc, int len, Ipp32s* pNorm, 




 

pSrc 

pNorm 

len 

scaleFactor 

pSrc 

 

 

2 

 

ippsNorm pSrc CL1 L2 

pNorm 

ippsNorm_Inf ippsNorm_Inf C 

Norm C = max len – 1 

ippsNorm_L1 ippsNorm_L1 L1 

Norm L1 = 

n = 0 

pSrc[ n] 

ippsNorm_L2 ippsNorm_L2 L2 

Norm L2 = 

len – 1 

∑ 

n = 0 

len – 1 

∑ 

n = 0 

pSrc[ n] 

pSrc[ n] 2 


 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrc pNorm NULL 

len 

5-105


NormDiff 

2 CL1 L2 

 

IppStatus ippsNormDiff_Inf_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

int len, Ipp32f* pNorm); 

IppStatus ippsNormDiff_Inf_16s32s_Sfs(const Ipp16s* pSrc1, 

const Ipp16s* pSrc2, int len, Ipp32s* pNorm, int scaleFactor); 

IppStatus ippsNormDiff_Inf_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 


IppStatus ippsNormDiff_Inf_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


IppStatus ippsNormDiff_L1_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsNormDiff_L1_16s32s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

int len, Ipp32s* pNorm, int scaleFactor); 

IppStatus ippsNormDiff_L1_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 




IppStatus ippsNormDiff_L2_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 


IppStatus ippsNormDiff_L2_16s32s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

int len, Ipp32s* pNorm, int scaleFactor); 





5-106


 

pSrc1pSrc2 

pNorm 

len 

scaleFactor 

2 pSrc2 NULL 

 

 

 

2 

 

ippsNorm CL1 L2 

pNorm 

ippsNormDiff_Inf ippsNormDiff_Inf C 

 

Norm Inf = max len – 1 

pSrc1[ n] – pSrc2[ n] 

n = 0 

ippsNormDiff_L1 ippsNormDiff_L1 L1 

 

Norm L1 

= 

n = 0 

pSrc1[ n] – pSrc2[ n] 

ippsNormDiff_L2 ippsNormDiff_L2 L2 

 

Norm L2 = 

len – 1 

∑ 

len – 1 

∑ 

n = 0 

pSrc1[ n] pSrc2[ n] 

– 

2 


 

5-27 ippsNormDiff 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrc1pSrc2 pNorm NULL 

len 

5-107


5-27 

ippsNorm 

int norm( void ) { 

Ipp16s x[LEN]; 

Ipp32f Norm[3]; 

IppStatus st; 

int i; 

for( i=0; i


IppStatus ippsDotProd_64fc(const Ipp64fc* pSrc1, const Ipp64fc* pSrc2, 

int len, Ipp64fc* pDp); 

IppStatus ippsDotProd_64f64fc(const Ipp64f* pSrc1, const Ipp64fc* pSrc2, 


IppStatus ippsDotProd_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

int len, Ipp16s* pDp, int scaleFactor); 

IppStatus ippsDotProd_16sc_Sfs(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 

int len, Ipp16sc* pDp, int scaleFactor); 

IppStatus ippsDotProd_16s16sc_Sfs(const Ipp16s* pSrc1, const Ipp16sc* pSrc2, 

int len, Ipp16sc* pDp, int scaleFactor); 

IppStatus ippsDotProd_16s64s(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

int len, Ipp64s* pDp); 

IppStatus ippsDotProd_16sc64sc(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 

int len, Ipp64sc* pDp); 

IppStatus ippsDotProd_16s16sc64sc(const Ipp16s* pSrc1, const Ipp16sc* pSrc2, 

int len, Ipp64sc* pDp); 

IppStatus ippsDotProd_16s32f(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

int len, Ipp32f* pDp); 

IppStatus ippsDotProd_16sc32fc(const Ipp16sc* pSrc1, const Ipp16sc* pSrc2, 


IppStatus ippsDotProd_16s16sc32fc(const Ipp16s* pSrc1, const Ipp16sc* pSrc2, 


 

pSrc1 

pSrc2 

pDp 

len 

scaleFactor 

 

2 

 

 

2 

5-109


 

ippsDotProd 2 pSrc1 pSrc2 ( ) 

pDp 

 

dp = 

len – 1 

∑ 

n = 0 

pSrc1[ n] ⋅ pSrc2[ n] 

1 

ippsConj pSrc1 pSrc2 

5-28 ippsDotProd_64f 

2 2 

 

 

ippStsNoErr 


ippStsSizeErr 

 

pDppSrc1 pSrc2 NULL 

len 

5-28 

ippsDotProd 

void dotprod(void) { 

Ipp64f x[10], dp; 

int n; 

for (n = 0; n> n = 0:9; x = sin(2*pi*n/8); a = x(1:8); b = x(3:10); a*b' 

5-110


 

 

 

 

• ( ) 

• ( ) 

6 

 

SampleUp 

 

 

IppStatus ippsSampleUp_32f (const Ipp32f* pSrc, int srcLen, Ipp32f* pDst, 

int* pDstLen, int factor, int* pPhase); 

IppStatus ippsSampleUp_32fc (const Ipp32fc* pSrc, int srcLen, Ipp32fc* pDst, 


IppStatus ippsSampleUp_64f (const Ipp64f* pSrc, int srcLen, Ipp64f* pDst, 


IppStatus ippsSampleUp_64fc (const Ipp64fc* pSrc, int srcLen, Ipp64fc* pDst, 


IppStatus ippsSampleUp_16s (const Ipp16s* pSrc, int srcLen, Ipp16s* pDst, 


IppStatus ippsSampleUp_16sc (const Ipp16sc* pSrc, int srcLen, Ipp16sc* pDst, 


 

pSrc 

srcLen 

pDst 

pDstLen 

( ) 

pSrc 

 

pDst 

5-111


factor factor -1 

pSrc 

pPhase pSrc pDst factor 

 

 

pPhase [0; factor-1] pPhase 

factor pSrc 

 

 

ippsSampleUp srcLen pSrc factor pPhase 

pDst pDstLen 

 

factor-1 pSrc 

pPhase factor 

pPhase [0; factor-1] 

factor 

factor-1 

pDstLen 

pPhase 

 

pPhase 

ippsSampleUp 

pDst[factor * n + phase] = pSrc[n], 0 ≤ n < srcLen 

pDst[factor * n + m] = 0, 0 ≤ n < srcLen, 0 ≤ m < factor, m ≠ phase 

pDstLen = factor * srcLen 

 

ippStsNoErr 


ippStsSizeErr 

 

pDstpSrcpDstLen pPhase 

NULL 

srcLen 

ippStsSampleFactorErr factor 

ippStsSamplePhaseErr 


5-112


SampleDown 

 

 

IppStatus ippsSampleDown_32f(const Ipp32f* pSrc, int srcLen, Ipp32f* pDst, 


IppStatus ippsSampleDown_32fc(const Ipp32fc* pSrc, int srcLen, Ipp32fc* pDst, 


IppStatus ippsSampleDown_64f(const Ipp64f* pSrc, int srcLen, Ipp64f* pDst, 


IppStatus ippsSampleDown_64fc(const Ipp64fc* pSrc, int srcLen, Ipp64fc* Dst, 


IppStatus ippsSampleDown_16s(const Ipp16s* pSrc, int srcLen, Ipp16s* pDst, 


IppStatus ippsSampleDown_16sc(const Ipp16sc* pSrc, int srcLen, Ipp16sc* pDst, 


 

pSrc 

srcLen 

pDst 

pDstLen 

 

pSrc 

ippsSampleDown 

pDst 

factor factor - 1 

pSrc factor 

 

pPhase pSrc factor 

pDst 

pPhase [0; factor-1] 

pPhase factor pSrc 

( ) 

5-113


 

ippsSampleDown srcLen pSrc factor pPhase 

pDst pDstLen 

 

pSrc factor - 1 

factor 1 pSrc pDst 

pPhase 

factor pPhase 

[0; factor-1] pDstLen 

 

pPhase 

 

pPhase 

FIR 

 

 


pDstLen = (srcLen + factor - 1 - phase)/factor 

pDst[n] = pSrc[factor * n + phase], 0 ≤ n < pDstLen 

phase = (factor + phase - srcLen %factor)%factor 

5-29 ippsSampleDown 

 

ippStsNoErr 


ippStsSizeErr 

 

pDstpSrcpDstLen pPhase 

NULL 

srcLen 

ippStsSampleFactorErr factor 

ippStsSamplePhaseErr 


5-114


5-29 


void sampling( void ) { 

Ipp16s x[8] = { 1,2,3,4,5,6,7,8 }; 

Ipp16s y[8] = {9,10,11,12,13,14,15,16}, z[8]; 

int dstLen1, dstLen2, phase = 2; 

IppStatus st = ippsSampleDown_16s(x, 8, z, &dstLen1, 3, &phase); 

st = ippsSampleDown_16s(y, 8, z+dstLen1, &dstLen2, 3, &phase); 

printf_16s(“down-sumpling =”, z, dstLen1+dstLen2, st); 

} 

Output: 

down-sampling = 3 6 9 12 15 

5-115


5-116

6 

 


 

LTI(linear time-invariant) 

 

2 

 

Conv 

2 

 

IppStatus ippsConv_32f(const Ipp32f* pSrc1, int lenSrc1, 

const Ipp32f* pSrc2, int lenSrc2, Ipp32f* pDst); 

IppStatus ippsConv_16s_Sfs(const Ipp16s* pSrc1, int lenSrc1, 

const Ipp16s* pSrc2, int lenSrc2, Ipp16s* pDst, int scaleFactor); 

 

pSrc1pSrc2 

lenSrc1 

lenSrc2 

2 

pSrc1 

pSrc2 

6-1


pDst 

scaleFactor 

pDst 

 

2 

 

ippsConv 2 

lenSrc1 pSrc1 lenSrc2 pSrc2 

(lenSrc1 + lenSrc2 - 1) pDst 

 

n 

pDst[ n] = ∑ pSrc1[ k] ⋅ pSrc2 [ n – k ], 0 ≤ n < lenSrc1 + lenSrc2 – 1 

k = 0 

i ≥ lenSrc1 pSrc1[i] = 0j ≥ lenSrc2 pSrc2[j] = 0 

6-1 ippsConv_16s_Sfs 2 

 

6-1 

ippsConv 2 

IppStatus convolution(void) { 

Ipp16s x[5] = {-2,0,1,-1,3}, h[2] = {0,1}, y[6]; 

IppStatus st = ippsConv_16s_Sfs(x, 5, h, 2, y, 0); 

printf_16s(“conv =”, y, 6, st); 

return st; 

} 

Output: 

conv = 0 -2 0 1 -1 3 


>> x = [-2,0,1,-1,3]; h = [0,1]; y = conv(x,h) 

6-2


 

ippStsNoErr 


ippStsSizeErr 

ippStsMemAllocErr 

 


len 

 

ConvCyclic 

2 

 

IppStatus ippsConvCyclic8x8_32f(const Ipp32f* x, const Ipp32f* h, 

Ipp32f* y); 

IppStatus ippsConvCyclic8x8_16s_Sfs(const Ipp16s* x, const Ipp16s* h, 

Ipp16s* y, int scaleFactor); 

IppStatus ippsConvCyclic4x4_32f32fc(const Ipp32f* x, const Ipp32fc* h, 

Ipp32fc* y); 

 

xh 

y 

scaleFactor 

 

y 

2 

 

ippsConvCyclic 2 

 

 

ippStsNoErr 

 

6-3


AutoCorr 

 

2 

 

IppStatus ippsAutoCorr_32f( const Ipp32f* pSrc, int srcLen, Ipp32f* pDst, 

int dstLen ); 

IppStatus ippsAutoCorr_NormA_32f( const Ipp32f* pSrc, int srcLen, 

Ipp32f* pDst, int dstLen ); 

IppStatus ippsAutoCorr_NormB_32f( const Ipp32f* pSrc, int srcLen, 


IppStatus ippsAutoCorr_64f( const Ipp64f* pSrc, int srcLen, Ipp64f* pDst, 

int dstLen ); 

IppStatus ippsAutoCorr_NormA_64f( const Ipp64f* pSrc, int srcLen, 


IppStatus ippsAutoCorr_NormB_64f( const Ipp64f* pSrc, int srcLen, 


IppStatus ippsAutoCorr_32fc( const Ipp32fc* pSrc, int srcLen, Ipp32fc* pDst, 

int dstLen ); 

IppStatus ippsAutoCorr_NormA_32fc( const Ipp32fc* pSrc, int srcLen, 

Ipp32fc* pDst, int dstLen ); 

IppStatus ippsAutoCorr_NormB_32fc( const Ipp32fc* pSrc, int srcLen, 


IppStatus ippsAutoCorr_64fc( const Ipp64fc* pSrc, int srcLen, Ipp64fc* pDst, 

int dstLen ); 

IppStatus ippsAutoCorr_NormA_64fc( const Ipp64fc* pSrc, int srcLen, 


IppStatus ippsAutoCorr_NormB_64fc( const Ipp64fc* pSrc, int srcLen, 


IppStatus ippsAutoCorr_16s_Sfs( const Ipp16s* pSrc, int srcLen, Ipp16s* pDst, 

int dstLen, int scaleFactor ); 

IppStatus ippsAutoCorr_NormA_16s_Sfs( const Ipp16s* pSrc, int srcLen, 

Ipp16s* pDst, int dstLen, int scaleFactor ); 

IppStatus ippsAutoCorr_NormB_16s_Sfs( const Ipp16s* pSrc, int srcLen, 

Ipp16s* pDst, int dstLen, int scaleFactor ); 

6-4


 

pSrc 

srcLen 

 

 

pDst 

 

 

dstLen ( ) 

scaleFactor 

2 

 

ippsAutoCorr srcLen pSrc 

dstLen pDst 

ippsAutoCorr_NormA ippsAutoCorr_NormA 

pDst 

 

srcLen – 1 

∑ 

pDst[ n] = 

conj( pSrc[ i] 

) ⋅ pSrc[ i + n] , 0 ≤ n < dstLen 

i = 0 

( ) 

pDst[ n] 

= 

srcLen – 1 

1 

---------------------- conj( pSrc[ i] 

) ⋅ pSrc[ i + n] , 0 ≤ n < dstLen ( ) 

srcLen ∑ 

i = 0 

pDst[ n] 

= 

srcLen – 1 

1 

srcLen 

-------------------------------- – n ∑ conj ( pSrc [ i ]) ⋅ pSrc[ i + n] , 0 ≤ n < dstLen 

i = 0 

( 

) 

 

pSrc[ i] 

⎧pSrc[ i] , 0 ≤ i < srcLen 

= ⎨ 

⎩ 

0, 

otherwise 

6-5


 

 

 

 

ippStsNoErr 


ippStsSizeErr 

 


srcLen dstLen 

 

CrossCorr 

2 

CrossCorr 

2 

IppStatus ippsCrossCorr_32f(const Ipp32f* pSrc1, int len1, 

const Ipp32f* pSrc2, int len2, Ipp32f* pDst, int dstLen, int lowLag); 

IppStatus ippsCrossCorr_64f(const Ipp64f* pSrc1, int len1, 

const Ipp64f* pSrc2, int len2, Ipp64f* pDst, int dstLen, int lowLag); 

IppStatus ippsCrossCorr_32fc(const Ipp32fc* pSrc1, int len1, 

const Ipp32fc* pSrc2, int len2, Ipp32fc* pDst, int dstLen, int lowLag); 

IppStatus ippsCrossCorr_64fc(const Ipp64fc* pSrc1, int len1, 

const Ipp64fc* pSrc2, int len2, Ipp64fc* pDst, int dstLen, int lowLag); 

IppStatus ippsCrossCorr_16s_Sfs(const Ipp16s* pSrc1, int len1, 

const Ipp16s* pSrc2, int len2, Ipp16s* pDst, int dstLen, int lowLag, 


 

pSrc1 

len1 

pSrc2 

 

pSrc1 

2 

6-6


len2 

pDst 

dstLen 

lowLag 

scaleFactor 

pSrc2 

pSrc1 pSrc2 

 

pDst 

 

 

2 

 

ippsCrossCorr len1 pSrc1 len2 pSrc2 

pDst 

pDst 

len1 – 1 

∑ 

pDst[ n] = conj( pSrc1[ i] 

) ⋅ pSrc2[ n+ i + lowLag] 

i = 0 

, 

 

0 ≤ n < dstLen , 

 

pSrc2[ j] 

⎧pSrc2[ j] , 0 ≤ j < len2 

= ⎨ 

⎩0, 

otherwise 

6-2 ippsCrossCorr 

6-7


6-2 

ippsCrossCorr 

void crossCorr(void) { 

#undef LEN 

#define LEN 11 

Ipp32f win[LEN], y[LEN]; 

IppStatus st; 

ippsSet_32f (1, win, LEN); 

ippsWinHamming_32f_I (win, LEN); 

st = ippsCrossCorr_32f (win, LEN, win, LEN, y, -(LEN-1)); 

printf_32f(“cross corr =”, y,7,st); 

} 

Output: 

cross corr = 0.006400 0.026856 0.091831 0.242704 0.533230 

1.009000 1.672774 

Matlab* analog: 

>> x = hamming(11)'; y = xcorr(x,x); y(1:7) 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrc1 pSrc2 NULL 

len1 len2 

6-8


 

 

• (FIR) 

• 2 (LMS) FIR 

• (IIR) 

• 

FIR 

FIR ( ) 

 

FIR 

1. ippsFIRInitAlloc 

ippsFIRMRInitAlloc 

 

2. ippsFIROne 

ippsFIR 

 

 

3. ippsFIRGetTaps ( ) 

ippsFIRGetDlyLine ippsFIRSetDlyLine 

 

4. ippsFIRFree FIR 

6-9


FIRInitAllocFIRMRInitAlloc 

 

FIR 

IppStatus ippsFIRInitAlloc_32f(IppsFIRState_32f** pState, 

const Ipp32f* pTaps, int tapsLen, const Ipp32f* pDlyLine); 

IppStatus ippsFIRMRInitAlloc_32f(IppsFIRState_32f** pState, 

const Ipp32f* pTaps, int tapsLen, int upFactor, int upPhase, 

int downFactor, int downPhase, const Ipp32f* pDlyLine); 

IppStatus ippsFIRInitAlloc_32fc(IppsFIRState_32fc** pState, 

const Ipp32fc* pTaps, int tapsLen, const Ipp32fc* pDlyLine); 

IppStatus ippsFIRMRInitAlloc_32fc(IppsFIRState_32fc** pState, 

const Ipp32fc* pTaps, int tapsLen, int upFactor, int upPhase, 

int downFactor, int downPhase, const Ipp32fc* pDlyLine); 

IppStatus ippsFIRInitAlloc32f_16s(IppsFIRState32f_16s** pState, 

const Ipp32f* pTaps, int tapsLen, const Ipp16s* pDlyLine); 

IppStatus ippsFIRMRInitAlloc32f_16s(IppsFIRState32f_16s** pState, 


int downFactor, int downPhase, const Ipp16s* pDlyLine); 

IppStatus ippsFIRInitAlloc32fc_16sc(IppsFIRState32fc_16sc** pState, 

const Ipp32fc* pTaps, int tapsLen, const Ipp16sc* pDlyLine); 

IppStatus ippsFIRMRInitAlloc32fc_16sc(IppsFIRState32fc_16sc** pState, 


int downFactor, int downPhase, const Ipp16sc* pDlyLine); 

IppStatus ippsFIRInitAlloc_64f(IppsFIRState_64f** pState, 


IppStatus ippsFIRMRInitAlloc_64f(IppsFIRState_64f** pState, 



IppStatus ippsFIRInitAlloc_64fc(IppsFIRState_64fc** pState, 


6-10


IppStatus ippsFIRMRInitAlloc_64fc(IppsFIRState_64fc** pState, 



IppStatus ippsFIRInitAlloc64f_32f(IppsFIRState64f_32f** pState, 


IppStatus ippsFIRMRInitAlloc64f_32f(IppsFIRState64f_32f** pState, 



IppStatus ippsFIRInitAlloc64fc_32fc(IppsFIRState64fc_32fc** pState, 


IppStatus ippsFIRMRInitAlloc64fc_32fc(IppsFIRState64fc_32fc** pState, 























6-11


IppStatus ippsFIRInitAlloc32s_16s(IppsFIRState32s_16s** pState, 

const Ipp32s* pTaps, int tapsLen, int tapsFactor, const Ipp16s* pDlyLine); 

IppStatus ippsFIRMRInitAlloc32s_16s(IppsFIRState32s_16s** pState, 

const Ipp32s* pTaps, int tapsLen, int tapsFactor, int upFactor, 

int upPhase, int downFactor, int downPhase, const Ipp16s* pDlyLine); 

IppStatus ippsFIRInitAlloc32s_16s32f(IppsFIRState32s_16s** pState, 


IppStatus ippsFIRMRInitAlloc32s_16s32f(IppsFIRState32s_16s** pState, 

const Ipp32f *pTaps, int tapsLen, int upFactor, int upPhase, 


IppStatus ippsFIRInitAlloc32sc_16sc(IppsFIRState32sc_16sc** pState, 

const Ipp32sc* pTaps, int tapsLen, int tapsFactor, 

const Ipp16sc* pDlyLine); 

IppStatus ippsFIRMRInitAlloc32sc_16sc(IppsFIRState32sc_16sc** pState, 

const Ipp32sc* pTaps, int tapsLen, int tapsFactor, int upFactor, 

int upPhase, int downFactor, int downPhase, const Ipp16sc* pDlyLine); 

IppStatus ippsFIRInitAlloc32sc_16sc32fc(IppsFIRState32sc_16sc** pState, 


IppStatus ippsFIRMRInitAlloc32sc_16sc32fc(IppsFIRState32sc_16sc** pState, 



 

pTaps 

tapsLen 

tapsFactor 

downFactor 

downPhase 

upFactor 

upPhase 

 

tapsLen 

 

Ipp32s 

( ) 

ippsFIRMRInit 

 

ippsFIRMRInit 

 

ippsFIRMRInit 

 

ippsFIRMRInit 

 

6-12


pDlyLine 

pState 

 

tapsLen 

(tapsLen + upFactor - 1) / upFactor 

FIR 

 

ippsFIRInitAlloc ippsFIRMRInitAlloc 

FIR 

tapsLen pTaps pState 

tapsFactor pDlyLine 

pDlyLine NULL 

pState 

 

 

ippsFIRInitAlloc ippsFIRInitAlloc 

pState tapsLen pTaps 

pDlyLine NULL tapsLen 

ippsFIRMRInitAlloc ippsFIRMRInitAlloc ( 

 

) pState 

 

 

 

upFactor 

(5-111 SampleUp ) 

upFactor-1 

upPhase upFactor 

 

downFactor FIR 

(5-113 SampleDown 

) downFactor 

downFactor-1 

6-13


downPhase 

 

pDelay NULL 

(tapsLen + upFactor - 1) / upFactor 

 

ippStsNoErr 



ippStsFIRLenErr 

 

 

