Three-Dimensional Analysis of the Photogrammetric Methods PPK, DRTK2 and DRTK + GCP

Abstract

The development of drones has given new impetus to modern photogrammetry. These unmanned aerial vehicles (UAVs) have become highly valued for 3D modeling. They allow cartographers, surveyors, and others to more easily collect the data necessary for a perfect three-dimensional reconstruction of objects on the Earth’s surface. The advent of real-time kinematic positioning (RTK) systems and their use by drones allows for time savings, a reduction in ground control points (GCPs), a reduction in cost and facilitates data acquisition while maintaining good accuracy. The objective of this study is to perform a three-dimensional comparative analysis of three modern photogrammetric positioning approaches: DRTK2, DRTK2 combined with GCPs, and short-baseline PPK obtaining corrections from a GNSS base station. Data was acquired using a DJI Phantom 4 RTK. These different approaches were used in both relatively flat and rugged terrain to help users choose the most appropriate one for a given project. The evaluation focused on their accuracy in relation to the topography and their ease of implementation. The methodology was initially based on establishing an altimetric reference system through direct leveling from a reference point linked to the 1953 General Levelling Survey of West Africa (NGAO53) and a planimetric reference system using the Senegal West African Navigation (SWAN) network of permanent stations. Subsequently, the data obtained by each positioning method were processed using Pix4D Mapper software and compared with the reference data. The results obtained reveal that the DRTK2 + GCP combination (RMSE xyz = 0.062 m) ensures more consistent accuracy, particularly in rugged terrain where acquisition conditions are more complex. PPK (RMSE xyz = 0.046 m) provides better results in a non-rugged area, with simpler implementation. These results offer a better understanding of the advantages and limitations of these three GNSS positioning techniques applied to drone photogrammetry.