I. Introduction
Face recognition has been a research hotspot in pattern recognition and image processing in the past few years due to its friendliness and convenience. In recent years, many algorithms have been proposed by scholars [1] –[4]. Most previous researches can obtain satisfying recognition performance under uniform illumination conditions with frontal face images. However, it is still a very challenging research area because even the images of same person seem different due to occlusion, illumination, expression and pose variation, which can cause sharp decline in recognition rate [5, 6]. Improvement of face recognition performance in complex light environment is still a difficult problem in the field of artificial intelligence and computer vision.