Investigation into enabling machine vision and machine learning technologies for surface defect detection of pit support systems

Abstract

Cracks and water seepage are common structural safety hazards in excavation and pit support system. Traditional methods usually rely on a lot of manpower and material resources, and there are some problems in the monitoring process such as low efficiency, long time, incomplete data collection and insufficient accuracy, which cannot meet the needs of modern engineering construction. In recent years, the construction industry has gradually changed to the trend of intelligence and automation, and machine vision has entered the field of vision. It can not only effectively reduce labor costs, but also improve the overall accuracy of monitoring. However, previous machine learning framework usually uses a two-stage monitoring method, which takes a long time including the collection and process of data separately. This paper focuses on pit support systems and provides an overview and comparison of the application of machine vision and machine learning technologies. Furthermore, a real-time defect detection method based on the improved YOLOv8 algorithm, which can process the collected crack data and water seepage pictures, give the physical characteristics of the crack, and mark the location of water seepage, has been proposed and verified. Additionally, a practical project in Huzhou serves as a case study, where the established method has been applied. The actual implementation shows that the model also has good robustness under complex foundation pit conditions.

Keywords

Excavation and pit support / Defect detection / Machine vision / Machine learning / YOLOv8

References

[1] Ali ML, Zhang Z. The YOLO framework: A comprehensive review of evolution, applications, and benchmarks in object detection. Computers, 2024, 13(12): 336.

[2] Azarang A, Manoochehri H, Kehtarnavaz N. Convolutional autoencoder-based multispectral image fusion. IEEE Access, 2019, 7: 35673-35683.

[3] Cha YJ, Choi W, Büyüköztürk O. Deep learning-based crack damage detection using convolutional neural networks. Computer-Aided Civil and Infrastructure Engineering, 2017, 32(5): 361-378.

[4] Cha YJ, Choi W, Suh G, Mahmoudkhani S, Büyükztürk O. Autonomous structural visual inspection using region-based deep learning for detecting multiple damage types. Computer-Aided Civil and Infrastructure Engineering, 2018, 33(9): 731-747.

[5] Chen B, Wu ZR, Liang JC, Dou YH. Time-varying identification model for crack monitoring data from concrete dams based on support vector regression and the bayesian framework. Mathematical Problems in Engineering, 2017, 2017(1): 5450297.

[6] Choi W, Cha YJ. SDDnet: Real-time crack segmentation. IEEE Transactions on Industrial Electronics, 2020, 67(9): 8016-8025.

[7] Dai Y, Li C, Su X, Liu H, Li J. Multi-scale depthwise separable convolution for semantic segmentation in street–road scenes. Remote Sensing, 2023, 15(10): 2649.

[8] Deng Y, Deng YK. A method of sar image automatic target recognition based on convolution auto-encode and support vector machine. Remote Sensing, 2022, 14(21): 5559.

[9] Fang ZQ, Jia T, Chen QS, Xu M, Yuan X, Wu CD. Laser stripe image denoising using convolutional autoencoder. Results in Physics, 2018, 11: 96-104.

[10] Farahat L, Azar ER. Content annotation in images from outdoor construction jobsites using YOLO V8 and Swin transformer. Smart Construction and Sustainable Cities, 2024, 2(1): 10.

[11] Gao C, Zhang Q, Tan Z. Applying optimized YOLOv8 for heritage conservation: Enhanced object detection in Jiangnan traditional private gardens. Heritage Science, 2024, 12(1): 1-20.

[12] GB 50204-2015. Code for quality acceptance of concrete structure construction, 2014China Construction Industry Press

[13] Guo XF, Liu XW, Zhu E, Yin JP, Liu DRLiu D, Xie S, Li Y, Zhao D, El-Alfy ES. Deep clustering with convolutional autoencoders. Neural information processing, 2017Springer373-382.

