Today’s world demands higher standards for securing public utility facilities. These buildings are often monitored remotely, without a security guard on-site. These are primarily closed drinking water intakes, transformer substations, water towers, or gas and fuel pumping stations. Due to their importance, they are particularly vulnerable to all kinds of sabotage. Additionally, a large number of false alarms, caused by animal interference or trees moved by the wind, lull security personnel into a false sense of security. GE Infrastructure has developed a unique system that helps security staff monitor protected facilities more efficiently and effectively. Seemingly, everything happens as in a traditional video monitoring system. Meanwhile, VideoIQ analyzes the image transmitted from cameras in the same way the human brain does. After installation, VideoIQ begins learning, memorizing what the surroundings look like and where stationary objects are located. It learns to ignore signals coming from the “background” of the image, such as the movements of trees or bushes swayed by the wind. Thanks to a new technology developed in GE Infrastructure laboratories called Concept Coding, the system analyzes the full spectrum of behaviors typical of humans and incorporates facial and gesture recognition algorithms. In the event that VideoIQ surrounds a monitored object on the screen with a characteristic red frame, we can be sure that a human has been recognized in the protected zone, and not another moving object, as is the case in traditional systems. Thanks to the use of an advanced data analysis and processing algorithm, the system works perfectly in conditions of limited visibility – at night and during heavy precipitation. “We have taught VideoIQ how to recognize people, not just by shape and color, but by the entire spectrum of human behavior,” says Piotr Laskowski, General Manager, GE Infrastructure.
The article comes from the magazine “e-Fakty” 2/2005.