Evaluation and calibration of fixed-wing multisensor uav mobile mapping system: Improved results

Abstract

Unmanned Aerial Vehicles (UAVs) are willingly used in photogrammetry and remote sensing, especially for image acquisition, and are characterised by high spatial resolution. UAVs can be used for the fast and, if necessary, frequent acquisition of spatial data, especially for small areas. In recent years, new trends in the development of UAVs have emerged, including the integration of various sensors and the application of ultralight laser scanners. Within the described experiment, UAV data, i.e. RGB and NIR imagery, as well as ALS data were obtained over three test areas. For one test area, the flight calibration was performed. 3 strips were oriented perpendicularly to another 3 strips and the flight was performed on two different heights: 120 and 150 m. In order to process the data acquired for the next 3 test areas, the determined calibration parameters were utilised. The oriented images were used to generate RGB and NIR ortophotos, as well as the point cloud using the Dense Image Matching (DIM) algorithm. Height differences between UAV Laser Scanning (ULS) and DIM clouds were calculated for all test areas. Experiment data from Terrestrial Laser Scanning and check points measured with GPS RTK have been used. Finally, an accuracy of less than 10 cm was achieved for the DTM. The results were improved by eliminating the problem of horizontal accuracy, but its influence is still slightly visible on the vertical accuracy of the data. The experiment proved the quality of data obtained with the ultralight scanner mounted on the platform moving with much more speed, being an alternative to manned flight missions and multi-rotors UAVs.