3d computer animation, open source technology, software application development, IT network security, and random awesomeness!

From the Blog

Mar
13
Posted by BrianH at 9:27 am

Some new research has sparked a great deal of interest for optimizing my multi-touch table and it’s design.  My last post hinted that I would start looking around for other solutions for detecting the inferred blobs (touches) on the surface of the screen.  So, I started my homework and I found a lot more then I could chew at the moment.

Before I started slapping search queries into Google, I started brainstorming of ways to detect inferred light with a compact design.  I’ve never been much of hardcore electronics geek, at least not at the soldering level.  So, much of this research is a learning experience.  I always love to think of solutions before I start looking at what others have done because it keeps a fresh perspective of the evolution of my progress.  Let’s start looking at the problem.

The multi-touch table I’m building is intended to be coffee table sized, much like the Microsoft Surface.  I really want a scalable design so that I can utilize space on the tables framework more effectively, and maintain a low cost solution.  I’ve already partially done this by replacing the projector, mirrors, and diffuser layer of the table with an LCD panel.  We covered the pro’s and con’s already on that issue.  But now I’m still left with a camera that consumes a large portion of the table’s underside due to it’s view distance needed to see the entire surface.

A solution would be to build an Inferred Sensor Matrix (or Array) that can detect touches, interpolate them and calculate touch positions based from sensor input.  So, I start searching for “IR sensor matrix”, and it looks like the NUI group has some users on the task of building one themselves.  The link in the thread also points to some valuable resources, which lead me to learning about Arduino micro-controllers.  The Arduino is used to handle sensor input and provide an interface for the sensors to the CPU (through USB).

Arduino really catches my attention for several reasons.  First, they’re completely open source.  From the software to the actual circuit boards, it’s all released under a Creative Commons license.  That alone has already got me hooked.  Second, it’s cross-platform.  As many may know, I’m a Mac fanatic, but I intend to build my table with a Linux system as well.  And Windows users, well, you can be happy too.

So, now I’m back to brain storming.  What true power do I have with the Arduino?  What other benefits could it provide to the multi-touch table, other then an interface to the IR Sensor Matrix?  Depending on the speed, I wonder how well it (if at all) it could handle processing more of the IR sensor input, taking the stress off the CPU.  I’ll have to do some more digging to find other projects using the Arduino in a multi-touch setup, but the real results will come with the actual testing.

Now, where’d I put my solder gun?

Mar
03
Posted by BrianH at 9:10 pm

Laser Light Plane (LLP) Illumination, Liquid Crystal Displays (LCDs), and Light Emitting Diodes (LEDs) may seem like a bunch of computer jargon to many, but when combined make for an impressive setup for a multi-touch table design.  Understanding how these 3 systems work has given me a much better and different perspective to designing the multi-touch table I’ve been developing over the past several months.  Since I’ve actually been ranting about LLP for some time now (without actually knowing what it was officially called), lets have a look at one of my more recent problems and how LCDs may be the solution I’ve been looking for.

Creating rear projection onto an acrylic surface requires a digital projector.  As I scavenge local computer stores and online vendors for parts to build the multi-touch table, I haven’t been able to find any projectors under $500.  This is outrageous and a huge hole to burn in my pocket.  Not only is it expensive, but it’s very difficult to design a way to project the display onto the surface the size of a coffee table.  Complex mirror arrangements are needed and make for a lot of work.  How do I shoot two birds with one stone?  I started looking for other projection methods, and even DIY (Do It Yourself) projectors to possibly elevate the cost.  It was then I realized what I needed.  The DIY projector plans I found use actual dissected LCD monitors with an overhead projector setup and lenses to create cheap digital projectors.  I never really thought about using an actual LCD monitor for displaying the UI because I never really understood how they worked.

Lots of people recognize the term LCD, especially if you’ve been browsing for a new computer or television.  You would recognize is as the “flat screen monitor” as compared to the older and bulky CRT (Cathode Ray Tube) monitors and TV’s.  But looking deeper into how LCD’s actually work has changed my perspective entirely on designing my multi-touch table.

To learn more about how LCDs work, read this: http://en.wikipedia.org/wiki/LCD

Using an LCD will completely change my table design, and relieve me from a lot of design issues and headaches.  I can find cheap LCD monitors at local computer stores (even the one I work at) and possibly cheaper through online vendors.  A 19″ LCD monitor usually sells for about $99 on sale at Staples.  Of course I’ll need a larger monitor for the final design, but a 19″ display would suffice for a prototype.  Not only is it cheaper, but it’s much more space efficient.  I no longer need to worry about fitting a computer, projector, camera, and a complex mirror setup into a coffee table sized enclosure.  The LCD display only needs one light source with a diffuser to evenly illuminate the entire display.  This is where the LED’s will fall into play.  I should be able to illuminate the LCD from it’s sides (much like laptop LCD monitors are designed).  Instead of using expensive florescent bulbs that are as bright as the sun, I can use energy efficient LED’s that will do the job just as well.

Since the LCD is a flat panel display, the only depth restrictions now will be due to the camera detecting the inferred blobs created by the LLP system.  Or, at least until I learn of a new way to detect IR without using a camera . . .