Accuracy Assessment and Optimization of the Photogrammetric Process Variables for 3D Mapping Using Unmanned Aerial Vehicle (Drones)

Abstract

The rapid advancement of unmanned aerial vehicles (UAVs) or drone technologies has enabled the extensive use of 3D point clouds, Digital Surface Models, and orthophotos. With the proliferation of UAV technology across diverse domains, there has been an increase in investigations validating the precision of image processing outcomes. This study investigates optimal Ground Control Point (GCP) distributions and overlap conditions in two contrasting environments: a sparse area and a dense area (Urban). Four GCP distributions—stratified, half-of-the-area, corner, and center—were assessed for their impact on accuracy, using checkpoints collected using the Differential Global Positioning System surveying technique for validation. In the sparse area, the stratified distribution was the most efficient, achieving optimal results with 10–13 GCPs and a 60–80% overlap. In urban areas, the corner distribution with an 80% PPK overlap minimized GCP use to around 10–11 GCPs while maintaining low RMSE values, especially for vertical accuracy. Higher overlap rates were crucial in both environments, while poorly placed GCPs adversely affected accuracy. Although limited to two study sites, the findings provide valuable insights into GCP strategy and overlap configuration. Future research should expand this analysis to more sites and varied terrains for broader applicability. © Indian Society of Remote Sensing 2025.