反投影(BP) 算法是一类经典的合成孔径雷达(SAR)成像处理方法。BP在已知航迹条件下可以实现数据的高精度聚焦。但是，BP重建图像中，运动误差导致的目标散焦通常沿着不同的方向存在，残留距离徙动不能严格限制在一个距离分辨率单元内，运动补偿困难。文中提出了一种修正投影栅格的子带宽相位梯度自聚焦(PGA)处理方案。其中，基于数据采集平面的非均匀栅格重建图像，完全去除了目标散焦方向的空变特性。然后，对子带宽分解的反投影数据进行PGA运动补偿并拼接得到全分辨率图像。最后，点目标仿真验证了该方法的有效性。

Backprojection(BP) algorithm is one classical imaging processing methodology in synthetic aperture radar(SAR), which can focus data with high accuracy when the flight path is exactly known. However, defocus due to motion error in BP imagery usually along variant directions and residual range cell migration can not be restricted in one range resolution. So the motion compensation of BP imagery is difficult to implement and can not be realized by traditional autofocus methods. To solve this problem, this paper proposes a sub bandwidth phase gradient autofocus method based on the modified backprojection grid. The imagery is reconstructed with a modified non-uniform gird along the data collection plane in order that the space variance of directions of defocus can be eliminated completely. Then PGA processing and combination are employed successively on the decomposed sub bandwidth backprojection data to obtain the refocused full resolution imagery. Finally, point target simulation validates this method.

TN957

江苏省高校自然科学研究项目资助;中央高校基本科研业务费专项资金资助项目;毫米波国家重点实验室开放课题资助项目;雷达成像与微波光子技术教育部重点实验室开放课题