【摘要】 人脸识别问题涵盖了多个领域如图像处理、计算机视觉和模式识别等,文中主要研究了PCA算法的原理,并基于MATLAB软件平台实现了人脸识别系统。本系统对图像进行预处理后,用ORL人脸库中的部分图像作为样本,通过K-L变换获取训练样本的特征值和特征向量,从而得到特征脸向量,然后将测试样本投影到已知的特征脸空间,计算出坐标系数,进而得出分类识别的结果。实验证明了本系统有较高的人脸识别率,具有一定的实际应用价值。更多还原

【Abstract】 Face recognition problem covers many fields such as image processing,computer vision and pattern recognition,etc.,this paper mainly analyzed the principle of PCA algorithm,based on MATLAB software platform and implements a face recognition system. After the system of image preprocessing,with part of the image as a sample,ORL face image database by K-L transform for the training sample of eigenvalue and eigenvector, face vector is obtained, and then will test sample projection to the characteristics of the known face space,to calculate the coordinate coefficient,classification results are obtained. The experiment proves that this system has a higher human face recognition rate,has certain actual application value.更多还原