基于TLS的树高转换模型构建研究

Construction of Tree Height Conversion Model Based on Terrestrial Laser Scanning

树高是森林资源调查中重要的结构参数,也是森林蓄积量、树木生长模拟及森林碳汇计算的基础因子。以安徽省休宁县西田国有林场的杉木人工林为研究对象,使用地基激光雷达(TLS)采集单株样木的激光点云数据,通过HDScene软件完成测站拼接、去噪等预处理,并直接提取单株杉木树高。同时,联立获取的伐倒样木树高实测值,运用相关分析和回归分析方法构建树高转换模型。结果表明,树高转换模型关系式为y=0.970x+0.312,R^(2)=0.973,RMSE=0.441,并且当树高超过9 m时,TLS树高测量值与伐倒木真实值之间的误差会降低且趋于平稳。该研究提出一种获取高精度树高参数的方法,可为解决传统方法获取树高时精度不高或工作量大等不足提供科学依据。

Tree height is an important structural parameter in forest resource survey and a fundamental factor in forest stock,tree growth simulation and forest carbon sink calculation.In this paper,laser point cloud data of single sample trees were collected using terrestrial laser scanning(TLS).Pre-processed by HDScene software,such as station stitching and denoising,and extracted directly from single fir tree heights in Xitian state-owned forest in Xiuning County,Anhui Province.At the same time,the measured height of the felled sample trees was obtained by the correlation analysis and regression analysis to construct a tree height conversion model.The results showed that the relationship between the tree height conversion model was y=0.970x+0.312,R^(2)=0.973,RMSE=0.441,and it was found that the error between the TLS tree height measurement and the real value of the felled wood decreased and becomed smooth when the tree height exceeded 9 m.This study proposed a method to obtain high-precision tree height parameters to solve the shortcomings of traditional methods in obtaining tree height with low accuracy or high workload.