Automated fingerprint identification is the process of automatically matching one or many unknown fingerprints against a database of known and unknown prints. Automated fingerprint identification systems are primarily used by law enforcement agencies for criminal identification initiatives, the most important of which include identifying a person suspected of committing a crime or linking a suspect to other unsolved crimes.
Automated fingerprint verification is a closely-related technique used in applications such as attendance and access control systems. On a technical level, verification systems verify a claimed identity (a user might claim to be John by presenting his PIN or ID card and verify his identity using his fingerprint), whereas identification systems determine identity based solely on fingerprints.
With greater frequency in recent years, automated fingerprint identification systems have been used in large scale civil identification projects. The chief purpose of a civil fingerprint identifications system is to prevent multiple enrollments in an electoral, welfare, driver licensing, or similar system. Another benefit of a civil fingerprint identifications system is its use in background checks for job applicants for highly sensitive posts and educational personnel who have close contact with children.
The U.S. Integrated Automated Fingerprint Identification System holds all fingerprint sets collected in the country, and is managed by the FBI. Many states also have their own AFIS system. AFIS systems have capabilities such as latent searching, electronic image storage, and electronic exchange of fingerprints and responses.
Many other entities, including Canada, the European Union, the United Kingdom, Israel, Pakistan, Argentina, Turkey, Algeria, Italy, Chile, Venezuela, Australia, the International Criminal Police Organization, and various states, provinces, and local administrative regions have their own systems, which are used for a variety of purposes, including criminal identification, applicant background checks, receipt of benefits, and receipt of credentials (such as passports).
Fingerprints may be scanned into automated fingerprint identification systems by rolling prints or placing flat impressions onto a glass platen above a camera unit. Alternatively, prints may be obtained by placing a tenprint card (prints taken using ink) onto a flatbed or high-speed scanner. In addition to these devices, there are other devices to capture prints from crime scenes (latent prints), as well as devices (both wired and wireless) to capture one or two live finger impressions. The most common method of acquiring fingerprint images remains the inexpensive ink pad and paper form.
To match a print, a fingerprint technician scans in the print in question, and computer algorithms are utlized to mark all minutia points, cores, and deltas detected on the print. In some systems, the technician is allowed to perform a review of the points that the software has detected, and submits the feature set to a one-to-many (1:N) search. The better commercial systems provide fully automated processing and searching ("lights-out") of print features. The fingerprint image processor will generally assign a "quality measure" that indicates if the print is acceptable for searching.
Fingerprint matching algorithms vary greatly in terms of Type I (false positive) and Type II (false negative) error rates. They also vary in terms of features such as image rotation invariance and independence from a reference point (usually, the "core", or center of the fingerprint pattern). The accuracy of the algorithm, print matching speed, robustness to poor image quality, and the characteristics noted above are critical elements of system performance.
Fingerprint matching has an enormous computational burden. Some larger AFIS vendors deploy custom hardware while others use highly optimized software to attain matching speed and throughput. In general, it is desirable to have, at the least, a two stage search. The first stage will generally make use of global fingerprint characteristics while the second stage is the minutiae matcher.
In any case, the search systems return results with some numerical measure of the probability of a match (a "score"). In tenprint searching using a "search threshold" parameter to increase accuracy, there should seldom be more than a single candidate unless there are multiple records from the same candidate in the database. Many systems use a broader search in order to reduce the number of missed identifications, and these searches can return from one to ten possible matches. Latent to tenprint searching will frequently return many (often fifty or more) candidates because of limited and poor quality input data. The confirmation of system suggested candidates is usually performed by a technician in forensic systems. In recent years, though, "lights-out" or "auto-confirm" algorithms produce "identified" or "non-identified" responses without a human operator looking at the prints, provided the matching score is high enough. "Lights-out" or "auto-confirm" is often used in civil identification systems, and is increasingly used in criminal identification systems as well.