dental biometrics used in forensic science - Call for papers= Asian ...

JERS/Vol.III/ Issue I/January-March, 2012/26-29
Journal of Engineering Research and Studies E-ISSN0976-7916
Research Article
DENTAL BIOMETRICS USED IN FORENSIC SCIENCE
Shubhangi Jadhav 1 , Revati Shriram 2
Address for Correspondence
1 Dept. of Instrumentation and Control, D Y Patil College of Engineering, Pimpri, Pune, INDIA
2 Cummins College of Engineering for Women, Karvenagar, Pune, Maharashtra, INDIA
ABSTRACT
Dental biometrics used in forensic science for human identification. It utilizes dental radiographs. This radiograph provides
information related to teeth shape, teeth contour and relative position of neighboring teeth, also it gives shapes of dental
work like crowns, filling & bridges etc. This paper includes different method used for dental biometrics and related
information. Dental biometrics requires antemortem (AM) and postmortem (PM) radiographs for finding unidentified
subject. Dental biometrics having three stages: Preprocessing and segmentation of radiographs, Contour extraction or dental
work extraction, Atlas registration and matching. Segmentation can be done by various methods that are mentioned in this
paper. Contour or shape of teeth and dental work can be extracted by using active contour model (ACM) or active shape
model (ASM) methods. Atlas registration is the method used for labeling to teeth, which will help in the matching stage.
Matching of AM radiograph with PM radiograph can be done by using algorithms.
KEYWORDS Biometrics, Dental Radiograph, Forensic Science, Teeth contour, Dental work.
1. INTRODUCTION
Determining identity of an individual is becoming
very important now days. For this purpose biometric
system is used. A biometric is a measurable physical
characteristic which are reliable than a password.
Many biometric systems are employed which work
on the basis of image analysis. “Biometrics” is a
general term used alternatively to describe a
characteristic or process.
The biometric systems are divided into two
categories as: behavioral biometrics and
physiological biometrics. In behavioral biometric
systems a person is identified based on how he
performs something. The most popular technique
under this category is the voice verification
technique. Here user has to actively participate in the
process of identification i.e. he needs to perform
some action in front of the machine. Other examples
of physiological biometric characteristic are making a
signature, walking, typing on a keyboard. In
physiological biometric system a person is identified
based on a unique characteristic of some body organ
of that person. It’s a passive process. All the user has
to do is stand in front of the system or at max touch
the sensor of the system with a finger tip (finger
biometry) and wait till identification process is
completed. The typical examples of physiological
biometric system are Iris biometry, face biometry,
finger biometry, ear biometry, dental biometry etc.
Dental biometry used in forensic identification. This
technique requires ante mortem and postmortem
radiographs. In this both radiographs are segmented
and matched for identification of unidentified victim.
There are some various techniques of dental
biometric that are proposed by different authors as in
below section.
1.1. Forensic Identification
Forensic identification is used for suspect
identification and victim identification. Victim
identification is done by physical biometrics. Dental
radiograph can be used for victim identification based
on dental evidences. [1]
1.2. Types of Dental Radiographs
There are three types of dental radiograph (X-ray):
• Bitewing X-ray- Bitewing x-ray is taken at
routine check-ups.
• Periapical x-ray- It shows entire tooth, including
crown, root and bone.
• Panoramic x-ray- It gives broader overview of
entire dentition. It shows not only teeth also
sinus, upper and lower jawbone.
1.3. Dental Biometry
Forensic dental biometrics used to identify
unidentified victims. Automatic forensic
identification utilizes dental radiographs [1].
Biometrics can be classified in two category based on
characteristic like behavioral and physical. Physical
biometric represents iris, fingerprint, face recognition
etc. Behavioral biometric represent voice, gait,
signature and all behavioral traits of individual. [1]
Evaluation of biometrics features requires
characteristics such as universality, uniqueness,
permanence, performance, collectability and
acceptability. [1]
2. DENTAL IDENTIFICATION SYSTEM
The components of Dental identification system
are:
1) Dental Radiograph
2) Radiograph preprocessing and segmentation
3) Contour extraction
4) Atlas registration
5) Matching of Radiograph
In block diagram of Dental identification system
(Fig.1), dental radiograph of patients are collected for
dental identification system. These radiographs firstly
preprocessed for filter out unwanted background
present with teeth. Then radiographs segmented for
region of interest. Contour of teeth are extracted from
radiograph and also contour of dental work present
on radiograph is extracted using active contour

model. Atlas registration is used to give labeling or
numbering to the teeth present in jaw, it will help in
matching stage. Radiograph from database selected
and this radiograph are preprocessed, segmented and
contour extracted from it. Also registration of teeth is
given as per atlas registration. And last stage of this
system is matching, in this feature extracted from
these two radiograph matched with each other using
algorithm and final result is identification of person
based on matching distance between radiographs[1]
[5].
