A Robust Illumination Normalization Method Based on Mean Estimation for Face Recognition

Abstract

An illumination normalization method for face recognition has been developed since it was difficult to control lighting conditions efficiently in the practical applications. Considering that the irradiation light is of little variation in a certain area, a mean estimation method is used to simulate the illumination component of a face image. Illumination component is removed by subtracting the mean estimation from the original image. In order to highlight face texture features and suppress the impact of adjacent domains, a ratio of the quotient image and its modulus mean value is obtained. The exponent result of the ratio is closely approximate to a relative reflection component. Since the gray value of facial organs is less than that of the facial skin, postprocessing is applied to the images in order to highlight facial texture for face recognition. Experiments show that the performance by using the proposed method is superior to that of state of the arts.

Addressing the illumination challenge in two‐dimensional face recognition: a survey

“…From the UFE methods, multi‐OGPF [86], wGAP [93] and NDF [83] provided 100% accuracy. From the UP group, mean‐LDA [54], DD‐DTCWT [78] and fuzz + LFDCT [60] had the same results. In the same way, the 9‐basis [38] relighting approach had a full recognition result.…”

“…Wang et al [53] decomposed the image into two new images, and then applied a minimum‐and‐maximum filter to the high‐frequency part and a shadow removal to the low‐frequency part. Luo et al [54] suggested the use of a mean estimation method to simulate the illumination component of the image. Then, a post‐processing step is performed to reduce noise and highlight facial texture features.…”