A hierarchical finger selection method for finger vein recognition using sift

Finger vein has currently observed as an competent biometric strategy. In this paper, we proposed a hierarchical preferred finger authentication using finger vein, which is main source over the traditional score-level fusion methods. In this proposed technique, initially we consider feature extraction of the 6 fingers of center, and ring finger for both right and left hand as the essential feature for finger vein recognition and then a new verdict rule is defined based on the sift algorithm. Finally, we find that the finger will provide maximum number robust feature of points for the matching algorithm. Considerable experiments are conducted to evaluate using SIFT algorithm and results of conducted experiments show that it can achieve a favorable performance.