Hybridization of Divide-and-Conquer technique and Neural Network algorithm for better contrast enhancement in medical images

Abstract

The aim of this work is to propose a new method for optimal contrast enhancement of a medical image. The main idea is to improve the Divide-and-Conquer method to enhance the contrast, and highlight the information and details of the image, based on a new conception of the Neural Network algorithm. The Divide-and-Conquer technique is a suitable method for contrast enhancement with an efficiency that directly depends on the choice of weights in the decomposition subspaces. A new hybrid algorithm was used for the optimal selection of weights, considering the optimization of the enhancement measure (EME). To evaluate the proposed model's effectiveness, experimental results were presented showing that the proposed hybrid technique is robustly effective and produces clear and high contrast images.