Research On Key Technologies Of Haze Removal Of Sea-Sky Background Image Based On Full Polarization Information Detection

In recent years,with the development of photoelectric imaging technology and the increasing demand of application,the diversified demand for imaging methods is increasingly strong.Polarization imaging,as a new method of target detection,has its unique advantages over traditional imaging methods such as visible light imaging and infrared imaging.Based on the acquisition of two-dimensional intensity information of targets,polarization imaging technology increases the dimension of polarization information,including the degree of polarization,polarization direction and polarization ellipticity.The effective use of polarization information can improve the ability of aerial reconnaissance to detect and identify sea targets.The use of the polarization characteristics of scattered light,background light and target light can suppress the influence of sea clouds,solar flares and other factors on the imaging quality,improve the working distance of the camera,break through the use conditions and environmental restrictions on the camera,which has important application potential.Full polarization imaging technology can detect the circular polarization information of the target scene on the basis of linear polarization imaging.In general,the circular polarization of the natural target scene is very weak,so in non-man-made target detection,only linear polarization detection is generally used.In the sea background,both circular polarization characteristics and linear polarization characteristics have strong response.Therefore,full polarization detection is more suitable for Highlight the targets from the sea background.At the same time,the full polarization information including circular polarization has a longer action distance under the transmission environment of fog,haze,smoke and dust,so it is more suitable for the application under special weather conditions.In this paper,several key technologies for haze removal of sea-sky background mist images by using full polarization imaging detection technology are studied theoretically and experimentally.The main research contents and innovation achievements are summarized as follows:1.Based on the basic principle of polarization of light propagation,aiming at the improvement demand of haze image quality degradation in sea-sky scene,the polarization state components of sky light and field light in sea-sky scene are analyzed in depth,and the polarization bidirectional reflection distribution of calm sea surface and rough sea surface is deduced respectively.This paper analyzes and compares the technical means of linear polarization imaging and full polarization imaging,and their respective characteristics and advantages.According to the polarization characteristics of sea and air scenes,the liquid crystal modulated full polarization imaging detection technology is selected as the main technical means of polarization information acquisition and imaging.2.Based on the analysis of the modulation mechanism and the modulation accuracy of the liquid crystal modulation polarization imaging system,the influence of the noise caused by the non-ideal error on the polarization image is simulated,the nonideal polarization response model of the liquid crystal modulation polarization imaging system is established,and the non-ideal correction algorithm of the system polarization response is proposed.The optimized Stokes vector is used to reconstruct the polarization parameters.Based on the calibration of the radiation source in the laboratory and the actual scene,it is verified that the algorithm can improve the detection accuracy of the liquid crystal modulation full polarization information detection system and effectively reduce the system error.3.Aiming at the problem that the effective information of sea-sky scene is seriously weakened due to the strong scattering effect of atmospheric medium,which leads to the serious degradation of imaging quality,based on the analysis of the polarization characteristics of the scene,the polarization transmission model of sea-sky scene and the haze restoration algorithm of sea-sky background haze image based on full polarization information detection are proposed.This paper analyzes the applicability of traditional image dehazing algorithm for sea and sky scene image dehazing,and designs an image restoration algorithm for sea and sky scene polarization characteristics,which makes full use of the full polarization information to effectively estimate the sky light and scene light in the scene,so as to recover the real intensity of the scene and improve the image clarity of the scene.4.Experiments are carried out to verify the real sea-sky scene in the haze weather.Combined with the subjective visual analysis and the quantitative evaluation of objective indicators,the effect of the algorithm on improving the image information richness,contrast,detail and edge information is verified,and the effectiveness of the algorithm on the fog sea-sky scene image dehazing is proved.Furthermore,this algorithm is compared with the traditional dehaze algorithm for the sea-sky scene dehaze effect map and objective image quality evaluation index,which proves that the algorithm proposed in this paper has relatively better applicability for the sea-sky scene image dehaze restoration.In conclusion,the paper analyzes the detection accuracy of the polarization information detection system which can be applied to the sea-sky scene image dehazing,constructs the system polarization non-ideal response model,and puts forward the nonideal correction algorithm for the liquid crystal modulated polarization detection system.At the same time,aiming at the problem of image quality degradation in haze detection,a haze polarization transmission model of sea scene is proposed and a sea-sky background image dehazing algorithm based on full polarization information detection is designed.The relevant achievements of this paper provide theoretical basis and practical experience for improving the ability of remote sensing imaging to detect and identify sea area targets and have reference significance.