Hand-Eye Calibration Using Invariant Calibrator Placed to a Robotic Arm

The development of proper hand-eye calibration procedure is essential for the application of RGB-D camera in industrial context. Here, a hand-eye calibration approach is proposed using a colored marker placed on a robotic arm as an invariant calibrator. The procedure combines several algorithms, namely color segmentation, iterative approaches and geometrical constraint approaches, to identify in accurate manner point cloud data belonging to the marker and compute the geometrical transformation between the RGB-D camera and the robotic-arm end effector reference frames. The calibration procedure has been tested over multiple robotic arm/calibrator poses to obtain a reasonable transformation matrix that in a future perspective will allow its applicability in an industrial scenario.