Preliminary work on rotation-invariant algorithms for contactless palm vein biometrics

Contactless palm vein biometrics is an emerging technology that identifies a person based on his/her palm's vascular pattern without needing the person to have contact with the recording device. The vascular pattern is underlying beneath the human palm, so the palm vascular vein pattern information maintains secrecy. Most of the existing contactless palm vein systems are unable to locate the region of interest (ROI) when the rotation of the user's hand is large and rotated to certain angles. Large rotation in the hand position will adversely affect the performance of the contactless palm vein biometrics system. In this paper, we propose a novel rotation-invariant algorithm called Zig-Zag valley detection (Z2VDA) to solve this problem. The proposed Z2VDA uses a circular intersection technique to locate the points on the root of the fingers which serve as the landmark points to correct the rotation of the hand. An empirical study shows that the proposed algorithm works well and is able to detect a hand effectively in eight different rotation angles.