稀疏轨道条件下SAR几何校正轨道拟合策略

Strategy of Orbit Fitting with Sparse Trajectory Data in SAR Geometry Correction

针对SAR影像头文件中提供的轨道矢量分布较为稀疏时,使用低阶多项式拟合轨道参数会存在较大误差,从而导致基于距离-多普勒模型的星载SAR影像几何校正精度较差的问题,提出了一种稀疏轨道条件下实用的轨道拟合策略。通过对轨道拟合误差进行分析,综合考虑距离-多普勒模型的解算与现有几何校正资源,选择使用高阶插值方法生成卫星成像时刻处足够的加密点轨道状态矢量,再对局部的轨道进行二阶多项式拟合,利用拟合的轨道模型系数进行几何校正。利用实测GPS点数据进行了SAR几何校正精度验证,结果表明该策略能够明显改善ALOS2PALSAR影像几何校正精度,证明了该方法的有效性和良好的适用性。

There are errors in orbit parameters fitted by low-order polynomial while the vectors of satellite trajectory provided in the image header file are sparse,resulting in poor accuracy of the geometric correction of spaceborne SAR images based on the Range-Doppler model.This paper presents a practical orbit fitting strategy when trajectory data is sparse.By analyzing the errors of fitting orbit and considering the requirement and program to solve the Range-Doppler model,we use high-order interpolation method to generate enough quantity of orbit vectors at satellite imaging time ,firstly.Then we use the second-order polynomial to fit local satellite orbit with the interpolated vector data.Finally,geometric correction is performed with the fitted second order orbit model.The result of experiments in this paper has been verified by using the measured data. The result demonstrates that the method is useful to improve the accuracy of the geometric correction of the ALOS2PALSAR image.It illustrates that the method presented in this paper is valid and applicable.