Fog Visibility Detection of Highway Based on Improved Dark Channel Prior Algorithm

The low visibility near the ground in foggy weather can easily lead to vicious traffic accidents, so research on visibility detection in foggy weather is fundamental for traffic control. This paper presents an algorithm for detecting fog visibility in expressway based on improved dark channel prior method, which transforms the difficult problem of detecting visibility through image processing into the problem of calculating atmospheric extinction coefficient. Firstly, Canny operator is used to extract the image contour and divide the sky region. The average value of all pixels in the sky region is regarded as the atmospheric light value, thus increasing the accuracy of atmospheric light estimation. Then, guided filtering is used to refine the atmospheric transmittance. Finally, the extinction coefficient of the image can be calculated according to the lane line, so that the visibility value of the image can be calculated. In the experiment, the video image of a section of Rylan Expressway under fog is used. Experimental results show that the relative error of the improved algorithm is obviously reduced.