Implement OpenCV's Hand-Eye Calibration for Comparison

[Kazadhum]

Hello! I'm trying to "ATOM-ize" OpenCV's Hand-Eye calibration so we can compare our results with theirs. This issue is related to #891.

I've seen people implement OpenCV's calibrateHandEye method [1] [2] but I don't think this is the same as our method and, as such, is not comparable.

In this method, the problem is formulated as $AX = XB$, and the only transformation estimated is the transformation between the camera and the gripper.

I think what we want to use is the calibrateRobotWorldHandEye() method, where the $AX=ZB$ problem is addressed.

I'm currently trying to get from ATOM's dataset the necessary transformations to perform the calibrations (right now I'm going to try getting the transform from the camera to the calibration pattern).