Article

Received 14.07.2024

Revised 08.11.2024

Accepted 14.12.2024

Retrieved from Iss. 116, P. 2, 2024

Pages 74 -80

Gameliak, I., Kharchenko, A., & Dmytrychenko, A. (2024). FUNDAMENTALS OF THE METHODOLOGY FOR RECOGNIZING DEFECTS IN CEMENT CONCRETE PAVEMENTS. Automobile Roads and Road Construction, (116.2), 74-80. https://doi.org/10.33744/0365-8171-2024-116.2-074-080

FUNDAMENTALS OF THE METHODOLOGY FOR RECOGNIZING DEFECTS IN CEMENT CONCRETE PAVEMENTS

Igor Gameliak Anna Kharchenko Andrii Dmytrychenko

The article is devoted to the results of the development of the foundations for building a methodology for recognizing defects in cement concrete coatings. An analysis of foreign sources regarding progressive methods for processing the results of photofixation of defects on critical infrastructure objects has been performed. A general algorithm for building a methodology for recognizing defects has been developed, which is based on five main steps, in particular, on collecting and summarizing data with subsequent preprocessing and filtering; improving image quality and eliminating background noise using image processing methods; using machine learning models for detection and classification; checking the correctness of the detected defect; calibrating the machine learning algorithm. It is proposed to use the harmonic mean index to assess the efficiency of the machine learning model and as a metric for its calibration. It was determined that the practical testing of the study will be carried out within the framework of the authors' participation in the project 2023.04/0097 "Creation of a technology and a system of operational analysis and management of cement-concrete pavement condition of critical infrastructure objects based on spectral photoinformational images", financed by the National Research Fund of Ukraine

recognition methodology, defects in cement-concrete pavements, machine learning, recognition, calibration, average harmonic index
  1. Kolappan Geetha, G., & Sim, S. (2022). Fast identification of concrete cracks using 1D deep learning and explainable artificial intelligence-based analysis. Automation in Construction, advance online publication.doi: 10.1016/j.autcon.2022.104572.
  2. Choi, Y., Park, H.W., Mi, Y., & Song, S. (2024). Crack detection and analysis of concrete structures based on neural network and clustering. Sensors, 24(6), article number 1725. doi: 10.3390/s24061725.
  3. Kim, H., Ahn, E., Shin, M., & Sim, S.-H. (2019). Crack and noncrack classification from concrete surface images using machine learning. Structural Health Monitoring, 18(3), 725-738. doi: 10.1177/1475921718768747.
  4. Iraniparast, M., Ranjbar, S., Rahai, M., & Nejad, F.M. (2023). Surface concrete cracks detection and segmentation using transfer learning and multi-resolution image processing. Structures, 54, 386-398. doi: 10.1016/j.istruc.2023.05.062.
  5. Padmapoorani, P., & Senthilkumar, S. (2023). Application of machine learning for crack detection on concrete structures using CNN architecture. Matéria (Rio de Janeiro), 28(1), article number e20230010. doi; 10.1590/1517-7076-rmat-2023-0010.
  6. Wikipedia. (n.d.). Digital image noise. Retrieved from https://en.wikipedia.org/wiki/Image_noise.
  7. The Transmitted. (2023). F1 score in machine learning. Retrieved from https://thetransmitted.com/adlucem/pokaznyk-f1-umashynnomu-navchanni/.

https://doi.org/10.33744/0365-8171-2024-116.2-074-080