At kickoff it was stated that this year’s robots will be allowed an illumination source to detect the reflective targets on the end of the scoring pegs. Does it make sense to be able to turn the light source on/off a frame or two at a time and take the difference between the frames for targeting information? Incandescent lights are probably too slow, but an LED is easily quick enough. Also, the light source doesn’t have to be white, nor would it have to stay the same color (there are plenty of superbright red and green LEDs out there).
Any thoughts? Am I interpreting the rules correctly?
You have correctly identified several good approaches for illuminating the target. More info can be found http://decibel.ni.com/content/docs/DOC-8923 in the lower right More section.
The speed of the robot and objects in the scene defines the speed of the flash. LEDs are a great choice for many reasons, but I suspect incandescent would be fine, perhaps even using something to build a shutter to hide the light.
The other approach is to use a known light color, again LEDs are a great pure light color source.
Greg McKaskle
Thanks, Greg,
It occurred to me after I posted that taking the difference between frames assumes that the robot (the camera, actually, if it’s mounted on the end effector) is moving slowly relative to the frame rate, or else everything gets flagged as different. I hadn’t thought of a shutter, I just assumed (being an electrical guy) it’d be easier to switch (or even synch) an LED at the video frame rate.
Personally, I’d probably go with LEDs too, but I was just pointing out there are many ways to do this.
Greg McKaskle
Suer bright lights aimed at your driver are probably not a good idea.
Luckily, they don’t need to be bright. The material is rated as 600 times brighter than a white surface. Even tiny LEDs are effective if placed close to the eye or camera lens. Test it with a keychain LED or other small lights to determine for yourself.
Greg McKaskle