改进YOLOv4-tiny的安全帽佩戴检测算法

doi:10.3778/j.issn.1002-8331.2205-0137

摘要/Abstract

摘要： 针对已有的安全帽检测方法存在的模型参数量大，难以部署在边缘设备上，以及对较小目标检测效果不好等问题，提出一种改进YOLOv4-tiny的轻量级安全帽检测模型。针对小目标丢失过多问题，增加了检测小目标的尺度，提升模型关注小目标的能力。提出了一种轻量级特征融合结构，缓解特征融合部分的语义混叠问题，并且在模型中融入了优化的注意力模块，提升模型捕获上下文信息的能力。针对分类与回归任务之间的冲突，将模型预测头替换为解耦合的预测头，采用并行的卷积分别进行分类与回归任务。将改进的模型命名为HM-YOLO，通过实验验证了HM-YOLO算法的有效性，相比YOLOv4-tiny模型，HM-YOLO模型平均精度提升了14.2个百分点，参数量减少了19%，检测速度为为63?FPS，具有良好的检测精度和实时性，更易于部署在边缘设备上。

关键词: 小目标检测, 频域注意力, 解耦头, YOLOv4, 轻量级网络

Abstract: Aiming at the problems of the existing safety helmet detection methods, such as large number of model parameters, difficulty to deploy on edge devices, and poor detection effect on small targets, an improved YOLOV4-tiny lightweight safety helmet detection model is proposed. Firstly, to solve the problem of too many small targets missing, the detection scale of small targets is increased to improve the ability of the model to focus on small targets. Secondly, a lightweight feature fusion structure is proposed to alleviate semantic confusionin feature fusion. In addition, an optimized attention module is integrated into the model to improve the ability of the model to capture context information. Finally, for the conflicts between classification and regression tasks, the model prediction head is replaced by the decoupled prediction head, and the parallel convolution is used for classification and regression tasks respectively. The improved model is named HM-YOLO, and the effectiveness of the HM-YOLO algorithm is verified by experiments. Compared with the YOLOv4-tiny model, the average accuracy of the HM-YOLO model is improved by 14.2 percentage points, the number of parameters is reduced by 19%, and the speed is 63 FPS. It has good detection accuracy and real-time performance, and is easier to deploy on edge devices.

Key words: small object detection, frequency channel attention, decoupled head, YOLOv4, lightweight network

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 Jianbo, WU Youxin. Safety Helmet Wearing Detection Algorithm of Improved YOLOv4-tiny[J]. Computer Engineering and Applications, 2023, 59(4): 183-190.

参考文献

[1] 国家统计局.中华人民共和国2021年国民经济和社会发展统计公报[EB/OL].（2022-02-28）[2022-04-16].http：//www.stats.gov.cn/xxgk/sjfb/zxfb2020/202202/t20220228_1827971.html.
National Bureau of Statistics of China.Statistical communique of the PRC on the 2021 national economic and social development[EB/OL].（2022-02-28）[2022-04-16].http：//www.stats.gov.cn/xxgk/sjfb/zxfb2020/202202/t20220228_1827971.html.
[2] DALAL N，TRIGGS B.Histograms of oriented gradients for human detection[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition，San Diego，2005：886-893.
[3] VIOLA P，JONES M.Rapid object detection using a boosted cascade of simple features[C]//2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition，Kauai，2001.
[4] KRIZHEVSKY A，SUTSKEVER I，HINTON G.ImageNet classification with deep convolutionalneural networks[J].Communications of the ACM，2017，60（6）：84-90.
[5] ZHAI S，SHANG D，WANG S，et al.DF-SSD：an improved SSD object detection algorithm based on densenet and feature fusion[J].IEEE Access，2020，8：24344-24357.
[6] WANG C，BOCHKOVSKIY?A，LIAO H.Scaled-YOLOv4：scaling cross stage partial network[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition，2021：13024-13033.
[7] GE Z，LIU S，WANG F，et al.YOLOX：exceeding YOLO series in 2021[J].arXiv：2107.08430，2021.
[8] 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，2014，40（3）：365-370.
[9] 徐守坤，王雅如，顾玉宛，等.基于改进Faster RCNN的安全帽佩戴检测研究[J].计算机应用研究，2020，37（3）：901-905.
XU S K，WANG Y R，GU Y W，et al.Safety helmet wearing detection study based on improved Faster RCNN[J].Application Research of Computers，2020，37（3）：901-905.
[10] 肖体刚，蔡乐才，高祥.改进YOLOv3的安全帽佩戴检测方法[J].计算机工程与应用，2021，57（12）：216-223.
XIAO T G，CAI L C，GAO X.Improved YOLOv3 helmet wearing detection method[J].Computer Engineering and Applications，2021，57（12）：216-223.
[11] HU J，SHEN L，ALBANIE S，et al.Squeeze-and-excitation networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2020，42（8）：2011-2023.
[12] QIN Z，ZHANG P，WU F，et al.FcaNet：frequency channel attention networks[C]//2021 IEEE/CVF International Conference on Computer Vision，Montreal，2021：783-792.
[13] WOO S，PARK J，LEE J，et al.CBAM：convolutional block attention module[C]//15th European Conference on Computer Vision，Munich.Cham：Springer，2018：3-19.
[14] ZHANG X，ZHOU X，LIN X，et al.ShuffleNet：an extremely efficient convolutional neural network for mobile devices[J].arXiv：1707s.01083，2017.
[15] SONG G，LIU Y，WANG X.Revisiting the sibling head in object detector[C]//2021 IEEE Computer Society Conference on Computer Vision and Pattern Recognition，Seattle，2020：11560-11569.
[16] WU Y，CHEN Y，YUAN L，et al.Rethinking classification and localization for object detection[C]//2020 IEEE Computer Society Conference on Computer Vision and Pattern Recognition，Seattle，2020：10183-10192.
[17] PENG D，SUN Z，CHEN Z，et al.Detecting heads using feature refine net and cascaded multi-scale architecture[C]//2018 International Conference on Pattern Recognition，Beijing，2018：2528-2533.
[18] XU Z P，ZHANG Y，CHENG J，et al.Safety helmet wearing detection based on YOLOv5 of attention mechanism[J].Journal of Physics：Conference Series，2022，13（1）：012038.
[19] 张锦，屈佩琪，孙程，等.基于改进YOLOv5的安全帽佩戴检测算法[J].计算机应用，2022，42（4）：1292-1300.
ZHANG J，QU P Q，SUN C，et al.Safety helmet wearing detection algorithm based on improved YOLOv5[J].Journal of Computer Applications，2022，42（4）：1292-1300.