Haze Removal in Color Images Using Hybrid Dark Channel Prior and Bilateral Filter

Abstract

Haze formation is the combination of airlight and attenuation. Attenuation decreases the contrast and airlight increases the whiteness in the scene. Atmospheric conditions created by floting particles such as fog and haze, severely degrade image quality. Removing haze from a single image of a weather-degraded scene found to be a difficult task because the haze is dependent on the unknown depth information. Haze removal algorithms become more beneficial for many vision applications. It is found that most of the existing researchers have neglected many issues; i.e. no technique is accurate for different kind of circumstances. The existing methods have neglected many issues like noise reduction and uneven illumination which will be presented in the output image of the existing haze removal algorithms. This dissertation has proposed a new haze removal technique HDCP which will integrate dark channel prior with CLAHE to remove the haze from color images and bilateral filter is used to reduce noise from images. Poor visibility not only degrades the perceptual image quality but it also affects the performance of computer vision algorithms such as surveillance system, object detection, tracking and segmentation. The proposed algorithm is designed and implemented in MATLAB. The comparison between dark channel prior and the proposed algorithm is also drawn based upon some standard parameters. The comparison has shown that the proposed algorithm has shown quite effective results.