FB pixel

New imaging software from Israeli university could improve smartphone facial recognition

Ben-Gurion University of the Negev (BGU) has developed a new kind of imaging software called Light Invariant Video Imaging (LIVI) which improves the clarity of images captured in low light and could improve smartphone facial recognition systems and augmented reality (AR) apps.

The software uses amplitude-modulated (AM) light separation, which allows the camera to eliminate the effects of background or dynamic lighting conditions, according to the announcement. This results in images which are free of shadows, with constant colour output and sharpened contrast, regardless of the lighting conditions in which they are taken.

“Our invention produces a ‘flash’ effect that clears the backlight, removes shadows and improves contrast, making all captured frames much clearer,” says Professor Hugo Guterman of the Department of Electrical and Computer Engineering, and head of the BGU Laboratory of Autonomous Robotics. “This can have numerous applications, from smart security cameras through cell phone or computer face recognition apps, augmented reality and video game applications and military use.”

The technology filters the backlight for each pixel in the image similarly to the way a radio receiver filters radio stations, according to Ph.D student Amir Kolaman, whose work on underwater photography led to the innovation.

BGN Technologies, BGU’s technology company, is seeking a partner to further develop and commercialize the technology, which a BGN representative says is inexpensive and can be integrated into various devices.

Several companies unveiled technologies aimed at improving the performance of facial recognition systems in different lighting conditions at the recent MWC 2018.

Article Topics

 |   |   | 

Latest Biometrics News


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Most Read This Week

Featured Company

Biometrics Research

Biometrics White Papers

Biometrics Events

Explaining Biometrics