Classification of foggy images for vision enhancement

For vision enhancement and minimization of road accidents in bad weather, modern vehicles are often equipped with a camera connected to head-up display (HUD), which captures and displays the scene in front of a vehicle. Classification is a methodology to identify the type of fog and their optical characteristics for vision enhancement algorithms to make them more efficient. In this reported work, mean intensity value and the entropy are proposed for the classification of camera foggy images into homogeneous and heterogeneous nature of fog due to turbid weather. The use of average intensity along with entropy as new classification parameter of synthetic foggy images with different density of fog conditions which are taken as reference to classify camera images with fog. Classification parameters are suitable for synthetic as well as camera foggy images. This paper, therefore presents images with homogeneous fog have always higher mean intensity and lower entropy than heterogeneous fog. This knowledge will finally help to classify camera foggy images using mean intensity and entropy.