Having fun with Processing. Still a proof of concept, there are many things to improve.
How it works:
The application is fed live video from the webcam, processes the frame difference, detect edges, and uses the points on the edge as vertices for a building a mesh.
The result is a realtime, pseudo-3D reconstruction of the video feed.
Other more accurate techniques for doing this exist, like structured-light scanning et all, but my goal is to achieve something simple enough to be realtime and requiring only a notebook with webcam.
Made with Processing, a programming language geared to visual arts applications developed on MIT.
Learn about on processing.org/about/
Loading more stuff…
Hmm…it looks like things are taking a while to load. Try again?