Skip to main content

Components of Biometric Systems

Pedro Couto, Research Engineer at Vision-Box


In our previous instalment of this series, we took a look at the different types of biometrics currently in use. We also discussed the various biometric or inherence factors that are utilised to verify identity and authenticate subjects

In this article, we shall consider the processes involved in biometric measurement for verification and authentication and the components that make up biometric systems.


Standard Biometric Processes

The science of biometrics relies on the cataloguing and on-demand recognition of unique personal identifiers, such as fingerprint patterns, face, voice, or typing rhythm. Biometric systems leverage these unique characteristics for the purposes of identification, authentication, verification, and authorisation. All these functions rely on the outcome of the matching process, which involves comparing a captured biometric sample with one or more templates stored in a database to establish a level of similarity or identity. This matching process is foundational to performing identification, authentication, and verification.

In identification and authentication, systems subject biometric samples to one-to-many matching and database comparisons, to establish whether an individual is who they claim to be (authentication), or to establish them as a known entity (identification).

Verification is a one-to-one matching process whereby a live sample taken from the subject is compared with a previously stored template in the biometric database.

In authorisation scenarios, biometric matching serves to provide access rights to authenticated or verified users. Hierarchical systems often exist, with authorised users having access levels in line with their job description or security clearance rating.

For each of these scenarios, certain processes are standard to all biometric systems. They include the following:


Biometric enrolment is the process of capturing biometric data from an individual to create a unique template or reference that can be later used for identification or verification purposes.


Rather than saving an entire image or video, most biometric systems convert discriminative biometric attributes into an encoded template, which is held in the system database.


This is the on-demand verification aspect of a biometric system. Each time a subject presents him/herself to the system, it compares their biometric features they offer to the information on record. The results of this comparison determine whether the system accepts or rejects the subject.


Biometric System Components

biometrics system components

Biometric system components comprise the technology which takes an individual’s physiological or behavioural characteristics as input, analyses them, and then identifies the individual as a genuine or fraudulent user. In essence, they consist of four basic components:

1. Input Interface (Scanners or Sensors)

The input interface consists of the scanner or sensor that identifies a biometric trait the system will use for identification. This is the sensing component of a biometrics system which converts human biometric data into a digital form.

Fingerprint systems typically use an optical scanner, while voice recognition systems employ a microphone. Facial recognition systems primarily rely on RGB cameras for their ability to capture colour information and high-resolution facial images, crucial for accurate identification. Additionally, depth cameras may be also utilised to capture depth information, enhancing the system's ability to detect facial features and improve accuracy. Handprint recognition systems may also use RGB cameras for capturing hand geometry or palm vein patterns, although depth cameras offer additional benefits by providing depth information alongside colour. In contrast, iris and retina recognition systems commonly use near-infrared (NIR) sensors, which detect unique NIR reflections to capture detailed biometric patterns. NIR sensors excel in low-light conditions and can penetrate ocular tissues, ensuring reliable recognition.

2. Processing Unit

A computing system must then collect and analyse information from the input interface. The processing unit of a biometric system typically consists of a microprocessor, a Digital Signal Processor (DSP), or a dedicated computer that processes the data captured from the sensors.

Biometric sample processing will usually involve some form of enhancement of the sample image and normalisation of this image. Subsequently, the processor extracts key features from the normalised image and compares the biometric sample with stored samples in the system database.

3. Database Store

The database of a biometric system stores enrolled samples in an encoded and space-efficient format, allowing for quick retrieval when needed to perform a match during authentication.

4. Output Interface

The output interface communicates the decision of a biometric system regarding the verification/authentication or otherwise of each user. Several communication protocols can serve this function, depending on the biometric system.


Biometric System Components in Action

The standard workflow for a biometric system involves the following steps: scanners or sensors acquire a live biometric sample from the candidate; the processing unit extracts prominent features from the sample; using relevant algorithms, the computing system compares the live sample with samples stored in the biometric database, the output interface presents the decision to accept or reject the candidate.

biometrics system components in action

Know more about our Seamless Journey Platform here


Measurement of Accuracy

False Acceptance Rate (FAR) and False Rejection Rate (FRR) are inversely related in biometric systems. As one decreases, the other tends to increase, and vice versa. This relationship is crucial for understanding the trade-offs involved in setting the operating threshold of a biometric system. What are they? FAR measures the probability that the system incorrectly accepts an unauthorised user. In other words, it indicates the likelihood of a system incorrectly identifying an imposter as a legitimate user. FRR measures the probability that the system incorrectly rejects a legitimate user. It indicates the likelihood of a system failing to identify a legitimate user.

How do they influence each other?


Biometric systems use a threshold to decide whether a biometric sample matches the stored template of an enrolled user. This threshold determines the balance between FAR and FRR. A higher threshold decreases FAR but increases FRR, leading to more stringent security but potentially more false rejections. A lower threshold decreases FRR but increases FAR, resulting in more false acceptances but potentially fewer false rejections.

Adjusting the threshold allows administrators to tune the system's security level according to specific requirements. For example, in high-security environments, a lower FAR may be prioritised, even if it leads to a higher FRR. Conversely, in applications where user convenience is critical, such as access control in a busy office building, a higher FAR may be acceptable to minimise inconvenience to legitimate users.

Equal Error Rate (EER) is a key metric used to evaluate the performance of biometric systems. It represents the point where the FAR and FRR are equal. In other words, EER is the point of operation setting where the system makes an equal number of false acceptances and false rejections. At this point, there's a balance between security and convenience. EER provides a single metric to compare the performance of different biometric systems or to evaluate the performance of a single system under different conditions.

In summary, FAR and FRR are interrelated, and adjusting the threshold of a biometric system allows administrators to strike a balance between security and convenience. EER serves as a reference point for assessing the overall performance of the system.Top of Form

Bottom of Form

Know more about our Facial Recognition Engine here


Understanding the processes behind biometric measurement and the components of biometric systems reveals the intricate mechanisms at play in identifying and authenticating individuals. From capturing biometric data using specialised sensors to processing and matching features through algorithms, biometric systems form a critical foundation for security, efficiency, and personalisation across various domains. Biometric systems play a crucial role in enhancing security, accuracy, and convenience across diverse applications, from access control and border security to healthcare and finance.

This concludes our look at the various components of biometric systems. Keep watching this space to learn more about biometrics, as we continue our series of articles.

Share this post