How Facial Recognition Works – And Why We Don’t Store Your Face

Facial recognition technology is often misunderstood. A common misconception is that systems store photographs of people and use those images to identify them later. Modern facial recognition systems don’t work this way. Here’s what actually happens.

What Information Is Captured During Enrolment?

When a user enrols onto an Idency-supported facial recognition terminal, the device analyses key characteristics of the face and converts them into a mathematical template. This template is then used for future comparisons when the user presents themselves to the device.

The characteristics measured during enrolment include:

The distance between the eyes
The shape and position of facial features
The contours and proportions of the face
Other biometric reference points

These measurements are processed by the device’s biometric algorithm and converted into a digital template. The original image used during enrolment is not required for future recognition.

What Is Actually Stored?

The device stores an encrypted biometric template — not a photograph. Think of it like a password hash: information goes in, a secure mathematical representation comes out, but the original cannot be reconstructed from the stored data.

The template is:

Randomised and encrypted
Mathematical in nature
Designed solely for matching purposes
Not a photograph or image
Not human-readable

Can Someone Recreate My Face From the Stored Data?

No. The biometric template is a one-way representation of facial characteristics. While the system can compare a live face against the stored template to determine a match, it cannot reverse the process and recreate the original face or image.

In practical terms:

A face can be converted into a template
The template can be used to verify identity
The template cannot be converted back into a face

Even if someone gained access to the stored template, they would not be able to reconstruct a photograph or visual image of the individual.

Do the Devices Store Photograph?

No. Idency-supported facial recognition devices do not store photographs of users for identification purposes. The system stores only the encrypted biometric template generated during enrolment, which contains the mathematical data required for recognition, not an image of the person.

How Recognition Works Day to Day

When a user presents themselves to the device:

The device captures a live image of the face
Facial characteristics are analysed in real time
A temporary biometric template is created from the live scan
The temporary template is compared against the stored template
If a match is found, the user is authenticated

The live image is used only for the recognition process and is not retained as part of the biometric template.

Why Use Facial Recognition

Modern facial recognition offers several advantages over traditional authentication methods:

Fast and contactless authentication
No cards, fobs or passwords to forget
Reduced buddy punching and time fraud
Excellent performance for manual workers where fingerprints may be worn or difficult to read
Reliable operation in both indoor and outdoor environments

Today’s facial recognition technology is significantly more accurate and reliable than older biometric methods, making it the preferred choice for many workforce management and access control applications.

Privacy By Design

Facial recognition systems supplied and supported by Idency are designed with privacy and security in mind. The biometric data stored is limited to encrypted mathematical templates used solely for authentication purposes. No photographs, images or visual representations of individuals are stored as part of the facial recognition process.

At Idency we provide Biometric Statements to our customers that describe the above for each of our devices, giving assurance to their employees about how the process works.

Get in touch

Time & Attendance Systems

Authentication devices