NULL 

tapsLen 

ippStsFIRMRFactorErr upFactor (downFactor) 

ippStsFIRMRPhaseErr 

upPhase (downPhase) upFactor 

(downFactor) 

ippStsContextMatchErr 

 

FIRFree 

FIR 

IppStatus ippsFIRFree_32f(IppsFIRState_32f* pState); 

IppStatus ippsFIRFree_32fc(IppsFIRState_32fc* pState); 

IppStatus ippsFIRFree32f_16s(IppsFIRState32f_16s* pState); 

IppStatus ippsFIRFree32fc_16sc(IppsFIRState32fc_16sc* pState); 

IppStatus ippsFIRFree_64f(IppsFIRState_64f* pState); 

IppStatus ippsFIRFree_64fc(IppsFIRState_64fc* pState); 

IppStatus ippsFIRFree64f_32f(IppsFIRState64f_32f* pState); 

IppStatus ippsFIRFree64fc_32fc(IppsFIRState64fc_32fc* pState); 



6-14




IppStatus ippsFIRFree32s_16s(IppsFIRState32s_16s* pState); 

IppStatus ippsFIRFree32sc_16sc(IppsFIRState32sc_16sc* pState); 

 

pState 

FIR 

 

ippsFIRFree ippsFIRInitAlloc ippsFIRMRInitAlloc 

FIR 

ippsFIRFree 

 

 

ippStsNoErr 


 

NULL 


 

FIROne 

FIR 

 

IppStatus ippsFIROne_32f(Ipp32f src, Ipp32f* pDstVal, 

IppsFIRState_32f* pState); 

IppStatus ippsFIROne_32fc(Ipp32fc src, Ipp32fc* pDstVal, 

IppsFIRState_32fc* pState); 

IppStatus ippsFIROne32f_16s_Sfs(Ipp16s src, Ipp16s* pDstVal, 

IppsFIRState32f_16s* pState, int scaleFactor); 

6-15


IppStatus ippsFIROne32fc_16sc_Sfs(Ipp16sc src, Ipp16sc* pDstVal, 

IppsFIRState32fc_16sc* pState, int scaleFactor); 

IppStatus ippsFIROne_64f(Ipp64f src, Ipp64f* pDstVal, 


IppStatus ippsFIROne_64fc(Ipp64fc src, Ipp64fc* pDstVal, 


IppStatus ippsFIROne64f_32f(Ipp32f src, Ipp32f* pDstVal, 

IppsFIRState64f_32f* pState); 

IppStatus ippsFIROne64fc_32fc(Ipp32fc src, Ipp32fc* pDstVal, 

IppsFIRState64fc_32fc* pState); 









IppStatus ippsFIROne32s_16s_Sfs(Ipp16s src, Ipp16s* pDstVal, 

IppsFIRState32s_16s* pState, int scaleFactor); 

IppStatus ippsFIROne32sc_16sc_Sfs(Ipp16sc src, Ipp16sc* pDstVal, 

IppsFIRState32sc_16sc* pState, int scaleFactor); 

 

pState 

src 

pDstVal 

scaleFactor 

FIR 

ippsFIROne 

ippsFIROne 

2 

6-16


 

ippsFIROne src 

pDstVal pState 

scaleFactor 

FIR 

x(n) h(i) 

y(n) 

ippsFIROne ippsFIRInitAlloc 

tapsLenpTaps 

pDelay 

 

tapsLen – 1 

∑ 

yn ( ) = 

hi ( ) ⋅ xn ( – i) 

i = 0 

ippStsNoErr 


 

NULL 


 

FIR 

 

 

 

IppStatus ippsFIR_32f(const Ipp32f* pSrc, Ipp32f* pDst, int numIters, 


IppStatus ippsFIR_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int numIters, 


IppStatus ippsFIR32f_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, int numIters, 


IppStatus ippsFIR32fc_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int numIters, IppsFIRState32fc_16sc* pState, int scaleFactor); 

6-17


IppStatus ippsFIR_32f_I(Ipp32f* pSrcDst, int numIters, 


IppStatus ippsFIR_32fc_I(Ipp32fc* pSrcDst, int numIters, 


IppStatus ippsFIR32f_16s_ISfs(Ipp16s* pSrcDst, int numIters, 


IppStatus ippsFIR32fc_16sc_ISfs(Ipp16sc* pSrcDst, int numIters, 


IppStatus ippsFIR_64f(const Ipp64f* pSrc, Ipp64f* pDst, int numIters, 


IppStatus ippsFIR_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int numIters, 


IppStatus ippsFIR_64f_I(Ipp64f* pSrcDst, int numIters, 


IppStatus ippsFIR_64fc_I(Ipp64fc* pSrcDst, int numIters, 


IppStatus ippsFIR64f_32f(const Ipp32f* pSrc, Ipp32f* pDst, int numIters, 


IppStatus ippsFIR64fc_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int numIters, 


IppStatus ippsFIR64f_32f_I(Ipp32f* pSrcDst, int numIters, 


IppStatus ippsFIR64fc_32fc_I(Ipp32fc* pSrcDst, int numIters, 














6-18






IppStatus ippsFIR32s_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 

int numIters, IppsFIRState32s_16s* pState, int scaleFactor); 

IppStatus ippsFIR32sc_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

int numIters, IppsFIRState32sc_16sc* pState, int scaleFactor); 

IppStatus ippsFIR32s_16s_ISfs(Ipp16s* pSrcDst, int numIters, 

IppsFIRState32s_16s* pState, int scaleFactor); 

IppStatus ippsFIR32sc_16sc_ISfs(Ipp16sc* pSrcDst, int numIters, 

IppsFIRState32sc_16sc* pState, int scaleFactor); 

 

pState 

pSrc 

pDst 

pSrcDst 

numIters 

scaleFactor 

FIR 

ippsFIR 

ippsFIR 

ippsFIR ( 

) 

ippsFIR 

 

numIters numIters 

 

(numIters * downFactor) 

(numIters * upFactor) 

 

2 

6-19


 

ippsFIR 

pSrc pSrcDst pDst 

pSrcDst pState 

pSrc pSrcDst numIters 

numIters pDst 

pSrcDst ippsFIROne numIters 

 

FIR x(n) 

h(i) y(n) 

y(n) 

tapsLen – 1 

∑ 

yn ( ) = 

hi ( ) ⋅ xn ( – i), 0 ≤ n < numIters 

i = 0 

 

numIters 

 

ippsFIR (numIters*downFactor) 

(numIters*upFactor) 

 

FIR 3 

3 

 

(upFactor*srcLen) 

ippsFIR ippsFIRInitAlloc ippsFIRMRInitAlloc 

tapsLen 

pTaps pDelay 

 

6-3 ippsFIR_32f 

 

ippStsNoErr 


ippStsSizeErr 

 

NULL 

len 

6-20



 

6-3 

ippsFIR 

IppStatus fir(void) { 

#undef NUMITERS 

#define NUMITERS 150 

int n; 

IppStatus status; 

IppsIIRState_32f *ictx; 

IppsFIRState_32f *fctx; 

Ipp32f *x = ippsMalloc_32f(NUMITERS), 

*y = ippsMalloc_32f(NUMITERS), 

*z = ippsMalloc_32f(NUMITERS); 

const float taps[] = { 

0.0051f, 0.0180f, 0.0591f, 0.1245f, 0.1869f, 0.2127f, 0.1869f, 

0.1245f, 0.0591f, 0.0180f, 0.0051f, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 

}; 

for (n =0;n> F = 0.2; N = 150; n = 0:N-1; x = sin(2*pi*n*F); 

y = filter(fir1(10,0.15),1,x); y(121:125) 

6-21


FIRGetTaps 

FIR 

IppStatus ippsFIRGetTaps_32f(const IppsFIRState_32f* pState, Ipp32f* pTaps); 

IppStatus ippsFIRGetTaps_32fc(const IppsFIRState_32fc* pState, 

Ipp32fc* pTaps); 

IppStatus ippsFIRGetTaps32f_16s(const IppsFIRState32f_16s* pState, 

Ipp32f* pTaps); 

IppStatus ippsFIRGetTaps32fc_16sc(const IppsFIRState32fc_16sc* pState, 


IppStatus ippsFIRGetTaps_64f(const IppsFIRState_64f* pState, Ipp64f* pTaps); 

IppStatus ippsFIRGetTaps_64fc(const IppsFIRState_64fc* pState, 


IppStatus ippsFIRGetTaps64f_32f(const IppsFIRState64f_32f* pState, 


IppStatus ippsFIRGetTaps64fc_32fc(const IppsFIRState64fc_32fc* pState, 










IppStatus ippsFIRGetTaps32s_16s(const IppsFIRState32s_16s* pState, 

Ipp32s* pTaps, int* tapsFactor); 

IppStatus ippsFIRGetTaps32sc_16sc(const IppsFIRState32sc_16sc* pState, 

Ipp32sc* pTaps, int* tapsFactor); 

IppStatus ippsFIRGetTaps32s_16s32f(const IppsFIRState32s_16s* pState, 


6-22


IppStatus ippsFIRGetTaps32sc_16sc32fc(const IppsFIRState32sc_16sc* pState, 


 

pState 

pTaps 

tapsFactor 

FIR 

 

Ipp32s 

( ) 

 

ippsFIRGetTaps pState tapsLen pTaps 

tapsFactor 

ippsFIRInit 

ippsFIRMRInit 

ippsFIRGetTaps ippspFIRInitAlloc ippspFIRMRInitAlloc 

 

 

ippStsNoErr 


 

NULL 


 

FIRGetDlyLine, FIRSetDlyLine 

FIR 

 

IppStatus ippsFIRGetDlyLine_32f(const IppsFIRState_32f* pState, 

Ipp32f* pDlyLine); 

IppStatus ippsFIRSetDlyLine_32f(IppsFIRState_32f* pState, 

const Ipp32f* pDlyLine); 

6-23


IppStatus ippsFIRGetDlyLine_32fc(const IppsFIRState_32fc* pState, 

Ipp32fc* pDlyLine); 

IppStatus ippsFIRSetDlyLine_32fc(IppsFIRState_32fc* pState, 

const Ipp32fc* pDlyLine); 

IppStatus ippsFIRGetDlyLine32f_16s(const IppsFIRState32f_16s* pState, 

Ipp16s* pDlyLine); 

IppStatus ippsFIRSetDlyLine32f_16s(IppsFIRState32f_16s* pState, 

const Ipp16s* pDlyLine); 

IppStatus ippsFIRGetDlyLine32fc_16sc(const IppsFIRState32fc_16sc* pState, 

Ipp16sc* pDlyLine); 

IppStatus ippsFIRSetDlyLine32fc_16sc(IppsFIRState32fc_16sc* pState, 


IppStatus ippsFIRGetDlyLine_64f(const IppsFIRState_64f* pState, 


IppStatus ippsFIRSetDlyLine_64f(IppsFIRState_64f* pState, 


IppStatus ippsFIRGetDlyLine_64fc(const IppsFIRState_64fc* pState, 


IppStatus ippsFIRSetDlyLine_64fc(IppsFIRState_64fc* pState, 


IppStatus ippsFIRGetDlyLine64f_32f(const IppsFIRState64f_32f* pState, 


IppStatus ippsFIRSetDlyLine64f_32f(IppsFIRState64f_32f* pState, 


IppStatus ippsFIRGetDlyLine64fc_32fc(const IppsFIRState64fc_32fc* pState, 


IppStatus ippsFIRSetDlyLine64fc_32fc(IppsFIRState64fc_32fc* pState, 










6-24










IppStatus, ippsFIRGetDlyLine32s_16s, (const IppsFIRState32s_16s* pState, 


IppStatus, ippsFIRSetDlyLine32s_16s, (IppsFIRState32s_16s* pState, 


IppStatus, ippsFIRGetDlyLine32sc_16sc, (const IppsFIRState32sc_16sc* pState, 


IppStatus, ippsFIRSetDlyLine32sc_16sc, (IppsFIRState32sc_16sc* pState, 


 

pState 

pDlyLine 

FIR 

 

 

ippsFIRGetDlyLine ippsFIRSetDlyLine FIR 

 

ippsFIRGetDlyLine ippsFIRGetDlyLine pState 

pDlyLine 

ippsFIRSetDlyLine ippsFIRSetDlyLine pDlyLine 

pState 

ippsFIRGetDlyLine ippsFIRSetDlyLine ippsFIRInitAlloc 

ippsFIRMRInitAlloc 

 

6-25


 

ippStsNoErr 


 

NULL 


FIR LMS 

 

• FIR LMS 

• 

• ( ) 

• 

• 

FIR LMS 

1. ippsFIRLMSInitAlloc FIR LMS 

 

2. ippsFIRLMSOne 1 FIR 

1 ippsFIRLMS 

 

3. ippsFIRLMSGetTaps ( ) 

ippsFIRLMSGetDlyLine ippsFIRLMSSetDlyLine 

 

4. ippsFIRLMSFree FIR LMS 

 

6-26


FIRLMSInitAlloc 

2 (LMS) 

FIR 

IppStatus ippsFIRLMSInitAlloc_32f(IppsFIRLMSState_32f** pState, 

const Ipp32f* pTaps, int tapsLen, const Ipp32f* pDlyLine, int dlyIndex); 

IppStatus ippsFIRLMSInitAlloc32f_16s(IppsFIRLMSState32f_16s** pState, 

const Ipp32f* pTaps, int tapsLen, const Ipp16s* pDlyLine, int dlyIndex); 

 

pTaps 

tapsLen 

pDlyLine 

dlyIndex 

pState 

 

tapsLen 

 

 

2*tapsLen 

 

 

 

ippsFIRLMSInitAlloc FIR LMS 

ippsFIRLMSInitAlloc tapsLen pTaps 

pState 2*tapsLen pDlyLine 

pDlyLine dlyIndex 

pDlyLine pTaps NULL 

 

 

ippStsNoErr 


 

NULL 


 

6-27


FIRLMSFree 

2 (LMS) 

FIR 

IppStatus ippsFIRLMSFree_32f(IppsFIRLMSState_32f* pState); 

IppStatus ippsFIRLMSFree32f_16s(IppsFIRLMSState32f_16s* pState); 

 

pState 

FIR LMS 

 

ippsFIRLMSFree ippsFIRLMSInitAlloc 

FIR LMS 

ippsFIRLMSFree 

 

ippStsNoErr 


 

NULL 


 

FIRLMSOne 

FIR LMS 

 

IppStatus ippsFIRLMSOne_Direct_32f(Ipp32f src, Ipp32f ref, Ipp32f* pDstVal, 

Ipp32f* pTapsInv, int tapsLen, float mu, Ipp32f* pDlyLine, int* pDlyIndex); 

IppStatus ippsFIRLMSOne_Direct32f_16s(Ipp16s src, Ipp16s ref, Ipp16s* pDstVal, 

Ipp32f* pTapsInv, int tapsLen, float mu, Ipp16s* pDlyLine, int* pDlyIndex); 

6-28


IppStatus ippsFIRLMSOne_DirectQ15_16s(Ipp16s src, Ipp16s ref, Ipp16s* pDstVal, 

Ipp32s* pTapsInv, int tapsLen, int mu15, Ipp16s* pDlyLine, int* pDlyIndex); 

 

src 

pDstVal 

ref 

pTapsInv 

tapsLen 

pDlyLine 

pDlyIndex 

mu 

muQ15 

 

 

 

FIR 

 

 

 

 

 

 

 

ippsFIRLMSOne FIR 1 

tapsLen pTapsInv FIR 

2*tapsLen pDlyLine pDlyIndex 

pDstVal 

 

err[n] = ref[n] - *pDstVal. 

ippsFIRLMSOne mu FIR 1 

 

 

ippsLMSOne 

 

 

6-29


DirectQ15 ippsFIRLMSOne FIR 

1 muQ15 

 

muQ15 = (int)(mu *(1


FIRLMS 

FIR LMS 

 

IppStatus ippsFIRLMS_32f(const Ipp32f* pSrc, const Ipp32f* pRef, 

Ipp32f* pDst, int len, float mu, IppsFIRLMSState_32f* pState); 

IppStatus ippsFIRLMS32f_16s(const Ipp16s* pSrc, const Ipp16s* pRef, 

Ipp16s* pDst, int len, float mu, IppsFIRLMSState32f_16s* pState); 

 

pState 

pSrc 

pRef 

pDst 

len 

mu 

FIR LMS 

 

 

 

 

 

 

ippsFIRLMS pRef mu pSrc 

 

len 2 1 

ippsLMS pSrc 

pDst 2 pRef pDst 

mu 

ippsFIRLMS ippsFIRLMSInitAlloc pState 

 

6-5 ippsFIRLMS_32f 

 

6-31


 

ippStsNoErr 


ippStsSizeErr 

 

NULL 

len 


 

6-5 

ippsFIRLMS 

IppStatus firlms(void) { 

IppStatus st; 

Ipp32f taps = 0, x[LEN], y[LEN], mu = 0.03f; 

IppsFIRLMSState_32f* ctx; 

int i; 

/// no taps and no delay line from outside 

ippsFIRLMSInitAlloc_32f( &ctx, 0, 1, 0, 0 ); 

/// make a const signal of amplitude 1 and noise it 

for(i=0; i


FIRLMSGetTaps 

FIR LMS 

IppStatus ippsFIRLMSGetTaps_32f(const IppsFIRLMSState_32f* pState, 

Ipp32f* pOutTaps); 

IppStatus ippsFIRLMSGetTaps32f_16s(const IppsFIRLMSState32f_16s* pState, 

Ipp32f* pOutTaps); 

 

pState 

pOutTaps 

FIR LMS 

 

 

ippsFIRLMSGetTaps pState tapsLen 

pOutTaps 

ippsFIRLMSInitAlloc 

ippsFIRLMSGetTaps ippsFIRLMSInitAlloc 

 

 

ippStsNoErr 


 

NULL 


 

6-33


FIRLMSGetDlyLineFIRLMSSetDlyLine 

FIR LMS 

 

IppStatus ippsFIRLMSGetDlyLine_32f(const IppsFIRLMSState_32f* pState, 

Ipp32f* pDlyLine, int* pDlyIndex); 

IppStatus ippsFIRLMSGetDlyLine32f_16s(const IppsFIRLMSState32f_16s* pState, 

Ipp16s* pDlyLine, int* pDlyIndex); 

IppStatus ippsFIRLMSSetDlyLine_32f(IppsFIRLMSState_32f* pState, 

const Ipp32f* pDlyLine, int dlyIndex); 

IppStatus ippsFIRLMSSetDlyLine32f_16s(IppsFIRLMSState32f_16s* pState, 

const Ipp16s* pDlyLine, int dlyIndex); 

 

pState 

pDlyLine 

pDlyIndex 

dlyIndex 

FIR LMS 

tapsLen 

ippsFIRLMSGetDlyLine pState 

 

ippsFIRLMSSetDlyLine pState 

 

 

ippsFIRLMSGetDlyLine ippsFIRLMSSetDlyLine FIR LMS 

 

ippsFIRLMSGetDlyLine ippsFIRLMSGetDlyLine pState 

pDlyLine 

pDlyIndex 

ippsFIRLMSSetDlyLine ippsFIRLMSSetDlyLine pDlyLine 

dlyIndex 

pState 

ippsFIRLMSGetDlyLine ippsFIRLMSSetDlyLine 

ippsFIRLMSInitAlloc 

 

6-34


 

ippStsNoErr 


IIR 

 

NULL 


 

(IIR) 

 

IIR 

1. ippsIIRInitAlloc IIR 

ippsIIRInitAlloc_BiQuad 

 

2. ippsIIROne IIR 

ippsIIR 1 

 

3. ippsIIRGetDlyLine ippsIIRSetDlyLine IIR 

 

4. ippsIIRFree 

 

IIRInitAllocIIRInitAlloc_BiQuad 

 

IppStatus ippsIIRInitAlloc_32f(IppsIIRState_32f** pState, 

const Ipp32f* pTaps, int order, const Ipp32f* pDlyLine); 

IppStatus ippsIIRInitAlloc_32fc(IppsIIRState_32fc** pState, 

const Ipp32fc* pTaps, int order, const Ipp32fc* pDlyLine); 

IppStatus ippsIIRInitAlloc32f_16s(IppsIIRState32f_16s** pState, 


6-35


IppStatus ippsIIRInitAlloc32fc_16sc(IppsIIRState32fc_16sc** pState, 


IppStatus ippsIIRInitAlloc_64f(IppsIIRState_64f** pState, 


IppStatus ippsIIRInitAlloc_64fc(IppsIIRState_64fc** pState, 


IppStatus ippsIIRInitAlloc64f_32f(IppsIIRState64f_32f** pState, 


IppStatus ippsIIRInitAlloc64fc_32fc(IppsIIRState64fc_32fc** pState, 










IppStatus ippsIIRInitAlloc_BiQuad_32f(IppsIIRState_32f** pState, 

const Ipp32f* pTaps, int numBq, const Ipp32f* pDlyLine); 

IppStatus ippsIIRInitAlloc_BiQuad_32fc(IppsIIRState_32fc** pState, 

const Ipp32fc* pTaps, int numBq, const Ipp32fc* pDlyLine); 

IppStatus ippsIIRInitAlloc32f_BiQuad_16s(IppsIIRState32f_16s** pState, 


IppStatus ippsIIRInitAlloc32fc_BiQuad_16sc(IppsIIRState32fc_16sc** pState, 


IppStatus ippsIIRInitAlloc_BiQuad_64f(IppsIIRState_64f** pState, 


IppStatus ippsIIRInitAlloc_BiQuad_64fc(IppsIIRState_64fc** pState, 


IppStatus ippsIIRInitAlloc64f_BiQuad_32f(IppsIIRState64f_32f** pState, 


6-36


IppStatus ippsIIRInitAlloc64fc_BiQuad_32fc(IppsIIRState64fc_32fc** pState, 










IppStatus ippsIIRInitAlloc32s_16s(IppsIIRState32s_16s** pState, 

const Ipp32s* pTaps, int order, int tapsFactor, const Ipp32s* pDlyLine); 

IppStatus ippsIIRInitAlloc32s_16s32f(IppsIIRState32s_16s** pState, 

const Ipp32f* pTaps, int order, const Ipp32s* pDlyLine); 

IppStatus ippsIIRInitAlloc32sc_16sc(IppsIIRState32sc_16sc** pState, 

const Ipp32sc* pTaps, int order, int tapsFactor, const Ipp32sc* pDlyLine); 

IppStatus ippsIIRInitAlloc32sc_16sc32fc(IppsIIRState32sc_16sc** pState, 

const Ipp32fc* pTaps, int order, const Ipp32sc* pDlyLine); 

IppStatus ippsIIRInitAlloc32s_BiQuad_16s(IppsIIRState32s_16s** pState, 

const Ipp32s* pTaps, int numBq, int tapsFactor, const Ipp32s* pDlyLine); 

IppStatus ippsIIRInitAlloc32s_BiQuad_16s32f(IppsIIRState32s_16s** pState, 

const Ipp32f* pTaps, int numBq, const Ipp32s* pDlyLine); 

