Skip to content


New Idea: Insanely Cheap Multitouch

So I’ve been thinking about a new design for a Multitouch system. I’ve googled it a bit, and it seems original.

Right now, there are a couple popular multitouch designs. The most popular one right now is probably FTIR, or Frustrated Total Internal Reflection. This is the one used by Jeff Han in his TED demonstrations. There are several variations of FTIR, like Diffused Surface Illumination. Then there is Diffused Illumination, which powers the Microsoft Surface, and a variation of DI is Front DI (where the light source is in front) like the simple DIY MTmini system (where the light source is ambient light). The problem with FTIR, DSI, and DI is that they require the camera to be in the back of the screen. This makes it impossible to retrofit a surface.

The Wiimote tricks by Johnny Chung Lee aren’t exactly multitouch. They involve wearing special things to interact. They are interesting nonetheless, but not true multitouch. It’s virtually a completely different market, though the Wiimote IR camera may be used with the LaserTouch system theoretically instead of the camera (I was planning to try this out originally).

Laser Light Plane, or LLP is usually similarly used as the ones above. A variant of LLP is the Microsoft Research LaserTouch system (apparently used in Touchwall as well). In LLP, a laser hooked up to a line generator creates a “plane” of infrared light only millimeters above the surface. When something interacts with that plane, light is scattered in all directions. Most systems take light from the bottom, but LaserTouch looks at the light from the top. Wherever your finger touches the plane, it appears to have something like a thin halo around it.

LLP is interesting (especially the LaserTouch variant) because it allows for comparatively really cheap multitouch. The Aixiz 780nm 5-10mW laser (the one(s) most commonly used around nuigroup for LLP rigs). cost less than $10 (though normally 4 or so are used together, and buying goggles for protection from the dangerous light may cost close to a hundred, and the visible light filter is also a slight tax, along with disassembling a webcam to remove the IR filter, making it closer to the $100 estimate by microsoft).

Well, I have a relatively simple idea. You just have a very thin mirror angled just right off to the side of the surface (actually, 2 mirrors, for two coordinates). You are probably thinking that this is only going to be like the normal single-touch systems, which suffer from not being able to detect where you actually pushed when there is multiple points. Actually, the mirrors are only used to determine whether you’ve contacted the screen yet. The position is determined by some magical image processing that hasn’t been implemented yet.

So what do you think of this idea? Did I explain it enough? I’m probably gonna elaborate on this later.

Posted in MirrorTouch, Touchscreens.

Tagged with , , .


3 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

  1. lrbabe says

    We need some visual illustrations I guess.

  2. MartinJ says

    If your mirrors are top and right, you might need a row of LED’s on the bottom and left to make it easier to track from your camera/optical sensor.

    The problem is that you would need your camera to be directly behind the user, so the mirrors could reflect the touches properly. That would wreck your overall image processing.

    I think there may be some temporary solutions that help us develop the apps and UI for multitouch, but ultimately I think we’ll be able to buy a huge sheet of acrylic with a USB cable that will plug into our pc’s. Hang it on the wall, calibrate the corners, and you’re in business.

    The technology is available, it just needs to be implemented as an input device for PC’s.

  3. admin says

    Well, I think ambient light should perfectly suffice. If anything, a piece of cardboard on the left and bottom. And I think the camera could exist anywhere in front of the monitor. The angles of the mirrors can be calculated to have the camera at just about any position



Some HTML is OK

or, reply to this post via trackback.