🔐How Fingerprint Recognition Works?
Fingerprint System Working Internals
Hi, I’m Wajid Khan. I am trying to explain computer stuff in a simple and engaging manner, so that even non-techies can easily understand, and delivered to your inbox weekly. Join me on an under-the-hood tech journey.
The process of creating a digital image of a fingerprint involves the use of a special-purpose scanner, which illuminates the fingertip with a light source.
With fingerprints, the raised areas known as ridges reflect a greater amount of light compared to the valleys that separate them.
Fingerprint Identification Core Structure
Fingerprint Identification (FID) software analyzes the ridges by searching for two types of specific features. One is the core, or center, of the print. The other is minutia-points at which ridges end or divide.
The FID calculates the distances and angles among the minutia. Even if a finger is off center or rotated during the scan, the relationships among the minutia, shown here in diagram, do not change vastly and are complex enough to be unique. Ridges and minutia are used to perform one of two possible identifications, one-to-one or one-to-many.
A one-to-one search occurs in an identity verification system (IVS), which is used for security purposes. The IVS knows in advance who you claim to be, and compares the pattern of minutia of your freshly scanned fingerprint only with a known record of the fingerprint stored in the computer’s database.
Because not all minutia may be captured due to a cut or other aberration, the best the software can do is figure out the probability that the print matches some other fingerprint. When a known print matches within an acceptable margin, the software permits access to computer, locked door, or whatever it’s guarding.
The best known example of a one-to-many search is performed by the AFIS (Automated Fingerprint Identification System) used by law enforcement. Also called a cold search, it is most commonly used to compare a fingerprint from an unknown person, such as one left at a crime scene, against a database of known persons and fingerprints, to determine whether the person is already in the database.
Fingerprint Identification one-to-many search
To narrow the number of prints, the unknown must be compared against the FID first counts the ridges in one direction from the core to the rim of the print. Although counting ridges is quick, it is inherently inaccurate. The number of ridges on the same finger can easily change from print to print, depending on how hard the finger is pressed against the scanner. Also, the range of ridges on all fingers is not that varied—usually yielding a count between 10 and 20. But ridge counting eliminates obvious mismatches.
Fingerprint Identification Ridge Counting
The FID examines the possible matches generated by ridge counting, looking for further similarities in the pattern of minutia. Because the science is not perfect and the same fingerprint can vary from sample to sample, a one-to-many search often comes up with a list of candidate fingerprints that closely resemble the unknown print. A human makes the final selection.
Hi, I’m Wajid Khan. I am trying to explain computer stuff in a simple and engaging manner, so that even non-techies can easily understand, and delivered to your inbox weekly.
AASAN || Tech by Wajid Khan is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.