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 19, 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

799

Computer Vision-Based Bridge Damage Detection Using Deep Convolutional Networks with Expectation Maximum Attention

Wenting Qiao1,2, Biao Ma3, Qiangwei Liu3

  • 1School of Highway, Chang'an University, Xi'an 710064, Shaanxi, China.

Sensors (Basel, Switzerland)
|February 3, 2021
PubMed
Summary

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

Curcumin-loaded tetrahedral framework nucleic acids reshape intestinal homeostasis in ulcerative colitis by coordinating barrier repair, microbiota reconstruction, and immune modulation.

Materials today. Bio·2026
Same author

Biphasic Memory Impairment and Recovery After Sevoflurane Exposure Are Associated With Time-Dependent Hippocampal α5-GABAAR Remodeling.

CNS neuroscience & therapeutics·2026
Same author

[Research on a muscle fatigue quantification based on an improved critical power model].

Sheng wu yi xue gong cheng xue za zhi = Journal of biomedical engineering = Shengwu yixue gongchengxue zazhi·2026
Same author

Mas-Related G Protein-Coupled Receptor Member D Sustains Hypertension.

MedComm·2026
Same author

Effect of career calling on learning engagement among medical students: moderated chain mediation model.

Frontiers in psychology·2026
Same author

Multimodal radiomics prediction of long-term postoperative prognosis in non-small cell lung cancer patients based on CT imaging.

Current problems in surgery·2026
This summary is machine-generated.

A new algorithm, EMA-DenseNet, improves bridge surface damage detection by integrating an expected maximum attention (EMA) module and a novel pixel connectivity loss function. This enhances accuracy and robustness in identifying cracks and exposed steel bars, crucial for bridge longevity.

Area of Science:

  • Civil Engineering
  • Computer Vision
  • Artificial Intelligence

Background:

  • Bridge integrity is threatened by surface damage like cracks and exposed steel.
  • Automated detection of bridge surface damage is challenging due to complex structures and environmental factors.
  • Convolutional neural networks show promise but struggle with real-world environmental impacts.

Purpose of the Study:

  • To propose a novel algorithm, EMA-DenseNet, for accurate and robust detection of bridge surface damage.
  • To enhance feature extraction for bridge damage using an expected maximum attention (EMA) module.
  • To improve fracture prediction accuracy by reducing break points with a new pixel connectivity-aware loss function.

Main Methods:

  • Redesigned the structure of densely connected convolutional networks (DenseNet).
Keywords:
bridge damage detectiondeep convolutional networksdensely connected networksexpected maximum attention

Related Experiment Videos

Last Updated: Nov 19, 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

799
  • Integrated an expected maximum attention (EMA) module after the last pooling layer.
  • Developed and applied a new loss function that considers pixel connectivity.

    • Main Results:

    • EMA-DenseNet achieved high performance on an open concrete crack dataset (MPA: 87.42%, MIoU: 92.59%, Precision: 81.97%, FPS: 25.4).
    • On a challenging bridge damage dataset, EMA-DenseNet demonstrated strong results (MPA: 79.87%, MIoU: 86.35%, Precision: 74.70%, FPS: 14.6).
    • The proposed algorithm outperformed current state-of-the-art methods in accuracy and robustness.

    Conclusions:

    • The EMA-DenseNet algorithm effectively detects bridge surface damage.
    • The EMA module significantly aids in bridge damage feature extraction.
    • The novel loss function improves prediction accuracy and reduces fracture break points, enhancing bridge inspection capabilities.