Simultaneous Calibration and Mapping

Abstract

We present evaluation experiments of a hand-eye calibration and camera- camera calibration method, which is applicable to cases where classical calibration methods fail. As described in our earlier works, the calibration works by performing rotational movements with the robot and estimating the rotational axes by tracking fiducial markers or other static parts of the environment (if the camera is moved with the robot, as in this experiment; otherwise tracking the robot itself or markers on it with a static camera). We extend our earlier work by virtually increasing the field of view of the cameras by using a mapping approach. We compare our results with an extended classical approach for the challenging case of calibrating a compliant humanoid robot having cameras with non-overlapping fields of view. We also show another application of this method: By observing markers attached to a robot's end effector, we can calibrate the markers to each other as well as to the robot's frame of reference.