Switchable edge detection and depth perception with a phase-change image-processing metasurface.

Abstract

Recent advances in machine vision, three-dimensional (3D) reconstruction, and biomedical imaging underscore the critical importance of edge extraction and depth perception. However, these applications demand increasingly compact, multifunctional, and highly controllable optical systems. In this work, we propose a reconfigurable multifunctional metasurface based on the phase-change material Ge₂Sb₂Se₄Te₁ (GSST) for near-infrared optical image processing. In its crystalline state of GSST, the metasurface facilitates direct edge-enhanced imaging without a conventional 4-f system. Conversely, in the amorphous state, the same structure implements depth-sensitive point spread functions (PSFs) via a double-helix phase mask, translating axial displacement into angular variations for depth perception. This study provides a theoretical foundation and design methodology for developing highly integrated, programmable optical information processors based on metasurfaces, demonstrating great application potential for near-infrared intelligent imaging and 3D sensing applications.