Computationally Efficient Vision-based UAV Positioning

Joakim Rydell, Erika Bilock, Michael Tulldahl

Abstract:	We present a vision-based positioning system for unmanned aerial vehicles (UAV), which enables autonomous flight (e.g. waypoint navigation) even when GNSS is unavailable. The system uses a combination of optical flow and map matching, and runs in real-time on a small, low-power computer, which can easily be mounted on-board the UAV. Results from flight experiments show good positioning performance.
Published in:	Proceedings of the 2019 International Technical Meeting of The Institute of Navigation January 28 - 31, 2019 Hyatt Regency Reston Reston, Virginia
Pages:	856 - 861
Cite this article:	Rydell, Joakim, Bilock, Erika, Tulldahl, Michael, "Computationally Efficient Vision-based UAV Positioning," Proceedings of the 2019 International Technical Meeting of The Institute of Navigation, Reston, Virginia, January 2019, pp. 856-861. https://doi.org/10.33012/2019.16729
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In