Finger vascular pattern imaging - A comprehensive evaluation

Highly unique pattern of the vein within the finger makes it robust biometric characteristic which is instrumental in achieving low error rates for verification. Vascular pattern is reliable in many terms as they are characterized by very low error rates, good spoofing resistance and user convenience when compared with other existing biometric modalities. In this paper, we present an extensive study of the state-of-the-art algorithms for vein recognition on a newly developed vein sensor. To this extent we collected a database using our newly designed finger vein imaging sensor. The database comprises of 1780 finger vein samples corresponding to 89 unique fingers from 32 subjects and benchmark the verification results on the newly composed database.