Article Details

Abstract

To address the limitations of existing algorithms in fine crack detection, this paper proposes a pavement crack detection method based on improved YOLOv11, aiming to improve the accuracy and real-time performance of road crack detection and provide strong support for road maintenance. By introducing the ContextGuidedDown module into the Backbone, the extraction of texture and contextual features is enhanced. By adding the hyper-MfM module to the Neck, the efficiency of multi-scale fusion and semantic representation are improved. By adopting the WIoU loss function, the inherent shape and scale of pavement cracks are focused on and the robustness of the model is enhanced. The experiments are conducted on the China subset of the RDD2022 dataset. Results show that the improved model outperforms the original one in metrics such as mAP, Precision and Recall while maintaining a favorable inference speed. The method effectively improves the accuracy and robustness of crack detection and achieves satisfactory detection performance.

Keywords

YOLOv11; crack detection; ContextGuidedDown; hyper-MfM; multi-scale feature fusion; WIoU Loss Function

References

1. [1] MA N，SONG Z，HU Q，et al. WP-CrackNet: A Collaborative Adversarial Learning Framework for End-to-End Weakly-Supervised Road Crack Detection [J/OL].Neurocomputing，2026，659：131845（2025-10-24）.https://www.sciencedirect.com/science/article/abs/pii/S0925231225025172.
2. [2] Tursun Maimaiti, Qiu Jianzhuo, Liu Jian, etc Road crack detection algorithm based on improved YOLOv8n [J]. Science, Technology and Engineering, 2025, 25 (14): 6044-6053
3. [3] Du Lei, Chen Xi, Bai Pengpeng, etc Asphalt pavement crack detection method based on improved YOLOv5 [J]. Science, Technology and Engineering, 2024, 24 (31): 13579-13591
4. [4] Yin Xuehui, Fu Linlin, Zhou Shangbo Concrete pavement crack detection network based on progressive contextual interaction and attention mechanism [J]. Computer Applications, 2025, 45 (10): 3353-3362
5. [5] Yuan Hongshuai, Li Qi, Wang Yueming Road crack detection based on YOLO-CD [J]. Science, Technology and Engineering, 2025, 25 (9): 3888-3895
6. [6] LIU W，DRAGOMIR ANGUELOV D，ERHAN D，et al. SSD: Single Shot MultiBox Detector [C]//Computer Vision–ECCV 2016. Amsterdam：Springer，2016：21-37.
7. [7] ZOU Q，ZHANG Z，LI Q，et al. DeepCrack: Learning Hierarchical Convolutional Features for Crack Detection [J].IEEE Transactions on Image Processing，2018，28（3）：1498-1512.
8. [8] HUYAN J，LI W，TIGHE S，et al. Detection of Sealed and Unsealed Cracks with Complex Backgrounds Using Deep Convolutional Neural Network [J].Automation in Construction，2019，107：102946（2019-10-05）.https://doi.org/10.1016/j.autcon.2019.102946.
9. [9] ZHANG C，CHANG C，JAMSHIDI M. Concrete Bridge Surface Damage Detection Using a Single-Stage Detector [J].Computer-Aided Civil and Infrastructure Engineering，2020，35（4）：389-409.
10. [10] XIANG X，ZHANG Y，EL SADDIK A. Pavement Crack Detection Network Based on Pyramid Structure and Attention Mechanism [J].IET Image Processing，2020，14（8）：1580-1586.

Funding

Jilin Province Science and Technology Development Plan Project (YDZJ202401527ZYTS)

About the authors

Yang Yi (2002-), female, Han ethnicity, from Bozhou, Anhui Province, is currently pursuing a master's degree. Her research focuses on intelligent perception and advanced control; Zhang Yuying (1993-), female, Han ethnicity, from Shulan, Jilin, lecturer, PhD, research direction: intelligent recognition algorithm research; Zhao Bin (2002-), male, Han ethnicity, from Suihua, Heilongjiang Province, currently pursuing a master's degree in detection technology and automation devices; Ji Yufang (2001-), female, Han ethnicity, from Weifang, Shandong Province, currently pursuing a master's degree in detection technology and automation devices; Xu Fei (2001-), female, Han ethnicity, from Tonghua, Jilin, is currently pursuing a master's degree in Intelligent Perception and Advanced Control.