A visual servoing system with feature prediction using a motion observer

A visual servo is an eye-in-hand robot system that directly uses real-time visual information for control. This paper proposes a novel visual servoing system with motion observer (MOB)-based feature prediction. The position of the target object cannot be observed directly. The motion observer is designed as a minimum-order observer and is useful for high-performance servoing. The effectiveness of the proposed method is shown by simulations and experiments. © 2003 Wiley Periodicals, Inc. Electr Eng Jpn, 145(1): 67–73, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/eej.10154