Hi, we have been trying to use the camera to track the minibot pole in order to place the robot in the right place to release the minibot. The problem is that the code is realy slow. Any suggestions on how to improve the code?
It looks like you use some pretty expensive image processing functions. I suspect they are quite accurate, but too slow to drive a robot using them.
Once you have a convolved Value plane, you can likely look through for the big transitions. I suspect the caliper and line detection are the expensive parts, especially on the cRIO.