Research On Defogging Of Traffic Video Images

Outdoor traffic video surveillance systems are always subject to many adverse weather conditions such as fog,rain,smoke,etc.These conditions degrade the color and contrast of video images,making it difficult to analyze and process information within video images.Therefore,the real-time de-atomization processing of foggy video images is of great significance for improving the accuracy of the work on traffic monitoring systems and promoting the development of intelligent transportation systems.In the current dehazing research,the more mature development is the defogging of single images.The researchers have proposed a number of algorithms and improved methods for some algorithms to restore the defogged image and improve its clarity.However,the existing single image defogging algorithm is more complicated.If the existing defogging algorithm is used to recover,the calculation takes a long time,and it is difficult to satisfy the video.The real-time processing may also cause the video image to play discontinuously.In order to improve the real-time performance of video processing,it is necessary to find a more effective method of defogging.For these issues,the main work of this paper is as follows:Firstly,the moving vehicles are detected.In the traffic video,the image can be divided into two parts: the background and the foreground.When the position of the shooting lens is fixed,the background content is fixed,and the background of the foggy day can be completely replaced by the background when the fog is sunny.The prospect is the focus of the traffic video surveillance system,and it is necessary to focus on the defogging process,so this paper extracts the moving target.By comparing three typical target tracking algorithms,the RPCA algorithm is used to complete the motion vehicle detection.Secondly,the moving vehicle is defogged.After accurately detecting the target area,in order to make the target clearer,it is necessary to perform defogging treatment firstly.Through the analysis and comparison of the existing defogging algorithm,the dark primary color prior algorithm of the guide filter is used to refine the transmission map to defogg the target area.Because the calculation of the large background is neglected,the calculation amount is greatly reduced,which is beneficial to improve the real-time performance of video processing.Thirdly,the goal after defogging is clearer.If only the initial defogging operation is performed on the detected target,the improvement in sharpness is limited.It has been observedthat traffic video varies little between frames with only small displacements between moving vehicles.According to this feature,the target can be reconstructed by using the associated information between adjacent frames.Image super-resolution reconstruction technology is used to fuse the information contained in each frame of the video to obtain a clearer image.In the image super-resolution technology,the reconstruction reference frame is obtained by interpolation technique.And the displacement generated by the motion is calculated between each frame image and the reference frame,and then the registration is performed,and finally the target image is reconstructed by using the convex projection method.After data evaluation,using this method effectively improves the quality of the image.The experimental results show that the method can not only improve the computational efficiency and real-time performance of video images,but also significantly enhance the defogging effect and clarity of moving targets.And this system lay the foundation for the next step of extracting moving target information.