[Niko_Riko_BODE]

Dear RoboDK,

I am currently using an Intel RealSense D455 camera for object detection, mounted on a UR5e robot. While the object detection is functioning correctly, I am facing challenges in accurately determining the camera's TCP. The program is designed to align the camera's TCP with the center of the detected object. Would performing a hand-eye calibration help resolve this issue, or are there alternative methods for accurately calibrating the camera's TCP that you would recommend?

In my current setup, I connect my computer to the UR5e control box via a UTP cable and run a Python program on the robot. After detecting the object, the program switches the TCP to another tool and executes a specific movement. I would need this whole program in URScript so I can then compensate it using my post-processor. Is this possible to achieve?

Thank you for your assistance.

[Sergei]

Yes, it's possible, and the hand-eye calibration method will allow you to calibrate your camera.

Please take a look at the related documentation topics:
https://robodk.com/doc/en/PythonAPI/examples.html#camera-hand-eye-calibration
https://robodk.com/doc/en/PythonAPI/examples.html#depth-camera-3d

Also, the Zivid company who uses RoboDK for their solutions has excellent explanatory materials:
https://support.zivid.com/en/latest/acad...oblem.html

[Niko_Riko_BODE]

Thanks for the fast answer and the attached documentation.
Best Regards