基于PCA、SVM算法实现人脸识别

Face Recognition Based on PCA and SVM Algorithm

人脸识别是视觉图像识别领域的一项重要技术。本文结合目前主流的两种算法PCA和SVM,先通过PCA提取低维度的人脸图像特征向量,再用SVM进行分类识别。使用MATLAB对ORL人脸库进行仿真实验,测试仿真结果正确率在90%以上,表明该方法行之有效。

Face Recognition is an import technology in the image pattern recognition field. In this paper, PCA （Prin-cipal component analysis） is combined with SVM （Support vector machine） algorithm. After extracting the feature from face image by PCA, it uses SVM to classify face. MATLAB is used to simulate experiment based on ORL face database. The accuracy of the simulation results is above 90%, which indicates that the method is effective.