What question did this study set out to answer?

This research aims to develop a high-precision, non-contact method for monitoring cracks in heritage buildings over time.

Crack Detection and Evolution Monitoring for Heritage Buildings: A Systematic Computer Vision-Based Approach

View Full Paper

Ask AI

Bookmark

Authors

Systematic computer vision method monitors crack evolution in heritage buildings, supporting preservation efforts.

This research aims to develop a high-precision, non-contact method for monitoring cracks in heritage buildings over time.
Proposed a framework integrating deep learning, image stitching, and skeleton-based analysis.
Utilized U 2 -Net, Mask R-CNN, and DeepCrack for crack detection and evaluated their performance.
Implemented camera calibration for millimeter-scale crack measurements.
Applied sliding-window matching for temporal monitoring of crack evolution.
Achieved robust crack detection even in complex backgrounds.
DeepCrack outperformed other models in connectivity and stability for crack extraction.
Successfully monitored crack evolution with quantitative metrics including mean and standard deviation.

Ye et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark