A Review on the Enhancement of Haze Image Techniques by Dark Channel Prior

The bad weather conditions like the presence fog, heavy rain, mist other reasons moreover time to time. Images of scenes captured in bad weather have poor contrasts and colors. This may cause difficulty in detecting the objects in the captive hazy images the general problem for imaging in the atmosphere is the introduction of fog and also the introduction of atmospheric bluster in the images. Haze is formed because of the two fundamental phenomena which might be attenuation and the air light. Attenuation decrease contrast and air light increase more light inside the region. We use color attenuation prior for de-hazing single hazy image. We break image into small sized patches. We explain atmospheric scattering model, transmission map and patch size also. We calculate depth map of hazy image. For calculating the comparison between different techniques. We explain various parameters like Mean square error (MSE), peak signal to noise ratio (PSNR), structural similarity (SSIM) and computational time which shows which technique is best. In last part of work, we explain our result and explain its shortcoming and future scope of my research process. This paper gives a review on the various haze removal techniques. First, the famous dark channel prior, a statistics of the haze-free outdoor images, can be used to estimate the thickness of the haze; and second, gradient prior law of transmission maps, which is based on dark channel prior. We proposed to implement various haze removal algorithm using color attenuation prior model. These techniques are used in many vision applications. The overall objective of this paper is explaining the different methods for efficiently eliminating the haze from digital images.