Kinect returns skeleton data in the form of SkeletonStream. We can set up either the default tracking or the seated tracking
mode using the SkeletonTrackingMode enumeration during the initialization of the skeleton stream. The process flow for capturing skeleton data will remain the same as the one we used for the color and depth data streams. We can capture the data by using either event model or polling model. The KinectSensor object has an event named SkeletonFrameReady, which fires each time new skeleton data becomes available from the sensor. Each frame of SkeletonStream produces a collection of Skeleton objects. Each Skeleton object contains the data for a series of Joint points, which are wrapped inside the JointCollection object. Each joint has its own type of tracking mode and additional information to represent the positions.
Our initial focus is to write an application that can leverage the skeleton tracking APIs provided in the Kinect for Windows...