Tuesday, March 5, 2013

A More Powerful Surveillance State?

Via nc links, MIT releases video enhancement technology to analyze invisible changes in recorded data:
Scientists at MIT have developed open-source software that can reveal details in videos that are otherwise invisible. This software — which works with videos on YouTube or DVD — can reveal the blood pumping beneath someone’s skin, with such detail that you can accurately measure their heart rate. The software could also be released for mobile devices, such as smartphones or Google Glass, so that you can see the heartbeat of those around you in real time — a boon for gamblers, FBI lie detection agents, and doctors alike.
The underlying technology used by the software is called Eulerian Video Magnification (EVM), which essentially tracks the variation of individual pixels over time — and then exaggerates those differences. As your heart pumps blood around your body your arteries swell with bright red blood, which changes the color of your skin slightly. To the human eye, no matter how long and hard you stare at your wrist or someone else’s face, you would struggle to detect a change in color (unless they blush, of course). For a computer, however, the tiniest per-pixel fluctuations (between white and slightly-redder-white, say) are easy to detect. In the case of detecting someone’s heartbeat, EVM picks up these slightly redder pixels and exaggerates them, turning them violet.
MIT originally developed the software to measure the vital signs of neonatal babies without physical contact, but as you can see in the video, there are other, far-ranging applications. Not only can EVM detect changes in color, but it can also exaggerate movements — such as a crane or building swaying in the wind, or the tiny movements made by your eyes as they scan an environment. The scientists at MIT say that their software might act as an early warning system, if the crane is swaying too much, or if a bolt is working its way loose from a machine.
In an interview with The New York Times, Michael Rubinstein, co-author of the EVM research paper presented at Siggraph last year, says he has been asked if it would be possible to run the software on Google Glass. “People wanted to be able to analyze their opponent during a poker game or blackjack and be able to know whether they’re cheating or not, just by the variation in their heart rate,” he says. The research was partly funded by Royal Dutch Shell, too, so presumably the EVM algorithm can help with oil exploration — though we’re not entirely sure how.
Wow.  The possibilities of this could be terrifying.  And it is open code.  Nothing bad can come of that.

No comments:

Post a Comment