Biometrics
Facial recognition, through photographs or video, is the biometric information we use to verify a person’s identity. Modern facial recognition systems are fast, accurate and secure, reducing the time and cost of identity verification.
DIA’s biometrics systems can identify or verify an identity in seconds. When the system cannot make a match, trained staff assess the image.
Biometric information, such as photographs, play a crucial role in verifying a person's identity and preventing fraud. Biometric information protects the integrity of New Zealand’s identity products and services.
The benefits of biometrics
Using biometrics makes it easier, faster, and more convenient for New Zealanders to access services and information they need.
Biometrics enhance the security of personal information by using physical traits which are unique to individuals.
This helps combat fraudulent activity – such as stealing someone’s identity – because it makes it harder for personal information to be stolen.
DIA’s use of biometrics
DIA uses biometric information to make it easier and faster to verify people’s identity and issue passports and other identity documents. Facial recognition, through photographs or video, is the biometric information DIA uses to do verify a person’s identity.
DIA services biometrics are use in:
- Passports
- Citizenship
- RealMe verified login
- Identity Check service
Privacy considerations
Biometric information is personal information and is regulated by the Privacy Act.
Privacy Impact Assessments are undertaken for any services involving biometrics.
Algorithms
Three algorithms are used when DIA verifies a customer’s identity across our products:
- A one-to-many algorithm which searches for matches of an image against the records held in the DIA passport database to find if a face exists.
- A 1:1 facial recognition algorithm that matches the image captured during the verification process with a passport or driver licence image.
- A liveness algorithm, which tests the images of the applicant to detect false presentations such as a person holding up a photo of someone else or a deepfake.
