Vision Tracking, now what :)

[Teamcodeorange]

Ok, so, Code Orange won the San Diego regional with the Super Nurds and team 4161. We did well but I want to optimize the robot as much as possible before championship. At San Diego I had to manually line up with the hoop using the camera feed. It took a while longer than I would have liked.

So... I would like to use some simple vision tracking turning the robot side to side (x axis of the joystick) to align with the hoop. Code Orange bought a second cRio this year so that isn't a problem. I have already calibrated the camera to be able to consistently find the x and y values from the vision tracking example. I need help telling the drivetrain to line up with the top vision target. I understand how I would need to sort through the 4 hoop's coordinates and choose the x value of the target with the highest y value. I made a proof of concept that tracks the target with a servo, but it always overshoots so I know I need a PID loop to make the drive train act how I want it to. The thing is I don't know how to do any of that.

tl;dr I need help figuring out how to make the -1 to 1 "x" values of the vision tracking .vi into something that tells the drivetrain to line up with the target. (preferably with a PID loop)

If you need my code please ask.

Thanks for your time

[Teamcodeorange]

Well, looks like I need this to be in the new posts page to get any help.

[Chris Hibner]

I don't think anyone is answering since there have been numerous threads on this topic already. Search the forums a bit and if you have any specific questions, post them.

[Alan Anderson]

Originally Posted by Teamcodeorange

tl;dr I need help figuring out how to make the -1 to 1 "x" values of the vision tracking .vi into something that tells the drivetrain to line up with the target. (preferably with a PID loop)

How would you do it as a driver controlling the robot by hand? Come up with a good answer to that, and you might be able to implement it in code so the robot can control itself.

[Brandon_L]

We are having the same problem. Ill get you up to speed with where we are right now -

It seems like you filtered out all the target info to get what you want. Now from there, look up camera specs for field of vision. We are using the axis 206, which has 48 degrees of vision. So from -1 to 1 is 48 degrees. 48/2=24, so its 24 degrees each way from the center. So what you want to do is take the X value from the camera for your target, and multiply it by 24. This will give you how many degrees you want to turn. (example - X value of 0.5 * 24 = 12 degrees).

From here, it depends if your using a gyro to sense robot turning, or encoders on your drivetrain. We tried both and got farther with the gyro but its still not great.