Monocular Visual Odometry for Fixed Wing Small Unmanned Aircraft Systems
Abstract
Monocular visual odometry (VO) has seen significant advances in the last decade, recently producing several real-time solutions that can be used for navigation without reliance on GPS. However, the majority of current research has focused on developing and applying these algorithms on ground-based or quadrotor platforms in constrained environments. This research compares the performance of three open-source and current state-of-the-art VO algorithms on a fixed-wing SUAS through simulation and real-world flight tests in large, outdoor environments and under high-speed maneuvers. The algorithms tested are Direct Sparse Odometry, Semi-Direct VO and ORB-SLAM2 with loop closures disabled.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 21, 2019
- Accession Number
- AD1075631
Entities
People
- Kyung M. Kim
Organizations
- Air Force Institute of Technology