UAV Block Geometry Design and Camera Calibration: A Simulation Study

Area of Science:

• Geomatics Engineering
• Photogrammetry
• Remote Sensing

Background:

• Established guidelines for analogue aerial photogrammetry are absent in Unmanned Aerial Vehicle (UAV) photogrammetry.
• Numerous project parameters (camera, flight plan, georeferencing, calibration) contribute to this complexity.
• Camera calibration methods (pre- vs. on-the-job) present a significant decision point.

Purpose of the Study:

• To evaluate the accuracy of camera parameters and ground coordinates using Monte Carlo simulations.
• To assess the impact of various project parameters on photogrammetric accuracy.
• To investigate the bias (dome effect) caused by inaccurate pre-calibration.

Main Methods:

• Generation of synthetic Unmanned Aerial Vehicle (UAV) photogrammetric blocks (144 combinations).
• Simulation of varying terrain and area shapes, block control types (ground and aerial), strip configurations, and data precision.
• On-the-job self-calibration and investigation of pre-calibration bias.

Main Results:

• Accuracy differences between configurations can reach an order of magnitude.
• Oblique imaging is essential for accurate results in flat terrain.
• Aerial control (camera station positions) outperforms Ground Control Points (GCPs) in efficiency and accuracy, especially in flat terrain.
• Ground control density showed less impact than expected.

Conclusions:

• Project parameter selection significantly influences UAV photogrammetric accuracy.
• On-the-job self-calibration with appropriate configurations is vital for reliable results.
• Aerial control offers a more accurate and efficient georeferencing solution for specific terrains compared to traditional GCPs.