360° structured light with learned metasurfaces

E. Choi

doi:10.1038/s41566-024-01450-x

2024-06-11

PRE

DOI：10.1038/s41566-024-01450-x

原文链接

原文求助

摘要

En 中文

Structured light has proven instrumental in three-dimensional imaging, LiDAR and holographic light projection. Metasurfaces, comprising subwavelength-sized nanostructures, facilitate 180 degrees-field-of-view structured light, circumventing the restricted field of view inherent in traditional optics like diffractive optical elements. However, extant-metasurface-facilitated structured light exhibits sub-optimal performance in downstream tasks, due to heuristic design patterns such as periodic dots that do not consider the objectives of the end application. Here we present 360 degrees structured light, driven by learned metasurfaces. We propose a differentiable framework that encompasses a computationally efficient 180 degrees wave propagation model and a task-specific reconstructor, and exploits both transmission and reflection channels of the metasurface. Leveraging a first-order optimizer within our differentiable framework, we optimize the metasurface design, thereby realizing 360 degrees structured light. We have utilized 360 degrees structured light for holographic light projection and three-dimensional imaging. Specifically, we demonstrate the first 360 degrees light projection of complex patterns, enabled by our propagation model that can be computationally evaluated 50,000x faster than the Rayleigh-Sommerfeld propagation. For three-dimensional imaging, we improve the depth-estimation accuracy by 5.09x in root-mean-square error compared with heuristically designed structured light. Such 360 degrees structured light promises robust 360 degrees imaging and display for robotics, extended-reality systems and human-computer interactions. A single-metasurface-based holographic light projection covering the whole 360 degrees field of view is realized by optimizing the metasurface design through a neural network and applying 360 degrees structured light for holographic light projection and three-dimensional imaging.

Keyword:

RAYLEIGH-SOMMERFELD

PHASE

KIRCHHOFF

期刊

Nature Photonics

IF:

32.9

论文数:

4.3K

被引数:

6.1W

机构

暂无机构信息

引用论文

Neural Holography with Camera-in-the-loop Training 具有相机在环训练的神经全息术

ACM TRANSACTIONS ON GRAPHICS

IF9.5

2020-11-27

232

PREAI

Peng, Yifan; Choi, Suyeon; Padmanaban, Nitish; Wetzstein, Gordon

Forebrain contribution to the induction of a cellular correlate of conditioned taste aversion in the nucleus of the solitary tract

The Journal of Neuroscience

IF0

1995-10-01

OAAI

GE Schafe; RJ Seeley; IL Bernstein

Helicity multiplexed broadband metasurface holograms

NATURE COMMUNICATIONS

IF15.7

2015-09-10

890

OAAI

Wen, Dandan; Yue, Fuyong; Li, Guixin; Zheng, Guoxing; Chan, Kinlong; Chen, Shumei; Chen, Ming; Li, King Fai; Wong, Polis Wing Han; Cheah, Kok Wai; Pun, Edwin Yue Bun; Zhang, Shuang; Chen, Xianzhong

Characterization and mobilization of toxic metals from electrolytic zinc waste 电解锌废料中有毒金属的表征和动员

Chemosphere

IF0

2019-10-01

PREAI

M.J. Martínez-Sánchez; A.M. Solan-Marín; A.M. Hidalgo; C. Pérez-Sirvent

Probability distribution function and multiscaling properties in the Korean stock market

Physica A: Statistical Mechanics and its Applications

IF0

2007-09-01

PREAI

Kyoung Eun Lee; Jae Woo Lee