IppStatus ippsIIRInitAlloc32sc_BiQuad_16sc(IppsIIRState32sc_16sc** pState, 

const Ipp32sc* pTaps, int numBq, int tapsFactor, const Ipp32sc* pDlyLine); 

IppStatus ippsIIRInitAlloc32sc_BiQuad_16sc32fc(IppsIIRState32sc_16sc** pState, 

const Ipp32fc* pTaps, int numBq, const Ipp32sc* pDlyLine); 

 

pTaps 

tapsFactor 

 

2*(order+1)BQ 6*numBq 

 

Ipp32s ( 

) 

6-37


numBq 

order 

pDlyLine 

pState 

numBq ippsIIRBQInit 

 

IIR ippsIIRInit 

 

orderBQ 2*numBq 

 

IIR 

 

ippsIIRInitAlloc ippsIIRInitAlloc_BiQuad 

(BQ) IIR 

pTaps pState 

tapsFactor pDlyLine 

pDlyLine NULL 

 

 

ippsIIRInitAlloc ippsIIRInitAlloc IIR 

order pDlyLine 

order 

2*(order + 1) pTaps 

 

B 0 , B 1 , .. ., B order , A 0 , A 1 , .. ., A order 

A 0 ≠ 0 

32s ippsIIRInitAlloc 

ippsIIRInitAlloc_BiQuad ippsConvert 

 

32s_32f 

 

 

6-7 

 

6-38


ippsIIRInitAlloc_BiQuad ippsIIRInitAlloc_BiQuad IIR 

( ) 

2*numBq pDlyLine 

numBq 6*numBq pTaps 

 

B 0,0 , B 0,1 , B 0,2 , A 0,0 , A 0,1 , A 0,2 ; B 1,0 , B 1,1 , B 1,2 , A 1,0 , A 1,1 , A 1,2 ; .. . A numBq-1,2 

A n,0 ≠ 0, B n,0 ≠ 0 

 

ippStsNoErr 



ippStsIIROrderErr 


 

 

NULL 

order numBq 

A 0 A n,0 B n,0 


 

IIRFree 

IIR 

IppStatus ippsIIRFree_32f(IppsIIRState_32f* pState); 

IppStatus ippsIIRFree_32fc(IppsIIRState_32fc* pState); 

IppStatus ippsIIRFree32f_16s(IppsIIRState32f_16s* pState); 

IppStatus ippsIIRFree32fc_16sc(IppsIIRState32fc_16sc* pState); 

IppStatus ippsIIRFree_64f(IppsIIRState_64f* pState); 

IppStatus ippsIIRFree_64fc(IppsIIRState_64fc* pState); 

IppStatus ippsIIRFree64f_32f(IppsIIRState64f_32f* pState); 

IppStatus ippsIIRFree64fc_32fc(IppsIIRState64fc_32fc* pState); 


6-39





IppStatus ippsIIRFree32s_16s(IppsIIRState32s_16s* pState); 

IppStatus ippsIIRFree32sc_16sc(IppsIIRState32sc_16sc* pState); 

 

pState 

IIR 

 

ippsIIRFree ippsIIRInitAlloc ippsIIRBQInitAlloc 

IIR 

ippsIIRFree 

 

 

ippStsNoErr 


 

NULL 


 

IIROne 

IIR 

 

IppStatus ippsIIROne_32f(Ipp32f src, Ipp32f* pDstVal, 

IppsIIRState_32f* pState); 

IppStatus ippsIIROne_32fc(Ipp32fc src, Ipp32fc* pDstVal, 

IppsIIRState_32fc* pState); 

6-40


IppStatus ippsIIROne32f_16s_Sfs(Ipp16s src, Ipp16s* pDstVal, 

IppsIIRState32f_16s* pState, int scaleFactor); 

IppStatus ippsIIROne32fc_16sc_Sfs(Ipp16sc src, Ipp16sc* pDstVal, 

IppsIIRState32fc_16sc* pState, int scaleFactor); 

IppStatus ippsIIROne_64f(Ipp64f src, Ipp64f* pDstVal, 


IppStatus ippsIIROne_64fc(Ipp64fc src, Ipp64fc* pDstVal, 


IppStatus ippsIIROne64f_32f(Ipp32f src, Ipp32f* pDstVal, 

IppsIIRState64f_32f* pState); 

IppStatus ippsIIROne64fc_32fc(Ipp32fc src, Ipp32fc* pDstVal, 

IppsIIRState64fc_32fc* pState); 









IppStatus ippsIIROne32s_16s_Sfs(Ipp16s src, Ipp16s* pDstVal, 

IppsIIRState32s_16s* pState, int scaleFactor); 

IppStatus ippsIIROne32sc_16sc_Sfs(Ipp16sc src, Ipp16sc* pDstVal, 

IppsIIRState32sc_16sc* pState, int scaleFactor); 

6-41


 

pState 

src 

pDstVal 

scaleFactor 

IIR 

ippsIIROne 

ippsIIROne 

 

2 

 

ippsIIROne IIR src 

pDstVal 

pState scaleFactor 

 

scaleFactor 

ippsIIR ippsFIROne ippsIIRInitAlloc 

ippsIIRBQInitAlloc 

tapsLenpTaps pDlyLine 

order numBq 

 

ippStsNoErr 


 

NULL 


 

IIR 

IIR 

 

IppStatus ippsIIR_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 


IppStatus ippsIIR_32f_I(Ipp32f* pSrcDst, int len, IppsIIRState_32f* pState); 

6-42


IppStatus ippsIIR_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len, 


IppStatus ippsIIR_32fc_I(Ipp32fc* pSrcDst, int len, 


IppStatus ippsIIR32f_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


IppStatus ippsIIR32f_16s_ISfs(Ipp16s* pSrcDst, int len, 


IppStatus ippsIIR32fc_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


IppStatus ippsIIR32fc_16sc_ISfs(Ipp16sc* pSrcDst, int len, 


IppStatus ippsIIR_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 


IppStatus ippsIIR_64f_I(Ipp64f* pSrcDst, int len, IppsIIRState_64f* pState); 

IppStatus ippsIIR_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, int len, 


IppStatus ippsIIR_64fc_I(Ipp64fc* pSrcDst, int len, 


IppStatus ippsIIR64f_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 


IppStatus ippsIIR64f_32f_I(Ipp32f* pSrcDst, int len, 


IppStatus ippsIIR64fc_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, int len, 


IppStatus ippsIIR64fc_32fc_I(Ipp32fc* pSrcDst, int len, 






6-43














IppStatus ippsIIR32s_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, int len, 


IppStatus ippsIIR32sc_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, int len, 


IppStatus ippsIIR32s_16s_ISfs(Ipp16s* pSrcDst, int len, 


IppStatus ippsIIR32sc_16sc_ISfs(Ipp16sc* pSrcDst, int len, 


 

pState 

pSrc 

pDst 

pSrcDst 

len 

scaleFactor 

IIR 

ippsIIR 

ippsIIR 

ippsIIR ( 

) 

ippsIIR 

2 

6-44


 

ippsIIR IIR pSrc 

pSrcDst len pDst 

pSrcDst pState 

scaleFactor 

scaleFactor 

ippsIIR ippsIIRInitAlloc ippsIIRBQInitAlloc 

tapsLen 

pTaps pDlyLine order numBq 

 

6-6 ippsIIR_32f 60Hz 

6-7 ippsIIR 

ippsCnvrt_64f32s_Sfs ippsIIRInitAlloc_32s 

 

 

ippStsNoErr 


ippStsSizeErr 

 

NULL 

len 


 

6-45


6-6 

ippsIIR_32f 60Hz 

IppStatus iir( void ) { 



int n; 


IppsIIRState_32f *ctx; 

Ipp32f *x = ippsMalloc_32f( NUMITERS ), *y = ippsMalloc_32f( NUMITERS ); 

/// A second-order notch filter having notch freq at 60 Hz 

const float taps[] = { 

0.940809f,-1.105987f,0.940809f,1,-1.105987f,0.881618f 

}; 

/// generate a signal having 60 Hz freq sampled with 400 Hz freq 

for(n=0;n> B = [0.940809,-1.105987,0.940809]; A = [1,-1.105987,0.881618]; 

n = 0:150; x = sin(2*pi*n*60/400); y = filter(B,A,x); y(121:125) 

6-46


6-7 

ippsIIR 

IppStatus iir16s( void ) { 



int n, tapsfactor = 30; 


IppsIIRState32s_16s *ctx; 

Ipp16s *x = ippsMalloc_16s( NUMITERS ), *y = ippsMalloc_16s( NUMITERS ); 

/// A second-order notch filter having notch freq at 60 Hz 

Ipp64f taps[6] = { 

0.940809f,-1.105987f,0.940809f,1,-1.105987f,0.881618f 

}; 

Ipp32s taps32s[6]; 

Ipp64f tmax, tmp[6]; 

ippsAbs_64f( taps, tmp, 6 ); 

ippsMax_64f( tmp, 6, &tmax ); 

tapsfactor = 0; 

if( tmax > IPP_MAX_32S ) 

while( (tmax/=2) > IPP_MAX_32S ) ++tapsfactor; 

else 

while( (tmax*=2) < IPP_MAX_32S ) --tapsfactor; 

if(tapsfactor >0 ) 

ippsDivC_64f_I( (float)(1


IIRGetDlyLineIIRSetDlyLine 

IIR 

 

IppStatus ippsIIRGetDlyLine_32f(const IppsIIRState_32f* pState, 


IppStatus ippsIIRSetDlyLine_32f(IppsIIRState_32f* pState, 


IppStatus ippsIIRGetDlyLine_32fc(const IppsIIRState_32fc* pState, 


IppStatus ippsIIRSetDlyLine_32fc(IppsIIRState_32fc* pState, 


IppStatus ippsIIRGetDlyLine32f_16s(const IppsIIRState32f_16s* pState, 


IppStatus ippsIIRSetDlyLine32f_16s(IppsIIRState32f_16s* pState, 


IppStatus ippsIIRGetDlyLine32fc_16sc(const IppsIIRState32fc_16sc* pState, 


IppStatus ippsIIRSetDlyLine32fc_16sc(IppsIIRState32fc_16sc* pState, 


IppStatus ippsIIRGetDlyLine_64f(const IppsIIRState_64f* pState, 


IppStatus ippsIIRSetDlyLine_64f(IppsIIRState_64f* pState, 


IppStatus ippsIIRGetDlyLine_64fc(const IppsIIRState_64fc* pState, 


IppStatus ippsIIRSetDlyLine_64fc(IppsIIRState_64fc* pState, 


6-48


IppStatus ippsIIRGetDlyLine64f_32f(const IppsIIRState64f_32f* pState, 


IppStatus ippsIIRSetDlyLine64f_32f(IppsIIRState64f_32f* pState, 


IppStatus ippsIIRGetDlyLine64fc_32fc(const IppsIIRState64fc_32fc* pState, 


IppStatus ippsIIRSetDlyLine64fc_32fc(IppsIIRState64fc_32fc* pState, 


















IppStatus ippsIIRGetDlyLine32s_16s(const IppsIIRState32s_16s* pState, 


IppStatus ippsIIRSetDlyLine32s_16s(IppsIIRState32s_16s* pState, 


6-49


IppStatus ippsIIRGetDlyLine32sc_16sc(const IppsIIRState32sc_16sc* pState, 


IppStatus ippsIIRSetDlyLine32sc_16sc(IppsIIRState32sc_16sc* pState, 


 

pState 

pDlyLine 

IIR 

 

orderBQ 2*numBq 

NULL 

 

 

ippsIIRGetDlyLine ippsIIRSetDlyLine IIR 

 

ippsIIRGetDlyLine ippsIIRGetDlyLine pState 

pDlyLine 

ippsIIRSetDlyLine ippsIIRSetDlyLine pDlyLine 

pState 

ippsIIRGetDlyLine ippsIIRSetDlyLine ippsIIRInitAlloc 

ippsIIPBQInitAlloc 

 

 

ippStsNoErr 


 

NULL 


 

6-50


 

 

( ) 

 

 

( Window ) 


 

 


 

 

 

FilterMedian 

 

IppStatus ippsFilterMedian_32f_I(Ipp32f* pSrcDst, int len, int maskSize); 

IppStatus ippsFilterMedian_64f_I(Ipp64f* pSrcDst, int len, int maskSize); 

IppStatus ippsFilterMedian_16s_I(Ipp16s* pSrcDst, int len, int maskSize); 

IppStatus ippsFilterMedian_8u_I(Ipp8u* pSrcDst, int len, int maskSize); 

IppStatus ippsFilterMedian_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

int len, int maskSize); 

IppStatus ippsFilterMedian_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsFilterMedian_16s(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsFilterMedian_8u(const Ipp8u* pSrc, Ipp8u* pDst, 


6-51


 

pSrcDst 

pSrc 

pDst 

len 

maskSize 

( ) 

ippsFilterMedian 

 


 

 

 

1 

 

 

ippsFilterMedian pSrc pSrcDst 

pDst pSrcDst 

 

x[-1]=x[0] x[len]=x[len-1] 

6-8 ippsFilterMedian_16s_I 

 

 

ippStsNoErr 


ippStsSizeErr 

 

NULL 

len 

ippStsEvenMedianMaskSize 

 

6-52


6-8 


void median(void) { 

Ipp16s x[8] = {1,2,127,4,5,0,7,8}; 

IppStatus status = ippsFilterMedian_16s_I(x, 8, 3); 

printf_16s(“median =”, x,8, status); 

} 

Output: 

median = 1 2 4 5 4 5 7 8 


>> x = [1 2 127 4 5 0 7 8]; medfilt1(x) 

6-53


6-54

7 

 

IPP IPP 

(DFT) (FFT) 

(DCT) 

 

 

flag hint 

PermPackCCS 

 

 

flag hint 

flag hint 

flag 7-1 flag 

flag 1 A B DFT 

 

7-1


7-1 

flag 

A B 

IPP_FFT_DIV_FWD_BY_N 1/N 1 1/N 

IPP_FFT_DIV_INV_BY_N 1 1/N 1/N 

IPP_FFT_DIV_BY_SQRTN 1/N 1/2 1/N 1/2 1/N 1/2 

 

IPP_FFT_NODIV_BY_ANY 1 1 1/N 

1/N 1/2 

hint 

7-2 hint 

hint 4-5 

8-27 

7-2 

 

hint 

ippAlgHintNone 

ippAlgHintFast 

ippAlgHintAccurate 

Pack 

 

 

 

 

Pack 

FFT 

( 2 FFT )Pack 

FFT 7-3 

ippsConjPack 7-6 

ConjPack 

7-2


Perm 

Perm FFT FFT 

Perm Pack 

1 Perm 1 

 

Perm FFT 7-3 

ippsConjPerm 7-4 

ConjPerm 

CCS 

CCS FFT 

 

CCS FFT 7-3 CCS 

1 

ippsConjCCS 7-8 

ConjCCS 

7-3 

PackPerm 

CCS FFT 

FFTReal 0 1 2 3 . . . N-2 N-1 N N+1 

Pack R 0 R 1 I 1 R 2 . . . I N/2-1 R N/2 

Perm R 0 R N/2 R 1 I 1 . . . R N/2-1 I N/2-1 

CCS R 0 0 R 1 I 1 . . . R N/2-1 I N/2-1 R N/2 0 

 

ConjPermConjPack ConjCCS FFT 

 

 

2 CCS 

 

 

 

7-3


ConjPerm 

Perm 

 

IppStatus ippsConjPerm_64fc_I(Ipp64fc* pSrcDst, int lenDst); 

IppStatus ippsConjPerm_32fc_I(Ipp32fc* pSrcDst, int lenDst); 

IppStatus ippsConjPerm_16sc_I(Ipp16sc* pSrcDst, int lenDst); 

IppStatus ippsConjPerm_64fc(const Ipp64f* pSrc, Ipp64fc* pDst, int lenDst); 

IppStatus ippsConjPerm_32fc(const Ipp32f* pSrc, Ipp32fc* pDst, int lenDst); 

IppStatus ippsConjPerm_16sc(const Ipp16s* pSrc, Ipp16sc* pDst, int lenDst); 

 

pSrc 

pDst 

pSrcDst 

lenDst 

 

 

( ) 

 

 

 

ippsConjPerm pSrc Perm 

pDst 

ippsConjPerm pSrcDst Perm 

pSrcDst 

7-4 Perm FFT 

 

 

7-1 ippsConjPerm 

 

7-4


 

ippStsNoErr 


ippStsSizeErr 

 

pSrcDstpDst pSrc NULL 

lenDst 

7-4 FFT 

 

 

 

 

FFT([1]) 1 {1, 0} 1 

FFT([1 2]) 3, -1 {3, 0}, {-1, 0} 2 

FFT([1 2 3]) 6, -1.5, 0.86 {6, 0}, {-1.5, 0.86}, {-1.5, -0.86} 3 

FFT([1 2 3 9]) 15, -7, -2, 7 {15, 0}, {-2, 7}, {-7, 0}, {-2, -7} 4 

7-1 

ippsConjPerm 

void ConjPerm(void) { 

Ipp16s x[8] = {1,2,3,5,6,7,8,9}; 

Ipp16sc zero={0,0}, y[6]; 

IppStatus st; 

ippsSet_16sc( zero, y, 6 ); 

st = ippsConjPerm_16sc( x, y, 6 ); 

printf_16sc(“Perm 6:”, y, 6, st ); 


st = ippsConjPerm_16sc( x, y, 5 ); 

printf_16sc(“Perm 5:”, y, 5, st ); 

} 

Output: 

Perm 6: {1,0} {3,5} {6,7} {2,0} {6,-7} {3,-5} 

Perm 5: {1,0} {2,3} {5,6} {5,-6} {2,-3} 

7-5


ConjPack 

Pack 

 

IppStatus ippsConjPack_64fc_I(Ipp64fc* pSrcDst, int lenDst); 

IppStatus ippsConjPack_32fc_I(Ipp32fc* pSrcDst, int lenDst); 

IppStatus ippsConjPack_16sc_I(Ipp16sc* pSrcDst, int lenDst); 

IppStatus ippsConjPack_64fc(const Ipp64f* pSrc, Ipp64fc* pDst, int lenDst); 

IppStatus ippsConjPack_32fc(const Ipp32f* pSrc, Ipp32fc* pDst, int lenDst); 

IppStatus ippsConjPack_16sc(const Ipp16s* pSrc, Ipp16sc* pDst, int lenDst); 

 

pSrc 

pDst 

pSrcDst 

lenDst 

 

 

( ) 

 

 

 

ippsConjPack pSrc Pack 

pDst 

ippsConjPack pSrcDst Pack 

pSrcDst 

7-5 Pack FFT 

 

 

7-2 ippsConjPack 

 

7-6


 

ippStsNoErr 


ippStsSizeErr 

 


lenDst 

7-5 Pack 

 

 

 

 

FFT([1]) 1 {1, 0} 1 

FFT([1 2]) 3, -1 {3, 0}, {-1, 0} 2 

FFT([1 2 3]) 6, -1.5, 0.86 {6, 0}, {-1.5, 0.86}, {-1.5, -0.86} 3 

FFT([1 2 3 9]) 15, -2, 7, -7 {15, 0}, {-2, 7}, {-7, 0}, {-2, -7} 4 

7-2 

ippsConjPack 

void ConjPack(void) { 

Ipp16s x[8] = {1,2,3,5,6,7,8,9}; 

Ipp16sc zero={0,0}, y[6]; 

IppStatus st; 


st = ippsConjPack_16sc( x, y, 6 ); 

printf_16sc(“pack 6”, y, 6, st ); 


st = ippsConjPack_16sc( x, y, 5 ); 

printf_16sc(“pack 5”, y, 5, st ); 

} 

Output: 

Pack 6: {1,0} {2,3} {5,6} {7,0} {5,-6} {2,-3} 

Pack 5: {1,0} {2,3} {5,6} {5,-6} {2,-3} 

7-7


ConjCCS 

CCS 

 

IppStatus ippsConjCCS_64fc_I(Ipp64fc* pSrcDst, int lenDst); 

IppStatus ippsConjCCS_32fc_I(Ipp32fc* pSrcDst, int lenDst); 

IppStatus ippsConjCCS_16sc_I(Ipp16sc* pSrcDst, int lenDst); 

IppStatus ippsConjCCS_64fc(const Ipp64f* pSrc, Ipp64fc* pDst, int lenDst); 

IppStatus ippsConjCCS_32fc(const Ipp32f* pSrc, Ipp32fc* pDst, int lenDst); 

IppStatus ippsConjCCS_16sc(const Ipp16s* pSrc, Ipp16sc* pDst, int lenDst); 

 

pSrc 

pDst 

pSrcDst 

lenDst 

 

 

( ) 

 

 

 

ippsConjCCS pSrc CCS 

pDst 

ippsConjCcs pSrcDst CCS 

pSrcDst 

7-6 CCS FFT 

 

 

CCS 

2 7-3 ippsConjCCS 

7-8


 

ippStsNoErr 


ippStsSizeErr 

 


lenDst 

7-6 CCS 

 

 

 

 

FFT([1]) 1, 0 {1, 0} 1 

FFT([1 2]) 3, 0, -1, 0 {3, 0}, {-1, 0} 2 

FFT([1 2 3]) 6, 0, -1.5, 0.86 {6, 0}, {-1.5, 0.86}, {-1.5, -0.86} 3 

FFT([1 2 3 9]) 15, 0, -2, 7, -7, 0 {15, 0}, {-2, 7}, {-7, 0}, {-2, -7} 4 

7-3 

ippsConjCCS 

void ConjCCS(void) { 

Ipp16s x[8] = {1,2,3,5,6,7,8,9}; 

Ipp16sc zero={0,0}, y[6]; 

IppStatus st; 


st = ippsConjCCS_16sc( x, y, 6 ); 

printf_16sc(“CCS 6:”, y, 6, st ); 


st = ippsConjCCS_16sc( x, y, 5 ); 

printf_16sc(“CCS 5:”, y, 5, st ); 

} 

Ouput: 

CCS 6: {1,2} {3,5} {6,7} {8,9} {6,-7} {3,-5} 

CCS 5: {1,2} {3,5} {6,7} {6,-7} {3,-5} 

7-9


 

Pack Perm 

ippsFFTFwd_RToPack 

ippsFFTInv_PackToR 

ippsMul Pack Perm 

 

• Pack 2 

• Perm 

 

order FFT N 2 (N = 2 order ) 

MulPackMulPerm 

Pack Perm 

2 

IppStatus ippsMulPack_16s_ISfs(const Ipp16s* pSrc, Ipp16s* pSrcDst, 

int length, int scaleFactor); 

IppStatus ippsMulPack_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

Ipp16s* pDst, int length, int scaleFactor); 

IppStatus ippsMulPack_32f_I(const Ipp32f* pSrc, Ipp32f* pSrcDst, int length); 

IppStatus ippsMulPack_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 

Ipp32f* pDst, int length); 

IppStatus ippsMulPack_64f_I(const Ipp64f* pSrc, Ipp64f* pSrcDst, int length); 

IppStatus ippsMulPack_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


IppStatus ippsMulPerm_16s_ISfs(const Ipp16s* pSrc, Ipp16s* pSrcDst, 

int length, int scaleFactor); 

