Study On Moving Targets SAR Imaging Technology

As a kind of active microwave remote sensing equipment, Synthetic Aperture Radar(SAR) could work in all-weather and day-and-night situations where the traditional optical imaging radars cannot work irregularly. And it can also obtain high-resolution imaging in azimuth direction by the “virtual long aperture” caused by the motion of the aircraft platform.Nowadays, the SAR imaging technologies aiming at ground static targets have been able to achieve high-resolution focused imaging and they have already been widely used in both military and civil fields. However, if we use those technologies to process the return signals of the ground moving targets, there will be blurs or smearing in the SAR imaging generally. Therefore, the traditional SAR imaging technologies are no longer applicable to the moving targets. This thesis conducted an in-depth research in SAR imaging of moving targets, the main content is as follows:1. This thesis studies the effects of different motion parameters of moving targets on the result of SAR imaging, and reveals the cause of moving targets’ misregistration and smearing in SAR image through theoretical derivation and analysis.2. It studies three methods based on Sinc kernel interpolation, Knab kernel interpolation and Chirp-Z transform respectively to realize Keystone Transform, and concludes the moving targets’ SAR imaging algorithm based on Keystone Transform. The results of the simulation shows that the Keystone algorithm based on Chirp-Z Transform has some advantages than those interpolations, such as smaller computation and better numerical stability.3. To solve the problem of large time-consuming and low compensation accuracy for phase error in the process of traditional high-order phase fitting, this thesis provides a modified high-order phase fitting approach which is able to remove the range curvature for multi-targets. This modified approach can improve the compensation accuracy effectively without large computation through processing the image area in segments, and the effectiveness of this approach has been verified by the simulation results. Otherwise, in order to remove the second-order and higher-order phase errors, we use PHAF to estimate the coefficients of high-order phase terms and compensate the phase errors. The simulation results show that the SAR imaging algorithm based on HAF could improve the degree of the SAR image focus.4. The classical time-frequency analysis Short-Time-Fourier-Transform has low time-frequency gathering density, and the Wigner-Ville Distribution will cause terrible cross-term interference for multi-component signal processing. In this thesis, we study in depth on another time-frequency analysis which is called S-Method. It contains the advantages of both STFT and WVD. Based on the excellent properties of S-Method, we propose a new Doppler parameter estimation approach which is combined S-Method and Radon-Mapping with two-step search. This new approach can estimate more accurate Doppler parameters by S-Method and reduce the computation significantly by two-step search. The results of simulation prove that the SAR imaging algorithm based on S-Method can generate more focused SAR images than STFT and WVD.