Each of us is born with a unique barcode: our fingerprints. A combination of genes and environment help shape the individual ridges you see encircling your fingertip. Just like the plethora of factors that make up mountainous landscapes, the biological influences of a developing embryo are so complex that no two humans will carry the same pattern on their fingertips.
As any CSI-show will tell you (perhaps one lesson you’ve successfully learned from the infamously inaccurate TV shows), being able to identifying fingerprints forms a huge part of any crime scene investigation. How does the scanner generate this information, just seconds after the culprit presses the finger down? If you take a closer look, you’ll be surprised at the sophisticated steps that all have to occur within minutes.
Shining a brighter light: a lesson in optics
The main type of fingerprint scanner is called an ‘optical scanner’. As the name suggests, it works by using optics – the physics of light – to generate an image.
It does this by using a light sensor system called Charged Couple Device (CCD), which is also found in your camera. The CCD is made up of a series of metallic, cylindrical diodes called ‘photosites’. When hit with light, these photosites produce an electrical signal. It then records a tiny dot (pixel) that represents the light that hit it. Light bounces off the ridges of the fingertip. In contrast, light is absorbed into the valleys of the fingerprint. The image appears as an arrangement of dark and light pixels. The CCD system actually creates an image that is an upside-down version of your fingertip — probably bewildering for us to look at, but we’re clearly not done with this magnificent process just yet.
Picture perfect: scanner processors and quality control
The scanner processor needs to make sure the CCD produces a good quality image. It’ll checks for image clarity (no fuzziness) and rejects the image if it’s too dark or too light. If the scanner discards the image, it is most likely because the exposure time – the amount of light emitting from the LED – was not right. If this happens, the scanner adjusts the exposure time and takes another picture.
Next, the scanner checks the sharpness of the image. It does this by running a series of vertical and horizontal lines across the image. A sharp image is one with a lot of contrast: A line running perpendicular to the ridge of an image, for instance, will be made up of alternating sections of very dark and very light pixels.
Matching game: it’s all in the details
Once the scanner has a clear, sharp image it is ready to actually identify your fingerprint. It does this by comparing the image to a series of prints it has stored in its database. Rather than comparing the entire fingerprint, the scanner zooms in at very high detail. It does this by using complex algorithms to analyze specific parts of the fingerprint called minutiae. These are often where the ridge line ends or where it splits into two lines on the fingerprint. The idea is to compare the positions of these minutaes to each other. One way to think about it is to imagine you accidently veered off your hiking trail while on a camp trip in a national park. How do you figure out where you are? You’d naturally compare your position to landmarks – the position of the sun or a river for instance – to figure out which part of the park you’re in. Likewise, is how the scanner distinguishes the fingerprint by comparing the relative distance between minutiae.
If two prints share the same pattern of minutiae (for example, the same dimensions) there is high likelihood they are the same fingerprint. While fingerprint scanners are not perfect, combining it with other forms of identification (a pin number for instance) can make for a reliable security system. One day we may find them as prevalent as ATM passwords and credit card chips are today.
Your Turn: Feel like there are reasonable limitations to fingerprint scanning technology? Think this is the way to identify yourself in the future? Leave us a comment. We’d love to hear from you.