21.02.2013 Views

Advances in Fingerprint Technology.pdf

Advances in Fingerprint Technology.pdf

Advances in Fingerprint Technology.pdf

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.

5. Traur<strong>in</strong>g 19<br />

6. K<strong>in</strong>gston 20<br />

7. Osterburg et al. 21<br />

8. Stoney and Thornton 22,23<br />

9. Champod 11,24<br />

10. Meagher, Budowle, and Ziesig 25<br />

Each of these models focuses on f<strong>in</strong>gerpr<strong>in</strong>t m<strong>in</strong>utiae. Alternative approaches<br />

based on the f<strong>in</strong>gerpr<strong>in</strong>t pore structure are of corollary <strong>in</strong>terest but are<br />

outside the scope of this chapter.<br />

Galton Model (1892)<br />

Description of the Galton Model<br />

Galton 14 made the first attempt to quantify f<strong>in</strong>gerpr<strong>in</strong>t <strong>in</strong>dividuality. His<br />

approach was to divide a f<strong>in</strong>gerpr<strong>in</strong>t <strong>in</strong>to small regions, such that the ridge<br />

detail with<strong>in</strong> each region could be treated as an <strong>in</strong>dependent variable. Galton<br />

worked with photographic enlargements of f<strong>in</strong>gerpr<strong>in</strong>ts. The enlargements<br />

were placed on the floor and paper squares of various sizes were allowed to<br />

fall haphazardly on the enlarged f<strong>in</strong>gerpr<strong>in</strong>t. Galton then attempted to reconstruct<br />

the ridge detail that was masked by the paper squares, given the<br />

surround<strong>in</strong>g ridges. He sought the size of square region where he could<br />

successfully predict the actual ridge detail with a frequency of 1/2. Galton<br />

found that for a square region “six ridge <strong>in</strong>tervals” on a side, he was able to<br />

correctly predict the hidden detail with a frequency of 1/3, and consequently<br />

concluded that a square region with five ridge <strong>in</strong>tervals on a side was very<br />

nearly the size he was seek<strong>in</strong>g.<br />

To ensure that any errors would overestimate the chance of f<strong>in</strong>gerpr<strong>in</strong>t<br />

duplication, Galton used a six ridge <strong>in</strong>terval square region and then assumed<br />

a probability of 1/2 for f<strong>in</strong>d<strong>in</strong>g the exist<strong>in</strong>g m<strong>in</strong>utiae configuration, given the<br />

surround<strong>in</strong>g ridges. The total area of a complete f<strong>in</strong>gerpr<strong>in</strong>t was estimated<br />

to consist of 24 such square regions. Assum<strong>in</strong>g <strong>in</strong>dependence among these<br />

regions, Galton calculated the probability of a specific f<strong>in</strong>gerpr<strong>in</strong>t configuration,<br />

given the surround<strong>in</strong>g ridges, P(C/R), us<strong>in</strong>g Equation (9.1):<br />

P(C/R) = (1/2) 24 = 5.96 × 10 –8 (9.1)<br />

Galton next estimated the chance that a particular configuration of surround<strong>in</strong>g<br />

ridges would occur. Two factors were considered: (1) the occurrence<br />

of general f<strong>in</strong>gerpr<strong>in</strong>t pattern type, and (2) the occurrence of the correct

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

Saved successfully!

Ooh no, something went wrong!