Related Experiment Video
Updated: Dec 10, 2025
12:45
Detection and Quantification of Tunneling Nanotubes Using 3D Volume View Images
Published on: August 31, 2022
3.4K
Vision Measurement of Tunnel Structures with Robust Modelling and Deep Learning Algorithms
1School of Rail Transit, Soochow University, Suzhou 215006, China.
Sensors (Basel, Switzerland)
|September 5, 2020
Summary
This study introduces an automated tunnel health monitoring system using image data. It combines 3D modeling and deep learning for accurate deformation and crack detection, enhancing railway safety.
Area of Science:
- Civil Engineering
- Geotechnical Engineering
- Computer Vision
Background:
- Tunnel structural health is critical for railway safety.
- Traditional inspection methods are labor-intensive, costly, and subjective.
- Increasing tunnel infrastructure necessitates advanced monitoring techniques.
Purpose of the Study:
- To develop an automated tunnel health monitoring method using image data.
- To create a robust 3D tunnel model for deformation analysis.
- To intelligently detect tunnel cracks using deep learning.
Main Methods:
- Utilized a moving vision measurement unit with a camera array for data acquisition.
- Reconstructed a 3D tunnel model using a B-spline method for geometric analysis.
- Employed a Mask R-CNN deep learning network for automated crack identification.
Main Results:
- Successfully generated a robust 3D model of a multi-kilometer subway tunnel.
- Automatically identified cracks within the tunnel structure using image data.
- Demonstrated the system's capability for comprehensive deformation and crack assessment.
Conclusions:
- The proposed method offers an accurate and reliable approach to tunnel health monitoring.
- Combining 3D modeling and deep learning enhances the efficiency and objectivity of inspections.
- This automated system improves the safety and maintenance of critical tunnel infrastructure.
