A novel improved method of RMSHE-based technique for mammography images enhancement

Abstract

Contrast improvement is one of the most important steps in medical image enhancement procedures such as mammography. In this paper, a combination of best features related to direct and indirect histogram equalization techniques is proposed in a two dimensional workspace. Using different advantages of these methods, while the proposed algorithm is able to improve the contrast and brightness of mammography images, it could decrease different effects of noises, too. On the other hand, in order to reduce undesirable effects of traditional histogram equalization techniques, an improvement of recursive mean-separate histogram equalization using a fusion of contrast-limited adaptive histogram equalization is proposed, too. Evaluation results using four effective measurement techniques e.g. peak signal-to-noise ratio, mean squared error, absolute mean brightness error and effective measure of enhancement, shows that the suggested method has significant results in contrast enhancement.