Impact of Atmospheric Light on Haze Removal using Single Image

Fog or haze are the common phenomenon which creates unpleasant weather condition. This unpleasant weather condition is not favorable for outdoor photography or imaging, because image characteristics are drastically deteriorated by fog or haze. To enhance the image features, fog or haze removal is required. Fog removal is also required in different applications like recognition, tracking and navigation and consumer electronics. During the fog removal process, estimation of atmospheric light plays main role. In this paper we performed an analysis on an estimation of atmospheric light by taking 0.01%, 0.02% and 0.05% most haze-opaque pixels from foggy image and its effect on results. To perform analysis, we used dark channel prior concept with guided filter for fog or haze removal.