Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Aug 25, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

608

Single-pixel imaging for edge images using deep neural networks.

Ikuo Hoshi, Masaki Takehana, Tomoyoshi Shimobaba

    Applied Optics
    |October 18, 2022
    PubMed
    Summary
    This summary is machine-generated.

    Related Concept Videos

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    Hologram computation based on sparse matrix multiplication.

    Optics express·2026
    Same author

    Modified split-Lohmann holography: a shift- and ringing-free approach.

    Optics letters·2026
    Same author

    Sub-sampled single-step Fresnel diffraction for efficient computation of high-resolution holograms.

    Optics letters·2026
    Same author

    Optical initialization and manipulation of higher-order electron states with spin and orbital angular momentum in a semiconductor quantum disk.

    Optics express·2025
    Same author

    Asymmetric point-spread function in the tilted plane.

    Optics express·2025
    Same author

    Quantized neural network for complex hologram generation.

    Applied optics·2025

    This study introduces two deep neural networks for single-pixel imaging (SPI) edge image generation without shifting illumination patterns. This method enhances object identification in computer vision and low-light imaging applications.

    Area of Science:

    • Computer Vision
    • Optical Imaging
    • Machine Learning

    Background:

    • Edge images are crucial for object identification in computer vision, cellular morphology, and surveillance.
    • Single-pixel imaging (SPI) offers a promising approach for wide-wavelength, low-light-level measurements.
    • Conventional SPI edge enhancement requires shifting illumination patterns, increasing complexity.

    Purpose of the Study:

    • To develop novel deep neural networks for generating SPI-based edge images.
    • To eliminate the need for shifting illumination patterns in SPI edge imaging.
    • To improve the efficiency and applicability of SPI for edge detection.

    Main Methods:

    • Proposed two deep neural network architectures for SPI edge image reconstruction.

    More Related Videos

    Deep Neural Networks for Image-Based Dietary Assessment
    13:19

    Deep Neural Networks for Image-Based Dietary Assessment

    Published on: March 13, 2021

    9.3K
    Deep Learning-Based Segmentation of Cryo-Electron Tomograms
    10:25

    Deep Learning-Based Segmentation of Cryo-Electron Tomograms

    Published on: November 11, 2022

    9.2K

    Related Experiment Videos

    Last Updated: Aug 25, 2025

    Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
    03:31

    Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

    Published on: December 15, 2023

    608
    Deep Neural Networks for Image-Based Dietary Assessment
    13:19

    Deep Neural Networks for Image-Based Dietary Assessment

    Published on: March 13, 2021

    9.3K
    Deep Learning-Based Segmentation of Cryo-Electron Tomograms
    10:25

    Deep Learning-Based Segmentation of Cryo-Electron Tomograms

    Published on: November 11, 2022

    9.2K
  • The first network performs end-to-end mapping from measured intensities to edge images.
  • The second network uses dual-path convolutional layers to reconstruct horizontal and vertical edges separately.

    • Main Results:

    • Successfully generated SPI-based edge images without requiring shifting illumination patterns.
    • Demonstrated the ability to reconstruct comprehensive edge information, comparable to traditional methods like the Sobel filter.
    • The proposed networks provide an efficient alternative for SPI edge image acquisition.

    Conclusions:

    • The developed deep neural networks offer a significant advancement in SPI-based edge imaging.
    • This technique simplifies SPI by removing the need for pattern shifting, broadening its practical applications.
    • The method holds potential for enhanced object recognition and analysis in various imaging domains.