The Basics Of Building A Multitouch Table

Here is a bare-bones multitouch table setup. We looked in on [Seth Sandler’s] multitouch work a few years ago when he completed the MTmini build. He’s scaling up the size a bit with the MTbiggie, and showing you how easy it is to put together. The demo rig seen above is just a couple of chairs, a sheet of acrylic, a mirror, a projector, a computer, and a diy infrared webcam.

The rig uses ambient infrared light to detect the outlines of your fingers when they touch the acrylic surface. A webcam with an exposed camera film filter feeds an image of the infrared light received below the surface to the computer. The incoming video is processed using Community Core Vision, where each individual point is isolated and mapped. Once the data is available the sky’s the limit on what you can develop. [Seth’s] demo packages include a mouse driver, some physics applications, an Angry Birds implementation, and a few others. See for yourself in the video after the break.


15 thoughts on “The Basics Of Building A Multitouch Table

  1. Nope can’t be done with a kiddie projector I bought the eyeclops projector to try to build one. It has very poor resolution and only works if the room is pitch black dark. which means it would have to be completely enclosed and used in a dark room which eliminates the ambient light.

  2. Nice setup.
    I thought the same here…
    With the lack of the ir leds, the camera can only see the shadows cast by the fingers,
    which means that this setup only works at day time or with a strong upper light source.
    IR leds will help it to work in the darkness as well.

  3. Good build. The only bad point is, as usual, that cameras are too laggy to deliver quick feedback : whether it is kinect, camera, wiimote or whatever, there is always this lag that entirely kills the whole user experience.

    I would really enjoy cameras with <20ms delay : interaction would feel much more "real". Do they even exist at reasonable prices ?

  4. @fontFiend:

    ‘an exposed camera film filter’ = 35MM film from a non-digital camera that has been exposed. Basically, take a chunk of used film and expose it – makes a visible-light filter.

  5. @ DudeGuy

    That would make sense seeing as how the video capture software uses a gradient gate to detect the outline of “blobs” or finger presses. If there was a bright back lighting the outline of your finger would be more defined. Thus, resulting in better blob detection.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.