A face recognition method based on fuzzy data fusion is presented. In traditional principle component analysis method, operating directly on the whole face image leads to only global information about face image can be extracted and local one may be neglected. It is not very effective under variations of facial expression, pose and illumination. To solve this problem, in proposed scheme, each original image sample is divided into a certain number of sub-images and all training sub-images from the same position construct a series of new training sub-pattern sets where PCA method is used to extract local projection sub-feature vectors separately, then a set of projection sub-spaces can be obtained. To an unknown face image, after the same partition, projected sub-feature vectors of corresponding sub-space are gained. The Euclidean distances between test sub-imagespsila eigenvectors and trainingspsila are obtained to calculate their membership grade. After fuzzy classification of local projected sub-features, strategy of fuzzy synthetic is adopted to fuse each of them. At last the result of classification is determined by maximum membership principle. Simulation experiments indicate that the proposed scheme does can suitably fuse local sub-feature of face images, improve recognition rate effectively and robust.
[1]
M. Turk,et al.
Eigenfaces for Recognition
,
1991,
Journal of Cognitive Neuroscience.
[2]
Lawrence Sirovich,et al.
Application of the Karhunen-Loeve Procedure for the Characterization of Human Faces
,
1990,
IEEE Trans. Pattern Anal. Mach. Intell..
[3]
Jung-Hsien Chiang,et al.
Aggregating membership values by a Choquet-fuzzy-integral based operator
,
2000,
Fuzzy Sets Syst..
[4]
Vijayan K. Asari,et al.
An improved face recognition technique based on modular PCA approach
,
2004,
Pattern Recognit. Lett..
[5]
Jun Zhang,et al.
Pace recognition: eigenface, elastic matching, and neural nets
,
1997,
Proc. IEEE.
[6]
Yulian Zhu,et al.
Subpattern-based principle component analysis
,
2004,
Pattern Recognit..