A New Approach to Visual Servoing in Robotics

This paper describes a new approach of vision-based control in robotics. The basic idea consists in considering a vision system as a specific sensor dedicated to a task, and included in a control servo-loop. Once the necessary modeling stage is performed, the framework becomes the one of automatic control, and naturally stability and robustness questions arise. The paper is organized as follows: in the introduction, a short state-of-the-art in the area of visual servoing is reviewed. Then, the basic concepts allowing to model the concerned interactions are given. The Interaction Screw is thus defined in a general way, and the application to images follows. Starting from the concept of task function, the general frame-work of the control is then described, and stability results are recalled. The concept of hybrid task is also presented and then applied to visual sensors. The paper ends with the presentation of several simulation and experimental results, and some guidelines for future work are drawn in the conclusion.