LPR-YOLO：轻量化钢筋截面检测方法

沈益俊; 刘吉敏; 鲁向阳; 季燕春

doi:10.11835/j.issn.2096-6717.2026.033

您当前的位置：

首页 >

文章列表页 >

LPR-YOLO：轻量化钢筋截面检测方法

更新时间：2026-05-11

- LPR-YOLO：轻量化钢筋截面检测方法
- LPR-YOLO: A lightweight method for dense rebar section
- 土木与环境工程学报 2026年页码：1-10
- 作者机构：
  
  1.安徽理工大学土木建筑学院，安徽淮南 232000
  2.矿山深井建设技术国家工程研究中心，北京 100013
- 作者简介：
  
  沈益俊（2002-），男，主要从事视觉识别和井壁稳定性研究，E-mail：xuexishen632@gmail.com。
  刘吉敏（通信作者），女，博士，教授，E-mail：jimliu@aust.edu.cn。
- 基金信息：
  
  矿山深井建设技术国家工程研究中心开放研究基金(NERCDSC-2024001);建筑结构与地下工程安徽省重点实验室开放课题(KLBSUE-2022-01);国家级一流本科课程(2023240791)
- DOI：10.11835/j.issn.2096-6717.2026.033
  中图分类号： TU511.3
- 收稿：2026-01-20，
  
  网络首发：2026-05-11，
- 稿件说明：
移动端阅览
沈益俊,刘吉敏,鲁向阳等.LPR-YOLO：轻量化钢筋截面检测方法[J].土木与环境工程学报,

SHEN Yijun,LIU Jimin,LU Xiangyang,et al.LPR-YOLO: A lightweight method for dense rebar section[J].Journal of Civil and Environmental Engineering,
沈益俊,刘吉敏,鲁向阳等.LPR-YOLO：轻量化钢筋截面检测方法[J].土木与环境工程学报, DOI：10.11835/j.issn.2096-6717.2026.033.

SHEN Yijun,LIU Jimin,LU Xiangyang,et al.LPR-YOLO: A lightweight method for dense rebar section[J].Journal of Civil and Environmental Engineering, DOI：10.11835/j.issn.2096-6717.2026.033.

摘要

针对预制梁厂钢筋盘点场景中存在的密集堆叠、相互遮挡、难以清点以及现有深度学习检测模型参数量大、难以在边缘设备部署等问题，提出一种基于YOLOv8n的轻量化密集钢筋截面检测方法：LPR-YOLO。该方法利用Ghost模块重构YOLOv8n的主干网络，通过高效线性运算替代部分标准卷积，在保证特征提取有效性的同时大幅降低模型的参数量与计算冗余；在颈部网络嵌入卷积块注意力机制（CBAM），从通道和空间两个维度增强模型对钢筋截面关键特征的聚焦能力，抑制复杂背景噪声及光照对钢筋截面的干扰。结果表明，LPR-YOLO的mAP@0.5达到94.8%，相较于原始基线模型提高了2.2%，同时模型参数量和计算量分别降低了1.3×10⁶和3GFLOPs。在强光干扰、泥污遮挡等复杂工况下，LPR-YOLO也拥有较优秀的泛化能力。

Abstract

To address the challenges of dense stacking

mutual occlusion

and the difficulty of deploying heavy deep learning models on edge devices in prefabricated beam yard rebar inventory scenarios

this paper proposes a lightweight dense rebar cross-section detection method named LPR-YOLO

based on YOLOv8n. Firstly

the method reconstructs the backbone network of YOLOv8n using Ghost modules

which replace standard convolutions with cheap linear operations to significantly reduce model parameters and computational redundancy while maintaining feature extraction effectiveness. Secondly

the Convolutional Block Attention Module (CBAM) is integrated into the neck network to enhance the model's focus on key rebar cross-section features from both channel and spatial dimensions

effectively suppressing complex background noise and lighting interference. The experimental results show that LPR-YOLO achieved an mAP@0.5 of 94.8%

which is a 2.2% improvement compared to the original baseline model

while the model parameters and GFLOPs were reduced by 1.3M and 3

respectively. LPR-YOLO also demonstrates excellent generalization ability under complex conditions such as strong light interference and mud occlusion.

