Biometrics for System Builders

Identity theft last year was one of the fastest growing crimes in America, according to the National Crime Prevention Council. Amazingly, the majority of ID theft takes places in the work-place. But the market for prevention of ID theft is growing fast as well—as banks, credit-card companies, e-commerce businesses, and any other company entrusted with protecting their customers’ identities, passwords and other confidential information, are clamoring for solutions to thwart ID theft.

Because ID theft security is still a fairly new market, there's opportunity for system builders to compete effectively with the major PC and device manufacturers. To do so, you must offer what the tier-one vendors don’t, namely, complete biometric security solutions.

This TechBuilder Recipe will show you how to get started in this market. After a brief primer, I’ll provide a prime example of how to install and configure a simple biometric fingerprint scanner at a reasonable price. I'll also show what to look for when selecting a biometric scanner. This will ensure that you offer your customers a system that offers the greatest possible security, reliability and ease of use.

Biometrics: Not Just for Spies and Sci-Fi Anymore

Biometrics refers to the process by which a computer peripheral can identify one person from another by using a unique metric of their body. Eye retinas, irises, voice characteristics and fingerprints are among the unique body features used by biometrics systems.

You may have seen biometric scanners in spy and science fiction movies, but in the real world, biometric technology has today become an affordable reality. Real-world IT shops commonly use fingerprint recognition, hand/palm scanners, retina scanners, voice recognition, and signature recognition systems.

While affordable consumer-based fingerprint biometric devices began appearing in late 1999, the technology initially left much to be desired. Issues with finger pressure and pigment sensitivity made the devices difficult for many to use, and the accompanying software was often difficult to use, unreliable and offered limited security features.

But now in its third generation, biometric technology has greatly matured. Reliable fingerprint scanning devices (the focus of this Recipe) are now available from several manufacturers. These devices range in price from less than $40 to more than $175.

A Fingerprint Scanner Primer

Fingerprint scanning is near bullet-proof. Given the physiological fact that everyone has unique fingerprints (even identical twins), the likelihood of someone having a fingerprint similar enough to yours to trick a fingerprint scanner is very close to impossible.

Since 1999, when the first fingerprint systems appeared, the technology has gone through three generations. With each generation have come improvements in accuracy, reliability, ease of use, and ease of installation. The older generations, still available, are cheaper than the newer ones, but they are also less effective.

Generation 1: Optical Scanning: This method works by literally taking a photograph of the user's fingerprint with a small CCD (charge-coupled device) similarly to those found in digital cameras and camcorders. The system then compares specific elements of the photo with specific elements of photos it has on file to determine if any match.

For a detailed technical description on how fingerprint scanners work, there’s a terrific primer on the How Stuff Works site.

This generation has at least two serious shortcomings. First, in theory at least, you could have an actual-size grey-scale or color picture of a fingerprint and place it into the scanner as you would an actual finger—which could fool the scanner into thinking you are the system's authorized user. Second, people who have dark skin on the pads of their fingers may find that this type of fingerprint scanner cannot take accurate photos of their fingerprints, which could lead to the system denying them access, even if they are an authorized user.

Generation 2: Capacitance Scanning: This method calls for a small electrical current to measure and determine where the ridges and valleys exist on the pad of the user's finger. In essence, it's a tiny radar system.

Because pictures are flat, (they have no depth, or ridges, like an actual fingerprint) the capacitance across the surface will have no changes and therefore, a photograph of a fingerprint will not work with this type of fingerprint scanner.

Second-generation systems have their pluses and minuses. On the plus side, they tend to be smaller and less obtrusive than optical devices. That's because they use semiconductors instead of a CCD device. On the minus side, second-generation systems can still be fooled. Thieves can use a gelatin mold of an authorized user's fingerprint (much harder to produce than a photo) to trick a capacitance-based fingerprint scanner.

Generation 3: E-Field Technology: Electric Field, or E-Field technology, looks beyond the easily obscured outer surface of the skin to the living layer below, where the unique ridge and valley patterns of the fingerprint originate. This allows proper image acquisition of dirty and difficult finger types while, at the same time, ensuring the image supplied is not a photocopy or replica of a finger.

The advantage to this system is that even if the user doesn't have fingerprints because of some bizarre accident or birth defect, the system will still be able to uniquely recognize the user's actual fingerprint which resides behind the outer layer of the skin.

An E-Field system uses a signal generator on-chip that applies a small RF signal between the finger and the adjacent semiconductor. The signal is coupled into the live conductive layer of the skin by a conductive surface (called the finger drive ring) positioned around the outside of the active imaging region of the sensor. A field is then created between the finger and the semiconductor that mimics the shape of the finger’s epidermal layer. Sensors then operate together to glean an image that accurately corresponds to the pattern of the fingerprint, which results in a more precise image than both optical or capacitance-based scanners produce.