Research of Face Recognition Based on SVD
Through the analysis of the singular value vector and left-right orthogonal Characteristic Matrix, this paper affirms that the singular vectors from imaging matrix decomposition can show image gray, which is related to gray distribution density; the Maximum component of the singular vectors presents the position of image levels, and other components of the singular vectors show the width of image levers. Moreover, combination the maximum singular value and the other components of the singular vectors are exactly able to ascertain image gray. The decomposed left-right orthogonal Characteristic Matrix can show structure information of picture contour. Finally, a kind of recognition algorithm based on basis space is brought forward, emulated in the ORL and ORLIC database. The conclusion affirms the validity of the analysis.
Face recognition singular value decomposition (SVD) Eigenvector Characteristic Matriz Basis Space
Song Hai-zhou Qian Ying
The Research Centre of Medical Image and Information System, College of Computer and Science Technology Chongqing University of Posts and Telecommunications
国际会议
2009图像分析与信号处理国际会议(2009 International Conference on Image Analysis and Signal Processing)
浙江台州
英文
255-259
2009-04-11（万方平台首次上网日期，不代表论文的发表时间）