关键词

Keywords

references

伊东波 , 张阔 . 预制梁场智慧中枢驱动智能化生产研究 [J ] . 绿色建造与智能建筑 , 2025 ( 8 ): 73 - 76 .

Yi D B , Zhang K . Research on the intelligent production driven by the smart hub of prefabricated beam yard [J ] . Green Construction and Intelligent Building , 2025 ( 8 ): 73 - 76 . (in Chinese)

李长春 , 郝艳海 . 智能化预制梁厂功能区布局及优化设计 [J ] . 建筑施工 , 2023 , 45 ( 6 ): 1199 - 1202 .

Li C C , Hao Y H . Functional area layout and optimization design of intelligent precast beam factory [J ] . Building Construction , 2023 , 45 ( 6 ): 1199 - 1202 . (in Chinese)

刘天文 . 建筑工程施工技术与现场施工管理 [J ] . 建材发展导向 , 2025 , 23 ( 6 ): 97 - 99 .

Liu T W . Construction technology and site construction management of building engineering [J ] . Development Guide to Building Materials , 2025 , 23 ( 6 ): 97 - 99 . (in Chinese)

Wang H , Polden J , Jirgens J , et al . Automatic rebar counting using image processing and machine learning [C ] // 2019 IEEE 9th Annual International Conference on CYBER Technology in Automation, Control, and Intelligent Systems (CYBER) . July 29-August 2 , 2019 . Suzhou, China . IEEE , 2019: 900 - 904 .

倪富陶 , 李倩 , 聂云靖 , 等 . 基于改进YOLOv8的轻量化钢筋端面检测算法研究 [J ] . 太原理工大学学报 , 2024 , 55 ( 4 ): 696 - 704 .

Ni F T , Li Q , Nie Y J , et al . Lightweight rebar end detection algorithm based on improved YOLOv8 [J ] . Journal of Taiyuan University of Technology , 2024 , 55 ( 4 ): 696 - 704 . (in Chinese)

臧晶 , 刘冠楠 , 杨慧英 . Otsu算法的打捆钢筋计数方法研究 [J ] . 沈阳理工大学学报 , 2017 , 36 ( 6 ): 47 - 50, 60 .

Zang J , Liu G N , Yang H Y . A method of counting packed steel bar based on Otsu algorithm [J ] . Journal of Shenyang Ligong University , 2017 , 36 ( 6 ): 47 - 50, 60 . (in Chinese)

袁中锦 , 刘宇 , 时娟 . 基于空间相关性的成捆钢筋图像计数算法研究 [J ] . 电脑与信息技术 , 2019 , 27 ( 4 ): 10 - 13 .

Yuan Z J , Liu Y , Shi J . Research on counting algorithms of bundled steel bar image based on spatial-correlation [J ] . Computer and Information Technology , 2019 , 27 ( 4 ): 10 - 13 . (in Chinese)

刘啸宇 , 黄永 , 徐峰 , 等 . 基于轻量化卷积神经网络的桥梁斜拉索PE护套损伤识别方法 [J ] . 土木与环境工程学报(中英文) , 2025 , 47 ( 1 ): 167 - 178 .

Liu X Y , Huang Y , Xu F , et al . Damage identification method of PE sheath of bridge stay cable based on lightweight convolutional neural network [J ] . Journal of Civil and Environmental Engineering , 2025 , 47 ( 1 ): 167 - 178 . (in Chinese)

姚刚 , 廖港 , 杨阳 , 等 . 基于卷积神经网络的预制叠合板多目标智能化检测方法 [J ] . 土木与环境工程学报(中英文) , 2024 , 46 ( 1 ): 93 - 101 .

