printer friendly version

Putting a face to facial biometrics

Facial recognition has become mainstream. The technology and its potential has been around for years, but it’s only recently that we have seen it take off in the real world.

From Apple’s facial recognition that had all Apple fans salivating, through to new access control units that open doors after just a glimpse of one’s face, through to China’s plans to play Big Brother in a number of cities where citizens will be tracked via facial recognition through citywide surveillance cameras, it’s all the rage. But is it a reality we can rely on?

There is a vast difference between access control via facial recognition in a controlled internal environment and making important decisions based on facial recognition out in the open without lighting controls or clear shots of the individual. Yet, some say that is not a problem and today’s combination of camera technology and software can make it happen.

Hi-Tech Security Solutions decided to ask a few people what their take on facial recognition is and what the companies they represent are doing with it. We received input from two companies, representing global brands.

The first entry was from Forbatt, the local representative of Kedacom as well as NUUO. Vaughn Tempelhoff from Forbatt says the company offers two alternatives that allow clients to recognise people in public spaces.

“Kedacom has its own Recognitive Technology, and the newly acquired (a new addition to the Forbatt stable) NUUO offers NuFace. Both work with their own niche set of features and advantages.

If one is searching for a suspect or a lost child, for example, Tempelhoff says by incorporating Kedacoms HDI series NVRs, you have the ability to search for colour, direction and type via video synopsis.

Moreover, he says Kedacom’s technology requires only an image from a 2-megapixel camera, but features such as WDR are most important in generating a clean and clear facial image. “The age-old tradition of distance to target and pixel density applies to trying to identify facial features. The general rule is the more specific the camera the better.”

On the NUUO side, the software features blacklisting, age, gender and distance estimation, through which one can easily setup alarms for people residing in your blacklists. NUUO requires a facial picture with 100 x 100 pixel density, with the benefit that this is an open platform that would accept most mainstream cameras, giving one the versatility to implement facial recognition on existing sites.

The second company is Panasonic, a company that claims a core facial recognition engine with the world’s highest face recognition performance. The system can identify faces that are difficult to recognise with conventional technologies, including faces at an angle of up to 45-degrees to the left or right, or 30-degrees up or down, and those partially hidden by sunglasses.

Panasonic’s facial software features iA (intelligent Auto) mode that automatically adjusts settings for the camera to shoot optimal images best suited for face recognition. When it is used with Panasonic’s i-PRO EXTREME series network cameras installed with the ‘Best Shot License Key’ that comes bundled with the software, only the ‘Best Shots’ will be sent to the server for face recognition.

The benefit of facial recognition is enormous, people can be identified without requiring and form of identity or access control. It will allow authorities to track people wherever there is a camera – which may also lead to abuses of the technology – changing the concept of identity management forever.

For more information, contact:

• Forbatt SA, +27 (0)11 469 3598, [email protected], www.forbatt.co

• Panasonic South Africa, +27 (0)11 312 7015, [email protected], www.panasonic.com

Share this article:

Further reading: