基于人工角反射器的InSAR沉降场与地形图的配准研究

Research on registration between subsidence field of InSAR and topographic map based on corner reflectors

InSAR技术可以获取高精度的沉降场,然而,一方面它与我国使用的坐标系统和投影方式不同,成果之间较难转换;另一方面,由于轨道和大气误差等影响,InSAR结果在地理编码时存在误差。上述两个原因导致InSAR结果在我国不能直接应用于生产实际,限制了InSAR技术的实用化发展。针对这一问题,本文研究一种基于人工角反射器(CR点)的InSAR沉降场与地形图的配准算法。该算法首先通过探测和识别SAR影像上CR点位置,然后对地理编码后的InSAR影像进行校正,最后利用仿射变换对InSAR沉降场进行坐标和投影转换,从而实现与地形图的配准。实验结果表明,该算法精度可靠,原理简单,易于编程实现。

We can obtain high-precision subsidence field by the technique of InSAR, while its practical application is limited by its different coordinate system and projection method used, as well as the deviations of its geocoding caused by satellite orbital and atmospheric errors. To solve these problems, we research on the registration algorithm between subsidence field of InSAR and topographic map based on corner reflectors （CR）. Firstly, we detect SAR image to identify the positions of CR points, which can be regarded as control points for registration and rectification. Then, we use a quadratic polynomial to rectify the geocoded InSAR image. Finally, the method of affine is used to convert coordinate reference frames between InSAR image and vector map. By this way, the subsidence fields of InSAR not only have accurate coordinates, but also conversion relationship between the topographic maps. The results show that, the algorithm is accurate and reliable, and the method is easy to be realized in programming.