search

VUKOZ-OEL / 3d-forest-classic

software for analysis of Lidar data from forest environment.
GNU General Public License v3.0
84 stars 26 forks source link

UAV LiDAR #23

Open MichaelCHHang opened 3 years ago

MichaelCHHang commented 3 years ago

Hi,

First of all, big thanks for the open-source program. I am going to use this program for extracting the parameters of trees, including DBH, tree height, and tree crown spread, from the point cloud data acquired by the UAV LiDAR. As the point density is normally lower than that acquired by the Terrestrial Laser Scanner, the performance of the segmentation of the tree is not as good as the result from the TLS point cloud. So, is there any tips for me to enhance the performance of the tree segmentation, such as the settings of the built-in functions and the workflow of the point cloud processing apart from the modification of the flight planning and the site selection for the tree (i.e. a site with less structural complexity)?

Best regards Michael

janekT commented 2 years ago

Hi, for segmentation of UAV pointclouds we have some results with diferent type of input. For terrain segmentation, you can use values around 0,5 - 1 m of voxel or octree node.
For tree segmentation we used bigger voxel size ( depends on cloud density) so at least 3 point should be present in voxel. Since tree segmentation is biased to data source my advice is just to try various parameters ot obtain the possible results.