期刊文献+

基于三维点云模型获取单木树高和胸径的研究

Study on Obtaining Height and DBH of Single Tree Based on 3D Point Cloud Model
下载PDF
导出
摘要 使用无人机和智能手机分别从空中和地面拍摄的样地林分影像构建三维点云模型,并从三维点云模型中获取样地内单木树高和胸径参数。本研究以池杉人工林为研究对象,利用PhotoScan Agisoft软件对无人机倾斜摄影和智能手机近景摄影的样地影像进行三维重建,通过对齐照片、控制点刺点、对齐优化、建立密集点云等步骤,构建出与样地实景相符的三维点云模型;通过LiDAR 360软件从样地三维点云模型中获取单木的树高和胸径参数,将其与实地测量获取的单木树高和胸径参数进行对比分析。利用无人机和智能手机影像构建的三维模型可以满足《数字航空摄影测量测图规范》的精度要求。通过实测数据和点云数据获取的树高和胸径的平均差值分别为-0.9 m和-0.8 cm,平均相对误差分别为5.4%和7.1%。以实测数据作为自变量x,以点云数据作为因变量y,树高和胸径回归模型的R2分别为0.809 5和0.918 4。将倾斜摄影和近景摄影的点云模型统一在同一空间参考基准下可构建出与样地实景相匹配的三维点云模型,从样地三维点云模型中获取的单木树高和胸径与实地测量结果具有较好的线性相关性,本研究所使用的方法可以代替传统林业调查手段应用于人工林。 The 3D point cloud model was constructed from aerial and ground stand images taken by uav and smart phone respectively, and the height and DBH parameters of individual trees in the sampling area were obtained by the 3D point cloud model.The plantation Taxodium chinensis var.imbricatum was studied in this study, PhotoScan Agisoft software was used to conduct 3D reconstruction of the sample site images from UAV oblique photography and smart phone close-range photography.Through the steps of photo alignment, control point thorn, alignment optimization, and establishment of dense point cloud, a 3D model consistent with the real scene of the sample site was constructed;The height and DBH parameters of individual trees were obtained from the point cloud data of the 3D model of the sample plot by LiDAR 360 software, and then compared with the height and DBH parameters obtained by field measurement.The 3D model constructed by uav and smart phone images can meet the accuracy requirements of Digital Aerial Photogrammetry Mapping Specification.The average difference of tree height and DBH obtained from the measured data and point cloud data were-0.9 m and-0.8 cm, respectively, and the average relative errors were 5.4% and 7.1%,respectively.With measured data as independent variable X and point cloud data as dependent variable Y,R~2 of tree height and DBH regression models were 0.809 5 and 0.918 4,respectively.A 3D model matching the real scene of the sample site can be constructed based on the point cloud data of oblique photography and close-range photography under the same spatial reference.The height and DBH of individual trees obtained from the point cloud data of the 3D model of the sample site have a good linear correlation with the field measurement results.The method used in this study can replace the traditional forestry survey method and be applied to the artificial forest.
作者 王怡 汤景明 孙拥康 袁传武 Wang Yi;Tang Jingming;Sun Yongkang;Yuan Chuanwu(Hubei Academy of Forestry,Wuhan 430075)
出处 《湖北林业科技》 2023年第1期22-29,共8页 Hubei Forestry Science and Technology
基金 湖北省林业科学研究院科技基金项目“基于无人机航拍影像的森林资源调查与监测研究”(2021YJJ03) 中德财政合作林业重大科研项目“近自然森林经营示范与成效监测”(zdhz2017ky02)。
关键词 倾斜摄影 近景摄影 三维点云模型 树高 胸径 oblique photograph close-range photography 3D point cloud model Tree height DBH
  • 相关文献

参考文献25

二级参考文献221

共引文献556

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部