Reality modeling and visualization from multiple video sequences

A reality modeling and visualization system called Immersive Video uses multiple videos of an event from different perspectives to generate a 3D digital sequence of object movement. Given appropriate camera coverage, full 3D digital videos can be generated using today's technology. The potential for photorealistic videos from arbitrary perspectives exists, although the coverage and degree of realism in virtual views will ultimately be determined by the application. Many issues demand resolution before one will see practical and realistic 3D digital video, including more robust shape recovery methods, broadcast-quality virtual views, and real-time interactive rendering of long 3D sequences.