Author:
Rahok Sam Ann,Oneda Hirohisa,Nakayama Taichi,Inoue Kazumichi,Osawa Shigeji,Tanaka Akio,Ozaki Koichi, , ,
Abstract
Scan matching is one of the most reliable localization methods for mobile robots in known environments. However, an unexpected shift in posture remains its major issue. A method that uses an environmental magnetic field, a magnetic field that occurs in the environment, is presented to address this issue. The environmental magnetic field, which mostly refers to the geomagnetic field, is rarely changed by time. This unique property provides a means to enhance scan matching to provide a more robust localization method by using it to compensate the mobile robot’s pose. In this study, we describe how to compensate the mobile robot’s pose with the environmental magnetic field. Through experiments, we show that a mobile robot with the proposed method can recover, even if irregular changes in posture occur during the navigation.
Publisher
Fuji Technology Press Ltd.
Subject
Electrical and Electronic Engineering,General Computer Science
Reference11 articles.
1. F. Gustafsson, F. Gunnarsson, N. Bergman, U. Forssell, J. Jansson, R. Karlsson, and P.-J. Nordlund, “Particle Filter for Positioning, Navigation, and Tracking,” IEEE Trans. Signal Process, Vol.50, No.2, pp. 425-437, 2002.
2. S. Thrun, “Particle Filters in Robotics,” UAI’02 Proc. of the 18th Conf. on Uncertainty in Artificial Intelligence, pp. 511-518, 2002.
3. M. D. Breitenstein, F. Reichlin, B. Leibe, E. Koller-Meier, and L. Van Gool, “Robust Tracking-by-detection using a detector confidence particle filter,” Proc. of IEEE Int. Conf. on Computer Vision, pp. 1515-1522, 2009.
4. S. Thrun, D. Fox, W. Burgard, and F. Dellaert, “Robust Monte Carlo Localization for Mobile Robots,” Artificial Intelligence J., Vol.128, pp. 99-141, 2011.
5. S. Kristensen and P. Jensfelt, “An Experimental Comparison of Localisation Methods, the MHL Sessions,” Proc. of IEEE/RSJ Int. Conf. on Intelligent Robots and Systems, pp. 992-997, 2003.
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献