POSITION-BASED VISUAL SERVOING IN INDUSTRIAL MULTI-ARM ROBOTIC CELLS USING MULTIPLE CAMERAS

This paper deals with position-based visual servoing in a multi-arm robotic cell equipped with a hybrid eye-in-hand/eye-to-hand multi-camera system. The proposed approach is based on the real-time estimation of the pose of a target object using the extended Kalman filter. The data provided by all the cameras are selected by a suitable algorithm on the basis of the prediction of the object self occlusions and of the mutual occlusions caused by the robot links and tools. Only an optimal subset of image features is considered for feature extraction, thus ensuring high estimation accuracy with a computational cost independent of the number of cameras. An experimental case study is presented for the case of two industrial robots performing a vision-guided grasping task.