多感受野增强的爆破现场安全帽检测算法

doi:10.3778/j.issn.1002-8331.2311-0409

摘要/Abstract

摘要： 针对安全帽检测任务中存在的目标面积小、目标被不同程度遮挡、复杂背景干扰目标等问题，提出了基于YOLOX的多感受野增强的安全帽检测算法（multiple receptive field enhancement-YOLOX，MRFE-YOLOX）。在特征融合网络中增加浅层特征融合分支，提升小目标特征信息流通效率，提高了小目标的检测精度；设计基于空洞卷积组与卷积注意力机制的感受野增强模块（receptive field augmentation module，RFAM），捕获了更大范围的感受野和图像特征，改善了遮挡目标漏检率高的问题；根据三分支注意力机制构建特征增强网络（feature enhancement network，FENet），抑制背景噪音对目标区域的干扰，降低了复杂背景下的目标误检率；引入空间到深度卷积（space to depth-conv，SPD-Conv）得到无信息损失的二倍下采样特征图，保留了更多的特征信息，同时减少了模型的参数量。实验结果表明，所提算法的均值平均精度相较于基线算法提升了2.78个百分点，FPS达到了102.67，满足了爆破现场安全帽实时检测的需要。

关键词: 安全帽检测, YOLOX-s算法, 感受野, 空洞卷积, 注意力机制

Abstract: To solve the problems of small area of targets, different degrees of overlapped targets, targets interfered by complex images background, the multiple receptive field enhancement safety helmet detection algorithm based on YOLOX (MRFE-YOLOX) is proposed. A shallow feature fusion branch is added to feature fusion network to improve the small target feature fusion information circulation efficiency and the small targets detection accuracy. Receptive field augmentation module (RFAM) based on dilated convolution groups and convolution attention mechanism are designed to capture wider range receptive field and images feature, which improvs the high miss detection rate of overlapped targets. Feature enhancement network (FENet) based on triplet attention mechanism is constructed to suppress the background noise interference to the target area and make the model focus on the targets feature region and decrease targets false detection rate in complex background. The space to depth-conv (SPD-Conv) module is introduced to obtain the double down-sampling feature maps without information loss, which retains more feature information and reduces the model’s parameters number. The experimental results demonstrate that MRFE-YOLOX’s mean average precision promotes 2.78 percentage points compared with YOLOX and FPS reaches 102.67, which meets the need for real-time safety helmet detection in blasting sites.

Key words: safety helmet detection, YOLOX-s algorithm, receptive field, dilated convolution, attention mechanism

王新良, 王璐莹. 多感受野增强的爆破现场安全帽检测算法[J]. 计算机工程与应用, 2025, 61(7): 315-324.

WANG Xinliang, WANG Luying. Helmet Detection Algorithm Based on Multiple Receptive Field Enhancement in Blasting Scene[J]. Computer Engineering and Applications, 2025, 61(7): 315-324.

