Camera (Kinect Cam) Calibration routine for a KUKA iiwa Robot.
This module was designed for our invited live demo in IROS 2017, Vancouver, Canada.
- You should have a KUKA iiwa Robot.
- If you have other robots, you can change codes in the robot side, since UDP communication is universal.
- We are using AR Markers, the ROS package: ar_track_alvar, http://wiki.ros.org/ar_track_alvar
- No need to worry about how to make your KUKA iiwa robot working with ROS. We make it possible.
- If you plan to use KUKA iiwa Robot for a more complex and long duration task, you may need a standalone wrapper: our ROS_UDP_Wrapper.
- But for this simple application, we just integrate UDP Comm inside.
- That's it.
- Assume you are familiar with KUKA Sunrise JAVA Dev.
- Assume you are familiar with ROS.
- Assume you have a basic knowledge of using UDP communication.
- You should know immediately how to use the codes.
- Enjoy :)
- Yes, we do support auto-calibration.
- previously, we did calibration in 20 minutes.
- Now, it's within 5 minutes.
- And we like it VERY MUCH, especially when we are in a live demo setup.
- Assume you know what's SE(3).
- We save correspondent coordinates in "result.txt".
- e.g., [x,y,z] in camera's frame; [x,y,z] in robot's frame.
- We do have a python script that can automatically read the txt file and calculate the transformation matrix. But I can't find it. It's a python script, I'm sure you can find it elsewhere.
- As a result, I recommend My other solver.
- I plan to incorporate this C++ solver into this repository in the future when I have a particular project.