Research On Front View Single Face Recognition Technology

With the thorough of reform and opening-up, the development of economy construction and the period of civilization ,the complicate of social public order and the rigorous situation of international anti-terror. National security and city safety have been focuses increasingly. System security is of vital importance with the premise to make our personal identification. Since some biological characteristics are intrinsic and stable to people and are strongly different from others, they can be used as features for identity check. Among all the characteristics of human, the characteristics of face are the most direct tools which are friendly and convenient and can be accepted by the customers easily. Naturally, face identification has bright application future in the field of criminal distinguishing, safety check system, credit card identification, medicine, file management, video meeting, people-machine exchange system, and has been a popular study direction of information science. Face recognition can be divided into: face detection and location, feature extraction and recognition.In the part of face location, research on the interference within the face aiming at face's complexity when locating the face, this paper advances a new algorithm of multiple-step hiding genetic algorithm to locate the face, which doesn't serve the result obtained from the first genetic algorithm as face contour, but as region parameters to locate a hiding region based on the parameters and hide this region. On the basis of the hided image, searching for the next hiding region until finding the optimal region, and locate the contour ultimately based on the processed region. This algorithm can solve the interference within the face effectively when searching the face edge adopting genetic algorithm.In the part of features abstraction, combining Gabor wavelet transform and eigenface to abstract face feature because Gabor wavelet is sensitive to the image edge but not to the illumination change and can offer good direction choice and scalechoice characteristic. Discrete Gabor wavelet is applied to face images with parameters deliberately selected. A method named two-step dimension reduction is used to reduce the dimension of Gabor feature vector. Obtaining transform feature bases on the basis of principal component analysis method and projecting wavelet transform to these bases and obtaining the face transform features finally.Finally, in the step of face recognition, analyzing some distance measure problems when judging and classifying. And practicing a method which is based on the combination of Gabor wavelet and principal components analysis.Experimental results show that this paper puts forward the algorithm of multiple-step hiding genetic algorithm to locate the face which is exact and effective, and the method on the basis of Gabor wavelet and principal components analysis is tried and true, which has vital application foreground.