Single hazy image restoration based on fields of experts model and guided filtering

A kind of haze removal algorithm is proposed, so as to improve the visibility of degraded image with only single image. Firstly, the raw atmospheric transmission map is obtained using the dark channel prior. Then, Fields of Experts model with guided filtering is integrated to get the priors from the degraded image, so as to amend the raw atmosphere transmission map. Finally, the scene albedo is restored based on the atmosphere scattering model. Results on a variety of outdoor haze images show the feasibility and validity of our algorithm. Moreover, a comparative study is also proposed with several other state-of-the-art algorithms demonstrate that our algorithm has similar or even better haze removal results.