基于c++的人脸识别系统的设计与实现-软件工程专业论文.docxVIP

Abstract Face recognition technology is a classic and novel technology, which is widely used in many fields. The basic principles and algorithms of face recognition are studied in the dissertation, and the demand for face recognition system is analyzed. The overall framework and the main functional modules of face recognition system is designed, which Includes face image input, preprocessing modules, face detection, face recognition and so on. Finally, a face recognition system is implemented on the basis of VC ++, and the system is tested and the test results is analyzed in the thesis. The main work of the thesis are: The basic principle and algorithm of face recognition are studied firstly. The basic process of face recognition and the impact of various image parameters on performance in face recognition are analyzed. The geometric characteristics of face recognition technology face recognition method is disscussed, which includs face recognition method basing on Eigen faces (PCA), neural networks’ face recognition method, elastic graph matching face recognition method and Supporting Vector Machine (SVM)’s recognition method. The advantages and disadvantages of these methods are given in the thesis. The requirent analysis of a face recognition system is gaven. It includes five functional modules of the system, which are the image acquisition module, image pre-processing module, face location module, feature extraction module, identification module, and gives processes of each module. The main modules of the face recognition system are designed, which covers image pre-processing module, face location module, feature extraction module and identification module. Giving the main functions’ code of each module and classes designed for the system. The face recognition system is tested, and the test results are analyzed. Test results show that the system meets the demand. Keywords: Face Recognition Image preprocessing Face location Signature extracting 目录 第一章 绪论 1 1.1