主成成份分析( PCA) 方法是人脸识别技术中常用的一种一维特征抽取方法。传统PCA 方法用于人脸识别常常面临图像维数高,直接计算量的问题。为了解决这2 个问题,人们对PCA 进行了改进,提出并实现了多种基于PCA 的人脸识别。对3 种基于PCA 的人脸识别方法做了理论上的研究和实验上的性能比较。实验结果表明PCA + 2DPCA 是其中综合效果最好的一种方法。-Principal component analysis into (PCA) is a commonly used face recognition feature extraction method of one-dimensional. The traditional PCA method for face recognition are often faced with images of high dimensionality, direct calculation of the problem. To solve this two problems, one of the PCA has been improved, proposed and implemented a variety of PCA-based Face Recognition. On 3 Face Recognition Based on PCA do theoretical research and experiments on the performance comparison. Experimental results show that the PCA+ 2DPCA combined effect is one of the best methods.