Color image enhancement using evolutionary principles and the Retinex theory of color constancy

The paper investigates the use of Evolutionary Computation (EC) to automate the process of color image enhancement, for a large category of color images taken with different illuminants. Our approach employs the model of Multi-Scale Retinex with Color Restoration (MSRCR) for which the parameters of the model are adapted to each given image with a Real-Coded Genetic Algorithm (GA). Results on a diverse set of images show the efficiency of our approach compared to other classical enhancement methods.