Finger-vein Verification Using Gabor Filter and SIFT Feature Matching

In this paper, a novel method to verify the infrared finger-vein patterns is proposed for biometric purposes. Firstly, we select parameters for Gabor filter with eight orientations to exploit finger-vein network, then we extract vein patterns by the fusion of two distinct orientation results. Secondly, we utilize SIFT features to offset the effect of images rotation and shift impact during finger-vein verification. Finally, the number of matching SIFT features between the registered and test finger vein patterns is calculated as the similarity measurement to verify the personal identification. The experiment results show that EER is low to 0.46%, which demonstrates our proposed approach is valid and effective for finger-vein verification.