Related Experiment Video
Updated: Jan 10, 2026
12:45
Detection and Quantification of Tunneling Nanotubes Using 3D Volume View Images
Published on: August 31, 2022
3.5K
A lightweight fully convolutional neural network for efficient and accurate tunnel crack detection.
Hao Huang1, G T S Ho2, V Tang2
1School of Integrated Circuits, Harbin Institute of Technology, Shenzhen, 518055 China.
Journal of Advanced Research
|November 26, 2025
Summary
FCN CrackDetect offers a fast and accurate solution for automated tunnel crack detection. This lightweight model achieves high performance without needing data from unseen environments, addressing computational limitations in tunnel maintenance.
Area of Science:
- Civil Engineering
- Computer Vision
- Artificial Intelligence
Background:
- Tunnel structural integrity relies on effective crack detection.
- Existing methods are often computationally intensive, limiting their use during brief inspection windows.
- Automated detection is crucial for efficient tunnel maintenance.
Purpose of the Study:
- To develop a lightweight, high-accuracy model for automated tunnel crack detection.
- To ensure reliable performance across diverse tunnel environments despite data gaps.
- To achieve strong generalization without explicit domain adaptation modules.
Main Methods:
- A U-Net-based architecture incorporating EfficientNetV2S for feature extraction.
- Integration of DropBlock regularization and Efficient Channel Attention (ECA) for implicit domain generalization.
- Training models to learn domain-invariant features for enhanced resilience.
Main Results:
- FCN CrackDetect achieved superior accuracy and a rapid runtime of 29 ms on 6000 tunnel crack images.
- Ablation studies confirmed the effectiveness of integrated components like EfficientNetV2S and attention mechanisms.
- The model demonstrated robust performance without requiring data from unseen tunnel environments during training.
Conclusions:
- FCN CrackDetect presents a practical and high-speed method for automated tunnel crack detection.
- It serves as an effective alternative to computationally demanding adaptation frameworks.
- The model's design promotes generalization for reliable performance in real-world tunnel inspections.
Related Concept Videos
01:20Microcracking in Concrete
414
Microcracking in concrete refers to the tiny cracks that can form within the material even before any external load is applied. These microcracks typically occur at the interface between the coarse aggregate and the hydrated cement paste, often as a result of differential volume changes prompted by variations in stress-strain behavior, as well as thermal and moisture movement. Initially, these microcracks remain stable and do not grow substantially until the concrete is stressed to about 30...
414
01:28Types of Non-structural Cracks in Concrete
465
Non-structural cracks are primarily of three types: plastic, early-age thermal, and drying shrinkage cracks. Plastic cracks are further classified into plastic shrinkage cracks and plastic settlement cracks.
Plastic shrinkage cracks typically form within hours after the concrete is poured. The concrete's surface dries faster than the bottom, creating tensile stress that the still-plastic concrete cannot withstand, leading to diagonal or randomly patterned cracks on the concrete surface.
Plastic shrinkage cracks typically form within hours after the concrete is poured. The concrete's surface dries faster than the bottom, creating tensile stress that the still-plastic concrete cannot withstand, leading to diagonal or randomly patterned cracks on the concrete surface.
465