Study On The Extraction Method Of DBH And Tree Skeleton Based On Terrestrial Laser Scanning Point Cloud Data

In recent years,terrestrial laser scanning(TLS)technology has developed rapidly in the fields of industry,architecture,archaeology and so on.Terrestrial laser scanning can quickly obtain high-precision,high-density three-dimensional point cloud data,so its application in forest direction,such as forest measurement has become a hot research in recent years.The main purpose of this paper is to explore and discuss the method of extracting DBH and trunk skeleton based on single station terrestrial laser data.Artificial tree data and RIEGL VZ-400 TLS data were used to validate the tree diameter estimation method and tree skeleton extraction method.The results of DBH estimation showed that when scanning distance was 26m sample range,DBH estimation RMSE was 1.1cm.And when scanning distance was 56m sample range,DBH estimation RMSE was 1.99cm.For that trees regarded as elliptic distribution,the mean error of DBH estimation results were reduced by 4.7%compared with the simple circular fitting results,and the method could be used to estimate the DBH quickly and effectively.An adaptive circle-ellipse fitting method based on point cloud slices is proposed for estimating tree diameter.First the point cloud data of the forest site was sliced at breast height location,and then the point cloud of breast height transaction was clustered,at last circle-ellipse fitting adaptive DBH estimation method was used to distinguish whether they are trunk points.DBH of the trunk point in line with the circular distribution can be calculated directly,while DBH of the trunk point in line with the elliptic distribution can be calculated after breast height position correction.Tree skeleton extraction method firstly separates the trunk points from the ground point cloud from the point cloud data,and then constructs the adjacent graph of the tree points according to the neighbor relation,and the key nodes of the tree skeleton are obtained using multi-source breadth-first search method and clustering method,and the tree skeleton is obtained according to the neighbor relation of key nodes.The method of estimating the breast diameter was verified by the point cloud data of toon trees and peach trees collected by using RIEGL VZ-400 terrestrial laser scanning,and artificial simulation of Huangshan Pine peach point cloud data.The results showed that the extracted trunk skeleton is basically consistent with the real trunk skeleton,and it can be more abundant in the detail of branches.