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: Nov 27, 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

824

A New Deep Learning Based Multi-Spectral Image Fusion Method.

Jingchun Piao1, Yunfan Chen1, Hyunchul Shin1

  • 1Department of Electrical Engineering, Hanyang University, Ansan 15588, Korea.

Entropy (Basel, Switzerland)
|December 3, 2020
PubMed
Summary

This study introduces a novel deep neural network for infrared and visible image fusion. The method enhances image fusion by automatically generating weight maps, improving pedestrian detection and visual quality.

Related Concept Videos

Deconvolution01:20

Deconvolution

425
Deconvolution, also known as inverse filtering, is the process of extracting the impulse response from known input and output signals. This technique is vital in scenarios where the system's characteristics are unknown, and they must be inferred from the observable signals.
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
425

You might also read

Related Articles

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

Sort by
Same author

Post-IMvigor011 Era: CtDNA-Guided Postoperative Adjuvant Treatment Stratification in Muscle-Invasive Bladder Cancer.

Cancer management and research·2026
Same author

TFDF: Self-Supervised Time-Frequency Dynamic Fusion with Dual Constraints for Atrial Fibrillation Detection.

IEEE journal of biomedical and health informatics·2026
Same author

A machine learning-assisted Ag-TiO<sub>2</sub> SERS platform for intraoperative osteomyelitis diagnosis.

Nanoscale·2026
Same author

Multiscale Feature Enhancement and Bidirectional Temporal Dependency Networks for Arrhythmia Classification.

Biology·2026
Same author

Data Augmentation-Enhanced Myocardial Infarction Classification and Localization Using a ResNet-Transformer Cascaded Network.

Biology·2025
Same author

A Multi-Domain Feature Fusion CNN for Myocardial Infarction Detection and Localization.

Biosensors·2025

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Image Processing

Background:

  • Image fusion combines information from multiple source images.
  • Traditional methods struggle with activity level measurement and fusion rule design.
  • Deep learning offers potential for automated image fusion.

Purpose of the Study:

  • To develop an effective infrared (IR) and visible (VIS) image fusion method using deep neural networks.
  • To address key challenges in image fusion, namely activity level measurement and fusion rule design.
  • To improve the perceptual quality and utility of fused images for downstream tasks like object detection.

Main Methods:

  • A Siamese convolutional neural network (CNN) is employed to automatically generate pixel-wise weight maps indicating image saliency.
Keywords:
Siamese networkconvolutional neural networkimage fusioninfraredvisible

Related Experiment Videos

Last Updated: Nov 27, 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

824
  • The CNN encodes images into a feature domain for effective analysis.
  • Fusion is performed using multi-scale image decomposition via wavelet transform.

    • Main Results:

    • The proposed method automatically determines activity levels and fusion rules, simplifying the process.
    • Fused images exhibit improved perceptual quality, benefiting human visual perception.
    • Evaluation using pedestrian detection with YOLOv3 on a public dataset demonstrated competitive quantitative and visual results.

    Conclusions:

    • The deep neural network-based image fusion method is effective and efficient.
    • The approach successfully addresses limitations of traditional image fusion techniques.
    • The method shows promise for enhancing visual information and improving object detection performance.