Fingerprint Detection and Matching. Finger print recognition is one of the most used biometric in present day society – with its use cases ranging from forensics, security, and identification. In this project I implemented a fingerprint matching software.
Probe Fingerprint (ie. fingerprint we are trying to identify) Sample Fingerprint Minutiae (ie. fingerprint we are using to match the probe fingerprint minutiae points overlayed over probe fingerprint) 1. 2. 3. Probe Fingerprint = Sample #2 (as you can see the blue minutiae points from #2 fairly match the probe images' shown in red)
Project Keywords: | MATLAB | Biometrics | Fingerprint Recognition | RANSAC
Implementation Process: This project was implemented in two parts - part one was to extract features and part two was to match features to classify from whom the fingerprint came from. For feature selections, I decided to use minutiae points. Minutiae features are ridges and bifurcations in a fingerprint, in particular they are the points where these ridges and bifurcations start and end. To extract the minutiae points, I used a off the shelf software to detect minutiae points from the National Institutes of Standards and Technologies (NIST), called MINDTCT. Once the features had been extracted I wrote a matcher program in MATLAB. To do this I used the RANSAC classification algorithm and compared the locations and orientations of the minutiae points from the probe and test samples to get the total number of matching ridges and bifurcations in the fingerprints.