Unsupervised Color and Contrast Enhancement based on the Intensity Difference between Adjacent Pixels

A computational color constancy algorithm for digital image enhancement is proposed. This algorithm imitates the inner complex behaviors of the human visual system and ends up expressing one cost function. The algorithm approaches the human visual system with mathematical optimization and reduces the number of parameters that are needed to enhance digital images compared with automatic color equalization by including an automatic thresholding called Otsu’s method. It is shown through experiments that the proposed algorithm has good performance in terms of color/lightness constancy and image contrast.