So this year there’s a heated debate going on among some of our team members as to where the camera should go on the robot. I believe that it would be best to have it pointed in the same direction as the shooter, and follow it if it moves Some of my other teammates however believe it would be better to put the camera on its own rotating gimbal, Separate from the shooter entirely. Thoughts on this? What are other teams doing?
Thanks!
I would assume having the camera fixed to the shooter would help with aiming shots. Unless there is a significant advantage to having it point elsewhere, keep it fixed on the shooter. Drivers rarely look at the camera feed when driving. I know our driver only used it to double check that we were aligned before shooting last year.
Putting the camera on a rotating turret sounds like a good idea, but any sort of lag time in either receiving the image or processing it will make it hard to implement a good aiming strategy. If the camera is on a relatively stationary platform you can use it to determine the desired turret direction, then use simple position feedback like a potentiometer to aim the turret quickly.
Put it on the shooter locked into place. You can get a pretty good idea of what the robot is doing by looking at it – less so for figuring out where it’s aiming. That’s where you’ll get the most benefit out of it.
We’ll try to use 2 (bandwidth testing will answer the question of whether that’s realistic) because we’re planning to be able to pick up from the floor (another difficult task to do from the driver station pov).
Thanks, guys. I’m the driver, and you’re right, I seldom look at it. We’re thinking of using a raspberry pi to do our image processing, so lag shouldn’t be so much of an issue. But keeping it on the shooter in a fixed place was my feeling, I’ll forward your thoughts to the rest of my team.