| The restoration algorithm for fog image and fog video is a research focusing on digital image processing and computer vision, which have been widily applied on transportation, military and other aspects. The adverse weather conditions, such as fog or haze, caused serious degradation of outdoor images and video, contrast to deteriorate, and difficulty of extracting the image feature. It seriously impacts on monitoring, navigation and other outdoor computer vision applications.Therefore, the processing of the fog-degraded images and video has been great significance.The image fog removal algorithm based dark channel prior has poor real-time performance, dark defogged image and bad effect in light area. A fast image defogging improved algorithm has been proposed. Firstly, the inaccurate positioning of atmospheric light has been improved by narrowing the scope of search area. Secondly, the guide filter optimized transmittance shorten processing time by down-sampled. Transmittance of correction can solve the color distortion in bright areas. Thirdly, for the problem of the dark defogged image, the slight mist superimposed on the clear image. Finally a user interface application is programed to implement this image fog removal algorithm. Video fog removal algorithm is an extension of image fog removal algorithms. In order to improve the efficiency of video processing and suppress changes of light, a video fog removal algorithm of the inter-frame difference method based on dynamic thresholds is proposed by combining with improved defogged image algorithms and inter-frame difference. The experimental results show that the fog removal algorithm can effectively improve the effect of fog images and video and improve the overall processing efficiency. |
PDF Full Text Request