Crack is an important indicator for evaluating the damage level of concrete structures. However, traditional crack detection algorithms have complex implementation and weak generalization. The existing crack detection algorithms based on deep learning are mostly window-level algorithms with low pixel precision. In this article, the CrackUnet model based on deep learning is proposed to solve the above problems. First, crack images collected from the lab, earthquake sites, and the Internet are resized, labeled manually, and augmented to make a dataset (1200 subimages with 256 × 256 × 3 resolutions in total). Then, an improved Unet-based method called CrackUnet is proposed for automated pixel-level crack detection. A new loss function named generalized dice loss is adopted to detect cracks more accurately. How the size of the dataset and the depth of the model affect the training time, detecting accuracy, and speed is researched. The proposed methods are evaluated on the test dataset and a previously published dataset. The highest results can reach 91.45%, 88.67%, and 90.04% on test dataset and 98.72%, 92.84%, and 95.44% on CrackForest Dataset for precision, recall, and F1 score, respectively. By comparing the detecting accuracy, the training time, and the information of datasets, CrackUnet model outperform than other methods. Furthermore, six images with complicated noise are used to investigate the robustness and generalization of CrackUnet models.
XiaQChengYYZhangJ, et al. In-service condition assessment of a long-span suspension bridge using temperature-induced strain data. J Bridge Eng2017; 22(3): 04016124.
2.
AiedHGonzálezACanteroD.Identification of sudden stiffness changes in the acceleration response of a bridge to moving loads using ensemble empirical mode decomposition. Mech Syst Signal Pr2016; 66–67: 314–338.
3.
KunwarAJhaRWhelanM, et al. Damage detection in an experimental bridge model using Hilbert–Huang transform of transient vibrations. Struct Control Health Monit2013; 20(1): 1–15.
4.
ChenagXJiXHenryRS, et al. Coupled axial tension-flexure behavior of slender reinforced concrete walls. Eng Struct2019; 188: 261–276.
5.
KimHAhnEShinM, et al. Crack and noncrack classification from concrete surface images using machine learning. Struct Health Monit2019; 18(3): 725–738.
6.
LeiBRenYWangN, et al. Design of a new low-cost unmanned aerial vehicle and vision-based concrete crack inspection method. Struct Health Monit. Epub ahead of print3February2020. DOI: 10.1177/1475921719898862.
7.
ZhangAWangKCPLiB, et al. Automated pixel-level pavement crack detection on 3D asphalt surfaces using a deep-learning network. Comput Aid Civ Infrastruct Eng2017; 32(10): 805–819.
8.
ShiYCuiLQiZ, et al. Automatic road crack detection using random structured forests. IEEE Trans Intell Transp Syst2016; 17: 3434–3445.
9.
ZhangLYangFZhangYD, et al. Road crack detection using deep convolutional neural network. In: Proceedings of the IEEE international conference on image processing (ICIP 2016), Phoenix, AZ, 25–28 September 2016, pp. 3708–3712. New York: IEEE.
10.
ChaYJChoiWSuhG, et al. Autonomous structural visual inspection using region-based deep learning for detecting multiple damage types. Comput Aid Civ Infrastruct Eng2018; 33(9): 731–747.
11.
XueYLiY.A fast detection method via region-based fully convolutional neural networks for shield tunnel lining defects. Comput Aid Civ Infrastruct Eng2018; 33(8): 638–654.
12.
Abdel-QaderIAbudayyehOKellyME.Analysis of edge-detection techniques for crack identification in bridges. J Comput Civ Eng2003; 17(4): 255–263.
13.
FujitaYMitaniYHamamotoY.A method for crack detection on a concrete structure. In: Proceedings of the IEEE 18th international conference on pattern recognition (ICPR 2006), Hong Kong, China, 20–24 August 2006, vol. 3, pp. 901–904. New York: IEEE.
14.
KimHAhnEChoS, et al. Comparative analysis of image binarization methods for crack identification in concrete structures. Cem Concrete Res2017; 99: 53–61.
15.
JahanshahiMRMasriSFPadgettCW, et al. An innovative methodology for detection and quantification of cracks through incorporation of depth perception. Mach Vis Appl2013; 24(2): 227–241.
LiGZhaoXDuK, et al. Recognition and evaluation of bridge cracks with modified active contour model and greedy search-based support vector machine. Automat Constr2017; 78: 51–61.
18.
LecunYBottouLBengioY, et al. Gradient-based learning applied to document recognition. Proc IEEE1998; 86(11): 2278–2324.
19.
ChaYJChoiWBuyukozturkO.Deep learning-based crack damage detection using convolutional neural networks. Comput Aid Civ Infrastruct Eng2017; 32(5): 361–378.
20.
XuYLiSZhangD, et al. Identification framework for cracks on a steel structure surface by a restricted Boltzmann machines algorithm based on consumer-grade camera images. Struct Control Health Monit2018; 25(2): e2075.
21.
