基于PolInSAR相干区域的最优正规矩阵近似解的地形与树高估计

发布时间：2018-05-12 06:11

  本文选题：极化干涉合成孔径雷达 + 相干区域　； 参考：《电子与信息学报》2017年05期

【摘要】：森林区域林下地形及树高的反演是极化干涉雷达的一个重要应用。该文首先对极化干涉SAR数据的相干区域进行建模及运用最优正规矩阵近似干涉互相关矩阵,得到白化正规干涉互相关矩阵。白化正规干涉互相关矩阵的相干区域为一条直线,任意求得两个不同极化状态下的相干系数进行直线拟合,完成地表的估计,再结合体散射去相干与树高之间的关系,运用查表方法完成树高的估计。该方法回避了传统方法中求解所有极化状态下的相干系数估计及相干区域边缘提取的步骤,在简化参数反演提升估计效率的同时获得正确地表与树高估计,最后运用仿真数据完成算法有效性与可靠性的验证。
[Abstract]:The inversion of forest understory topography and tree height is an important application of polarimetric interferometric radar. In this paper, the coherent region of polarimetric interference SAR data is modeled and the whitening normal interference cross-correlation matrix is obtained by using the optimal normal matrix to approximate the interference cross-correlation matrix. The coherent region of the whitening normal interference cross-correlation matrix is a straight line. The coherence coefficients in two different polarization states are obtained to fit the coherent coefficients in a straight line. The estimation of the ground surface is completed, and the relation between the decoherence and the tree height is obtained. The estimation of tree height is completed by the method of look-up table. This method avoids the steps of solving the coherent coefficient estimation and the edge extraction of coherent region in all polarization states in the traditional method. It simplifies the parameter inversion to improve the efficiency of estimation and obtains the correct estimation of surface and tree height at the same time. Finally, the validity and reliability of the algorithm are verified by simulation data.
【作者单位】： 北京理工大学光电学院;北京理工大学信息与电子学院;
【分类号】：TN958
，

本文编号：1877444