07-Biometrics-Lecture3-Day2-Part2-14-00-16-30

07-Biometrics-Lecture3-Day2-Part2-14-00-16-30 - Continuing...

Info iconThis preview shows pages 1–7. Sign up to view the full content.

View Full Document Right Arrow Icon
1 Biometric Identity Verification http://scgwww.epfl.ch/courses Continuing Education – COST 2101 Training School Dr. Andrzej Drygajlo Speech Processing and Biometrics Group Signal Processing Institute Ecole Polytechnique Fédérale de Lausanne (EPFL) Center for Interdisciplinary Studies in Information Security (ISIS)
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 Fingerprint Recognition Generalities Applications Fingerprints and their images History of fingerprints Fingerprint sensing Fingerprint features Fingerprint matching Fingerprint enhancement Performance evaluation Synthetic fingerprint generation
Background image of page 2
3 5 10 15 20 25 5 30 35 40 5 5 Smooth orientation fields Fingerprints from the same finger • Having a noisy fingerprint (right). The direct use of the local structure for matching is very difficult. Fingerprint matching with relative pre-alignment •However ,the global structure is rather stable against noise. colour = local orientation of the global pattern
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
4 Global ridge feature-based matching Most frequently used features for fingerprint matching: -Orientation image -Singular points (loop and delta) -Ridge line flow -Gabor filter responses
Background image of page 4
5 Local texture analysis The fingerprint area of interest is tesselated with respect to the core point The local texture information in each sector is decomposed into separate channels by using a Gabor filterbank Feature vector: 80 cells (5 bands and 16 sectors) Filterbank of 8 Gabor filters (8 orientations, 1 scale =1/10 for 500 dpi fingerprint images) Each fingerprint is represented by a 80x8 = 640 fixed-size feature vector, called the FingerCode Computation of average absolute deviation (AAD)
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
6 FingerCode Approach
Background image of page 6
Image of page 7
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 06/25/2009 for the course MATH MAT 400 taught by Professor Jamespotvein during the Fall '08 term at University of Toronto- Toronto.

Page1 / 37

07-Biometrics-Lecture3-Day2-Part2-14-00-16-30 - Continuing...

This preview shows document pages 1 - 7. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online