On foggy days, the light reaching a camera is severely scattered by the particles in the atmosphere which arises the poor quality and low contrast of images. In viewing the shortcoming of the traditional image processing techniques which could not remove foggy weather conditions from the images completely, we present a physics-based atmospheric model that describes the appearances of scenes under bad weather conditions and then develop scene depth structure of the image from two foggy images captured under two different weather conditions from which we can calculate the scene point that would appear on a clear day with the help of Hough Transformation. The experiments of images from synthetic and real scenes which have been done in Matlab prove the effectiveness of the method in recovering contrast of images suffering from poor weather conditions.
展开▼
