Single-metalens-assisted polarization imaging and edge detection for target recognition

Yandong Fan,Chunqi Jin,Jiayu Yang,Fei Zhu,Wei Li
DOI: https://doi.org/10.1063/5.0192217
IF: 4
2024-04-29
Applied Physics Letters
Abstract:Simultaneous capture of various light information, including polarization and edge information of the objects, has consistently been a fundamental concern within the field of target recognition. However, these tasks are typically accompanied by bulky optical components and active illumination methods, which significantly restricts their use in compact and lightweight applications. Here, we demonstrate a metalens-assisted imaging system that can simultaneously achieve polarization imaging and optoelectronic edge detection in a single shot with low consumption. The dielectric metalens is designed to achieve polarization imaging by dispersing the input polarized light into two orthogonal components, resulting in optoelectronic isotropic edge detection of two-dimensional images after digital post-processing. Compared with the algorithmic methods using a convolution kernel, the proposed system has a much lower computational complexity. The work presented in this study demonstrates the potential applications in machine vision and paves the way for the development of compact target recognition and real-time image processing systems.
physics, applied
What problem does this paper attempt to address?