Metasurface- and PCSEL-Based Structured Light for Monocular Depth Perception and Facial Recognition

Area of Science:

• Optics and Photonics
• Computer Vision
• Materials Science

Background:

• Structured light (SL) depth-sensing systems traditionally use diffractive optical elements (DOEs) and collimators, leading to bulky and power-hungry devices.
• Existing systems face limitations in dot projection density, size, and power efficiency, hindering integration into compact applications like wearables.

Purpose of the Study:

• To develop a novel, compact, and power-efficient depth-sensing system for facial recognition using metasurfaces and PCSELs.
• To overcome the limitations of conventional SL depth-sensing technologies.

Main Methods:

• Utilized a GaAs-based metasurface and photonic crystal surface-emitting lasers (PCSELs) for infrared dot projection.
• Designed a lens-free, lightweight optical architecture.
• Measured the system's field-of-view (FOV) and dot sampling angle.

Main Results:

• Projected approximately 45,700 infrared dots from a 297-μm metasurface, significantly outperforming DOE-based projectors.
• Achieved a wide FOV of 158° and a dot sampling angle of 0.611°.
• Demonstrated 5-10 times lower power consumption compared to vertical-cavity surface-emitting laser (VCSEL) arrays.

Conclusions:

• The developed system offers a revolutionary approach to SL depth-sensing, enabling efficient facial recognition.
• The compact size, lens-free design, and low power consumption make it ideal for integration into wearable devices.
• This innovation paves the way for widespread adoption of advanced depth-sensing technology in consumer electronics.