Research And Application Of Single Fog-degraded Image Clearness Algorithm

The form of information dissemination mainly includes text,sound,image,etc.,about 80% of which comes from the image information received by the vision.The image has become the main carrier of information dissemination because of the more intuitive,vivid and efficient information it contains.This has led to a growing demand for image clarity from imaging devices.However,in the severe fog and haze weather,the quality of the image collected by the imaging equipment will deteriorate,and many features contained in the image will be covered or blurred,which will have a very negative impact on the subsequent image processing.This paper focuses on the single image defogging algorithm based on dark channel priori theory.Through a lot of experimental analysis,we find out the shortcomings of traditional dark channel priori theory defogging algorithm and propose improvement measures.The main works of this thesis are as follows:(1)Adaptively matches the dark channel filter window size based on the original size of the foggy image.Replacing the original fixed 15×15 size window with the designed adaptive filtering window can eliminate the interference of other light sources to the atmospheric light value A to some extent,and achieve the purpose of optimizing the defogging image quality.(2)Since the dark channel prior principle fails the highlight area,if there is a large sky area in the fog image,the sky area in the image after defogging is likely to produce color cast,and the defogging effect is not ideal.In order to solve this problem,this paper adopts the method of first segmenting the sky domain and then obtaining the atmospheric light values of the sky and non-sky regions respectively,and then obtaining the total atmospheric light value of the whole image,so as to obtain the optimized atmospheric light value.(3)By setting the dynamic tolerance value K to optimize the improved fog image transmittance,and to improve the defogging speed of the algorithm,the first mean downsampling of the foggy image is taken,and the intermediate process uses the guided filter to refine the coarse transmittance map,and then uses the double Linear interpolation restores the transmittance map to improve the defogging speed while ensuring the defogging quality.(4)The GUI tool in Matlab is used to design a single image defogging visual operation interface,which simplifies the cumbersome process of image defogging operation and greatly improves work efficiency.Subjective and objective overall evaluation of the improved dehazing algorithm and other dehazing algorithms,and applying the dark channel prior image dehazing algorithm proposed in this paper to the foggy traffic sign recognition system,effectively improving the recognition success rate of traffic signs in foggy days.Through the optimization and improvement of the traditional dark channel algorithm,the applicability of the algorithm to deal with the foggy image containing the sky region is expanded,and the image clarity after defogging is greatly improved,and the running speed of the algorithm is improved.The actual detection and application are obtained in the traffic sign recognition system.