Related Experiment Video
Updated: Jul 5, 2025

07:03
In Situ Measurement of Vacuum Window Birefringence using 25Mg+ Fluorescence
Published on: June 13, 2020
3.9K
Vision Measurement Method Based on Plate Glass Window Refraction Model in Tunnel Construction
Zhen Wu1, Junzhou Huo1,2, Haidong Zhang1
1School of Mechanical Engineering, Dalian University of Technology, Dalian 116024, China.
Sensors (Basel, Switzerland)
|January 11, 2024
Summary
This study introduces a novel vision system for tunnel construction, correcting for refractive errors caused by protective glass. The new model significantly improves depth and positioning accuracy in harsh environments.
Area of Science:
- Optics and Vision Systems
- Robotics and Automation
- Geotechnical Engineering
Background:
- Tunnel construction environments necessitate explosion-proof vision systems with protective glass.
- Plate glass windows introduce optical refraction, invalidating standard pinhole camera models.
- Existing binocular vision systems fail due to refraction, impacting measurement accuracy.
Purpose of the Study:
- To develop a comprehensive solution for non-parallel plane refraction in vision measurement systems.
- To establish accurate imaging and triangulation models that account for refractive effects.
- To enhance the precision of depth and positioning measurements in challenging construction settings.
Main Methods:
- Established a non-parallel plane refraction imaging model using dynamic virtual focal length and the Rodriguez formula.
- Developed an epipolar constraint model specifically for independent refractive plane imaging.
- Proposed an independent refraction plane triangulation model to overcome measurement failures.
Main Results:
- The independent refraction plane triangulation model achieved a Root Mean Square Error (RMSE) of 0.3187 mm for depth errors after correction (down from 2.9902 mm).
- Positioning errors were reduced to an RMSE of 0.3465 mm post-correction (from 3.5661 mm).
- The proposed models effectively correct for refractive distortions in vision systems.
Conclusions:
- The developed models provide a robust solution for vision measurement systems operating behind refractive surfaces.
- Significant improvements in accuracy demonstrate the practical applicability in tunnel construction and similar environments.
- This research enhances the reliability and precision of vision-based measurements under adverse conditions.

