Comparative Study of Forest Survey Based on Handheld LiDAR and Terrestrial Laser Scanning
In order to verify the applicability of handheld LiDAR and terrestrial laser scanning under different for-est types and improve the efficiency and accuracy of forest resource survey,the forest land around Zhejiang Agri-cultural and Forestry University was taken as the study area.Three representative plots were set according to dif-ferent factors,including terrain slope,understory shrubs,and stand density.The time efficiency,point cloud quali-ty,and accuracy of single tree segmentation and parameter extraction of the two scanning modes were comprehen-sively compared.The results showed that the scanning time of handheld LiDAR was more than twice that of ter-restrial laser scanning,but the single-tree point cloud obtained showed large noises.Under gentle slope terrain conditions,more than 80％of trees have a point cloud closure at breast height ≥50％when terrestrial laser scan-ning was used,which was relatively higher than that of handheld LiDAR.Under the same conditions,the point cloud single tree segmentation accuracy obtained by the two scanning methods tends to be consistent,both above 75％.For the extraction of breast diameter,the root mean square error corresponding(1.499～1.774 cm)to the ter-restrial laser scanning point cloud data was smaller than that of the handheld LiDAR,while for the extraction of tree height,the root mean square error corresponding(1.439～1.953 m)to the terrestrial laser scanning point cloud data was greater than that of the handheld LiDAR.Both handheld LiDAR and terrestrial laser scanning have good results in forestry research.Handheld LiDAR has high scanning efficiency but slightly lower accuracy than mounted LiDAR.Terrestrial laser scanning point cloud data can more effectively reflect tree morphology but has low operating efficiency.
handheld LiDARterrestrial laser scanningindividual tree segmentationindividual tree parameter extraction
NETL
NSTL
万方数据
