[1]张震,张晨稳,张俊杰,等.改进YOLOv7-tiny的施工现场安全衣帽穿戴检测算法[J].郑州大学学报(工学版),2026,47(XX):1-8.[doi:10.13705/j.issn.1671-6833.2025.05.001]
　ZHANG Zhen,ZHANG Chenwen,ZHANG Junjie,et al.Improved YOLOv7-tiny Safety Clothing and Hat Detection Algorithm For Construction Site[J].Journal of Zhengzhou University (Engineering Science),2026,47(XX):1-8.[doi:10.13705/j.issn.1671-6833.2025.05.001]
点击复制

改进YOLOv7-tiny的施工现场安全衣帽穿戴检测算法()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
47
期数:
2026年XX
页码:
1-8
栏目:
出版日期:
2026-09-10

文章信息/Info

Title:
Improved YOLOv7-tiny Safety Clothing and Hat Detection Algorithm For Construction Site
作者:
张震1; 张晨稳2 ; 张俊杰3; 裴胜利3; 王文娟4
1.郑州大学 电气与信息工程学院,河南 郑州450001;2郑州大学 河南先进技术研究院,河南 郑州450001;3河南汇融油气技术有限公司,河南 郑州450001;4广东省轻工业技师学院 机电工程学院,广东 广州511330
Author(s):
ZHANG Zhen1 ;  ZHANG Chenwen2 ;  ZHANG Junjie3 ;  PEI Shengli3 ;  WANG Wenjuan4
1. School of Electrical and Information Engineering, Zhengzhou University, Zhengzhou 450001, China ; 2. School of Henan Institute of Advanced Technology , Zhengzhou University, Zhengzhou 450001, China; 3. Henan Huirong Oil & Gas Technology Co. , Ltd . , Zhengzhou 450001, China; 4 . School of Mechanical and Electrical Engineering, Guangdong Province Technician College of Light Industry, Guangzhou, Guangdong 511330, China
关键词:
YOLOv7-tiny; 注意力机制; RFEM; Shape-IoU; 安全衣帽检测
Keywords:
YOLOv7-tiny;  Attention mechanism;  RFEM;  Shape-IoU;  safety helmet and vest delection
分类号:
TP391
DOI:
10.13705/j.issn.1671-6833.2025.05.001
文献标志码:
A
摘要:
针对当前施工现场安全衣帽穿戴检测算法在复杂背景、弱光环境及目标遮挡情况下的抗干扰能力不足，导致检测精度低、漏检率高和误检现象频繁等问题，提出了一种改进YOLOv7-tiny的施工现场安全衣帽穿戴检测算法。首先，在特征提取区域引入了EMA注意力机制增强网络特征提取能力，弱化复杂背景干扰；其次，在特征融合部分插入RFEM模块提升网络感受野，获取更广泛的上下文信息，增强对小目标的感知能力；最后，采用Shape-IoU替换IoU边界回归损失函数，提升检测准确性。实验结果表明：改进模型在自制数据集上的mAP@0.5达到90.4%，相比原模型提高3.0百分点；检测速度达到了93帧/s，模型参数仅为6.1×10⁶。相比YOLOv8s、YOLOv9s等模型，所提算法在检测精度、速度和模型轻量化方面更具优势，更适合施工现场的实时检测应用。
Abstract:
In response to the insufficient robustness to interference of current safety helmet and clothing detection algorithms in complex backgrounds, weak light, and target occlusion, which lead to low detection accuracy, high false negative rates, and frequent false positive rates, an improved YOLOv7-tiny detection algorithm for safety helmets and clothing in construction sites was proposed. Firstly, the EMA attention mechanism was introduced into the feature extraction module to enhance the network’s feature extraction capability and mitigate complex background interference. Secondly, the RFEM module was integrated into the feature fusion stage to improve the network’s receptive field, acquire broader contextual information, and enhance perception for small targets. Finally, Shape-IoU was employed to replace the IoU boundary regression loss function, improving detection accuracy. Experimental results showed that the improved model achieved a mAP@0.5 of 90.4% on the proprietary dataset, 3.0 percentage points higher than the original model. The detection speed reached 93 frames/s, and the model size comprised only 6.1 million parameters. Compared to YOLOv8s, YOLOv9s, and other models, the proposed algorithm demonstrated superior performance in detection accuracy, speed, and model efficiency, making it suitable for real-time detection applications on construction sites.

参考文献/References:

