Assessing the spatial accuracy of UAV-derived products based on variation of flight altitudes


Turkish Journal of Engineering, vol.5, no.1, pp.35-40, 2021 (Scopus) identifier identifier

  • Publication Type: Article / Article
  • Volume: 5 Issue: 1
  • Publication Date: 2021
  • Doi Number: 10.31127/tuje.653631
  • Journal Name: Turkish Journal of Engineering
  • Journal Indexes: Scopus, TR DİZİN (ULAKBİM)
  • Page Numbers: pp.35-40
  • Keywords: Digital surface model (DSM), Point cloud, Unmanned aerial vehicle, Volumetric analysis
  • Istanbul Technical University Affiliated: Yes


Unmanned Aerial Vehicles (UAVs), which can carry a variety of payloads, and be operated automatically or manually with ground control stations. Nowadays, UAVs can make photogrammetric flight plans and obtain photogrammetric data with existing sensor systems. Automatic data acquisition processes provide lower cost, and high spatial and temporal resolution images in a short period of time compared to other measurement methods. As a result, orthomosaics, dense point clouds and digital surface models (DSMs) are produced and these UAV-derived data are used in various disciplines such as constructions, geomatics, earth sciences, etc. In this study, the same flight plans were realized with an UAV at different altitudes and all aerial images were obtained with the same integrated digital camera. As a result of the processing of images acquired from different altitudes, orthomosaics, DSMs and point cloud were produced. In this study, it is aimed to compare the length, areal and volumetric differences of a small geostationary object. Ground control points (GCPs), which were collected by RTK-GPS (Real-Time Kinematic) in conjunction with the flight integrated into data production process in order to highly accurate product. Ultimately, cross-correlation has been done with the produced data and the terrestrial measurement. Results show that the dimension of the object depend on the flight altitude as expected, however the volumetric changes vary due to the uncertainties in the raw point cloud data.