利用主元分析识别人脸  

How to Use the Principle Component Analysis Method to Recognize Human Face

作　　者：李丽[1] 易从琴[1] 孟传良[1] 

机构地区：[1]贵州工业大学计算机科学与信息技术学院,贵州贵阳550003

出　　处：《贵州工业大学学报（自然科学版）》2003年第4期69-72,共4页Journal of Guizhou University of Technology(Natural Science Edition)

摘　　要：简要叙述了人脸识别方法,用PCA算法实现了人脸的识别,并对实验结果进行分析,认为:PCA算法的识别率不是很高,达到78.6%,还需要改进提高,但它对于光照、姿势、是否戴眼镜、脸部表情有一定的鲁棒性。This paper briefly introduces some face recognition algorithms and carefully describes how to use the principle component analysis method in a face recognition system.Experimental results show that PCA algorithm does not enjoy a high recognition rate,but the recognition rate is less affected by lighting,glasses wearing,posture or face expressions.All things considered,PCA algorithm is a more effective algorithm for face recognition.

关 键 词：主元分析 人脸识别 PCA算法 图像标准化 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]