文献类型：期刊文献

中文题名：用地基激光雷达提取单木结构参数——以白皮松为例

英文题名：Retrieving structural parameters of individual tree through terrestrial laser scanning data

作者：刘鲁霞[1] 庞勇[1] 李增元[1] 徐光彩[1] 李丹[1] 郑光[2]

第一作者：刘鲁霞

机构：[1]中国林业科学研究院资源信息所;[2]南京大学国际地球系统科学研究所

年份：2014

卷号：18

期号：2

起止页码：365-377

中文期刊名：遥感学报

外文期刊名：Journal of Remote Sensing

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金：国家高技术研究发展计划(863计划)(编号:2012AA12A306);国家重点基础研究发展计划(973计划)(编号:2013CB733404)~~

语种：中文

中文关键词：地基激光雷达;体元化;2维凸包;枝下高;树冠轮廓;树冠体积

外文关键词：terrestrial laser scanning, voxel, convex hull, crown base height, crown outline, crown volume

分类号：S757

摘要：以白皮松(Pinus bungeana Zucc)为研究对象,针对地基激光雷达TLS扫描的3维点云数据在单株木垂直方向的分布特征,提出了一种基于体元化方法的树干覆盖度变化检测方法,获取单木枝下高;然后根据获取的枝下高引入2维凸包算法获取垂直方向分层树冠轮廓,并计算树冠体积和冠幅;同时获取的单木参数还有胸径与树高。结果表明:单木枝下高的估测精度较高,R2与RMSE分别为0.97 m和0.21 m;胸径估测结果的R2与RMSE分别为0.79 cm和1.07 cm;采用逐步线性回归方法建立单木树冠体积与其他单木参数的相关关系,模型变量包括冠幅、叶子填充树冠长度和胸径,样本数为20,模型的R2与RMSE分别是0.967 m3和2.64 m3。本文方法能较准确地估测枝下高,TLS数据具有对树冠结构3维建模的潜力。
Terrestrial Laser Scanning （TLS） has shown great potential in obtaining the structural parameters of individual trees in past decades. In this work, three-dimensional point cloud data generated from TLS were obtained for a homogeneous forest plot with 20 Pinus bungeana （Pinus bungeana Zucc） trees. Through the analysis of the vertical profile of the point cloud data of an indi- vidual tree, a voxel-based method was developed to estimate crown base height. Further, the tree crowns were delineated by calcu- lating the convex hull of each horizontal sliced point cloud data at different heights. In addition, crown volume, crown width, tree height, and Diameter at Breast Height （DBH） were estimated at the same time. The TLS-based estimation captured 97% variation of the manually measured clear bole height （n=20, Root Mean Square Error （RMSE）= 0.21 m, p =0.01）. The R^2 and RMSE of DBH provided the coefficient of R^2 was 0.79 and 1.07 cm （n=20）, respectively. Crown length, DBH, and crown width were select- ed to predict crown volume by the stepwise linear regression method, and statistical analysis showed that the linear regression model explained 96.7% of the crown volume variance with an RMSE of 2.64 m^3（n=20, p=0.01）. The method developed in this investigation accurately estimated crown base height and crown geometry and can thus facilitate the application of TLS in precision forestry.