[1] 宋健, 黄建军. 建筑工人安全参与行为演化博弈分析 [J]. 安全, 2021, 42(10): 62-67.
SONG J, HUANG J J. Evolutionary game analysis on safety participation behavior of construction workers [J]. Safety & Security, 2021, 42(10): 62-67.
[2] 张震, 王晓杰, 晋志华, 等. 基于轻量化YOLOv5的交通标志检测 [J]. 郑州大学学报(工学版), 2024, 45(2): 12-19.
ZHANG Z, WANG X J, JIN Z H, et al. Traffic sign detection based on lightweight YOLOv5 [J]. Journal of Zhengzhou University (Engineering Science), 2024, 45(2): 12-19.
[3] GIRSHICK R. Fast R-CNN [C]// 2015 IEEE International Conference on Computer Vision (ICCV). Piscataway: IEEE, 2015: 1440-1448.
[4] DAI J F, LI Y, HE K M, et al. R-FCN: Object detection via region-based fully convolutional networks [EB/OL]. (2016-05-20) [2024-11-22]. https://doi.org/10.48550/arXiv.1605.06409
[5] HE K M, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN [C]//2017 IEEE International Conference on Computer Vision (ICCV). Piscataway: IEEE, 2017: 2980-2988.
[6] Ultralytics. YOLOv5 [EB/OL]. (2020-05-18) [2024-04-22]. https://github.com/ultralytics/yolov5.
[7] LI C Y, LI L L, GENG Y F, et al. Yolov6 v3. 0: a full-scale reloading [EB/OL]. (2023-01-13) [2024-04-22]. https://arxiv.org/abs/2301.05586.
[8] WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors [C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2023: 7464-7475.
[9] Ultralytics. YOLOv8 [EB/OL]. (2023-01-10) [2024-11-22]. https://github.com/ultralytics/ultralytics.
[10] WANG C Y, YEH I H, LIAO H Y M. YOLOv9: learning what you want to learn using programmable gradient information [EB/OL]. (2024-2-21) [2024-04-22]. https://arxiv.org/abs/2402.13616v2.
[11] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector [C]// European Conference on Computer Vision. Cham: Springer, 2016: 21-37.
[12] LAW H, DENG J. CornerNet: detecting objects as paired keypoints [EB/OL]. (2024-2-21) [2024-11-22]. https://doi.org/10.48550/arXiv.1808.01244.
[13] ZHAO Q J, SHENG T, WANG Y T, et al. M2Det: a single-shot object detector based on multi-level feature pyramid network [J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2019, 33(1): 9259-9266.
[14] OUYANG D L, HE S, ZHANG G Z, et al. Efficient multi-scale attention module with cross-spatial learning [C]//2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Piscataway: IEEE, 2023: 1-5.
[15] YU Z P, HUANG H B, CHEN W J, et al. YOLO-FaceV2: a scale and occlusion aware face detector [J]. Pattern Recognition, 2022, 155: 110714.
[16] ZHANG H, ZHANG S J. Shape-IoU: more accurate metric considering bounding box shape and scale [EB/OL]. (2023-12-29) [2024-11-22]. https://arxiv.org/abs/2312.17663.
[17] JIANG B, LUO R, MAO J, et al. Acquisition of localization confidence for accurate object detection [C]//European Conference on Computer Vision. Cham: Springer, 2018: 816-832.
[18] MA W B, GUAN Z, WANG X, et al. Light-YOLO-FL: a target detection algorithm for reflective clothing wearing inspection [J]. Displays, 2023, 80: 102561.
[19] 赵红成, 田秀霞, 杨泽森, 等. YOLO-S: 一种新型轻量化的安全帽佩戴检测模型 [J]. 华东师范大学学报(自然科学版), 2021(5): 134-145.
ZHAO H C, TIAN X X, YANG Z S, et al. YOLO-S: a new lightweight helmet wearing detection model [J]. Journal of East China Normal University (Natural Science), 2021(5): 134-145.
[20] WANG L L, ZHANG X J, YANG H L. Safety helmet wearing detection model based on improved YOLO-M [J]. IEEE Access, 2023, 11: 26247-26257.
[21] PENG J, PENG F, JIN S Z, et al. Research on safety helmet wearing detection based on YOLO [C]//ECON 2023- 49th Annual Conference of the IEEE Industrial Electronics Society. Piscataway: IEEE, 2023: 1-6.
[22] MA X J, JI K F, XIONG B L, et al. Light-YOLOv4: an edge-device oriented target detection method for remote sensing images [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14: 10808-10820.
[23] LIU Y C, SHAO Z R, HOFFMANN N. Global attention mechanism: retain information to enhance channel-spatial interactions [EB/OL]. (2021-12-10) [2024-11-22]. 2021: 2112.05561. https://arxiv.org/abs/2112.05561v1.
[24] HU J, SHEN L, SUN G. Squeeze-and-excitation networks [C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 7132-7141.
[25] WANG Q L, WU B G, ZHU P F, et al. ECA-net: efficient channel attention for deep convolutional neural networks [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA, USA: IEEE, 2020: 11531-11539.
[26] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[EB/OL]. (2018-7-11) [2024-11-22]. https://doi.org/10.48550/arXiv.1807.06521.
[27] TONG Z, CHEN Y, XU Z, et al. Wise-IoU: bounding box regression loss with dynamic focusing mechanism [EB/OL]. (2023-01-24) [2024-11-22]. https://arxiv.org/abs/2301.10051.
[28] ZHANG Y F, REN W Q, ZHANG Z, et al. Focal and efficient IOU loss for accurate bounding box regression [J]. Neurocomputing, 2022, 506: 146-157.
[29] GEVORGYAN Z. SIoU loss: more powerful learning for bounding box regression [EB/OL]. (2022-05-25) [2024-11-22]. https://doi.org/10.48550/arXiv.2205.12740.

相似文献/References:

[1]郑建兴,郭彤彤,申利华,等.基于评论文本情感注意力的推荐方法研究[J].郑州大学学报(工学版),2022,43(02):44.[doi:10.13705/j.issn.1671-6833.2022.02.007]
　ZHENG Jianxing,GUO Tongtong,SHEN Lihua,et al.Research on Recommendation Method Based on Sentimental Attention of Review Text[J].Journal of Zhengzhou University (Engineering Science),2022,43(XX):44.[doi:10.13705/j.issn.1671-6833.2022.02.007]

备注/Memo

备注/Memo:
收稿日期:2025-01-06;修订日期:2025-02-20
基金项目:河南省重点研 发专项(231111211600) ;河南省重大公益专项(201300311200)
作者简介:张震(1966— ) ,男,河南郑州人,郑州大学教授,博士,博士生导师,主要从事计算机视觉研究,E-mail:zhangzhen66@126.com。
更新日期/Last Update: 2026-01-16