欢迎访问林草资源研究

林草资源研究 ›› 2023›› Issue (6): 105-112.doi: 10.13466/j.cnki.lczyyj.2023.06.013

• 技术应用 • 上一篇    下一篇

基于机载激光雷达点云数据的马尾松单木材积估测

唐佳俊(), 汪刚, 柴宗政()   

  1. 贵州大学 林学院,贵阳 550025
  • 收稿日期:2023-08-14 修回日期:2023-11-13 出版日期:2023-12-28 发布日期:2024-02-21
  • 通讯作者: 柴宗政,副教授,硕士生导师,主要研究方向:森林经营与生态建模。Email:chaizz@126.com
  • 作者简介:唐佳俊,硕士研究生,主要研究方向:林业遥感。Email:tangjj0706@163.com
  • 基金资助:
    国家自然科学基金(32001314);贵州省林业科研项目“马尾松中幼林近自然抚育关键技术研究”(黔林科合[2022]38号);贵州大学培育项目“近自然经营对马尾松群落植物功能性状调控机制”(贵大培育[2019]38号)

Single-TreeVolume Estimation of Pinus massoniana based on Airborne LiDAR Point Cloud

TANG Jiajun(), WANG Gang, CHAI Zongzheng()   

  1. College of Forestry,Guizhou University,Guiyang 550025,China
  • Received:2023-08-14 Revised:2023-11-13 Online:2023-12-28 Published:2024-02-21

摘要:

为提高森林单木材积估测精度和效率,选取贵州省织金县城郊典型马尾松林为研究对象,基于机载激光雷达点云和样地调查数据,以提取的树高、冠幅、树冠投影面积和树冠体积等单木结构参数为变量,构建基于机载激光雷达点云数据的马尾松单木材积估测模型。结果表明:1)基于点云数据提取的马尾松单木树高和冠幅因子与实际调查数据之间存在良好的相关性,决定系数R2在0.7以上,精度相对较高,可用于构建马尾松单木材积模型。2)在经典非线性CAR模型基础上,利用枚举法对树高、冠幅、树冠投影面积、树冠体积等4个变量组合构建的11个模型中,包含树高、冠幅及树冠体积三个林分因子的模型表现最佳,R2为0.774 1。3)树高、冠幅及树冠体积被确定为马尾松单木材积估测的关键因子,其中,树高的贡献最大且与单木材积呈极显著正相关关系(P<0.001)。利用机载激光雷达点云数据提取单木结构参数,并基于非线性CAR模型构建单木材积模型估测马尾松单木材积的方法是可行的,该方法不仅能满足森林资源调查的精度要求,且能有效提高调查效率。

关键词: 机载激光雷达, 马尾松, 单木, 结构参数, 材积估测

Abstract:

In order to improve the accuracy and efficiency of forest single-tree volume estimation,the typical Pinus massoniana forest in the suburban area of Zhijin County,Guizhou Province was selected as the research object.Based on airborne LiDAR point cloud and sample survey data,a single-tree structure parameter estimation model for P.massoniana based on airborne LiDAR point cloud data was constructed using extracted tree height,crown width,crown projection area,and crown volume as variables.The research results indicate that:1)There was a good correlation between the height and crown width factors of P.massoniana single-tree extracted from point cloud data and actual survey data,with a determination coefficient of R2 above 0.7 and relatively high accuracy,which could be used to construct a single-tree volume model of P.massoniana.2)On the basis of the classic nonlinear CAR model,an enumeration method was used to construct 11 models that combined four variables:tree height,crown width,crown projected area,and crown volume.Among them,the model containing three stand factors,tree height,crown width,and crown volume,performed the best,with an R2 of 0.774 1.3)Tree height,crown width and crown volume were identified as key factors in estimating the individual volume of P.massoniana,with tree height contributing the most and showing a highly significant positive correlation with individual volume (P<0.001).It is feasible to extract single tree structural parameters using airborne LiDAR point cloud data and construct a single tree volume model based on nonlinear CAR model to estimate the single tree volume of P.massoniana.This method not only meets the accuracy requirements of forest resource surveys but also effectively improves survey efficiency.

Key words: Airborne LiDAR, Pinus massoniana, single-tree, structural parameter, volume estimation

中图分类号: