Radar/IMU Integration Using Visual Feature Matching
Mohamed Elkholy, Mohamed Elsheikh, Naser El‐Sheimy · Proceedings of the Satellite Division's International Technical Meeting (Online)/Proceedings of the Satellite Division's International Technical Meeting (CD-ROM) · 2022
In this paper, a 360^o Frequency Modulated Continuous Wave (FMCW) Radar was integrated with IMU, nonholonomic constraints (NHC), and an odometer through a closed-loop extended Kalman filter (EKF) to compensate for the Global Navigation Satellite System (GNSS) signal outages, overcome INS drift and errors, and smooth the navigation solution. The Oriented Fast and Rotated Brief (ORB) algorithm was adopted for Radar ego-motion estimation. An open access dataset was used in the experimental work. The data was collected in a GNSS challenging environment containing trees and high buildings around the Daejeon Convention Center (DCC) in South Korea. To access the performance of the proposed methodology, simulated GNSS outages were introduced for two trajectories at three different starting outage points. The horizontal position and 3D position percentage errors range between 1.07% and 2.79% of the traveled distance during two minutes of GNSS signal outage, where the traveled distance was around 1.1 km.