This week I got the CV camera working with the cup detection, I also polished off ball detection some more. What ended up working the best was getting all the points that the LIDAR generates and then projecting those onto a 2d plane that is perpendicular to the gravity vector. Then I go through all the planes from 2 meters below the camera to the same plane as the camera and I use a simple circle detection algorithm to find the top of the cups. From there, extracting the location and the plane that defines the cups is trivial.
This puts me behind schedule, but I am still confident in my ability to finish everything.
Next week I will be combining ball + cup detection and getting an MVP for the launching subsystem.
