Eye-in-hand camera calibration

Technical Questions
1

Hi everyone. I’m doing a project where a UR10 with the help of an Intel RealSense Depth camera has to grab some objects stored on a shelf. So for the vision system I am using Python and also to communicate with the robot.

My difficulties at the moment are transforming the coordinates of the vision system to the robot’s base coordinate system. What I understood is that I must find the transformation matrix (4x4) and then multiply it by the matrix of the object’s coordinates in relation to the camera. So it is something like this (and please correct me if I’m wrong):

image

So my first question is: How to find the transformation matrix by using python? Can I obtain that by using a chessboard?
And my second question is how can I get the complete tcp pose [X, Y, Z, RX, RY, RZ]

I hope you can help me :slight_smile:

2
3

https://support.zivid.com/en/latest/academy/applications/hand-eye/hand-eye-calibration-solution.html

4

Thank you for your reply @forbech However I don’t fully understand the equation of matrixes that are in the link you provide. Is it the same as the one that I put in the question? I still dont know to achieve the transformation matrix. I already saw the zivid examples. And I also don’t have robodk, I have process simulate. Sorry I am new at this

5

Eye-Hand calibration is based on this paper:

R. Y. Tsai, R. K. Lenz, et al.
A new technique for fully autonomous and efficient 3D robotics hand/eye calibration
IEEE Transactions on Robotics and Automation, 1989
Published 1 May 1988 in Computer Science

  • But it’s probably not a big help for you
6

I’ll take a look anyway @forbech Thank you!