The Video Of Face Recognition System Based On Local Feature Fusion

After a long period of research and development,face recognition technology has become more and more mature,and has been widely applied to public security and daily life.However,because of the influence of various external factors such as illumination and pose,the application of face recognition is still facing many challenges.Feature extraction is the key of face recognition,and face image is obtained after face detection.Extracting the feature of image and comparing the extracted feature vector with the feature vector of training human face image,we can know which category the face belongs,and obtain the final recognition result.This thesis introduces the application of face detection,image preprocessing,local features in face feature extraction algorithm and proposes a new feature extraction algorithm of face recognition algorithm which is adaptive threshold local feature fusion in order to solve the problems of Local Binary Pattern.Finally,design and complete a face recognition system with face detection and recognition.The main work and research results are as follows:(1)A new feature extraction algorithm which is adaptive threshold local feature fusion is proposed.Based on the original LBP,the paper proposes ULBP algorithm with adaptive threshold.Inspired by the facial expression recognition,a fusion of the local features of the adaptive threshold algorithm will involve the local gradient coding.New algorithm can solve the problem which LBP is not adaptive threshold,describing the texture from different directions the characteristics of face images.The texture feature extraction is more abundant,and the experiment is carried out in ORL,YALE and FERET face database,improve the recognition rate.(2)The algorithm of feature extraction based on adaptive threshold feature fusion is encapsulated.In the MFC platform,a face recognition that contains face detection and face recognition module system using C++ and Open CV is built.