Adaptive visual tracking for motions on smooth surfaces

Abstract

A robot visual servoing system using an uncalibrated fixed camera is considered. An adaptive controller is proposed for visually tracking a predefined trajectory on some unknown smooth surface. Adaptation is designed to compensate the camera uncertainties and the lack of knowledge about the surface. The key idea is to extend the use of a recently proposed hierarchy of control method to solve the adaptive control problem which now consists of a non-linear visual servoing system. Experimental results illustrate the robustness and viability of the proposed scheme.