Mulittouch Table - The Beginning

So I have been doing a ton of research over the last few months and learning anything and everything I can about the future and current progress in mulit-touch hardware. In 2005, Jeff Han brought the concept of personal, cheap multitouch to the masses, using much cheaper components for larger applications.

The main parts are:

• infrared light source
• web camera modified to see only IR
• special software to turn reflected IR light into points of contact with the surface being touched.

Other than that, there are more specific things needed depending on what setup you'd like to create.

One of the main contributors to the movement (Seth S.) just graduated from UCSD and seeing his table last spring was a huge influence into making my own. Keep in mind that popular companies such as Microsoft have created professional setups to demonstrate the technology, as seen in the Microsoft Surface. The idea though is to make it even cheaper.

For a screen, two types of "screens" can be used; either a projector or a gutted LCD. Since I'm a starving college student and can only afford used LCD panels vs a used computer projector, i chose LCD.

My setup is going to be a Front Diffused Illumination Multitouch Box. Here is a short breakdown of what I mean (can also be seen here) under Front DI:

• A LCD crystal is placed flat on top of a box.
• The backlight from the LCD is placed at the bottom of the box facing up towards the LCD to illuminate it.
• A web camera for a computer is pointed straight up to "look at" the underside of the LCD.
• A piece of vellum or other type of diffuser paper is placed under the LCD to diffuse the light and evenly light up the LCD display.
  
• Two pieces of Acrylic sandwich the vellum and the LCD, which all sit on top of the box.
• Infrared light via LED rails (LEDs arranged in rows) are placed on top of the top-most layer of acrylic, arranged so that they shine across the area that is occupied by the LCD.

When you place you fingers on top of the LCD, the LEDs light up your finger and makes them show up as white "blobs". Yes blob is the technical term. So the fingers stop the light, and the light travels down to the camera. Software on the computer that the web camera is hooked to turns those blobs into points of touch, and thus allows you to control things.

**My method is more accurately called LED-LP (LED light plane), or the Gap Method...but no one has come up with an official name for it. (It works on the same principle as the Front DI, but a little differently)**

These points of contact can control all sorts of neat stuff. Many people have created really cool applications that work off of multitouch gestures.
For example:
Nasa World Wind control (Like Google Earth with Multitouch):

Interactive Bars:

Giant Interactive Walls for Marketing:

Play Computer Video Games:

Music Beat Mixing (This video demonstrates a concept using blocks with fiducials (patterns) on the bottom that the camera sees and recognizes when they are placed on the table or when they are turned/moved. These movements cause changes in the sound that is produced in this setup)


My goal it to make a table and then work on learning flash so that I can create some neat programs to be used in our livingroom. Such things as easy media browsers, pretty music/video/picture player GUIs, fiducials recognization, etc.