摘要： 传统基于主成分分析(PCA)的人脸识别算法不能最优区分不同种类样本。为此，提出一种新的基于PCA的人脸识别算法。利用PCA降维方法提取人脸的个体差异特征，并采用最近邻距离分类器对该特征进行分类。在ORL人脸数据库上的实验结果表明，与传统算法相比，该算法的正确识别率较高。

关键词: 人脸识别, 特征提取, 个体差异, 主成分分析, 最近邻分类

Abstract: The Principal Component Analysis(PCA) is not the best method to extract features for recognition because the difference between different kinds is not considered. Aiming at this problem, a new face recognition algorithm based on PCA is proposed. It uses PCA reducing dimensions method to extract the individuality difference. A nearest neighbor classifier is employed to classify the extracted features. The method in the paper is evaluated on the ORL face image database, a series of experiments to compare the proposed approach with traditional PCA method. Experimental results demonstrate the efficacy of the algorithm.

Key words: face recognition, feature extraction, individuality difference, Principal Component Analysis(PCA), nearest neighbor classification

中图分类号: 

• TP18