So today as we were building our hoops, our main programmer brought up the point that the competition blackboard is clear while the plywood isn’t. And that will effect vision tracking, however because the main target is the reflective tape I believe it wouldn’t be an issue but not 100% sure.
Seeing that clear polycab is expensive vs plywood. Would this be an issue?
I used inexpensive acrylic, Plexiglass brand I think, from the hardware store. In the same section, they had small pieces of diamond plate aluminum. For under $20, I was able to replicate a small piece of the field backboard.
To answer your original question, the example code given was developed to be a bit more immune to the glare and things that will show through the backboard. Simplified processing may indeed fail more often once the background goes transparent.
The retroreflective material is flanked by opaque black tape. That helps a lot.
By the way, please don’t shoot at the acrylic backboard. Mine cracked just sliding it over the bolts. My suggestion was for vision testing only. I’d stick with plywood for the real deal. But this will allow you to verify that your code works equally well on both.
We’re considering painting the backboard grey to try to replicate the darker color of the smoked polycarbonate. Any ideas on whether this might work or whether another color might be better?
It really shouldn’t matter. Don’t paint it the same color as the LEDs, and it shouldn’t interfere. The retro material is 300 times brighter than white paint, so white paint will work, black paint will work. I used plexi to ensure that the glare and backlighting wouldn’t have a huge negative impact on the algorithm in the sample code.
Just an opinion, but polycarbonate should be the standard you use for everything pertinent. You don’t want a piece of plexi - GLASS winding up on your competition robot. it shatters on impact creating sharp edges and dulls cutting tools. hese materials are easily confused by the inexperienced.
So examining the labview code we were having issues with the code finding other rectangles, or the inside rectangle of the back board. The example code expands using the rubber band method to look for squares, does no one else think that on the polycarbonate this will be worse? especially with the aluminum bars behind the board making rectangles.
This looks like a helpful clue…use color LEDs…look for their reflection…