Related Experiment Video
Updated: Jan 17, 2026
12:45
Detection and Quantification of Tunneling Nanotubes Using 3D Volume View Images
Published on: August 31, 2022
3.6K
Research on Tunnel Crack Identification Localization and Segmentation Method Based on Improved YOLOX and UNETR+
Wei Sun1,2, Xiaohu Liu3, Zhiyong Lei4
1School of Mechatronic Engineering, Xi'an Technological University, Xi'an 710021, China.
Sensors (Basel, Switzerland)
|September 19, 2025
Summary
This study introduces an advanced method for detecting, locating, and segmenting fine tunnel cracks using improved YOLOX and UNETR++ models. The approach enhances accuracy for subtle crack features, improving tunnel safety inspections.
Area of Science:
- Civil Engineering
- Computer Vision
- Artificial Intelligence
Background:
- Identifying fine, irregular cracks in tunnels is crucial for structural integrity.
- Existing methods often struggle with the subtle features of these cracks.
Purpose of the Study:
- To develop a robust method for crack identification, localization, and segmentation in tunnels.
- To improve the detection accuracy of fine and irregular cracks.
Main Methods:
- Utilized an improved YOLOX algorithm with EfficientNet backbone and ECA module for enhanced feature extraction and sensitivity.
- Employed the UNETR++ network for efficient segmentation via global feature capture and multi-scale fusion.
Main Results:
- The proposed method effectively integrates identification, localization, and segmentation of tunnel cracks.
- Demonstrated high precision in detecting and segmenting fine and irregular cracks.
Conclusions:
- The novel approach significantly advances the automated analysis of tunnel crack defects.
- Offers a promising solution for improving tunnel structural health monitoring and maintenance.
