I'm trying to detect certain positions or gestures of the user using Kinect and the Kinect SDK 1.8. For this purpose I intend to use the information provided by the Joint.Position attribute. However, I can't make sense of the values I'm reading. What do they mean? Are they the distance between the joint and the sensor? What unit are they being measured with?
Joint.Position is a 3D vector which contains X, Y and Z properties. Each of these properties represents the distance in meters (along the X, Y and Z directions) of the joint from the origin of the coordinate system used.
In Microsoft Kinect SDK 1.8, the coordinate system is centered on the Kinect sensor, and the following image explains how it is defined:
Take a look to the official documentation and this post on MSDN forum if you need more details.
来源：https://stackoverflow.com/questions/32102243/what-does-joint-position-refer-to