Nonlocal phase-change metaoptics for reconfigurable nonvolatile image processing

Area of Science:

• Photonics and Optical Engineering
• Materials Science
• Computer Vision

Background:

• Next-generation smart imaging demands compact, high-speed, low-power optical computing hardware.
• Computing metasurfaces are crucial for analog optical pre-processing, particularly for edge detection.
• Current metasurfaces lack reconfigurability, limiting their system-level impact.

Purpose of the Study:

• To propose and demonstrate a reconfigurable flat optical image processor.
• To utilize low-loss phase-change nonlocal metasurfaces for tunable optical computing.
• To enable switchable functionalities like edge detection and bright field imaging on a single device.

Main Methods:

• Fabrication of a nonlocal metasurface incorporating phase-change materials.
• Configuring the metasurface to exhibit distinct transfer functions by transitioning phase-change material between amorphous and crystalline states.
• Experimental validation of the metasurface's reconfigurable imaging capabilities.

Main Results:

• Demonstrated a reconfigurable flat optical image processor with switchable functionalities.
• Achieved edge detection and bright field imaging modes using the same metasurface device.
• Metasurface is compatible with a numerical aperture of ~0.5, suitable for high-resolution microscopy.

Conclusions:

• Phase-change reconfigurable nonlocal metasurfaces offer a promising platform for advanced optical computing.
• This technology can enable AI-assisted imaging devices with switchable multitasking capabilities.
• The developed processor paves the way for more versatile and efficient smart imaging systems.