Add color-based token recognition

[philipbelesky]

Continues on from #25, but in the context of incorporating this functionality into the current mega-component. A work in progress as part of #99. I have the color-isolation code working in a demo project, it just needs to be connected up to the image/camera feed of the Kinects and have the image mask translated into Rhino geometry.

[mariuszhermansdorfer]

Nice that you're picking this topic up. Let me know if you need my support in any way.