Lightweight Concrete Crack Detection Method Based on Multi-Dimensional Attention Module

doi:10.19678/j.issn.1000-3428.0068622

Abstract

Abstract:

Most current concrete crack detection models are too large to be deployed in mobile devices, and the crack detection is inaccurate, with cracks being missed; to solve these problems, a lightweight concrete crack detection method based on a multi-dimensional attention module is proposed. In this method, because most current mainstream crack detection models are large, depth-separable convolution is used to reconstruct the Conv-BN-SiLU (CBS) feature extraction module in YOLOv5s to obtain a Lightweight (LCBS) module. To solve the problem of inaccurate crack detection, a Multi-Scale Feature (MSF) module is proposed to replace the convolution layer of the first layer of YOLOv5s to enhance the ability of the network to extract the features of cracks of different sizes. To address the problem of missed cracks, a Multi-Dimensional Attention (MDA) module, which fuses spatial and channel information, is proposed to enhance the ability of crack feature extraction and reduce the number of missed cracks. Experiments show that, compared with YOLOv5s, the proposed method reduces the number of parameters by 35.2%, computation amount by 50.9%, and model size by 32.8% and increases the average accuracy (mAP@0.5) by 4.2 percentage points. Compared with other mainstream target detection methods of the same type, the proposed method has fewer parameters and higher detection accuracy.

Key words: crack detection, attention module, lightweight, YOLOv5s model, object detection

摘要：

为解决当前混凝土裂缝检测模型庞大难以部署到移动端设备且裂缝检测不准及漏检问题, 提出一种基于多维注意力模块的轻量化混凝土裂缝检测方法。该方法针对当前主流的裂缝检测模型庞大的问题, 采用深度可分离卷积对YOLOv5s中的CBS特征提取模块进行重构, 得到轻量化CBS(LCBS)特征提取模块, 以减少网络的参数量及计算量; 针对裂缝检测不准的问题, 提出一种多尺度特征(MSF)提取模块用于替换YOLOv5s第1层的卷积层, 以增强网络对不同尺寸裂缝特征的提取能力; 针对裂缝漏检问题, 提出融合空间及通道信息的多维注意力(MDA)模块, 以增强裂缝特征提取能力和减少裂缝漏检。实验结果表明, 所提方法比YOLOv5s参数量减少了35.2%, 计算量减少了50.9%, 模型规模减小了32.8%, 且平均精度均值(mAP@0.5)提高了4.2百分点, 与目前主流的同类目标检测方法相比, 具有较低的参数量和较高的检测精度。

关键词: 裂缝检测, 注意力模块, 轻量化, YOLOv5s模型, 目标检测

XU Huajie, ZHENG Liwen, ZHANG Pin, QIN Yuanzhuo. Lightweight Concrete Crack Detection Method Based on Multi-Dimensional Attention Module[J]. Computer Engineering, 2025, 51(5): 351-360.

许华杰, 郑力文, 张品, 秦远卓. 基于多维注意力模块的轻量化混凝土裂缝检测方法[J]. 计算机工程, 2025, 51(5): 351-360.

Save 0 / Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0068622

https://www.ecice06.com/EN/Y2025/V51/I5/351

Figures/Tables 12

Fig.1 LMMDA structure

Fig.2 Structure of main modules in LMMDA

Fig.3 MSF module structure

Fig.4 MDA module structure

Fig.5 Examples of the original image and ground truth

Fig.6 LMMDA model training loss function curves

Fig.7 The corresponding detection images of different attention modules and the effect diagrams superimposed with heat map

