Extraction Of Tree Diameter At Breast Height Based On Terrestrial Laser Scanning And Backpack Baser Scanning

Diameter at breast height(DBH)is one of the most basic forest structure parameters.It is the basic data needed to calculate volume and establish tree growth models.Terrestrial laser scanning(TLS)can accurately obtain the surface information of the surrounding area,and has the characteristics of accurately obtaining the diameter of trees at breast height.However.TLS single-station scanning is easily blocked by the tree trunks and the scanning information is not completed.Multi-station scanning requires a target ball,and multiple scanning of a sample takes a long time.In Backpack laser scanning(BLS)This problem was solved after it appeared Compared with TLS,BLS can scan 3D point cloud information of woodland more quickly,and avoids the problem of mutual occlusion of trees.However,the scanning accuracy of BLS is lower than TLS,and the accuracy of breast diameter extraction is not ideal.Aiming at the problem that BLS has a certain gap in data quality compared with TLS,there is a certain gap in DBH extraction,and the accuracy of DBH extraction is not ideal.By analyzing the principle of BLS scanning data,it is proposed that it is more suitable for BLS to extract tree DBH.Methods and TLS data were used to extract the changes in breast diameter.The specific research contents are as follows:Extract the ground points from the TLS data through the progressive triangulation algorithm,complete the elevation normalization,and separate the trunk points through the principal component analysis method,fit the position of the individual tree according to the trunk point,and find the smallest total determined by the position of the individual tree between the stations.The initial matching point pair is obtained by equal quadrilateral,and finally the diameter at breast height is extracted by nonlinear least squares circle fitting.Then,the BLS data is denoised,the elevation is normalized,and the ICP algorithm is used to complete the registration of the sample plot with the measured data.The slice is sliced within the range of 0.8m-1.8m with a step length of 0.1m,and the difference is filtered out.Point cloud intensity point cloud to extract diameter at breast height,select the minimum diameter of the fitting circle of all slices in each tree as the diameter of this tree and compare it with the measured data to complete the BLS diameter extraction study.Finally,the TLS data of the experimental forest farm of Northeast Forestry University is used as the first phase data,and the BLS data is used as the second phase data.The registration is completed through the ntr algorithm and the previous research method is used to complete the extraction of the ground Lidar tree diameter at breast height,and the first phase TLS data Comparing the accuracy of breast diameter extraction with the BLS data of the same plot scanned in the same period.Through the above experimental method,the TLS breast diameter extraction result is rmse=1.03cm,R2=0.98.After BLS point cloud filtering and extraction,the DBH results are that the rmse of the flat plot is reduced from 2.63cm to 0.99cm,and the rmse%is reduced from 11.79%to 4.44%;the gentle slope can reduce the rmse from 5.43cm to 4.72cm,and the rmse%is reduced from 28.04%.To 24.38%;the steep slope plot can reduce rmse from 4.01cm to 2.15cm,and rmse%from 19.1%to 10.24%.The results of the change in breast diameter were the RMSE of the first-stage TLS fitting breast diameter-0.99.cm,the BLS extraction accuracy RMSE=2.43cm,the second-stage BLS fitting RMSE=1.31cm,and the average breast diameter of the second-stage plots was higher than that of the first stage.The period has increased from 20.01cm to 21.18cm.The low-intensity point cloud in the backpack-type lidar data will affect the breast diameter extraction accuracy.By filtering out the low-intensity point cloud,the breast diameter extraction accuracy can be improved.The proportion of low-quality point clouds in different point cloud data is different,and the slope is high.The proportion of low-quality point clouds in sample plots is higher than that on flat ground.Ground-based lidar can be used to extract breast diameter changes,and it is not limited to using the same equipment.