分块稀疏表示和2DPCA的人脸识别方法

Face recognition method based on modular sparse representation and 2DPCA

稀疏表示提出了一种分块稀疏表示和二维主成分分析(2DPCA)的人脸识别方法.该方法应用了逐像素分块的与2DPCA技术相结合的方式,充分地考虑了图像中相邻的多个像素间的相关性.实验结果表明,其中提出的新算法具有可行性以及在识别精度上的优越性.进一步的研究还表明,所提出的分块识别的方法较之于以往传统算法在存在位置偏移、单色遮挡问题的人脸图像误判率上也有显著降低.

A new face recognition method based on modular sparse representation and two-dimensional principal component analysis(2DPCA) was proposed.The segmentation of pixel and 2DPCA were combined to take full consideration of the correlation between the adjacent pixels.The experimental results demonstrated that the proposed method is feasible and effective.Further research showed that the new method reduces the false acceptance rate significantly on the images with position migration and black blocks,compared with traditional algorithms.