Self-calibration technique for on-machine spindle-mounted vision systems

On-machine measuring (OMM) systems are being more and more applied in machine tools in order to measure workpieces on the machine itself. Many of these systems are directly mounted in the machine spindle, so the measuring uncertainty is affected by clamping positioning and orientation variations, especially when integrating optical systems based on machine vision. This paper presents a self-calibration technique for vision systems by using redundant information of on machine measurements, avoiding extra mechanical anchoring or calibration means. It has been applied to a vision system with the angular placement uncertainty of a tool holder coupling being the main uncertainty contributor. A milling machine pilot case has been selected for demonstration, showing an effective self-calibration capability both in laboratory and industrial conditions.