[1]钟铭恩,谭佳威,袁彬淦,等.复杂交通环境下二轮机动车乘员头盔检测算法[J].华侨大学学报(自然科学版),2023,44(3):301-308.[doi:10.11830/ISSN.1000-5013.202212028]
　ZHONG Mingen,TAN Jiawei,YUAN Bingan,et al.Helmet Detection Algorithm of Two-Wheeled Motor Vehicle Occupant in Complex Traffic Environment[J].Journal of Huaqiao University(Natural Science),2023,44(3):301-308.[doi:10.11830/ISSN.1000-5013.202212028]
点击复制

复杂交通环境下二轮机动车乘员头盔检测算法()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第44卷
期数:
2023年第3期
页码:
301-308
栏目:
出版日期:
2023-05-12

文章信息/Info

Title:
Helmet Detection Algorithm of Two-Wheeled Motor Vehicle Occupant in Complex Traffic Environment
文章编号:
1000-5013(2023)03-0301-08
作者:
钟铭恩1;  谭佳威1;  袁彬淦2;  吴志华1;  冯妍1;  朱程林1
1. 厦门理工学院 机械与汽车工程学院, 福建 厦门 361024;2. 厦门大学 航空航天学院, 福建 厦门 361104
Author(s):
ZHONG Ming’en1;  TAN Jiawei1;  YUAN Bin’gan2; WU Zhihua1;  FENG Yan1;  ZHU Chenglin1
1. School of Mechanical and Automotive Engineering, Xiamen University of Technology, Xiamen 361024, China; 2. School of Aeronautics and Astronautics, Xiamen University, Xiamen 361104, China
关键词:
二轮机动车;  头盔检测;  YOLOv7;  轻量级网络;  感受野;  注意力机制
Keywords:
two-wheeled motor vehicle;  helmet detection;  YOLOv7;  lightweight network;  receptive field;  attention mechanism
分类号:
TP391.4;U483
DOI:
10.11830/ISSN.1000-5013.202212028
文献标志码:
A
摘要:
针对现有二轮机动车乘员头盔检测算法在目标密集分布、随机遮挡等情况下效果较差且难以在边缘设备上应用的问题,制作了具有针对性的数据集,对比现有模型后,以YOLOv7为参考提出一种复杂交通环境下二轮机动车乘员头盔检测算法.首先,采用EfficientNet-B3作为主干网络,可提高特征提取能力且更为轻量化;其次,将增大感受野模块(RFB)引入特征融合结构中,以增大模型感受野,提升小目标头盔检测能力;最后,在检测头嵌入SimAM机制,在不增加参数的前提下提高算法精度.结果表明:相较于YOLOv7,文中算法的准确率、召回率和平均准确率分别提高了2.84%,2.26%和3.26%,参数量和运算量分别为YOLOv7的33.1%,23.5%,可实现当前主流模型算法的最佳检测性能和效率;在NVIDIA Jetson Nano开发板上的处理速度达到47.58 F·s-1,可满足边缘设备部署需求.
Abstract:
The existing helmet detection algorithms for two-wheeled motor vehicle occupant are less effective in the case of dense object distribution and random occlusion, and difficult to apply on edge devices. To address this problem, a targeted dataset is created and a helmet detection algorithm of two-wheeled motor vehicle occupant in complex traffic environment is proposed using YOLOv7 as a reference after comparing existing models. Firstly, EfficientNet-B3 is used as the backbone network to improve the feature extraction capability and make it more lightweight. Secondly, receptive field block(RFB)is introduced into the feature fusion structure to increase the receptive field of the model, and improve the detection capability of small target helmets. Finally, SimAM mechanism is embedded in the detection head to improve the accuracy of the algorithm without increasing the number of parameters. The results show that compared to YOLOv7, the accuracy, recall rate and average accuracy of the proposed algorithm have been improved by 2.84%, 2.26% and 3.26% respectively, and the number of parameters and operations are 33.1% and 23.5% of YOLOv7 respectively, achieving the best detection performance and efficiency of current mainstream model algorithms. The processing speed on the NVIDIA Jetson Nano development board reaches 47.58 frames per second, which can meet the requirements of edge device deployment.

参考文献/References:

[1] World Health Organization.Global status report on road safety 2018: Summary [R].Geneva:WHO,2018.
[2] LI Xiaofei,FABIAN F,YANG Yue,et al.A new benchmark for vison-based cyclist detection[C]//Proceedings of the IEEE Intelligent Vehicles Symposium(Ⅳ).Gothenburg:IEEE Press,2016:1109-1114.
[3] LIN Hanhe,FELIX W S.HELMET dataset[EB/OL].(2020-03-06)[2022-10-23] .https://osf.io/4pwj8/.
[4] 陈闯闯,胡绍方.密集场景下头盔佩戴智能检测研究[J].智能计算机与应用,2020,10(9):223-224.
[5] WANG Wei,GAO Song,SONG Renjie.A safety helmet detection method based on the combination of SSD and HSV color space[M]//KIM H.IT Convergence and Security.Washington D C:Mineralogical Society of America,2021:117-211.DOI:10.1007/978-981-15- 9354-3_12.
[6] 刘琛,王江涛,王明阳.引入视觉机制的SSD网络在摩托车头盔佩戴检测中的应用[J].电子测量与仪器学报,2021,35(3):144-151.DOI:10.13382/j.jemi.B2003332.
[7] 冉险生,陈卓,张禾.改进YOLOv2算法的道路摩托车头盔检测[J].电子测量技术,2021,44(24):105-115.DOI: 10.19651/j.cnki.emt.2107718.
[8] 冉险生,张之云,陈卓,等.基于改进DeepSORT算法的摩托车头盔佩戴检测[EB/OL].(2022-07-27)[2022-10-27] .http://kns.cnki.net/kcms/detail/11.2127.TP.20220726.1653.016.html.
[9] REN Shaoqing,HE Kaiming,GIRSHICK R,et al.Faster R-CNN: Towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,39(6):1137-1149.DOI:10.1109/tpami.2016.2577031.
[10] LIU Wei,ANGUELOV D,ERHAN D,et al.Ssd: Single shot multibox detector[C]//European Conference on Computer Vision.Amsterdam:Springer,2016:21-37.DOI:10.1007/978-3-319-46448-0_2.
[11] REDMON J,FARHADI A. YOLOv3: An incremental improvement[EB/OL].(2018-04-08)[2022-10-23] . https://doi.org/10.48550/arXiv.1804.02767.
[12] BOCHKOVSKIY A,WANG C Y,LIAO H Y M.YOLOv4: Optimal speed and accuracy of object detection[EB/OL].(2020-04-23)[2022-10-23] .https://doi.org/10.48550/arXiv.2004.10934.
[13] 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[EB/OL].(2022-07-06)[2022-10-23] .https://doi.org/10.48550/arXiv.2207.02696.
[14] ZHOU Xingyi,WANG Dequan,KR?HENBüHL P. Objects as points[EB/OL].(2019-04-16)[2022-10-23] .https://doi.org/10.48550/arXiv.1904.07850.
[15] LIN T Y,GOYAL P,GIRSHICK R,et al.Focal loss for dense object detection[C]//Proceedings of the IEEE International Conference on Computer Vision.Venice:IEEE Press,2017:2980-2988.DOI:10.1109/iccv.2017.324.
[16] DOSOVITSKIY A,BEYER L,KOLESNIKOV A,et al.An image is worth 16×16 words: Transformers for image recognition at scale[ED/OL].(2020-10-22)[2022-10-25] .http://www.arxiv-vanity.com/papers/2010.11929/.
[17] TAN Mingxing,LE Q V.Efficientnet: Rethinking model scaling for convolutional neural networks[C]//International Conference on Machine Learning.Long Beach:[s.n.],2019:6105-6114.DOI:10.48550/arXiv.1905.11946.
[18] 柳长源,何先平,毕晓君.融合注意力机制的高效率网络车型识别[J].浙江大学学报(工学版),2022,56(4):775-782.DOI:10.3 785/j.issn.1008-973X.2022.04.017.
[19] 陶英杰,张维纬,马昕,等.面向无人机视频分析的车辆目标检测方法[J].华侨大学学报(自然科学版),2022,43(1):111-118.DOI:10.11830/ISSN.1000-5013.202011014.
[20] LIU Songtao,HUANG Di.Receptive field block net for accurate and fast object detection[C]//Proceedings of the European Conference on Computer Vision.Munich:[s.n.],2018:385-400.DOI:10.48550/arXiv.1711.07767.
[21] WANG Qilong,WU Banggu,ZHU Pengfei,et al.ECA-Net: Efficient channel attention for deep convolutional neural networks[C]//IEEE Conference on Computer Vision and Pattern Recognition.Salt Lake City:IEEE Press,2020:13-19.DOI:10.48550/arXiv.1910.03151.
[22] YANG Lingxiao,ZHANG Ruyuan,LI Lida,et al.Simam: A simple, parameter-free attention module for convolutional neural networks[C]//International Conference on Machine learning.[S.l.]:IEEE Press,2021:11863-11874.
[23] 王年涛,王淑青,黄剑锋,等.基于改进YOLOv5神经网络的绝缘子缺陷检测方法[J].激光杂志,2022,43(8):60-65.DOI:10.14016/j.cnki.jgzz.2022.08.060.
[24] LIU Ze,LIN Yutong,CAO Yue,et al.Swin transformer: Hierarchical vision transformer using shifted windows[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Montreal:IEEE Press,2021:10012-10022.DOI:10.48550/arXiv.2103.14030.

备注/Memo

备注/Memo:
收稿日期: 2022-12-23
通信作者: 钟铭恩(1980-),男,教授,博士,主要从事机器视觉、人工智能和智能交通的研究.E-mail:zhongmingen@xmut.edu.cn.
基金项目: 国家自然科学基金资助项目(51978592); 福建省自然科学基金资助项目(2019J01859)http://www.hdxb.hqu.edu.cn
更新日期/Last Update: 2023-05-20