A CSAIL graduate student Robert Wang shows off a new system, that can tracked the motion and position of the hand using colored gloves.
The system used things like in NATAL. Namely a database of template image with associated hand skeleton. So a template matching algorithm is used to track and recognize the hand position and movement. The color is only here to decrease ambiguity cases and so give a more accurate result. It seems to work fine, but used a huge database of 40*40 pixels images. So it take a lot of places in the ram memory.
NATAL integrate a compressed version of the database contrary to this method.
Source : MIT