Finger vein recognition method based on GLCM-HOG and SVM

Abstract

Finger vein recognition is an emerging biometric technology that plays an increasingly important role in daily life. The finger vein texture information is rich but the image contrast is poor, and it is easily affected by background lighting during image acquisition; it is aimed at the problem of incomplete utilization of image information when the traditional feature extraction algorithm is used for finger vein feature extraction. A finger vein recognition algorithm based on GLCM-HOG and SVM is proposed. Firstly, CLHE enhancement and improved adaptive median filtering are applied to the image to enhance the texture of the image and remove noise interference. Secondly, a serial feature fusion method of finger veins based on GLCM-HOG is proposed. The principal component analysis method is used to reduce the dimensionality of the serial fusion feature vectors to reduce redundant feature vectors. Finally, support vector machines (SVM) are used for classification to achieve feature matching and authentication. Simulation experiment results show that the recognition rate can reach 97.3%. Compared with traditional algorithms, the recognition rate and recognition speed of this method are significantly improved, and it has better stability and robustness.