Fig.8 Comparison of the detection effect between different models

References 25

1	WANG Y Y, SONG K C, LIU J, et al. RENet: rectangular convolution pyramid and edge enhancement network for salient object detection of pavement cracks. Measurement, 2021, 170, 108698. doi: 10.1016/j.measurement.2020.108698
2	LI B, WANG K C, ZHANG A, et al. Automatic classification of pavement crack using deep convolutional neural network. International Journal of Pavement Engineering, 2020, 21(4): 457- 463. doi: 10.1080/10298436.2018.1485917
3	郝巨鸣, 杨景玉, 韩淑梅, 等. 引入Ghost模块和ECA的YOLOv4公路路面裂缝检测方法. 计算机应用, 2023, 43(4): 1284- 1290.
	HAO J M, YANG J Y, HAN S M, et al. YOLOv4 highway pavement crack detection method using Ghost module and ECA. Journal of Computer Applications, 2023, 43(4): 1284- 1290.
4	DU Y C, ZHONG S, FANG H Y, et al. Modeling automatic pavement crack object detection and pixel-level segmentation. Automation in Construction, 2023, 150, 104840. doi: 10.1016/j.autcon.2023.104840
5	SHOLEVAR N, GOLROO A, ESFAHANI S R. Machine learning techniques for pavement condition evaluation. Automation in Construction, 2022, 136, 104190. doi: 10.1016/j.autcon.2022.104190
6	ZHU X K, LYU S C, WANG X, et al. TPH-YOLOv5: improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios[C]//Proceedings of IEEE/CVF International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2021: 2778-2788.
7	李鹏程, 孙立双, 谢志伟, 等. 基于改进MobileNet-SSD的路面裂缝图像检测算法. 激光杂志, 2022, 43(7): 123- 127.
	LI P C, SUN L S, XIE Z W, et al. Improved pavement crack image detection algorithm based on MobileNet-SSD. Laser Journal, 2022, 43(7): 123- 127.
8	WANG C Y, BOCHKOVSKIY A, LIAO H M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[EB/OL]. [2023-09-10]. https://arxiv.org/abs/2207.02696v1.
9	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of IEEE International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2017: 2980-2988.
10	TAN M X, PANG R M, LE Q V. EfficientDet: scalable and efficient object detection[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2020: 10781-10790.
11	CHEN Z H, YANG C, LI Q F, et al. Disentangle your dense object detector[C]//Proceedings of the 29th ACM International Conference on Multimedia. New York, USA: ACM Press, 2021: 4939-4948.
12	REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137- 1149. doi: 10.1109/TPAMI.2016.2577031
13	MA D, FANG H Y, WANG N N, et al. Automatic detection and counting system for pavement cracks based on PCGAN and YOLO-MF. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(11): 22166- 22178. doi: 10.1109/TITS.2022.3161960
14	ALI R, CHUAH J H, ABU TALIP M S, et al. Structural crack detection using deep convolutional neural networks. Automation in Construction, 2022, 133, 103989. doi: 10.1016/j.autcon.2021.103989
15	YAO H, LIU Y, LI X, et al. A detection method for pavement cracks combining object detection and attention mechanism. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(11): 22179- 22189. doi: 10.1109/TITS.2022.3177210
16	廖延娜, 李婉. 基于卷积神经网络的桥梁裂缝检测方法. 计算机工程与设计, 2021, 42(8): 2366- 2372.
	LIAO Y N, LI W. Bridge crack detection method based on convolution neural network. Computer Engineering and Design, 2021, 42(8): 2366- 2372.
17	LI R X, YU J Y, LI F, et al. Automatic bridge crack detection using Unmanned aerial vehicle and Faster R-CNN. Construction and Building Materials, 2023, 362, 129659. doi: 10.1016/j.conbuildmat.2022.129659
18	袁磊, 唐海, 陈彦蓉, 等. 改进YOLOv5的复杂环境道路目标检测方法. 计算机工程与应用, 2023, 59(16): 212- 222.
	YUAN L, TANG H, CHEN Y R, et al. Improved YOLOv5 for road target detection in complex environments. Computer Engineering and Applications, 2023, 59(16): 212- 222.
19	HOU Y, LIU S, CAO D, et al. A deep learning method for pavement crack identification based on limited field images. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(11): 22156- 22165. doi: 10.1109/TITS.2022.3160524
20	WANG Q L, WU B G, ZHU P F, et al. ECA-net: efficient channel attention for deep convolutional neural networks[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2020: 358-366.
21	HOU Q B, ZHOU D Q, FENG J S. Coordinate attention for efficient mobile network design[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2021: 13713-13722.
22	HUANG J R, SHEN Q, WANG M, et al. Multiple attention Siamese network for high-resolution image change detection. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60, 3127580.
23	谌婷婷, 魏怡. 基于改进YOLOv4的混凝土裂缝检测方法. 激光杂志, 2024, 45(1): 80- 85.
	CHEN T T, WEI Y. Concrete crack detection method based on improved YOLOv4. Laser Journal, 2024, 45(1): 80- 85.
24	WANG J K, HE X H, SHAO F M, et al. A real-time bridge crack detection method based on an improved inception-resnet-v2 structure. IEEE Access, 2021, 9, 93209- 93223. doi: 10.1109/ACCESS.2021.3093210
25	YU Z W, SHEN Y G, SHEN C K. A real-time detection approach for bridge cracks based on YOLOv4-FPM. Automation in Construction, 2021, 122, 103514. doi: 10.1016/j.autcon.2020.103514

