Research on Finger Vein Recognition Based on Improved Convolutional Neural Network

Abstract

Aiming at the problems of low recognition accuracy and poor generalization performance of finger vein recognition method, a finger vein recognition method based on the combination of deep convolutional network and extreme learning machine was proposed. Use deep convolutional networks to automatically extract feature from finger veins to reduce the large amount of effective information lost in traditional method feature extraction. At the same time, in order to enhance generalization, the deep convolutional network has been improved to remove the original fully connected layer Add extreme learning machine layers to identify the extracted feature vectors. An experimental analysis of the proposed method was performed on a common finger vein dataset. Experimental results show that, compared with other finger vein recognition methods, this method has higher accuracy and stronger generalization performance in finger vein recognition.