Seillinienplanung: Genauigkeit der Einzelbaumdetektion mit drohnengenerierten Luftbildern

Abstract

Skyline planning: accuracy of singletree detection with dronegenerated aerial photos Information collected with drones has the potential to simplify the planning and marking of skylines. The prerequisite for this is that 1) the tree coordinates of potential intermediate support and anchor trees and 2) their diameter at breast height (DBH) can be determined with sufficient accuracy in the surface model obtained from the drone data. To analyse the achievable accuracies, two Swiss marteloscopes were surveyed with a senseFly eBee Classic drone at a height of 180 m and aerial photos were taken with a Sony camera with a resolution of 18.2 megapixels. In the resulting normalised surface model (nDOM) the local maxima (treetops) were determined with the single-tree detection software “FINT”. For these treetops the coordinates and the DBH were determined. The detection rate for both marteloscopes was 65%. On average, the coordinates deviated less than 1.4 m from the terrestrial reference tree coordinates. The predominant and dominant trees could be located even more precisely. The DBH was determined with an average accuracy of 5 cm. A practical test with nine skylines showed that the coordinates were accurate enough to use the support trees determined in the nDOM for the technical realisation of the skylines. However, an on-site inspection is still necessary to check the potential intermediate support trees for damages, invisible in the aerial view.