Radiograph Collection
Radiograph segmentation
Matching tooth
contour
Contour Extraction
Dental work Extraction
Atlas Registration
Matching of Radiograph
Fusion
Subject Identification
Fig.1 Block diagram of Dental
Identification System
3. HUMAN IDENTIFIACTION BASED ON
DENTAL RESTORATIONS
Preprocessing and Segmentation:
Dental radiograph initially converted into gray scale
image. Region of interest (ROI) are decided on
radiograph. Algorithm is used to determined gray
threshold value of ROI. Then histogram is prepared
for the grayscale threshold. To smooth the histogram
filter is used to filter out unwanted intensities from
image .After this gray image converted into binary
image. Active contour model (Snake method) is used
for segmentation of image [6] [15].
Dental Code:
Dental code is used to locate teeth, size of teeth and
distance between DR or teeth.
Dental Atlas registration:
Method or code is developed to locate teeth this is
known as dental atlas registration. In this we can give
numbering to teeth from left to right of jaw and also
we can differentiate between upper jaw and lower
jaw. As per [2], HMM and Markov model was used
JERS/Vol.III/ Issue I/January-March, 2012/26-29
Journal of Engineering Research and Studies E-ISSN0976-7916
Matching dental
work
for this registration purpose. This registration will
help in matching stage.
Size of DR:
The method uses to resize DR in same size. It means
amount of pixels should be same for all DR. So it
will help in matching.
Distance between DR:
This is more useful in matching stage to create more
sensitive algorithm of matching. In this distance of
neighboring teeth is calculated.
Matching:
After creation of dental code (DC) ,this DC will
match to the DC of database radiograph and distance
between this two DCs calculated. So finally we
received matching percentage between the AM and
PM radiograph. Based on result subject get identified
[6] [15].
2.1 Nomir & Adlab-Mottaleb
In [2], Nomir & Adlab-Mottaleb introduce a fully
automated approach for dental x-ray images. The
technique depends on applying the following stages:
iterative threshold to divide the image into two parts
teeth & background, adaptive threshold in order to
increase the accuracy and remove teeth interfering,
horizontal integral part in order to separate the upper
jaw and lower jaw, final, vertical integral projection
in order to separate individual tooth.
3.2 Eyad Haj Said, Diaa Eldin M Nassar, Gamal
Fabry & Hany Ammar
In [3], Eyad Haj Said, Diaa Eldin M Nassar, Gamal
Fabry & Hany Ammar presented method of teeth
segmentation using mathematical morphology
approach offers fully automated teeth segmentation
and reduces segmentation error due to inherent to that
improve the definition of teeth versus background
and segmentation performance.
3.3 Supaporn Kiathsin, Adison Leelasantitham,
Kosin Chamnogthai & Kohji Higuchi
A match of x-ray teeth films using image processing
based on special features of teeth by Supaporn
Kiathsin, Adison Leelasantitham, Kosin
Chamnogthai & Kohji Higuchi [4]. This method can
help dentist to match simply pair of teeth using
special features present on teeth film. This presents
some steps: teeth’s picture is scanned and adjusted by
scanner & computer, converted into binary code, and
decoded in chain code (Direction code)
3.4 Hong Chen & A.K. Jain
Dental biometry is used to identify human. In [5],
Hong Chen & A.K. Jain introduced dental biometrics
using active contour extraction model (ACM). In this
they proposed a new dynamic energy term i.e.
directional snake to extract contours of teeth. As per
this paper traditional snake cannot able to
discriminate edges of multiple adjacent objects. So
there can be presence of overlapping images. To
remove this problem Hong Chen & A.K. Jain utilized
direction gradients. The contour extraction process
having three steps: initialization- In this gum line is

used to separate the crown and roots of teeth for the
snake initialization, convergence of Gradient, fine
adjustment
Hong Chen & A.K. Jain [6] presented Dental
Biometrics: Alignment and Matching of Dental
Radiographs. This proposed system has main two
stages: feature extraction, matching.
In this to extract contours of dental work the intensity
histogram of the tooth image is automated with the
mixture of Gaussian model.
In the matching stage three steps given: Tooth level
matching, tooth contours are matched using a shape
registration method, and the dental work is matched
on overlapping areas.
Distance between postmortem and ante mortem
radiographs provide candidates identities to estimate
subject identification.
Hofer M & Marana
In [6], Hofer M & Marana AN presented Dental
biometrics for human identification based on dental
work. Dental biometrics used for forensic
identification, in this dental work information is used
to identify human. This paper three processing steps
were given as per following: Segmentation (Feature
Extraction), creation of dental code, matching
Primarily segmentation is done by detecting
threshold and finally snake (active contour model)
algorithm used. Dental code is prepared as per the
position, size of dental work and neighboring dental
work. Matching done by editing distance between
dental works.
