CONTRAST ENHANCEMENT TECHNIQUES FOR FOGGY VIDEOS

One of the major causes for vehicle accidents on the road is the presence of fog. Driving a vehicle while foggy road can create havoc. The presence of fog disturbs visual sight, because of lights scattering by the airborne water droplets. As we have seen the density of the fog very low while it comes a camera and hence the low density creates nonlinear increase with the distance. Due to this objects are not clear for large distance. In this paper, the contrast enhancement techniques for fog videos are proposed, particularly videos are captured in fog conditions. An algorithm uses a mathematical model for calculating the atmospheric veil and variation in fog density with the distance. By estimating this atmospheric veil using various constraints we can reconstruct the original fog free videos, which gives better results when compared with Dark channel prior method. A qualitative comparison is performed on various videos. The proposed algorithms are suitable for both color and grayscale videos.