Skip to content
Breakthrough in Facial Recognition Energy Consumption and Size Using Metasurfaces

Breakthrough in Facial Recognition Energy Consumption and Size Using Metasurfaces

Facial recognition systems are part of our daily lives, being used to unlock smartphones, gaming systems, and to safely access our bank accounts online. However, traditional technology requires bulky projectors and lenses, leading researchers to seek a more streamlined solution. They have now developed a 3D surface imaging system with less complex and flatter optics.

3D surface imaging is an essential component of smartphone facial recognition, computer vision, and autonomous driving systems. These systems usually comprise a dot projector that contains multiple components: a laser, lenses, a light guide and a diffractive optical element (DOE). The DOE acts as a unique lens that dissects the laser beam into an array of nearly 32,000 infrared dots. These dots project onto a user's face, producing a unique pattern which the device interprets to confirm the user's identity. However, this system's bulky nature poses a problem for smaller devices.

In response, a team of researchers sought to create a more compact facial recognition system that would require less energy to operate, developing a streamlined system. They replaced the traditional dot projector with a low-power laser and a flat gallium arsenide surface, considerably reducing the imaging device’s size and power requirements. The researchers also etched a nanopillar pattern into the metallic surface to create a metasurface, which scatters light as it passes through.

The new system's prototype features a low-powered laser that scatters into 45,700 infrared dots, projected onto a person's face or object in front of the light source. Like the traditional dot projector, this new system also incorporates a camera to read the patterns created by the infrared dots.

The researchers tested their prototype using a 3D replica of Michelangelo's David, comparing the infrared dot patterns to online photos of the famous statue. Impressively, it achieved this feat using 5 to 10 times less power and with a platform 230 times smaller than a standard dot-projector system, demonstrating the potential of metasurfaces for effective facial recognition, robotics and extended reality that are small-scale and low-power.

The research was funded by Hon Hai Precision Industry, the National Science and Technology Council in Taiwan, and the Ministry of Education in Taiwan.

Disclaimer: This article was written using an AI tool and based on an original article by the American Chemical Society. Visit the full source here.