Research On Imaging Characteristics And Image Restoration Method Of Large Field Circular Scanning System

Large field circular scanning system adopts the rotating imaging mode,which is conducive to the realization of large aperture and high-resolution imaging.It has the advantages of light weight,high phase coherence and large coverage field of view.It is one of the most potential ways of large aperture remote sensing imaging technology in geostationary orbit.However,due to the special rotating push broom imaging method,it is easy to be affected by the coupling factors such as detector rotation sampling,satellite platform vibration,optical system aberration and so on.The degradation characteristics are complex,and there are image quality problems such as non-uniform blur and annular geometric distortion.Therefore,in order to improve the application ability of image products,it is necessary to restore the image according to the imaging characteristics of the system.But there are few reports on the imaging characteristics of the above-mentioned systems and the corresponding image restoration methods at home and abroad.In order to solve this problem,this paper studies the imaging characteristics and image restoration methods of the large field circular scanning system.The main research work is as follows:(1)The on-orbit imaging characteristics of large field circular scanning system are analyzed.Based on the full-link imaging mechanism of large field circular scanning system,the coupling model of large field circular scanning system imaging simulation is established by considering the elements of detector rotation sampling,satellite platform vibration,optical system aberration and the coupling relationship among them.On this basis,simulation experiments are carried out to analyze the characteristics of the system,and the influence of platform amplitude,vibration frequency,detector rotation rate,TDI series and other key parameters on the image quality is analyzed,which provides a theoretical basis and data set input for the follow-up research of image restoration methods.(2)The image restoration model and method of large field circular scanning system are proposed.Aiming at the annular space variant characteristics of the system image,this paper proposes a fan-shaped segmentation method of space variant image.In order to solve the image quality degradation problems such as geometric distortion and non-uniform blur,the non-uniform sampling matrix and system degradation matrix are constructed respectively,and the image regularization restoration model based on gradient domain is established.The weighted total variation regularization term is introduced to improve the restoration ability of the model for edge and detail information in all directions.Finally,based on the alternating direction multiplier method,a fast solution to the model is proposed.(3)The experimental verification and analysis are completed.In order to accurately evaluate the performance of the image restoration model,the system imaging simulation model is used to generate the simulation data set of large field circular scanning system,and the experimental scheme of image restoration method is designed.The correctness and effectiveness of the restoration method are verified by comparing the data set before and after restoration.At the same time,comparative experiments and applicability analysis are carried out to analyze the image restoration performance under different scenes and different imaging parameters.The experimental results show that the image restoration method can effectively solve the image quality degradation problems such as geometric distortion and non-uniform blur in the large field of view circular scanning system,so as to provide support for the future space application of the system.