Authentication with a wave of your hand, a “large fingerprint”


Two former Leap Motion employees think they’ve solved the perennial problem of passwords and identity management.

Under the nom de guerre of Redrock Biometrics, the company’s technology is waging war on the world’s legions of identification cards, pin numbers, passwords, and voice identification technologies with a combination of off-the-shelf hardware and proprietary software to identify palm-prints to verify user identity.

Redrock Biometrics’ chairman, Lenny Kontsevich, said that the company sees broad applications in authenticating payments in virtual worlds, physical security, and cash withdrawals among other transactions.

“You can think about the palm as a very large fingerprint,” says Kontsevich. “It has a rich structure and can be captured by any camera touchlessly.”

After Kontsevich worked with the startup Kaching!, the graduate of the Moscow Institute of Physics and Technology went to the lab to demonstrate (using only existing database technology) the matching software algorithms between palm prints and their unique signatures that would become the basis for Redrock Biometrics.

The issue, for Kontsevich, was capturing the images and processing them from a background. That’s where Hua Yang, Kontsevich’s Leap Motion colleague came in.

The two absconded from Leap Motion in 2015 and founded Redrock biometrics on the basis of Kontsevich’s work and Hua Yang’s background in machine learning and visualization.

“There is no other commercially available palm biometric which works with RGB camera,” Kontsevich told me.

Redrock’s technology converts a palm image into a unique signature and authenticates the user in 10- to 100-milliseconds depending on CPU speed, according to a statement. The technology uses machine vision techniques to detect a palm in a video stream and pass its descriptor for enrollment or verification, according to a statement.

The technology runs on either a client or a server and matches a verification request against an enrollment template, using proprietary algorithms tested on thousands of palms.

Positioning palm biometrics vs. either fingerprint or iris scanners, Konsevich says that the palm scanning technology he’s developed doesn’t require special equipment. And it’s far more secure than scans like the face scanning technology currently the rage thanks to Apple’s new iPhones.

There are thousands of images of people’s faces publicly available on social media that can be used to fake a face scan, says Konsevich. “A palm, in this regard, is much harder to get. People don’t make photographs of their palm in good resolution.”

Read the source article at TechCrunch.