4. DISCUSSION
Dental biometry is used in forensic identification.
Some other methods are also present for forensic
identification. But it is more useful when there is no
any other identity like iris [11], face, fingerprints [16]
etc of victim remaining. Dental atlas can remain in
well condition for some time period after the death of
person so by taking radiograph at the time of PM and
matching with it to database having with forensic
department it can give you identification of victim or
subject. Also dental biometry can be used in human
identification system. Also matching of x-ray films
which is the part of dental biometry can help dentist
to their clinical diagnosis [1] [5]. Dental biometry
requires AM and PM radiograph only for
identification. Dental biometry cannot require
additional information. DNA requires additional
information of subject for identification. Dental
biometry comes under physical biometry so it is
easier to measure quantitatively as compared to
behavioral biometry. Hand vein pattern and retina
scan are physical biometry but it having low
acceptability because it may reveal person’s health
status and it is invasive method. In disasters, victim
or subject’s physical characteristics like iris, face,
fingerprints not available that time biometry related
to these characteristic not worked. So there is then
need of dental biometry because dental jaw
JERS/Vol.III/ Issue I/January-March, 2012/26-29
Journal of Engineering Research and Studies E-ISSN0976-7916
remaining in well condition in massive disasters.
Dental biometry is used for forensic identification.
Also matching of dental radiograph can be used in
clinical diagnosis forensic medicine (Forensic
Dentistry) [1].
Dental biometry having some limitation based on
below mentioned points so in some cases it is not
good choice for identification:
• Positive identification: AM and PM radiographs
should have sufficient data to match, with no any
discrepancy like missing tooth.
• Possible identification: If consistent features are
present in AM or PM radiograph but if there is
problem in quality of either AM or PM
radiograph then this may create no positive
identification.
• Insufficient evidence: Sufficient information is
not available for the basis of conclusion.
• Exclusion: The radiograph information is
inconsistent [1].
5. CONCLUSION
Segmentation of Radiographs: A fast marching based
algorithm is used to segment radiographs into
homogeneous regions so that each region contains
only one tooth [9]. This method utilizes the
characteristics of dental radiographs in that soft tissue
and bone have different pixel intensity. This method
can be extended to analyze other types of
radiographs. ASM-Based Tooth Contour Extraction:
An active shape model (ASM) is used to extract tooth
contours. The traditional ASMs have some problems
that are solved by new ASM. Extraction of Contours
of Dental Work: An anisotropic diffusion used to
enhance the images and extract the contours of dental
work with a reasonable assumption that pixel values
of dental restoration follow a Gaussian distribution
[1] [2] [3] [5] [10] [17]. Hybrid HMM/SVM Model
for Representation of Dental Atlas: Tooth states in
the HMM combined with SVMs model the shape of
teeth, and distance states in the HMM model the
inter-teeth distance caused by missing teeth.
Construction of Observation sequences: An
observation sequence is an alternating sequence of
tooth shapes and distances between successive teeth.
To compensate for the difference between image
resolutions of AM and PM images, the tooth shapes
are normalized to affixed width, and the distances
between teeth are regularized with the width of
neighboring teeth [1] [7] [17]. Location of DW
(Dental Work) can be finding out by implementing
algorithm. In this all DWs in the DWM from left to
right based on the center of mass point of each
individual DW sorted. This is requiring creating
dental code. The upper jaw and lower jaw is
identified by the valley location with respect to upper
jaw and lower jaw [6] [15].
Matching of tooth contours: Tooth contours are open
curves. To match tooth contours, correspondence
between the tooth contours has to be established. An

iterative procedure is present to align tooth contours
and compute distances based on the corresponding
segments of the two curves. Algorithm is presented to
align dental work and by using we can compute the
distance between two images using a metric of
overlapping pixels. Dental restorations are not always
present in the dental radiographs. To fuse the
matching distances between tooth contours and
dental restoration contours, a scheme presented based
on posterior probability. A scheme used to compute
the distance between an image and a subject and the
distance between two subjects [1].
REFERENCES
1. Hong Chen, “Automatic Forensic Identification
based on dental radiographs”, Michigan State
University, 2007.
2. O. Nomir and M. Abdel-Mottaleb, “Human
Identification from Dental X-Ray Images Based
on the Shape and Appearance of the Teeth”,
IEEE Transactions on Information Forensics and
Security, vol. 2, Issue 2, pp. 188 – 197, 2007.
