Digital Terrain Model (DTM) Extraction from Leaf off UAV Images in a Dense Broadleaved Mixed Forest

Vahid Nasiri and Carmen Maftei

Abstract – The Digital Elevation Models (DEMs) are essential source of
topographic information in many disciplines. The goal of this research was to exploit
leaf-off UAV images as an alternative method for DTM development in a dense forest.
In this regard, the low-altitude images acquired by a consumer-grade camera mounted
on a drone was used for analysis. For DTM generation we used a photogrammetry
based workflow to produce and classify 3D point clouds. Based on ground 3D point
clouds and interpolation technique a high resolution DTM was generated. The results
indicated that, there is a strong relationship between the estimated (leaf-off UAV
images) and measured (DGPS) Z values (R2 = 0.972 and RMSE = 1.47 m). We
concluded that the VHR leaf-off UAV aerial imagery can be used for DTM extraction
from dense forests.

Keywords – digital elevation model, digital terrain model, UAV photogrammetry

Pages: 150-154

DOI: 10.2478/ouacsce-2022-0016