Hi Openni Group!
I'm using openni/kinect for a skeleton-tracking application. I wonder why Openni labels the actual right hand of a user as the left hand, the actual right elbow as the left elbow, and so forth. Only if mirroring is enabled, the accurate labels are set. In my view, the left hand of a user is always the left hand, regardless of standing in front of a mirror or not. Also, I can't see that labeling a user's skeleton from his/her own viewpoint would be more difficult to code.
I assume there is a good argument for the chosen implementation? Could anyone tell me, why?
Thanks and best regards
Thomas