Hand-Eye Calibration Object

The hand-eye calibration process requires using an object of known geometry that can be detected and localized from the camera image.

../../../_images/hand-eye-common-calibration-objects.png

2D calibration objects

2D patterns are most commonly used as calibration objects. OpenCV libraries use a checkerboard, while HALCON software uses a dot pattern.

3D calibration objects

Having a 3D camera enables using a 3D calibration object. An example is the asymmetrical object used in Halcon software that is straightforward for 3D printing (see image above).

However, the key for good hand-eye calibration is in utilizing 3D features of a calibration object and that can also be achieved by using a checkerboard as a 3D calibration object. For example, Zivid Hand-Eye Calibration software uses a checkerboard pattern as a calibration object, but it utilizes the 3D features.

../../../_images/infield-calibration-boards.png

Continue reading about Hand-Eye Calibration Process.