Understanding Biometrics

Want a lighter way to understand LightSpeed?
Watch this video.

Biometric Technology Terminology

The language of biometrics is an evolving lexicon of science and technology. Below are some commonly used terms and their meanings.

2D Facial Landmarking

An advanced biometric process that captures and analyzes face image data, combined with facial feature tracking, facial modeling and animation, and expression analysis.  It increases the accuracy of biometric systems by an order of magnitude.


ABIS stands for Automated Biometric Identification System which is a large scale customizable software solution similar to AFIS, but with the added advantages of utilizing multiple biometric modalities including iris and facial recognition to provide fast, secure, and reliable results.

Active Impostor Acceptance

Active Impostor Acceptance is when an access control system incorrectly recognizes and accepts a biometric sample which has been altered, modified, or cloned.


A sequence of instructions that instructs a biometric system on how to solve a problem. It could have a finite number of steps in the instruction for computing whether the sample and the template are matched.


The moment a biometric sample is being submitted for verification. An “attempt” may happen more than once in cases where the sample is denied or rejected.


When biometric data being used is considered to be correct and valid. “Validation” is the preferred term for this.


Physical traits or patterns which are unique to every individual that are used to verify and authenticate a person’s identity who is enrolled into a system. Biometric patterns can be anything from fingerprints, iris scans, facial recognition and even voice recognition.

Biometric Application

The implementation of any system that involves biometric data.

Behavioral Biometrics

Behavioral Biometrics identify individuals based on human actions and behaviors (e.g. keystroke dynamics, signature analysis, and gesture biometrics).

Biometric De-Duplication

Biometric De-Duplication is a process of identifying duplicate biometric templates from the database to ensure every biometric template is unique.

Biometric Engine

The portion of the biometric software system that processes the gathered data. It can start to operate from the data capture, extraction, and comparison down to the matching.

Biometric Enrollment

Biometric Enrollment is the process of collecting an individual’s unique physical or behavioral characteristics using biometric hardware which is converted into a mathematical file using a complex algorithm and saved into the biometric database for future identification/authentication purposes.

Biometric Identification Device

A Biometric Identification Device is a device which gathers, reads, and compares biometric data. “Biometric System” is the term more often used.

Biometric Template

The file containing the mathematical representation of a person’s biometric trait  After a person’s unique physical or behavioral characteristics are captured by a biometric device, it is converted into a mathematical file using a complex algorithm and saved into the biometric database for future identification/authentication purposes.

Biometric Sample Data

The data captured by a system collected from a person of interest or a user.

Biometric SSO (Single Sign On)

An independent software system which authenticates access with a single biometric attribute within the network.

Biometric System

An automated system which does the following:

  1. Collects or captures biometric data via a scanner.
  2. Extracts the data from the actual submitted sample.
  3. Compares the scanned data from those capture for reference.
  4. Matching the submitted sample with the templates.
  5. Determining or verifying whether the identity of the biometric data holder is authentic.

The process of collecting biometric data from the end user or enrollee. Most biometric data are “captured” by the use of an image scanner in cases of fingerprints, palm vein patterns, or a camera to collect facial and iris scans.


Certification is testing gathered biometric data against a system or software to verify its ability to perform. The application will then be tested according to set standards for certifications, which are then issued by testing organizations.


A person who submits his biometric sample for identity verification. Claimants may either have true or false identities.

Claimed Identity

A biometric sample of an enrolled user of the system.

Closed Set Identification

Closed Set Identification is the process by which users need to be enrolled into a biometric system and verified in order for access to be granted.


Comparing a biometric sample with previously gathered samples or against a template or templates for verification of the identity.

Contactless Biometrics

Biometric technology that can identify an individual without having to make physical contact with a hardware device (e.g. iris and facial recognition), which is beneficial to certain hygienic environments such as patient and provider identification in healthcare.

D Prime

Statistical measure which grades the ability of a system to distinguish between biometric samples or individuals. The higher D prime number means that the system is more capable of distinguishing between samples.

Degrees of Freedom

The number of independent features in a biometric system.


The conversion of any biometric data into a cypher code which cannot be easily read. A password or biometrics itself may be used as the key to decrypt or decode the data.

End User

A currently enrolled or about to be enrolled individual who has his/her biometric data submitted for identification or verification.

End User Adaptation

End User Adaptation is the way in which users of a biometric system are able to adjust accordingly to it after becoming familiar with the test.


The user who has their biometric template entered into the system.


Gathering and processing of biometric data with the intent of storing records into a database.

Enrollment Time

The time spent on collecting the biometric data and successfully processing it.

Equal Error Rate

The frequency in which the rate of false rejection is almost equal to the rate of false acceptance.


The moment a biometric sample is converted into data which is later compared to a biometric template.

Face Recognition

A process in which facial features are analyzed and gathered as biometric data.

