Research On Clearing Method Of Sea Fog Video Image Based On Gradient Domain Guided Filtering

With the development of China’s shipping industry,on the premise of safe production,it is crucial to improve its management and technical level through information and intelligent means.CCTV video,as an important tool in water traffic safety management,can increase maritime supervision and improve work efficiency.Due to the effect of the natural environment,the sea area of China has been affected by sea fog of different concentrations for a long time,which has reduced the visibility of the sea,severely reduced the monitoring quality of closed-circuit surveillance video,created hidden dangers of water traffic safety,and reduced the navigation efficiency of ports.However,most CCTV video clearing algorithms are designed for inland.Some algorithms have limitations for sea fog images and videos with single color,dark overall brightness and complex background.Therefore,this paper will carry out the research on the defogging algorithm in order to avoid the impact of poor visibility and algorithm limitations,and to obtain clear closed-circuit surveillance video.First,by analyzing the color histogram,information entropy and edge intensity of the sea fog image,the characteristics of the sea fog image are quantified.It is found that the improved sea fog image and video clearing algorithm can enrich the color of a single frame image and improve the sky area the phenomenon of color distortion,increase the content contained in a single frame image and enhance the details of the information.Secondly,the research of video defogging algorithm is based on the research of single image defogging algorithm.In this paper,it is found through experiments that the gradient domain guided filter is more suitable for processing sea fog images,and the dark channel with color restoration based on gradient domain guided filtering Retinex(Multi-Scale Retinex with Color Restoration,MSRCR)algorithm for single frame image dehazing is proposed.The key of this algorithm lies in two places.One is to use the image denoising function of gradient domain-oriented filtering to refine the transmittance of the image and improve the dark channel prior algorithm to obtain the restored image.Second,in the space of hexagonal cone(Hue,Saturation,Value,HSV),the gradient domain steering filter is used again to enhance the image detail’s ability to estimate the illumination component.Then the adaptive linear stretching is performed on the reflection component to enhance the saturation of the image to obtain the final fog-free image.The experimental results show that the dark channel MSRCR sea fog image defogging algorithm based on gradient domain-oriented filtering can enrich the color of the image,improve the brightness of the image,enhance the details of the image,and make the image clearer,while avoiding color distortion in the sky area.Finally,based on the improved sea fog image defogging algorithm,this paper proposes a video defogging algorithm based on gradient domain guided filtering to estimate the optical path propagation map.The key of this algorithm is for the video segment with little scene change,the dark channel MSRCR algorithm of the gradient domain-oriented filtering is used to process the first frame image of the video to obtain its fog-free image and "universal" optical path propagation map.Guided filtering processes grayscale images and "universal" optical path propagation maps of subsequent frames to obtain subsequent fog-free frames.Experiments show that the algorithm not only realizes the function of the above image defogging algorithm,but also avoids the phenomenon of color mutation in the video after defogging.