测绘学报 ›› 2014, Vol. 43 ›› Issue (10): 1061-1067.doi: 10.13485/j.cnki.11-2089.2014.0170

• 学术论文 • 上一篇    下一篇

极化干涉SAR植被高反演复数最小二乘平差法

付海强,朱建军,汪长城,解清华,赵蓉   

  1. 中南大学 地球科学与信息物理学院
  • 收稿日期:2013-12-05 修回日期:2014-02-20 出版日期:2014-10-20 发布日期:2014-10-24
  • 通讯作者: 付海强 E-mail:15111144067@163.com
  • 基金资助:

    国家自然科学基金项目;国家973计划资助项目;湖南省自然科学基金;国家自然科学基金项目;国家自然科学基金项目

Polarimetric SAR Interferometry Vegetation Height Inversion Method of Complex Least Squares Adjustment

FU HaiqiangZHU Jianjun2,WANG Changcheng2,XIE Qinghua2,ZHAO Rong2   

  • Received:2013-12-05 Revised:2014-02-20 Online:2014-10-20 Published:2014-10-24
  • Contact: FU Haiqiang E-mail:15111144067@163.com

摘要:

提出一种PolInSAR植被高反演的复数最小二乘法。在考虑植被层时间去相干的条件下,将RVoG扩展为RVoG+VTD模型;之后,将模型解算问题概括为复数的实部、虚部联合平差问题,提出了该模型线性化的方法及参数解算方法;最后利用解算得到的更为准确的“纯”体去相干系数反演植被高度。采用覆盖德国Oberpfaffenhofen地区的2景E-SAR L波段数据进行试验并与经典的非线性迭代及三阶段算法进行对比分析。结果表明,本文提出的方法结果精度优于其他两种方法,运算效率方面明显优于非线性迭代算法,略低于三阶段算法。

关键词: 极化干涉SAR, 树高反演, 复数最小二乘平差, 时间去相干

Abstract:

A PolInSAR vegetation height inversion method of complex least squares adjustment (CLSA) is proposed. Considering the temporal decorrelation, the RVoG is extended into RVoG+VTD. And then, the issue of resolving model is summarized as combined adjustment of complex real and imagine. The linearization method and parameter retrieval method are also given. At last, the robust volume-only coherence is used to compute the vegetation height. The CLSA method is validated on 2 E-SAR L band data of Oberpfaffenhofen, Germany and compared with the results of nonlinear iteration and three-stage methods. It demonstrates that the result of CLSA method is more accurate. The computational efficiency of CLSA method is higher than that of nonlinear iteration method and slightly lower than three-stage method.

Key words: polarimetric SAR interferometry(PolInSAR), vegetation height inversion, complex least squares adjustment, temporal decorrelation

中图分类号: