Gerchberg-Saxton Optimized Silicon Metasurface for Sinusoidal Structured-Light Three-Dimensional Imaging

Abstract

Three-dimensional (3D) optical imaging based on structured illumination has attracted considerable attention in applications including machine vision, robotics, and computational sensing. However, conventional structured-light systems generally rely on bulky projection components, such as digital micromirror devices and diffractive optical elements, which limit system miniaturization and integration. Here, we demonstrate a compact structured-light 3D imaging platform enabled by a silicon dielectric metasurface operating at a near-infrared wavelength of 940 nm. The desired sinusoidal fringe illumination is encoded into the metasurface through a Gerchberg-Saxton (GS) phase optimization algorithm, which converts the target far-field distribution into a nanoscale phase profile and subsequently into silicon nanopillar geometries. The fabricated metasurface generates high-quality sinusoidal structured illumination, which is integrated with a four-step phase-shifting reconstruction method to recover 3D object profiles. The proposed approach provides a compact and scalable solution for miniaturized structured-light projection and optical 3D sensing.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.