3. E.H. Said; D.E.M. Nassar; G. Fahmy and H.H.
Ammar, “ Teeth segmentation in digitized dental
X-ray films using mathematical morphology”,
IEEE Transactions on Information Forensics and
Security, vol. 1, Issue 2, pp. 178 – 189, 2006
4. Supaporn Kiattism, Adisorn Leelasantitham,
Kosin Chamnongthai and Kohji Higuchi, “ A
match of x-ray teeth films using image
processing based on special features of teeth”,
SICE Annual conference 2008, August 20-
22,2008, The University Electro-
Communications, Japan.
5. Hong Chen and A.K. Jain, “Dental biometrics:
alignment and matching of dental radiographs”,
IEEE Transactions on Pattern Analysis and
Machine Intelligence, vol. 27, Issue 8, pp. 1319 –
1326, 2005.
6. Hofer M and Marana AN, “Dental biometrics :
Human identification based on dental work
information” eHealth2008 – Medical Informatics
meets eHealth. Tagungsband der eHealth2008 –
Wien, 29.-30. Mai 2008
7. Andrea Castellani, Debora Botturi, Manuele
Bicago, and Paolo Fiorini “ Hybrid HMM/SVM
model for the analysis and segmentation of
teleoperation tasks”, In Proc. IEEE International
Conference on Robotics & Automation, page
2918-2923,New Orleans, LA, April 2004.
8. G. Fahmy, D Nassar, E. Haj-Said, H. Chen, O.
Nomir, J. Zhou, R. Howell, H.H. Ammar, M.
Abdel-Mottaleb and A. K. Jain “Towards an
automated dental identification system (ADIS).In
Proc. ICBA(International Conference on
Biometric Authentication),volume LNCS 3072,
pages 789–796, Hong Kong, July 2004.
9. Hong Chen and Carol Novak, “ Segmentation of
hand radiographs using fast marching methods”,
In Proc. SPIE conference on Medical Imaging,
volume 6144, pages 68-76,San Diego, California,
2006.
10. Hao Huang and Jughua Wang, “ Anistropic
diffusion for object segmentation”, In Proc.
IEEE International conference on systems, Man
and Cybernetics, volume3 pages 1563–1567,
Nashville, TN, October 2000.
11. John Daughman, “How Iris Recognition Works”,
IEEE Transactions on circuits and systems for
video technology, Vol.14, January 2004.
JERS/Vol.III/ Issue I/January-March, 2012/26-29
Journal of Engineering Research and Studies E-ISSN0976-7916
12. A.K. Jain and H. Chen, “Matching of Dental Xray
Images for Human Identification”, Pattern
Recognition, vol37, no.7, pp.1519-1532, 2004.
13. H. Chen and A.K. Jain, “Tooth Contour
Extraction for Matching Dental Radiographs”,
Proc. 17 th international conference. Pattern
Recognition, vol.III, pp.522-525, Aug2004.
14. P.Perona and J.Malik, Scale-Space and Edge
Detection Using Anistropic Diffusion, IEEE
Trans.Pattern Analysis and Machine Intelligence,
Vol.12, no.7, pp. 629-639, July 1990.
15. Michael Hofer, Aparecido Nilceu, “ Marana2
Dental Biometrics: Human Identification Based
on Dental Work Information”, Carinthia
University of Applied Sciences, Klagenfurt,
Austria UNESP, Faculdade de Cincias,
Departmento de Computao, Bauru SP, Brazil
16. R.Khanna and W. Shen, “Automated fingerprint
identification system (AFIS) benchmarking using
the National Institute of Standards and
Technology (NIST) special database4”, in Proc.
Inst. Elect.Electron. Eng. 28 th Annu. Int.
Carnahan Conf., Albuquerque, NM, Oct.12-14,
1994, pp. 188–194.
17. Juri Lember, Kristi Kuljus and Alexe
Koloydenco, “Theory of segmentation”,
University Tartu, Swedish University of
Agricultural Sciences, Royal Holloway,
University of London, Estonia, Sweden, UK.
18. Anil K. Jain, Hong Chen and Silviu Minut,
“Dental Biometrics: Human Identification using
dental radiographs”, Department of Computer
Science, Michigan State University, East
Lansing, MI 48824.
19. Marcelo Cavalcanti, Axel Ruprecht, William
Johnson, Southard Thomas, Jane Jakobsen, and
Sao Paulo, “Oral and maxillofacial radiology”,
Oral Surg Oral Med Oral Pathol, Volume 88:
pages 353.357, 1999.
20. Greth Jonasson and Gudrun Bankvall,
“Estimation of skeletal bone mineral density by
means of the trabecular pattern of the alveolar
bone, its interdental thickness, and the bone mass
of the mandible”, Oral Surg Oral Med Oral
Pathol Oral Radiol Endod, Volume92:pages 346-
352 Ba2001.