This is the basic information backing a presentation on multitouch displays given for CS-395, Multimedia Systems, Fall 2008. The document is organized as follows:
The authoritative version of this document with embedded media is available at odin.himinbi.org/classes/cs395/multitouch_survey.xhtml.
Multitouch displays are a type of computer interface that permit interaction using multiple styluses, generally fingers, applied directly to the interface surface. The most familiar form of multitouch for most users will be Apple's iPhone. Those familiar with the cinema of the early 2000's might remember a large-scale multitouch interface that played prominently in Minority Report.
The touch sensor was invented in 1971 by Dr. Sam Hurst at the University of Kentucky. The company he founded when he was awarded a patent, Elographics, later migrated the technology to a transparent panel and the touchscreen was essentially born.
The earliest touch screens used the resistive characteristics of transparent metal oxides to determine the location of a touch.
Resistive technologies are still in use today. There have been a variety of other technologies developed including utelizing the absorbtion of ultrasonic waves and measuring surface tension using strain gagues.
The most popular method in use today utilizes the capacitive properties of the human body. A grid of transparent metal oxide is etched into a surface and the electrostatic field generated by contact with a finger is used to localize the touch.
Apple uses capacitive interfaces extensively which is why the iPod and iPhone will not work without skin contact.
Capacitative interfaces require a dense array of sensors and is not practically scalable for large interfaces.
Most larger multitouch displays use a camera to visually detect the presence of multiple styluses on the surface. The projected image causes fluxuations in the visible light spectrum, making it difficult to track the stylus. To avoid this, the camera imaging the stylus generally operates in infrared spectrum. The basic setup is:
There are a variety of methods used for illuminating the stylus with infrared. The earliest was based on a physical property known as total internal reflection. Light travels at different speeds through different media. When a beam of light hits the border of two media, the top of the beam will speed up sooner than the bottom of the beam causing a bend in the beam. Past a certain angle of incidence, this will cause the beam to be reflected back into the original medium:
Frustrated total internal reflection uses a set of LEDs to fill a surface with reflecting infrared rays. When contact is made with the surface, light rays are reflected off that point of contact and shine through the other side of the material. These are then picked up buy the camera and used to drive the multitouch:
One of the disadvantages of FTIR is that the object perception is based solely on touch. In particular, this means that it is not possible to read fiducial markers (markings on objects) through the surface.
An alternative is to make the surface out of a special material called EndLighten from Plexiglass. EndLighten contains particles throughout the material that scatter light rays and cause a uniform illumination across the surface:
Using EndLighten then, objects placed on the surface not only frustrate internal reflection, but are also illuminated by the scattered light making it possible to view patterns on the surface. This is known as Distributed Surface Illumination:
EndLighten, however, is, in Nashville $450 per sheet with a 20 sheet minimum. A cheaper method is to simply point banks of LED lights at the screen in a method known as Diffused Illumination: