摘要
树高与林木的生物量、碳储量息息相关,测量树高是传统森林调查的痛点和难点,需要大量的工作量。使用机载LiDAR能方便地提取森林树高,极大地减少外业工作量。为了研究机载LiDAR提取树高的精度,以云南省红塔区为试验区,采集3个样地的样木调查信息(包括树高、胸径、树种和位置)和点云数据,通过对点云数据进行去噪、重采样、地面点分类、点云归一化和单木分割等处理提取单株树高,并用外业调查数据进行验证。结果表明,机载LiDAR提取的树高信息具有极高精度,可达95%以上,完全可以满足实际调查的要求;在对林分样木进行单木分割时,F评分在70%~85%范围,且在不同样地中使用相同方法其参数分割效果存在显著差异,这可能与林分树高的分布特征和林木的形态有关。建立一个普适性较强、精度较高的机载LiDAR分割模型可极大地提升森林生物量和碳储量的反演精度。
Tree height is closely related to the biomass and carbon stock of trees,and measuring tree height is a challenging and labor-intensive aspect of forest surveys.The use of airborne LiDAR can conveniently extract forest tree height and greatly reduce the workload of fieldwork.In order to study the accuracy of tree height extraction by airborne LiDAR,this study selected Hongta District in Yunnan Province as the test area,and collected sample tree survey information(including tree height,diameter at breast height,tree species and location)and point cloud data from three sample plots.The point cloud data were processed through denoising,resampling,ground point classification,normalization,and individual tree segmentation to extract individual tree heights,which were then validated using field survey data.The results indicated that tree heights extracted from airborne LiDAR achieved high accuracy,exceeding 95%,fully meeting the requirements for practical surveys.During individual tree segmentation for sample trees,the F-score ranged from 70%to 85%.However,segmentation performance using the same parameters varied significantly across different plots,likely due to differences in the distribution characteristics of tree heights and tree morphology.Developing a more universal and accurate airborne LiDAR segmentation model would greatly enhance the inversion accuracy for forest biomass and carbon stock.
作者
冷鸿天
施凯泽
黄运荣
吴颜奎
昝建春
廖聪宇
LENG Hongtian;SHI Kaize;HUANG Yunrong;WU Yankui;ZAN Jianchun;LIAO Congyu(Yunnan Institute of Forest Inventory and Planning,Kunming 650051,China)
出处
《林业调查规划》
2026年第1期1-6,共6页
Forest Inventory and Planning
基金
云南省林业调查规划院青年联合会项目“机载激光雷达森林参数反演研究”(2022311026)
云南省科技厅科技计划项目(林草联合专项)“云南省桤木等五个主要树种的林分碳计量技术研究”(202404CB090005).
