Whole-body modelling of people from multiview images to populate virtual worlds

In this paper a new techniques is introduced for automatically building rec ognisable, moving 3D models of individual people. A set of multiview colour images of a person is captured from the front, sides and back by one or mo re cameras. Model-based reconstruction of shape from silhouettes is used to transform a standard 3D generic humanoid model to approximate a person's s hape and anatomical structure. Realistic appearance is achieved by colour t exture mapping from the multiview images. The results show the reconstructi on of a realistic 3D facsimile of the person suitable for animation in a vi rtual world. The system is inexpensive and is reliable for large variations in shape, size and clothing. This is the first approach to achieve realist ic model capture for clothed people and automatic reconstruction of animate d models. A commercial system based on the approach has recently been used to capture thousands of models of the general public.