How is everyone setting the angle for the shooter/camera?

[waialua359]

I just wanted to know how everyone is setting up the shooter/ camera angle?
Other than just guessing is there a accurate way?


Waialua Robotics Team 359

[DonRotolo]

Originally Posted by waialua359

I just wanted to know how everyone is setting up the shooter/ camera angle?
Other than just guessing is there a accurate way?

Yes, there is a more accurate way than guessing. It goes something like this:

1. Move the robot 8 feet from the center goal
2. Fire the shooter and change the angle until the ball goes in, through the center.
3. Note the camera's angle (what number does it report?) and the shooter angle (what number does it report?
4. Repeat for 9 feet, 10 feet, 11 feet, and so on.
5. Write software such that, if camera angle is "X", set shooter angle to "Y".

All done

Feel free to start and end at whatever distances you like - 8 feet is an example.

Also feel free to see if there is a difference when you are X feet away, but not directly in front of the goal, instead being off to the side.

This method is known as the Empirical method.

(If the 'bot isn't ready for this just yet, make up some values just for now and get it to work, you can plug in the real values later)

Good luck, too bad about the team not making it to NJ, that would've been cool.

Don

[Cuog]

you can also use the cuog_cam program that i wrote here:http://www.chiefdelphi.com/forums/papers.php?s=&action=single&paperid=589 u just add it to whatever code u have that utilizes the camera and it does the work for you.

[Jared Russell]

We don't

We have a camera and a ball shooter accurate from all over the field, but neither change in tilt.

[half geek]

Originally Posted by Abwehr

We don't

We have a camera and a ball shooter accurate from all over the field, but neither change in tilt.

Do you control the velocity?
Or maybe you send a pwm signal to magically adjust the height of the goal? That'd be really neat.

[Kevin Sevcik]

Originally Posted by Abwehr

We don't

We have a camera and a ball shooter accurate from all over the field, but neither change in tilt.

Is the camera tilted up slightly so the light can't move out of the top of the frame, or does it stay in at a reasonably short range?

[Jared Russell]

Originally Posted by Kevin Sevcik

Is the camera tilted up slightly so the light can't move out of the top of the frame, or does it stay in at a reasonably short range?

The default camera lens has a vertical field of view of about 34 degrees in our estimation. If you statically mount the camera at a given angle (say for example 28 degrees) from vertical, and mount it high (~4.5 ft up), you can see the target from 6 feet (the target is at the top of the frame) to 31 feet (the image is at the bottom of the frame). Moreover, the "my" data (y coordinate of the image centroid) is related to range by simple mathematics in the fixed tilt case.

Keeping in mind that the CMUcam2 uses S-Mount lenses, you can easily increase this operating range even more by purchasing lenses with higher fields of view, or you can adjust the entire bracket to be shorter or longer ranges by changing the tilt.