Assessment of positional accuracies of UAV-based coordinates derived from orthophotos at varying times of the day- A case study

Abstract

Positional accuracy is one of the important factors which determines acceptability of survey work. Apart from the equipment and method used which affect the accuracy of surveys, time of the day in which the equipment operates can equally affect the accuracy of a survey. In this study, the performance of Unmanned Aerial Vehicle (UAV) surveys as well as the appropriate time in the day to apply the technology in Tarkwa, Ghana, has been investigated. The paper assessed the positional accuracies of ground features on UAV-based orthophotos (with emphasis on horizontal coordinates), captured at different times of the day, keeping all other parameters unchanged for capturing, production and processing of all orthophotos each time. The positional accuracies of selected features on the orthophotos were determined by calculating the Root Mean Square Error (RMSE) between the feature coordinates on the ground measured with GNSS Receivers and those derived from the UAV-based orthophotos. The results show that coordinates derived from orthophotos captured in the morning, with average temperatures between 21 ℃ and 23 ℃, and average wind speed of not more than 10 m/s, produced images with the highest positional accuracies, with RMSE values between 0.0047 m and 0.0283 m. These RMSE are within the range of values recommended for standard mapping surveys as well as GIS.