Research Of Virtual Human Motion Control Methods Based On Depth Map

Virtual human is the human's vivid expression in the computer digital world. With the continuous development of computer and industry technology, virtual human technology has improved very fast in recent years. Nowadays virtual human technology is applied in lots of fields, such as aeronautics and space, military affairs, medical treatment and education. With the wide spread internet and electronic commerce, people's life ask for virtual level higher and higher. It generates higher demand for virtual human technology.Virtual human technology is mainly divided into two aspects. One is virtual human model building, the other is virtual human movement simulation. With the swift development, although there's the big advancement in the virtual human technology, it still doesn't reach the standard the field of industry wants.In this paper, a new way to control virtual human movement through gaining human movement information by depth map is given. The depth map gotten by depth camera is a kind of two-dimension gray map. The information with every pixel is different from that with pixels in the normal image, like brightness, color and hue. It's the object's physical distance from the camera. There are some unreachable advantages, if we use depth map to capture and track human motion, instead of using the normal image. Use the depth camera, we can get the motion of people under any illumination situation, and we don't have to put on any additional equipment or marker. The second treatment after getting the map is much simpler, and we can find the noise is smaller, and the threshold is higher.Virtual human has been a common concern in industry and academia by its wide application. Motion control for virtual human is now the focus of virtual human technology. In this paper, by induction of the main methods, we analyzed the content and requirements of motion control for virtual human, summarized the methods'features, content of research and key technologies. Finally, we pointed out the advantages and disadvantages of each methodThen research the methods of getting and treating the depth map, and do the marginal separation by operation Sobel.Finally, use the way of combining VC++ and 3ds max, and build the human skeleton system to control virtual human movement in 3ds max. Map the human motion information from the depth map to virtual human under the 3ds max, in order to realize controlling the virtual human movement.