View Single Post
  #3   Spotlight this post!  
Unread 12-02-2010, 17:24
Greg McKaskle Greg McKaskle is offline
Registered User
FRC #2468 (Team NI & Appreciate)
 
Join Date: Apr 2008
Rookie Year: 2008
Location: Austin, TX
Posts: 4,752
Greg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond reputeGreg McKaskle has a reputation beyond repute
Re: Tracking Ball with Camera

If they've achieved results that amazing, they must have implemented one of those CSI filters the policemen use on TV all the time. Perhaps they can also zoom in and estimate the air pressure of the ball and determine which team's robot kicked it by using the shadows in the image .

The circle tracker will be quite slow, needs sharply defined edges, doesn't like shadows. The carpet technique sill be pretty fast and more resistant to blur. The key is to use some of the particle filters before you get the particle analysis and to look at some of the measurements that indicate how circular the particle is.

Note that this technique and the circle detection both need the camera to point downwards in order to be effective. This is pretty easy to verify simply by moving a ball around in front of the robot and viewing on the dashboard.

Greg McKaskle
Reply With Quote