利用辐射传输模型和随机森林回归反演LAI

针对辐射传输模型与查找表结合反演叶面积指数的方法存在反演工作量大且反演速度缓慢的问题,提出利用辐射传输模型和随机森林组合模型对路域植被叶面积指数进行估算的方法。该模型定义一种辐射传输模型和随机森林回归模型结合反演叶面积指数的方法。以研究区实测高光谱数据和模拟光谱数据为数据源,在相关性分析和敏感性分析的基础上,选取适宜作为反演因子的植被指数,而后进行随机森林算法回归,反演得到预测叶面积指数。结果表明:基于辐射传输模型和随机森林算法反演的路域植被叶面积指数与实测结果一致,准确及时的反映路域植被叶面积指数信息,可以较好地应用在路域环境植被参数反演中。

LAI inversion using radiation transfer model and random forest regression

In order to overcome the shortcomings of large inversion workload and slow inversion speed in the method of combining radiation transfer model with look-up table,a method of estimating the vegetation leaf area index in road area by using radiation transfer model and random forest combination model is proposed.The model defines a method of retrieving leaf area index by combining radiation transfer model with stochastic forest regression model.Based on the correlation analysis and sensitivity analysis,the vegetation index suitable for the inversion factor is selected,and then the random forest algorithm regression is carried out to obtain the predicted leaf area index.The result shows that the retrieved vegetation leaf area index based on the radiation transfer model and the stochastic forest algorithm is consistent with the measured result.It can reflect the vegetation leaf area index information accurately and timely,and can be better applied to the retrieved vegetation parameters of the road environment.vegetation parameters.