Image Contrast Enhancement Using Weighted Uniform Histogram for Maximum Information

Contrast enhancement is a fundamental and important process in image based applications, which often require the image to provide the viewer with the maximum amount of details from the captured scene. This objective can be accomplished by maximizing the information content of the image. Instead of modifying the image histogram as in most equalization methods, we propose to firstly process the image intensity to follow an exact uniform distribution for information maximization. Then this intermediate image is combined with the original input in a weighted sum to preserve the features. A resultant image of high information content, while maintaining the original features, can then be produced. Experimental results had shown the effectiveness of our developed method with regard to maximizing the information content.

I. Introduction

Imagery and computer vision have constituted an indispensable modern technology that has wide applications. Examples include industrial robot operations [1], machine wear condition monitoring [2], remote sensing of the environment [3], medical imaging [4], and object tracking [5]. All these applications demand an input image of high information content in order to fulfill the problem-specific requirement.