An intelligent non-destructive method to identify the quality of self-compacting concrete based on convolutional neural networks via image recognition
Source abstract
Abstract
Self-compacting concrete (SCC) was used in numerous structures of harbor and offshore engineering because of its high production efficiency, less labor cost and strong resistance to chloride ion erosion in seawater. Sampling inspection not only cannot realize the comprehensive coverage of the inspected concrete, but also is complicated, damaging and costly. An intelligent non-destructive method to identify the quality of self-compacting concrete based on convolutional neural networks (CNN) via image recognition was proposed in this study. In order to correlate the image recognition of SCC with its characteristics, the fluidity, compressive strength and chloride ion diffusion coefficient of SCC were tested while images of SCC were acquired using a smartphone. The efficiency of CNN models including GoogleNet, ResNet18, ResNet34, ResNet50 and ShuffleNet were compared in terms of prediction accuracy and computation time, and the results demonstrated superiority of the ResNet18 model over all competing models. In addition, the performance of the proposed CNN model was evaluated with respect to the input image types and the parameters in CNN including dropout ratio, learning rate, size and resolution of the input images. The feasibility of training the CNN model with the images of the polished surfaces was demonstrated. Surface polishing does not compromise the strength of the concrete, and is a normal concrete construction operation to remove floating aggregates. The optimal model employed the largest size of polished surface images with a pixel resolution of 1266 × 1266 as the input data under the situation that learning rate and dropout ratio are 0.1 and 0, respectively. The R 2 of the proposed optimal CNN model can achieve 0.9987 with superior accuracy and strong generalization ability. It can provide a convenient and cheap method for the project management departments to achieve the intelligent, non-destructive and efficient supervision of concrete quality inspection.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Infrastructure Maintenance and Monitoring · Industrial Vision Systems and Defect Detection · Geophysical Methods and Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (41)
Show source-provided references
1. An investigation of mechanical properties of Self Compacting Concrete (SCC) after addition of Ground Granulated Blast Furnace Slag (GGBFS) and Fly Ash (FA)
2. Autonomous detection of concrete damage under fire conditions10.1016/j.autcon.2022.104364 ↗
3. Image-based reinforced concrete component mechanical damage recognition and structural safety rapid assessment using deep learning with frequency information10.1016/j.autcon.2023.104839 ↗
4. Assessment of concrete compressive strength by image processing technique10.1016/j.conbuildmat.2012.07.055 ↗
5. The bonding performance of desert sand self-compacting concrete overlay on normal strength concrete substrate: macro, micro, and ultrasonic testing
6. Determination of concrete compressive strength from surface images with the integration of CNN and SVR methods10.1016/j.measurement.2024.115331 ↗
7. To predict the compressive strength of self compacting concrete with recycled aggregates utilizing ensemble machine learning models
8. Deep Learning approach for estimating workability of self-compacting concrete from mixing image sequences10.1155/2018/6387930 ↗
9. Concrete compressive strength detection using image processing based new test method10.1016/j.measurement.2017.05.051 ↗
10. Experimental studies on behavior of hybrid materials concrete using non-destructive testing (NDT) methods10.1016/j.matpr.2023.04.061 ↗
11. He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep Residual Learning for Image Recognition, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV, USA, 770-778. http://dx.doi.org/10.1109/CVPR.2016.90.10.1109/CVPR.2016.90 ↗
12. Improving neural networks by preventing co-adaptation of feature detectors
13. Ioffe, S., Szegedy, C., 2015. Batch normalization: accelerating deep network training by reducing internal covariate shift, Proceedings of the 32nd International Conference on Machine Learning. Lille, France, 37, 448-456. https://doi.org/10.48550/arXiv.1502.03167.
