监控视角下密集人群口罩佩戴检测算法

doi:10.19678/j.issn.1000-3428.0065697

摘要/Abstract

摘要：

针对密集人群场景的口罩佩戴检测面临着监控视角下目标密集、互相遮挡、目标小、人脸透视畸变等难题, 同时，涵盖不规范佩戴口罩场景的公开数据集也较为缺乏。提出一种基于YOLO-v5改进的监控视角下密集人群口罩佩戴检测算法MDDC-YOLO。利用空洞卷积构造多分支感受野模块MRF-C3替换YOLO-v5中常规C3模块，解决密集人群中小目标占比大的问题。使用Repulsion Loss基于样本边界框排斥吸引的原则提高模型抗遮挡能力，并充分利用训练过程中的遮挡正样本。在此基础上, 引入ECA注意力机制进行特征通道最优化选择，并提出基于透视变换的离线数据增强方法, 结合使用更适用于生成更多小目标样本的Mosaic-9数据增强方法，解决监控视角下密集人群口罩佩戴数据集缺乏的问题。实验结果表明，MDDC-YOLO算法相较于YOLO-v5算法mAP提升6.5个百分点，并达到32帧/s的检测速度，满足密集人群口罩佩戴检测的应用需求。

关键词: 口罩佩戴检测, 密集人群, 多分支空洞卷积, ECA注意力机制, Repulsion Loss

Abstract:

In dense crowds scenario, dense targets under the monitoring perspective, mutual occlusion, small targets, and face perspective distortion cause problems in mask wearing detection. Meanwhile, public datasets covering incorrectly worn masks are also lacking. Therefore, this paper proposes a mask wearing detection algorithm from a monitoring perspective, MDDC-YOLO, based on the YOLO-v5 improvement. In view of the large proportion of small- and medium-sized targets in dense population, the conventional C3 module in YOLO-v5 is replaced with the MRF-C3 module of the atrous convolutional structure. The anti-occlusion ability of the model is also improved by using Repulsion Loss based on the principle of repulsion attraction of the sample bounding box, and the masking positive sample is fully utilized during the training process. An Efficient Channel Attention(ECA) mechanism is further introduced for optimal selection of feature channels. Finally, to address the lack of mask wearing data in the crowd from a monitoring perspective, an offline data enhancement method based on perspective transformation is proposed. The proposed Mosaic-9 data enhancement generates additional small target samples to address this problem. The experimental results show that the MDDC-YOLO algorithm provides 6.5 percentage points mAP improvement compared with YOLO-v5, thereby reaching a detection speed of 32 frame/s, which satisfies the application requirements of mask-wearing detection in dense populations.

Key words: mask wearing detection, dense crowds, multi-branch atrous convolution, Efficient Channel Attention(ECA) mechanism, Repulsion Loss

孙龙, 张荣芬, 刘宇红, 饶庭漓. 监控视角下密集人群口罩佩戴检测算法[J]. 计算机工程, 2023, 49(9): 313-320.

Long SUN, Rongfen ZHANG, Yuhong LIU, Tingli RAO. Mask Wearing Detection Algorithm for Dense Crowds from a Monitoring Perspective[J]. Computer Engineering, 2023, 49(9): 313-320.

