Two months ago, e-Passports started to go through an authentication process at John F. Kennedy International Airport with the help of a biometric facial recognition technology. Today, we find out who are the guys behind the tech.
The vb e-pass desktop was provided by Vision-Box as a way to match the image of the traveler’s face from the passport with an instant picture of the real person. The technology reads the chip of the e-passport first, then takes a quality photo of the passenger’s face to see if it matches the photo stored. In the end, it gives a match by analyzing facial features. Those who do not pass the test are further screened.
The system was first tested at Washington’s Dulles International Airport and it was a huge success. That is due to the system’s strict compliance with ISO standards for photography and, of course, the use of a remarkable face matching algorithm.
The solution utilizes software-controlled multidirectional lighting and automated camera adjustment after the face position. Apparently, it works even when the photo in the passport’ chip is low quality. But could a mobile version of it deliver in day-to-day situations, where you can’t control light and environmental factors?
In unpredictable contexts such as those we need solutions more than ever; ones that work in low light and are indifferent to glasses, hats and other accessories that cover parts of the face. Such algorithms have already been developed, by companies as FotoNation; face recognition solutions that use analysis, registration and enhancement techniques to match faces on still images or video. Most importantly, working on mobile platforms and offering accuracy without affecting processing power, they open a plethora of possibilities that go beyond the perfect conditions from an airport.