Research Of Face Recogntion System Based On OpenCV

With the progress of science and technology, and people's living standards continue to improve, people pursue a more convenient way of life and expect to have a intelligent future life. Face recognition is an ideal automatic authentication method because of its advantages of fast and simple authentication, and its wide application prospect makes it a hotspot in the field of computer vision. This paper is surrounded with the face recognition technology,research and design an automatic face recognition system with the function of face recognition.In this paper, doing step by step and block method research face recognition technology related content according to the general process. In more than one part relates to face recognition technology, analysis and experimental test, and puts forward and design some solutions in the related research directions, not only promoting the development of face recognition technology, but also provides reference for future innovative applications of face recognition.Through the deep research of face recognition technology, this paper summarizes the main technical route of face recognition, and the problems encountered in the practical application of face recognition. The main research achievements of this paper are: 1) in the part of image pre-processing, designed a correction method for face images based on face key point alignment algorithm ; 2) in the face detection part, proposed a multi stage scale factor of face image retrieval traversal strategy; 3) in the face recognition, researched the effect of SVM of the kernel function on the classification accuracy of facial feature extraction algorithm based on PCANet model; 4) in the face of attribute recognition and the "real face"in anti-spoofing detection part, designed a face gender recognition scheme with HOG features which based on SVM, designed a face detection scheme based on an interaction action and in the game interaction; 5) Using OpenCV and Qt library based on C++ language development a graphical interface of the face recognition system.Through the research we found that the design of the face image for face alignment algorithm correction scheme can correct the face image based on the purpose of the multi stages; scale factor of face image detection traversal strategy can optimize the face detection performance in a certain extent; PCANet face recognition algorithm on ORL face database,Yale face database and FERET face database the rate reached 98.5%, 100% and 95.75%respectively based on face recognition can meet the requirements under simple conditions;the design desired effect of face gender recognition based on SVM features and HOG scheme can reach the experimental research; design based on interaction and face detection scheme based on live interactive games can meet the basic needs of the detection system in anti-spoofing detection.