YOLO-based Approach for Pavement Damage Detection

Xue Lin; Huanxin Zhou; He Zhang; Mingjian Liu; Jiaqi Li

doi:10.6919/ICJE.202410_10(10).0003

Authors

Xue Lin
Huanxin Zhou
He Zhang
Mingjian Liu
Jiaqi Li

DOI:

https://doi.org/10.6919/ICJE.202410_10(10).0003

Keywords:

Pavement; Damage Detection; Object Detection; YOLO; Computer Vision.

Abstract

As the important role of highway in modern social and economic development becomes increasingly prominent, maintaining the good condition of road surface is crucial to ensure traffic safety and extend the service life of road. Although the traditional pavement damage detection method is effective, there are limitations in convenience and practicability, especially in the allocation of personnel and cost control. In view of this, a new pavement damage detection method based on computer vision is proposed in this paper. This study uses the latest version of YOLOv8, an object detection framework, and a homemade datasets that covers different types of pavement damage, including lateral cracks, longitudinal cracks, potholes, cracks, and damage around manhole covers. The datasets collected 5,965 images from an iPhone 7 Plus smartphone and shot under different lighting and backgrounds to enhance the model's generalization. The experimental results show that the mean Average Precision of the model reaches 0.98 in the task of target detection using YOLO v8, which has satisfactory detection effect. In addition, the study also uses YOLOv8 to carry out a case segmentation task, which can not only accurately locate the damage area, but also provide pixel-level segmentation results, which is helpful to evaluate the specific scope and severity of damage. The mean Average Precision of the instance segmentation model reaches 0.604. In conclusion, the method proposed in this study not only improves the efficiency and accuracy of pavement damage detection, but also provides strong support for subsequent road maintenance.

Downloads

Download data is not yet available.

References

[1] A. Haghighattalab, A. Mohammadzadeh, M. J. V. Zoej, et al.: Post-earthquake road damage assessment using region-based algorithms from high-resolution satellite images, Proc. Image and Signal Processing for Remote Sensing XVI (SPIE, 2010), Vol. 7830, p.439-447. DOI:10.1117/12.864538.

[2] X. Zhang, Y. Chen, M. Jia, et al.: The study of road damage detection based on high-resolution SAR image, Proc. 2013 IEEE International Geoscience and Remote Sensing Symposium - IGARSS (2013), p.2633-2636. DOI:10.1109/IGARSS.2013.6723363.

[3] J.Wang,Q.Qin, J.Zhao, et al.:A knowledge-based method for road damage detection using high-resolution remote sensing image, Proc. 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS)(2015),p.3564-3567.DOI:10.1109/IGARSS.2015.7326591.

[4] A. Benedetto, F. Benedetto, M. R. D. Blasiis, et al.: Reliability of Radar Inspection for Detection of Pavement Damage, Road Materials and Pavement Design (2004).DOI:10.1080/14680629.2004.9689964

[5] T. Yamada, T. Ito, A. Ohya: Detection of road surface damage using mobile robot equipped with 2D laser scanner, Proc. 2013 IEEE/SICE International Symposium on System Integration (2013), p.250-256. DOI:10.1109/SII.2013.6776679.

[6] D. H. Kil, f. b. Shin: Automatic road-distress classification and identification using a combination of hierarchical classifiers and expert systems-subimage and object processing, Proc. International Conference on Image Processing(1997), Vol. 2, p.414-417. DOI:10.1109/ICIP.1997.638795.

[7] C. Yanli: Algorithm Study on Thinning and Keeping Connectivity of Bituminous Pavement Crack Images, Proc. 2010 International Conference of Information Science and Management Engineering (2010), Vol. 1, p.550-553. DOI:10.1109/ISME.2010.265.

[8] W. Huang, N.Zhang: A novel road crack detection and identification method using digital image processing techniques, Proc. 2012 7th International Conference on Computing and Convergence Technology (ICCCT) (2012), P.397-400. DOI:10.1109/ICCCT.2012.6530365.

[9] Y. Jo, S. Ryu: Pothole Detection System Using a Black-box Camera, Vol. 15 (2015) No.11, p.29316-29331. DOI:10.3390/s151129316.

[10] L. Zhang, F. Yang, Y. Daniel Zhang, et al.: Road crack detection using deep convolutional neural network, Proc. 2016 IEEE International Conference on Image Processing (ICIP) (2016), p.3708-3712. DOI:10.11 09/ICIP.2016.7533052.

[11] A. Zhang, K. C. P. Wang, Y. Fei, et al: Deep Learning-Based Fully Automated Pavement Crack Detection on 3D Asphalt Surfaces with an Improved CrackNet, Journal of Computing in Civil Engineering, Vol. 32 (2018) No.5, p.04018041.DOI:10.1061/(ASCE)CP.1943-5487.0000775.

[12] H. Maeda, Y. Sekimoto, T. Seto, et al.: Road Damage Detection and Classification Using Deep Neural Networks with Smartphone Images: Road damage detection and classification, Computer-Aided Civil and Infrastructure Engineering, Vol. 33 (2018). DOI:10.1111/mice.12387.

[13] B. Li, K. Wang, A. Zhang, et al.: Automatic classification of pavement crack using deep convolutional neural network,International Journal of Pavement Engineering, Vol. 21 (2018), p.1-7. DOI:10.1080/ 10298436.2018.1485917.

[14] L. Li, B. Fang, J. Zhu: Performance Analysis of the YOLOv4 Algorithm for Pavement Damage Detection with Different Embedding Positions of CBAM Modules, Applied Sciences, Vol. 12 (2022), p.10180 DOI:10.3390/app121910180.

[15] Y. Jiang, D. Pang, C. Li, et al.: Two-step deep learning approach for pavement crack damage detection and segmentation, International Journal of Pavement Engineering, Vol. 24 (2022), p.1-14. DOI:10.1080/ 10298436.2022.2065488.

[16] Y. Ni, J. Mao, Y. Fu, et al.:Damage Detection and Localization of Bridge Deck Pavement Based on Deep Learning, Sensors, Vol. 23 (2023), p.5138. DOI:10.3390/s23115138.

[17] Z. Li, X. Xiao, J. Xie, et al.: Cycle-YOLO: A Efficient and Robust Framework for Damage Detection, 2024.

[18] Y. LeCun, L. Bottou, Y. Bengio, et al.:Gradient-Based Learning Applied to Document Recognition, Proceedings of the IEEE, Vol. 86 (1998), p.2278-2324. DOI:10.1109/5.726791.

[19] Y. LeCun, L. Bottou, Y. Bengio, G. Hinton: Deep learning, Nature, Vol. 521 (2015) No.7553, p436-444. DOI:10.1038/nature14539.

[20] G. Jocher, A. Chaurasia, J. Qiu: Ultralytics YOLO, (2023-01). https://github.com/ultralytics/ultralytics.