Making Interactive Whiteboards Smarter 
by Andrew Kerr
The
interactive whiteboard has been touted as a sort of educational
salvation, a piece of technology that will dynamically transform
the classroom. Interactive whiteboards basically look like
"regular" whiteboards, but you can project whatever's on your computer onto them, and your hand becomes a mouse, allowing you to double click items with two taps of your finger.
In an office in the Technology Square Research Building, Georgia Tech College of
Computing graduate student Jay Summet is conducting extensive work into interactive whiteboad
technology.
One particular area of focus for Jay is Virtual Rear Projection (VRP). To help better explain VRP, I enlisted Sizemik the flying squirrel, who usually hangs out at our Family Learning Nook, to give a short lecture
along the right-hand side of this page.
An example of
"real" rear projection is the computer monitor upon which you're reading this sentence, wherein light is emitted from behind the screen. If you wave your hand in front of your monitor you cannot block any of the light that appears on the screen, because you are not between the monitor and the light source.
But a movie projector is different. It beams light onto a surface, where an image then appears. The problem with this sort of projection is that objects, people, or (ahem) squirrels can wind up between the projector and the display surface. And that means: shadows. That's great if you want to make shadow puppets, but it can be annoying if you are blocking part of a movie that you want to see.
Although rear projection gets around a lot issues with regards to display quality, shadows, etc., Jay describes the drawbacks. "To
create a rear projection system, you could cut a hole in the wall behind the screen and take out part of the room behind it," he says, "but that's very space inefficient, and it's also pretty expensive because space costs money. Plus, you have to have a special screen."
The solution is "Virtual Rear Projection." Obviously, it's not actual "rear projection," as the "virtual" part tells you. So, a trick is somehow involved. The trick is to use multiple projectors, as opposed to a single one. The projectors are set up at angles to the screen, and each one shows exactly the same image. The images from the two projectors overlap perfectly so that you feel as though you're looking at a single image. The simple beauty of this concept is that if you should block the light from one projector, the other projector will fill in the missing portion (the filled in part is
half as bright, since each projector provides 50% of the image's total brightness, but the information is there and it's visible to the presenter).
"People who are watching can see the entire screen; they can essentially see through your shadow," Jay explains.
(As a fun aside, because there are two different projectors involved, one could theoretically create large, 3D images which could be viewed through the proper 3D glasses.)
This particular type of VRP is called "Passive" VRP. But Jay has been working on ways to create an "Active" VRP system that does even more. The word "active" is used because in this form of projection the computer plays an active role in creating an even more realistic rear projection experience.
An infrared camera mounted on the projector and aimed at the
screen tells the computer is if there is a "blob" blocking part
of the image. The computer then removes image pixels wherever
the blob is, replacing those with black pixels. So, the image does not appear on top of the presenter. Meanwhile, the computer can also increase the brightness of pixels that lie in the shadow of the blob. In Passive VRP, as you may recall, the image in a shadow is half as bright as the un-blocked image. But by cranking up the pixel brightness in the shadow area, the entire image looks uniformly bright.
While all this sounds like a lot of work for a computer to do, especially one laden with a PowerPoint presentation full of bouncing words and spinning graphics, Jay explains that the processing is done mostly on a graphics card, which frees up the computer to do its own thing. "A lot of the heavy lifting is done on the graphics card. What we're using is a $150 -$200 graphics card. I suspect in about two years all new computers will come with them built-in."
Jay's research began when SMART Technologies asked his lab to conduct research in the future of interactive whiteboads.
"Regular" interactive whiteboards are becoming increasingly common in the classroom. Regarding the massive three panel version in his own office (the idea being that it will replace the classic chalkboard that currently spans the classroom wall), Jay says, "I believe [SMART Technologies] is trying to sell [the larger interactive whiteboads] to governmental agencies, military command centers, city emergency response teams, that kind of thing. But eventually they're going to be provided for classroom use."
Georgia Performance Standards (-- Show --)
Web Resources (-- Show --)