Published since 1923
DOI: 10.33622/0869-7019
Russian Science Citation Index (RSCI) на платформе Web of Science


  • BUILDING STRUCTURES, BUILDINGS AND FACILITIES
  • Convolutional Neural Networks For Detecting Defects And Structural Damage
  • UDC 69.07:004.8
    doi: 10.33622/0869-7019.2024.09.52-58
    Dmitry V. STEPANOV1, soyzstal@mail.ru
    Alexander V. MAKAROV1, soyzstal@mail.ru
    Alexey M. MOLOTOV2, molotov@a-2-m.ru
    Evgeny N. OBLETOV1,3, evg.obletov@gmail.com
    1 Soyuzstal'konstrukciya, ul. Maksima Gor'kogo, 262, office P250, Nizhniy Novgorod 603155, Russian Federation
    2 A2M, k. 313A, pom. I, kom. 13, Moscow, Zelenograd 124482, Russian Federation
    3 Nizhny Novgorod State University of Architecture and Civil Engineering (NNGASU), ul. Il'inskaya, 65, Nizhniy Novgorod 603950, Russian Federation
    Abstract. The possibilities of the NEIMARKER digital platform for automatic recognition of defects and damages of building structures based on computer vision algorithms and deep learning of neural networks are presented.The main functional advantages of the developed automated system are described, including recognition of various types of structures, identification of the main types of defects, the possibility of retraining neural networks to identify new types of damage. Information is provided on the achieved high accuracy of defect recognition, which ensures the practical applicability of the solution in engineering companies. The implementation of the proposed system makes it possible to reduce the time for processing the results of the inspection of objects, improve the quality of the final documentation and reduce the number of errors. The prospects for further development and application of the developed digital platform in various industries are outlined.
    Keywords: NEIMARKER digital platform, convolutional neural networks (CNN), neural network training, structural defect recognition, technical inspection of objects, machine vision
  • REFERENCES
    1. Naumov A. E., Yudin D. A., Dolzhenko A. V. Improvement of construction technology and technical expertise using the hardware and software complex of automated flaw detection. Vestnik BGTU im. V. G Shuhova, 2019, no. 4, pp. 61-69. (In Russ.).
    2. Kim B., Cho S. Automated vision-based detection of cracks on concrete surfaces using a deep learning technique [Автоматизированное визуальное обнаружение трещин на бетонных поверхностях с использованием метода глубокого обучения]. Sensors (Basel), 2018, vol. 18 (10), pp. 1-18. doi: 10.3390/s18103452
    3. Knyazeva N. V., Nazojkin E. A., Orekhov A. A. The use of artificial intelligence to detect defects in building structures. Stroitel'stvo i arhitektura, 2023, no. 3, pp. 18. (In Russ.). doi: 10.29039/2308-0191-2023-11-3-18-18
    4. Zou Q., Zhang Z., Li Q. et al. DeepCrack: Learning hierarchical convolutional features for crack detection [DeepCrack: изучение иерархических сверточных функций для обнаружения трещин]. IEEE Transactions on Image Processing, 2019, vol. 28(3), pp. 1498-1512. doi: 10.1109/TIP.2018.2878966
    5. Zhao W., Chen F. et al. A new steel defect detection algorithm based on deep learning [Новый алгоритм обнаружения дефектов стали на основе глубокого обучения]. Computational Intelligence and Neuroscience, 2021, vol. 10, pp. 1-13. doi: 10.1155/2021/5592878
    6. Kyong N. T., Syryamkin V. I., Thui N. Ch. H. A model of the object recognition method in images using a "convolutional neural network - CNN". Sovremennye naukoemkie tekhnologii, 2020, no. 12-2, pp. 269-280. (In Russ.).
    7. Zhang L., Yang F., Zhang Y. D., Zhu Y. J. Road crack detection using deep convolutional neural network [Обнаружение дорожных трещин с помощью глубокой сверточной нейронной сети]. IEEE International Conference on Image Processing, 2019, pp. 2869-2873. doi: 10.1109/IVS.2019.8814000
    8. Xu Y., Ding Z. et al. The steel surface multiple defect detection and size measurement system based on improved YOLOv5 [Система обнаружения множественных дефектов поверхности стали и измерения размеров на основе усовершенствованного YOLOv5]. Journal of Electrical and Computer Engineering, 2023, pp. 1-16. doi: 10.1155/2023/5399616
    9. Samohvalov I. A., Tryanina N. Yu., Obletov E. N. Stress-strain state of a steel tower taking into account corrosion damage to elements and joints. Promyshlennoe i grazhdanskoe stroitel'stvo, 2024, no. 5, pp. 54-60. (In Russ.). doi: 10.33622/0869-7019.2024.05.54-60
    10. Lao W., Cui Ch. et al. Computer vision-based autonomous method for quantitative detection of loose bolts in bolted connections of steel structures [Автономный метод компьютерного зрения для количественного обнаружения ослабленных болтов в болтовых соединениях стальных конструкций]. Structural Control and Health Monitoring, 2023, pp. 1-17. doi: 10.1155/2023/8817058
    11. Yao N., Zhao Y., Wu X. et al. Bolt defect detection based on centerNet model [Обнаружение дефектов болтов на основе модели CenterNet]. IEEE International Conference on Dependable, Autonomic and Secure Computing, 2021, pp. 731-737. doi: 10.1109/DASC-PICom-CBDCom- CyberSciTech52372.2021.00121
    12. Zhou J., Huo L. Computer vision-based detection for delayed fracture of bolts in steel bridges [Обнаружение замедленного разрушения болтов в стальных мостах с помощью компьютерного зрения]. Journal of Sensors, 2021, pp. 1-12. doi: 10.1155/2021/8325398 13. Брекоткин И. А., Рабовская М. Я. Применение нейронной сети для распознавания сварочных дефектов // Молодой ученый. 2024. № 19(518). С. 9-13.
    13. Brekotkin I. A., Rabovskaya M. Ya. The use of a neural network for the recognition of welding defects. Molodoj uchyonyj, 2024, no. 19(518), pp. 9-13. (In Russ.).
  • For citation: Stepanov D. V., Makarov A. V., Molotov A. M., Obletov E. N. Convolutional Neural Networks for Detecting Defects and Structural Damage. Promyshlennoe i grazhdanskoe stroitel'stvo [Industrial and Civil Engineering], 2024, no. 9, pp. 52-58. (In Russ.). doi: 10.33622/0869-7019.2024.09.52-58

BACK