Research On Motion Compensation And Polarimetric Information Processing For High Resolution Synthetic Aperture Radar

As an advanced active microwave remote sensing system,synthetic aperture radar(SAR)has very broad application prospects,which can observe the earth all weather and all time.Motion compensation(MOCO)is a key procedure to obtain high-quality images for SAR applications.However,with the improvement of resolution,the performance of the widely used two step MOCO algorithm has been seriously degraded.In the case of high-resolution imaging,it is still a great challenge to correct space-variant motion errors accurately.Moreover,the development trend of high resolution and wide swath means a larger amount of data,and conventional MOCO algorithms will consume a lot of computing and storage resources,which are difficult to meet the needs of real-time processing.How to improve the real-time processing performance of approaches on the premise of ensuring the processing accuracy is an important consideration in the design of MOCO approaches.In addition,polarimetric synthetic aperture radar(Pol SAR)can acquire more precise and comprehensive scattering characteristics of targets,which has important application value in target recognition,classification and parameter inversion.Polarimetric target decomposition,as the main tool to interpret the scattering characteristics of targets,has always been the emphasis of Pol SAR research.In this dissertation,the key problems of motion compensation and polarimetric information processing for high resolution synthetic aperture radar are studied in depth.The main work is summarized as follows.1.The performance degradation of the widely used two step MOCO algorithm is studied in the case of high-resolution imaging,and an analysis is made from the perspective of range cell migration correction error.In this dissertation,the general expression of the echo signal spectrum in the presence of trajectory deviations is derived,and the range cell migration correction error and azimuth compression error caused by motion errors are analyzed.On this basis,the influence of residual range-variant error on range cell migration correction is studied,and the performances of two-step MOCO algorithm and one-step MOCO algorithm are compared.It is verified that one-step MOCO algorithm can avoid additional range cell migration correction errors and is more suitable for high-resolution imaging.Meanwhile,the one-step MOCO algorithm can well integrate with the conventional Omega-K algorithm more succinctly.2.The real-time processing performance of one-step MOCO algorithm and precise topography-and Aperture-Dependent(PTA)algorithm is improved based on subswath processing.In this dissertation,an improved multi-level space-variant MOCO approach is proposed,taking into account both precision and efficiency.By subswath processing,an approximate range envelope compensation without interpolation can be implemented efficiently,improving the efficiency of one step MOCO algorithm.In addition,the conventional processing flow of one-step MOCO algorithm is modified,guaranteeing that the phase compensation is not affected by the approximate range envelope compensation.Moreover,based on subswath processing,the range variance of residual azimuth-variant errors can also be degraded.Thus,it is no longer necessary to calculate residual azimuth-variation errors pixel by pixel,reducing the computational burden of the PTA algorithm.Compared with conventional MOCO algorithms,the proposed approach can reduce the consumption of computing and storage resources,and has high engineering application value.It provides an effective idea for real-time processing approach design.3.The MOCO algorithm for sliding spotlight mode is studied,and a real-time imaging processing approach based on coherent superposition of subaperture images is proposed.Subaperture technology is used to solve the problem of insufficient PRF in imaging processing for sliding spotlight mode.Omega-K algorithm is selected for subaperture data processing to ensure the focusing quality in the case of high resolution.According to the concept of equivalent motion error curve,continuously changing line of sight is selected for MOCO,which is more ideal than the fixed direction MOCO.This approach implements azimuth compression in each subaperture.Then,the adjacent subaperture images are registered and superposed coherently in the image domain to obtain the full-aperture image.Compared with the conventional framework of subaperture processing,the proposed approach does not require full-aperture azimuth compression and improves the real-time performance of imaging processing.4.A hybrid polarimetric target decomposition algorithm based on urban area extraction is studied,in order to improve the adaptability of the polarimetric target decomposition algorithm based on scattering model to targets with different scattering characteristics.Polarimetric coherence enhancement technique is used to distinguish urban area from natural distribution area.Different polarimetric target decomposition algorithms are used to deal with these two areas with different scattering characteristics.Wire scattering model and modified volume scattering model are introduced to improve the interpretation accuracy of scattering mechanism in urban area.Moreover,based on the analysis of scattering characteristics,the restriction conditions of negative power pixels are derived to eliminate negative power pixels,which is more in line with physical reality.