基于模块C-2DPCA算法的人脸识别方法

提出了基于模块化完全二维主成分分析(modular C-2DPCA)算法的人脸识别方法,该方法首先对图像矩阵进行分块,将分块得到的子图像矩阵直接用于鉴别分析,由于直接基于二维子图像矩阵,能方便地降低鉴别特征的维数,在特征提取过程中可以避免使用矩阵的奇异值分解,方法简便.该方法与改进前完全二维主成分分析(C-2DPCA)方法在ORL人脸数据库上的仿真识别效果比较表明,改进后的方法在保持较高识别率的前提下鲁棒性有很大提高.

Human face recognition method based on modular C-2DPCA

A novel method,the modular complete two dimensional principal component analysis(modular C-2DPCA),is proposed for face recognition.The original images are divided into modular images and then the modular C-2DPCA method can be directly used to two dimensional sub-images obtained from the previous step,and dimension reduction of discriminant features can be done conveniently.At the same time,singular value decomposition(SVD) of the image matrix is avoided in the process of features extraction.Compared with the C-2DPCA through an examination of simulation recognition in the ORL face database,it is shown that the modular C-2DPCA method is more robust in the face recognition with a high recognition rate.