http://www.ecice06.com/CN/Y2023/V49/I9/313

图/表 16

图1 MDDC-YOLO网络结构

Fig.1 Structure of MDDC-YOLO network

图2 C3与MRF-C3模块

Fig.2 C3 and MRF-C3 modules

图3 目标遮挡示意图

Fig.3 Schematic diagram of target occlusion

图4 ECA注意力插入位置示意图

Fig.4 Schematic diagram of ECA attention insertion position

图5 ECA注意力模块结构

Fig.5 Structure of ECA module

图6 离线数据增强示意图

Fig.6 Schematic diagram of offline data augmentation

图7 Mosaic-9数据增强示意图

Fig.7 Schematic diagram of Mosaic-9 data augmentation

图8 各类别关键点匹配示意图

Fig.8 Schematic diagrams of matching each class of landmarks

图9 训练结果

Fig.9 Train results

图10 MDDC-YOLO检测效果1

Fig.10 Detection effect 1 of MDDC-YOLO

图11 MDDC-YOLO检测效果2

Fig.11 Detection effect 2 of MDDC-YOLO

参考文献 29

1	国家卫生健康委员会. 关于印发新型冠状病毒感染不同风险人群防护指南和预防新型冠状病毒感染的肺炎口罩使用指南的通知[EB/OL]. [2022-05-28]. http://www.gov.cn/x-inwen/2020-01/31/content_5473401.htm.
	National Health Commission. Notice on the issuance of guidelines for the protection of people at different risks of novel coronavirus infection and guidelines for the use of pneumonia masks for preventing new coronavirus infection[EB/OL]. [2022-05-28]. http://www.gov.cn/x-inwen/2020-01/31/content_5473401.htm. (in Chinese)
2	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2016: 779-788.
3	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]//Proceedings of European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 21-37.
4	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal Loss for dense object detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(2): 318- 327. doi: 10.1109/TPAMI.2018.2858826
5	GIRSHICK R. Fast R-CNN[C]//Proceedings of IEEE International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2016: 1440-1448.
6	REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137- 1149. doi: 10.1109/TPAMI.2016.2577031
7	CAI Z W, VASCONCELOS N. Cascade R-CNN: delving into high quality object detection[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2018: 6154-6162.
8	春雨童, 韩飞腾, 何明珂. 新冠肺炎疫情背景下聚集性传染风险智能监测模型. 计算机工程, 2022, 48(8): 45-52, 61. URL
	CHUN Y T, HAN F T, HE M K. Intelligent monitoring model for aggregated infection risk against the background of COVID-19 epidemic. Computer Engineering, 2022, 48(8): 45-52, 61. URL
9	HE Y M, WANG Z H, GUO S Y, et al. Face mask detection algorithm based on HSV+HOG features and SVM. Journal of Measurement Science and Instrumentation, 2022, 13(3): 267- 275.
10	李雨阳, 沈记全, 翟海霞, 等. 基于改进SSD的口罩佩戴检测算法. 计算机工程, 2022, 48(8): 173-179, 186. URL
	LI Y Y, SHEN J Q, ZHAI H X, et al. Mask wearing detection algorithm based on improved SSD. Computer Engineering, 2022, 48(8): 173-179, 186. URL
11	曹城硕, 袁杰. 基于YOLO-Mask算法的口罩佩戴检测方法. 激光与光电子学进展, 2021, 58(8): 211- 218. URL
	CAO C S, YUAN J. Mask-wearing detection method based on YOLO-Mask. Laser & Optoelectronics Progress, 2021, 58(8): 211- 218. URL
12	白梅娟, 查祖福水, 王杨洋, 等. 可见光与红外图像联合口罩佩戴检测研究[J/OL]. 控制工程: 1-9[2022-05-28]. https://doi.org/10.14107/j.cnki.kzgc.20220073.
	BAI M J, CHAZU F S, WANG Y Y, et al. Mask-wearing detection based on combination of visible image and infrared image[J/OL]. Control Engineering: 1-9[2022-05-28]. https://doi.org/10.14107/j.cnki.kzgc.20220073. (in Chinese)
13	LI Y H, CHEN Y T, WANG N Y, et al. Scale-aware trident networks for object detection[C]//Proceedings of IEEE/CVF International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2020: 6053-6062.
14	WANG X L, XIAO T T, JIANG Y N, et al. Repulsion Loss: detecting pedestrians in a crowd[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2018: 7774-7783.
15	HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2018: 7132-7141.
16	WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of European Conference on Computer Vision. Berlin, Germany: Springer, 2018: 3-19.
17	WANG J, CHEN Y, CHAKRABORTY R, et al. Orthogonal convolutional neural networks[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2020: 11505-11515.
18	LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2017: 936-944.
19	LIU S, QI L, QIN H F, et al. Path aggregation network for instance segmentation[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2018: 8759-8768.
20	ZHONG Z, ZHENG L A, KANG G L, et al. Random erasing data augmentation. Proceedings of AAAI Conference on Artificial Intelligence, 2020, 34(7): 13001- 13008. doi: 10.1609/aaai.v34i07.7000
21	BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4: optimal speed and accuracy of object detection[EB/OL]. [2022-05-28]. https://arxiv.org/abs/2004.10934.
22	CABANI A, HAMMOUDI K, BENHABILES H, et al. MaskedFace-Net—a dataset of correctly/incorrectly masked face images in the context of COVID-19. Smart Health, 2021, 19, 100144. doi: 10.1016/j.smhl.2020.100144
23	JIANG X B, GAO T H, ZHU Z C, et al. Real-time face mask detection method based on YOLOv3. Electronics, 2021, 10(7): 837. doi: 10.3390/electronics10070837
24	KARRAS T, LAINE S, AILA T M. A style-based generator architecture for generative adversarial networks[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2020: 4396-4405.
25	CHEN Q, WANG Y M, YANG T, et al. You only look one-level feature[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition Washington D. C., USA: IEEE Press, 2021: 13034-13043.
26	GE Z, LIU S T, WANG F, et al. YOLOX: exceeding YOLO series in 2021[EB/OL]. [2022-05-28]. https://arxiv.org/abs/2107.08430.
27	LI C Y, LI L L, JIANG H L, et al. YOLOv6: a single-stage object detection framework for industrial applications[EB/OL]. [2022-05-28]. https://arxiv.org/abs/2209.02976.
28	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]. [2022-05-28]. https://arxiv.org/abs/2207.02696.
29	GLENN J. YOLO-v5[EB/OL]. [2022-05-28]. https://github.com/ultralytics/yolov5.

[1]	张欣怡, 张飞, 郝斌, 高鹭, 任晓颖. 基于改进YOLOv5的口罩佩戴检测算法[J]. 计算机工程, 2023, 49(8): 265-274.
[2]	李雨阳, 沈记全, 翟海霞, 冯伟华. 基于改进SSD的口罩佩戴检测算法[J]. 计算机工程, 2022, 48(8): 173-179,186.
[3]	叶茂, 马杰, 王倩, 武麟. 多尺度特征融合的轻量化口罩佩戴检测算法[J]. 计算机工程, 2022, 48(7): 42-50.
[4]	彭成, 张乔虹, 唐朝晖, 桂卫华. 基于YOLOv5增强模型的口罩佩戴检测方法研究[J]. 计算机工程, 2022, 48(4): 39-49.
[5]	王艺皓, 丁洪伟, 李波, 杨志军, 杨俊东. 复杂场景下基于改进YOLOv3的口罩佩戴检测算法[J]. 计算机工程, 2020, 46(11): 12-22.

选择文件类型/文献管理软件名称

选择包含的内容