Related Experiment Video
Updated: Aug 31, 2025

Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
Published on: January 5, 2024
Concrete Surface Crack Recognition Based on Coordinate Attention Neural Networks
1School of Logistics and Transportation, Central South University of Forestry and Technology, Changsha 410004, China.
This study introduces CaNet, a deep learning model for concrete crack identification in transportation infrastructure. CaNet enhances crack detection accuracy, improving infrastructure maintenance and safety.
Area of Science:
- Civil Engineering
- Computer Vision
- Artificial Intelligence
Background:
- Concrete cracks are a prevalent issue in highway and tunnel infrastructure, compromising structural integrity and safety.
- Effective identification and reporting of concrete cracks are crucial for infrastructure maintenance, preventing economic losses and ensuring public safety.
Purpose of the Study:
- To propose a novel deep learning network, CaNet, for accurate and efficient identification of concrete cracks.
- To enhance the detection capabilities for narrow crack areas within concrete structures.
Main Methods:
- Developed CaNet, a deep learning network utilizing ResNet50 as its backbone.
- Integrated coordinate attention mechanism into ResNet50's residual units to improve capture of cross-channel, direction-aware, and position-sensitive information.
- Focused on enhancing the network's ability to locate small crack regions.
Main Results:
- CaNet achieved an accuracy rate of 89.6% in experiments, outperforming compared networks.
- The model demonstrated strong performance with a recall of 86%, F1 score of 85%, and precision of 87%.
Conclusions:
- The CaNet model is highly effective for identifying concrete cracks in transportation infrastructure.
- The integration of coordinate attention significantly improves the network's precision in locating narrow crack areas.
Related Concept Videos
Types of Non-structural Cracks in Concrete
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.
Microcracking in Concrete
Design Example: Joints in Concrete Pavements
Contraction joints are typically formed by sawing a groove into the concrete shortly after it has hardened. This creates a weakened vertical plane, deliberately encouraging cracking at...
Structural Classification of Joints
A fibrous joint is where the adjacent bones are united by fibrous connective...
Creep in Concrete
Tensile Strength Considerations of Concrete
The dimensions and shape of a concrete specimen...

