A Framework for Haze removal using nearest neighbor intensity difference technique

In this paper, we present a study and review of various haze [22] removal techniques [1] for Image processing that are existing today and then we present a framework for haze removal for an image overcoming the pitfalls of the existing ones. The technique aims to remove/reduce the degradation of the image, which happened while capturing the image. The degradation can be caused by various factors like any atmospheric disturbance such as air light or environment attenuation, haze due to miss-focus of the camera or others [2]. The existing techniques for haze removal increases the contrast and recovers the color without affecting the quality, these techniques can be further classified into two categories, i.e. image restoration [9] and image enhancement techniques. These techniques are applied in various applications like outdoor surveillance and consumer electronics etc. The aim of this paper is to go through the already existing techniques for haze removal and then summarizing with their shortcomings at end of the paper and propose a new system for haze removal. Keywords— Haze removal, Attenuation, Air light, Image restoration, Image enhancement.