Single Image Defogging Method Based on Adaptive Modified Dark Channel Value
Peiling Li, Feng Wang, Yitao Liang, Xiao Zhang · 2019
In the conventional image defogging algorithm based on dark channel prior, the calculation deviation of dark channel value of image edge position could lead to halo effect on image.In this paper, a method for correcting dark channel value is proposed.Firstly, according to the gray difference between the minimum image and dark channel image, an adaptive threshold is obtained to determine the edge image pixels with drastic depth gray value changes.Secondly, with the threshold, the gray values are actively modified.Finally, clear fog-free image can be restored according to the atmospheric scattering model.The test results of several groups of foggy images show that after the application of the method, the halo effect effectively removed, the image color deviation corrected, and the contrast improved.The restored image looks clearer and more natural.