Technology On Geometric Precision Correction Of Space-borne SAR Imagery Based On DEM

With the development of Spaceborne Synthetic Aperture Radar(SAR),the research on spaceborne SAR image is becoming more and more deeply.Synthetic Aperture Radar is widely used in SAR images because of its all-day and all-weather observation characteristics.It is necessary to know exactly the spatial position information of SAR images when SAR images are used for various monitoring and analysis statistic,as well as the fusion and registration of SAR images with optical images and geographic information systems.However,there are some geometric distortions in the SAR image due to the influence of radar sidetracking and terrain fluctuation.Therefore,before the application of SAR image,we must first carry out geometric correction on the SAR image to eliminate the various geometric deformations in the image,and finally get the image consistent with the actual geographical location.With the improvement of orbit accuracy and measurement accuracy,how to effectively suppress the influence of radar system error and terrain fluctuation on the geometric characteristics of SAR images and improve the geometric quality of SAR images become the key issues in current high-resolution spaceborne SAR image processing one.In this paper,aiming at the geometric correction model selection based on scene feature,geometric correction model improvement,and different resolution of a priori DEM effects on geometric correction and other aspects,the theoretical and experimental research are made.The main work and innovation of this thesis are as follows:In the second chapter,four kinds of spaceborne SAR image geometric correction models are studied.The accuracy and efficiency of each model are analyzed and compared,and the experimental results are validated by using multiple spaceborne SAR images.For the scene features and user requirements,Proposed the model selection decision-making plan.In the third chapter,we study the polynomial geometric correction model based on elevation and the geometric correction method of large scene SAR image based on RPC model.Firstly,the geometric correction method of the polynomial model is studied.The effect of the elevation on the azimuth and range orientation accuracy of the spaceborne SAR image geometry correction is analyzed.Then,based on the correlation between the elevation and the position error,polynomial models are built.The experimental results show that the new model has a high accuracy in the terrain.Aiming at the low accuracy of the RPC model in the whole scenes,the method of block correction is proposed,and the effect of the sub-block size on the positioning accuracy is proposed,and a sub-block partitioning strategy is proposed.The experiments show that using block RPC method can effectively improve the positioning accuracy.In chapter 4,the effect of prior resolution DEM with different resolution on geometric correction of SAR image is studied.The spaceborne SAR image simulation based on DEM and the geometric correction method of spaceborne SAR based on the simulation image correction system error are realized.The effects of different resolution DEM on the geometric correction of SAR image are simulated by using the simulated data and TerraSAR-X satellite data are analyzed.The fifth chapter is conclusion and prospect,summarize the main work of this paper,and look forward to the next step in combination with the current research and engineering application.