University of Twente Student Theses
Multi-UAV Real-Time Collaborative Aerial Mapping for 3D Reconstruction
Shastry, Pragna (2024) Multi-UAV Real-Time Collaborative Aerial Mapping for 3D Reconstruction.
|
PDF
29MB |
| Abstract: | UAVs are being increasingly used to perform Simultaneous Localization and Mapping (SLAM) tasks due to being compact, low cost, lightweight, and their ability to be used in complex environments. Multiple UAVs can be used to map the area rapidly and in real-time, especially in disaster areas. However, developing and testing SLAM systems in the real-world conditions can be challenging, so a simulation environment provides a controlled platform for creating datasets. A 3D model is created using Blender and 3D map tiles. A gazebo simulation is built with this model and a drone. ArduPilot is used as the FCU. Using QGC and mavros, missions are planned, and the drone captures camera and GNSS data. The Blender model and Gazebo plugins provide ground truth data. SLAM is performed, comparing the individual agent scenario with the multi-agent scenario. The SLAM output has arbitrary scale and orientation. To improve that, the GNSS data is used to georeference the trajectory. The percent RMSE with respect to trajectory length reduces in collaborative SLAM. The 3D reconstruction point clouds are examined. Collaborative SLAM merges and optimizes maps of different agents, so the map is denser and more accurately aligned with the ground truth. |
| Item Type: | Essay (Master) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 38 earth sciences |
| Programme: | Robotics MSc (60973) |
| Link to this item: | https://purl.utwente.nl/essays/104702 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page