Filter-based control of a gripper-to-object positioning movement

One of the basic operations in the manipulation of an object with a robotic gripper is the positioning of the gripper with respect to the object, so that the fingers are placed at some selected grasp points on the object when they are closed. A two-fingered gripper and a stereo pair of cameras mounted in an eye-in-hand configuration on a robot arm are considered in this paper for such a task. The guidance of this positioning movement is based on a visual servoing system in which the grasp points are used as control features. An invariant grasp description is used to track them in image space within the control loop. The development of this control system is based on a modular architecture, with which the whole system is built as a chain of basic components.