Research Of Detecting Moving Targets In SAR Imagery By Phase Error Estimation

Most target motions induce phase errors in the SAR signal history. These phaseerrors cause smearing in the image. If we can measure these errors we couldfind the moving target.In the preface to the article, the development of GMTI technique in the fieldof SAR are introduced briefly. And the principal contribution of thisdissertation is summarized here.Chapter 2 analyzes the characteristic of SAR signal of moving targetstheoretically. The simulation data and real SAR data validate the result.Chapter 3 introduces two different phase error estimation algorithm: shearaveraging algorithm and PGA (Phase Gradient Autofocus) algorithm.In chapter 4, we discussed the method which involves segmenting acomplex-valued SAR image into patches, focusing each patch separately in twoautofocus ways, PGA and Shear Averaging, and measuring the sharpness increasein the focused patches. Then judge if there're moving targets through thesharpness increase. At last we give out the result of plenty of simulationdata and real SAR data. And we also analyze the performance of the algorithmfrom the point of view of different velocity, different signal/noise energyratio, different aotufocus algorithm, different patch size.In chapter 5 we discuss how to realize the shear averaging algorithm on theADSP21060-SHARC parallel DSP processing system.Chapter 6 is the conclusion. And we also put forward some problems which needto be solved.