Bounding Visual Odometry Drift with Radio Range Data for Small Unmanned Aerial Vehicles
Abstract
Vision-aided navigation for unmanned aerial vehicles (UAVs) is a well-researched topic. Global Navigation Satellite System (GNSS) information is valuable and used on a routine basis, however having multiple methods of obtaining navigation information for small UAVs is desirable. One popular method of utilizing vision information for UAV navigation is that of Visual Odometry (VO), even though without additional aides, position drift is unbounded for this approach. This study investigates the amount of additional information that is required to bound the position drift for a UAV that is using a VO algorithm at the heart of its navigation solution. The proposed method is to utilize a pair of ranging radios between a GNSS denied UAV navigating with VO and another UAV that has access to good GNSS information, giving it a known location. A flight test recreating this scenario was performed in July 2017 and the data are used to demonstrate the proposed concept. In addition to test data, an analytic observability analysis is presented.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 21, 2018
- Accession Number
- AD1219942
Entities
People
- Adam Rutkowski
- Jeremy Gray
- John Raquet
- Robert C. Leishman
- Will Myrick