摘要： 在人脸识别中，特征提取技术被广泛应用于减少数据量和增强数据可分性。该文依据最大类间边缘准则，提出一种加权最大类间边缘准则的特征提取方法，引入加权函数，对类内和类间散布矩阵分别进行加权。并设计了一个基于离散小波分解、主成分分析和加权最大类间边缘准则的人脸识别系统。在ORL人脸库上的测试结果证实，该方法提高了识别率，最高识别率达98.25%。

关键词: 人脸识别, 特征提取, 加权最大类间边缘准则

Abstract: Feature extraction techniques are widely employed in face recognition system to reduce the dimensionality of data and to enhance the discriminatory information. In this paper, a new feature extraction method——Weighted Maximum Margin Criterion(WMMC) is proposed, which is an extension of Maximum Margin Criterion(MMC). The inter-class and between-class scatter matrix are weighted separately by a weighting function. An efficient face recognition system is designed by using Discrete Wavelet Transform(DWT), Principal Component Analysis(PCA) and WMMC. Experimental results on ORL database show that the proposed algorithm improves the performance of face recognition significantly and achieves the highest recognition rate to 98.25%.

Key words: face recognition, feature extraction, Weighted Maximum Margin Criterion(WMMC)

中图分类号: 

• TP391.41