Science

OLG facial scans to help gambling addicts

Ontario casinos and slots at racetracks are getting ready to introduce a facial recognition system for people who have identified themselves as gambling addicts, hoping to help them stay out of trouble.

Ontario casinos and slots at racetracks are getting ready to introduce a facial recognition system for people who have identified themselves as gambling addicts, hoping to help them stay out of trouble.

In a project that has received the blessing of the Ontario Information and Privacy Commissioner, the Ontario Lottery and Gaming Corporation will install facial biometric video cameras at the entrance of all 27 of its slots rooms and casinos in 2011.

Starting next spring with the slots at its Woodbine Racetrack in Toronto, the OLG cameras will link to a system that alerts staff when someone registered with the service walks through the doors.

The system makes use of an emerging technology called biometric encryption that scans a person's face. When it picks out someone who has registered with the OLG service, it alerts staff and "unlocks" information provided by the registrant.

"The security team would be alerted and would talk to the person and remind them about the requirements and counsel them to leave," said OLG spokesman Rui Brum. If the person refused to leave, security staff would then issue a trespass notice, a move the person would have agreed to in advance.

While video cameras at the entrance will scan all customers, OLG said there is no danger of misuse of that information.

"OLG will not keep [video] records of anyone not self-excluded," said Brum. "They are captured and released immediately. If there is no match, they're released right away."

In addition, the system relies on an algorithm developed by University of Toronto engineer Kostas Plataniotis and his team that protects private information supplied by registrants.

Karl Martin, one of Plataniotis's former PhD students who helped develop the system, said the algorithm ensures there is no permanent link between the facial scan and a registrant's private data.

"If any data gets stolen, what they will not have is a bunch of personal information and biometrics data. They'll have biometric data that doesn't connect with anything. It's useless to anyone," said Martin, now president of KMKP Engineering, a University of Toronto spinoff company created to commercialize the system.

Provincial privacy commissioner Ann Cavoukian lauded the OLG facial recognition system earlier this month for making privacy integral, rather than an afterthought.

"This biometric encryption-based technology will offer dramatically improved privacy protection over simple facial recognition, without compromising any functionality, security or performance — the hallmarks of a privacy by design application," she said in a release.