Using 2DGabor values and kernel fisher discriminant analysis for face recognition

A novelty method of 2DGabor-KDA(kernel Fisher discriminant analysis) for face recognition is proposed. First of all, every facial image is segmented into several sub-areas according to the five particular face parts and then the features of five key parts are extracted through 2DGabor wavelet, average values are calculated from feature vectors gained from the corresponding pixel of each test sample and then the eigenvectors are gained, in the next place, KDA is applied to kernel-process the gained eigenvectors, and then SVM(Support Vector Machine) is adopted to recognize the face images. The numerical experiments on face database of ORL demonstrate that this method achieves better results of face recognition than other methods and shows stronger robustness to changes of illumination, expressions, poses and so on.