• Advanced Imaging
  • Vol. 1, Issue 3, 033001 (2024)
Guanghui Yuan*
Author Affiliations
  • Department of Optics and Optical Engineering, School of Physical Sciences, University of Science and Technology of China, Hefei, China
  • show less
    DOI: 10.3788/AI.2024.30001 Cite this Article Set citation alerts
    Guanghui Yuan, "Diffractive neural networks enabling superoscillatory imaging without sidelobes," Adv. Imaging 1, 033001 (2024) Copy Citation Text show less
    Superoscillatory diffractive neural networks (SODNNs) for super-resolution imaging without sidelobes[2]. (a) SODNN optimization procedures with 3D optical field constraints. (b) SODNN optimized optical superoscillatory spots and (c) an optical needle with long depth-of-focus. (d) Height distribution (column 1), scanning electron microscope (SEM) image (column 2) of the fabricated SODNN diffractive modulation layer, and its simulated (column 3) and experimental (column 4) focusing performance. (e) Schematic of experimental setup for imaging resolution testing. (f) Imaging results by the commercial objective and SODNN.
    Fig. 1. Superoscillatory diffractive neural networks (SODNNs) for super-resolution imaging without sidelobes[2]. (a) SODNN optimization procedures with 3D optical field constraints. (b) SODNN optimized optical superoscillatory spots and (c) an optical needle with long depth-of-focus. (d) Height distribution (column 1), scanning electron microscope (SEM) image (column 2) of the fabricated SODNN diffractive modulation layer, and its simulated (column 3) and experimental (column 4) focusing performance. (e) Schematic of experimental setup for imaging resolution testing. (f) Imaging results by the commercial objective and SODNN.