Finger vein recognition method based on ResNet and DenseNet

The invention discloses a finger vein recognition method based on ResNet and DenseNet. Two networks of ResNet and DenseNet are adopted to respectively process vein images of the same vein image in different modes. Wherein the ResNet is used for processing an original vein image; and the DenseNet is used for processing the vein image after the region of interest (ROI) is positioned, and is used for processing the vein image after the region of interest (ROI) is positioned. The two networks are mutually independent, and feature extraction is carried out on the two networks respectively. And finally, performing feature fusion on output results of the two networks by using a self-adaptive threshold fusion method to obtain a final recognition result. According to the method, richer feature information can be obtained, and the recognition accuracy can be improved; effects exerted by the two networks can be accurately evaluated, integration is carried out on the basis, and the accuracy of a final result can be impro