Face Recognition by Combining Wavelet Transform and k-Nearest Neighbor

Face Recognition by Combining Wavelet Transform and k-Nearest Neighbor

A novel technique for face recognition is presented in this paper. Wavelet transform and k-nearest neighbor rule are combined in this technique. Wavelet transform is adopted to obtain the low pass images of the original face images in order to reduce the impact of facial expressions. When the scales of the head of an individual in sample images are different, wavelet transform is employed to extract the position of racial features. Then a new method is adopted to construct feature vectors using the grey values of face images. Finally, k-nearest neighbor rule is employed to complete the recognition process. Experimental results show that the proposed technique is efficient at computation and robust to the facial expressions.