考虑多普勒参数的SAR影像距离—共面方程及其定位

Range- Cocone equation with Doppler parameter for SAR imagery positioning

通过对合成孔径雷达(SAR)影像成像几何机理的分析,以传感器位置和姿态作为定向参数,构建了考虑多普勒参数的SAR影像距离—共面方程及基于该方程的SAR影像空中三角测量模型。以登封地区99景0.5 m分辨率机载非0多普勒SAR影像为试验数据,进行了单片定向、无控制点的自由网平差、稀少控制点的区域网平差及带有自检校参数的区域网平差等系列试验。与R-D模型相比,本文模型减少了定向参数的数目,实现了SAR影像与POS数据的联合平差处理,建立的带有自检校参数的SAR影像区域网平差模型,对提高稀少或无地面控制点条件下未进行(或低精度)几何定标SAR影像的区域网平差精度起到了显著作用。较少控制点条件下模型满足了丘陵和山区1∶5000/1∶10000航摄测图对空中三角测量精度的要求。

In this paper we analyze the geometry mechanism of Synthetical Aperture Radar （SAR） imaging, and propose the Range-Cocone equation with Doppler parameter （RCD） for SAR imagery positioning taking linear and angular elements as orientation parameters. A serial of tests are taken that include orientation with single image, free-net adjustment without Ground Control Point （GCP）, block adjustment with sparse GCPs, and block adjustment with self-calibration parameters. The test datasets covered a part of Dengfeng suburb are 99 airborne SAR images with 0.5 m resolution and non-zero doppler parameter. Compared to R-D model, the presented RCD model decreases the number of orientation parameters and is convenient for adjustment combining SAR imagery with POS data. RCD based block adjustment model with self-calibration parameters can increase the accuracy greatly in the case where SAR sensor is not calibrated, or calibrated coarsely Tests also indicate that the airborne SAR imagery and the proposed positioning model can meet the accuracy requirement for aerial triangulation at the scale of 1∶5000/1∶10000 over hill and mountainous areas with sparse GCPs.