主元分析变换空间上的鉴别共同矢量人脸识别方法

Discriminant Common Vectors in Principal Component Analysis Transformed Space for Face Recognition

给出了在主元分析(PCA)变换空间上求取DCV投影矩阵的方法(PCA+DCV),在保留所有鉴别信息的条件下,显著降低了算法复杂度,提高了运算效率。进一步提出了依据主元成分对应特征值进行适度权重的DCV识别方法(WPCA+DCV),一定程度上减小因光照、饰物遮挡等造成的面部变化带来的识别影响,增强表征信息,提高识别率。在ORL、YALE和AR人脸库上的实验结果证实了本方法的性能。

In this paper, we first offer an efficient algorithm to perform DCV in principal component analysis （PCA） transformed space. In this way, we reduce the algorithm＇s complexity and improve the efficiency whilst preserving the whole discriminant information. Then, the new algorithm further facilitates us to subtly weight the facial components in PCA space according to corresponding eigenvalues, which is potential to enrich the representative information and thus improves DCV＇s recognition performance. The experiments conducted on ORL, YALE and AR face database demonstrate the effectiveness of our method.