[1]	CHEN Ziyan, WANG Xiaolong, HE Di, AN Guocheng. Lightweight Vehicle Detection Network Based on Improved YOLOv8 [J]. Computer Engineering, 2025, 51(5): 314-325.
[2]	ZHOU Siyu, XU Huiying, ZHU Xinzhong, HUANG Xiao, SHENG Ke, CAO Yuqi, CHEN Chen. Mobile Phone Screen Defect Detection Algorithm Based on Improved YOLOv8n: PGS-YOLO [J]. Computer Engineering, 2025, 51(5): 326-339.
[3]	WANG Xiaolong, JIANG Bo, LUO Runshu, AN Guocheng. Congestion Detection Algorithm of Highway Toll Station Based on Multi-Information Fusion [J]. Computer Engineering, 2025, 51(5): 377-386.
[4]	HUANG Kun, QI Zhaojian, WANG Juanmin, HU Qian, HU Weichao, PI Jianyong. Aggregation Pedestrian Detection Model Based on Improved YOLOv8 [J]. Computer Engineering, 2025, 51(5): 133-142.
[5]	YANG Ping, ZHANG Xi. Improved DeepLabv3+ Road Surface Crack Detection Method [J]. Computer Engineering, 2025, 51(4): 261-270.
[6]	WANG Zeyu, XU Huiying, ZHU Xinzhong, HUANG Xiao, LIANG Jiajie, LI Chen. Lightweight Fry Detection Algorithm Based on Improved YOLOv8: FD-YOLO [J]. Computer Engineering, 2025, 51(4): 327-338.
[7]	LI Shuwei, HUANG Zhengxiang, HU Yun, LIU Xing, LU Xiao, GUO Chang, WU Chengzhong, WANG Yaonan. Low-Light Salient Object Detection Based on Source-Free Domain Adaptation [J]. Computer Engineering, 2025, 51(4): 75-84.
[8]	CUI Jinrong, YE Weihao, ZHENG Hong, LIU Tonglai, QI Long, XU Yong. Rice Seedling Counting in Complex Environments Based on Domain-Adaptive NWD-YOLOv5 [J]. Computer Engineering, 2025, 51(3): 320-333.
[9]	ZHANG Zhaoxin, HUANG Shize, ZHANG Bingjie, SHEN Tuo. Camouflaged Adversarial Example Generation Method for the Form of Motion Blur in Traffic Scenes [J]. Computer Engineering, 2025, 51(3): 45-53.
[10]	YUAN Yajian, MAO Li. A Lightweight Traffic Sign Detection Model with Enhanced Foregrounds [J]. Computer Engineering, 2025, 51(3): 54-63.
[11]	HUANG Shuyi, TAN Guang. Efficient Video Object Detection Based on Partitioning [J]. Computer Engineering, 2025, 51(2): 65-77.
[12]	HU Chaoju, GUO Fengyi. MODF Port State Detection Algorithm Based on Improved YOLOv7 [J]. Computer Engineering, 2025, 51(2): 78-85.
[13]	SUN Haomiao, LI Zongmin, XIAO Qian, SUN Wenjie, ZHANG Wenxin. AI-Curling: An On-Site Curling Analysis and Decision-Making Method [J]. Computer Engineering, 2025, 51(2): 102-110.
[14]	YANG Wangda, WAN Yaping, ZOU Gang, MIN Xiaoshan, WANG Yi, LU Yucheng. Research on Deep Learning Classification Method for Testing Eye Status of Driving Quality Deficiency [J]. Computer Engineering, 2025, 51(2): 149-158.
[15]	LIU Shengjie, HE Ning, WANG Xin, YU Haigang, HAN Wenjing. Human Pose-Estimation Algorithm Based on Lightweight High-Resolution Network [J]. Computer Engineering, 2025, 51(2): 278-288.

Please choose a citation manager

Content to export