What does your camera do?

[Qbranch]

I've seen varied responses out of different people's camera systems. After you vote, please tell what you extrapolate from the camera data (i.e. run a turret yaw, shooter pitch, shooter speed,...).

[Stuart]

fritzes out and causes a 1.5 sec delay on all controls . .

but before it did that it controlled the speed of our shooter and our turret

[JamesBrown]

Originally Posted by Qbranch

I've seen varied responses out of different people's camera systems. After you vote, please tell what you extrapolate from the camera data (i.e. run a turret yaw, shooter pitch, shooter speed,...).

Our camera controls our pan and tilt for our turret and also controls the distance we travel when using our shooting autonomous.

[Qbranch]

The poll is up now. sorry for the delay.

[Michael Corsetto]

Do a little dance...
Make a little love...
Get down tonight...
Get down tonight...

Sorry, just had to get that out.

Anyways our CMU camera controlled both pan and tilt for our shooter. When it was working, it was great, but sometimes it decided to "get jiggy with it" and refused to work properly. Good times.

"Its locked on. I think."
(waves hand in front of lens)
"Nevermind, its not locked on"

or better yet

"Searching... searching... Its got it! Wait, now its searching... searching.."

Also, we have Mark Leon quoted reffering to our robot as "The CMU cam: Its just that easy"

Mike C.

[Qbranch]

Originally Posted by Rex114

Do a little dance...
Make a little love...
Get down tonight...
Get down tonight...

Sorry, just had to get that out.

Anyways our CMU camera controlled both pan and tilt for our shooter. When it was working, it was great, but sometimes it decided to "get jiggy with it" and refused to work properly. Good times.

"Its locked on. I think."
(waves hand in front of lens)
"Nevermind, its not locked on"

or better yet

"Searching... searching... Its got it! Wait, now its searching... searching.."

Also, we have Mark Leon quoted reffering to our robot as "The CMU cam: Its just that easy"

Mike C.

During part of the build season, we had the opposite problem. I quote:

"Ok turn it on. (CLUNK, fans spin up) Searching...... ok it found the lights on the ceiling again."

But I don't mean to hurt the little guy's feelings. We have come sooo far since a few months ago. Camera works awesome now, has been treating us well since the season started and continues to.

-Q

[mechanicalbrain]

We have our camera working, we used it in conjunction with a HUD to give the driver pretty accurate alignment directions (via LEDs in the glasses).

[Wayne Doenges]

If the target was within a 40 degree window our camera would lock on. Our whole robot was the turret Once locked on, the program would take over the shooting.
The camera team was outstanding.
One of our students found a solution to the 'lock on any light source' syndrome. He programmed the camera to ONLY lock on the green light. It was so good that you could pass a lighted fluorescent work light in front of the target and the camera would ignore it. It showed up as a black band. As a safety bonus, we could only shoot with a good lock. Thus we would not shoot into the crowd or at the judges
Also, at our playing field, we had pieces of tape laid out on the floor for every 5 degrees of camera-to-target tilt. We than programmed our shooter for each increment to shoot into the center of the goal. This way the program would vary the speed of our shooter to shoot into the goal at any angle. We used a gear tooth sensor to keep the speed constant. We could shoot while moving forward and while stationary. We were working on a program to shoot while moving backwards.
Work great!

[Chriszuma]

We gave a valiant effort to get our camera working. There was a lot of code written before we came to the conclusion that it was simply not reliable enough to take control of the robot for shooting. We ended up just coding up a PID loop similar to our original plan, except using the gyroscope to hold a heading. (We even got the Radioshack Innovation in Control award for it at Palmetto)

Ideally I wanted to use the two in concert to hold a heading, and adjust that heading depending on feedback from the camera. Oh well, I can dream can't I?

[Donut]

Our camera controlled shooter speed and turret position, plus in autonomous it would control firing (since it would not fire until the turret was in position). Once we worked out all our problems with the turret wiggling we got the camera to work well enough that I shot blindly with 1 or 2 robots in the way of my view and would make over half the shots, unfortunately it took us until Saturday morning to get it working right.

[Eldarion]

We had ours working great, but mechanical issues made it difficult to test.

It took all the practice matches and the first day of competition to get it honed in, however. We had "shooter creep" (the PD loop compensating for more balls being fed through by increasing shooter speed).

[MrForbes]

We have our new programmer working on getting around to learning about the camera by next January. He figured out limit switches and ultrasound this afternoon.

[Donut]

Originally Posted by squirrel

We have our new programmer working on getting around to learning about the camera by next January. He figured out limit switches and ultrasound this afternoon.

If he needs help I'm sure he can inquire anywhere in the programming forums here, or he can contact our team or other willing bodies out here in AZ.

Did anyone have any terrible accidents with the camera this year, like tracking something in the stands and shooting bystanders?

[Qbranch]

Originally Posted by Donut

If he needs help I'm sure he can inquire anywhere in the programming forums here, or he can contact our team or other willing bodies out here in AZ.

Did anyone have any terrible accidents with the camera this year, like tracking something in the stands and shooting bystanders?

Yes, herd about it and have seen it live. No one gets hurt though, the speed limit keeps it down. Once I was in the stands and caught a ball from a robot.

Our vision and tracking system actually ended up working much better than expected. Our operator now has very little to do, since they have never used manual shooting during a match yet. Won't explain my whole program or anything, but all i can say is that the whole system working together targets very well under pressure, to the point that we can recover from a defensive autonomous hit by another team.

By the way, I'm pretty good with the camera, so If you need some help with it I'd be glad to. I will be at Atlanta if you need help there as well... I'll be sorta busy there though, I have new code I've written that needs to be tested.

1501: Are you going to atlanta? You can reply here or email me, either way's fine. Thanks.

[Nuttyman54]

well we have it working, but we ended up modifying our robot to shoot for the low goals, so we don't use it. It's currently on loan to team 1560. However LAST year we had it tracking and on the robot, and that was something to see.