结合MMC与BPSO优化的人脸特征提取

为了提高最大间距准则法表征人脸特征空间的能力,提出了一种融合最大间距准则和二进制粒子群优化算法的人脸识别方法。利用离散二进制粒子群算法对最大间距准则变换后的特征向量进行选择优化,获得有利于分类的最优特征子空间。在ORL和Yale标准人脸库上的实验结果表明,该方法不但降低了特征空间的维数,而且更好的发挥了最大间距准则算法的优点,提高了人脸识别的速度和精度。

Face Feature Extraction Combined MMC with BPSO

In order to improve the ability of Maximum Margin Criterion (MMC),this paper proposes a face recognition method which fusion the MMC and binary particle pwarm optimization algorithm(BPSO).Then BPSO is used in feature selection after the transformation of MMC,which can find out feature subspace that is beneficial to classification.Experimental results on ORL face database and Yale face database show that the proposed method not only reduces the dimensions of face feature space ,but also expresses the advantages of MMC, improves the speed and accuracy of face recognition.