Research On Defog Algorithm Of Night Fog Image Based On Infrared Imaging Model

The infrared imaging system is based on the thermal radiation information emitted by the target object itself,so that the acquired infrared image can display the scene information with high thermal radiation amount,and the scene information with low thermal radiation amount will be annihilated in the background thermal radiation.In recent years,infrared imaging technology has been widely used in target recognition,power system detection and maintenance,outdoor traffic monitoring and other fields.However,the thermal radiation emitted by the target object is easily affected by the atmospheric medium during the transmission process,which leads to the attenuation of radiation intensity,and causes the infrared image to have low peak signal noise,poor contrast,loss of detailed information,and image distortion.Those problems greatly affected the information extraction and subsequent processing.In order to facilitate the visual observation of the human eyes and the subsequent information processing of the computer,these images need to be restored and enhanced to achieve the purpose of highlighting the goals and details and improve the visual effect.Therefore,domestic and foreign scholars have proposed many anti-fog algorithms on visible light images and have been widely used.For example,Kaiming He uses the dark channel prior anti-fog algorithm to effectively filter out fog noise in the visible light environment;Robby T.Tan uses a local maximum contrast based the image defogging algorithm to effectively highlight the texture details and structure of the defogging image;Brian Eriksson uses the curvelet transform to detect the vanishing point to realize the automatic defogging of the image,so that the scene edge information of the restored image is greatly preserved.Since the imaging principles and scenes of infrared images and visible light images are different,the defogging algorithm based on visible light images cannot be directly applied to the defoggingprocessing based on infrared foggy images.Hence,it is extremely important to study the defogging algorithm of infrared images in the night fog environment,which has certain practical significance.At present,when using mainstream image defogging algorithms to deal with infrared image noise in the night fog environment,the texture details and edge information about the image are incomplete,and it will cause the problem of noise amplification in the image.Thus when processing infrared images in the night fog environment,not only the details of the images are must be highlighted,but the influence of noise in the images are must be reduced as much as possible.Therefore,in order to solve the problems of obvious noise and non-prominent texture details in infrared images in the night fog environment,this paper deeply studies the mechanism between the current mainstream infrared camera frequency band(3-5 ?m ?8-12 ?m)and common scattering media,builds an infrared device imaging model,and proposes an infrated image defogging algorithm based on infrared imaging model;aiming at the problem that the edge sharpness of the infrared foggy image is not obvious,the guide filter algorithm is improved,and an edge weight factor is introduced into the original filter to adjust the adaptive ability of the algorithm,so as to highlight the details of the infrared image and enhance the edge of the image;for the problems of low overall contrast of the restored image and poor visual effects,the method of limiting contrast adaptive histogram equalization(CLAHE)was adopted to enhance.Finally,through the processing of a large number of measured infrared fog images,a variety of subjective and objective evaluation methods are used to perform the rendering of the infrared image defogging algorithm processed by the algorithm of this paper,the dark channel prior algorithm and the atmospheric scattering model and retinex algorithm,so as to verify the effectiveness and feasibility of the improved algorithm in this paper.