Abstract
The human body is now the ultimate security system.
Prone to losing your keys? The advent of biometrics may help: The “keys” of the future will be part of you–your finger, retina, face, or other unique physical (or even behavioral) trait. “The emerging trend is the need for ‘identity management,’” says Lawrence Hornak, biometrics professor and director of the Center for Identification Technology Research at West Virginia University.
Step into a grocery store today and you may be able to pay the bill by scanning your fingerprint; similar technology might keep you locked out of a bank's automatic teller room whose access is controlled by biometric door locks. Some manufacturers are installing fingerprint reading-devices on keyboards so that users no longer need to memorize passwords, and the U.S. government has incorporated biometric requirements–a photo and an inkless fingerprint scan–into its visa program. Enhancing convenience and security will continue to be the two main drivers of biometric technologies, Hornak says. In the United States alone, industry revenues from biometrics are expected to reach $4 billion by 2007–with around 75 percent of the demand coming from government-related investments.
If having your eyeball or hand scanned strikes you as a little sci-fi, consider that researchers are exploring even more unusual ways to recognize people with biometrics, including by an individual's lip movements, skin, the way they walk–and even by their unique odor. For a breakdown of selected biometrics, see the descriptions below and at right.
Retina
This “eye-catching” biometric is fairly reliable but intrusive. Unlike the iris, which is visible externally, the retina lies at the back of the eye, and capturing an image of it requires a person to stay still for a longer time. Unlike an iris pattern, which is stable throughout life, a retina can be altered by disease, such as glaucoma.
Hand
A device scans a hand and records thousands of data points related to dimensions of width, length, thickness, and surface area. The Defense Department has been using hand geometry biometrics (coupled with PIN identification) to regulate access to Scott Air Force Base in Illinois for over a year and says the system saves time, money, and manpower.
Gait
Most biometric identification methods rely on scanning physical features at close range. Not so with gait recognition. In 2002, researchers at Georgia Tech hoped to develop tools that, using a powerful radar system not unlike that used by police to nab speeders, could identify people from distances of up to 500 feet–simply by the way they walked. The research, which at one point was able to identify 80-95 percent of subjects by their individual gait cycles, ended when the funding did, in fall 2003.
Face
Face recognition technology got a lot of press in January 2001, when police scanned surveillance images of Super Bowl attendees, unbeknownst to them, and compared their mugs against a database of facial profiles. (No wanted criminals were identified.) Casinos also use the tech, which renders a “map” of the face's unique measurements, as a way to spot card counters.
Iris
The iris (above), with its random pattern of freckles and ridges, has more than 250 “points” of reference, compared to fewer than 60 for other biometrics, and is therefore considered the most reliable. Iris recognition technology is unaffected by users wearing glasses or colored contacts, or even by people who have had laser eye surgery.
Fingerprint
Print recognition, the oldest modern biometric, looks at the arrangement of tiny details like bifurcation and ridge endings within a print's larger features (whorls, loops, and arches). Prints are considered unique and immutable but can actually be altered through damage, and the prints of some people, like manual laborers, can wear off. Print recognition systems can be duped, too, by fake “gummy” fingers–or even a real, severed one.
Voice
Voice recognition technology is fairly pervasive–from Furbies (the talking toys) to the “operator” on the other end of the line. It digitizes a person's speech patterns and turns them into a voiceprint. Like gait recognition technology, this is a behavioral, not physiological, biometric. Voice patterns can be affected by illness, excitement, even background noise, and simple voice recognition systems can be defeated by recordings.
Blood vessel
This biometric really gets under your skin. An array of LED lights illuminates a hand or finger with near-infrared light, and a camera captures the resulting image of the unique blood vessel structure (see left). Unlike fingerprint scanning, blood vessel recognition requires a live subject.