Biometric screening — using biological characteristics such as fingerprints, iris recognition or facial features — is a high priority for researchers who are working to develop future security solutions.
The vascular pattern in your fingers is as unique to you as a fingerprint. Finger vein recognition can be used both to identify you and – by sensing your blood flow pattern – prove that you are a real living person. Using a finger vein system will make it harder for cheaters to use a false fingerprint to get past security checks.
Rubber mask and fake passport
Authentication methods are made safer by increasing the number of factors that are used to identify an individual. Airports, for example, use a two-step identification method of passport and fingerprints to identify travellers.
Both face and fingerprint biometrics, however, are relatively easy to fake — a rubber mask and a dummy finger fitted with a copied fingerprint will do the trick.
“Fake passports and masks can fool today’s identification system; it’s easy and it happens,” says NTNU associate professor Raghavendra Ramachandra. “These fraudulent attempts make it all the more important to develop methods that ensure everyone is who they claim to be, and that the biological characteristics we measure are real.”
Ramachandra is affiliated with the Center for Cyber and Information Security at NTNU’s Department of Information Security and Communication Technology in Gjøvik.
Ensuring that the biological feature – whether fingerprint or face – is not a fake, is calledliveness detection in the language of the profession. Ramachandra and his colleagues have now developed a sensor prototype that scans and detects both fingerprints and blood flow in the finger veins.
Difficult to create fake blood vessels
A multimodal biometric system involves identifying several biometric features simultaneously, like the sensor that detects both fingerprints and veins. This authentication method provides a double check that the identifiers and the person match.
But it’s tough to check multiple identifiers manually, and moving queues along quickly when checking several features at once is challenging with existing methods. Ramachandra is proposing a solution for this problem.
“We envision a corridor where you don’t have to stop and wait to identify yourself, but instead, you swipe your finger across the sensor. The sensor takes 80 frames per second, so it can quickly detect and verify that the fingerprint and veins match and that blood is flowing through the finger. We’re developing algorithms to make the process faster and more seamless,” says Ramachandra.
“Since the sensor also checks your finger for blood flow, the chances of getting through with a fake fingerprint are minimal. Faking veins is very difficult,” he adds.
Source : Gemini Research