none
Identify joints and bones in 3D point cloud using kinect RRS feed

  • Question

  • Hi,

     Please anybody have answers for these questions:-

    1.        how to place joints and bones in 3D point cloud using kinect?
    2.        how to capture poses and gestures of human body using 3D point cloud using kinect?
    3.        how to translate skeleton into network structure in 3D space with nodes by using point cloud in kinect?

    Thanks in advance

    Tuesday, April 30, 2013 12:21 PM

All replies

  • Hi Samkisen,<o:p></o:p>

    1) A joint is a point at which "two skeleton bones are joined". This point is defined by the SkeletonPoint structure, which defines a 3D position—or point defined in meters by the three values (x,y,z)—in the skeleton space.
    With a Kinect sensor and  Kinect SDK you can easily utilize the 3d-coordinates of 20 joint:
    http://msdn.microsoft.com/en-us/library/microsoft.kinect.jointtype.aspx<o:p></o:p>

    If you utilize WPF you easily represent joints and bones. See the examples of Kinect Developer Toolkit.<o:p></o:p>

    2) You need to implement by yourself a gesture/poses recognizer or utilize a 3rd party API. In literature there are many approaches for implementing gesture recognizer and not all are easily to implement. 
    With the last SDK (1.7) you can utilize "Kinect Interactions", a combination of new controls such as "push" to select virtual objects or "grip" to pan and scroll selection.

    If you need a better overview on Skeleton tracking and poses/gestures you can read the chapter 3 of "Kinect in Motion - Audio and Visual Tracking by Example" ;)

    3) please explain better you question... I am not sure of what you mean.<o:p></o:p>

    Tuesday, April 30, 2013 1:17 PM
    Moderator