南京林业大学学报(自然科学版) ›› 2021, Vol. 45 ›› Issue (4): 41-48.doi: 10.12302/j.issn.1000-2006.202006023

• 专题报道I (执行主编李凤日) • 上一篇    下一篇

基于TLS数据的杨树削度方程建立及材积估算

花伟成(), 田佳榕, 孙心雨, 徐雁南*()   

  1. 南京林业大学,南方现代林业协同创新中心,南京林业大学林学院,江苏 南京 210037
  • 收稿日期:2020-06-18 接受日期:2020-08-14 出版日期:2021-07-30 发布日期:2021-07-30
  • 通讯作者: 徐雁南
  • 基金资助:
    国家重点研发计划(2019YFD1100404)

Assessing the stem taper function and volume estimation of poplar (Populus) by terrestrial laser scanning

HUA Weicheng(), TIAN Jiarong, SUN Xinyu, XU Yannan*()   

  1. Co-Innovation Center for the Sustainable Forestry in Southern China, College of Forestry, Nanjing Forestry University, Nanjing 210037, China
  • Received:2020-06-18 Accepted:2020-08-14 Online:2021-07-30 Published:2021-07-30
  • Contact: XU Yannan

摘要:

【目的】削度方程可以很好地描述树干直径随树高变化的情况,基于地基激光雷达(terrestrial laser scanner, TLS)的高精度三维点云数据建立准确的削度方程并进行立木材积估算,对活立木尺度的材积估计具有重要意义。【方法】以江苏省黄海海滨国家森林公园杨树人工林为研究对象,获取4块样地的TLS点云数据,通过MATLAB 2020a软件计算点云平坦度和法向量以提取单木主干,采用圆拟合方法进行不同高度处的直径拟合,利用32株样木的数据,选取6种削度模型进行建模,得到杨树树干削度方程最优拟合模型,并进行材积估算。【结果】利用TLS数据提取的胸径能替代实测胸径,其平均误差小于0.90 cm。通过对6种模型的拟合优度检验,Schumacher and Hall模型为该地区杨树削度方程最优拟合模型,模型的决定系数R2=0.984,均方根误差为1.00 cm,相对百分误差为2.79%,平均预估误差为0.271%。利用Schumacher and Hall 削度方程最优拟合模型进行活立木材积的估算,经与二元材积方程估计结果进行对比,其相对差异为3.34%,二者在统计上无显著差异。【结论】该方法可以减少地面调查对树木造成的永久性破坏,为人工林的蓄积量调查提供有效的技术支持。

关键词: 杨树, 地基激光雷达, 削度方程, 材积, 主干提取, 点云数据

Abstract:

【Objective】The taper function can describe variations in the trunk diameter and tree height, and it is important to estimate standing tree volume using an accurate taper function based on the high-precision three-dimensional point cloud data of the terrestrial laser scanner (TLS). 【Method】 A poplar (Populus) artificial forest in Huanghai Haibin National Forest Park, Jiangsu Province, was examined, and TLS was used to obtain point cloud data from four sampling plots. The single tree trunk point cloud was extracted by calculating flatness and a normal vector using MATLAB 2020a software. The circular fitting method was used to fit diameters at different heights. Using data of 32 sample trees, six types of taper functions were selected for modeling. The poplar tree volume was estimated using an optimized taper function. 【Result】The DBH(diameter at breast height) produced from TLS data corresponds to the measured DBH values, with an average error of less than 0.90 cm. According to the goodness of fit test of six models, a Schumacher and Hall model was the optimal taper function for poplars in this region, with an R2 value of 0.984, RMSE is 1.00 cm, MAPE is 2.79%, and MPE is 0.271%. Volumes of standing trees were estimated using the optimal taper function. Compared with the estimation results of the binary volume equation, the relative difference was 3.34%, and there was no significant difference between them. 【Conclusion】This method may help reduce ground surveys and provide an effective technical support for artificial forest volume investigations.

Key words: poplar (Populus spp.), terrestrial laser scanner (TLS), taper function, tree volume, trunk detection, point cloud data

中图分类号: