基于无人机倾斜摄影的单木参数提取研究

Research on extraction of individual tree parameters based on UAV oblique photography

随着城市树木保护工作的持续开展,有效提取树木参数的需求日益迫切。本文采用无人机倾斜摄影测量技术,对广佛交界地带一处林地进行单木参数提取。首先,进行低空倾斜摄影测量,继而获取目标区域三维密集点云模型,并利用布料滤波算法去除地面点云;其次,提出一种基于点云光谱信息的树干—树冠点云分离方法,在此基础上,通过树干点云聚类和外轮廓拟合确定树木株数、位置和胸径;最后,通过标记树木位置对树冠点云水平投影网格进行分割,进一步提取树高和冠幅。实验结果表明,本文提出的单木参数提取方法具有较强的可行性,胸径、树高、冠幅的均方根误差分别为10.14cm、1.04m、1.63m,尤其是株数识别召回率达到97%,可为单木参数提取提供高效、准确的解决方案。

With the continuous development of urban tree protection,the need for effectively extraction of tree parameters is becoming more and more urgent.In this paper,Unmanned Aerial Vehicle(UAV)oblique photography is used to extract single tree parameters from a plantation forest in the Guangzhou-Foshan junction area.We firstly carry out low-altitude tilt photogrammetry,and then obtain a 3D dense point cloud model of the target area.Then,we remove the ground point cloud by using cloth simulation filter algorithm.Secondly,we propose a trunk-canopy point cloud separation method based on the spectral information of the point cloud,on the basis of which,the number of trees,their positions and diameter at breast height are determined by clustering of the trunk point cloud and fitting of the outer contour.Finally,the canopy point cloud horizontal projection grid is segmented by labeling the tree positions to further extract the tree height and crown width.The experimental results show that the single-tree parameter extraction method proposed in this paper has strong feasibility.The root mean square errors of diameter at breast height,tree height,and crown width are 10.14cm,1.04m,and 1.63m,respectively.Especially the accuracy of plant number identified reaches 97%,which also provides an efficient and accurate solution for the individual tree parameters estimation.