Biometric identification through palm and dorsal hand vein patterns

Hand vein patterns are among the biometric traits being investigated today for identification purposes, attracting interest from both the research community and industry. This paper presents a multimodal system that combines hand-palm vein and hand-dorsal vein biometrics information at the score level. The palm and dorsal veins are considered as texture samples being automatically extracted from the user's hand image. A 2D Gabor filter is employed for texture feature extraction. For matching, the newly acquired biometric samples are compared with those stored in the system database, at the enrolment stage. The metric used is based on the Hamming distance. A palm and dorsal vein database is built using the proposed acquisition system. This paper proposes a novel multimodal system that combines palm and dorsal vein information at the score level, with the experimental results showing that with the proposed system a much lower Equal Error Rate (EER) can be achieved, in comparison to existing unimodal systems.