IppStatus ippsMulPerm_16s_Sfs(const Ipp16s* pSrc1, const Ipp16s* pSrc2, 

Ipp16s* pDst, int length, int scaleFactor); 

IppStatus ippsMulPerm_32f_I(const Ipp32f* pSrc, Ipp32f* pSrcDst, int length); 

7-10


IppStatus ippsMulPerm_32f(const Ipp32f* pSrc1, const Ipp32f* pSrc2, 


IppStatus ippsMulPerm_64f_I(const Ipp64f* pSrc, Ipp64f* pSrcDst, int length); 

IppStatus ippsMulPerm_64f(const Ipp64f* pSrc1, const Ipp64f* pSrc2, 


 

pSrc1pSrc2 

pDst 

pSrc 

pSrcDst 

length 

scaleFactor 

 

pSrc1[n] * pSrc2[n] 

 

pSrcDst 

 

( ) 

 

 

2 

 

ippsMulPack ippsMulPerm pSrc1 pSrc2 

pDst 

ippsMulPack ippsMulPerm( ) pSrc 

pSrcDst pSrcDst 

 

Perm Pack 

 

FFT 

 

ippsMulPack ippsMulPack Pack 

ippsMulPerm ippsMulPerm Perm 

CCS 


 

7-4 ippsMulPack_32f_I 

7-11


 

ippStsNoErr 


ippStsSizeErr 

 

pSrcDstpDstpSrc1pSrc2 

pSrc NULL 

length 

7-4 

ippsMulPack 

void mulpack( void ) { 

Ipp32f x[8], X[8], h[8]={1.0f/3,1.0f/3,1.0f/3,0,0,0,0,0}, H[8]; 

IppStatus st; 

IppsFFTSpec_R_32f* spec; 

st = ippsFFTInitAlloc_R_32f(&spec, 3, IPP_FFT_DIV_INV_BY_N, 

ippAlgHintNone); 

ippsSet_32f( 3, x, 8 ); 

x[3] = 5; 

st = ippsFFTFwd_RToPack_32f( x, X, spec, NULL ); 

st = ippsFFTFwd_RToPack_32f( h, H, spec, NULL ); 

ippsMulPack_32f_I( H, X, 8 ); 

st = ippsFFTInv_PackToR_32f( X, x, spec, NULL ); 

printf_32f("filtered =", x, 8, st ); 

ippsFFTFree_R_32f( spec ); 

} 

Output: 

filtered = 3.0 3.0 3.0 3.666667 3.666667 3.666667 3.0 3.0 

Matlab* analog: 

>> x=3*ones(1,8); x(4)=5;h=zeros(1,8); h(1:3)=1/3; 

real(ifft(fft(x).*fft(h))) 

7-12


 

(FFT) 

FFT (DFT) 

FFT 2 

 

FFT 

 

 

 

hint ( 

) flag 

 

2 FFT PermPack 

CCS 

SIMD FFT 

 

 

 

FFTInitAlloc_CFFTInitAlloc_R 

 

 

IppStatus ippsFFTInitAlloc_C_16sc(IppsFFTSpec_C_16sc** pFFTSpec, int order, 

int flag, IppHintAlgorithm hint); 

IppStatus ippsFFTInitAlloc_C_16s(IppsFFTSpec_C_16s** pFFTSpec, int order, 


IppStatus ippsFFTInitAlloc_R_16s(IppsFFTSpec_R_16s** pFFTSpec, int order, 


IppStatus ippsFFTInitAlloc_C_32fc(IppsFFTSpec_C_32fc** pFFTSpec, int order, 


7-13


IppStatus ippsFFTInitAlloc_C_32f(IppsFFTSpec_C_32f** pFFTSpec, int order, 


IppStatus ippsFFTInitAlloc_R_32f(IppsFFTSpec_R_32f** pFFTSpec, int order, 


IppStatus ippsFFTInitAlloc_C_64fc(IppsFFTSpec_C_64fc** pFFTSpec, int order, 


IppStatus ippsFFTInitAlloc_C_64f(IppsFFTSpec_C_64f** pFFTSpec, int order, 


IppStatus ippsFFTInitAlloc_R_64f(IppsFFTSpec_R_64f** pFFTSpec, int order, 


 

flag 

hint 

order 

pFFTSpec 

flag flag 

hint 

hint 

flag hint 

FFT N=2 order 

FFT 

 

ippsFFTInitAlloc_C ippsFFTInitAlloc_R order 

flag hint 

FFT pFFTSpec order 

2 order 

ippsFFTInitAlloc_C ippsFFTInitAlloc_C FFT 

 

ippsFFTInitAlloc_R ippsFFTInitAlloc_R FFT 

 

 

ippStsNoErr 

ippsStsNullPtrErr 

ippStsFftOrderErr 

 

pFFTSpec NULL 

order 

7-14


ippStsFftFlagErr 


flag 

 

FFTFree_CFFTFree_R 

 

 

IppStatus ippsFFTFree_C_16sc(IppsFFTSpec_C_16sc* pFFTSpec); 

IppStatus ippsFFTFree_C_16s(IppsFFTSpec_C_16s* pFFTSpec); 

IppStatus ippsFFTFree_R_16s(IppsFFTSpec_R_16s* pFFTSpec); 

IppStatus ippsFFTFree_C_32fc(IppsFFTSpec_C_32fc* pFFTSpec); 

IppStatus ippsFFTFree_C_32f(IppsFFTSpec_C_32f* pFFTSpec); 

IppStatus ippsFFTFree_R_32f(IppsFFTSpec_R_32f* pFFTSpec); 

IppStatus ippsFFTFree_C_64fc(IppsFFTSpec_C_64fc* pFFTSpec); 

IppStatus ippsFFTFree_C_64f(IppsFFTSpec_C_64f* pFFTSpec); 

IppStatus ippsFFTFree_R_64f(IppsFFTSpec_R_64f* pFFTSpec); 

 

pFFTSpec 

FFT 

 

ippsFFTFree ippsFFTInit_C ippsFFTInit_R 

FFT pFFTSpec 

ippsFFTFree 

ippsFFTFree_C ippsFFTFree_C FFT 

 

ippsFFTFree_R ippsFFTFree_R FFT 

7-15


 

ippStsNoErr 


 

pFFTSpec NULL 


pFFTSpec 

FFTGetBufSize_CFFTGetBufSize_R 

FFT ( ) 

 

IppStatus ippsFFTGetBufSize_C_16sc(const IppsFFTSpec_C_16sc* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_C_16s(const IppsFFTSpec_C_16s* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_R_16s(const IppsFFTSpec_R_16s* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_C_32fc(const IppsFFTSpec_C_32fc* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_C_32f(const IppsFFTSpec_C_32f* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_R_32f(const IppsFFTSpec_R_32f* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_C_64fc(const IppsFFTSpec_C_64fc* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_C_64f(const IppsFFTSpec_C_64f* pFFTSpec, 

int* pSize); 

IppStatus ippsFFTGetBufSize_R_64f(const IppsFFTSpec_R_64f* pFFTSpec, 

int* pSize); 

7-16


 

pFFTSpec 

pSize 

FFT 

FFT 

 

ippsFFTGetBufSize pFFTSpec FFT 

( ) pSize 

ippsFFTGetBufSize_C ippsFFTGetBufSize_C FFT 

 

ippsFFTGetBufSize_R ippsFFTGetBufSize_R FFT 

 

 

ippStsNoErr 


 

pFFTSpec NULL 


pFFTSpec 

FFTFwd_CToCFFTInv_CToC 

 

(FFT) 

IppStatus ippsFFTFwd_CToC_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

const IppsFFTSpec_C_16sc* pFFTSpec, int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsFFTInv_CToC_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

const IppsFFTSpec_C_16sc* pFFTSpec, int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsFFTFwd_CToC_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 

Ipp16s* pDstRe, Ipp16s* pDstIm, const IppsFFTSpec_C_16s* pFFTSpecx, 

int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsFFTInv_CToC_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 

Ipp16s* pDstRe, Ipp16s* pDstIm, const IppsFFTSpec_C_16s* pFFTSpec, 


7-17


IppStatus ippsFFTFwd_CToC_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

const IppsFFTSpec_C_32fc* pFFTSpec, Ipp8u* pBuffer); 

IppStatus ippsFFTInv_CToC_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 


IppStatus ippsFFTFwd_CToC_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 

Ipp32f* pDstRe, Ipp32f* pDstIm, const IppsFFTSpec_C_32f* pFFTSpec, 

Ipp8u* pBuffer); 

IppStatus ippsFFTInv_CToC_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 



IppStatus ippsFFTFwd_CToC_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 


IppStatus ippsFFTInv_CToC_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 


IppStatus ippsFFTFwd_CToC_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 



IppStatus ippsFFTInv_CToC_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 



 

pFFTSpec 

pSrc 

pDst 

pSrcRe 

pSrcIm 

pDstRe 

pDstIm 

pBuffer 

scaleFactor 

FFT 

 

 

 

 

 

 

FFT 

2 

7-18


 

ippsFFTFwd_CToC ippsFFTInv_CToC pFFTSpec ( 

order flag 

hint) FFT 

(32fc ) 

pSrc pDst 

pSrcRe pSrcIm 

(32f ) pDstRe 

pDstIm 

scaleFactor 

pBuffer FFT 

 

FFT 

2 

ippsFFTFwd_CToC ippsFFTFwd_CToC FFT 

FFT 2 

ippsFFTInv_CToC ippsFFTInv_CToC FFT 

FFT 2 

 

ippStsNoErr 


 

NULL 


pFFTSpec 


 

7-19


FFTFwd_RToPermFFTInv_PermToR 

FFTFwd_RToPackFFTInv_PackToR 

FFTFwd_RToCCSFFTInv_CCSToR 

 

(FFT) 

IppStatus ippsFFTFwd_RToPerm_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 

const IppsFFTSpec_R_16s* pFFTSpec, int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsFFTFwd_RToPack_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsFFTFwd_RToCCS_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsFFTInv_PermToR_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsFFTInv_PackToR_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsFFTInv_CCSToR_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsFFTFwd_RToPerm_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

const IppsFFTSpec_R_32f* pFFTSpec, Ipp8u* pBuffer); 

IppStatus ippsFFTFwd_RToPack_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsFFTFwd_RToCCS_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsFFTInv_PermToR_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsFFTInv_PackToR_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsFFTInv_CCSToR_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


7-20


IppStatus ippsFFTFwd_RToPerm_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsFFTFwd_RToPack_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsFFTFwd_RToCCS_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsFFTInv_PermToR_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsFFTInv_PackToR_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsFFTInv_CCSToR_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


 

pFFTSpec 

pSrc 

pDst 

pBuffer 

scaleFactor 

FFT 

 

 

 

 

 

2 

 

pFFTSpec ( order 

flag hint) 

FFT 

( ) PackPerm 

CCS 

FFT 

scaleFactor 

pBuffer 

FFT 

FFT 

2 

7-21


ippsFFTFwd_RToPermippsFFTInv_PermToR 

FFT Perm FFT 2 

 

ippsFFTFwd_RToPackippsFFTInv_PackToR 

FFT Pack FFT 2 

 

ippsFFTFwd_RToCCSippsFFTInv_CCSToR 

FFT CCS FFT 2 

 

7-7 PackPerm CCS 

7-5 ippsFFTFwd_RToCCS_32f 

 

 

ippStsNoErr 


 

NULL 


pFFTSpec 


 

7-7 

PackPerm 

CCS FFT 

FFTReal 0 1 2 3 . . . N-2 N-1 N N+1 

Pack R 0 R 1 I 1 R 2 . . . I N/2-1 R N/2 

Perm R 0 R N/2 R 1 I 1 . . . R N/2-1 I N/2-1 

CCS R 0 0 R 1 I 1 . . . R N/2-1 I N/2-1 R N/2 0 

7-22


7-5 

ippsFFTFwd_RToCCS 

IppStatus fft( void ) { 

Ipp32f x[8], X[10]; 

int n; 


IppsFFTSpec_R_32f* spec; 

status = ippsFFTInitAlloc_R_32f(&spec, 3, IPP_FFT_DIV_INV_BY_N, 

ippAlgHintNone ); 

for(n=0; n> n=0:7; x=sin(2*pi*n*16/64); X=abs(fft(x)); X(1:4) 

7-23


 

 

(DFT) DFT 

DFT 

hint ( 

) flag 

 

2 FFT PermPack 

CCS 

DFT 

 

 

[Mit93]8-2 Fast Computation of the DFT 

 

DFTInitAlloc_CDFTInitAlloc_R 

 

 

IppStatus ippsDFTInitAlloc_C_16sc(IppsDFTSpec_C_16sc** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_C_16s(IppsDFTSpec_C_16s** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_R_16s(IppsDFTSpec_R_16s** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_C_32fc(IppsDFTSpec_C_32fc** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_C_32f(IppsDFTSpec_C_32f** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_R_32f(IppsDFTSpec_R_32f** pDFTSpec, int length, 


7-24


IppStatus ippsDFTInitAlloc_C_64fc(IppsDFTSpec_C_64fc** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_C_64f(IppsDFTSpec_C_64f** pDFTSpec, int length, 


IppStatus ippsDFTInitAlloc_R_64f(IppsDFTSpec_R_64f** pDFTSpec, int length, 


 

flag 

hint 

length 

pDFTSpec 

flag flag 

hint 

hint 

flag hint 

DFT 

DFT 

 

ippsDFTInitAlloc_C ippsDFTInitAlloc_R length 

flag hint 

DFT pDFTSpec length 

 

ippsDFTInitAlloc_C ippsDFTInitAlloc_C DFT 

 

ippsDFTInitAlloc_R ippsDFTInitAlloc_R DFT 

 

 

ippStsNoErr 

ippsNullPtrErr 

ippStsFftOrderErr 

ippStsFftFlagErr 


 

pDFTSpec NULL 

order 

flag 

 

7-25


DFTFree_CDFTFree_R 

 

 

IppStatus ippsDFTFree_C_16sc(IppsDFTSpec_C_16sc* pDFTSpec); 

IppStatus ippsDFTFree_C_16s(IppsDFTSpec_C_16s* pDFTSpec); 

IppStatus ippsDFTFree_R_16s(IppsDFTSpec_R_16s* pDFTSpec); 

IppStatus ippsDFTFree_C_32fc(IppsDFTSpec_C_32fc* pDFTSpec); 

IppStatus ippsDFTFree_C_32f(IppsDFTSpec_C_32f* pDFTSpec); 

IppStatus ippsDFTFree_R_32f(IppsDFTSpec_R_32f* pDFTSpec); 

IppStatus ippsDFTFree_C_64fc(IppsDFTSpec_C_64fc* pDFTSpec); 

IppStatus ippsDFTFree_C_64f(IppsDFTSpec_C_64f* pDFTSpec); 

IppStatus ippsDFTFree_R_64f(IppsDFTSpec_R_64f* pDFTSpec); 

 

pDFTSpec 

DFT 

 

ippsDFTFree ippsDFTInitAlloc_C ippsDFTInitAlloc_R 

DFT 

pDFTSpec ippsDFTFree 

ippsDFTFree_C ippsDFTFree_C DFT 

 

ippsDFTFree_R ippsDFTFree_R DFT 

7-26


 

ippStsNoErr 


 

pDFTSpec NULL 


pDFTSpec 

DFTGetBufSize_CDFTGetBufSize_R 

DFT ( ) 

 

IppStatus ippsDFTGetBufSize_C_16sc(const IppsDFTSpec_C_16sc* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_C_16s(const IppsDFTSpec_C_16s* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_R_16s(const IppsDFTSpec_R_16s* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_C_32fc(const IppsDFTSpec_C_32fc* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_C_32f(const IppsDFTSpec_C_32f* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_R_32f(const IppsDFTSpec_R_32f* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_C_64fc(const IppsDFTSpec_C_64fc* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_C_64f(const IppsDFTSpec_C_64f* pDFTSpec, 

int* pSize); 

IppStatus ippsDFTGetBufSize_R_64f(const IppsDFTSpec_R_64f* pDFTSpec, 

int* pSize); 

7-27


 

pDFTSpec 

pSize 

DFT 

DFT 

 

ippsDFTGetBufSize pDFTSpec DFT 

( ) pSize 

ippsDFTGetBufSize_C ippsDFTGetBufSize_C DFT 

 

ippsDFTGetBufSize_R ippsDFTGetBufSize_R DFT 

 

 

ippStsNoErr 


 

pDFTSpec NULL 


pDFTSpec 

DFTFwd_CToCDFTInv_CToC 

 

(DFT) 

IppStatus ippsDFTFwd_CToC_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

const IppsDFTSpec_C_16sc* pDFTSpec, int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsDFTInv_CToC_16sc_Sfs(const Ipp16sc* pSrc, Ipp16sc* pDst, 

const IppsDFTSpec_C_16sc* pDFTSpec, int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsDFTFwd_CToC_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 

Ipp16s* pDstRe, Ipp16s* pDstIm, const IppsDFTSpec_C_16s* pDFTSpec, 


IppStatus ippsDFTInv_CToC_16s_Sfs(const Ipp16s* pSrcRe, const Ipp16s* pSrcIm, 

Ipp16s* pDstRe, Ipp16s* pDstIm, const IppsDFTSpec_C_16s* pDFTSpec, 


7-28


IppStatus ippsDFTFwd_CToC_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 

const IppsDFTSpec_C_32fc* pDFTSpec, Ipp8u* pBuffer); 

IppStatus ippsDFTInv_CToC_32fc(const Ipp32fc* pSrc, Ipp32fc* pDst, 


IppStatus ippsDFTFwd_CToC_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 

Ipp32f* pDstRe, Ipp32f* pDstIm, const IppsDFTSpec_C_32f* pDFTSpec, 


IppStatus ippsDFTInv_CToC_32f(const Ipp32f* pSrcRe, const Ipp32f* pSrcIm, 



IppStatus ippsDFTFwd_CToC_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 


IppStatus ippsDFTInv_CToC_64fc(const Ipp64fc* pSrc, Ipp64fc* pDst, 


IppStatus ippsDFTFwd_CToC_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 



IppStatus ippsDFTInv_CToC_64f(const Ipp64f* pSrcRe, const Ipp64f* pSrcIm, 



 

pDFTSpec 

pSrc 

pDst 

pSrcRe 

pSrcIm 

pDstRe 

pDstIm 

pBuffer 

scaleFactor 

DFT 

 

 

 

 

 

 

 

2 

7-29


 

ippsDFTFwd_CToC ippsDFTInv_CToC pDFTSpec ( 

length flag 

hint) DFT 

(32fc ) 

pSrc pDst 

pSrcRe pSrcIm 

(32f ) pDstRe 

pDstIm 

scaleFactor 

pBuffer DFT 

 

DFT 

2 

DFT 

Xk ( ) = A ∑ x( n) ⋅ exp – j2π kn ------ , 

⎝ 

⎛ N ⎠ 

⎞ 

k n N 

lengthA B flag x(n) 

pSrc[n]X(k) pDst[k] x(n) pDst[n]X(k) 

pSrc[k] 

 

ippsDFTFwd_CToC ippsDFTFwd_CToC DFT 

ippsDFTInv_CToC ippsDFTInv_CToC DFT 

 

N – 1 

n = 0 

N – 1 

xk ( ) = B ∑ X( n) ⋅ exp j2π kn ------ 

⎝ 

⎛ N ⎠ 

⎞ 

n = 0 

ippStsNoErr 


 

NULL 


pDFTSpec 


 

7-30


DFTFwd_RToPermDFTInv_PermToR 

DFTFwd_RToPackDFTInv_PackToR 

DFTFwd_RToCCSDFTInv_CCSToR 

 

(DFT) 

IppStatus ippsDFTFwd_RToPerm_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 

const IppsDFTSpec_R_16s* pDFTSpec, int scaleFactor, Ipp8u* pBuffer); 

IppStatus ippsDFTFwd_RToPack_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsDFTFwd_RToCCS_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsDFTInv_PermToR_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsDFTInv_PackToR_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus, ippsDFTInv_CCSToR_16s_Sfs(const Ipp16s* pSrc, Ipp16s* pDst, 


IppStatus ippsDFTFwd_RToPerm_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

const IppsDFTSpec_R_32f* pDFTSpec, Ipp8u* pBuffer); 

IppStatus ippsDFTFwd_RToPack_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsDFTFwd_RToCCS_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsDFTInv_PermToR_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsDFTInv_PackToR_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


IppStatus ippsDFTInv_CCSToR_32f(const Ipp32f* pSrc, Ipp32f* pDst, 


7-31


IppStatus ippsDFTFwd_RToPerm_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsDFTFwd_RToPack_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsDFTFwd_RToCCS_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsDFTInv_PermToR_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsDFTInv_PackToR_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


IppStatus ippsDFTInv_CCSToR_64f(const Ipp64f* pSrc, Ipp64f* pDst, 


 

pDFTSpec 

pSrc 

pDst 

pBuffer 

scaleFactor 

DFT 

 

 

 

 

 

2 

 

pDFTSpec ( length 

flag hint) 

DFT 

( ) PackPerm 

CCS 

DFT 

scaleFactor 

pBuffer DFT 

 

DFT 

2 

7-32


DFT 

Xk ( ) = A ∑ x( n) ⋅ exp – j2π kn ------ 

⎝ 

⎛ , 

N ⎠ 

⎞ 

k n N 

lengthA B flag x(n) 

pSrc[n]X(k) pDst[k] x(n) pDst[n]X(k) 

pSrc[k] 

 

ippsDFTFwd_RToPermippsDFTInv_PermToR 

DFT Perm 

ippsDFTFwd_RToPackippsDFTInv_PackToR 

DFT Pack 

ippsDFTFwd_RToCCSippsDFTInv_CCSToR 

DFT CCS 

7-6 ippsDFTFwd_RToCCS_32f 

 

N – 1 

n = 0 

N – 1 

xk ( ) = B ∑ X( n) ⋅ exp j2π kn ------ 

⎝ 

⎛ N ⎠ 

⎞ 

n = 0 

ippStsNoErr 


 

NULL 


pDFTSpec 


 

7-33


7-6 

ippsDFTFwd_RToCCS 

IppStatus dft( void ) { 

Ipp32f x[7], X[8]; 

int n; 


IppsDFTSpec_R_32f* spec; 

status = ippsDFTInitAlloc_R_32f(&spec, 7, IPP_FFT_DIV_INV_BY_N, 

ippAlgHintNone); 

for( n=0; n> N=7;F=14/49;n=0:N-1;x=cos(2*pi*n*F);y=abs(fft(x));y(1:4) 

7-34


DFT(Goertzel) 

1 

DFT 01/N2/N 

... (N-1)/N N 

 

IPP Goertzel [Mit98] ( ) 

DFT 

1 2 

( 

) SIMD 

 

Goertz 

DFT 

 

IppStatus ippsGoertz_32fc(const Ipp32fc* pSrc, int len, Ipp32fc* pVal, 

Ipp32f freq); 

IppStatus ippsGoertz_64fc(const Ipp64fc* pSrc, int len, Ipp64fc* pVal, 

Ipp64f freq); 

