[RyanCahoon]

I believe all three languages come with example vision programs, so you can check out the code there. The ScreenStepsLive documentation from 2014 also has some explanation: http://wpilib.screenstepslive.com/s/3120/m/8731

If your team still has the license code for RoboRealm that was given out in the KoP last year, I would highly recommend downloading it. At the very least, it's a great way to rapidly experiment with a bunch of different vision algorithms to give you a sense of when each is useful. You can use RoboRealm as the final vision solution, or you can convert the algorithm to something else (like NI Vision or OpenCV).