星载高分辨率雷达摄影测量与干涉测量联合提取地形信息方法研究

As two major approaches to use SAR remote sensing data for topographic mapping, both StereoSAR and InSAR techniques have their own advantages and disadvantages. StereoSAR can be used with less constrains than InSAR, however, its accuracy in height measurement is usually worse than that of the later due to difficulty in stereo image matching over terrain relief areas. In contrast, although InSAR can achieve higher accuracy of height measurement, it is often disturbed by decorrelation effects as well as problems of integer cycle uncertainty and systematic errors for which GCPs are needed to perform calibrations. Consequently, for topographic mapping in terrain relief areas it is not easy to obtain accurate and reliable height information through the use of only either technique. Aiming at such a problem, in this project we are going to carry out a study on joint utilization of StereoSAR and InSAR for topographic mapping. The major goal is to make full use of advantages of the two complementary techniques to substantially improve our capability of height measurement in areas of rough terrain. Key problems to be solved include stereo/interfeometric image matching, atmospheric phase screen correction, phase unwrapping and calibration, DEM fusion, etc. The processing chain developed for high-resolution spaceborne SAR datasets by this project is expected to be able to act as an efficient tool for large-area topographic mapping in Western China.

作为雷达遥感提取地表高程信息的两种主流技术，雷达摄影测量和干涉测量具有各自不同的优缺点。摄影测量应用限制条件少，但由于立体像对匹配难度大，高程测量精度通常不如干涉测量；干涉测量虽然高程测量精度高，但易受到去相干效应影响，同时存在整周模糊度和系统误差问题，需借助地面控制点进行定标。因此在困难地区地形测绘中单独采用一种技术往往难以提取到准确可靠的高程信息。针对这一问题，本项目将从建立两种技术各自的严格成像几何模型入手，在深入理解掌握地形高程与立体视差和干涉相位之间转化机理的基础上，开展两种技术联合提取地表高程信息方法的研究，充分发挥各自优势，相互弥补对方弱点，重点解决立体像对精确匹配和视差提取、干涉像对精确配准、大气相位估计去除、干涉相位解缠和定标、结果DEM融合处理等关键问题，建立二者联合从高分辨率星载SAR数据集中提取地表高程信息的处理流程，从而为我国困难地区地形测绘提供实用可靠的技术。

作为雷达遥感提取地表高程信息的两种主流技术，雷达摄影测量和干涉测量具有各自不同的优缺点。摄影测量应用限制条件少，但由于立体像对匹配难度大，高程测量精度通常不如干涉测量；干涉测量虽然高程测量精度高，但易受到去相干效应影响，同时存在整周模糊度和系统误差问题，需借助地面控制点进行定标。因此在困难地区地形测绘中单独采用一种技术往往难以提取到准确可靠的高程信息。针对这一问题，本项目开展了两种技术联合提取地表高程信息方法的研究，重点研究解决了立体像对精确匹配和视差提取、大气相位估计去除、干涉相位解缠和定标、结果DEM融合处理、结果DEM质量评价等关键问题，提出了基于外部DEM辅助的InSAR大气相位改正方法和SAR立体像对匹配方法、多基线InSAR地形高程重建方法、升降轨InSAR DEM优化融合方法、基于CUDA并行计算的立体SAR直接匹配测高方法等一系列技术方法，初步建立了二者联合从高分辨率星载SAR数据集中提取地表高程信息的处理流程，可为困难地区地形测绘提供实用可靠的技术储备。此外，本项目还针对山区峡谷地带滑坡形变监测的应用需求，探索研究了基于单对数据/时间序列差分干涉测量和偏移量分析技术的滑坡表面形变探测方法，初步解决了线性基础设施沿线形变探测、快速移动滑坡表面形变监测等难点问题，为进一步全面开展雷达遥感在地质灾害监测预警中的应用研究打下了坚实的基础。