I’m creating an interaction design project that will read the colour combinations of a rubiks cube. Is this at all possible to do with webcams and jitter. I know that it’s possible to do with single colours, but any tips or links to similar projects/code would bee very much appreciated.
you should be able to get a long ways with jit.findbounds, it’s well-suited for this. just use nine to analyze each side. However, if there are multiple squares with the same color, you’ll get the overall bounding box.
Actually, for a more or less static image of a side, you could just pick a point in each section and get the straight RGB values, may not need findbounds at all. Look to [suckah] to help determine the colors—as they are actually "seen" by the camera, that is.
Now, making an interactive virtual one with OpenGL would be really cool… especially if it solved itself!
You could use [jit.findbounds] to find the entire face of the rubik’s cube and then downsample the resolution to 3 3 and use the colour values for each of the nine squares. I really hope you intend to hook this up to some sort of robot arm for automatic, extremely fast puzzle solving!
Thanks guys, for your fantastic suggestions.
C74 RSS Feed | © Copyright Cycling '74