2D LiDAR-Based Human Pose Tracking for a Mobile Robot

Zhenyu Gao, Ze Wang, Ludovic Saint-Bauzel, Faïz Ben Amar

2023

Abstract

Human pose tracking is a practical feature for service robots, which allows the robot to predict the user’s trajectory and behavior and thus provide appropriate assistance for them. In this paper, we propose a human pose tracking method based on a knee-high 2D LiDAR mounted on the mobile robot. Inspired by human gait, a motion intention zoning, and a walking gait model are proposed to adapt to various motion patterns and achieve accurate orientation estimation. We propose a Kalman Filter-based human pose tracker that considers the leg occlusion problem and the data association of legs. We evaluate the proposed method’s performance in various complex scenarios and demonstrate robustness to leg occlusion. We released our implementation as open-source code∗ .

Paper Citation

in Harvard Style

Gao Z., Wang Z., Saint-Bauzel L. and Ben Amar F. (2023). 2D LiDAR-Based Human Pose Tracking for a Mobile Robot. In Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO; ISBN 978-989-758-670-5, SciTePress, pages 511-519. DOI: 10.5220/0012255600003543

in Bibtex Style

@conference{icinco23,
author={Zhenyu Gao and Ze Wang and Ludovic Saint-Bauzel and Faïz Ben Amar},
title={2D LiDAR-Based Human Pose Tracking for a Mobile Robot},
booktitle={Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO},
year={2023},
pages={511-519},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012255600003543},
isbn={978-989-758-670-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO
TI - 2D LiDAR-Based Human Pose Tracking for a Mobile Robot
SN - 978-989-758-670-5
AU - Gao Z.
AU - Wang Z.
AU - Saint-Bauzel L.
AU - Ben Amar F.
PY - 2023
SP - 511
EP - 519
DO - 10.5220/0012255600003543
PB - SciTePress