Illumination Compensation Algorithm for Face Images Based on Combination of Several Algorithms
多方法结合的人脸图像光照补偿算法

Combining the illumination preprocessing algorithm based on Retinex theory with two common preprocessing algorithms, which are the histogram equalization and the gamma,gray-transformation, will result in two new illumination preprocessing algorithms. According to experimental findings, the resulting algorithms significantly improve side-lighting correction. This improvement can considerably increase face detection rate under complex photo environment. At the same time, as the new methods are based on two simple preprocessing algorithms, the preprocessing efficiency is ensured.