基于PCA主成分分析的人脸检测实现与分析

Human Face Recognition with PCA Method

介绍了基于PCA主成分分析的人脸检测原理,提出使用直方图均衡和切割图像的方法弱化非人脸特征信息干扰,提升主成分分析效果.并通过对35幅原始灰度人脸样本进行训练,使用Matlab实现全部分析步骤,并成功通过信噪比阈值判定完成了对样本空间外的人脸图像与非人脸图像的区分.

As a subspace projection technique within the context of a baseline system for human face recognition,the principal component analysis（PCA） is introduced in this paper.With the histogram equalization and human face normalization methods,an algorithm with a higher recognition rate is achieved by using MATLAB.In this study,35 grayscale human faces were used as sample images.