Research On High-resolution SAR Imaging Processing And Visualization Platform

Synthetic aperture radar(SAR)is a radar based on microwave remote sensing.Its principle is to synthesize a virtual aperture larger than the actual antenna aperture through the relative motion between the radar and the target,coherently accumulate the echo signal of the point target in the virtual aperture,process the data through the imaging algorithm,and finally complete a high-resolution SAR image.Because SAR earth observation has the advantages of resisting rain,snow,dust and fog and other complex meteorology,and can maintain stable work for a long time,it is widely used in military and civil fields.Combining SAR with satellite and other carriers,with the advantage of high satellite orbit,the imaging coverage of SAR is larger.On the basis of improving the imaging range,the high-resolution imaging algorithm of SAR is an important research direction at present.However,there are some errors in the existing SAR imaging geometric model,resulting in the inability to further improve the resolution.However,in the conventional high-resolution imaging algorithm,the amount of data calculation is large,the operation time is long,and the internal process of SAR imaging is invisible,so it is impossible to further optimize the quality and efficiency of the imaging algorithm.Therefore,the research on the above-mentioned problems is carried out in the following three aspects:1.An improved range Doppler-Circle Model is studied.In the range preprocessing,a processing method combining Linear Range Cell Migration Correction(LRCMC)and Keystone Transformation(KT)is adopted to completely remove the LRCM component.By analyzing the two-dimensional spatial variation characteristics of signal echo after Bulk Range Cell Migration Correction(BRCMC),a model to accurately describe the spatial variation characteristics of SAR echo under squint is established.Through experimental comparison,compared with the original oblique range model,the improved new model is more accurate in describing the azimuth spatial variation characteristics of the target beam center oblique range and the azimuth spatial variation characteristics of Doppler parameters,and the applied imaging algorithm has higher resolution.2.A method of parallel computing after SAR echo data segmentation is studied.Based on the model and algorithm in 1,this thesis analyzes the amount of computation of the optimized ENLCS(Extended Non Linear Chirp Scaling algorithm),discusses the parallelism of the algorithm,and finally gives the parallel design scheme according to the function division.The experimental results show that the method of data cutting and parallel computing can effectively improve the utilization of computer processor and reduce the time-consuming of imaging without affecting the imaging quality.3.The high-resolution SAR high-performance imaging platform is studied and developed.Based on the work in 1 and 2,an efficient and high-resolution imaging visualization platform based on single machine architecture is designed.The whole platform is divided into six modules:spaceborne SAR system mode selection and verification module,spaceborne SAR system orbit simulation module,target scene generation module,spaceborne SAR echo simulation generation module,spaceborne SAR echo imaging processing module and spaceborne SAR echo data stage processing visualization module.The platform has the advantages of low deployment and maintenance cost,supporting offline operation,protecting data security,high operation efficiency and visualization of the whole imaging process.