Visual Servoing

 

The use of visual feedback for closed loop control of a robot motion termed visual servoing has received a significant amount of attention during the last two decades. Most of the key research problem have been related to the performance of visual servoing methods in the presence of measurement and system modeling errors. As a result, a number of variations and hybrid visual servoing approaches have been proposed in the literature to cope with the inherit problems of image and position based visual servoing. Specificc problems such as the effect of camera calibration errors have been studied in. The convergence properties of the control part of the systems are known for most cases.

While the convergence of the system is an essential performance property, it does not reveal much about the generated robot trajectory and its uncertainty. The procedures of camera calibration have improved enormously over the last decade. However, even perfect calibration does not overcome the restriction of the image resolution and the imaging process causes an uncertainty in the control. Motivated by this fact, in our work we have proposed the use of error propagation in the analysis and comparison of different types of visual servoing methods, i.e., position-based and hybrid. Other work related to visual servoing is cited below.

Related Publications

Measurement Errors in Visual Servoing
(V. Kyrki, D. Kragic, H. Christensen)
Robotics and Autonomous Systems (in press)

Nonholonomic epipolar visual servoing
(G. Lopez-Nicolas, C. Sagues, J.J. Guerrero, D. Kragic and P. Jensfelt)
In IEEE International Conference on Robotics and Automation 2006. ICRA'06, Orlando, USA

New Shortest-Path Approaches to Visual Servoing
(V. Kyrki, D. Kragic and H. I. Christensen)
In IEEE/RSJ International Conference on Intelligent Robots and Systems , 2004. IROS'04, Sendai, Japan

Measurement Errors in Visul Servoing
(V. Kyrki, D. Kragic and H. Christensen)
In IEEE International Conference on Robotics and Automation , 2004. ICRA'04, New Orleans, USA

Robust Visual Servoing
(Danica Kragic, Henrik I Christensen)
International Journal of Robotics Research, Volume 22, Number 10-11, pp 923-939, Oct-Nov 2003.

Visually guided manipulation tasks
(Danica Kragic, Lars Petersson and Henrik I Christensen)
Robotics and Autonomous Systems, vol. 40, Issue 2-3, pp.193-203, August, 2002.

A Coarsely Calibrated Vision System for 3D Grasping
(Danica Kragic and Henrik I. Christensen)
In Masoud Mohammadian (Ed.), New Frontier in Computational Intelligence and its Applications,
pp 283-292, ISBN 90 5199 476 I, IOS Press

A Framework for Visual Servoing Tasks
(Danica Kragic and Henrik I Christensen)
In Intelligent Autonomous Systems 6, IAS-6, E. Pagello et al. (Eds.),
pp 835-842, July 2000, Venice, Italy

Robust Vision for Manipulation and Mobility
(Henrik I. Christensen and Danica Kragic)
In Proc. 23rd OAGM/AAPR Workshop,
Ed. M. Vincze, pp 1-22, May 1999. Steyr, Austria

Using a Redundant Coarsely Calibrated Vision System for 3D Grasping
(Danica Kragic and Henrik I. Christensen)
In Computational Intelligence for Modelling, Control and Automation , CIMCA'99,
vol. Intelligent Image Processing, Data Analysis and Information Retrieval,
pp 91-97, February 1999. Wien, Austria


Back to research