Automating video surveillance
OCT 31, 2012
DOI: 10.1063/PT.5.026489
CNET : New computer software is being developed to monitor security camera feeds for suspicious activity. Although such video surveillance has traditionally relied on human operators, it has proven to be expensive and fallible. Using advances in machine vision, two researchers at Carnegie Mellon University, Alessandro Oltramari and Christian Lebiere, have designed an automated system that can not only recognize whether any illicit activities are taking place but also predict what’s going to happen next. It does this by calculating what humans are most likely to do based on their physical environment. The researchers presented their findings at last week’s Semantic Technology for Intelligence, Defense, and Security conference near Washington, DC.
© 2012 American Institute of Physics
