It is important that the camera images of the calibration object are well exposed and that they are in focus. To ensure the calibration object is within the depth of field it is recommended to use higher f-number values and compensate with larger exposure times. Also, the same aperture setting should be used for all the images to avoid unnecessary fluctuations due to different aperture sizes. It is important not to choose similar views for the hand-eye calibration images, but excite all six degrees of freedom of the calibration object and that of the robot.
During the eye-to-hand calibration process, the calibration object is mounted on the robot end-effector and moves with the robot. It can be fixed directly to a flange or held by a gripper. The exact location for mounting is not important because the relative pose between the calibration object and the end-effector does not have to be known. It is important that the calibration object does not move relative to the flange or the gripper during the motion; it has to be is fixed or held tightly. It is recommended that mounting brackets, as well as the calibration plate itself, are made of rigid materials. It is also crucial that the calibration object is still during image acquisition. A good routine is to wait a couple of seconds after the robot moves with the calibration object to allow the whole structure to stabilize in case there are post motion vibrations. Robot motion should be smooth with controlled accelerations to avoid shaking and dislocation of the calibration object.
During the eye-in-hand calibration process, the calibration object is stationary, placed in robot’s workspace where the end-effector mounted camera can see it from different perspectives. The exact location of the calibration object is not important because it is not necessary to know its pose relative to the robot's base. However, the calibration object must not move during the calibration and should, therefore, be tightly fixed.
It is beneficial to ensure that the temperature during the hand-eye calibration is somewhat stable and similar to the working conditions when the system is going to be set in operation.
Choosing the correct method
Some hand-eye calibration methods involve estimating camera intrinsic parameters along with extrinsic parameters and the relative pose between the camera and the robot frame. Each Zivid camera unit goes through an extensive calibration process, a part of which includes determining the intrinsic parameters of its 2D color camera. These parameters are available from the Zivid API. Therefore, it is beneficial to use a hand-eye calibration method that allows making use of these parameters. Besides, since Zivid cameras provide 3D data, it is possible to calculate camera extrinsics from the point cloud, and this is also highly recommended. Using a hand-eye calibration method that does not use your Zivid camera’s intrinsics parameters and its ability to provide extrinsic parameters boils down to treating a Zivid camera as an uncalibrated 2D camera rather than a well-calibrated 3D camera.
Accuracy and recalibration
The picking accuracy of a vision guided-robotic system depends on the combined accuracy of the camera, hand-eye calibration, machine vision software, and robot’s positioning. Robots are in general highly repeatable but not accurate. Temperature, joint friction, payload, and manufacturing tolerances are some of the factors that cause the robot to deviate from its preprogrammed positioning. However, robot pose accuracy can be improved by calibrating the robot itself, which is highly recommended for complex systems with multiple factors that affect the picking accuracy. If the robot loses the calibration, the picking accuracy will deteriorate. Repeating the calibration (robot and/or hand-eye) can compensate for such deteriorated performance. It is also necessary to repeat the hand-eye calibration after dismounting the camera from a fixed structure or a robot and mounting it back on.
Read more about hand-eye calibration residuals.