A Novel Single Training Sample Face Recognition Algorithm Based on Modular Weighted (2D)2PCA

Single training sample face recognition technique is an emerging research hotspot in the fields of computer vision and pattern recognition for its practical application and theoretical research value. In this paper, we propose MW(2D)(2)PCA (Modular Weighted (2D)(2)PCA) algorithm based on the study of (2D)(2)PCA, in which weighting method is introduced to emphasize the different recognition results influenced by the eigenvector of different eigenvalue, and image blocking method is used to obtain more detail face information. Finally, maximum membership degree principle is used to recognize unknown face sample. Plenty of simulation has been fulfilled, including the experiments about influences of weighting method and image blocking method. And comparative analyses of various algorithms show that the proposed algorithm can achieve better recognition effects.