This paper explores the use of radar odometry for Global Positioning System-denied navigation. The range progression from arbitrary ground-based point scatterers is used to estimate an unmanned aerial vehicle's relative motion. In high clutter environments, the recursive-RANSAC algorithm provides robust and efficient feature identification, data association, and tracking. The tracked feature range measurements are integrated with inertial measurement unit measurements in an extended Kalman filter. Real flight data from a cluttered environment are used to validate the results.
Radar odometry with recursive-RANSAC
IEEE Transactions on Aerospace and Electronic Systems ; 52 , 4 ; 1618-1630
2016-08-01
1330176 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Combining Edge and One-Point RANSAC Algorithm to Estimate Visual Odometry
BASE | 2018
|Real-Time Monocular Visual Odometry for On-Road Vehicles with 1-Point RANSAC
BASE | 2011
|British Library Online Contents | 2017
|British Library Conference Proceedings | 2010
|