Latest Wave of Security Technology Relies on Facial Patterns

While passengers mill around a bustling airport terminal – grappling with luggage, rushing to their gates – cameras rigged with special software analyze faces in search of known criminals.

Science fiction? No. Digital face recognition technology is no longer the stuff of James Bond spy flicks and Orwellian novels. It’s another terrorist-trapping tool that is already becoming part of our post-Sept. 11, high-security reality.

"It’s the wave of the future," said technology and operations expert Andrew Rudin, CEO of Eisner Information Solutions. "The fact that 1984 has finally caught up with us and Big Brother is watching really doesn’t upset me."

The "FaceIt" software was in use this Memorial Day weekend on ferries to the Statue of Liberty and Ellis Island, after the federal government issued vague warnings of terrorist attacks on New York landmarks – including Lady Liberty.

Four U.S. airports are testing the technology, and several others overseas, like Keflavik Airport in Iceland and an undisclosed one in Europe, have already implemented it.

It’s also being used for surveillance in high-crime districts that are difficult to patrol, like the Ybor City section of Tampa, Fla. It was first unveiled along the streets of Newham, England. And police in Belgium and Australia rely on it to identify suspects of crimes.

"It will help law enforcement," said Frances Zelazny, director of corporate communications for Visionics, the New Jersey company that developed FaceIt. "It will reduce fraud, protect the public safety, prevent terrorism and protect privacy."

When its function is surveillance, the software is installed in security cameras, and identifies people using what are called the nodal points on the face, associated with bone structure and the relationship and distance between features. Those 80 points form a unique facial pattern on each person, except in the case of identical twins.

The technology compares the faces of people who pass by or stop in front of it with those in a database of criminals compiled by local or national law enforcement, depending on the venue. Visionics says it works even when subjects are moving or not facing the camera straight on.

If 14 to 22 of the nodal points are the same, the system generates the top 10 matches. The human operator scans the list and decides whether to alert authorities by sounding an alarm in a central control room or sending a wireless emergency signal through a cell phone.

"You might not catch every single person, but it still does more than a human can do," Zelazny said.

The Jersey City company says eyeglasses, beards and mustaches, hairstyle changes or even plastic surgery – unless it’s of the severe, reconstructive variety – don’t foil the system.

FaceIt is being tested at Dallas Fort Worth Airport, Fresno Airport and Palm Beach International Airport. Boston’s Logan Airport just finished its trial run with the technology.

Because most airport operations were turned over to the federal government’s new Transportation Security Administration after the Sept. 11 terrorist attacks, it’s unclear if, when or where the $1 million software will be purchased and installed permanently.

"At the end of the test, the airport can apply for funding of it or leave it up to the TSA," Zelazny said. "We’ve heard they (the airports) are pleased with the results and think it could be used as an investigation tool at security checkpoints."

Its application isn’t limited to airports. Aside from surveying passersby on the street in certain cities, it’s already being used in casinos to catch card-cheaters and in police stations in Australia and Belgium to match suspects they arrest with wanted criminals on file, according to Zelazny.

The health care and banking industries have also begun relying on the software for internal operations to prevent unauthorized staff from accessing medical or financial records. In those cases, the authorized employee’s facial code is installed on his specific computer so that only he can access it.

Other biometrics technology that identifies people by biological traits has crept into the real world, including fingerprint imprints, retina scans and voice recognition devices.

Those are mainly used for background checks during hiring or for entry into extreme-security facilities like prisons. The airline industry uses voice recognition for pilots calling in for their schedules, according to Zelazny.

Since Sept. 11, the public has adapted to the heightened security. One Long Island, N.Y., woman said she isn’t bothered by face recognition technology, as long as it’s used fairly and doesn’t get into the wrong hands.

"I am for something like that," said Cristina Barden. "We need to find a happy medium where we’re not abusing individual rights but protecting people too."

Zelazny said Visionics has tackled the right-to-privacy issue with guidelines suggesting the public be notified when FaceIt is at work, except in cases of national security.

Company officials and other experts envision a future where face recognition technology will be used everywhere from ATM machines to shopping malls.

"We all want to be protected. We want to feel safe," said Rudin. "This is where we’re heading."