Failure to Acquire

The process in which a biometric system fails to capture, extract, and store the data.

Failure to Acquire Rate

The number of times that a Failure to Acquire occurs.

Failure to Capture (FTC)

Failure to Capture (FTC) occurs when the system failed detect a biometric input, even though the input is correct.

Failure to Enroll (FTE)

Failure to Enroll (FTE) happens when the biometric system fails to enroll a person’s biometric attribute due to technical or environmental or damage due to accidental reasons.

False Acceptance

The biometric system accepts either a false identity or incorrectly identifies a wrong identity against a claimed one.

False Accept Rate (FAR)

A metric used to measure the accuracy of biometric systems to correctly or incorrectly identify individuals that represents the statistical chance of incorrect matches. This is synonymous with the terms False Positive Identification Rate (FPIR) or False Match Rate (FMR).

False Match

An instance of an incorrect yet positive match outcome between submitted data and the enrolled database.

False Rejection

False Rejection occurs when a newly acquired sample of an enrolled identity is rejected by the system or when it fails to verify a legitimate identity.

False Rejection Rate (FRR)

The probability that a biometric system will fail to identify a legitimate identity due to the system failing to match the biometric input with the template.  This is synonymous with the terms False Negative Identification Rate (FNIR) or False Non-Match Rate (FNMR).

Identification or Identity

A biometric sample which is matched against templates and other biometric references.


A person who poses as a verified user by submitting his own biometric sample.

In House Test

A series of testing done in a closed facility or laboratory. It may or may not involve the use of external participants or subjects.

Iris Recognition

A method to identify someone by a single iris or both irises (the colored, ring-shaped region surrounding the pupil of the eye).

Live Capture

The process of gathering biometric sample from a live user using a biometric system

Match or Matching

The process of matching a template to a submitted biometric sample. It is then rejected or accepted based on the whether the score has met the threshold or not.

Multimodal Biometrics

A system that uses more than one physiological or behavioral characteristic such as iris and facial imaging, fingerprint and finger vein for enrollment, verification, and identification of a person.

Multi-Factor Authentication/Two Factor Authentication

A biometric system where more than one biometric credential is required to identify or authenticate a person. It increases the accuracy of biometric systems by an order of magnitude.

Open-Set Identification

Identifying users who are not enrolled in the system, the opposite of closed set identification.

Original Equipment Manufacturer or Module

An organization which assembles a biometric system from different parts or an independent module which can be integrated into a biometric system.

Passive Impostor Acceptance

When an impostor’s submitted sample is verified and accepted by the system.

Performance Criteria

A set of standards or criteria which is used to evaluate the performance of the system.

Personal Identification Number (PIN)

A preset number is entered into a secured system to gain access. Usually a four-digit value.

Physiological Biometrics

Physiological Biometrics use parts of the human body for individual identification (e.g. fingerprints, finger vein, facial or iris recognition).

Receiver Operating Curves

A graph showing how the false rejection and false acceptance rates varies with one another


The widely used term is identification, due to a preset standard or set of traits.

Response Time

The amount of time in which a biometric system analyzes a sample and returns with a decision.

Segmented identification

Segmented identification (one-to-few, or 1: Few) involves confirming or denying a person’s claimed identity using a biometric scan followed by verbal confirmation to a general identification question (e.g., What is your date of birth?) or the entry of some known general information (gender, race, eye color).

Template Capacity

The maximum capability of stored data in a biometric system.

Template or Reference Template Data

A biometric measurement which is used to verify succeeding biometric data.

Third Party Test

A test done by an independent party in a controlled environment.

Threshold or Decision Threshold

The acceptance level of any given biometric system. It may be tightened or widened accordingly to make the system meet certain requirements. If the data falls above or below the threshold, it is rejected. If the sample falls within the acceptable range it is accepted.

Throughput Rate

The number of users a biometric system can successfully process within a given time

Type 1 Error

See “False Rejection.”

Type 2 Error

See “False Acceptance.”


The client of any biometric vendor. Essentially, they are the clients that purchase the technology but may or may not enroll themselves into the system. End users are those who enroll their biometric data into the system.


The process of comparing a biometric sample with the biometric data in the system whose identity is claimed.

Verification (1:1 Matching)

Biometric verification is an identity authentication process used to confirm a person’s identity by matching their uniquely identifiable biometric traits such as fingerprints, palm vein, iris, and face recognition with a specific stored biometric template.

Xenon Illumination

A new generation of more effective near infrared light bands compared to traditional narrow-band IRLED light. Light is produced by passing electricity through ionized Xenon gas. It is recognized for its high illumination power output and long life.

Zero Effort Forgery

An impostor uses the actual biometric sample of an enrolled user.

Sources and Further Reading

While biometric technologies like LightSpeed are cutting-edge in many ways, there has been extensive analysis of their importance and applications going back well over a decade. Here are a few seminal papers and articles for further reading.

Request a Demo