测绘学报

• 学术论文 • 上一篇    下一篇

基于ESPRIT算法的极化干涉SAR植被高度反演研究

谈璐璐1,杨立波2,杨汝良2   

  1. 1. 中国电子科技集团第38研究所
    2. 中国科学院电子学研究所
  • 收稿日期:2008-10-27 修回日期:2010-12-20 出版日期:2011-06-25 发布日期:2011-06-25
  • 通讯作者: 谈璐璐

Investigation on Tree Height Retrieval with Polarimetric SAR Interferometry based on ESPRIT Algorithm

TAN Lulu 2, 2   

  • Received:2008-10-27 Revised:2010-12-20 Online:2011-06-25 Published:2011-06-25
  • Contact: TAN Lulu

摘要: 由于存在着去相干分量,利用ESPRIT(旋转不变技术)算法对植被区域的极化干涉SAR(POLInSAR)数据进行反演的结果存在着较严重的偏差。针对这一问题,结合极化干涉相干最优理论及其物理散射机制,引入新的散射矢量——相干最优化散射矢量,提出了一种改进的基于ESPRIT的植被高度反算法。最后,利用欧空局(ESA)提供的模拟L波段全极化干涉数据对改进算法的有效性进行了验证,试验结果证明改进算法能很大程度地提高反演的精度。

Abstract: Tree height retrieval results making use of polarimetric interferometry SAR data based on ESPRIT algorithm is seriously biased because of the existence of depolarized components. In order to solve this problem, a modified inversion algorithm combining coherence optimization theory and its physical scattering mechanism is presented. Also, a new scattering vector called coherence optimized scattering vector is introduced to help understanding the algorithm. Finally, the proposed and traditional ESPRIT algorithm is applied to simulated L-Band PolInSAR data provided by ESA respectively, and the experiment results suggest the validity of the inversion algorithm.