Rank One Computing announced the issuance of patent US 10,839,251 B2 by the US Patent Office covering a method to determine biometric liveness and prevent spoofing using micro-texture analysis to validate image authenticity when performing image comparison tasks.
The patent, “Method and system for implementing image authentication for authenticating persons or items”, Patent Number US 10,839,251 B2, covers authentication methods with several practical advantages over other methods for performing facial liveness verification:
- No special hardware is required; it works on generic RGB cameras that are deployed in mobile phones, embedded devices, and security systems
- The method is passive, meaning that users will not need to perform any active steps (such as a turning their head) to validate themselves
- The method operates on a single image or video frame, reducing processing and / or network bandwidth requirements
The anti-spoof biometric liveness algorithm covered by this patent has been available in the lightweight and easy-to-integrate ROC SDK face recognition libraries for several years and is already integrated into multiple operational products for access control.
This is the first patent awarded to Rank One Computing, as the company typically uses trade secret protections to cover the vast majority of its intellectual property. Rank One currently has another patent pending on efficiency improvements for biometric encrypted matching.
As an increasing number of biometric integrators are deploying systems for contactless payment, access to online banking services, voting, and a wide range of other identity authentication technologies, anti-spoofing methods are becoming critical to the reliability and security of these systems. Rank One is committed to remaining at the forefront of face anti-spoofing technologies and other related methods for validating image authenticity, and licensees of ROC software will continue to benefit from these innovations.