Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 17, 2026PLoS ONEOpen Access

CrackNet: A novel multi-scale architecture for crack segmentation

View Full Paper
Ask AI
Bookmark
Share

Authors

WZWubiao ZhuMYMengcai YeJYJiawei YinShanghai University

Discussion

Loading...

Member takes

Implication

Novel segmentation network improves crack detection accuracy in engineering applications, suggesting enhanced safety measures.

Key Points

  • The aim is to develop an efficient network for accurate concrete crack detection amidst challenging conditions.
  • Designed CrackNet with specific modules for crack detection, including LightMSCBlock, SAF, and MSFF.
  • Conducted extensive experiments on three public datasets: CFD, Crack500, and DeepCrack.
  • Performed ablation studies to evaluate the contribution of individual modules.
  • Achieved F1 and IoU improvements of 6.37% and 7.1% on CFD compared to SegFormer.
  • Increased F1 score by 3.86% on Crack500 compared to MobileNetV3-UNet.
  • Enhanced F1 and IoU gains of 5.7% and 2.5% respectively on DeepCrack.

Cite This Study

Zhu et al. (2026) studied this question.

synapsesocial.com/papers/69e1cfcb5cdc762e9d858b9bhttps://doi.org/10.1371/journal.pone.0346889
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Improved U-net network asphalt pavement crack detection method2024 · 26 citations
  2. 2VGG‐EffAttnNet: Hybrid Deep Learning Model for Automated Chili Plant Disease Classification Using VGG16 and EfficientNetB0 With Attention Mechanism2025 · 35 citations
  3. 3An Improved ResNet-Based Algorithm for Crack Detection of Concrete Dams Using Dynamic Knowledge Distillation2023 · 18 citations
  4. 4Semantic Enrichment for Building Information Modeling2015 · 192 citations
  5. 5Automatic pavement crack segmentation using a generative adversarial network (GAN)-based convolutional neural network2023 · 48 citations