Talk:Gesture recognition

From Wikipedia, the free encyclopedia

Gesture recognition does not require '2 video stereo cameras', it can be accomplished with just one camera. Does gesture recognition need to be distinguished from gesture tracking? It would be good to see information about the various techniques that can be used (computer vision-based & accelerometers are the two I know of). Mentions of HandVu and OpenCV wouldn't go astray either.

[edit] algorithm-based implementations overviews

I believe this article could do with some additional attention by someone who knows more about the variety of strategies and methods used for the actual implementation of gesture recognition, as depending on what forms of interaction are being used, different strategies can be taken to accomplish the goals of the software. For example, a program which tries to identify when a person is pointing and what they are pointing at could take on a very different strategy than that which identifies sign language symbols. And of course, analyzing mouse or accelerometer input has distinct differences over image-based methods, though they could have similar components. Roadm (talk) 03:47, 13 February 2008 (UTC)