Looking for Products and services? Then click here for our product showcase section

Featured article in the latest Security Middle East Magazine.

The face of the future

LogicaCMG director of global identity solutions Tim Best speaking to delegates in Dubai recently

The results of a recent live facial recognition trial could prove vital to the future of aviation security in the region. Ruth Edwards reports from Dubai

Cutting edge facial recognition which allows security staff to identify individuals was put through its paces in Dubai recently. Results of an innovative trial using biometric identification algorithms over CCTV to identify suspects, without their knowledge, in a public situation have proven promising, Middle East delegates were told at the Identity Forum which was held in the Emirate.

The cutting-edge facial recognition system, which was tested in several ‘live’ environments, including one of the world’s leading international airports, identified more than 85% of suspects – several of whom were disguised with facial hair and glasses. Announcing the results at the forum in Dubai earlier this month (April), Tim Best, director of global identity solutions at LogicaCMG, the technology’s developers, said that although biometric technology, such as finger printing or iris recognition to verify someone’s identity has been available for some time, the actual identification of a person is significantly harder. That is why it has had to be done manually for so long, despite the limitations.

“CCTV operators are often overwhelmed by the numbers of screens and information that is relayed to them. An average operator’s attention span, when they are deemed to be at their most focussed, is thought to be approximately 20 minutes, so on an average shift lasting several hours, the likelihood of suspects or persons being watched being missed is actually quite high,” said Best.

“And today, the new wide-bodied aircraft are carrying greater numbers of people around the world. Here, in the Middle East, they’re really racheting things up in terms of marketing the region and Dubai as prime business and tourism destinations. And in the next few years, the world’s largest airport will open to receive somewhere between 120- and 180-million passengers a year,” he added.

As passengers don’t want to be hanging around the airport in queues for extended periods while their identity is verified, and the authorities do not want potentially hazardous, overcrowded situations arising either – the faster the passengers can be processed through the airport, the better it is for everyone.

Airport trial

In the airport trial, LogicaCMG enrolled a number of volunteers’ photographs onto a database, along with a further 1,000 images. The system was then closely monitored to see if it would make any false identifications. Volunteers were asked to mix in with regular passengers and walk through selected locations. They were monitored by live CCTV feed, using biometric identification algorithms, which gave an identification score of between 0.0 and 1.0 according to characteristics matching those on a ‘suspect list’. The system successfully identified more than 85% of ‘suspects’.

“The advantages of being able to identify non-compliant people from a distance are obvious for the border control industry, but the potential for use in other situations is clearly enormous,” said Best, adding that the value of the technology is not only being recognised by airport operators, but is being seen as an invaluable tool in the fight against hooliganism at sports events and persistent shop-lifting in retail outlets.

“Facial recognition can be used in different areas of application, including the identification of undesirable individuals, such as those people who are banned from a sports stadium or blacklisted repeat offenders,” said Best.

The technology helps by automatically scanning incoming video streams, detecting faces and cross referencing them against a watch-list database of images. If a match is found, operators are alerted via the monitoring station and action can be taken. This enables the operator to focus on wanted individuals ensuring faster and more accurate response times, which could be critical in overcrowded areas such as airports.

In cooperation with Dutch soccer club, PSV Eindhoven, the city’s Justice Department and the Eindhoven Police Department, LogicaCMG conducted a pilot aimed at determining whether face recognition could contribute to recognising individuals at a distance, in a group, at the soccer stadium.

The trial used the existing infrastructure such as security cameras, lighting, police and security personnel at the PSV stadium, as well as police photographs and recordings made during previous soccer matches. Results yielded a greater than 90% success rate.

The method can be also used for ‘positive’ recognition for access control where, once recognised, a person is granted access to a specific space in, for example, an office block or industrial site. The technology can also be employed by stores, theatres, casinos, cinemas and amusement parks to enroll customers in loyalty programmes, such as discount entitlements, free access and special offers.

At the Zuidplein Shopping Centre in Rotterdam, which has more than 160 shops and in excess of 10 million visitors a year, LogicaCMG ran a trial to investigate whether facial recognition technology could help security staff track habitual shoplifters.

The facial recognition system was installed at the reception centre and connected to the shopping centre’s existing camera infrastructure. Images were picked up by the monitoring cameras and sent to the facial recognition system, which contained a database of photographs of volunteers who agreed to be registered as regular shoplifters.

Comparing the face of the person in the camera-generated image with the reference photographs stored in the database to identify the person, reliability exceeded 90%.