Computer Vision and Augmented Reality for Human-Centered Fatigue Crack Inspection

Area of Science:

• Structural Engineering
• Computer Vision
• Human-Computer Interaction

Background:

• Many US bridges exceed their design life, increasing vulnerability to fatigue cracks.
• Current visual inspection methods are slow, labor-intensive, and error-prone.
• Advanced technologies are needed to improve bridge structural health monitoring.

Purpose of the Study:

• To develop a novel human-centered bridge inspection methodology.
• To enhance the efficiency and accuracy of fatigue crack detection.
• To integrate computer vision and augmented reality (AR) for real-time analysis.

Main Methods:

• A computer vision algorithm analyzes structural surface motion from AR headset video.
• Feature point tracking and distance variation measurement detect crack opening/closing patterns.
• An AR environment visualizes crack detection results as holograms and supports human-in-the-loop decision-making.

Main Results:

• The method enables near-real-time, computationally efficient fatigue crack detection.
• Elimination of camera motion compensation enhances reliability.
• Integrated AR visualization aids inspectors in making informed field decisions.

Conclusions:

• The proposed methodology significantly improves fatigue crack detection efficiency and accuracy.
• The human-centered approach with AR enhances visualization and human-machine collaboration.
• The system demonstrates efficacy for real-world bridge inspection applications.