Automatic Docking System of Fuel Filler with CAD model-based tracking and Visual Servoing Control

In this paper, a high-precision docking method is proposed with CAD model-based tracking and position-based visual servoing controller for the automatic docking system of fuel filler. The deformable template matching algorithm is used to calculate the initial frame image pose. CAD-based tracking with Moving-Edge method in virtual visual servoing framework is utilized to estimate the current pose in real time. An adaptive gain is used to reduce the convergence time of the visual servoing control law. Experiments of a 6-DOF manipulator verify the robustness and effectiveness of the algorithm.