The ShapeWays blog interviews Qi Pan of Cambridge University and his initial go at getting a Webcam to generate 3D models. The process works like this:
1. Use a highly-textured object. (By this, I think he means any object that has sufficient contrast that the software can determine points of interest. In his example, he uses a pop [soda] can, which has a high contrast label.)
2. Rotate the can in front of the Web cam. The software notices changes, and begins to stitch together a 3D view.
The software is very rudimentary at this point. It requires a partial 3D model to begin with. He hopes to improve the software so that it can deal with objects of lower contrast.
