If you’ve seen any kind of spy movie or forensic file, you know the importance of a fingerprint.
For everything from phones and laptops to high-security prisons, fingerprint sensors are a way to verify a person’s identity and protect sensitive information. But what happens when these unique identifiers are taken from us?
Unfortunately, fingerprint identity theft is a real issue – and back in 2002, Japanese cryptographer Tsutomu Matsumoto proved just how easy it is. Using a simple technique and gelatin, Matsumoto demonstrated that fingerprint sensors would accept spoofed fingerprints at an alarmingly high rate – over 67 percent of the time.
In addition to accepting illegitimate fingerprints, these scanners have been known to reject legitimate fingerprints. If the finger is dirty, too wet or too dry, or even just flattened by age and wear, a standard scanner can’t read the print.Standard fingerprint sensors can't always differentiate between real fingers and spoofed ones, but the internal fingerprint sensor could negate this security issue. (Image courtesy of E. Auksorius and the Langevin Institute.)
For all intents and purposes, these sensors accept prints at face value and can’t distinguish between a real print and a spoofed one.
So how do we fix this problem?
A team at the Langevin Institute in Paris, France has come up with a potential solution. Researchers developed a fingerprint sensor designed to read internal fingerprints, which bear the same pattern as a visible fingerprint but lie about half a millimeter below the surface.
“In fact, the internal fingerprints serve as a ‘master template’ from which the surface regrows when it is damaged,” said Egidijus Auksorius, a postdoctoral researcher at the Langevin Institute. Auksorius worked with Claude Boccara, a professor specializing in scientific instruments, to develop the sensor.
The sensor uses a modified version of a medical imaging system called optical coherence tomography (OCT). Standard OCT systems work by analyzing the interference pattern created when a beam of light travels through a biological sample (like a finger) and is recombined with a reference beam of light.
These systems gather information in 3D format, which often means sophisticated and expensive laser systems and light detectors.
The modified version, full-field OCT (FF-OCT), was developed in the early 2000s by the same lab and generates its images in 2D, which saves imaging time and makes the process much simpler and more affordable.
An internal fingerprint will not always be exactly half a millimeter below the surface, so the team also developed a method to image fingertips at an angle. The first image taken is to determine the depth of the print and the second image is of the fingerprint itself.
While the internal fingerprint sensor will still be more expensive than a standard sensor, the team speculated that it could be especially useful in imaging problematic fingerprints or in heightening security where there is a concern.
Fingerprint identity theft is a real issue in today’s world. Modern security isn’t always able to tell the difference between a real finger and a spoofed one. Will we soon see a better alternative to this technology engineered into our devices and buildings?
For more information on the internal fingerprint sensor, read the team’s report.