Gray-scale image enhancement using differential evolution optimization algorithm

Abstract

Differential Evolution algorithm (DE) is a search method that iteratively searches for the solutions of machine learning and engineering problems that involve optimization. This paper aims at displaying the effectiveness and adaptability of the DE algorithm to find the global optimal solutions iteratively for contrast enhancement of grayscale images. An image's contrast can be modified by gray-level adjustments to the pixel intensities of the original image with the help of a parameterized intensity transformation function. A quality function is used to judge the quality of the enhanced images which incorporates various conditions of image enhancement and is used as the fitness criterion. The Differential Evolution algorithm aims at maximizing the fitness function through adjustments to variables of the pixel intensity transformation function.