A Quasi-automatic Rectification Method of SAR Image Based on Image Simulation
基于影像模拟的SAR几何校正准自动方法

SAR image usually becomes very difficult when precise orbit data of SAR platform is not available. A SAR rectification method based on image simulation is proposed in this paper. The simulated SAR image is generated by means of Digital Elevation Model (DEM) and the known orbit data, then image matching is performed between the simulated and the real SAR image to identify some pairs of corresponding points. The position differences of these corresponding points are caused by the imprecision of the orbit data, and can be used to correct the rectification residuals. In this paper, the common SAR image simulation algorithms are introduced briefly, and then the algorithm used to match real to the simulated image is presented, as well as the SAR rectification method based on simulated image. At last, an experiment on the proposed rectification method is made, in which a RADATSAR SGF image is rectified precisely by means of simulated SAR image. Moreover, except the image matching stage, all the procedures of this rectification can be done automatically, which is significant to the large-volume SAR image processing.