基于 PCA 与 LDA 方法的人脸识别讨论摘 要人脸识别技术是生物特征识别中一个重要的讨论领域，可以用于对人类身份的认证，该技术对新时代的综合安全具有非常重大的意义。人脸识别系统是以人脸做为识别身份的媒介，采纳计算机强大的图像处理技术，利用人脸与众不同的特征，讨论匹配和识别方法的系统。本文主要分析讨论了人脸识别及其相关技术（例如 LDA），主要包括基于 PCA 的算法、改进的 PCA 算法以及基于二维的 PCA 算法。首先介绍了人脸的预处理，其目的是增强图像的对比度与平滑度。预处理主要包括灰度变化、平滑处理和灰度变换；然后,利用经典的PCA 算法对预处理后的图像进行处理,得到特征子空间和各训练图像在特征子空间的投影系数。最后, 分析设计了一个基于 PCA 的人脸检测识别系统。关键词：人脸识别 特征脸 PCA 预处理 ABSTRACTFace recognition technology is an important area of research. It can be used in network authentication and biometric identification. Face recognition technology is of great significance for comprehensive security. Face recognition system is to study matching and identifying method for face wherein it uses computer vision and image processing technology by using facial features alone. In this paper, analysis of the face recognition and related technologies (for example LDA) are analyzed, including algorithms based on PCA, the modified PCA algorithm, and the two-dimensional PCA-based algorithm. Firstl, the preprocessing for face is introduced, its purpose is to enhance the image contrast and smoothness. Preprocessing mainly includes the gray changing, smooth handling and gray-scale transformation; and then, using the classic PCA algorithm to the image after preprocessing. By this way, sub-space with features and projection coefficients for training images in sub-space with features are obtained. Finally, a PCA-based face detection recognition system are analyzed and designed.Key words： Face Recognition feature Face ...
