According to a university press service obtained by RIA Novosti, the program operates by analyzing complex images from conventional security cameras. The program helps recognize and focus on suspicious individuals in a crowd environment based on the manner in which they walk, their physique, clothing style, etc. Development work aimed at providing the program with these capabilities is entering its final stages.
"We have plans to test this invention, tasking it with the collection and analysis of statistics on the behavior of buyers and sellers in large shopping centers," Danishin explained. "In the future, our algorithms can be used to identify unusual behavior, affixed to surveillance cameras in airports, train stations and on public transport," he added. "This is a priority for contemporary IT security systems."
In addition to its security applications, the new program could be used for virtual reality gaming, interactive advertising in shopping centers, and to create simulators providing automatic classification of behavior for drivers, pilots, doctors, firefighters, military personnel.
