EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Image recognition technology for bituminous concrete reservoir panel cracks based on deep learning

Kai Hu iD, Yang Ling, Jie Liu

PLOS ONE · 2025

DOI: 10.1371/journal.pone.0318550 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS ONE, 2025. Verfasst von Kai Hu und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Civil Engineering and Architecture, Xi'an Technological University, Xi 'an, Shaanxi, China. Hauptthemen laut MEDLINE: Deep Learning, Construction Materials, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Detecting cracks in asphalt concrete slabs is challenging due to environmental factors like lighting changes, surface reflections, and weather conditions, which affect image quality and crack detection accuracy. This study introduces a novel deep learning-based anomaly model for effective crack detection. A large dataset of panel images was collected and processed using denoising, standardization, and data augmentation techniques, with crack areas labeled via LabelImg software. The core model is an improved Xception network, enhanced with an adaptive activation function, dynamic attention mechanism, and multi-level residual connections. These innovations optimize feature extraction, enhance feature weighting, and improve information transmission, significantly boosting accuracy and robustness. The improved model achieves a 97.6% accuracy and a Matthews correlation coefficient of 0.98, remaining stable under varying lighting conditions. This method not only provides a fresh approach to crack detection but also greatly enhances detection efficiency.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningConstruction MaterialsImage Processing, Computer-AssistedAlgorithmsHydrocarbons

Substanzen

asphaltHydrocarbons

Autor:innen und Institutionen

  • School of Civil Engineering and Architecture, Xi'an Technological University, Xi 'an, Shaanxi, China.
  • Shaanxi Qinyuan Tendering Co., Ltd, Xi 'an, Shaanxi, China.
  • Shanxi Yuanqu Pumped Storage Co., Ltd, Yuncheng, Shanxi, China.

ORCID-Profile: Kai Hu

Zitieren

APA

Hu, K., Ling, Y., & Liu, J. (2025). Image recognition technology for bituminous concrete reservoir panel cracks based on deep learning. PLOS ONE. https://doi.org/10.1371/journal.pone.0318550

Vancouver

Hu K, Ling Y, Liu J. Image recognition technology for bituminous concrete reservoir panel cracks based on deep learning. PLOS ONE. 2025. doi:10.1371/journal.pone.0318550

BibTeX
@article{Hu2025,
  author = {Hu, Kai and Ling, Yang and Liu, Jie},
  title = {Image recognition technology for bituminous concrete reservoir panel cracks based on deep learning},
  journal = {PLOS ONE},
  year = {2025},
  doi = {10.1371/journal.pone.0318550}
}
Kostenlos starten →