NEC Australia and vision analytics firm CrowdOptic have announced a live video streaming security system that enables real-time facial recognition of footage captured from fixed cameras and mobile camera sensors in body cams, smartphones, and drones.
CrowdOptic's technology uses triangulation to detect when two or more cameras are aimed at the same person, which, according to NEC, offers new capabilities in mobility to NEC's NeoFace solution.
NEC expects mobility to play an important role in smart city technology, and said using internet-connected mobile cameras integrated with NEC's biometrics is the "future of public safety".
"The integration with NEC's NeoFace facial-recognition software means that fixed cameras can interact with body cams and smart glasses to enhance identification," the local arm of NEC said in a statement. "This provides the ability to feed multiple perspectives of an individual to the NeoFace database, speeding and enhancing the accuracy of the facial-recognition process."
The company said the real-time identification enabled by NEC biometric technology is enhanced by mobile camera sensors that give first responders, police, and ambulance staff a clearer, real-time picture of the environment they're operating in.
In addition, NEC said CrowdOptic technology lets first responders on the ground aim smartphones, wearables, or cameras at the same target in order to triangulate its position. A command centre can then direct a drone to the target without knowing its exact location.
"For example, two fire rangers in the field can point to a plume of smoke and direct the drone to it, without the drone's operator needing to directly see or know the location," the company explained.
The new system is open initially to NEC Australia's customers.
Speaking at the NEC Advanced Recognition Systems Experience in Melbourne in October, NEC Europe head of Global Face Recognition Solutions Chris de Silva said there is no technical reason why a nation could not load its population into a watchlist and attempt to track them constantly in real time, but said such a system would throw out too many false positives.
According to de Silva, the simple reason why a system would fail is because with a large list of people to track, too many people look alike.
"We don't notice it, we don't see millions of people in one shot ... but how many times have people walked down the street following somebody that they thought was somebody they knew, only to find it isn't that person?" he told ZDNet.