Accuracy Assessment of UAV-Derived Data from Multi-Flight Plans with RTK GNSS Surveys

This study aimed to create an orthophotograph and digital surface model (DSM) using an unmanned aerial vehicle (UAV) with a multi-flight plan and to assess their accuracy by comparing the results to real-time kinematic (RTK) survey data. This study also focuses on the limited comparisons of various flight plans and their effects on the accuracy of UAV-derived data. The study area covered 4,000 m<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> and employed various flight plans with speed and trigger mode adjustments while maintaining a consistent flying height for a ground sample distance (GSD) of 5 cm and 70% overlap in both front and side directions. Aerial imagery was processed using Pix4D Mapper software, and RTK GNSS was utilized to obtain coordinates and elevations at predefined survey points. The methodology consisted of three main components: UAV surveying, RTK GNSS surveys, and accuracy assessments of both horizontal and vertical dimensions. Survey points were arranged in a grid pattern, totaling 30 survey points. The DJI Phantom 4 Multispectral RTK was employed for the UAV survey, while the V90 Plus GNSS RTK System was used for the RTK surveys. The accuracy assessment was evaluated using the Root Mean Square Error (RMSE) for the X and Y coordinates and overall horizontal accuracy, while vertical accuracy focused on Z values. The overall assessment provided a comprehensive view of data precision. Results showed that the FE flight plan achieved the highest accuracy for horizontal coordinates (RMSE of 0.19), whereas the NE flight plan demonstrated the highest vertical accuracy (RMSE of 0.15). Ultimately, the overall accuracy assessment (RMSE<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</inf>) indicated that the NE flight plan, utilizing Normal Flight Speed and Equal Distance Trigger modes, provided the best accuracy with an RMSE of 0.38. These results highlight the effectiveness of UAVs in generating accurate geospatial data for various applications.