基于几何光学模型的毛竹林郁闭度无人机遥感定量反演

Retrieval of crown closure of moso bamboo forest using unmanned aerial vehicle(UAV) remotely sensed imagery based on geometric-optical model

基于几何光学模型,探讨无人机遥感数据在毛竹林郁闭度定量反演中的应用,并分析了无约束和全约束两种混合像元分解对反演结果的影响.结果表明:利用无人机遥感数据与几何光学模型在一定程度上能够实现毛竹林郁闭度的估算,但不同混合像元分解方法反演精度差异较大;相对于无约束混合像元分解而言,全约束混合像元分解反演得到的郁闭度精度高,其反演郁闭度与野外实测数据的相关系数达显著水平,决定系数R2为0.63,且均方根误差也很小,为0.04左右,能够较真实地反映毛竹林的实际情况.

This research focused on the application of remotely sensed imagery from unmanned aerial vehicle（ UAV） with high spatial resolution for the estimation of crown closure of moso bamboo forest based on the geometric-optical model,and analyzed the influence of unconstrained and fully constrained linear spectral mixture analysis（ SMA） on the accuracy of the estimated results. The results demonstrated that the combination of UAV remotely sensed imagery and geometric-optical model could,to some degrees,achieve the estimation of crown closure. However,the different SMA methods led to significant differentiation in the estimation accuracy. Compared with unconstrained SMA,the fully constrained linear SMA method resulted in higher accuracy of the estimated values,with the coefficient of determination（ R2） of 0. 63 at 0. 01 level,against the measured values acquired during the field survey. Root mean square error（ RMSE） of approximate 0. 04 was low,indicating that the usage of fully constrained linear SMA could bring about better results in crown closure estimation,which was closer to the actual condition in moso bamboo forest.