Innovative Artificial Intelligence System for Monitoring CCTV Footage

Story link: Innovative Artificial Intelligence System for Monitoring CCTV Footage by Brian Turner

Rama Chellappa, a professor at the University of Maryland’s A. James Clark School of Engineering, is developing a real-time computer monitoring system to monitor surveillance images and detect suspicious movements or individuals. The system alerts human security personnel if it detects any suspicious activity.

The use of surveillance cameras is growing rapidly, but using them to prevent or stop crime means that they must be constantly monitored. Paying people to monitor surveillance cameras is expensive and people tend to become bored and loose concentration when watching a security TV monitor for an extended period of time.

Creating software programs that can recognize suspicious activities or suspect individuals has proved to be very difficult, but Mr Chellappa demonstrated a system for military leaders at the 25th Army Science Conference.

Using video data from digital surveillance cameras and corresponding algorithms, Chellappa has developed a digital signature for characterising human gait and other activities such as carrying a backpack or briefcase.

When a person’s limbs are allowed to move freely, gait movements are symmetrical and can be represented as a twisted helical pattern. The pattern, which varies slightly from person to person, has been named ‘human gait DNA’. A person’s gait pattern is changed by any activity, such as carrying a parcel.

By defining these signatures, the system can recognise unique patterns in human gait and detect asymmetric movements, such as an individual walking with a hidden object tied to an ankle or wrist. Hidden objects secured in such a way that movement symmetry is unaffected, cannot currently be detected by this technology.

Mr Chellappa and his team integrated human gait DNA into a real-time video surveillance system and used it to study and locate pedestrians. The system has performed effectively under a range of experimental conditions. They have also developed advanced face recognition software that can be combined with the gait recognition technology. This can be used to watch for known terrorists, spies or criminals.