ave a look at this project I am working on. I'm using Kinect motion and 3D scanning sensors to track the position and orientation of peoples faces. Then I use that information to change the camera perspective in virtual pinball table simulators to make it seem as if the flat image has real depth.