Beijing-based Vkansee could have the key to the biometric fingerprint conundrum

February 10, 2017 | By Paul Mutter of GeekTime.com

Optical technology picks up greater levels of detail, without imposing heavy screen redesigns. They also recently raised $10 million

Biometrics developer Vkansee has raised $10 million in a Series A that will allow the company to begin mass production of its new high resolution fingerprint sensor, and complete development of a second system. The company will debut its latest offerings, in 200 evolution kits, at the upcoming Consumer Electronics Show in Las Vegas, including a ~3mm glass panel that can be used for a washing machine control panel.

This would allow customers to pay for and access “all kinds of equipment with biometrics – if the sensor meets the application requirements,” according to Jason Chaikin, President of Vkansee. The company debuted these sensors last year, also at the Consumer Electronics Show, and has been making the rounds with original equipment manufacturers since.

Although the company has a small staff today, he says that going forward, “In addition to shipping tens of millions of sensors every quarter, I would like to see us combine biometrics with Identity Management principles.” Security is central to the biometrics industry, of course, and the company’s investment in pinhole imaging to register the most minute details of our fingerprints is coupled with a desire to further explore the larger ethical issues of identity management technology, which he’s said before is the “white elephant” issue in the biometrics industry.

Founded in 2013 and based between Beijing and New York City, Vkansee previously raised $7 million from Aviation Industry Corporation of China in 2015. This round includes participation from Infotech Ventures, Yunnan Huizhong Fund, Shenzhen Qianhai Greatwall Fund, and Superpix Micro Technology.

The company also recently announced it had selected Precise Biometrics as a partner for fingerprinting software.

Vkansee has focused on pinhole imaging to develop a system that is not dependent on “a prism, lens and light path that optical imaging systems require,” says Chaikin. This means that glass thickness is not an issue, because the high resolution imaging captures a picture of your fingerprint to lock and unlock the smartphone or laptop. As for where the technology could be used outside of those devices, Chaikin further notes to Geektime that the “sensor can be fully sealed for many industrial environments.”

“As long as light passes without refraction or interference, we can reliably capture fingerprints,” he says. The power demands are also quite low, with the device needing about as much battery power as taking a photo does – which, it basically is because it’s based on a CMOS image sensor. (This would work out to 1-2% of the battery life for 150 logins on a Samsung Galaxy S5 or iPhone 6.)

This resolves a number of issues with the technology available on the market today. Sensors, Chaikin notes, have trouble discerning the “ridges” and “valleys” of our fingerprints when scanning for their unique electrical patterns. The more glass between the sensor and the hand, the harder it is for the sensor to map out what force is being applied to it.

The two systems are the VK2108, with an an active under-the-glass scanning area of 10 x 6 mm, and the VK0208, with an even smaller active scanning area, at 3.5 x 6 mm. The former will go into mass production next year, while the latter will benefit from that $10 million in order to wrap up the company’s R&D process so it can be pushed out and shown to potential customers. (The VK0208 has a pixels per inch resolution quadruple, 2000, that of existing Apple and Samsung devices, at 500.)

One of the biggest challenges to developing biometric technology for mobile devices and personal computers is color matching. For the latter, there’s more room and less precise color demands (usually a shade of black) than mobile devices, which have less surface area and much more color specificity as part of the individual smartphone manufacturer’s design process. Such specificity can be discouraging, as Chaikin notes that, “This entails quite of bit of trial and error on the production line for component suppliers such as ourselves.”

According to Chaikin, there is an omnipresent need to be a balance between convenience and security, because, “Higher security is in demand but not at the expense of convenience,” and, “Convenience here means how many times you must present your finger to get a good sample.” Micro feature details like sweat pores, which cannot be read by most existing technology, can be handled by the VK0208 and that extra layer of basic security.

It is possible to fool many sensors with fingerprint casts made of wood glue, plaster, or even Play-Doh, since they generally don’t detect if the “hand” is alive. (Scanned images can also be used for “spoofing.”) Since these prints could be lifted off of surfaces without users’ consent, Chaikin “believes that high resolution provides a stronger defense,” even though it is very difficult “to take a ‘latent’ print off a doorknob or wine glass from an uncooperative subject.”

But there should be more steps in what is a layered security process, such as ensuring too many login attempts raise a flag, or requiring additional login details. “Opening your phone may be just fine with a fingerprint,” Chaikin says, “But getting a new mortgage or selling the farm should have additional factors to further reduce opportunity for things to go wrong.”

Looking ahead, Chaikin thinks that use of biometric sensors will double, from 40% now to 70-80% in 2020. He concludes, “I believe we can have more personal privacy using biometrics while at the same time making it seamless for us to automatically authenticate our car, phone, home,” as well as IoT devices and wearables. And, he reiterates his point from a Planet Biometrics interview that eye recognition technology shouldn’t be neglected either.

Source: http://www.geektime.com/2016/12/19/beijing-based-vkansee-could-have-the-key-to-the-biometric-fingerprint-conundrum/