一种基于BP神经网络的极化干涉SAR植被高度反演方法  

A BP Neural-network Based Method for Vegetation Height Inversion of the Polarimetric Interferometric SAR

作　　者：李廷伟[1] 梁甸农[1] 黄海风[1] 朱炬波[1] 

机构地区：[1]国防科技大学电子科学与工程学院,湖南长沙410073

出　　处：《国防科技大学学报》2010年第3期60-64,共5页Journal of National University of Defense Technology

基　　金：国家自然科学基金资助项目(60902092);国家部委基金资助项目(41307020203)

摘　　要：地面干涉相位估计偏差和植被散射模型偏差都将引起三阶段植被高度反演偏差,针对该问题,提出了基于BP神经网络的植被高度反演方法,该方法直接利用BP神经网络模拟极化复相关系数与植被高度之间的非线性映射关系,不仅可以避免地面干涉相位估计偏差导致的植被高度反演偏差,还能降低三阶段植被高度反演方法面临的散射模型偏差导致的植被高度反演偏差,具有比三阶段植被高度反演方法更高的反演精度。实验结果验证了新方法的优越性。Error in the estimation of the ground interferometric phase and that in the forest scattering model will bring about errors of the inversion of the vegetation height. Aiming at this problem,a new inversion method based on the BP neural network was proposed. The new method directly fits the nonlinear mapping relationship between the complex polarimetric correction coefficients and the vegetation height,so it can reduce the error caused by the error in the estimation of the ground interferometric phase,and that caused by the scattering model error. The new method has better performance than the three-stage vegetation height inversion method,and the simulated results validate the superiority of this new method.

关 键 词：植被高度反演 极化干涉SAR BP神经网络 三阶段植被高度反演 

分 类 号：TN958[电子电信—信号与信息处理]