Recognizing Odd Behavior
Researchers at Johns Hopkins University, the University of Maryland, and other facilities across the country are developing behavior recognition software that, in conjunction with video surveillance systems, alerts authorities to suspicious behaviors.
This research has applications in security, anti-terrorism, and surveillance, among other sectors, and the Homeland Security Department and Pentagon have taken notice.
The software developed by Johns Hopkins allowed authorities to arrest robbery suspects during an armed robbery that occurred near the campus of Johns Hopkins in June 2006. During the incident, the Hopkins software singled out an image from one of the campus’s 89 security cameras. The software highlighted the image because it recognized that one of the suspects was engaging in suspicious behavior on the street, with aberrant movements.
In turn, the security system automatically alerted an on-duty security guard, placing the image of the suspect on the security guard’s computer screen, allowing the security guard to zoom in on the image and record a license plate number.
Abstracted by the National Law Enforcement and Corrections Technology Center (NLECTC) from the Baltimore Sun (01/05/07); Roylance, Frank D.