星载SAR复数图像的配准

Abstract

星载干涉雷达测量已被证实用于生成大范围 (乃至全球 )的 DEM或监测地表位移具有广阔的应用前景 ,其中 SAR复数图像序列的精确配准是保证生成有效干涉相位图的重要环节之一。本文分析了卫星 In SAR系统中所涉及的几种坐标系统变换问题 ,并提出了基于此的几何配准方案 ,即基于卫星轨道状态矢量、SAR成像几何进行配准的算法 ;同时 ,还设计了基于几何配准结果和能量影像相关技术的配准算法。最后使用 ERS-1 / 2、JERS-1和RADARSAT卫星获取的 3种 SAR复数图像对分别做几何配准和相关配准试验和比较 ,并对卫星 SAR图像质量和轨道数据质量作了定性分析 Spaceborne Synthetic Aperture Radar Interferometry (InSAR) and Differential InSAR (DInSAR) have been demonstrated to be the very potential tools for generating large-area even or global DEM as well as measuring and monitoring Earth surface deformation, respectively. For the purpose of computing the accurate interferogram, the co-registration of satellite SAR complex images is one of the most important processing procedures involved in satellite SAR interferometry. Based on the discussion about transformations among several spatial coordinate systems associated with interferometric processes, this paper proposes the geometric co-registration scheme of SAR images depending on satellite orbit state vectors and SAR imaging geometry. On the other hand, the correlative co-registration algorithm that makes full use of amplitude information of image chips is also designed on the basis of initial results of geometric co-registration. Finally, the co-registration experiments are carried out using three types of complex SAR image pairs that were collected by ERS-1/2, JERS-1, and RADARSAT, respectively. Some significant comparisons and analyses are made.