The Research On Image Defogging Algorithm Based On Dark Channel Prior

Blurred images acquired under foggy conditions usually show the same characteristics: the overall image contrast decreases,the color shifts in the scene and different degrees of distortion and detailed information of some areas(especially the edge)are lost.These low-quality images will bring inconvenience to the application of outdoor vision system,resulting in the inability of computer system to carry out post-processing of images.In order to reduce the adverse effects caused by foggy images on the outdoor vision system,this paper conducts an in-depth study of the defogging method based on the Dark Channel Prior(DCP)theory,and proposes improvement measures for the deficiencies of the algorithm.The specific contents are as follows:(1)Aiming at the limitations of the dark channel map obtained based on the dark channel prior theory and the block effect phenomenon at the edges of objects in the image after defogging,an adaptive window filtering and double-channel method for obtaining atmospheric light value is proposed.Firstly,the adaptive window filter is used to obtain the dark channel graph of the image,which lays a foundation for the subsequent calculation of atmospheric light value and transmission map.Secondly,the method of linearly weighting the atmospheric light value with double-channel graph is used to reduce the influence of white objects on the atmospheric light value and improve the accuracy of estimation.Finally,the median filter is used to optimize the transmission map to reduce the color distortion and eliminate the block effect.(2)Aiming at the failure ofthe dark channel prior theory in bright areas such as the sky,a dark channel prior defogging algorithm based on double thresholdssegmentation is proposed.Firstly,the sky region and the non-sky region are segmented by double threshold segmentation method.Then,the quadtree algorithm is used to perform accurate block search on the area where the atmospheric light value exists,and the calculation method of the value is optimized and improved;Secondly,the transmittance result is modified of the sky region,the transmittances of two regions are merged and the the guide filtering algorithm is used to refines the fused transmittance;Finally,gamma correction is used to enhance the brightness of the defogging image to improve the image clarity and visual effect.(3)The comparison and simulation experiments in this paper and several classic defogging algorithms on foggy images with different characteristics are carried out respectively.Experimental results show that,compared with several other algorithms,the proposed algorithm can effectively improve the contrast and sharpness of the original image,while effectively eliminating the phenomenon such as blockiness,halo effect,color shift and distortion caused by the traditional methods of defogging.