Feedforward-output tracking regulation control for human-in-the-loop camera systems

Platforms like gantries, booms, aircraft and submersibles are equipped with cameras. People teleoperate such platforms to capture desired views of a scene or a target. To avoid collisions with the environment and occluded views, such platforms are often equipped with redundant degrees-of-freedom. The operator must manually coordinate multiple degrees-of-freedom in order to get desired views. Tracking moving targets becomes especially tedious and often requires several highly skilled operators. Visual-servoing some degrees-of-freedom may reduce operator burden and improve tracking performance. This paper builds upon previous successes with human-in-the-loop visual-servoing by integrating the platform's dynamics with computer vision. A broadcast boom is modelled and both simulation and experimental tests contrasting trained and untrained camera operators are presented.