基于激光雷达数据的森林蓄积量模型反演及精度估算

Inversion and Accuracy Estimation of Forest Volume Model Based on Lidar Data

森林资源管理逐渐走向全面数字化、信息化,森林蓄积量作为林业调查的一项重要指标,在遥感反演领域一直处在热门地位。利用激光雷达数据和样地数据,使用多元线性回归和随机森林数学模型,选取合适的参数及评价指标,得到对森林蓄积量的反演结果;再对两重回归估计算法进行优化,得到两步回归估计,使用该方法对森林蓄积量反演结果与人工验证结果进行精度估算,兼顾到了全覆盖的激光雷达数据反演结果以及人工修正和地面验证两个部分的验证数据,得到的精度估算结果较好。

Forest resource management is gradually moving towards comprehensive digitization and informatization.As an important index of forestry survey,forest volume has been in a hot position in the field of remote sensing inversion.By using lidar data and sample plot data,using multiple linear regression and random forest mathematical model,selecting appropriate parameters and evaluation indexes,the inversion results of forest volume are obtained;then the double regression estimation algorithm is optimized,and the two-step regression estimation is obtained.The method is used to estimate the accuracy of forest volume inversion results and manual verification results,taking into account the accuracy.The results of lidar data inversion,manual correction and ground verification are satisfactory.