双向二维局部保持鉴别投影应用于人脸识别

摘要

Recently, bidirectional two-dimensional Local Preserving Projection(2DLPP)is proposed for face representa-tion and recognition. Compared with two-dimensional Local Preserving Projection(2DLPP), the main idea behind bidirec-tional 2DLPP is that bidirectional 2DLPP simultaneously considering the row and column directions of images. Bidirec-tional 2DLPP needs fewer coefficients for image representation than 2DLPP which essentially works in the row direction of images. Furthermore, to enhance the classification ability of bidirectional 2DLPP, a new supervised algorithm is pro-posed:bidirectional 2DLPP plus LDA, in which images preprocessed by bidirectional 2DLPP are processed by LDA. The-oretical analyses show that bidirectional 2DLPP plus LDA algorithm has advantages over PCA+LDA, 2DLPP and bidirec-tional 2DLPP in computation complexity and memory requirements. The results of face recognition experimental on ORL and Yale face databases also show good performance of the methods with less computation.%双向二维局部保持映射（双向2DLPP）与二维局部保持映射（2DLPP）比较，双向2DLPP同时对图像的行方向和列方向进行降维处理，可以采用较少的系数有效地表示图像。为了进一步增强双向2DLPP算法的分类能力，将双向2DLPP所提取的特征采用线性判别式分析（LDA）进行分类，从而形成了一种新的监督算法：鉴别双向二维局部保持投影。理论分析表明，无论在计算量还是内存要求方面，所提鉴别双向二维局部保持投影算法比双向2DLPP和主成分分析+线性判别式分析（PCA+LDA）要少，而且在ORL 和Yale数据库上的人脸识别实验表明，新算法的识别性能比双向2DLPP和PCA+LDA算法要好，且具有较少的计算复杂度。