3D Face Recognition Algorithm Based on Deep Belief Network

Abstract

Although the depth learning algorithm reduces the workload of face recognition to a certain extent, the local characteristics of 3D face images is ignored, resulting in low accuracy of 3D face recognition. Therefore, this paper proposed a new 3D face recognition method using LBP algorithm improve depth belief network. Firstly, LBP algorithm and depth belief network are analyzed, and then LBP texture feature vector of 3D face image is obtained, which is used as the input feature of depth belief network to capture the local information of 3D face image. Finally, this paper designed a 3D face image recognition process and realized 3D face recognition based on improved depth belief network. The proposed method is trained on FERET face image database, and the simulation results show that the proposed method has higher 3D face recognition rate and shorter recognition time, compared with the comparison method, which shows that the application effect of the improved depth learning algorithm in 3D face recognition is better.