IppStatus ippsGoertz_16sc_Sfs(const Ipp16sc* pSrc, int len, Ipp16sc* pVal, 

Ipp32f freq, int scaleFactor); 

 

freq 

pSrc 

len 

pVal 

scaleFactor 

[0, 1.0) 

 

 

DFT 

2 

7-35


 

ippsGoertz freq len pSrc 

DFT pVal 

ippsGoertz 

N – 1 

yk ( ) = ∑ xn ( ) ⋅ exp – j2π kn ------ 

⎝ 

⎛ N ⎠ 

⎞ 

n = 0 

, 

k/N DFT freq 

7-7 DFT Goertzel 

 

 

ippStsNoErr 


ippStsRelFreqErr 

ippStsSizeErr 

 

NULL 

freq 

len 

7-36


7-7 

Goertzel 

IppStatus goertzel( void ) { 

#undef LEN 



Ipp32fc *x = ippsMalloc_32fc( LEN ), y; 

int n; 

///generate a signal of 60 Hz freq that 

/// is sampled with 400 Hz freq 

for( n=0; n> N=100;F=60/400;n=0:N-1;x=sin(2*pi*n*F);y=fft(x);n=N*F;y(n+1) 

7-37


GoertzTwo 

2 

DFT 

IppStatus ippsGoertzTwo_32fc(const Ipp32fc* pSrc, int len, Ipp32fc pVal[2], 

const Ipp32f freq[2]); 

IppStatus ippsGoertzTwo_64fc(const Ipp64fc* pSrc, int len, Ipp64fc pVal[2], 

const Ipp64f freq[2]); 

IppStatus ippsGoertzTwo_16sc_Sfs(const Ipp16sc* pSrc, int len, 

Ipp16sc pVal[2], const Ipp32f freq[2], int scaleFactor); 

 

freq 

pSrc 

len 

pVal 

scaleFactor 

2 [0, 1.0) 

 

 

DFT 

2 

 

ippsGoertzTwo 2 freq len pSrc 

DFT pVal ® Pentium ® III 

1 DFT 2 DFT 

DFT 

ippsGoertz 

N – 1 

yk ( ) = ∑ xn ( ) ⋅ exp – j2π kn ------ 

⎝ 

⎛ N ⎠ 

⎞ 

n = 0 

k/N DFT freq 1 

, 

7-38


 

ippStsNoErr 


ippStsRelFreqErr 

ippStsSizeErr 

 

NULL 

freq 

len 

 

 

DCTFwdInitAllocDCTInvInitAlloc 

 

IppStatus ippsDCTFwdInitAlloc_32f(IppsDCTFwdSpec_32f** pDCTSpec, int length, 

IppHintAlgorithm hint); 

IppStatus ippsDCTInvInitAlloc_32f(IppsDCTInvSpec_32f** pDCTSpec, int length, 


IppStatus ippsDCTFwdInitAlloc_64f(IppsDCTFwdSpec_64f** pDCTSpec, int length, 


IppStatus ippsDCTInvInitAlloc_64f(IppsDCTInvSpec_64f** pDCTSpec, int length, 


 

flag 

hint 

length 

pDCTSpec 

flag flag 

hint 

hint 

flag hint 

DCT 

DCT 

7-39


 

ippsDCTFwdInitAlloc ippsDCTInvInitAlloc length 

flag hint 

DCT pDCTSpec length 

 

ippsDCTFwdInitAlloc ippsDCTFwdInitAlloc DCT 

 

ippsDCTInvInitAlloc ippsDCTInvInitAlloc DCT 

 

 

ippStsNoErr 


ippStsSizeErr 


 

pDCTSpec NULL 

length 

 

DCTFwdFreeDCTInvFree 

 

IppStatus ippsDCTFwdFree_32f(IppsDCTFwdSpec_32f* pDCTSpec); 

IppStatus ippsDCTInvFree_32f(IppsDCTInvSpec_32f* pDCTSpec); 

IppStatus ippsDCTFwdFree_64f(IppsDCTFwdSpec_64f* pDCTSpec); 

IppStatus ippsDCTInvFree_64f(IppsDCTInvSpec_64f* pDCTSpec); 

 

pDCTSpec 

DCT 

7-40


 

ippsFFTFwdFree ippsFFTInvFree ippsDCTFwdInitAlloc 

ippsDCTInvInitAlloc 

DCT pDCTSpec ippsDCTFwdFree 

ippsDCTInvFree 

ippsDCTFwdFree ippsDCTFwdFree DCT 

 

ippsDCTInvFree ippsDCTInvFree DCT 

 

 

ippStsNoErr 


 

pDCTSpec NULL 


pDCTSpec 

DCTFwdGetBufSize 

DCTInvGetBufSize 

DCT ( ) 

 

IppStatus ippsDCTFwdGetBufSize_32f(const IppsDCTFwdSpec_32f* pDCTSpec, 

int* pSize); 

IppStatus ippsDCTInvGetBufSize_32f(const IppsDCTInvSpec_32f* pDCTSpec, 

int* pSize); 

IppStatus ippsDCTFwdGetBufSize_64f(const IppsDCTFwdSpec_64f* pDCTSpec, 

int* pSize); 

IppStatus ippsDCTInvGetBufSize_64f(const IppsDCTInvSpec_64f* pDCTSpec, 

int* pSize); 

7-41


 

pDCTSpec 

pSize 

DCT 

DCT 

 

ippsDCTFwdGetBufSize ippsDCTInvGetBufSize pDCTSpec 

DCT ( ) pSize 

 

ippsDCTFwdGetBufSize ippsDCTFwdGetBufSize DCT 

 

ippsDCTInvGetBufSize ippsDCTInvGetBufSize DCT 

 

 

ippStsNoErr 


 

pDCTSpec NULL 


pDCTSpec 

7-42


DCTFwdDCTInv 

 

(DCT) 

IppStatus ippsDCTFwd_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

const IppsDCTFwdSpec_32f* pDCTSpec, Ipp8u* pBuffer); 

IppStatus ippsDCTInv_32f(const Ipp32f* pSrc, Ipp32f* pDst, 

const IppsDCTInvSpec_32f* pDCTSpec, Ipp8u* pBuffer); 

IppStatus ippsDCTFwd_64f(const Ipp64f* pSrc, Ipp64f* pDst, 

const IppsDCTFwdSpec_64f* pDCTSpec, Ipp8u* pBuffer); 

IppStatus ippsDCTInv_64f(const Ipp64f* pSrc, Ipp64f* pDst, 

const IppsDCTInvSpec_64f* pDCTSpec, Ipp8u* pBuffer); 

 

pDCTSpec 

pSrc 

pDst 

pBuffer 

scaleFactor 

DCT 

 

 

DCT 

2 

 

ippsDCTFwd ippsDCTInv (DCT) 

length 2 DCT 

length 

 

 

7-43


DCT N = length 

1 

2 

Ck ( ) = ------- (k = 0 ) Ck ( ) = ------- (k > 0 ); 

N 

N 

DCT x(n) pSrc[n]y(k) pDst[k]; DCT 

x(n) pDst[n]y(k) pSrc[k] 

DCT 

N – 1 

( 2n + 1)πk 

yk ( ) = C( k) ∑ xn ( ) ⋅ cos ---------------------------- 

2N 

n = 0 

 

N – 1 

( 2n + 1)πk 

xn ( ) = ∑ C( k)yk 

( ) ⋅ cos ---------------------------- 

2N 

k = 0 

scaleFactor 

pBuffer DCT 

 

DCT 

2 

ippsDCTFwd ippsDCTFwd DCT 

ippsDCTInv ippsDCTInv DCT 

7-8 ippsDCTFwd_32f ippsDCTInv_32f 

 

 

ippStsNoErr 


 

pSrcpDst pDCTSpec NULL 


pDCTSpec 


 

7-44


7-8 ippsDCTFwd ippsDCTInv 

void dct( void ) { 


Ipp32f x[LEN], y[LEN]; 

int n; 

IppsDCTFwdSpec_32f* fspec; 

IppsDCTInvSpec_32f* ispec; 


/// data: Gaussian function, magn =1 and sigma=N/3 

for(n=0; n


 

 

 

 

“ ” “ ” 2 

a i,k 

1 “ ” 

1 2 n 

 

7-1 

 


(DWT) 

DWT 2 

DWT 

 

1 

DWT 7-2 DWT 

7-1 3 

7-3 1 

 

2 

 


 

7-46


2 

• 

 

• 

 

 

7-1 

 

Frequency 

8 

7 

6 

5 

4 

3 

2 

1 

0 

a 2,0 a 2,1 a 2,2 a 2,3 

a 1,0 

a 1,1 a 1,2 a 1,3 a 1,4 a 1,5 a 1,6 a 1,7 

a 3,0 

a 3,1 

Time 

0 1 2 3 4 5 6 7 8 9 

7-47


7-2 

3 

 

 

 

FWD 

1{a 1,0 a 1.1 ... a 1,6 a 1.7 } 

2{a 2,0 a 2.1 a 2,2 a 2.3 } 

FWD 

3 {a 2,0 a 2.1 } 

FWD 

{c 0 } 

7-3 

3 

1{a 1,0 a 1.1 ... a 1,6 a 1.7 } 

2{a 2,0 a 2.1 a 2,2 a 2.3 } 

INV 

 

 

 

3 {a 2,0 a 2.1 } 

INV 

{c 0 } 

INV 

7-48


 

 

 

WTHaarFwdWTHaarInv 

 

Haar 

IppStatus ippsWTHaarFwd_8s(const Ipp8s* pSrc, int lenSrc, 

Ipp8s* pDstLow, Ipp8s* pDstHigh); 







IppStatus ippsWTHaarFwd_32f(const Ipp32f* pSrc, int lenSrc, 

Ipp32f* pDstLow, Ipp32f* pDstHigh); 

IppStatus ippsWTHaarFwd_64f(const Ipp64f* pSrc, int lenSrc, 

Ipp64f* pDstLow, Ipp64f* pDstHigh); 

IppStatus ippsWTHaarFwd_8s_Sfs(const Ipp8s* pSrc, int lenSrc, 

Ipp8s* pDstLow, Ipp8s* pDstHigh, int scaleFactor); 


Ipp16s* pDstLow, Ipp16s* pDstHigh, int scaleFactor ); 





IppStatus ippsWTHaarInv_8s(const Ipp8s* pSrcLow, const Ipp8s* pSrcHigh, 

Ipp8s* pDst, int lenDst); 

7-49








IppStatus ippsWTHaarInv_32f(const Ipp32f* pSrcLow, const Ipp32f* pSrcHigh, 

Ipp32f* pDst, int lenDst); 

IppStatus ippsWTHaarInv_64f(const Ipp64f* pSrcLow, const Ipp64f* pSrcHigh, 

Ipp64f* pDst, int lenDst); 

IppStatus ippsWTHaarInv_8s_Sfs(const Ipp8s* pSrcLow, const Ipp8s* pSrcHigh, 

Ipp8s* pDst, int lenDst, int scaleFactor); 







 

pSrc 

lenSrc 

pDstLow 

pDstHigh 

pSrcLow 

pSrcHigh 

pDst 

lenDst 

scaleFactor 

ippsWTHaarFwd 

pSrc 

ippsWTHaarFwd“” 

 

ippsWTHaarFwd“” 

 

ippsWTHaarInv“” 

 

ippsWTHaarInv“” 

 

ippsWTHaarInv 

pDst 

2 

7-50


 

ippsWTHaar Haar 

 

{1/2, 1/2} 

{1/2, -1/2} 

 

{1, 1} 

{-1, 1} 

 

 

1 0.5 

 

1 

 

2 scaleFactor = -1 

 

( [Str96] ) 

{2 -1/2 , 2 -1/2 } {2 -1/2 , -2 -1/2 } 

{2 -1/2 , 2 -1/2 } {-2 -1/2 , 2 -1/2 } 

Haar N=lenSrc “ 

” c(k) pDstLow[k] “ ” d(k) pDstHigh[k] 

x(2k) x(2k+1) pSrc 

c(k)=(x(2k)+x(2k+1))/2 

d(k)=(x(2k+1)-x(2k))/2 

N=lenDst “ ” c(k) pSrcLow[k] 

“ ” d(k) pSrcHigh[k] y(2i) y(2i+1) 

pDst 

y(2i)=c(i)-d(i) 

y(2i+1)=c(i)+d(i) 

N 0≤k


N N+1 

2 x[N]=x[N-1] 

 

c((N+1)/2-1)=x(N-1) 

d((N+1)/2-1)=0 

 

y(N)=y(N-1)=c((N+1)/2-1) 

N 0≤k


 

ippStsNoErr 


ippStsSizeErr 

 


len ippsWinBalckmanOpt 4 

3 

 

7-9 

ippsWTHaarFwd 

IppStatus wthaar(void) { 

Ipp32f x[8], lo[4], hi[4]; 


ippsSet_32f(7, x, 8); --x[4]; 

status = ippsWTHaarFwd_32f(x, 8, lo, hi); 

printf_32f(“WT Haar low =”, lo, 4, status); 

printf_32f(“WT Haar high =”, hi, 4, status); 

return status; 

} 

Output: 

WT Haar low = 7.000000 7.000000 6.500000 7.000000 

WT Haar high = 0.000000 0.000000 0.500000 0.000000 

 

[Str96], P.153-157, Wavelet and Filter Banks, 

Wellesley-Cambridge Press [Bri94], Classification of Nonexpansive Symmetric Extension 

Transforms for Multirate Filter Banks, Los Alamos Report LA-UR-94-1747, 1994 

 

7-53


 

 

 

WTFwdInitAllocWTInvInitAlloc 

 

IppStatus ippsWTFwdInitAlloc_32f(IppsWTFwdState_32f** pState, 

const Ipp32f* pTapsLow, int lenLow, int offsLow, 

const Ipp32f* pTapsHigh, int lenHigh, int offsHigh); 

IppStatus ippsWTFwdInitAlloc_8s32f(IppsWTFwdState_8s32f** pState, 



IppStatus ippsWTFwdInitAlloc_8u32f(IppsWTFwdState_8u32f** pState, 



IppStatus ippsWTFwdInitAlloc_16s32f(IppsWTFwdState_16s32f** pState, 



IppStatus ippsWTFwdInitAlloc_16u32f(IppsWTFwdState_16u32f** pState, 



IppStatus ippsWTInvInitAlloc_32f(IppsWTInvState_32f** pState, 



IppStatus ippsWTInvInitAlloc_32f8s(IppsWTInvState_32f8s** pState, 



IppStatus ippsWTInvInitAlloc_32f8u(IppsWTInvState_32f8u** pState, 



7-54


IppStatus ippsWTInvInitAlloc_32f16s(IppsWTInvState_32f16s** pState, 



IppStatus ippsWTInvInitAlloc_32f16u(IppsWTInvState_32f16u** pState, 



 

pState 

 

pTapsLow 

lenLow 

offsLow ( ) 

pTapsHigh 

lenHigh 

 

 

offsHigh ( ) 

 

ippsWTFwdInitAlloc ippsWTInvInitAlloc 

pTapsLow pTapsHigh lenLow lenHigh 

offsLow offsHigh WT 

pState 

ippsWTFwdInitAlloc ippsWTFwdInitAlloc WT 

 

ippsWTInvInitAlloc ippsWTInvInitAlloc WT 

 

 

 

pState 

 

2 

 

 

pTapsHigh pTapsLow lenHigh lenLow 

7-55


pTapsHigh pTapsLow 

lenHigh lenLow 

offsLow offsHigh 

 

• 

• 

 

7-58 WTFwd 7-64 

WTInv -1 

 

7-58 WTFwd 7-64 WTInv 

pState 

 

 

 

 

7-61 

ippsWTFwdSetDlyLine 

 

 

 

 

 

ippStsNoErr 


ippStsSizeErr 

ippStsWtOffsetErr 

 

pStatepTapsHigh pTapsLow 

NULL 

lenLow lenHigh 

offsLow offsHigh 

-1 

7-56


WTFwdFreeWTInvFree 

 

IppStatus ippsWTFwdFree_32f(IppsWTFwdState_32f* pState); 

IppStatus ippsWTFwdFree_8s32f(IppsWTFwdState_8s32f* pState); 

IppStatus ippsWTFwdFree_8u32f(IppsWTFwdState_8u32f* pState); 

IppStatus ippsWTFwdFree_16s32f(IppsWTFwdState_16s32f* pState); 

IppStatus ippsWTFwdFree_16u32f(IppsWTFwdState_16u32f* pState); 

IppStatus ippsWTInvFree_32f(IppsWTInvState_32f* pState); 

IppStatus ippsWTInvFree_32f8s(IppsWTInvState_32f8s* pState); 

IppStatus ippsWTInvFree_32f8u(IppsWTInvState_32f8u* pState); 

IppStatus ippsWTInvFree_32f16s (IppsWTInvState_32f16s* pState); 

IppStatus ippsWTInvFree_32f16u(IppsWTInvState_32f16u* pState); 

 

pState 

 

 

ippsWTFwdFree ippsWTInvFree ippsWTFwdInitAlloc 

ippsWTInvInitAlloc 

WT pState 

ippsWTFwdFree ippsWTInvFree pState 

NULL ippStsNullPtrErr 

ippsWTFwdFree ippsWTFwdFree WT 

 

ippsWTInvFree ippsWTInvFree WT 

 

7-57


 

ippStsNoErr 


ippStsStateMatchErr 

 

pState NULL 

pState 

WTFwd 

 

IppStatus ippsWTFwd_32f(const Ipp32f* pSrc, Ipp32f* pDstLow, 

Ipp32f* pDstHigh, int dstLen, IppsWTFwdState_32f* pState); 

IppStatus ippsWTFwd_8s32f(const Ipp8s* pSrc, Ipp32f* pDstLow, 

Ipp32f* pDstHigh, int dstLen, IppsWTFwdState_8s32f* pState); 

IppStatus ippsWTFwd_8u32f(const Ipp8u* pSrc, Ipp32f* pDstLow, 

Ipp32f* pDstHigh, int dstLen, IppsWTFwdState_8u32f* pState); 

IppStatus ippsWTFwd_16s32f(const Ipp16s* pSrc, Ipp32f* pDstLow, 

Ipp32f* pDstHigh, int dstLen, IppsWTFwdState_16s32f* pState); 

IppStatus ippsWTFwd_16u32f(const Ipp16u* pSrc, Ipp32f* pDstLow, 

Ipp32f* pDstHigh, int dstLen, IppsWTFwdState_16u32f* pState); 

 

pSrc 

pDstLow 

pDstHigh 

dstLen 

pState 

 

“” 

“ ” 

 

pDstHigh pDstLow 

 

7-58


 

ippsWTFwd (2 * 

dstLen) pSrc “ ” pDstLow “ 

” pDstLow pState 

 

 

1 7-4 

“ ” “ ” 

 

pState 

 

 

 

 

 

1 

2 

 

 

 

6 2 

static const float decLow[6] = 

{ 

-6.25000000e-002f, 

6.25000000e-002f, 

5.00000000e-001f, 

5.00000000e-001f, 

6.25000000e-002f, 

-6.25000000e-002f 

}; 

static const float decHigh[2] = 

{ 

-5.00000000e-001f, 

5.00000000e-001f 

}; 

7-59


2 

 

offsLow = -1offsHigh = -1 + 2 = 1 

 

2 

 

1 

 

 

 

 

7-2 3 6 

2 

 

offsLow3 = -1offsHigh3 = 1 

2 

2 “ ” 

2*2 2 

offsLow2 = -1offsHigh2 = offsHigh3 + 4 = 5 

offsLow1 = -1offsHigh1 = offsHigh2 + 2*4 = 13 “ ” 

 

3 12 

 

ippStsNoErr 


 

NULL 


pState 

ippStsSizeErr 

dstLen srcLen 

7-60


7-4 

1 

( ) 

 

 

 

 

FIR 

2 

 

( ) 

 

 

 

 

FIR 

2 

 

( ) 

 

( ) 

 

 

WTFwdSetDlyLine 

WTFwdGetDlyLine 

 

 

IppStatus ippsWTFwdSetDlyLine_32f(IppsWTFwdState_32f* pState, 

const Ipp32f* pDlyLow, const Ipp32f* pDlyHigh); 

IppStatus ippsWTFwdSetDlyLine_8s32f(IppsWTFwdState_8s32f* pState, 


IppStatus ippsWTFwdSetDlyLine_8u32f(IppsWTFwdState_8u32f* pState, 


IppStatus ippsWTFwdSetDlyLine_16s32f(IppsWTFwdState_16s32f* pState, 


7-61


IppStatus ippsWTFwdSetDlyLine_16u32f(IppsWTFwdState_16u32f* pState, 


IppStatus ippsWTFwdGetDlyLine_32f(IppsWTFwdState_32f* pState, 

Ipp32f* pDlyLow, Ipp32f* pDlyHigh); 

IppStatus ippsWTFwdGetDlyLine_8s32f(IppsWTFwdState_8s32f* pState, 


IppStatus ippsWTFwdGetDlyLine_8u32f(IppsWTFwdState_8u32f* pState, 


IppStatus ippsWTFwdGetDlyLine_16s32f(IppsWTFwdState_16s32f* pState, 


IppStatus ippsWTFwdGetDlyLine_16u32f(IppsWTFwdState_16u32f* pState, 


 

pState 

pDlyLow 

pDlyHigh 

 

“ ” 

“ ” 

 

ippsWTFwdSetDlyLine ippsWTFwdSetDlyLine pDlyHigh pDlyLow 

pState 

ippsWTFwdSetDlyLine ippsWTFwdSetDlyLine WT 

 

ippsWTFwdGetDlyLine ippsWTFwdSetDlyLine WT 

 

 

 

 

 

7-62


 

 

 

 

 

 

 

“ ” 

dlyLowLen = lenLow + offsLow - 1 

lenLow offsLow “ ” 

 

“ ” 

dlyHighLen = lenHigh + offsHigh - 1, 

lenHigh offsHigh “ ” 

 

lenLowoffsLowlenHigh offsHigh 

7-54 WTFwdInitAllocWTInvInitAlloc 

 

 

ippStsNoErr 



 

pDlyLow pDlyHigh NULL 

pState 

7-63


WTInv 

 

IppStatus ippsWTInv_32f(const Ipp32f* pSrcLow, const Ipp32f* pSrcHigh, 

int srcLen, Ipp32f* pDst, IppsWTInvState_32f* pState); 

IppStatus ippsWTInv_32f8s(const Ipp32f* pSrcLow, const Ipp32f* pSrcHigh, 

int srcLen, Ipp8s* pDst, IppsWTInvState_32f8s* pState); 

IppStatus ippsWTInv_32f8u(const Ipp32f* pSrcLow, const Ipp32f* pSrcHigh, 

int srcLen, Ipp8u* pDst, IppsWTInvState_32f8u* pState); 

IppStatus ippsWTInv_32f16s(const Ipp32f* pSrcLow, const Ipp32f* pSrcHigh, 

int srcLen, Ipp16s* pDst, IppsWTInvState_32f16s* pState); 

