Velocity Estimation for Motorcycles Using Image-to-Road Mapping
- 13 December 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3199-3205
Abstract
The authors propose a visual-inertial approach to estimate the body-fixed lateral velocity of motorcycles traveling along extra-urban roads. The approach comprises the following steps: First, a monocular camera takes video of the road ahead. Key features from sequential images of the road surface are extracted using the Harris corner detector and matching features are identified using the Fast retina keypoint descriptor. The locations of these features on the road surface are determined using a mapping based on an intuitive ray-casting approach. Next, the feature locations on the road, the angular velocity measurements and the optical flow of the feature projection locations on the image plane are used to formulate the egomotion of the motorcycle as a system of linear equations from which a velocity estimate is solved for using the least-squares method. Finally, this estimate is fused with readings from an inertial navigation system using a Kalman filter to produce a filtered estimate and correct integrator drift. The approach is validated against simulation data generated using BikeSim and the results are compared against state observer approaches and previously published visual-inertial approaches from the authors.Keywords
This publication has 19 references indexed in Scilit:
- A Geometric approach for estimating sideslip angle for Powered Two- Wheeled Vehicles2022
- Advanced vehicle technologies and road safety: A scoping review of the evidenceAccident Analysis & Prevention, 2020
- Inverse Perspective Mapping Roll Angle Estimation for Motorcycles2018
- Modern Robotics2017
- Motorcycle Dynamics2014
- FREAK: Fast Retina Keypoint2012
- Motorcycle state estimation for lateral dynamicsVehicle System Dynamics, 2012
- 1-Point-RANSAC Structure from Motion for Vehicle-Mounted Cameras by Exploiting Non-holonomic ConstraintsInternational Journal of Computer Vision, 2011
- An Inertial and Visual Sensing System for a Small Autonomous HelicopterJournal of Robotic Systems, 2004
- Inverse perspective mapping simplifies optical flow computation and obstacle detectionBiological Cybernetics, 1991