Identification, 3D-Reconstruction, and Classification of Dangerous Road Cracks.

Area of Science:

• Civil Engineering
• Computer Vision
• Materials Science

Background:

• Automated infrastructure inspection is advancing, but road crack identification remains challenging due to image acquisition difficulties and small crack sizes.
• Manual road inspection is time-consuming, expensive, and subjective, necessitating efficient automated solutions.
• Pavement cracks and potholes degrade infrastructure value and pose safety risks, requiring accurate severity assessment.

Purpose of the Study:

• To develop an efficient automated system for road crack identification, extraction, and 3D reconstruction.
• To improve the accuracy and speed of pavement defect assessment.
• To contribute to safer road infrastructure through advanced detection methods.

Main Methods:

• Image processing techniques including Gaussian filtering, contrast adjustment, and threshold-based segmentation were employed.
• A novel five-stage automated system was designed for flaw detection.
• A dedicated database of road cracks was created for method validation.

Main Results:

• The proposed automated system demonstrated commendable performance in identifying and analyzing road cracks.
• The method significantly outperformed existing state-of-the-art approaches in crack detection and reconstruction.
• The system effectively addresses challenges related to image acquisition and the small scale of pavement defects.

Conclusions:

• The developed automated system offers an efficient and accurate solution for road crack identification and 3D reconstruction.
• This advancement has the potential to revolutionize road maintenance by providing objective and rapid assessments.
• Implementing such technologies is crucial for enhancing road safety and preserving infrastructure integrity.