Glossary of Common Biometric Terms

Biometrics The automatic recognition of persons based on unique combinations of measurable physical or behavioral characteristics. Examples include fingerprints, iris scanning, face and voice recognition, or hand geometry. All of these biometric techniques are differentiated by speed, durability, reliability, and cost effectiveness. Fingerprints are generally considered the most practical biometric identifier in use today.

Authentication Any systematic method of confirming the identity of an individual. Some methods are more secure than others. Simple authentication methods include user name and password, while more secure methods include token-based one-time passwords. The most secure authentication methods include layered or "multi-factor biometric procedures. This is independent of authorization.

Authorization The administration of person-specific rights, privileges, or access to data or corporate resources.

Biometric Authentication Mode The way biometric data (e.g. fingerprints) is used for authentication. The mode chosen for a biometric installation depends on the specific needs of a site, where either convenience or security may be emphasized. BioCert fingerprint devices may use either of two biometric authentication modes, identification or verification.

False Acceptance Rate Also known as FAR. Measures how frequently unauthorized persons are accepted by the system due to erroneous matching. Potentially serious. The FAR of BioCert devices is currently about .001%.

False Rejection Rate Also known as FRR. Measures how frequently registered users are rejected by the system. This usually amounts to nothing more than inconvenience, since it requires users to try again. The FRR of BioCert devices is currently about .01% and is usually improved by educating users on correct usage of fingerprint recognition devices, especially in high security environments.

Identification Also known as one-to-many or 1:n comparison. Authentication mode that compares the current biometric data set against all other reference data of persons previously recorded in the system. This method does not require any accompanying information to be provided with the fingerprint. It is user-friendly but inherently slower and less secure than the verification mode.

Latent Fingerprint Latent fingerprints are "left over" fragments usually caused by the build-up of oily residues on the optic sensor window after repeated use. The technique used by BioCert devices to defeat "faked" fingerprints also prevents latent fingerprints from being incorrectly validated by the system.

Matching Biometric data (e.g. fingerprints) are matched to another sample to confirm a person’s identity (authentication). For example, BioCert biometric systems use optic scanners to collect fingerprint minutiae, then create mathematical templates based on that information for storage. New input fingerprints are scanned and compared to the stored samples. If the minutiae matching threshold is met, the person is authenticated.

Minutiae The unique, measurable physical characteristics scanned as input and stored for matching by biometric systems. For fingerprints, minutiae include the starting and ending points of ridges, and ridge junctions among other features.

Template The biometric reference pattern of a person stored for matching. BioCert devices convert fingerprint minutiae into mathematical templates, so actual fingerprint images are not stored and cannot be reconstructed based on template data.

Verification Also known as one-to-one or 1:1 comparison. The verification procedure confirms whether the person in question is actually the person they claim to be. The person’s current biometric data are compared only with their own reference data. This authentication mode requires another unique identifier such as a User ID, PIN, or smart card. Verification is inherently faster and more secure than the identification method.