14. JGJ/T 283 · 2012
15. JTS 202-2011 Specifications for Concrete Construction of Port and Waterway Engineering, China, 2011.
16. Workability, testing, and performance of self-consolidating concrete
17. Convolutional neural network-based regression for predicting the chloride ion diffusion coefficient of concrete
18. Effect of content and particle size distribution of coarse aggregate on the compressive strength of concrete10.1016/j.conbuildmat.2009.10.009 ↗
19. Nair, V., Hinton, G.E., 2010. Rectified linear units improve restricted boltzmann machines, Proceedings of the 27th International Conference on Machine Learning. Haifa, Israel, 2010, 807-814. 〈https://dl.acm.org/doi/10.5555/3104322.3104425〉.
20. Micro-concrete crack detection of underwater structures based on convolutional neural network10.1007/s00138-022-01327-5 ↗
21. Hammer rebound index as an overall-mechanical-quality indicator of self-compacting concrete containing recycled concrete aggregate10.1016/j.conbuildmat.2022.128549 ↗
22. Models for compressive strength estimation through non-destructive testing of highly self-compacting concrete containing recycled concrete aggregate and slag-based binder10.1016/j.conbuildmat.2021.122454 ↗
23. Salman, S., Liu, X., 2019. Overfitting Mechanism and Avoidance in Deep Neural Networks. https://doi.org/10.48550/arXiv.1901.06566.
24. High strength self-compacting concrete original solutions associating organic and inorganic admixtures10.1016/S0008-8846(99)00037-X ↗
25. Scherer, D., Müller, A., Behnke, S., 2010. Evaluation of pooling operations in convolutional architectures for object recognition, Proceedings of the 20th International Conference on Artificial Neural Networks, Thessaloniki, Greece, 6354, 92–101. https://doi.org/10.1007/978-3-642-15825-4_10.10.1007/978-3-642-15825-4_10 ↗
26. Coarse aggregate effectiveness in concrete: Quantitative models study on paste thickness, mortar thickness and compressive strength10.1016/j.conbuildmat.2021.123171 ↗
27. Evaluating the performance of self compacting concretes made with recycled coarse and fine aggregates using non destructive testing techniques10.1016/j.conbuildmat.2018.06.039 ↗
28. Sivashankar · 2024 · 69
29. Dropout: a simple way to prevent neural networks from overfitting
30. Szegedy, C., Wei Liu, Yangqing Jia, Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., Rabinovich, A., 2015. Going deeper with convolutions, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, Boston, MA, USA, 1–9. https://doi.org/10.1109/CVPR.2015.7298594.10.1109/CVPR.2015.7298594 ↗
31. Taylor, L., Nitschke, G., 2018. Improving deep learning with generic data augmentation, 2018 IEEE Symposium Series on Computational Intelligence (SSCI). Bangalore, India, 2018, 1542-1547. http://dx.doi.org/10.1109/SSCI.2018.8628742.10.1109/SSCI.2018.8628742 ↗
32. Evaluation of a sustainable self compacting concrete using destructive and non-destructive testing10.1016/j.matpr.2021.09.389 ↗
33. Deep learning for computer vision: a brief review
34. Improved vanishing gradient problem for deep Multi-layer Neural Networks
35. Orthogonal experimental design for compressive strength of recycled coarse aggregate concrete with silica fume-slag-fly ash hybrid micro-powders10.1016/j.conbuildmat.2023.133669 ↗
36. Evaluation of compressive strength of concrete durability degradation in dry and wet environments using destructive and non-destructive testing10.1016/j.measurement.2023.113702 ↗
37. Estimation of concrete strength using thermography integrated with deep-learning-based image segmentation: case studies and economic analysis10.1016/j.eswa.2022.119249 ↗
38. Predictive and experimental assessment of chloride ion permeation in concrete subjected to multi-factorial conditions using the XGBoost algorithm
39. Investigation on shear strength properties of water-bearing concrete-rock interface based on convolutional neural network recognition method10.1016/j.conbuildmat.2024.137349 ↗
40. Zhang, X., Zhou, X., Lin, M., Sun, J., 2018. ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, Salt Lake City, UT, 6848–6856. https://doi.org/10.1109/CVPR.2018.00716.10.1109/CVPR.2018.00716 ↗
41. Effect of initial water-curing period and curing condition on the properties of self-compacting concrete10.1016/j.matdes.2011.09.053 ↗