I’ve been researching Multitouch for the last month or so over at the NUI Group website. There is a wealth of information on how to build multitouch tables using several different approaches all of which are fully documented on the forum and wiki. In addition to this there are open source frameworks, libraries and demo applications. Using all this information and free software, a fully functioning multitouch table can be assembled for less than a grand.
Before committing to full table, I decided to build a very basic version by following this superb tutorial:
Using only a webcam, a cardboard box, some tracing paper and a picture frame, I cobbled something together in about 10 minutes.
In terms of software, I decided to use tbeta A.K.A. CCV (Community Core Vision). This picks up and analyses the shadows on the diffused surface using blob detection. These results are broadcasted using the TUIC protocol and can be translated by platforms such as Flash, Processing, openFrameworks, C++, C#, Java etc. I was suprised how simple it was to set up. Within 30 minutes I dragging/resizing pictures, finger painting and all sorts using the demos.
So, time to start creating a few toys then.
*** UPDATE ***
Just to absolutely clear, I did not create this tutorial. It was by a very talented guy called Seth Sandler. I wish I had. The guy’s had over 1.3 million hits on YouTube.
this is amazing. I haven’t looked up the hardware requirements, but http://gestureworks.com/ looks pretty neato.
Really great Work. Anybody a link to a driver and opensource solution that will run under OSX?
Thanks Dick but I didn’t do the tutorial. I thought I made that clear in the post. I’ll add an update. Don’t want people thinking I’m taking credit for the great work of others.