Cameras have grown smarter in recent years -- better able to recognize faces at close distances, and pick up on strange behavior from a little farther out. Go in through an out door, or leave a suspicious package behind on a train platform, for example, and you'll be spotted, quick.
But figuring out what a group of people is doing, or being able to ID a face within that group, that takes brains today's digital video software still doesn't have. U.S. Special Operations Command (SOCOM) is hoping a new research grant can begin to change that.
SOCOM just gave Colorado Springs-based Securics, Inc. a $100,000 grant to start developing programs for "Monitoring of Crowd Activities." The idea is to train cameras to find faces from afar, and to "develop new algorithms explicitly for crowd management, rather than building on the traditional intelligent video surveillance algorithms that are focused on isolated targets." Oh, and by the way: this should all happen in a small, self-contained system that takes up barely any power at all -- 7 watts, maybe.
Securics will start small, looking at algorithms for a crowd's "vertical motion energy," like a group of people "pumping its fists, or raising signs," says company chief Terry Boult.
There will also be some comparisons to how much activity is usually in the area. "If normally, on Tuesdays, there are only three people on this corner, and now there are 50, maybe there's a problem," Boult adds.
In addition, Securics will build on the work it did for Darpa, as part of the agency's "Human ID at a Distance" program. Boult says the company developed for Darpa software to identify faces from 100 to 200 feet away. The SOCOM effort, he hopes, will far surpass that.
© Copyright 2024 Military.com. All rights reserved. This article may not be republished, rebroadcast, rewritten or otherwise distributed without written permission. To reprint or license this article or any content from Military.com, please submit your request here.
