An efficient face recognition for variant illumination condition

Abstract

The aim of this paper is to investigate how to preprocess the input face image for the task of robust face recognition, especially in changing illumination environment. Changing illumination poses a most challenging problem in face recognition. Previous research for illumination compensation has been investigated. We found that the performance of each preprocessing method for compensating illumination is highly affected by the working illumination environment. This paper proposes an adaptive filter block for efficient face recognition. Since no priori knowledge of the system working environment can be assumed, illumination environments are analyzed by an unsupervised learning method, fuzzy ART. The proposed method can decide an optimal configuration of filter block by exploring the filter combination and the associated parameters to unknown illumination conditions. The filter block includes retinex filter, histogram equalization filter. The proposed method has been tested in robust face recognition in varying illumination conditions. Extensive experiments show that the proposed system can achieve very encouraging performance in varying illumination environments.