Full-space Cloud of Random Points with a Scrambling Metasurface.

Area of Science:

• Optics and Photonics
• Materials Science
• Computer Science

Background:

• Full-space spot generation is crucial for AI-driven applications like facial recognition and augmented reality.
• Existing methods like diffractive optical elements (DOEs) and light detection and ranging (LIDAR) have significant limitations.
• DOEs are bulky and limited to half-space operation, while LIDAR requires complex scanning systems hindering miniaturization.

Purpose of the Study:

• To develop a compact and efficient full-space spot generator.
• To overcome the limitations of current technologies for 3D sensing and recognition.

Main Methods:

• Designed and fabricated a Hermitian-conjugate metasurface.
• The metasurface scrambles incident light into a cloud of random points in full space.
• Utilized amorphous silicon with subwavelength height and continuous phase coverage.

Main Results:

• Experimentally observed over 4044 random spots covering nearly 90° in the entire space.
• The metasurface operates in both transmission and reflection modes.
• Demonstrated a lightweight, flexible design with low-heat dissipation.

Conclusions:

• The developed metasurface offers a promising solution for miniaturized 3D recognition sensors.
• Potential for mass production and integration into semiconductor foundries.
• Enables advancements in motion sensing, facial recognition, and other AI applications.