Illumination invariant face recognition system based on a novel Lighting Equilibrium Distribution Map

In this paper, a novel Lighting Equilibrium Distribution Map (NLED Map)is proposed to solve this problem of environmental illumination. proposed method can be utilized to compensate illumination variations effectively, especially can alleviate the side lighting effect observably. Furthermore, NLED Map is also used to establish regional confidence model in recognition. In the stage of recognition, a Regional Confidence Weight 2-Dementional Principle Component Analysis (RCW-2DPCA) algorithm is proposed in this paper to improve the recognition rate. Compared with existing image preprocessing method, our method is shown more suitable for dealing with uneven illumination in face images. Experiment results prove that our method can effectively eliminate the effect of uneven illumination and greatly improve the recognition rate compared with other popular methods.