10.07.2015 Views

To the Graduate Council: I am submitting herewith a thesis written by ...

To the Graduate Council: I am submitting herewith a thesis written by ...

To the Graduate Council: I am submitting herewith a thesis written by ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Chapter 5: Analysis and Results 705.3 Results of our Informational ApproachIn Chapter 4 we discussed our CVM algorithm that quantifies surface shapecomplexity. We compute curvature based on <strong>the</strong> method suggested <strong>by</strong> [Abidi, 1995]and measure boundary complexity as <strong>the</strong> Shannon’s entropy of curvature on 2Dcontours. We have presented <strong>the</strong>se results in [Page et al., 2003b]. We discuss someimportant results on X-ray and range images. We also analyze some limitations ofusing Shannon entropy and <strong>the</strong> need for a normalized information measure beforediscussing <strong>the</strong> results of our graph representation on automotive components.5.3.1 Intensity and Range ImagesIn Figure 5.8(a) we show results on simple curves. We have made a few importantassumptions with <strong>the</strong>se curves. These curves are of <strong>the</strong> s<strong>am</strong>e resolution and areuniformly s<strong>am</strong>pled. We have computed <strong>the</strong> Shannon entropy of <strong>the</strong> turning angle ateach point on <strong>the</strong> boundary as <strong>the</strong> shape complexity measure (SCM). We note thatSCM and CVM are similar measures but are not equivalent.SCM inspired <strong>the</strong>development of CVM, and CVM represents <strong>the</strong> evolution of SCM from lessonslearned on scaling and resolution We would like to emphasize in <strong>the</strong>se results on howshape information behaves with symmetry and how important <strong>the</strong> assumption on sizeand resolution turns out to be. We would also like to note that <strong>the</strong> shape informationfrom <strong>the</strong> Shannon’s measure cannot be compared if <strong>the</strong> two images are not at <strong>the</strong> s<strong>am</strong>eresolution and comparable size. Hence for <strong>the</strong> real data we have normalized <strong>the</strong>segmented region of interest for size and resolution and <strong>the</strong>n computed <strong>the</strong> curvaturebasedmeasure on <strong>the</strong> normalized boundary contour. We would like to recall fromChapter 2 and note that our method falls under <strong>the</strong> boundary-based descriptionmethods. In Figure 5.8(b) we show an ex<strong>am</strong>ple with an X-ray image of a baggage. Thebag contains a few objects that we have segmented manually. We take each of <strong>the</strong>segmented objects and <strong>the</strong>n compute <strong>the</strong> shape information on each of <strong>the</strong>se contours.Our measure categorizes complex objects and simple ones with satisfactory ease.Next, we show some results on range images in Figure 5.8(c).We believe that we willbe able to distinguish between <strong>the</strong> man-made structures that have flat and nice edgeslike <strong>the</strong> building in Figure 5.8 (c) and natural vegetation that has rugged boundaries.5.3.2 Surface RuggednessIn terms of resolution we would like to present some results on syn<strong>the</strong>tic DEMs(Digital Elevation Maps) of <strong>the</strong> s<strong>am</strong>e resolution. The Shannon’s entropy of curvaturegives a consistent ruggedness measure of <strong>the</strong> surface. But we still face inconsistencywith resolution. We formulate our algorithm on <strong>the</strong> heuristic that <strong>the</strong> variation in <strong>the</strong>shape characteristics of surfaces is ma<strong>the</strong>matically <strong>the</strong> variation of curvature. Wedefine shape information as <strong>the</strong> entropy of <strong>the</strong> curvature density of <strong>the</strong> surface underconsideration.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!