Field-of-View Constrained Visual Servoing in Manipulators with No Structural Knowledge

We consider the problem of visual servoing subject to field-of-view (FoV) constraints for uncertain eye-in-hand manipulation. We assume that no knowledge is available about the structure of the terms in the Euler-Lagrange dynamics used to model the manipulator. The Jacobian that maps the joint space to the task space is also uncertain. The visual servoing problem is solved by making use of robust nonlinear control and funnel control tools, whose stability is proven analytically.