Measurement Pose Optimization for Joint Offset Calibration with a Hand-Eye Camera

A specified device or laser tracker accurately calibrates robot joints' offsets, but they require bothering setup for daily calibration. The offset calibration using a hand-eye camera and a marker makes it more convenient. However, the calibration accuracy reduces according to the limitation in pose estimation accuracy using a hand-eye camera and a marker. This paper proposes a method for optimizing the camera poses to capture the marker in joint offset calibration using a hand-eye camera to increase the calibration accuracy. We proposed an index to evaluate the effect of the error in the pose estimation based on the camera images on the offset calibration. The offset calibration with the marker measurement at the poses, which obtained the optimization to maximize the proposed index, realized higher accuracy than other approaches.