[14] He SX, Hu QZ, Zhuang XS. Influence of seepage on ground settlement of deep foundation pit. Chinese Journal of Rock Mechanics and Engineering, 2003, 2003(3): 1551-1554(in Chinese)

[15] Hou Y, Chen YH, Gu XY, Cao DD, Mao Q, Wang LB. Automatic identification of pavement objects and cracks using the convolutional auto-encoder. China Journal of Highway and Transport, 2020, 33(10): 288-303(in Chinese)

[16] Hu S, Zhou T, Zhong Y. Influence law of foundation pit excavation on stress of surrounding tunnel bolt. Applied Sciences, 2022, 12(22): 11479.

[17] Huang YT, Pai P. Using the least squares support vector regression to forecast movie sales with data from twitter and movie databases. Symmetry, 2020, 12(4): 625.

[18] Ilya, L., & Frank, H. (2017). SGDR: Stochastic Gradient Descent with Warm Restarts. arXiv:1608.03983.

[19] Iqbal SM, Park J, Jung K, Lee JS, Kang D. Long-term prediction of crack growth using deep recurrent neural networks and nonlinear regression: A comparison study. Applied Sciences, 2022, 12(20): 10514.

[20] Kang D, Benipal SS, Gopal DL, Cha YJ. Hybrid pixel-level concrete crack segmentation and quantification across complex backgrounds using deep learning. Automation in Construction, 2020, 118: 103291.

[21] Kim K, Kim K, Jeong S. Application of YOLO v5 and v8 for recognition of safety risk factors at construction sites. Sustainability, 2023, 15(20): 15179.

[22] Lai J, Wang XD, Xiang Q, Song YF, Quan W. Review on autoencoder and its application. Journal on Communications, 2021, 42(9): 218-230(in Chinese)

[23] Lan YH, Lv YJ, Xu JS, Zhang YQ, Zhang YH. Breast mass lesion area detection method based on an improved YOLOv8 model. Electronic Research Archive, 2024, 32(10): 5846-5867.

[24] Lee JU, Kim HS, Kim NY, Ryu EM, Kang JW. Learning to detect cracks on damaged concrete surfaces using two-branched convolutional neural network. Sensors, 2019, 19(21): 4796.

[25] Li YD, Li HG, Wang HR. Pixel-wise crack detection using deep local pattern predictor for robot application. Sensors, 2018, 18(9): 3042.

[26] Liang HY, Shen SL, Zhou AN, Zhao WW. Automated detection and quantification of leakage areas in shield tunnel linings using laser scanning data and deep learning network. Engineering Applications of Artificial Intelligence, 2025, 160: 111930.

[27] Liang MH, Liu RW, Li SC, Xiao Z, Liu X, Lu F. An unsupervised learning method with convolutional auto-encoder for vessel trajectory similarity computation. Ocean Engineering, 2021, 225: 108803.

[28] Liu N, Han JW, Yang MH. PiCANet: Learning pixel-wise contextual attention for saliency detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, 2018: 3089-3098

[29] Lv ZH, Cheng C, Lv HB. Automatic identification of pavement cracks in public roads using an optimized deep convolutional neural network model. Philosophical Transactions of the Royal Society a: Mathematical, Physical and Engineering Sciences, 2023, 381(2254): 20220169.

[30] Mohammed Z, Abeer A. Recognizing human activities with the use of convolutional block attention module. Egyptian Informatics Journal, 2024, 27: 100536.

[31] Qiu J, Li F. The application of foundation pit monitoring technology to the excavation. MATEC Web of Conferences, 2015, 22: 04004.

[32] Sapkota R, Ahmed D, Karkee M. Comparing YOLOv8 and Mask R-CNN for instance segmentation in complex orchard environments. Artificial Intelligence in Agriculture, 2024, 13: 84-99.

[33] Sun YN, Xue B, Zhang MJ, Yen G. A particle swarm optimization-based flexible convolutional autoencoder for image classification. IEEE Transactions on Neural Networks and Learning Systems, 2019, 30(8): 2295-2309.

[34] Tan JT, Huang LY, Chen ZH, Qu RK, Li CL. DarkSegNet: Low-light semantic segmentation network based on image pyramid. Signal Processing: Image Communication, 2025, 135: 117265

[35] Tong Z, Gao J, Yuan DD. Advances of deep learning applications in ground-penetrating radar: A survey. Construction and Building Materials, 2020, 258: 120371.

