This lecture presents a system that takes as input the video image from a webcam [monocular, RGB color]. The webcam is pointed at the user; and as the user moves his or her body, the system parses the video image and uses the motions to control a game character. Ideally, with systems like this one we can create games that impose less of a control barrier between users and their intentions (the usual barrier being keyboard, mouse, gamepad, and other traditional controllers). Because computer vision is an open area of research, there are limitations to when this technique can be used. This session discusses the constraints one should follow to design a control scheme that is not hampered by these limitations.