参考文献

[1] 国家质量监督检验检疫总局, 中国国家标准化管理委员会. 爆破安全规程: GB 6722—2014[S]. 北京: 中国标准出版社, 2015.
General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China, Standardization Administration of the People’s Republic of China. Safety regulations for blasting: GB 6722—2014[S]. Beijing: Standards Press of China, 2015.
[2] 蒙云琪, 吴剑锋, 张兆龙, 等. 爆破工程现场监理视频监控系统的应用[J]. 爆破, 2017, 34(2): 148-151.
MENG Y Q, WU J F, ZHANG Z L, et al. Application of video monitoring system in blasting engineering field[J]. Blasting, 2017, 34(2): 148-151.
[3] 高腾, 张先武, 李柏. 深度学习在安全帽佩戴检测中的应用研究综述[J]. 计算机工程与应用, 2023, 59(6): 13-29.
GAO T, ZHANG X W, LI B. Review on application of deep learning in helmet wearing detection[J]. Computer Engineering and Applications, 2023, 59(6): 13-29.
[4] 刘晓慧, 叶西宁. 肤色检测和Hu矩在安全帽识别中的应用[J]. 华东理工大学学报(自然科学版), 2014, 40(3): 365-370.
LIU X H, YE X N. Skin color detection and Hu moments in helmet recognition research[J]. Journal of East China University of Science and Technology (Natural Science Edition), 2014, 40(3): 365-370.
[5] 李太华, 王迪. 基于安全帽颜色识别的人员身份认证算法在变电站的应用[J]. 电脑知识与技术, 2014, 10(5): 1018-1020.
LI T H, WANG D. The algorithm of identity authentication based on the recognition of safety helmet color that is used in substation[J]. Computer Knowledge and Technology, 2014, 10(5): 1018-1020.
[6] 潘坚跃, 麻立群, 彭斌, 等. 人体及穿戴特征识别在电力设施监控中的应用[J]. 电子设计工程, 2015, 23(10): 68-71.
PAN J Y, MA L Q, PENG B, et al. Design of portable power control unit test system based on LabWindows/CVI[J]. Electronic Design Engineering, 2015, 23(10): 68-71.
[7] 刘云波, 黄华. 施工现场安全帽佩戴情况监控技术研究[J]. 电子科技, 2015, 28(4): 69-72.
LIU Y B, HUANG H. Research on monitoring of workers’ helmet wearing at the construction site[J]. Electronic Science and Technology, 2015, 28(4): 69-72.
[8] 赵震. 基于OpenCV的人体安全帽检测的实现[J]. 电子测试, 2017(14): 24-25.
ZHAO Z. Implementation of human helmet detection based on openCV[J]. Electronic Test, 2017(14): 24-25.
[9] RUBAIYAT A H M, TOMA T T, KALANTARI-KHANDANI M, et al. Automatic detection of helmet uses for construction safety[C]//Proceedings of the 2016 IEEE/WIC/ACM International Conference on Web Intelligence Workshops. Piscataway: IEEE, 2016: 135-142.
[10] 余震, 何留杰, 王振飞. 基于中智理论与方向α-均值的图像边缘检测算法[J]. 电子测量与仪器学报, 2020, 34(3): 43-50.
YU Z, HE L J, WANG Z F. Image edge detection based on intelligence theory and direction α-mean[J]. Journal of Electronic Measurement and Instrumentation, 2020, 34(3): 43-50.
[11] 张勇, 吴孔平, 高凯, 等. 基于改进型YOLOV3安全帽检测方法的研究[J]. 计算机仿真, 2021, 38(5): 413-417.
ZHANG Y, WU K P, GAO K, et al. Helmet detection based on modified YOLOV3[J]. Computer Simulation, 2021, 38(5): 413-417.
[12] 周敏新, 张方舟, 龚声蓉. 基于新型特征融合的安全帽佩戴检测方法[J]. 计算机工程与设计, 2021, 42(11): 3181-3187.
ZHOU M X, ZHANG F Z, GONG S R. Detection of non-hardhat-use based on new feature fusion[J]. Computer Engineering and Design, 2021, 42(11): 3181-3187.
[13] 李帅, 李丽宏, 王素刚, 等. 改进YOLOv4算法的安全帽检测[J]. 现代电子技术, 2022, 45(3): 103-110.
LI S, LI L H, WANG S G, et al. Helmet detection based on improved YOLOv4 algorithm[J]. Modern Electronics Technique, 2022, 45(3): 103-110.
[14] 王建波, 武友新. 改进YOLOv4-tiny的安全帽佩戴检测算法[J]. 计算机工程与应用, 2023, 59(4): 183-190.
WANG J B, WU Y X. Safety helmet wearing detection algorithm of improved YOLOv4-tiny[J]. Computer Engineering and Applications, 2023, 59(4): 183-190.
[15] 关雅琪, 侯群. 一种用于安全帽检测场景的深度学习算法[J]. 电子世界, 2021(22): 86-88.
GUAN Y Q, HOU Q. A deep learning algorithm for helmet detection scene[J]. Electronics World, 2021(22): 86-88.
[16] 伏德粟, 高林, 刘威, 等. 基于改进YOLOv5算法的电力工人作业安全关键装备检测[J]. 湖北民族大学学报(自然科学版), 2022, 40(3): 320-327.
FU D S, GAO L, LIU W, et al. Detection of key safety equipment for electric workers based on improved YOLOv5 algorithm[J]. Journal of Hubei Minzu University (Natural Science Edition), 2022, 40(3): 320-327.
[17] 刘颖, 刘红燕, 范九伦, 等. 基于深度学习的小目标检测研究与应用综述[J]. 电子学报, 2020, 48(3): 590-601.
LIU Y, LIU H Y, FAN J L, et al. A survey of research and application of small object detection based on deep learning[J]. Acta Electronica Sinica, 2020, 48(3): 590-601.
[18] 徐光达, 毛国君. 多层级特征融合的无人机航拍图像目标检测[J]. 计算机科学与探索, 2023, 17(3): 635-645.
XU G D, MAO G J. Aerial image object detection of UAV based on multi-level feature fusion[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(3): 635-645.
[19] 李杨, 武连全, 杨海涛, 等. 一种无人机视角下的小目标检测算法[J]. 红外技术, 2023, 45(9): 925-931.
LI Y, WU L Q, YANG H T, et al. A small target detection algorithm from UAV perspective[J]. Infrared Technology, 2023, 45(9): 925-931.
[20] WANG P Q, CHEN P F, YUAN Y, et al. Understanding convolution for semantic segmentation[C]//Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2018: 1451-1460.
[21] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2018: 3-19.
[22] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 770-778.
[23] 赵鹏飞, 谢林柏, 彭力. 融合注意力机制的深层次小目标检测算法[J]. 计算机科学与探索, 2022, 16(4): 927-937.
ZHAO P F, XIE L B, PENG L. Deep small object detection algorithm integrating attention mechanism[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(4): 927-937.
[24] MISRA D, NALAMADA T, ARASANIPALAI A U, et al. Rotate to attend: convolutional triplet attention module[C]//Proceedings of the 2021 IEEE Winter Conference on Applications of Computer Vision. Piscataway: IEEE, 2021: 3138-3147.
[25] SUNKARA R, LUO T. No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects[C]//Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases. Cham: Springer, 2023: 443-459.
[26] 胡启军, 潘学鹏, 余洋, 等. 复杂作业环境下安全帽实时检测算法研究[J]. 安全与环境学报, 2024, 24(5): 1904-1912.
HU Q J, PAN X P, YU Y, et al. Research on real-time detection algorithm of construction workers’ hard hats in complex operating environments[J]. Journal of Safety and Environment, 2024, 24(5): 1904-1912.
[27] 赵敏, 杨国亮, 王吉祥, 等. 改进YOLOv7-tiny的安全帽实时检测算法[J]. 无线电工程, 2023, 53(8): 1741-1749.
ZHAO M, YANG G L, WANG J X, et al. Improved YOLOv7-tiny algorithm for safety helmet real-time detection[J]. Radio Engineering, 2023, 53(8): 1741-1749.
[28] 丁田, 陈向阳, 周强, 等. 基于改进YOLOX的安全帽佩戴实时检测[J]. 电子测量技术, 2022, 45(17): 72-78.
DING T, CHEN X Y, ZHOU Q, et al. Real-time detection of helmet wearing based on improved YOLOX[J]. Electronic Measurement Technology, 2022, 45(17): 72-78.