[36] Wang F, Chen WL, Fakieh B, Ali BJA. Stock price analysis based on the research of multiple linear regression macroeconomic variables. Applied Mathematics and Nonlinear Sciences, 2022, 7(1): 267-274.

[37] Wang HY, Han X, Song XF, Su J, Li Y, Zheng WY, Wu XJ. Research on automatic pavement crack identification based on improved YOLOv8. International Journal on Interactive Design and Manufacturing (IJIDeM), 2024, 18(6): 3773-3783.

[38] Wang Z, Shen X, Tian X. Study on the influence and deformation control of rich water foundation excavation on adjacent buildings. Buildings, 2025, 15(1): 52.

[39] Wu H, Fu Q. An innovative deep learning approach to spinal fracture detection in CT images. Annali Italiani di Chirurgia, 2024, 95(4): 657-668.

[40] Wu ZY, Jia DW, Wang QA. Building crack identification based on convolutional neural network and regional growth method. Journal of Basic Science and Engineering, 2022, 30(2): 317-327(in Chinese)

[41] Xiao G, Hou S, Zhou H. PCB defect detection algorithm based on CDI-YOLO. Scientific Reports, 2024, 14(1): 7351.

[42] Xie JY, Hou Q, Cao JW. Image clustering algorithms by deep convolutional autoencoders. Journal of Frontiers of Computer Science and Technology, 2019, 13(4): 586-595(in Chinese)

[43] Xie SN, Tu ZW. Holistically-nested edge detection. IEEE International Conference on Computer Vision (ICCV), 2015, 2015: 1395-1403

[44] Xu Y, Cai Y, Li D. Building crack monitoring based on digital image processing. Fracture and Structural Integrity, 2020, 14(52): 1-8.

[45] Yan T, Shen SL, Zhou AN. Data augmentation-assisted muck image recognition during shield tunnelling. Underground Space, 2025, 21: 370-383.

[46] Zhao HY, Jin J, Liu Y, Guo YN, Shen Y. FSDF: A high-performance fire detection framework. Expert Systems with Applications, 2024, 238: 121665.

[47] Zhao WW, Shen SL, Yan T, Zhou AN. Intelligent approach for mucky soil identification during shield tunnelling by enhanced YOLO model. Journal of Rock Mechanics and Geotechnical Engineering, 2025, 17(6): 3327-3338.

[48] Zheng G, Zhu HH, Liu XR, Yang GH. Control of safety of deep excavations and underground engineering and its impact on surrounding environment. China Civil Engineering Journal, 2016, 49(6): 1-24

[49] Zheng ZY, Zhang SZ, Song HL, Yan Q. Deep clustering using 3D attention convolutional autoencoder for hyperspectral image analysis. Scientific Reports, 2024, 14(1): 4209.

[50] Zhou FY, Jin LP, Dong J. Review of convolutional neural network. Chinese Journal of Computers, 2017, 40(6): 1229-1251. DOI: (in Chinese)

[51] Zhou Y, Ali R, Mokhtar N, Harun SW, Iwahashi M. MixSegNet: A novel crack segmentation network combining CNN and Transformer. IEEE Access, 2024, 12: 111535-111545.

[52] Zhu GJ, Shen SL, Yao JJ, Wang MH, Zhuang JF, Fan Z. Automatic lightweight networks for real-time road crack detection with DPSO. Advanced Engineering Informatics, 2025, 68: 103610.

[53] Zhu YJ, Sekiya H, Okatani T, Tai M, Morichika S. B-CNN: A deep learning method for accelerometer-based fatigue cracks monitoring system. Journal of Civil Structural Health Monitoring, 2023, 13: 947-959.

Funding

Shanghai Municipal College Students' Innovation and Entrepreneurship Project(S202410247429)

Research project supported by Huzhou Transportation Bureau(No. 202408)

RIGHTS & PERMISSIONS

The Author(s)