Face recognition based on fusion of Gabor and 2DPCA features

A new face recognition algorithm based on fusion of 2DPCA and Gabor features with DCV method is presented. The method first extracts features by employed 2DPCA and Gabor wavelets respectively. And the `z-score´ method is applied to normalize the 2DPCA feature and Gabor feature. Then the 2DPCA feature is combined with the Gabor feature by the append rule. In order to overcome the small sample size (`SSS´) problem, the DCV method is then applied to the combined feature vector to extract discriminate nonlinear features for recognition. Finally, Nearest Neighbor (NN) method is used to classify. Experimental results on ORL database show that the proposed method achieves higher recognition rate compared with other methods, especially, when the number of the training set is small.