[Research] My first DIY multitouch tabletop
Last update: 24-Jan-2016
One of the reasons why I wanted to pursue a PhD is I'm fascinated by the use of large interactive surfaces as gateways to information. Not only does the larger size affords more content to be shown, it also allows novel ways of interaction for individuals and groups. Multitouch is one of the many ways that became very popular amongst researchers and hobbyists, as it naturally supports direct interaction (physical actions closely mapped to command, and in close proximity), and the components to achieve that are readily available in the consumer market.
In 2005, Jeff Han, a consultant for the NYU's CS department, published a paper titled: Low-cost multi-touch sensing through frustrated total internal reflection [1], which uses a number of off-the-shelf items and an infra-red (IR) camera to detect multitouch on a plexiglass. The basic idea is when a finger touches the sheet of plexiglass, it changes the refractive index at that area, causing any light traveling inside the glass to refract out of the glass on the other side. The refraction (acting like a reflection) is picked up by the IR camera at the back and translated into a touch event. Here are the basic ingredients:
The whole setup costs less than USD1000. And after his Jeff's talk at TED a lot of people followed his suit and made their own multitouch surfaces.
At that time there were no other affordable solutions (apart from Diamond-touch and Smart-table, which were either expensive or too small). So when my lab, Collaborative Systems Lab, decided to conduct some research on multitouch tabletop, we had to build our own. We had all the components except a table, and we wanted to make a coffee table. There were two main challenges:
One of the reasons why I wanted to pursue a PhD is I'm fascinated by the use of large interactive surfaces as gateways to information. Not only does the larger size affords more content to be shown, it also allows novel ways of interaction for individuals and groups. Multitouch is one of the many ways that became very popular amongst researchers and hobbyists, as it naturally supports direct interaction (physical actions closely mapped to command, and in close proximity), and the components to achieve that are readily available in the consumer market.
In 2005, Jeff Han, a consultant for the NYU's CS department, published a paper titled: Low-cost multi-touch sensing through frustrated total internal reflection [1], which uses a number of off-the-shelf items and an infra-red (IR) camera to detect multitouch on a plexiglass. The basic idea is when a finger touches the sheet of plexiglass, it changes the refractive index at that area, causing any light traveling inside the glass to refract out of the glass on the other side. The refraction (acting like a reflection) is picked up by the IR camera at the back and translated into a touch event. Here are the basic ingredients:
- a sheet of plexiglass, typically goes from 23" to 50" diagonally. Sides are filed to completely transparent for light to shine in
- a compliant material for the refraction to occur, typically a silicon coating
- a strip of IR leds attached to the sides of the plexiglass to produce the IR light
- an IR camera capturing the refracted IR light from a touch
- a computer taking the raw images from the IR camera and translate them into touch maps
- optional projector, and projection sheet on the plexiglass to superimpose the content
The whole setup costs less than USD1000. And after his Jeff's talk at TED a lot of people followed his suit and made their own multitouch surfaces.
At that time there were no other affordable solutions (apart from Diamond-touch and Smart-table, which were either expensive or too small). So when my lab, Collaborative Systems Lab, decided to conduct some research on multitouch tabletop, we had to build our own. We had all the components except a table, and we wanted to make a coffee table. There were two main challenges:
- A table with the centre hollowed out to house the plexiglass did not exist
- The projector needed at least 1.5m to project a reasonably sized screen, but a coffee table is less than 1m high
It worked as intended for some time, but as shown in the photos the projector took one side of the table, and the bottom of the table was off limits for a clear view for the IR camera and the projection -- not ideal for coffee tables. So we mostly used this setup for quick demos and tests.
Later on the technology caught up and a new breed of multitouch sensing technology came out, which squeezed all the necessary components into a frame that can be placed on top of any monitors or TVs. This eliminates the need for back space and is less susceptible to natural light, which also has IR in the spectrum. From then on we simply buy one from a manufacturer (PQ Labs) with a matching size to the TV we have.
Nevertheless, this little project let me realize the ins and outs of a multitouch setup, the pros and cons of various technologies. It's also great that I got to design and build something concrete, which I always enjoy.
Later on the technology caught up and a new breed of multitouch sensing technology came out, which squeezed all the necessary components into a frame that can be placed on top of any monitors or TVs. This eliminates the need for back space and is less susceptible to natural light, which also has IR in the spectrum. From then on we simply buy one from a manufacturer (PQ Labs) with a matching size to the TV we have.
Nevertheless, this little project let me realize the ins and outs of a multitouch setup, the pros and cons of various technologies. It's also great that I got to design and build something concrete, which I always enjoy.
[1] Jefferson Y. Han. 2005. Low-cost multi-touch sensing through frustrated total internal reflection. In Proceedings of the 18th annual ACM symposium on User interface software and technology (UIST '05). ACM, New York, NY, USA, 115-118. DOI=http://dx.doi.org/10.1145/1095034.1095054