A Self-Adaption Single Image Dehaze Method Based on Clarity-evaluation-function of Image

Abstract

The quality of image collected under severe weather such as fog and haze is badly damaged due to the atmospheric scattering. In order to solve the problem that image dehazing algorithms have poor adaptability, which will occur contrast distortion after restoration or cannot eliminating the influence of dense haze, a self-adaption single image dehaze method based on clarity evaluation is proposed to effectively recover the visual effects of the scene. The innovation points of this paper lied in that, first, the haze image is disposed separately according to average value, standard deviation, average gradient, information entropy and other clarity judgment features of the input image; then, the method of self-adaption image quality evaluation and coding decision is introduced to the dehaze results, to output the best effects obtained through comparison of many methods; finally, the clarity judgment is carried out again to output the final results. Experimental results demonstrate that the proposed method can achieve a better dehazing effect, and its chromaticity, luminance and contrast are improved to a certain extent. The universality of dehaze method is further improved.