The Application Of Classifier With Rejection Mechanism In Face Recognition

When identifying biometrics,face features have natural advantages due to their unique friendliness.However,some high correct recognition rates such as support vector machine(SVM)classifiers,support vector data description(SVDD)classifiers and Deep learning classifiers still have a misrecognition rate of about 2%;and for attendance system,there will be a situation of proxy attendance,that is,there is no suitable rejection mechanism.In face authentication recognition,in view of the problems that SVDD can not be tightly wrapped,there is no suitable rejection mechanism,and the correct recognition rate can not approach 100%,a high correct recognition rate classifier design algorithm based on similar feature set and close enclosed surface point set(CES-SVDD)is proposed.The algorithm is composed of "compact parcel set construction algorithm for homogeneous feature sets","algorithm for solving tightly wrapped surfaces of homogeneous feature regions based on homogeneous feature point sets and parcel point sets",and "suitable rejection region setting algorithm for multi-class classifiers",which solves the problems of the traditional classifier.In order to make the training samples robust,the extended methods of fuzzy,translation and rotation are applied to the training samples.The experimental results show that the correct recognition rate of CES-SVDD classifier can approach 100% under the condition of small rejection rate.In view of the effectiveness of face recognition,it is applied to the attendance system based on face recognition,and then carry out research to obtain attendance information,timely understand the attendance of students,which can effectively avoid the occurrence of attendance generation.