Research And Implementation On Face Recognition Technology

In all kinds of biological characteristics identification methods, Automatic FaceRecognition becomes the most receptive method of identity recognition methods for itsdirect, safe, reliable and effective characteristics, and it has has important research andapplication values. As one of the most important subjects in the field of patternrecognition, it has become a very active research direction.First, the research progress of face recognition and the common face recognitionmethods are analyzed in this paper. Then, Local Binary Pattern method is used to extractface feature and Principal Component Analysis (PCA) is used to reduce dimension.Finally, a real-time face recognition system is designed and realized. In this paper, mainresearch contents are as follows:①The typical integrated machine learning method, AdaBoost algorithm, isintroduced and the face detection method based on AdaBoost algorithm is researchedand realized. Face image preprocessing methods are also introduced.②Local Binary Pattern and Principal Component Analysis are researched indetail. To resolve the problem of high dimension feature vector that extracted by LocalBinary Pattern (LBP) for face recognition, Principal Component Analysis (PCA) is usedto reduce dimension. First, The original face image is divided into sub-images, and thenthe LBP operator is applied to extract the histogram features. The feature dimensionsare further reduced by PCA for face recognition. The proposed algorithm combines thelocal description ability of LBP and the global description ability of PCA, which caneffectively extract both the structural features and statistical features for facerecognition. Furthermore, the dimension of the extracted features is effectively reduced,which satisfies the requirement of real-time face recognition..③A real-time face recognition system is designed and realized. The experimentsare carried out based on ORL database and Self-built face database. and the validity isverified. The proposed method and traditional LBP method, PCA face recognitionmethod are compared. The experimental results show that the proposed method caneffectively reduce computational complexity, improve recognition accuracy and itsatisfies the requirement of real-time face recognition.