An application of Gabor based Kernel Fisher Discriminants to the online palmprint verification system

This paper presents an application of Gabor based Kernel Fisher Discriminant (KFD) palmprint feature extraction method to the online palmprint verification system. In this system, palmprint area as region of interest is selected after detecting a hand object in the image sequences given by a CCD camera. For palmprint representation, firstly, Gabor wavelets which gives robustness to the local distortions are performed. Then, with the usage of nonlinear KFD method most discriminative features are selected. Finally, mapping is done by using a weighted Euclidean distance. Because of proposed method is applied in the real environments the studies on the performance and optimality have also been achieved. The experimental results show that the proposed method gives the better biometric accuaricies on the feasibility of the online palmprint system.