Autonomous Trail Following

Masoud Hoveidar-Sefid, Michael Jenkin

2017

Abstract

Following off-road trails is somewhat more complex than following man-made roads. Trails are unstructured and typically lack standard markers that characterize roadways. Nevertheless, trails can provide an effective set of pathways for off-road navigation. Here we approach the problem of trail following by identifying trail-like regions; that is regions that are locally planar, contiguous with the robot’s current plane and which appear similar to the region in front of the robot. A multi-dimensional representation of the trail ahead is obtained by fusing information from an omnidirectional camera and a 3D LIDAR. A k-means clustering approach is taken based on this multi-dimensional signal to identify and follow off-road trails. This information is then used to compute appropriate steering commands for vehicle motion. Results are presented for over 1500 frames of video and laser scans of trails.

Paper Citation

in Harvard Style

Hoveidar-Sefid M. and Jenkin M. (2017). Autonomous Trail Following . In Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO, ISBN 978-989-758-264-6, pages 425-430. DOI: 10.5220/0006468404250430

in Bibtex Style

@conference{icinco17,
author={Masoud Hoveidar-Sefid and Michael Jenkin},
title={Autonomous Trail Following},
booktitle={Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,},
year={2017},
pages={425-430},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006468404250430},
isbn={978-989-758-264-6},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,
TI - Autonomous Trail Following
SN - 978-989-758-264-6
AU - Hoveidar-Sefid M.
AU - Jenkin M.
PY - 2017
SP - 425
EP - 430
DO - 10.5220/0006468404250430