Research On Clear Algorithm Of Foggy Color Video Image

Currently, the digital image processing technology has become a hot research in computer science and iconology, which has been widely applied in the field of road monitoring, navigation etc. The fog image enhancement and restoration technology is one of the key research fields in digital image; its main propuse is to facilitate the computer and human understanding of image effectively. This paper analyzes the current technology in the field of image enhancement, and for the Retinex algorithm and dark channel theory, carry out detailed research.According to the Retinex algorithm when processing the edge region of contrasting, it appears halo phenomenon. The scale is based on several image sizes which are fixed up, so it can not be selected in processing, and the threshold value also can not be freely determined. From the characteristics of the processing fog image, it will appear over-enhancement in the contrast part. This paper parents four innovative improvement algorithms, and has achieved good results.For the dark channel theory, when it processes the color regions of sudden change, there will be significantly lower brightness and color distortion, halo phenomenon. This paper gives six innovative improvement algorithms, including the improvement of the complexity and the real-time of the algorithm.The fog image contains lots of noisy; this paper analyzes the causes of noisy, while constantly to be tested. For the additive noise, the Retinex algorithm and its improved algorithm have a better effect. But for the multiplicative noise, the dark channel theory and its improved algorithm have a better effect.Finally, in view of the assessssment of the algorithm, from lots of experiments, through the subjective observation and objective evaluation, this paper, the fog video image processing, is better than the traditional classical algorithm which is included the center around Retinex algorithm and dark channel theory in the overall and details.