PSI - Issue 44

Agnese Natali et al. / Procedia Structural Integrity 44 (2023) 2020–2027 Agnese Natali et al./Structural Integrity Procedia 00 (2022) 000–000

2027

8

Fig. 11. Qualitative results on images from the PAVIS-FABRE dataset, using the trained model described in Fig. 11.

4. Conclusions and future development This paper presents an ongoing activity on the use of artificial intelligence and deep learning for damage detection of existing bridges. The AI tool is currently under development, facing the starting issues of recognizing the component under analysis and the individuation of the key defects. This tool may be a powerful instrument to support the process of the classification of the existing bridges, specifically referring to the prioritization and organization of the Level 0 and 1 activities which might be applied to a high quantity of activities. In this case, the criterion of prioritization is based on the conservation status of the structure, which is one of the most influencing parameters for the vulnerability parameter for the evaluation of the warning class of the bridge. This work is inserted in the general idea of computerisation and automated and intelligent management of actions for the management of existing structures. References Ansfisa (2022). Istruzioni Operative per l’applicazione delle Linee Guida per la classificazione e gestione del rischio, la valutazione della sicurezza ed il monitoraggio dei ponti esistenti. Ministero delle infrastrutture e della mobilità sostenibili (2020). Linee Guida per la classificazione e gestione del rischio, la valutazione della sicurezza ed il monitoraggio dei ponti esistenti. Mundt, M., Majumder, S., Murali, S., Panetsos, P., & Ramesh, V. (2019). Meta-Learning Convolutional Neural Architectures for Multi-Target Concrete Defect Classification With the COncrete DEfect BRidge IMage Dataset, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 11188-11197. Özgenel, Ç. F. (2019), “Concrete Crack Images for Classification”, Mendeley Data, V2. Özgenel, Ç.F., Gönenç Sorguç, A. (2018). Performance Comparison of Pretrained Convolutional Neural Networks on Crack Detection in Buildings, Proceedings of the 35th ISARC, 693—700. Zhang L., Yang F., Daniel Zhang Y., & Zhu Y. J. (2016). Road crack detection using deep convolutional neural network, 2016 IEEE International Conference on Image Processing (ICIP), 3708—3712. Simonyan, K. and Zisserman, A. (2015) Very Deep Convolutional Networks for Large-Scale Image Recognition. The 3rd International Conference on Learning Representations (ICLR2015). He, K., Zhang, X., Ren, S. & Sun, J. (2015). Deep Residual Learning for Image Recognition, Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition, 770—778. Chollet F. (2017). Xception: Deep Learning with Depthwise Separable Convolutions, Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition, 1800—1807. Padalkar M. G., Beltrán-González C., Bustreo M., Del Bue A., & Murino V. (2020). A Versatile Crack Inspection Portable System based on Classifier Ensemble and Controlled Illumination, 25th International Conference on Pattern Recognition (ICPR2020), 4009—4016. Sekachev B., Manovich N., and Zhavoronkov A. (2019). Computer Vision Annotation Tool, GitHub: https://github.com/opencv/cvat. Frangopol, D. M., Dong, Y., and Sabatino, S., 2017. Bridge life-cycle performance and cost: analysis, prediction, optimisation and decision making. Structure and Infrastructure Engineering, 13(10), 1239–1257.