A Novel Biometric System Based on Hand Vein

As an important member of biometric family, the vein patterns rely on the interior biological information of the body, and therefore, cannot be easily damaged, changed or falsified. In this paper, we present a new hand vein recognition system, which extracts and combines the dorsal, palm and finger vein for personal recognition. In the proposed system, the whole infrared frontal and back images of a hand are firstly captured. Secondly, the region of interest (ROI) of dorsal, palm and finger vein images are cropped. Thirdly, the veins in each ROI are extracted and matched by using multi-scale 2-D Gaussian matched filter. Finally, the matched distances are fused to form the final distance for decision by employing SVM classifier. The proposed system is tested on a large database (96 hands, 1900 images) and the result is satisfactory with equal error rate (EER) of 0.0223%.