I'm working in my thesis project, I want to know if is possible to
detect own body gestures beyond wave, click/push or the selectable
sliders. I'm planning to implement a trained classifier using the
skeleton joints to detect my own gestures, but I want to know if is
possible to do this directly with OpenNI or NITE.
Also I have been reading the posts of the group and I read something
about xnBoundingBox, so I understand this helps to bound some area in
the depth capture, but I don't understand what can I do un the volume
defined in the box, can I detect gestures in the region?, or the box
function is like a button, I mean, if I get close to the XnBoundingBox
with an articulation, can I start a process or event?
Thank you very much.