3D Scanning Without A 3D Scanner [Cambridge Develops Software to Create 3D Models with Regular Webcam]

Tweet

Qi Pan, a PhD engineering candidate at Cambridge University has developed software that allows users to create a fully 3D digital model with a regular, stationary webcam. No need for multiple cameras or 3D imaging lasers, as the software is able to render the model by simply having the model rotated in front of the webcam by hand.

The technology is called ProFORMA (obviously named by an engineer) and the secret to it, is that it renders the 3d model as the real life model is being collected. It seems to be highly accurate and what impresses me is that it even accurately captures the textures of the item being modeled into 3D.

The principal isn’t that much different than motion capture which has helped create realistic human motions in CGI movies and video games for some time, but unlike mo-cap, the target doesn’t have to be covered in small light-sensitive balls that help the computer and multiple cameras track the motion.

And what is equally impressive is that instead of the expensive motion-capture equipment (which can cost over $6,000 for a single motion-capture equipped glove, which doesn’t include the cost of the cameras of drivers), this is done with a single, off-the-shelf webcam, and, like I said earlier, also captures the textures.

The possible applications of this tech is pretty limitless at this point. This could seriously change the cost and effort needed for 3D modeling. Check out the full page at Cambridge below, including a YouTube video that shows the software capturing and rendering the 3D model ‘live’ (that is, as he rotates the target object).