Yao G , Liao G , Yang Y , et al . Multi-target intelligent detection method of prefabricated laminated board based on convolutional neural network [J ] . Journal of Civil and Environmental Engineering , 2024 , 46 ( 1 ): 93 - 101 . (in Chinese)

王超 , 张运楚 , 孙绍涵 , 等 . 改进YOLOv5算法的钢筋端面检测 [J ] . 计算机系统应用 , 2022 , 31 ( 4 ): 68 - 80 .

Wang C , Zhang Y C , Sun S H , et al . Steel-bar end face detection based on improved YOLOv5 algorithm [J ] . Computer Systems and Applications , 2022 , 31 ( 4 ): 68 - 80 . (in Chinese)

Su Z Q , Fang K L , Peng Z T , et al . Rebar automatically counting on the product line [C ] // 2010 IEEE International Conference on Progress in Informatics and Computing . December 10-12, 2010 , Shanghai, China . IEEE , 2010 : 756 - 760 .

石京磊 . 基于卷积神经网络的钢筋计数算法研究 [D ] . 武汉 : 华中科技大学 , 2019 .

Shi J L . Rebar counting algorithm research based on convolutional neural network [D ] . Wuhan : Huazhong University of Science and Technology , 2019 . (in Chinese)

Wang S , Kim M , Hae H , et al . The development of a rebar-counting model for reinforced concrete columns: using an unmanned aerial vehicle and deep-learning approach [J ] . Journal of Construction Engineering and Management , 2023 , 149 ( 11 ): 04023111 .

Sohan M , Sai Ram T , Rami Reddy C V . A review on YOLOv8 and its advancements [M ] // Data Intelligence and Cognitive Informatics . Singapore Springer Nature Singapore, 2024 : 529 - 545 .

翟永杰 , 朱浩硕 , 李佳蔚 , 等 . VDD-YOLO: 用于车辆变形损伤的评估方法 [J ] . 科学技术与工程 , 2025 , 25 ( 29 ): 12560 - 12569 .

Zhai Y J , Zhu H S , Li J W , et al . VDD-YOLO: method for vehicle deformation and damage detection [J ] . Science Technology and Engineering , 2025 , 25 ( 29 ): 12560 - 12569 . (in Chinese)

熊恩杰 , 张荣芬 , 刘宇红 , 等 . 面向交通标志的Ghost-YOLOv8检测算法 [J ] . 计算机工程与应用 , 2023 , 59 ( 20 ): 200 - 207 .

Xiong E J , Zhang R F , Liu Y H , et al . Ghost-YOLOv8 detection algorithm for traffic signs [J ] . Computer Engineering and Applications , 2023 , 59 ( 20 ): 200 - 207 . (in Chinese)

Han K , Wang Y H , Tian Q , et al . GhostNet: more features from cheap operations [C ] // 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . June 13 - 19 , 2020 . Seattle, WA, USA . IEEE , 2020: 1577 - 1586 .

Hou H T , Guo M Z , Wang W , et al . Improved lightweight head detection based on GhostNet-SSD [J ] . Neural Processing Letters , 2024 , 56 ( 2 ): 126 .

Paoletti M E , Haut J M , Pereira N S , et al . Ghostnet for hyperspectral image classification [J ] . IEEE Transactions on Geoscience and Remote Sensing , 2021 , 59 ( 12 ): 10378 - 10393 .

Woo S , Park J , Lee J Y , et al . CBAM: convolutional block attention module [M ] // Computer Vision – ECCV 2018 . ChamSpringer International Publishing2018: 3 - 19 .

Chen L J , Yao H D , Fu J Y , et al . The classification and localization of crack using lightweight convolutional neural network with CBAM [J ] . Engineering Structures , 2023 , 275 : 115291 .

赵本省 . 基于智能建造的装配式建筑施工关键技术研究与应用 [D ] . 郑州 : 郑州大学 , 2020 .

Zhao B S/X) . Research and application of key technologies of prefabricated building construction based on intelligent construction [D ] . Zhengzhou : Zhengzhou University , 2020 . (in Chinese)

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于改进DeepLabV3+的既有钢结构表面锈蚀图像分割方法