Sampling-based path planning for robust feature-based visual servoing

Classical image-based visual servo methods regulate error in the image space and undergo difficulties when the initial and desired robot positions are distant. It is not trivial to introduce constraints in the realized trajectories and to ensure convergence due to the nonlinearity of the system. This paper proposes a trajectory planning scheme based on Probabilistic Roadmaps (PRM) in order to achieve more robust visual servoing through the introduction of desired constraints at the task planning level such as visibility and occlusion avoidance constraints that ensure the object remains in the camera field of view (FOV). Off-line path planning is performed on a 5DOF robot arm to confirm the validity of the approach.