A Finger Vein Recognition Algorithm Based on Gradient Correlation

Finger vein authentication plays an important role in the field of vascular certification because of its simplification and high efficiency. In this paper, a finger vein recognition algorithm based on gradient-correlation is presented. Firstly, a method based on histogram statistics is given to recognize whether finger vein images or not. Secondly, matched filter based on maximum curvature model is adopted to extract the gradient image of finger vein. Finally, cross-correlation between two gradient images is figured out to estimate their similarity. The maximal correlation is used to decide whether matching or not with threshold. Experimental results show that FAR and FFR are 0.375% and 1.20% respectively. This algorithm is simple, efficient, insensitive to noise.