Mit ParaView

Mit der Visualisierungs-Software ParaView lassen sich auf einfache weise repräsentative Grafiken aus LiDAR Punktwolken erstellen.

Laden einer XYZ ASCII Punktwolke mit 'File → Open…, Have Headers und Field Delimiter Characters entsprechend dem Dateiformat anpassen. Umwandeln der geladenen ASCII Tabelle in eine Punktwolke mit Filters → Alphabethical → Table to Points unwandeln, X, Y und Z Spalten aus der Tabelle zuweisen. Mit einer 3D View kann die Punktwolke bereits visualisiert werden. Rekonstruieren der Oberfläche (feature preserving surface reconstruction) mittels Delaunay 2D Triangulierung mit Filters → Alphabethical → Delaunay 2D ====== Rendering LiDAR with MeshLab ====== Rendering point clouds originated from airborne laser scanning using MeshLab (cross platform) is a simple process. For using GRASS GIS see also http://grasswiki.osgeo.org/wiki/LIDAR. ===== Raw Data ===== The point cloud should be provided as ASCII XYZ data, a three column whitespace separated format. If the available dataset was provided in a different formatting, use an advanced editor like Kate (Linux) or Notepad++ (Windows) to reformat it. ===== Import ===== Use File → Import Mesh (Ctrl + H) to import an *.xyz'' point cloud.

Compute the vertice's normals using Filters → Point Set → Compute normals for point sets. For Numbers of neigbors select a value around 10 to keep the structure crisp enough. Make sure to disable the Flip normals w.r.t. viewport option, as the normals should align only in respect to each other.

Use Render → Show Vertex Normals to check the result.

To create the mesh use Filters → Remeshing, … → Surface Reconstruction: Ball Pivoting. Do not use the Poisson algorithm as the result require heavy mesh cleaning post processing.

Export the mesh using the PLY format, File → Export Mesh (Ctrl + E) as it can be import with various other 3D rendering application.

— Bernhard 2013/11/29 15:20