Realization of face recognition system based on Gabor wavelet and elastic bunch graph matching

This paper studies the face recognition system based on the Gabor wavelet and elastic bunch graph matching, and gives the realization method of the system. In this system, a face is expressed by a set of feature points, each of these feature point can be described as a Gabor Jet. For the training images, feature points can be located manually, and then get Gabor. For the sample face image, the corresponding feature points are searched automatically, which coarse positions are located by the training face structure knowledge, and then we use elastic bunch graph to adjust the accurate positions. At last Gabor Jets are obtained from the sample face images. The similar degree of the sample images is expressed by the related measure function of Gabor jets. The experiments show that the face recognition system tested on the face FERET database is efficiency.