A cool little project we’ve been working on in the Computational Imagery department of Academic Technologies is generating three-dimensional models using only a set of two dimensional photographs. The workflow usually looks somewhat like this: we take hundreds or even thousands of pictures of a room, an object, or natural landmark from all different angles and positions. Using a technology from Microsoft called Photosynth we upload the photographs to their servers to be compiled into a 3D scene that can be viewed from their web interface. A open sourced program called SynthExport has the capability to extract the point cloud data from the Photosynth and save it on your computer. It can then be opened in Meshlab to view and manipulate the points. After cleaning the clouds up, a mesh can be generated, and eventually colors from the verticals of the point clouds can be mapped on to the mesh to give it some element of color. Although we are working on creating finer quality meshes and a better way to apply color, the technique is a good work of concept, and best of all, it is entirely free!