基于极化干涉互相关矩阵的林高估计方法

Forest height estimation method using cross-variance matrix of polarimetric interferometric SAR data

基于噪声影响较小的极化干涉数据的互相关矩阵,提出了一种新的林高估计方法.该方法使用互相关矩阵的奇异值分解代替ESPRIT方法中相干矩阵的特征分解,获取森林散射中心的干涉相位信息,再由森林散射中心的干涉相位差估计森林高度.该方法不但能抑制噪声对森林散射中心干涉相位估计的影响,还提高了运算效率.L波段松树林极化干涉仿真数据验证该方法的有效性.

Using the cross-variance matrix of polarimetric interferometric SAR data which suffers less from noises, we proposed a new forest height estimation method. The method uses the singular decomposition of cross-variance matrix, instead of the eigen decomposition of coherence matrix in ESPRIT method, to obtain the interferometric phases of forest scattering centers. Then, forest heights are estimated from their interferometric phase differences. The proposed method not only suppresses the noise effects on the estimations for forest scattering centers, but also improves the computation efficiency. The L-band simulated polarlmetric intefferometric SAR data for the pine forest support the proposed method.