The Research Of Face Recognition Algrithm Based On Multi-view

With the suddenly development of the computer technology, and the complicated environment of society gradually, People’s daily lives security needs more and more advanced science and technology to guarantee.Undoubtedly,the technology of face recognition is the best choice, as a comparing face visual feature information for computer technology of identification,which bases on fixing-processing-analyzing the digital image with computer,then extracting the characteristic of the face image,lastly, determined the identity of the face by detecting.It has widely application value in the public security, information safe,human-computer interaction and so on.Based on analyzing and summarizing the research status and development trend of face recognition algorithm under the condition of multiple points of view, the article makes a deep and systematic research aiming at pose and illumination angle change and other factors. For appearing shade of critical areas of some collected images, we need to rely on key characteristics for identification, in order to avoid false recognition and leakage problem.Face recognition system has two main components:face localization and detection, face recognition. On the basis of these two aspects, the paper first introduces the face detection and location in a complex multi-angle statistical model based on probabilistic algorithms; then by size normalized grayscale normalization techniques for facial image pretreatment; then adaboost template detection method based on multi-angle, completed the pre-estimate of the multi-angle face recognition and face angle.