Finger-knuckle-print verification based on vector consistency of corresponding interest points

This paper proposes a novel finger-knuckle-print (FKP) verification method based on vector consistency among corresponding interest points (CIPs) detected from aligned finger images. We used two different approaches for reliable detection of CIPs; one method employs SIFT features and captures gradient directionality, and the other method employs phase correlation to represent the intensity field surrounding an interest point. The consistency of interframe displacements between pairs of matching CIPs in a match pair is used as a matching score. Such displacements will show consistency in a genuine match but not in an impostor match. Experimental results show that the proposed approach is effective in FKP verification.