Posted: 
Another season of robotics is upon us again! This year’s game, Rebound Rumble, involves shooting basketballs into baskets on the opposite end of the field. On the backboard behind each basket is a rectangle made from a retro-reflective material, similar to the reflective material that bikers or runners wear while out on the roads at night, so they are more visible to cars passing by. Shining a light at the retro-reflective backboards will cause the light to be reflected directly back at the light source.
On our robot, we can utilize this by placing a light source near a camera mounted on our robot. By doing this, the backboards will show up as bright rectangles in the image acquired from our camera. Using this image and running some vision processing code with it, we are able to pick out the rectangles from the image and calculate the x,y, and z coordinates of it. This data will allow us to track the targets and position the turret on our robot so it will point to the backboards.
So, while using a camera and vision tracking is something that we have never pursued in great length before, we are always up to the challenge. Having a robot that is able to automatically track the targets will allow us to have a reliable hybrid mode, as well as provide a much easier method of scoring for the drivers. Although it will be a lot of work, having this feature will prove to be invaluable at this year’s competition.
Comments