Research on Fast Smog Free Algorithm on Single Image
Lin Zhang, Xiaoping Li, Bo Hu, Xulong Ren · 2015
In the fog and haze weather, outdoor image is easily affected by optical noise which result in the degradation of image quality and it is difficult for target detection and analysis. Researching and analyzing some characteristics of smog images, we put forward a fast smog free algorithm which can be applied for single image. The algorithm is based on fog attenuation model, estimating ambient light and airlight by using mean filter. Because the defogged images look very dim, we adopt reconcile color algorithm to restore the final brightness uniformity image. Experimental result shows that compared with other algorithms, the algorithm is simple and effective, and it can effectively reduce the running time and handle both color image and gray scale image. The experiment shows that it can be used for real-time calculation and the result has the very good visibility.