Methods Of Face Recognition Based On Gabor Wavelets

Face recognition is one of the main identification technologies using biometric technology, applied to many fields. Face recognition technology field involves more professional field, including the field of artificial intelligence, image processing and analysis, image coding, pattern recognition, computer vision, biometrics technology, and is one of the focus of research in recent years. In addition, the safety of face recognition technology is high, the market demand for large, can be applied to the personnel file management, the suspect search to identify, security verification and identification of the security system, certificate and credit card verification, human-computer interaction system, access control and automatic teller machine etc. Therefore, face recognition has great theoretical research value and commercial development prospects.The paper based on the principle of several face recognition algorithm and Gabor wavelet transform theory and its application in human facial feature extraction in the face elastic graph, constructed by Gabor wavelet transform, in order to determine the face feature template; And then Gabor wavelet and PCA are combined to design a simple face recognition system, and the corresponding experiments were carried out on ORL and YALE face database. The main contents of the paper include:1. Studies and analyzes the process of face recognition method of face detection and face recognition method has carried on the detailed discussion.2. Research on the theory of Gabor wavelet transform, the Gabor wavelet transform is applied to extract the features of face images. First use of Gabor wavelet transform face elastic graph is constructed, which can determine the face feature template, to verify the Gabor wavelet transform in the role of facial feature extraction.3. Because after the Gabor wavelet transform, dimension will become very high, is not conducive to face recognition. Therefore, design a simple of a face recognition system based on Gabor+PCA. Through experiment verification, PCA through dimensionality reduction and by careful selection of dimension after dimension reduction can solve the learning problem in face recognition.