The Research Of Face Recognition Technique Based On Statistics

Face Recognition is one of the key techniques of the Digital Surveillance System. It has very high value of research and good future of application.Three face recognition method based on Statistics which include eigenface, fisherface and Singular value decomposition are implement on three standard face database. Some former conclusion is validated.An approach of the face recognition based on singular value decomposition in the wavelet domain. The key of improving the rate of face recognition is to reduce the affection of illumination, pose and expression. Singular value decomposition (SVD) is a effective face recognition method based on algebraic characteristic. But it isn't robust to variance of illumination, pose and expression. In this paper, we study the properties of different wavelet decomposition coefficients when illumination, pose and expression are changed. Then, Robustness of face recognition is improved by decomposing different coefficient.An approach to face recognition based on twice two-dimensional PCA (22DPCA). Robustness to variance of illumination, pose and expression of two-dimensional principal component analysis (2DPCA) is better than PCA, whereas 2DPCA takes up plentiful storage. The error rate of 22DPCA is same as 2DPCA, but its storage is far below 2DPCA.