Adaptive deep learning framework for crack contour recognition and dimensional measurement in concrete structures

Key Points

• Crack contour recognition is enhanced through a deep learning approach, enabling more precise dimensional measurements.
• The study reveals an adaptive framework that improves accuracy in crack detection and dimensional assessment.
• Utilizing deep learning techniques, the framework assesses concrete structures for structural integrity during inspections.
• The findings may enable more effective monitoring and maintenance of concrete infrastructures, reducing potential failures.