Metasurface-Enhanced Full-Space Structured Light for Depth Sensing

Laser-based imaging technology is regarded to be one of the most promising and accurate three-dimensional (3D) depth sensing technologies when combined alongside the rapid advances in computer science. To image 3D objects over a wide field of view (FOV), the laser beams illuminate the objects, and backscattered light is monitored through detectors. Structured light (SL)-based imaging system splits the light into an array of dots or lines over the FOV, enabling imaging of multiple objects simultaneously. So far, diffractive optical elements (DOEs) or spatial light modulators (SLMs) have been used to form the laser spot arrays. However, such architectures suffer from small FOVs, low efficiency, and bulkiness resulting from the large micron-scale pixel size. Here, we propose a metasurface-based SL imaging platform, allowing for subwavelength scale manipulation of incident light. The scattered light from the metasurface covers the full 180° FOV, with a high-density ~10K dot array. The metasurface is composed of a periodic supercell, and the properties of diffraction patterns are analyzed by convolution theorem considering the supercell as a kernel function. As a proof-of-concept, we place face masks one on the laser beam axis and the other 50° apart from axis within distance of 1 m and estimate the depth information from the backscattered light using a stereo matching algorithm. Furthermore, we demonstrate the replication of our metasurface on a glass surface using the nanoimprinting of a nanoparticle-embedded-resin (nano-PER) for high-throughput fabrication.