IppStatus ippsWTInv_32f16u(const Ipp32f* pSrcLow, const Ipp32f* pSrcHigh, 

int srcLen, Ipp16u* pDst, IppsWTInvState_32f16u* pState); 

 

pSrcLow 

pSrcHigh 

srcLen 

pDst 

pState 

“” 

“ ” 

pSrcHigh pSrcLow 

 

 

 

ippsDCTInv “ ” 

pSrcLow “ ” pSrcHigh (2 * srcLen) 

pDst 

pState 

7-64


 

1 “ 

” “ ” 2 7-5 

2 

 

 

 

 

 

2 

 

2 6 

 

static const float recLow[2] = 

{ 

1.00000000e+000f, 

1.00000000e+000f 

}; 

static const float recHigh[6] = 

{ 

-1.25000000e-001f, 

-1.25000000e-001f, 

1.00000000e+000f, 

-1.00000000e+000f, 

1.25000000e-001f, 

1.25000000e-001f 

}; 

 

7-58 WTFwd 

2 

2 

offsLow 

= 2offsHigh = 0 

2 

 

7-65


1 4 

 

1 

 

 

 

7-3 3 

 

7-58 WTFwd 

6 2 

offsLow3 = 2offsHigh3 = 0 

2 “ ” 

2 offsLow2 = 2 

offsHigh2 = offsHigh3 + 2*2 = 4 offsLow1 = -1 

offsHigh1 = offsHigh2 + 2*4 = 12 

 

3 12 3 

24 

 

ippStsNoErr 



ippStsSizeErr 

 

NULL 

pState 

dstLen srcLen 

7-66


7-5 

1 

 

( ) 

 

 

( ) 

 

 

( ) 

 

↑ 2 

↑ 2 

 

FIR 

 

FIR 

 

 

 

 

 

( ) 

 

WTInvSetDlyLine 

WTInvGetDlyLine 

 

 

IppStatus ippsWTInvSetDlyLine_32f(IppsWTInvState_32f* pState, 


IppStatus ippsWTInvSetDlyLine_32f8s(IppsWTInvState_32f8s* pState, 


IppStatus ippsWTInvSetDlyLine_32f8u(IppsWTInvState_32f8u* pState, 


IppStatus ippsWTInvSetDlyLine_32f16s(IppsWTInvState_32f16s* pState, 


7-67


IppStatus ippsWTInvSetDlyLine_32f16u(IppsWTInvState_32f16u* pState, 


IppStatus ippsWTInvGetDlyLine_32f(IppsWTInvState_32f* pState, 


IppStatus ippsWTInvGetDlyLine_32f8s(IppsWTInvState_32f8s* pState, 


IppStatus ippsWTInvGetDlyLine_32f8u(IppsWTInvState_32f8u* pState, 


IppStatus ippsWTInvGetDlyLine_32f16s(IppsWTInvState_32f16s* pState, 


IppStatus ippsWTInvGetDlyLine_32f16u(IppsWTInvState_32f16u* pState, 


 

pState 

pDlyLow 

pDlyHigh 

 

“ ” 

“ ” 

 

ippsWTFwdSetDlyLine ippsWTInvSetDlyLine pDlyHigh pDlyLow 

pState 

ippsWTFwdSetDlyLine ippsWTFwdSetDlyLine WT 

 

ippsWTInvSetDlyLine ippsWTInvSetDlyLine WT 

 

7-68


 

 

 

 

 

 

 

 

 

 

 

 

 

“ ” C 

( 2 ) 

dlyLowLen = (lenLow + offsLow - 1) / 2 

lenLow offsLow “ ” 

 

“ ” C 

 

dlyHighLen = (lenHigh + offsHigh - 1) / 2 

lenHigh offsHigh “ ” 

 

lenLowoffsLowlenHigh offsHigh 

7-54 WTFwdInitAlloc 

WTInvInitAlloc 

 

ippStsNoErr 



 

pDlyLow pDlyHigh NULL 

pState 

7-69


7-70

8 

 

 

 

 


 

SVD(Single Value Decomposition) 

 

AddAllRowSum 

 

( ) 

IppStatus ippsAddAllRowSum_32f_D2(const Ipp32f* pSrc, int srcStep, 

int srcHeight, Ipp32f* pSrcDst, int width); 

IppStatus ippsAddAllRowSum_32f_D2L(const Ipp32f** mSrc, int srcHeight, 

Ipp32f* pSrcDst, int width); 

 

pSrc 

mSrc 

[srcHeight*srcStep] 

[srcHeight][width] 

8-1


srcStep pSrc 

srcHeight 

pSrcDst [width] 

width 

mSrc pSrcDst 

 

ippsAddAllRowSum pScrDst 

 

D2 : 

srcHeight – 1 

pSrcDst[ j] = pSrcDst[ j] + 

pSrc[ i ⋅ srcStep + j] 

, 0≤j


AddMulCol 

 

IppStatus ippsAddMulCol_64f_D2L(Ipp64f** mSrcDst, int srcWidth, int srcHeight, 

int col1, int col2, int row1, const Ipp64f val); 

 

mSrcDst 

srcWidth 

srcHeight 

col1 

col2 

row1 

val 

[srcHeight][srcWidth] 

 

mSrcDst 

mSrcDst 

1 

2 

1 

 

 

AddMulCol val 

SVD 

 

mSrcDst[ i] [ col2] = mSrcDst[ i] [ col2] + mSrcDst[ i] [ col1] ⋅ val , 

row1≤i


AddMulRow 

 

ippsAddMulRow_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 

const Ipp64f val); 

pSrc 

[len] 

pDst 

len 

val 

 

[len] 

 

 

AddMulRow val 1 

pDst SVD 

 

pDst[ i] = pDst[ i] + pSrc[ i] ⋅ val , 0≤i


DotProdCol 

2 

IppStatus ippsDotProdCol_64f_D2L(const Ipp64f** mSrc, int srcWidth, 

int srcHeight, Ipp64f* pSum, int col1, int col2, int row1); 

 

mSrc 

srcWidth 

srcHeight 

pSum 

col1 

col2 

row1 


mSrc 

mSrc 

 

1 

2 

1 

 

ippsDotProdCol 2 

SVD 

sum = 


∑ 

i = row1 

mSrc[ i] [ col1] ⋅ mSrc[ i] [ col2] 

 

ippStsNoErr 


ippStsSizeErr 

 

mSrc pSum NULL 

srcHeightsrcWidthcol1 row1 

; 

row1 srcHeight ; 

col1 srcWidth 

8-5


MulCol 

 

IppStatus ippsMulCol_64f_D2L(Ipp64f** mSrcDst, int srcWidth, int srcHeight, 

int col1, int row1, const Ipp64f val); 

 

mSrcDst 

srcWidth 

srcHeight 

col1 

row1 

val 


 

mSrcDst 

mSrcDst 

1 

1 

 

 

ippsMulCol mSrcDst val SVD 

 

mSrcDst[ i] [ col1] = mSrcDst[ i] [ col1] ⋅ val , row1≤i


QRTransCol 

QR 

IppStatus ippsQRTransCol_64f_D2L(Ipp64f** mSrcDst, int srcWidth, int 

srcHeight, int col1, int col2, const Ipp64f val1, const Ipp64f val2); 

 

mSrcDst 

srcWidth 

srcHeight 

col1 

col2 

val1 

val2 


 

mSrcDst 

mSrcDst 

1 

2 

1 

2 

 

ippsQRTransCol QR SVD 

 

mSrcDst[ i] col2 

mSrcDst[ i] col1 

0≤i


SumCol 

 

IppStatus ippsSumCol_32f_D2(const Ipp32f* pSrc, int srcStep, Ipp32f* pDst, 

int height, int width); 

IppStatus ippsSumCol_64f_D2(const Ipp64f* pSrc, int srcStep, Ipp64f* pDst, 

int height, int width); 

IppStatus ippsSumCol_32f_D2L(const float** mSrc, int width, Ipp32f* pDst, 

int height); 

IppStatus ippsSumCol_64f_D2L(const Ipp64f** mSrc, int width, Ipp64f* pDst, 

int height); 

ippsSumCol_32f_D2 ippsSumCol_32f_D2L 

 

IppStatus ippsSumAllRow_32f_D2(const Ipp32f* pSrc, int 

srcStep, int srcHeight, Ipp32f* pDst, int width); 

 

IppStatus ippsSumAllRow_32f_D2L(const Ipp32f** mSrc, int 

srcHeight, Ipp32f* pDst, int width); 

 

pSrc 

[height*srcStep] 

mSrc 

[height][width] 

srcStep pSrc 

pDst 

[width] 

height(srcHeight) 

width 

pDst 

8-8


 

ippsSumCol pDst 

pDst 

D2 : 

height – 1 

∑ 

pDst[ j] = pSrc[ i ⋅ srsStep + j] 

, 0≤j


SumColAbs 

 

IppStatus ippsSumColAbs_64f_D2L(const Ipp64f** mSrc, int srcWidth, 

int srcHeight, Ipp64f* pSum, int col1, int row1); 

 

mSrc 

srcWidth 

srcHeight 

pSum 

col1 

row1 


mSrc 

mSrc 

 

1 

1 

 

ippsSumColAbs SVD 

 

 

sum = 


∑ 

i = row1 

pSrc[ i] [ col1] 

 

ippStsNoErr 


ippStsSizeErr 

 

mSrc pSum NULL 

srcWidthsrcHeightcol1 row1 

; 



8-10


SumColSqr 

2 

 

IppStatus ippsSumColSqr_64f_D2L(Ipp64f** mSrcDst, int srcWidth, int srcHeight, 

Ipp64f* pSum, int col1, int row1, const Ipp64f val); 

 

mSrcDst 

srcWidth 

srcHeight 

pSum 

col1 

row1 

val 


 

mSrcDst 

mSrcDst 

 

1 

2 

 

 

ippsSumColSqr mSrcDst val 2 

pSum SVD 

 

mSrcDst[ i] [ col1] = mSrcDst[ i] [ col1] ⋅ val , 0≤i


ippStsSizeErr 

srcWidthsrcHeightcol1 row1 

; 



RecSqrt 

 

( ) 

IppStatus ippsRecSqrt_32f_Th(Ipp32f* pSrcDst, int len, Ipp32f val); 

 

pSrcDst 

len 

val 

pSrcDst 

 

 

 

ippsRecSqrt pSrcDst 

 

⎧val, pSrcDst[ j] < val 

⎪ 

pSrcDst[ j] 

= ⎨ 1 

, 0≤j


SubRow 

 

IppStatus ippsSubRow_32f_D2(const Ipp32f* pSrc, int width, Ipp32f* pSrcDst, 

int dstStep, int height); 

IppStatus ippsSubRow_32f_D2L(const Ipp32f* pSrc, Ipp32f** mSrcDst, int width, 

int height); 

 

pSrc 

mSrcDst 

pSrcDst 

width 

dstStep 

height 

[height*width] 


 

[height*dstStep] 

 

 

pSrcDst 

 

 

ippsSubRow pSrc 

 

D2 : 

pSrcDst[ i ⋅ dstStep + j] = pSrcDst[ i ⋅ dstStep + j] – pSrc[ j] 

, 

0≤i


 

ippStsNoErr 


ippStsSizeErr 

ippStsStrideErr 

 

pSrcpDst mDst NULL 

height width 

width > dstStep 

SumRow 

 

IppStatus ippsSumRow_32f_D2(const Ipp32f* pSrc, int height, int srcStep, 

Ipp32f* pDst, int srcWidth); 

IppStatus ippsSumRow_64f_D2(const Ipp64f* pSrc, int height, int srcStep, 

Ipp64f* pDst, int srcWidth); 

IppStatus ippsSumRow_32f_D2L(const Ipp32f** mSrc, int height, Ipp32f* pDst, 

int srcWidth); 

IppStatus ippsSumRow_64f_D2L(const Ipp64f** mSrc, int height, Ipp64f* pDst, 

int srcWidth); 

 

pSrc 


mSrc 

[height][srcWidth] 

height 

srcStep pSrc 

pDst 

[height] 

dstWidth 

8-14


 

ippsSumRow 

 

D2 : 

srcWidth – 1 

∑ 

pDst[ i] = 


, 0≤i


 

ippsSumRowAbs pSrc SVD 

 

sum = 

len – 1 

∑ 

i = 0 

pSrc[ i] 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrc pSum NULL 

len 

SumRowSqr 

2 

 

IppStatus ippsSumRowSqr_64f(Ipp64f* pSrcDst, int len, Ipp64f* pSum, 

const Ipp64f val); 

 

pSrcDst 

len 

pSum 

val 

[len] 

 

pSrcDst 

 

 

8-16


 

ippsSumRowSqr pSrcDst val 2 

SVD 

 

pSrcDst[ i] = pSrcDst[ i] ⋅ val , 0≤i


width 

weight 

pDst 

 

 

ippsAddNRows N 

SDCHMM width pDst 

 

pDst[ k] = pSrc[ k + offset + srcStep ⋅ ( pInd[ i] + pAddIndex[ i]) 

] + weight , 

0≤k srcStep 

8-18


ZeroMean 

 

 

IppStatus ippsZeroMean_16s(Ipp16s* pSrcDst, int len); 

 

pSrcDst 

len 

[len] 

 

 

 

ippsZeroMean pSrcDst pSrcDst 

 

 

pSrcDst[n] = max(IPP_MIN_16s, min(IPP_MAX_16s, pSrcDst[n] - 

len – 1 

1 

- ---------- pSrcDst[ i] 

)), 0 ≤ n < len 

len ∑ 

i = 0 

IPP_MIN_16s = -32768IPP_MAX_16s = 32767 

 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcDst NULL 

len 

8-19


LMThreshold 

 

 

IppStatus ippsLMThreshold_32f(const Ipp32f floor, const Ipp32f scale, 

const Ipp32f base, const Ipp32f* pSrc, Ipp32f* pDst, int len); 

 

floor 

scale 

base 

pSrc 

pDst 

len 

pSrc 

 

 

[len] 

[len] 

pSrc pDst 

 

ippsLMThreshold pSrc 

 

⎧scale ⋅ pSrc[ n] + base , pSrc[ n] > floor 

pDst[ n] 

= ⎨ 

, 0 ≤ n < len 

⎩scale ⋅ floor + base , pSrc[ n] < floor 

 

 

ippStsNoErr 


ippStsSizeErr 

 


len 

8-20


 

 

MeanCol 

 

IppStatus ippsMeanCol_32f_D2(const Ipp32f* pSrc, int srcHeight, int srcStep, 

Ipp32f* pDstMean, int width); 

IppStatus ippsMeanCol_32f_D2L(const Ipp32f** mSrc, int srcHeight, 

Ipp32f* pDstMean, int width); 

 

pSrc 

mSrc 

srcHeight 

srcStep 

pDstMean 

width 



 

pSrc 

[width] 

pDstMean 

 

ippsMeanCol 

 

D2 : 


∑ 

i = 1 


pDstMean[ j] 

= ----------------------------------------------------------------------------------------------------- , 0≤j


D2L : 


∑ 

i = 1 

mSrc[ i] [ j] 

pDstMean[ j] 

= -------------------------------------------------------------------- , 0≤j srcStep 

VarCol 

 

IppStatus ippsVarCol_32f_D2(const Ipp32f* pSrc, int srcHeight, int srcStep, 

Ipp32f* pSrcMean, Ipp32f* pDstVar, int width); 

IppStatus ippsVarCol_32f_D2L(const Ipp32f** mSrc, int srcHeight, 

Ipp32f* pSrcMean, Ipp32f* pDstVar, int width); 

 

pSrc 

mSrc 

srcHeight 

srcStep 

pSrcMean 

pDstVar 

width 



 

pSrc 

[width] 

[width] 

pSrcMean 

pDstVar 

8-22


 

ippsVarCol 

 

D2 : 

pDstVar[ j] 

0≤j


MeanVarCol 

 

 

IppStatus ippsMeanVarCol_32f_D2(const Ipp32f* pSrc, int srcHeight, 

int srcStep, Ipp32f* pDstMean, Ipp32f* pDstVar, int width); 

IppStatus ippsMeanVarCol_32f_D2L(const Ipp32f** mSrc, int srcHeight, 

Ipp32f* pDstMean, Ipp32f* pDstVar, int width); 

 

pSrc 

mSrc 

srcHeight 

srcStep 

pDstMean 

pDstVar 

width 



 

pSrc 

[width] 

[width] 

pDstMean 

pDstVar 

 

ippsMeanVarCol 8-22 ippsVarCol 8-21 

ippsMeanCol width pDstMean 

pDstVar 

 

ippStsNoErr 


ippStsSizeErr 


 

pSrcmSrcpDstMean pDstVar 

NULL 

srcHeight width 

width > srcStep 

8-24


MeanVarAcc 

 

IppStatus ippsMeanVarAcc_32f(Ipp32f const* pSrc, Ipp32f const* pSrcMean, 

Ipp32f* pDstMeanAcc, Ipp32f* pDstVarAcc, int len, Ipp32f val); 

IppStatus ippsMeanVarAcc_64f(Ipp64f const* pSrc, Ipp64f const* pSrcMean, 

Ipp64f* pDstMeanAcc, Ipp64f* pDstVarAcc, int len, Ipp64f val); 

 

pSrc 

pSrcMean 

pDstMeanAcc 

pDstVarAcc 

len 

val 

[len] 

[len] 

[len] 

[len] 

 

 

 

ippsMeanVarAcc 

 

pDstMeanAcc[ i] = pDstMeanAcc[ i] + val ⋅ ( pSrc[ i] – pSrcMean[ i] 

) , 

pDstVarAcc[ i] = pDstMeanAcc[ i] + val ⋅ ( pSrc[ i] – pSrcMean[ i] 

) 2 , 0≤i


NormalizeCol 

 

 

IppStatus ippsNormalizeCol_32f_D2(const Ipp32f* pSrcMean, 

const Ipp32f* pSrcVar, int width, Ipp32f* pSrcDst, int step, int height); 

 

pSrcMean 

pSrcVar 

width 

pSrcDst 

step 

height 

[width] 

[width] 

 

[height*step] 

pSrcDst 

 

 

ippsNormalizeCol 

 

pSrcDst[ i ⋅ step + j] = ( pSrcDst[ i ⋅ step + j] – pSrcMean[ j]) ⋅ pSrcVar[ j] 

, 

0≤i


 

 

 

 

 

LogAdd 

2 

( ) 

IppStatus ippsLogAdd_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len, 


IppStatus ippsLogAdd_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len, 


 

pSrc 

pDst 

len 

hint 

[len] 1 

[len] 2 

 

hint 

flag hint 

 

ippsLogAdd pSrc pDst 

pDst 

pDst[ i] = ln( e pSrc[ i] 

+ e pDst[ i] 

), 0≤i


 

ippStsNoErr 


ippStsSizeErr 

 


len 

LogSub 

2 

( ) 

IppStatus ippsLogSub_32f(const Ipp32f* pSrc, Ipp32f* pDst, int len); 

IppStatus ippsLogSub_64f(const Ipp64f* pSrc, Ipp64f* pDst, int len); 

 

pSrc 

pDst 

len 

[len] 

[len] 

 

 

ippsLogSub pSrc pDst 

pDst 

pDst[ i] = ln( e pSrc[ i] 

– e pDst[ i] 

), 0≤i pSrc[i] 

8-28


 

1 2 

 

 

a 0 ,..., a N - 1 N ⋅M 1 

(N M ) 

b 0 ,..., b N - 1 N ⋅K 1 (K 

)M K 

ippsCopyCol K ≥ MippsDelta K ≥ 2M 

ippsDeltaDelta K ≥ 3M ippsEvalDelta 

( ) 

ippsCopyCol 

 

ippsCopyCol ippsEvalDelta 

ippsDelta ippsDeltaDelta 1 2 

 

 

 

b i, j 

= a i, 

j 

(1.1) 

b i, j+ 

M 

= a′ i j 

b ij , + 2M 

= a″ i j 

w 

∑ 

, 

= ν 1 

⋅ k ⋅ ( b max( i – k, 

0) 

, j 

– b min( i + k, 

N – 1) 

, j 

) 

k = 1 

w 

⋅∑ 

= ν , 2 

k ⋅( b max( i – k, 

0) 

, j + M 

– b min( i + k, 

N– 

1 ), 

j + M 

) 

k = 1 

(1.2) 

(1.3) 

i = 0,..., N-1; j = 0,..., M-1 ν 1 ν 2 w 

 

8-29


 

ippsDelta ippsDeltaDelta 

L L 

 

b i, j 

= a i, 

j 

, i = 0,..., L-1; (2.1) 

b i, j+ 

M 

= a′ i j 

b i, j+ 

2M 

= a″ i j 

w 

∑ 

, 

= ν 1 

⋅ k ⋅ ( b max( i – k, 

0) 

, j 

– b i + k, 

j 

) 

k = 1 

w 

∑ 

, 

= ν 2 

⋅ k⋅ 

( b max( i – k, 

0) 

, j + M 

– b i+ k, 

j + M 

) 

k = 1 

, i = 0,..., L-w-1; (2.2) 

i = 0,..., L-2w-1; 

, (2.3) 

L > w⋅d d ippsDelta 

d = 1 ippsDeltaDelta d = 2 w⋅d 

 

L L+w⋅d 

w⋅d 

 

= 

, i = w⋅d ,..., L+ w⋅d -1; (3.1) 

b i, j 

a i – w⋅d 

, j 

b i, j+ 

M 

= a′ i– w ⋅d 

j 

w 

∑ 

, 

= ν 1 

⋅ k ⋅ ( b i– k, 

j 

– b i + k, 

j 

) 

k = 1 

, (3.2) 

i = w⋅(d-1) ,..., L+ w⋅(d-1) -1; 

b i, j+ 

2M 

= a″ i– w⋅d 

j 

w 

∑ 

, 

= ν 2 

⋅ k ⋅ ( b i– k, 

j+ 

M 

– b i+ k, 

j + M 

) 

k = 1 

, (3.3) 

i = 0,..., L -1. 

8-30


L L+w⋅d 

w⋅d 

 

= 

, i = w⋅d ,..., L+ w⋅d -1; (4.1) 

b i, j 

a i – w⋅d 

, j 

b i, j+ 

M 

= a′ i– w ⋅d 

j 

w 

∑ 

, 

= ν 1 

⋅ k ⋅ ( b i– k, 

j 

– b min( i + k, 

N – 1) 

, j 

) 

k = 1 

, (4.2) 

i = w⋅(d-1) ,..., L+ w⋅d -1; 

b i, j+ 

2M 

= a″ i– w⋅d 

j 

w 

, 

= ν 2 ∑⋅ k ⋅ ( b i – k, 

j + M 

– b min( i + k, 

N – 1) 

, j + M 

) 

k = 1 

, (4.3) 

i = 0,..., L+2w-1. 

