A Novel Kemelized Face Recognition System

Face recognition is a quintessential biometric technique. It still remains challenging to accurately characterize the identity related features in face images. In this paper, we propose a novel classification method based on Kernel Fisher Discriminant Analysis using the distinctiveness of Gabor features and the robustness of ordinal measures. These parameters are derived from magnitude, phase, real and imaginary responses of Gabor filtering, respectively, and then are combined as visual primitive in local regions. The statistical distribution of these primitives in face image blocks are concatenated to obtain a feature vector whose dimension is reduced using PCA and variance. Finally, each feature vector is considered as a feature input for the proposed Multi-Class KFD classifier based on RBF Kernel. The proposed method is tested on the well-known ORL face database and the Yale face database. Then, it is evaluated and compared with linear classifier (LDA) in term of classification accuracy.