Location Extraction of Single Tree in Forest Based on Laser Point Cloud Data

doi:10.13466/j.cnki.lyzygl.2023.01.016

Abstract

Abstract:

This paper took the larch forest land of Saihanba Forest Farm in Hebei Province as the research object.Through setting cement piles and reflective sheets in the selected sampling sites,aiming at the low precision of forest single tree extraction,this paper conducted the research on the method of extracting the location of forest single tree based on laser point cloud data,which provided a reference for the location and quantity statistics of single tree in forest resource survey.Through the registration of the collected forest ground point cloud data with the cloud data of each survey site using the reflector model,an improved voxel space neighborhood and RANSAC cylinder fitting method was proposed to extract the single tree position from three different size sample plots.The results showed that the single tree position extraction effect of this method was good,and the extraction accuracy was more than 93%.

Key words: laser point cloud, single tree position extraction, RANSAC cylinder fitting, larch

CLC Number:

TP75

CAO Qingan, ZUO Yong, HE Yuanrong, LENG Peng. Location Extraction of Single Tree in Forest Based on Laser Point Cloud Data[J]. FOREST RESOURCES WANAGEMENT, 2023, 0(1): 133-140.

Figures/Tables 13

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Tab.1

Fig.6

Fig.7

Fig.8

Fig.9

Tab.2

Fig.10

Fig.11

References 21

[1]	徐新良, 曹明奎, 李克让. 中国森林生态系统植被碳储量时空动态变化研究[J]. 地理科学进展, 2007, 26(6):1-16.
[2]	刘鲁霞, 庞勇, 李增元. 基于地基激光雷达的亚热带森林单木胸径与树高提取[J]. 林业科学, 2016, 52(2):26-37.
[3]	Li Yumei, Guo Qinghua, Su Yanjun, et al. Retrieving the gap fraction,element clumping index,and leaf area index of individual trees using single-scan data from a terrestrial laser scanner[J]. Isprs Journal of Photogrammetry & Remote Sensing, 2017, 130:308-316.
[4]	刘月, 王君, 杨雨春, 等. 不同林分密度胡桃楸胸径、树高、材积与冠幅关系[J]. 森林工程, 2021, 37(3):28-35.
[5]	王臻, 王珠鹤. 结合Faster-RCNN和局部最大值法的森林单木信息提取[J]. 实验室研究与探索, 2022, 41(4):12-16.
[6]	余楚滢, 龚辉, 曹晶晶, 等. 基于无人机影像的无瓣海桑单木提取与地上生物量估算[J]. 热带地理, 2023, 43(1):12-22. doi: 10.13284/j.cnki.rddl.003609
[7]	张玉薇, 陈棋, 田湘云, 等. 基于UAV可见光遥感的单木冠幅提取研究[J]. 西部林业科学, 2022, 51(3):49-59.
[8]	林怡, 季昊巍, 叶勤. 基于LiDAR点云的单棵树木提取方法研究[J]. 计算机测量与控制, 2017, 25(6):142-147.
[9]	朱俊峰, 刘清旺, 崔希民, 等. 地基与无人机激光雷达结合提取单木参数[J]. 农业工程学报, 2022, 38(14):51-58.
[10]	谷志新, 裴方睿. 基于多层K-means在森林点云中的单木识别算法[J]. 林业资源管理, 2022(1):124-131.
[11]	张信杰, 郑焰锋, 温坤剑, 等. 融合机载和背包激光雷达的桉树单木因子估测[J]. 林业资源管理, 2022(6):131-137.
[12]	Yang Juntao, Kang Zhihong, Cheng Sai, et al. An individual tree segmentation method based on watershed algorithm and three-dimensional spatial distribution analysis from airborne LiDAR point clouds[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13:1055-1067. doi: 10.1109/JSTARS.4609443
[13]	戴永江. 激光雷达原理[M]. 北京: 国防工业出版社, 2002.
[14]	赵晓明, 洪波. 三维激光扫描仪应用技术研究[J]. 价值工程, 2010, 9(6):236-236.
[15]	冷鹏. 基于地面激光雷达的森林冠层LAI测量及验证[D]. 福州: 福州大学, 2019.
[16]	deConto T, Olofsson K, Görgens E B, et al. Performance of stem denoising and stem modelling algorithms on single tree point clouds from terrestrial laser scanning[J]. Computers and Electronics in Agriculture, 2017, 143:165-176. doi: 10.1016/j.compag.2017.10.019
[17]	Raumonen P, Kaasalainen M, Åkerblom M, et al. Fast automatic precision tree models from terrestrial laser scanner data[J]. Remote Sensing, 2013, 5(2):491-520. doi: 10.3390/rs5020491
[18]	Zou Jie, Leng Peng, Hou Wei, et al. Evaluating two optical methods of Woody-to-Total area ratio with destructive measurements at five Larix gmelinii Rupr. forest plots in China[J]. Forests, 2018, 9(12):746-772. doi: 10.3390/f9120746
[19]	庄崯国. 顾及冠层叶面积指数分布特征的单株树木几何建模研究[D]. 福州: 福州大学, 2017.
[20]	林怡, 季昊巍, 叶勤. 基于LiDAR点云的单棵树木提取方法研究[J]. 计算机测量与控制, 2017, 25(6):142-147.
[21]	Paris C, Valduga D, Bruzzone L. A hierarchical approach to three-dimensional segmentation of LiDAR data at single-tree level in a multilayered forest[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(7):4190-4203. doi: 10.1109/TGRS.2016.2538203

样方大小	实际单木数量/ 株	实验识别数量/ 株	识别率/ %
10m×10m	18	18	100.0
25m×25m	36	38	94.4

分块	实测数量/株	估计数量/株	精度/%
1	125	131	95.2
2	110	117	93.6
3	130	136	95.4
4	108	115	93.5
总	473	499	94.5