CopyCol 

 

 

 

IppStatus ippsCopyCol_32f_D2(const Ipp32f* pSrc, int srcWidth, Ipp32f* pDst, 

int dstStep, int height); 

 

pSrc 

srcWidth 

pDst 

dstStep 

height 

[height*srcWidth] 

 

 


 

 

 

8-31


 

ippsCopyCol (1.1) pSrc pDst 

 

 

ippStsNoErr 


ippStsSizeErr 


 


height srcWidth 

srcWidth > dstStep 

EvalDelta 

 

( ) 

IppStatus ippsEvalDelta_32f_D2(Ipp32f* pSrcDst, int height, int step, 

int width, int offset, int winSize, const Ipp32f val); 

 

pSrcDst 

height 

step 

width 

offset 

winSize 

val 

[height*step] 

 

pSrcDst 

pSrcDst 

 

 

 

 

8-32


 

ippsEvalDelta pSrcDst 

offset (offset + width - 1) 

(offset + width) (offset + 2*width - 1) 

(1.1) 

 

ippStsNoErr 


 

pSrcDst NULL 

ippStsSizeErr heightwidth winSize ; 

offset ; 

width offset ; 

height 2*winSize 


offset+2*width > step 

Delta 

 

IppStatus ippsDeltaW1_32f_D2(const Ipp32f* pSrc, int srcWidth, Ipp32f* pDst, 

int dstStep, int height, const Ipp32f val, int deltamode); 

IppStatus ippsDeltaW2_32f_D2(const Ipp32f* pSrc, int srcWidth, Ipp32f* pDst, 

int dstStep, int height, const Ipp32f val, int deltamode); 

 

pSrc 

srcWidth 

pDst 

dstStep 

height 


 

 

 

 

 

8-33


val 

deltamode 

 

2 ippsDeltaBegin 

ippsDeltaEnd 

 

ippsDelta pSrc pDst 

1 

 

ippsDelta deltamode 

 

• ippsDeltaBegin ippsDeltaEnd 

(1.1) (1.2) 

• ippsDeltaBegin (2.1) (2.2) 

• ippsDeltaEnd (4.1) (4.2) 

• (3.1) (3.2) 

ippsDeltaBegin 


[(height+winSize)*dstStep] 

winSize = 

⎧ 

⎨ 

⎩ 

1 

2 

( ippsDeltaW1 ) 

( ippsDeltaW2 ) 

 

ippStsNoErr 


 


ippStsSizeErr srcWidth ; 

height ; 


height≤ winSize+1 ; 

ippsDeltaBegin height < 0 

 

ippStsStrideErr 2*srcWidth > dstStep 

8-34


DeltaDelta 

1 2 

 

IppStatus ippsDeltaDeltaW1_32f_D2(const Ipp32f* pSrc, int srcWidth, Ipp32f* 

pDst, int dstStep, int height, const Ipp32f val1, const Ipp32f val2, 

int deltamode); 

IppStatus ippsDeltaDeltaW1_32f_D2(const Ipp32f* pSrc, int srcWidth, Ipp32f* 

pDst, int dstStep, int height, const Ipp32f val1, const Ipp32f val2, 

int deltamode); 

 

pSrc 

srcWidth 

pDst 

dstStep 

height 

val1, val2 

deltamode 


 

 

 

 

 

1 2 

2 ippsDeltaBegin 

ippsDeltaEnd 

 

ippsDeltaDelta pSrc pDst 

1 2 

 

ippsDeltaDelta deltamode 

 

• ippsDeltaBegin ippsDeltaEnd 

(1.1)(1.2) (1.3) 

• ippsDeltaBegin (2.1)(2.2) (2.3) 

 

8-35


• ippsDeltaEnd (4.1)(4.2) (4.3) 

 

• (3.1)(3.2) (3.3) 

 



[(height+2*winSize)*dstStep] 

winSize = 

⎧ 

⎨ 

⎩ 

1 

2 

( ippsDeltaDeltaW1 ) 

( ippsDeltaDeltaW2 ) 

 

ippStsNoErr 


 


ippStsSizeErr srcWidth ; 

height ; 

ippsDeltaBegin height≤ 

2*(winSize+1) ; 

ippsDeltaBegin height < 0 

 

ippStsStrideErr 3*srcWidth > dstStep 

8-36


 

 

(PDF) IPP 

 

PDF 

LogGaussSingle 

 

IppStatus ippsLogGaussSingle_32f(const Ipp32f* pSrc, const Ipp32f* pSrcMean, 

const Ipp32f* pSrcVar, int srcLen, Ipp32f* pResult, Ipp32f val); 

IppStatus ippsLogGaussSingle_64f(const Ipp64f* pSrc, const Ipp64f* pSrcMean, 

const Ipp64f* pSrcVar, int srcLen, Ipp64f* pResult, Ipp64f val); 

IppStatus ippsLogGaussSingle_32f64f(const Ipp32f* pSrc, const Ipp32f* 

pSrcMean, const Ipp32f* pSrcVar, int srcLen, Ipp64f* pResult, Ipp64f val); 

 

pSrc 

pSrcMean 

pSrcVar 

srcLen 

pResult 

val 

[srcLen] 

[srcLen] 

[srcLen] 

 

 

 

 

ippsLogGaussSingle 1 1 

 

8-37


 

srcLen – 1 

∑ 

result = val – 0.5 ⋅ pSrcVar[ i] ⋅ ( pSrc[ i] – pSrcMean[ i] 

) 

i = 0 

2 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcpSrcMeanpSrcVar 

pResult NULL 

srcLen 

MahDistSingle 

 

IppStatus ippsMahDistSingle_32f(const Ipp32f* pSrc, const Ipp32f* pSrcMean, 

const Ipp32f* pSrcVar, int srcLen, Ipp32f* pResult); 

IppStatus ippsMahDistSingle_64f(const Ipp64f* pSrc, const Ipp64f* pSrcMean, 

const Ipp64f* pSrcVar, int srcLen, Ipp64f* pResult); 

 

pSrc 

pSrcMean 

pSrcVar 

srcLen 

pResult 

[srcLen] 

[srcLen] 

[srcLen] 

 

 

8-38


 

ippsMahDistSingle 1 1 

 

result = 

srcLen – 1 

∑ 

i = 0 

pSrcVar[ i] ⋅ ( pSrc[ i] – pSrcMean[ i] 

) 

2 

 

ippStsNoErr 


ippStsSizeErr 

 


pResult NULL 

srcLen 

 

PDF 

LogGauss1 

 

IppStatus ippsLogGauss1_32f_D2(const Ipp32f* pSrc, int srcStep, 

const Ipp32f* pSrcMean, const Ipp32f* pSrcVar, int width, Ipp32f* pDst, 

int height, Ipp32f val); 

IppStatus ippsLogGauss1_32f_D2L(const Ipp32f** mSrc, const Ipp32f* pSrcMean, 

const Ipp32f* pSrcVar, int width, Ipp32f* pDst, int height, Ipp32f val); 

IppStatus ippsLogGauss1_64f_D2(const Ipp64f* pSrc, int srcStep, 

const Ipp64f* pSrcMean, const Ipp64f* pSrcVar, int width, Ipp64f* pDst, 


IppStatus ippsLogGauss1_64f_D2L(const Ipp64f** mSrc, const Ipp64f* pSrcMean, 

const Ipp64f* pSrcVar, int width, Ipp64f* pDst, int height, Ipp64f val); 

8-39


 

pSrc 

mSrc 

srcStep 

pSrcMean 

pSrcVar 

width 

pDst 

height 

val 



pSrc 

[width] 

[width] 

 

[height] 

pDst 

 

 

ippsLogGauss1 1 

 

D2 : 

pDst[ i] = val – 0.5 ⋅ pSrcVar[ j] ⋅ ( p Src[ i ⋅ srcStep + j] – pSrcMean[ j]) 2 

0≤i


LogGauss2 

 

IppStatus ippsLogGauss2_32f_D2(const Ipp32f* pSrcMean, const Ipp32f* pSrcVar, 

int srcStep, const Ipp32f* pSrc, int width, Ipp32f* pSrcDst, int height); 

IppStatus ippsLogGauss2_32f_D2L(const Ipp32f** mSrcMean, const Ipp32f** 

mSrcVar, const Ipp32f* pSrc, int width, Ipp32f* pSrcDst, int height); 

IppStatus ippsLogGauss2_64f_D2(const Ipp64f* pSrcMean, const Ipp64f* pSrcVar, 

int srcStep, const Ipp64f* pSrc, int width, Ipp64f* pSrcDst, int height); 

IppStatus ippsLogGauss2_64f_D2L(const Ipp64f** mSrcMean, const Ipp64f** 

mSrcVar, const Ipp64f* pSrc, int width, Ipp64f* pSrcDst, int height); 

 

pSrcMean 

pSrcVar 

mSrcMean 

mSrcVar 

pSrcDst 

srcStep 

pSrc 

height 

width 





[height] 

pSrcMean pSrcVar 

[width] 

pSrcDst 

pSrc 

 

ippsLogGauss2 1 

pSrcDst 

 

8-41


D2 : 

pSrcDst[ i] = pSrcDst[ i] - 

width – 1 

∑ 

– 0.5 ⋅ pSrcVar[ i ⋅ srcStep + j] ⋅ ( pSrc[ j] – pSrcMean[ i ⋅ srcStep + j]) 2 

j = 0 

0≤i


 

pSrc 

mSrc 

srcStep 

pSrcMean 

pSrcVar 

width 

pDst 

height 



pSrc 

[width] 

[width] 

 

[height] 

pDst 

 

ippsMahDist1 1 

 

D2 : 

pDst[ i] = pSrcVar[ j] ⋅ ( pSrc[ i ⋅ srcStep + j] – pSrcMean[ j] 

) 

0≤i


MahDist2 

 

IppStatus ippsMahDist2_32f_D2(const Ipp32f* pSrcMean, const Ipp32f* pSrcVar, 

int srcStep, const Ipp32f* pSrc, int width, Ipp32f* pDst, int height); 

IppStatus ippsMahDist2_32f_D2L(const Ipp32f** mSrcMean, const Ipp32f** 

mSrcVar, const Ipp32f* pSrc, int width, Ipp32f* pDst, int height); 

IppStatus ippsMahDist2_64f_D2(const Ipp64f* pSrcMean, const Ipp64f* pSrcVar, 

int srcStep, const Ipp64f* pSrc, int width, Ipp64f* pDst, int height); 

IppStatus ippsMahDist2_64f_D2L(const Ipp64f** mSrcMean, const Ipp64f** 

mSrcVar, const Ipp64f* pSrc, int width, Ipp64f* pDst, int height); 

 

pSrcMean 

pSrcVar 

mSrcMean 

mSrcVar 

srcStep 

pSrc 

width 

pDst 

height 





pSrc 

[width] 

pSrc 

[height] 

pDst 

 

1 

 

8-44


D2 : 

pDst[ i] = pSrcVar[ i ⋅ srcStep + j] ⋅ ( pSrc[ j] – pSrcMean[ i ⋅ srcStep + j] 

) 

0≤i


pSrcDst 

height 

val 

[height] 

pSrcDst 

 

 

ippsLogGaussMax1 1 

pSrcDst 

pSrcDst 

 

pSrcDst[ i] = max( pSrcDst[ i] , Vi [ ]) 

 

Vi [ ] = val – 0.5 ⋅ pSrcVar[ i] ⋅ ( pSrc[ i ⋅ srcStep + j] – pSrcMean[ j] 

) 2 

0≤i≤height 

width – 1 

∑ 

j = 0 

ippsLogGauss1 ippsMaxEvery_32f 

 

 

ippStsNoErr 


ippStsSizeErr 


 


pSrcDst NULL 

height width 


8-46


LogGaussAdd1 

 

2 

( ) 

IppStatus ippsLogGaussAdd1_32f_D2(const Ipp32f* pSrc, int srcStep, 

const Ipp32f* pSrcMean, const Ipp32f* pSrcVar, int width, Ipp32f* pSrcDst, 


 

pSrc 

srcStep 

pSrcMean 

pSrcVar 

width 

pSrcDst 

height 

val 


 

pSrc 

[width] 

[width] 

 

[height] 

pSrcDst 

 

 

ippsLogGaussAdd1 1 

pSrcDst 

pSrcDst 

 

pSrcDst[ i] = ln( e Vi [ ] + e pSrcDst[ i] 

), 

 

Vi [ ] = val – 0.5 ⋅ pSrcVar[ i] ⋅ ( pSrc[ i ⋅ srcStep + j] – pSrcMean[ j] 

) 2 

0≤i≤height 

width – 1 

∑ 

j = 0 

8-47


ippsLogGauss1 ippsLogAdd_32f 

 

 

ippStsNoErr 


ippStsSizeErr 


 


pSrcDst NULL 

height width 


 

 

 

GaussianDist 

 

IppStatus ippsGaussianDist_32f(const Ipp32f* pSrcMean1, const Ipp32f* 

pSrcVar1, const Ipp32f* pSrcMean2, const Ipp32f* pSrcVar2, int srcLen, 

Ipp32f* pResult, Ipp32f wgt1, Ipp32f det1, Ipp32f wgt2, Ipp32f det2); 

 

pSrcMean1 

pSrcVar1 

pSrcMean2 

pSrcVar2 

srcLen 

pResult 

wgt1 

det1 

1 [srcLen] 

1 [srcLen] 

2 [srcLen] 

2 [srcLen] 

 

 

1 

1 

8-48


wgt2 

det2 

2 

2 

 

ippsGaussianDist 

 

result = ( wgt1 ⋅ det1) + ( wgt2 ⋅ det2) –( wgt1 + wgt2) ⋅ ( srcLen ⋅ ln( 2π) – lnV) 

, 

V 

= 

srcLen – 1 

∏ 

i = 0 

W1[ i] + W2[ i] –( wgt1 ⋅ pSrcMean1[ i] 

+ wgt2 ⋅ pSrcMean2[ i] 

) 2 

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ 

( wgt1 + wgt2) 

, 

W1[ i] = wgt1 ⋅ ( pSrcVar1[ i] + ( pSrcMean1[ i] 

) 2 ) 


) 2 ) 

, 

 

ippStsNoErr 


ippStsSizeErr 

 

pSrcMean1pSrcMean2pSrcVar1 

pSrcVar2 pResult NULL 

srcLen 

8-49


GaussianMerge 

2 

IppStatus ippsGaussianMerge_32f(Ipp32f* pSrcMean1, Ipp32f* pSrcVar1, Ipp32f* 

pSrcMean2, Ipp32f* pSrcVar2, Ipp32f* pDstMean, Ipp32f* pDstVar, int len, 

Ipp32f* pDstDet, Ipp32f wgt1, Ipp32f wgt2); 

 

pSrcMean1 

pSrcVar1 

pSrcMean2 

pSrcVar2 

pDstMean 

pDstVar 

len 

pDstDet 

wgt1 

wgt2 

1 [len] 

1 [len] 

2 [len] 

2 [len] 

[len] 

[len] 

 

 

1 

2 

 

ippsGaussianMerge 2 (PDF) 

 

⎛ 

dstDet = len ⋅ ln( 2π) 

– ln⎜ 

⎜ 

⎝ 

len – 1 

∏ 

i = 0 

⎞ 

pDstVar[ i] 

⎟ 

⎟ 

⎠ 

, 

pDstMean[ i] 

= 

wgt1 ⋅ pSrcMean1[ i] 

+ wgt2 ⋅ pSrcMean2[ i] 

----------------------------------------------------------------------------------------------------------------------------------- 

( wgt1 + wgt2) 

, 0≤i



) 2 ) 


) 2 ) 

, 

 

ippStsNoErr 


ippStsSizeErr 

 


pSrcVar2pDstMeanpDstVar pDstDet 

NULL 

len 

GaussianSplit 

 

IppStatus ippsGaussianSplit_32f(Ipp32f* pSrcMean1, Ipp32f* pSrcVar1, 

Ipp32f* pSrcMean2, Ipp32f* pSrcVar2, int len, Ipp32f val); 

 

pSrcMean1 

pSrcVar1 

pSrcMean2 

pSrcVar2 

len 

val 

[len] 1 

[len] 1 

[len] 2 

[len] 2 

 

 

 

ippsGaussianSplit PDF 2 PDF 

 

pSrcMean1[ i] = pSrcMean1[ i] + val ⋅ pSrcVar1[ i] 

, 

pSrcMean2[ i] = pSrcMean1[ i] – val ⋅ pSrcVar1[ i] 

, 0≤i


 

ippStsNoErr 


ippStsSizeErr 

 


pSrcVar2 NULL 

len 

8-52

[Mit93] 27 

 

[Opp75][Opp89][Jac89] [Zie83] 

[Opp89] [Opp75] [Jac89] 

[Zie83] 

[Bri94] 

C. Brislawn Classification of Nonexpansive Symmetric Extension 

Transforms for Multirate Filter Banks(Los Alamos Report 

LA-UR-94-17471994 ) 

[Cap78] V. CappelliniA. G. ConstantinidesP. Emilani Digital Filters and 

Their Applications( Academic Press 1978 ) 

[CCITT] CCITT G.711 

[Cro83] R. E. CrochiereL. R. Rabiner Multirate Digital Signal Processing 

( 

1983 ) 

[Dau92] I. Daubechies Ten Lectures on Wavelets( 

1992 ) 

[Fei92] 

E. FeigS. Winograd Fast algorithms for DCT(IEEE Transactions on 

Signal Processingvol.40No.91992 ) 

[Har78] F. Harris On the Use of Windows(IEEE vol. 66No.1IEEE 

1978 ) 

[Hay91] S. Haykin Adaptive Filter Theory( 

1991 ) 

[Jac89] 

Leland B. Jackson Digital Filters and Signal Processing 

(Kluwer Academic Publishers 2 1989 ) 

-1


 

[Lyn89] Paul A. Lynn Introductory Digital Signal Processing with Computer 

Applications( John Wiley&Sons, Inc. 1993 ) 

[Mit93] 

Sanjit K. MitraJames F. Kaiser Handbook for Digital Signal 

Processing( John Wiley&Sons, Inc. 1993 ) 

[Mit98] S. K. Mitra Digital Signal Processing( 1998 ) 

[NIC91] Nam Ik ChoSang Uk Lee Fast algorithm and implementation of 2D 

DCT(IEEE Transactions on Circuits and Systemsvol. 31No.3 

1991 ) 

[Opp75] Alan V. OppenheimRonald W. Schafer Digital Signal Processing 

( 

1975 ) 

[Opp89] Alan V. OppenheimRonald W. Schafer Discrete-Time Signal 

Processing( 

1989 ) 

[Rab78] L.R. RabinerR.W. Schafer Digital Processing of Speech Signals 

( 

1978 ) 

[Rao90] K.R. RaoP. Yip Discrete Cosine Transform.Algorithms, Advantages 

and Applications( Academic Press 1990 ) 

[Str96] 

G. StrangT. Nguyen Wavelet and Filter Banks(Wellesley-Cambridge 

Press 1996 ) 

[Vai93] 

P. P. Vaidyanathan Multirate Systems and Filter Banks( 

) 

[Wid85] B. WidrowS.D. Stearns Adaptive Signal Processing( 

1985 ) 

[Zie83] 

Rodger E. ZiemerWilliam H. TranterD. Ronald Fannin Signals and 

Systems:Continuous and Discrete( Macmillan Publishing 

Co. 1983 ) 

-2

(adaptive filter) 

( ) 

 

 

 

 

(arithmetic operation) 

2 

 

BQ 

IIR 

 

CCS 

 

(companding functions) 

 

 

 

 

(complex conjugate-symmetric) 

 

x(-n) = x(n)* 

(* ) 

(conjugate) 

a + bj a - bj 

(conjugate-symmetric) 

complex conjugate-symmetric 

DCT 

Discrete Cosine Transform 

-1


 

(decimation) 

 

 

 

 

(down-sampling) 

 

 

 

(element-wise) 

 

{x 0 , x 1 , x 2 } 

{y 0 , y 1 , y 2 } 

{x 0 , x 1 , x 2 } + {y 0 , y 1 , y 2 } = {x 0 + y 0 , x 1 + y 1 , x 2 + y 2 } 

 

FIR 

Finite Impulse Response FIR 

 

FIR LMS 

Least Mean Squares Finite Impulse Response 

(fixed-point data format) 

1 

 

 

S.31 1 31 

S15.16 

1 15 16 

 

IIR 

Infinite Impulse Response 

(in-place) 

(interpolation) 

 

 

 

 

 

 

 

-2


 

LMS 

Least Mean Square 2 

LMS 

FIR 

 

LTI 

Linear Time-Invariant LTI 

 

( ) 

[Lyn89] 

MMX ® (MMX technology) 

 

® 

MMX 4 

8 64 MMX 

SIMD (single instruction, multiple data) 57 

 

MR 

1 

 

(multi-rate) 

 

 

(not-in-place) 

 

 

(polyphase) 

CCS 

Pack 

Perm 

 

 

Pack 

Perm 

 

FFT 

( 2 FFT 

) 

FFT 

RCPerm FFT 

 

 

-3


 

(saturation) 

 

 

7FFFh 7FFFh 

 

8000h 

8000h 

(sinusoid) 

SSE 

Streaming SIMD Extensions 

 

SSE 

 

 

 

 

(tone) 

 

 

(up-sampling) 

 

 

 

(Window) 

 

FFT 

-4

A 

Abs, 5-28 

Add, 5-13 

AddAllRowSum, 8-1 

AddC, 5-12 

AddMulCol, 8-3 

AddMulRow, 8-4 

AddNRows, 8-17 

ALawToLin, 5-77 

ALawToMuLaw, 5-80 

And, 5-2 

AndC, 5-1 

AutoCorr, 6-4 

C 

CartToPolarcomplex, 5-68 

CartToPolarreal, 5-69 

CCS , 7-3 

Conj, 5-42 

ConjCCS, 7-8 

ConjFlip, 5-43 

ConjPack, 7-6 

ConjPerm, 7-4 

Conv, 6-1 

ConvCyclic, 6-3 

Convert, 5-40 

Copy, 4-1 

CopyCol, 8-31 

CplxToReal, 5-53 

CrossCorr, 6-6 

D 

DCTFwd, 7-43 

DCTFwdFree, 7-40 

DCTFwdGetBufSize, 7-41 

DCTInv, 7-43 

DCTInvFree, 7-40 

DCTInvGetBufSize, 7-41 

DCTInvInitAlloc, 7-39 

DCT , 7-39 7-44 

DCTFwd, 7-43 

DCTFwdFree, 7-40 

DCTFwdGetBufSize, 7-41 

DCTFwdInitAlloc, 7-39 

DCTInv, 7-43 

DCTInvFree, 7-40 

DCTInvGetBufSize, 7-41 

DCTInvInitAlloc, 7-39 

Delta, 8-33 

DeltaDelta, 8-35 

DFTFree_C, 7-26 

DFTFree_R, 7-26 

DFTFwd_CToC, 7-28 

DFTFwd_RToCCS, 7-31 

DFTFwd_RToPack, 7-31 

DFTFwd_RToPerm, 7-31 

DFTGetBufSize_C, 7-27 

DFTGetBufSize_R, 7-27 

DFTInitAlloc_C, 7-24 

-1


 

DFTInitAlloc_R, 7-24 

DFTInv_CCSToR, 7-31 

DFTInv_CToC, 7-28 

DFTInv_PackToR, 7-31 

DFTInv_PermToR, 7-31 

DFT , 7-24 7-33 

DFTFree_C, 7-26 

DFTFree_R, 7-26 

DFTFwd_CToC, 7-28 

DFTFwd_RToCCS, 7-31 

DFTFwd_RToPack, 7-31 

DFTFwd_RToPerm, 7-31 

DFTGetBufSize_C, 7-27 

DFTGetBufSize_R, 7-27 

DFTInitAlloc_C, 7-24 

DFTInitAlloc_R, 7-24 

DFTInv_CCSToR, 7-31 

DFTInv_CToC, 7-28 

DFTInv_PackToR, 7-31 

DFTInv_PermToR, 7-31 

Div, 5-23, 5-24 

DotProd, 5-108 

DotProdCol, 8-5 

DTCFwdInitAlloc, 7-39 

E 

EvalDelta, 8-32 

Exp, 5-34 

F 

FFTFree_C, 7-15 

FFTFree_R, 7-15 

FFTFwd_CToC, 7-17 

FFTFwd_RToCCS, 7-20 

FFTFwd_RToPack, 7-20 

FFTFwd_RToPerm, 7-20 

FFTGetBufSize_C, 7-16 

FFTGetBufSize_R, 7-16 

FFTInitAlloc_C, 7-13 

FFTInitAlloc_R, 7-13 

FFTInv_CCSToR, 7-20 

FFTInv_CToC, 7-17 

FFTInv_PackToR, 7-20 

FFTInv_PermToR, 7-20 

FFT , 7-13 7-22 

FFTFree_C, 7-15 

FFTFree_R, 7-15 

FFTFwd_CToC, 7-17 

FFTFwd_RToCCS, 7-20 

FFTFwd_RToPack, 7-20 

FFTFwd_RToPerm, 7-20 

FFTGetBufSize_C, 7-16 

FFTGetBufSize_R, 7-16 

FFTInitAlloc_C, 7-13 

FFTInitAlloc_R, 7-13 

FFTInv_CCSToR, 7-20 

FFTInv_CToC, 7-17 

FFTInv_PackToR, 7-20 

FFTInv_PermToR, 7-20 

FilterMedian, 6-51 

FIR, 6-17 

FIR LMS , 6-26 6-35 

FIRLMS, 6-31 

FIRLMSFree, 6-28 

FIRLMSGetDlyLine, 6-34 

FIRLMSGetTaps, 6-33 

FIRLMSInitAlloc, 6-27 

FIRLMSOne, 6-28 

FIRLMSSetDlyLine, 6-34 

FIRFree, 6-26 

FIRGetDlyLine, 6-23 

FIRGetTaps, 6-22 

FIRInitAlloc, 6-10 

FIRLMS, 6-31 

FIRLMSFree, 6-28 

FIRLMSGetDlyLine, 6-34 

FIRLMSGetTaps, 6-33 

FIRLMSInitAlloc, 6-27 

FIRLMSOne, 6-28 

FIRLMSSetDlyLine, 6-34 

-2


 

FIRMRInitAlloc, 6-10 

FIROne, 6-15 

FIRSetDlyLine, 6-23 

FIR , 6-9 6-26 

FIR, 6-17 

FIRFree, 6-26 

FIRGetDlyLine, 6-23 

FIRGetTaps, 6-22 

FIRInitAlloc, 6-10 

FIRMRInitAlloc, 6-10 

FIROne, 6-15 

FIRSetDlyLine, 6-23 

flag hint , 7-1 7-2 

Free, 3-5 

G 

GaussianDist, 8-48 

GaussianMerge, 8-50 

GaussianSplit, 8-51 

GetLibVersion, 3-1 

GetStatusString, 3-3 

Goertz, 7-35 

GoertzTwo, 7-38 

I 

IIR, 6-42 

IIRFree, 6-39 

IIRGetDlyLine, 6-48 

IIRInitAlloc, 6-35 

IIRInitAlloc_BiQuad, 6-35 

IIROne, 6-40 

IIRSetDlyLine, 6-48 

IIR , 6-35 6-50 

IIR, 6-42 

IIRFree, 6-39 

IIRGetDlyLine, 6-48 

IIRInitAlloc, 6-35 

IIRInitAlloc_BiQuad, 6-35 

IIROne, 6-40 

IIRSetDlyLine, 6-48 

Imag, 5-51 

Intel Performance Primitives , 1-1 

ippsAbs()Abs 

ippsAdd()Add 

ippsAddAllRowSum()AddAllRowSum 

ippsAddC()AddC 

ippsAddMulCol()AddMulCol 

ippsAddMulRow()AddMulRow 

ippsAddNRows()AddNRows 

ippsALawToLin()ALawToLin 

ippsALawToMuLaw()ALawToMuLaw 

ippsAnd()And 

ippsAndC()AndC 

ippsAutoCorr()AutoCorr 

ippsCartToPolar()CartToPolarcomplex 

ippsCartToPolar()CartToPolarreal 

ippsConj()Conj 

ippsConjCCS()ConjCCS 

ippsConjFlip()ConjFlip 

ippsConjPack()ConjPack 

ippsConv()Conv 

ippsConvCyclic()ConvCyclic 

ippsConvert()Convert 

ippsCopy()Copy 

ippsCopyCol()CopyCol 

ippsCplxToReal()CplxToReal 

ippsCrossCorr()CrossCorr 

ippsDCTFwd()DCTFwd 

ippsDCTFwdFree()DCTFwdFree 

ippsDCTFwdGetBufSize()DCTFwdGetBufSize 

 

ippsDCTFwdInitAlloc()DCTFwdInitAlloc 

ippsDCTInv()DCTInv 

ippsDCTInvFree()DCTInvFree 

ippsDCTInvGetBufSize()DCTInvGetBufSize 

 

ippsDCTInvInitAlloc()DCTInvInitAlloc 

-3


 

ippsDelta()Delta 

ippsDeltaDelta()DeltaDelta 

ippsDFTFree_C()DFTFree_C 

ippsDFTFree_R()DFTFree_R 

ippsDFTFwd_CToC()DFTFwd_CToC 

ippsDFTFwd_RToCCS()DFTFwd_RToCCS 

ippsDFTFwd_RToPack()DFTFwd_RToPack 

ippsDFTFwd_RToPerm()DFTFwd_RToPerm 

 

ippsDFTGetBufSize_R()DFTGetBufSize_R 

ippsDFTInitAlloc_C()DFTInitAlloc_C 

ippsDFTInitAlloc_R()DFTInitAlloc_R 

ippsDFTInv_CCSToR()DFTInv_CCSToR 

ippsDFTInv_CToC()DFTInv_CToC 

ippsDFTInv_PackToR()DFTInv_PackToR 

ippsDFTInv_PermToR()DFTInv_PermToR 

ippsDiv()Div 

ippsDivC()DivC 

ippsDotProd ()DotProd 

ippsDotProdCol()DotProdCol 

ippsEvalDelta()EvalDelta 

ippsExp()Exp 

ippsFFTFree_C()FFTFree_C 

ippsFFTFree_R()FFTFree_R 

ippsFFTFwd_CToC()FFTFwd_CToC 

ippsFFTFwd_RToCCS()FFTFwd_RToCCS 

ippsFFTFwd_RToPack()FFTFwd_RToPack 

ippsFFTFwd_RToPerm()FFTFwd_RToPerm 

ippsFFTGetBufSize_C()FFTGetBufSize_C 

ippsFFTGetBufSize_R()FFTGetBufSize_R 

ippsFFTInitAlloc_C()FFTInitAlloc_C 

ippsFFTInitAlloc_R()FFTInitAlloc_R 

ippsFFTInv_CCSToR()FFTInv_CCSToR 

ippsFFTInv_CToC()FFTInv_CToC 

ippsFFTInv_PackToR()FFTInv_PackToR 

ippsFFTInv_PermToR()FFTInv_PermToR 

ippsFilterMedianFilterMedian 

ippsFIR()FIR 

ippsFIRFree()FIRFree 

ippsFIRGetDlyLine()FIRGetDlyLine 

ippsFIRGetTaps()FIRGetTaps 

ippsFIRInitAlloc()FIRInitAlloc 

ippsFIRLMS()FIRLMS 

ippsFIRLMSFree()FIRLMSFree 

ippsFIRLMSGetDlyLine()FIRLMSGetDlyLine 

 

ippsFIRLMSGetTaps()FIRLMSGetTaps 

ippsFIRLMSInitAlloc()FIRLMSInitAlloc 

ippsFIRLMSOne()FIRLMSOne 

ippsFIRLMSSetDlyLine()FIRLMSSetDlyLine 

 

ippsFIRMRInitAlloc()FIRMRInitAlloc 

ippsFIROne()FIROne 

ippsFIRSetDlyLine()FIRSetDlyLine 

ippsFree()Free 

ippsGaussianDist()GaussianDist 

ippsGaussianMerge()GaussianMerge 

ippsGaussianSplit()GaussianSplit 

ippsGetBufSize_C()DFTGetBufSize_C 

ippsGetLibVersion()GetLibVersion 

ippsGetStatusString()GetStatusString 

ippsGoertz()Goertz 

ippsGoertzTwo()GoertzTwo 

ippsIIR()IIR 

ippsIIRFree()IIRFree 

ippsIIRGetDlyLine()IIRGetDlyLine 

ippsIIRInitAlloc()IIRInitAlloc 

ippsIIRInitAlloc_BiQuad()IIRInitAlloc_BiQuad 

 

ippsIIROne()IIROne 

ippsIIRSetDlyLine()IIRSetDlyLine 

ippsImag()Imag 

ippsLinToALaw()LinToALaw 

ippsLinToMuLaw()LinToMuLaw 

ippsLMThreshold()LMThreshold 

-4


 

ippsLn()Ln 

ippsLogAdd()LogAdd 

ippsLogGauss1()LogGauss1 

ippsLogGauss2()LogGauss2 

ippsLogGaussAdd1()LogGaussAdd1 

ippsLogGaussMax1()LogGaussMax1 

ippsLogGaussSingle()LogGaussSingle 

ippsLogSub()LogSub 

ippsLShiftC()LShiftC 

ippsMagnitude()Magnitude 

ippsMahDist1()MahDist1 

ippsMahDist2()MahDist2 

ippsMahDistSingle()MahDistSingle 

ippsMalloc()Malloc 

ippsMax()Max 

ippsMaxIndx()MaxIndx 

ippsMaxOrder()MaxOrder 

ippsMean()Mean 

ippsMeanCol()MeanCol 

ippsMeanVarAcc()MeanVarAcc 

ippsMeanVarCol()MeanVarCol 

ippsMin()Min 

ippsMinIndxt()MinIndx 

ippsMul()Mul 

ippsMuLawToALaw()MuLawToALaw 

ippsMuLawToLin()MuLawToLin 

ippsMulC()MulC 

ippsMulCol()MulCol 

ippsMulPack()MulPack 

ippsMulPerm()MulPerm 

ippsNorm()Norm 

ippsNormalize()Normalize 

ippsNormalizeCol()NormalizeCol 

ippsNormDiff()NormDiff 

ippsNot()Not 

ippsOr()Or 

ippsOrC()OrC 

ippsPhase()Phasecomplex 

ippsPhase()Phasereal 

ippsPolarToCart()PolarToCartcomplex 

ippsPolarToCart()PolarToCartreal 

ippsPowerSpectr()PowerSpectrcomplex 

ippsPowerSpectr()PowerSpectrreal 

ippsPreemphasize()Preemphasize 

ippsQRTransCol()QRTransCol 

ippsRandGauss()RandGauss 

ippsRandUniform()RandUniform 

ippsReal()Real 

ippsRealToCplx()RealToCplx 

ippsRecSqrt()RecSqrt 

ippsRShiftC()RShiftC 

ippsSampleDown()SampleDown 

ippsSampleUp()SampleUp 

ippsSet()Set 

ippsSqr()Sqr 

ippsSqrt()Sqrt 

ippsStdDev()StdDev 

ippsSub()Sub 

ippsSubC()SubC 

ippsSubCRev()SubCRev 

ippsSubRow()SubRow 

ippsSum()Sum 

ippsSumCol()SumCol 

ippsSumColAbs()SumColAbs 

ippsSumColSqr()SumColSqr 

ippsSumRow()SumRow 

ippsSumRowAbs()SumRowAbs 

ippsSumRowSqr()SumRowSqr 

ippsThreshold()Threshold 

ippsThreshold_GT()Threshold_GT 

ippsThreshold_GTVal()Threshold_GTVal 

ippsThreshold_LT()Threshold_LT 

ippsThreshold_LTInv()Threshold_LTInv 

ippsThreshold_LTVal()Threshold_LTVal 

-5


 

ippsTone()Tone 

ippsTriangle()Triangle 

ippsVarCol()VarCol 

ippsVectorJaehne()VectorJaehne 

ippsWinBartlett()WinBartlett 

ippsWinBlackman()WinBlackman 

ippsWinHamming()WinHamming 

ippsWinHann()WinHann 

ippsWinKaiser()WinKaiser 

ippsWTFwd()WTFwd 

ippsWTFwdFree()WTFwdFree 

ippsWTFwdGetDlyLine()WTFwdGetDlyLine 

 

ippsWTFwdInitAlloc()WTFwdInitAlloc 

ippsWTFwdSetDlyLine()WTFwdSetDlyLine 

 

ippsWTHaarFwd()WTHaarFwd 

ippsWTHaarInv()WTHaarInv 

ippsWTInv()WTInv 

ippsWTInvFree()WTInvFree 

ippsWTInvGetDlyLine()WTInvGetDlyLine 

ippsWTInvInitAlloc()WTInvInitAlloc 

ippsWTInvSetDlyLine()WTInvSetDlyLine 

ippsXor()Xor 

ippsXorC()XorC 

ippsZero()Zero 

ippsZeroMean()ZeroMean 

IPP , 1-1 

L 

LinToALaw, 5-78 

LinToMuLaw, 5-76 

LMThreshold, 8-20 

Ln, 5-36 

LogAdd, 8-27 

LogGauss1, 8-39 


LogGaussAdd1, 8-47 

LogGaussMax1, 8-45 

LogGaussSingle, 8-37 

LogSub, 8-28 

LShiftC, 5-8 

M 

Magnitude, 5-44 

MahDist1, 8-42 


MahDistSingle, 8-38 

Malloc, 3-4 

Max, 5-97 

MaxIndx, 5-97 

MaxOrder, 5-72 

Mean, 5-101 

MeanCol, 8-21 

MeanVarAcc, 8-25 

MeanVarCol, 8-24 

Min, 5-98 

MinIndx, 5-99 

MMX , 1-1 

Mul, 5-16 

MuLawToALaw, 5-79 

MuLawToLin, 5-75 

MulC, 5-15 

MulCol, 8-6 

MulPack, 7-10 

MulPerm, 7-10 

N 

Norm, 5-104 

Normalize, 5-38 

NormalizeCol, 8-26 

NormDiff, 5-106 

Not, 5-7 

-6


 

O 

Or, 5-4 

OrC, 5-3 

P 

Pack , 7-2 

Perm , 7-3 

Phasecomplex, 5-46 

Phasereal, 5-47 

PolarToCartcomplex, 5-70 

PolarToCartreal, 5-71 

PowerSpectrcomplex, 5-48 

PowerSpectrreal, 5-49 

Preemphasize, 5-73 

Q 

QRTransCol, 8-7 

R 

RandGauss, 4-12 

RandUniform, 4-11 

Real, 5-51 

RealToCplx, 5-52 

RecSqrt, 8-12 

RShiftC, 5-9 

S 

SampleDown, 5-113 

SampleUp, 5-111 

Set, 4-2 

SIMD , 1-1 

Sqr, 5-29 

Sqrt, 5-31 

StdDev, 5-102 

Sub, 5-21 

SubC, 5-18 

SubCRev, 5-20 

SubRow, 8-13 

Sum, 5-95 

SumCol, 8-8 

SumColAbs, 8-10 

SumColSqr, 8-11 

SumRow, 8-14 

SumRowAbs, 8-15 

SumRowSqr, 8-16 

T 

Threshold, 5-54 

Threshold_GT, 5-58 

Threshold_GTVal, 5-62 

Threshold_LT, 5-58 

Threshold_LTInv, 5-65 

Threshold_LTVal, 5-62 

Tone, 4-5 

Triangle, 4-9 

V 

VarCol, 8-22 

VectorJaehne, 4-13 

W 

WinBartlett, 5-83 

WinBlackman, 5-85 

WinHamming, 5-88 

WinHann, 5-90 

WinKaiser, 5-92 

WTFwd, 7-58 

WTFwdFree, 7-57 

WTFwdGetDlyLine, 7-61 

WTFwdInitAlloc, 7-54 

WTFwdSetDlyLine, 7-61 

WTHaarFwd, 7-49 

WTHaarInv, 7-49 

-7


 

WTInv, 7-64 

WTInvFree, 7-57 

WTInvGetDlyLine, 7-67 

WTInvInitAlloc, 7-54 

WTInvSetDlyLine, 7-67 

WTInvInitAlloc, 7-54 

WTInvSetDlyLine, 7-67 

 

, 2-9 

X 

Xor, 5-6 

XorC, 5-5 

Z 

Zero, 4-4 

ZeroMean, 8-19 

 

, 5-74 5-80 

ALawToLin, 5-77 

bALawToMuLaw, 5-80 

LinToALaw, 5-78 

LinToMuLaw, 5-76 

MuLawToALaw, 5-79 

MuLawToLin, 5-75 

 

, 4-11 

RandUniform, 4-11 

 

, 7-46 7-69 

WTFwd, 7-58 

WTFwdFree, 7-57 

WTFwdGetDlyLine, 7-61 

WTFwdInitAlloc, 7-54 

WTFwdSetDlyLine, 7-61 

WTHaarFwd, 7-49 

WTHaarInv, 7-49 

WTInv, 7-64 

WTInvFree, 7-57 

WTInvGetDlyLine, 7-67 

 

, 1-3 

 

 

IPP , 2-6 

, 8-37 8-52 

GaussianDist, 8-48 

GaussianMerge, 8-50 

GaussianSplit, 8-51 



LogGaussAdd1, 8-47 

LogGaussMax1, 8-45 

LogGaussSingle, 8-37 



MahDistSingle, 8-38 

MeanVarAcc, 8-25 

, 8-48 8-52 

 

RandGauss, 4-12 

, 1-3 

, 2-2 

, 1-3 

 

, 8-27 

AddNRows, 8-17 

CopyCol, 8-31 

Delta, 8-33 

DeltaDelta, 8-35 

EvalDelta, 8-32 

LMThreshold, 8-20 

-8


 

LogAdd, 8-27 

LogSub, 8-28 

ZeroMean, 8-19 

, 8-37 8-39 

, 8-39 8-48 

 

, 

7-49 7-53 

 

, 1-2 

, 5-11 5-37, 8-1 8-20 

Abs, 5-28 

Add, 5-13 

AddC, 5-12 

Div, 5-24 

DivC, 5-23 

Exp, 5-34 

Ln, 5-36 

Mul, 5-16 

MulC, 5-15 

Sqr, 5-29 

Sqrt, 5-31 

Sub, 5-21 

SubC, 5-18 

SubCRev, 5-20 

, 8-1 8-26 

AddAllRowSum, 8-1 

AddMulCol, 8-3 

AddMulRow, 8-4 

DotProdCol, 8-5 

MeanCol, 8-21 

MeanVarCol, 8-24 

MulCol, 8-6 

NormalizeCol, 8-26 

QRTransCol, 8-7 

RecSqrt, 8-12 

SubRow, 8-13 

SumCol, 8-8 

SumColAbs, 8-10 

SumColSqr, 8-11 

SumRow, 8-14 

SumRowAbs, 8-15 

SumRowSqr, 8-16 

VarCol, 8-22 

, 5-111 5-115 

SampleDown, 5-113 

SampleUp, 5-111 

, 4-5 4-14 

 

, 1-4 

, 5-8 

, 3-1 3-3 

GetLibVersion, 3-1 

GetStatusString, 3-3 

, 1-4 

 

SIMD , 1-1 

 

, 4-12 

 

, 8-21 8-26 

 

, 8-27 8-28 

, 6-1 6-8 

Conv, 6-1 

ConvCyclic, 6-3 

CrossCorr, 6-6 

 

, 8-29 8-36 

, 5-95 5-110 

DotProd, 5-108 

Max, 5-97 

-9


 

MaxIndx, 5-97 

Mean, 5-101 

Min, 5-98 

MinIndx, 5-99 

Norm, 5-104 

Normalize, 5-38 

NormDiff, 5-106 

Phasecomplex, 5-46 

Phasereal, 5-47 

PowerSpectrcomplex, 5-48 

PowerSpectrreal, 5-49 

StdDev, 5-102 

Sum, 5-95 

, 4-5 4-7 

Tone, 4-5 

DFT(Goertzel) , 7-35 7-39 

Goertz, 7-35 

GoertzTwo, 7-38 

, 4-7 4-10 

Triangle, 4-9 

 

, 1-1 

, 7-3 7-9 

ConjCCS, 7-8 

ConjPack, 7-6 

ConjPerm, 7-4 

, 7-10 7-12 

MulPack, 7-10 

MulPerm, 7-10 

 

, 1-4 

, 1-4 

, 1-4 

, 1-5 

 

, 6-9 6-52 

, 1-5 

 

, 1-1 

Jaehne , 4-13 4-14 

VectorJaehne, 4-13 

, 4-1 4-4 

Copy, 4-1 

Set, 4-2 

Zero, 4-4 

 

AutoCorr, 6-4 

, 5-40 5-74 

CartToPolarcomplex, 5-68 

CartToPolarreal, 5-69 

Conj, 5-42 

ConjFlip, 5-43 

Convert, 5-40 

CplxToReal, 5-53 

Imag, 5-51 

Magnitude, 5-44 

MaxOrder, 5-72 

PolarToCartcomplex, 5-70 

PolarToCartreal, 5-71 

Preemphasize, 5-73 

Real, 5-51 

RealToCplx, 5-52 

Threshold, 5-54 

Threshold_GT, 5-58 

Threshold_GTVal, 5-62 

Threshold_LT, 5-58 

Threshold_LTInv, 5-65 

Threshold_LTVal, 5-62 

, 7-1 7-12 

 

, 1-2 

, 1-2 

, 1-3 

 

(Window) , 5-81 5-94 

WinBartlett, 5-83 

-10


 

WinBlackman, 5-85 

WinHamming, 5-88 

WinHann, 5-90 

WinKaiser, 5-92 

(Window) , 5-81 5-82 

 

, 6-51 6-52 

FilterMedian, 6-51 

, 3-4 3-5 

Free, 3-5 

Malloc, 3-4 

 

 

, 7-54 7-69 

 

, 5-1 5-10 

And, 5-2 

AndC, 5-1 

LShiftC, 5-8 

Not, 5-7 

Or, 5-4 

OrC, 5-3 

RShiftC, 5-9 

Xor, 5-6 

XorC, 5-5 

-11


 

-12

IPP - XLsoft.com

Create successful ePaper yourself

Delete template?

Save as template?