Human Tracking and Following in Dynamic Environment for Service Robots

A. H. Adiwahono,V. B. Saputra,K. P. Ng,W. Gao,Q. Ren,B. H. Tan,T. W. Chang
DOI: https://doi.org/10.1109/tencon.2017.8228388
2017-01-01
Abstract:Robot human interactions have significant applications in dealing with day-to-day robot service. This paper presents a novel approach for human detection, tracking and following. A flexible and robust detection method using a 2D laser scanner is proposed firstly. The detection method can be adapted to various leg characteristics. The detection result is then fed to a Kalman Filter based tracking algorithm. Owing to the inherent ability of the embedded Kalman Filter, the tracking algorithm is able to track specific persons and their motion. Then, a controller is designed to follow target persons while avoiding obstacles. Finally, a service robot, namely ISERA, is adopted in the experiments for verifying the proposed approach. The experimental results show the robot can following a target person both indoor and outdoor dynamic environment.
What problem does this paper attempt to address?