基于SSG-YOLOv8n模型的砖砌墙体裂缝形态识别方法

陈逵; 赵亚伟; 王光明; 梁建国

doi:10.3724/j.gyjzG25051205

基于SSG-YOLOv8n模型的砖砌墙体裂缝形态识别方法

doi: 10.3724/j.gyjzG25051205

陈逵^1,2,
赵亚伟^1, ,,
王光明³,
梁建国¹

1. 长沙理工大学土木与环境工程学院, 长沙 410000;
2. 长沙理工大学桥梁与建筑绿色建造及维护湖南省重点实验室, 长沙 410000;
3. 四川宏图都市设计咨询集团有限公司内蒙古设计咨询分公司, 内蒙古鄂尔多斯 017200

基金项目:

长沙理工大学土木工程优势特色重点学科创新性项目（16ZDXK07）。

详细信息

作者简介:
陈逵，博士，硕士生导师，主要从事工程检测与加固改造。

通讯作者:
赵亚伟，2356096317@qq.com。

计量
- 文章访问数: 60
- HTML全文浏览量: 11
- PDF下载量: 4
- 被引次数: 0
出版历程
- 收稿日期: 2025-05-12
- 网络出版日期: 2025-09-12

A Method for Crack Pattern Recognition of Brick Walls Based on the SSG-YOLOv8n Model

1. School of Civil and Environmental Engineering, Changsha University of Science and Technology, Changsha 410000, China;
2. Hunan Provincial Key Laboratory of Green Construction and Maintenance of Bridges and Buildings, Changsha University of Science and Technology, Changsha 410000, China;
3. Sichuan Hongtu Urban Design & Consulting Group Limited Inner Mongolia Design & Consulting Branch Company, Ordos 017200, China

摘要

摘要: 砖砌墙体裂缝的形成与发展是一个渐进过程，管理单位若在日常排查中忽视其动态变化，极易对人民生命财产安全构成威胁。针对传统人工检测方法存在的高风险、低效率、高成本等问题，提出一种改进的SSG-YOLOv8n模型。该模型在YOLOv8n基础架构上，通过引入SPD Conv卷积模块、增设小目标检测头并嵌入SAFM注意力机制模块，有效提升砖砌墙体裂缝形态的识别检测精度。试验数据显示，改进模型的mAP@50指标较原始YOLOv8n模型提升4.3%。为平衡检测精度提升与模型复杂度控制，进一步集成GhostNet轻量化模块，在保持精度优势的同时，显著降低浮点运算量、参数量及模型尺寸。试验结果表明：改进后的模型兼具检测精度与计算效率优势。
- 砖砌墙体裂缝 /
- 裂缝形态识别 /
- YOLOv8 /
- 深度学习 /
- 智能检测
Abstract: The formation and development of cracks in brick walls is a gradual process. If the management unit ignores dynamic changes during daily inspections， it is highly likely to pose a threat to the safety of people's lives and property. To address the problems of high risk， low efficiency， and high cost existing in traditional manual detection methods， this paper proposes an improved SSG-YOLOv8n model. Based on the basic architecture of YOLOv8n， this model significantly enhances the recognition and detection accuracy of the crack patterns of brick walls by introducing the SPD Conv convolution module， adding a small target detection head， and embedding the SAFM attention mechanism module. The experimental data showed that the mAP@50 index of the improved model was 4.3% higher than that of the original YOLOv8n model. To balance detection accuracy improvement and model complexity control， the GhostNet lightweight module was further integrated. While maintaining the accuracy advantage， it significantly reduced floating-point operations， parameter count， and model size. The experimental results indicated that the improved model achieved both high detection accuracy and computational efficiency， providing important technical support for automated crack pattern detection in brick walls.
- cracks in brick walls /
- crack pattern recognition /
- YOLOv8 /
- deep learning /
- intelligent detection

HTML全文

参考文献(19)

[1]	国家统计局. 中国统计年鉴［M］. 北京:中国统计出版社，2024.
[2]	SINGH S B，MUNJAL P. Bond strength and compressive stress-strain characteristics of brick masonry［J］. Journal of Building Engineering，2017，9:10-16.
[3]	CAO X L，WEI X L，HUO X Q，et al. Self-powered retractable reel sensor for crack monitoring and warning in civil infrastructures［J］. Chemical Engineering Journal，2023，478，147238.
[4]	CHEN W，CHEN C Y，LIU M，et al. Wall cracks detection in aerial images using improved mask R-CNN［J］. Computers，Materials& Continua，2022，73（1）. DOI: 10.32604/cmc.2022.028571
[5]	XIONG，C Q，ZAYED T，ABDELKADER E M. A novel YOLOv8-GAM-Wise-IoU model for automated detection of bridge surface cracks［J］. Construction and Building Materials，2024，414，135025.
[6]	THOHARI A N A，KARIMA A，SANTOSO K，et al. Crack detection in building through deep learning feature extraction and machine learning approch［J］. Journal of Applied Informatics and Computing，2024，8（1）:1-6.
[7]	CHEN Y L，ZHU Z L，LIN Z J，et al. Building surface crack detection using deep learning technology［J］. Buildings，2023，13（7），1814.
[8]	ZHU W，ZHANG H，EASTWOOD J，et al. Concrete crack detection using lightweight attention feature fusion single shot multibox detector［J］. Knowledge-Based Systems，2023，261，110216.
[9]	LAN M L，YANG D，ZHOU S X，et al. Crack detection based on attention mechanism with YOLOv5［J］. Engineering Reports，2025，7（1），e12899.
[10]	YAO J L，XU S，HUANG F J，et al. Improved lightweight infrared road target detection method based on YOLOv8［J］. Infrared Physics& Technology，2024，141，105497.
[11]	LIU Y F，GAO W L，ZHAO T T，et al. A rapid bridge crack detection method based on deep learning［J］. Applied Sciences，2023，13（17），9878.
[12]	LI Q J，ZHANG G Y，YANG P. CL-YOLOv8:crack detection algorithm for fair-faced walls based on deep learning［J］. Applied Sciences，2024，14（20），9421.
[13]	凌同华，贝政豪，张胜，等. 基于YOLO-Pipe和ByteTrack的排水管道缺陷检测［J］. 中国给水排水，2025，41（3）:125-130.
[14]	张汉钰. 基于深度学习的砌体结构房屋裂缝识别及安全评价研究［D］. 南京:南京理工大学，2020.
[15]	ZHANG S，BEI Z H，LING T H，et al. Research on high-precision recognition model for multi-scene asphalt pavement distresses based on deep learning［J］. Scientific Reports，2024，14（1），25416.
[16]	SUNKARA R，LUO T. No more strided convolutions or pooling:a new CNN building block for low-resolution images and small objects［C］// Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Cham:Springer Nature Switzerland，2022:443-459.
[17]	LI Y，XIN X Y. Crack detection method based on an improved YOLOv8 model［C］// 2024 IEEE 7th International Conference on Automation，Electronics and Electrical Engineering（AUTEEE）. IEEE，2024:226-231.
[18]	SUN L，DONG J X，TANG J H，et al. Spatially-adaptive feature modulation for efficient image super-resolution［C］// Proceedings of the IEEE/CVF International Conference on Computer Vision. 2023:13190-13199.
[19]	HAN K，WANG Y H，TIAN Q，et al. Ghostnet:More features from cheap operations［C］// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020:1580-1589.