Face recognition using fuzzy rough set and support vector machine
Shiyi Wang, Tao Liang · 2010
This paper proposes a method of face recognition using the support vector machine (SVM) based on the fuzzy rough set theory (FRST). Firstly, features from human face images are extracted by combining the 2-D wavelet decomposition technique with the grayscale integral projection technique. And then, the attribute reduction algorithm based on FRST is applied in face recognition. The reduction algorithm based on FRST can eliminate the redundant features of sample dataset and reduce the space dimension of the sample data. The proposed method avoids losing of information caused by dispersing before original rough set attribute reduction. Experimental results show that it can improve the classification accuracy in face recognition as compared with the method using the original rough set.