基于干涉极化SAR数据的森林树高反演

Forest tree height estimates based on polarimetric SAR interferometry

将Cloude极化干涉散射模型和ESPRIT算法结合来降低森林参数反演的运算量。利用ESPRIT算法对Cloude模型中的散射相位中心进行估计,再将估计得到的散射相位中心代入Cloude模型,从而得到树木高度等参数。采用中国天山地区SIR-C/X-SARL波段全极化干涉数据对森林地区植被树高进行了反演实验。实验表明,与直接采用Cloude的极化干涉模型进行反演相比,在反演精度不变的情况下,运算时间仅是单独使用Cloude模型的1%～2%。

Cloude's polarimetric interferometry model was combined with the ESPRIT algorithm to reduce the calculation cost of forest parameter extraction. Local scattering centers were formulated through a combination of the orientation scattering model and the ESPRIT algorithm. The interferometric phase of the scattering centers could be used in Cloude's model to estimate the forest parameters such as tree heights. Examples derived from SIR-C/X-SAR L-Band data pairs for the Tian Shan area were used to validate the effectiveness. For the same conditions, the calculation time of the new method is 1%2% that of Cloude's method.