用局部奇异值分解与改进LDA相结合的方法进行人脸识别

Combining Eigenface and LDA Method For Face Recognition

简要叙述人脸图像的标准化过程,特征脸,LDA算法的原理及实现过程。用PCA获取最佳描述特征,然后用LDA获取最佳分类特征,并用实验证明了这种方案的可行性。

This paper briefly discusses the process of face normalization,eigenface,the theory of LDA algorithm and the realization process. This paper gets the best description feature by utilizing the PCA and then gets the best classifying feature by utilizing the LDA. At last,this paper proves the feasibility of the method by experiment. The result is the method can get satisfying effect. The method can adapt to the changes of face and illumination.