Epsilon Archive for Student Projects

Abstract

The master ‘s thesis focused on forest inventory process based on the photogrammetric point cloud. The study deals with forest stand data collecting using terrestrial and aerial photogrammetry. The raw data – digital images were taken by walking through the forest stand with the imaging device carried on the measuring rod and by the drone with mounted camera.

This thesis describes data processing in the Agisoft PhotoScan, where are the raw images processed to the 3D point clouds. 3D point clouds were analysed in the ArcMap SW. From the modelled stems, the cuts at breast height were extracted, from which, the single stem diameter at breast height (DBH) was calculated. Based on ArcMap SW analysis, the tree height was calculated. For the tree height values calculation, the canopy height model (CHM) was created by so-called inversed watershed method. Volume of the individual stems was calculated from the values based on the photogrammetry data. Calculated stand characteristics were compared with ground true data, collected by the conventional methods. The calculation of the stand characteristics was achieved with RMSE 1,5 m error of the tree height, 2,9 cm error of the DBH and 0,14 m3 of the single stem volume. It was concluded that the presented technology can be used for forest inventory. Thanks to the possibility of its operative application, i.e. the possibility of updating the stand data at very short intervals is therefore possible to evaluate the actual thinning interventions in the premature stands, development of the regeneration of the stand, and also the potential threat of the stands by calamitous pests.