Robust and accurate global vision system for real time tracking of multiple mobile robots

Authors:
Mišel Brezak;Ivan Petrović;Edouard Ivanjko
Affiliations:
Department of Control and Computer Engineering, Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia;Department of Control and Computer Engineering, Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia;Department of Control and Computer Engineering, Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia
Venue:
Robotics and Autonomous Systems
Year:
2008

Citing 8
Cited 0

RoboCup: The Robot World Cup Initiative

AGENTS '97 Proceedings of the first international conference on Autonomous agents
A Flexible New Technique for Camera Calibration

IEEE Transactions on Pattern Analysis and Machine Intelligence
Color Processing in Digital Cameras

IEEE Micro
Real-time cooperative multi-target tracking by communicating active vision agents

Computer Vision and Image Understanding
Image Processing, Analysis, and Machine Vision

Image Processing, Analysis, and Machine Vision
Structure-based color learning on a mobile robot under changing illumination

Autonomous Robots
A calibration method for odometry of mobile robots based on the least-squares technique: theory and experimental validation

IEEE Transactions on Robotics
Demosaicing: image reconstruction from color CCD samples

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a new global vision system for tracking of multiple mobile robots. To the best knowledge of the authors it outperforms all existing global vision systems with respect to measurement precision and accuracy, high speed and real time operation and reliable tracking of large (theoretically unlimited) number of robots under light intensity changes. The originality of the proposed system lies mainly in specially designed robot marks and robots' poses measuring directly in Bayer format image delivered by the camera. These two measures enable robust pose estimation of the robots with subpixel precision, while the significant simplification of the image processing algorithms ensures tracking of many robots with very high framerates. With algorithms running on a 3 GHz Athlon 64 processor 65 robots can be tracked at 80 fps. Moreover, in order to perform a thorough analysis of the system performances related to defined requirements, we propose a new experimental procedure that can serve as a benchmark for evaluation of other systems for the same purpose.