The company behind Scotland Yard’s facial recognition cameras is exploring new technology that would allow officers to stay one step ahead of criminals.
NEC, whose NeoFace system is being deployed across London to spot wanted offenders, is investigating how to combine facial detection with crowd analysis to predict a suspect’s most likely next moves.
A patent published in the tech giant’s native Japan reveals details of a tracking system which would then calculate possible escape routes, which can be relayed to officers on the ground.
Algorithms cross-reference “time, position (and) probability” with crowd analysis of people’s “behaviour tendencies” to predict “future behaviour”, including where they could try to run.
The “face authentication” patent explains how the most likely getaway route of a wanted burglar picked up on CCTV could be predicted using maps, local traffic and public transport information. This could be cross-referenced against “criminal psychology” data that, it says, provides extra clues based on known criminal behaviours.
The technology could also be used to predict the path of someone being tracked. NEC suggests the technology could help track stalkers following people in crowds. Scotland Yard said its current NeoFace purchase was “not about the live facial recognition system predicting future behaviour but about helping the MPS locate people”.
NEC said: “We do not wish to comment on technical details including the usage of specific patented technology.”
The Met announced on Friday that facial recognition cameras are to be deployed across London following trials. It said “watch lists” will be developed to carry out live facial checks of people suspected of serious violence, gun and knife crime and sexual exploitation.
Concerns have been raised about privacy and the risk of unfair profiling.
According to the patent, the new technology could be used, if desired, to try to detect suspicious characters not on any police “watch list”. The patent states: “There are cases in which it is desired to track or monitor a person who has not been registered in advance.
“For example, it is when you want to track and monitor a person showing suspicious behaviour at a store or airport.”
