Image Enhancement Algorithm of Foggy Sky with Sky based on Sky Segmentation

Abstract In recent years, image defogging has become a research hotspot in the field of digital processing. Through enhancement processing, visual quality foggy images can be significantly improved, and it is also an important part subsequent To overcome limitations traditional dehazing, algorithm for with sky based on segmentation proposed. Firstly, K-means clustering feature analysis, region recognition performed fog sky. Secondly, according to pixels region, rough transmittance corrected, dehazing obtained by dark channel prior guided filter. Finally, equalized bi-histogram equalization. This effectively avoids problem color distortion halo caused area, makes restored have better global local contrast.