使用模拟数据和地面测量多角度反射率数据,探讨了反演均匀植被结构参数的可能性。BRDF模型的反演基本上是一个非适定问题,该文以贝叶斯推论为理论依据,通过引进先验知识,补偿数据中缺乏的信息量,使有噪声情况下反演的精度得到很好的控制。对于测量数据,进一步使用了多阶段目标决策(MSDT)抑制了系统测量误差和模型偏差对目标参数的影响,效果好于直接贝叶斯反演。

In this paper, we use Bayes inversion to retrieve structural parameters of horizontal homogeneous canopy. A priori constraints are introduced in order to add information into inversion procedure. Tests with both measured data show that the results become more stable and accurate. A multi-stage inversion further suppress the systematic noise due to measurement limitation or model defect.simulated and is also used to