XuYBaoYChenJ, et al. Surface fatigue crack identification in steel box girder of bridges by a deep fusion convolutional neural network based on consumer-grade camera images. Struct Health Monit2019; 18(3): 653–674.
22.
ChenFCJahanshamiMR.NB-CNN: deep learning-based crack detection using convolutional neural network and naïve Bayes data fusion. IEEE Trans Ind Electron2018; 65(5): 4392–4400.
23.
Mohtasham KhaniMVahidniaSGhasemzadehL, et al. Deep-learning-based crack detection with applications for the structural health monitoring of gas turbines. Struct Health Monit. Epub ahead of print11November2019. DOI: 10.1177/1475921719883202.
24.
MaedaHSekimotoYSetoT, et al. Road damage detection and classification using deep neural networks with smartphone images. Comput Aid Civ Infrastruct Eng2018; 33(12): 1127–1141.
25.
NiFTZhangJChenZQ.Zernike-moment measurement of thin-crack width in images enabled by dual-scale deep learning. Comput Aid Civ Infrastruct Eng2019; 34(5): 367–384.
26.
YangXLiHYuY, et al. Automatic pixel-level crack detection and measurement using fully convolutional network. Comput Aid Civ Infrastruct Eng2018; 33(12): 1090–1109.
27.
BangSParkSKimH, et al. Encoder–decoder network for pixel-level road crack detection in black-box images. Comput Aid Civ Infrastruct Eng2019; 34(8): 713–727.
28.
ZhangAWangKCPFeiY, et al. Automated pixel-level pavement crack detection on 3D asphalt surfaces with a recurrent neural network. Comput Aid Civ Infrastruct Eng2019; 34(3): 213–229.
29.
MeiQGülM.Multi-level feature fusion in densely connected deep-learning architecture and depth-first search for crack segmentation on images collected with smartphones. Struct Health Monit. Epub ahead of print3January2020. DOI: 10.1177/1475921719896813.
30.
FeiYWangKCPZhangA, et al. Pixel-level cracking detection on 3D asphalt pavement images through deep-learning-based CrackNet-V. IEEE Trans Intell Transp Syst2020; 21: 273–284.
31.
RonnebergerOFischerPBroxT.U-Net: convolutional networks for biomedical image segmentation. In: Proceedings of the international conference on medical image computing and computer-assisted intervention, Munich, 5–9 October 2015, pp. 234–241. Cham: Springer.
SimonyanKZissermanA.Very deep convolutional networks for large-scale image recognition, 2014, http://arxiv.org/abs/1409.1556
36.
SchererDMullerABehnkeS.Evaluation of pooling operations in convolutional architectures for object recognition. In: Proceedings of the 20th international conference on artificial neural networks (ICANN 2010), Thessaloniki, 15–18 September 2010, pp. 92–101. Berlin: Springer.
37.
NairVHintonGE.Rectified linear units improve restricted Boltzmann machines. In: Proceedings of the 27th international conference on machine learning (ICML 2010), Haifa, Israel, 21–24 June 2010, pp. 807–814. New York: ACM.
38.
IoffeSSzegedyC.Batch normalization: accelerating deep network training by reducing internal covariate shift. In: Proceedings of the international conference on machine learning (ICML 2015), Lille, 6–11 July 2015, pp. 448–456. New York: ACM.
39.
SrivastavaNHintonGKrizhevskyA, et al. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res2014; 15(1): 1929–1958.
40.
SudreCHLiWVercauterenT, et al. Generalized dice overlap as a deep learning loss function for highly unbalanced segmentations. In: CardosoMJArbelTCarneiroG, et al. (eds) Deep learning in medical image analysis and multimodal learning for clinical decision support. Cham: Springer, 2017, pp. 240–248.
41.
BengioY.Practical recommendations for gradient-based training of deep architectures. In: MontavonGOrrGBMüllerKR (eds) Neural networks: tricks of the trade. 2nd ed.Berlin: Springer, 2012, pp. 437–478.
42.
SutskeverIMartensJDahlG, et al. On the importance of initialization and momentum in deep learning. In: Proceedings of the 30th international conference on machine learning (ICML 2013), Atlanta, GA, 16–21 June 2013, pp. 1139–1147. New York: ACM.
WilsonDRMartinezTR. The need for small learning rates on large problems. In: Proceedings of the international joint conference on neural networks (IJCNN 2001), Cat. No. 01CH37222, Washington, DC, 15–19 July 2001, vol. 1, pp. 115–119. New York: IEEE.
46.
NguyenTSBegotSDucultyF, et al. Free-form anisotropy: a new method for crack detection on pavement surface images. In: Proceedings of the 18th IEEE international conference on image processing (ICIP 2011), Brussels, 11–14 September 2011, pp. 1069–1072. New York: IEEE.
47.
ZouQCaoYLiQ, et al. CrackTree: automatic crack detection from pavement images. Pattern Recogn Lett2012; 33: 227–238.
48.
LiHSongDLiuY, et al. Automatic pavement crack detection by multi-scale image fusion. IEEE Trans Intell Transp Syst2019; 20: 2025–2036.
