IMPLEMENTATION AND COMPARISON OF DIFFERENT IMAGE CONTRAST ENHANCEMENT TECHNIQUES

Abstract

Enhancement has a very important role in computer vision, pattern recognition and digital image processing (DIP). It is one of the most important and difficult techniques in image research [1]. The image enhancement aims at improving the visual looks of the image, or it may work on the better representation of the image so that it can be used for fi.iture auto mated image processing. Noise and poor contrast are two major factors which affect a variety of images such as aerial images, satellite images and also the images used in medical field. in addition to this the real life photographs are also affected by these two i.e. noise and poor contrast. Thus it becomes very important and also necessary to suppress these two issues and work on removing the noise from affected images, thus making noise free and enhancing the contrast of the poor contrast images. in this way we surely can enhance the quality of such images. Image enhancement plays the role an important stage for detection and analysis of' medical, satellite and aerial images. There are many image processing - operations, where image enhancement plays a vital role. Image Enhancement techniques improve the quality (clarity) of images for many applications such as highlighting the details, - removing blurring, increasing contrast, removing noise, human viewing etc. The enhancement technique differs from one field to another according to its objective. in this dissertation work various state-of-art contrast enhancement techniques are presented. Six different contrast enhancement techniques are implemented on MATLAB platform. A variety of test images (Natural, Medical, Aerial etc.) are used to procure the enhancement results in an exhaustive way. Experimental results demonstrate the features of these methods. A subjective analysis is carried out on the enhanced outputs. Observations are given for enhanced images. Objective analysis is also carried out by taking 5 different performance measuring parameters. A comparison is presented based upon the subjective and objective analysis