Learning motion from images

Describes a method of determining a robot end-effector´s motion required to achieve a standard position and orientation relative to an object through learning. By using a back-propagation network, the authors establish the direct mapping from `what is seen´ to `what should be done´. The method does not need camera calibration, nor hand-eye calibration, nor explicit object model. Some general rules for correct learning are presented. A recursive scheme of movement control is designed with convergence proof. The method is simulated on an application object and shows promising application potential