213 / 2017-01-11 19:45:52

Automated Hand Registration for Patented Finger Crease Pattern based Biometric Recognition

biometric,biometric registration and verification,Finger-Crease patterns

Abstract— Biometrics has wide applications in fields of automated identity authentication for insuring security and privacy of individuals or resources. Hand Geometry, Face, Iris, Fingerprint etc. are well-known biometric modalities. The presented research work reports two novel developments achieved over Image Pattern Based Biometric Technique, already patented by our organization, that authenticates the identity of individuals through their finger-crease patterns. These developments are: 1) Automated Hand Registration (AHR) to eliminate operator’s intervention and resulting subjectivity during hand registrations. 2) “Crease Based Compaction (CBC)” of image patterns to obtain compact, but at the same time feature rich, biometric templates. These enhancements have been tested successfully on 300 hand images of 21 individuals. The achieved success rate of AHR is over 99%. Also the templates achieved through CBC are 19.2 % more compact and 14.7% richer in the density of knuckle creases as computed during the performance evaluation.