WG: [robocup-small] field size experiment

[Raul Rojas]

Hi Raul,

there are new videocameras with much more resolution than 640 x 480. Digital
video resolution will continue increasing in the next years.

The golf ball can be tracked, definitely, we already did that. It is
definitely feasible to play in an 8 by 5.5 meters field. 

Changing the ball is an option, but then a) it should be available worldwide
b) it should roll well.

We have tested many balls: squash balls are too light and too small, tennis
balls too big, street hockey balls (with water inside) are better but very
difficult to obtain. They are around twice as heavy as the golf ball.

Raul Rojas
FU Fighters team

-----Ursprüngliche Nachricht-----
Von: "Raúl E. Torres" [mailto:rtorres at ece.uprm.edu] 
Betreff: Re: [robocup-small] field size experiment


Dear Mr. Rojas:

5.5 m / 640 pixels = 8.6 mm per pixel of resolution.  That is 1 pixel to 
see the line markers, and approximately 5x5 pixels to identify the ball 
under a well uniformly illuminated field.  Markers are not problem 
because they don't move.  The ball, however, will become more a 
difficulty to me than a challenge.

I insist that changing the ball to a sand-filled ball will keep the ball 
inside the field longer.  It required  changing the robots a little bit, 
though.

Sincerely,
Raul E. Torres

Raul Rojas wrote:

>Following on the discussion about perspectives for 2006 and beyond, I
>would like to report on an experiment we did a few days ago.
>
>We put one single camera in the middle of the current field, at 3.4
>meters of height, we calibrated the vision, and were able to play with 
>the current markers and ball. We just did so as if the field had been 
>increased to twice the area (instead of 4 by 5.5 meters, 8 m long by 
>5.5 m wide).
>
>That is, we can play on twice the current area with the same vision
>system, and one camera on each field half, at 3.4 meters height. Total 
>field size is the above mentioned 8 by 5.5 meters.
>
>At our lab we do not have that playing area, but we could train on the
>field we have now, as if it was one half of the new field.
>
>I took a video of the robots playing... but I lost it, so a new video
>will have to be made.
>
>As has been pointed out many times in this list, increasing field size
>is the more elegant solution to all the energy and speed issues 
>discussed so far. And the game looks so much more interesting, with 
>ball out of bounds becoming a smaller problem (althoug it is still 
>one).
>
>Raul Rojas
>FU-Fighters Team
>
>_______________________________________________
>robocup-small mailing list
>robocup-small at cc.gatech.edu
>https://mailman.cc.gatech.edu/mailman/listinfo/robocup-small
>  
>