A markerless human-manipulators interface using multi-sensors | Emerald Insight

To read this content please select one of the options below:

A markerless human-manipulators interface using multi-sensors

Ping Zhang (South China University of Technology, Guangzhou, China)
Xin Liu (South China University of Technology, Guangzhou, China)
Guanglong Du (South China University of Technology, Guangzhou, China)
Bin Liang (Tsinghua University, Beijing, China)
Xueqian Wang (Tsinghua University, Shenzhen, China)

Industrial Robot

ISSN: 0143-991X

Article publication date: 19 October 2015

413

Abstract

Purpose

The purpose of this paper is to present a markerless human–manipulators interface which maps the position and orientation of human end-effector (EE, the center of the palm) to those of robot EE so that the robot could copy the movement of the operator hand.

Design/methodology/approach

The tracking system of this human–manipulators interface comprises five Leap Motions (LMs) which not only makes up the narrow workspace drawback of one LM but also provides redundancies to improve the data precision. However, because of the native noises and tracking errors of the LMs, the measurement errors increase over time. To address this problem, two filter tools are integrated to obtain the relatively accurate estimation of the human EE, that is, Particle Filter for position estimation and Kalman Filter for orientation estimation. Because the operator has inherent perceptive limitations, the motions of the manipulator may be out of sync with the hand motions, so that it is hard to complete with the high performance manipulation. Therefore, in this paper, an over-damping method is adopted to improve reliability and accuracy.

Findings

A series of human–manipulators interaction experiments were carried out to verify the proposed system. Compared with the markerless and contactless methods(Kofman et al., 2007; Du and Zhang, 2015), the method described in this study is more accurate and efficient.

Originality/value

The proposed method would not hinder most natural human limb motion and allows the operator to concentrate on his/her own task, making it perform high-precision manipulations efficiently.

Keywords

Acknowledgements

The project was funded by “National Natural Science Foundation of China (Grant No:61403145)”, “China Postdoctoral Science Foundation (NO:2014M550436)” and “the Fundamental Research Funds for the Central Universities (NO:2014ZM0039)”.

Citation

Zhang, P., Liu, X., Du, G., Liang, B. and Wang, X. (2015), "A markerless human-manipulators interface using multi-sensors", Industrial Robot, Vol. 42 No. 6, pp. 544-553. https://doi.org/10.1108/IR-03-2015-0057

Publisher

:

Emerald Group Publishing Limited

Copyright © 2015, Emerald Group Publishing Limited

Related articles