Classification and recognition of unsafe behaviors of tobacco warehouse personnel based on improved YOLOv11

doi:10.16265/j.cnki.issn1003-3033.2025.03.1598

Abstract

Abstract:

To ensure the safety of personnel and property within the storage environment, the traditional YOLOv11 object detection algorithm was improved, and a method and model to identify unsafe behaviors of personnel in the complex environment of tobacco warehouses were proposed. First, a statistical analysis of common unsafe behavior types in tobacco storage was conducted, and the classification of unsafe behaviors of warehouse personnel was explored, including item-related, action-related, and area-related unsafe behaviors. Second, based on the characteristics of unsafe behaviors of warehouse personnel, a dataset augmentation and denoising preprocessing approach was proposed to enhance fine-grained feature extraction, and introduced to improve the saliency mapping of personnel behaviors. Then, the YOLOv11 algorithm was improved through functional enhancement modules and K-means++ anchor box optimization, and a fast detection method for unsafe behaviors of tobacco warehouse personnel was proposed. Finally, the proposed method's effectiveness was validated by comparing with self-built datasets and the open Microsoft COCO dataset. The results show that the method can quickly and effectively identify unsafe behaviors of warehouse personnel, with a significant improvement in recognition accuracy compared to traditional methods(accuracy rate is 94.91% and 88.69% respectively).

Key words: improved YOLOv11, tobacco warehouse personnel, unsafe behaviors, denoising, object detection

CLC Number:

X915.2安全社会学

KE Wei, ZHU Quanjie, CHEN Changmao, WU Chengyi, LIU Yan, ZHANG Yanlin. Classification and recognition of unsafe behaviors of tobacco warehouse personnel based on improved YOLOv11[J]. China Safety Science Journal, 2025, 35(3): 36-44.

Figures/Tables 8

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Table 1

Table 2

Fig.6

References 10

[1]	SREENU G, SALEEM DURAI M A. Intelligent video surveillance: a review through deep learning techniques for crowd analysis[J]. Journal of Big Data, 2019, 6 (1): DOI: 10.1186/s40537-019-0212-5.
[2]	HASSAN M M, UDDIN M D, MOHAMED A, et al. A robust human activity recognition system using smartphone sensors and deep learning[J]. Future Generation Computer Systems, 2018, 81: 307-313.
[3]	冯仕民, 刘忠育, 俞啸, 等. 矿山物联网人员情境描述与不安全行为识别[J]. 物联网学报, 2018, 2(4): 93-98. doi: 10.11959/j.issn.2096-3750.2018.00083
	FENG Shimin, LIU Zhongyu, YU Xiao, et al. Workers' context description and unsafe behavior recognition in internet of things for mines[J]. Chinese Journal on Internet of Things, 2018, 2(4): 93-98.
[4]	谢逸, 张竞文, 李韬, 等. 基于视频监控的地铁施工不安全行为检测预警[J]. 华中科技大学学报:自然科学版, 2019, 47(10): 46-51.
	XIE Yi, ZHANG Jingwen, LI Tao, et al. Research on detection and early warning of unsafe behavior in metro construction based on video monitoring[J]. Journal Huazhong University of Science and Technology: Natural Science Edition, 2019, 47(10): 46-51.
[5]	RAHMANI M, OSKOUEI A N, PETRUDI A M. Experimental and numerical study of the blast wave decrease using sandwich panel by granular materials core[J]. Defence Technology, 2021, 17(5): 1660-1670. doi: 10.1016/j.dt.2020.09.004
[6]	CAO Zhe, HIDALGO M G, SIMON T, et al. OpenPose: realtime multi-person 2D pose estimation using part affinity fields[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(1): 172-186.
[7]	王松, 党建武, 王阳萍, 等. 基于3D运动历史图像和多任务学习的动作识别[J]. 吉林大学学报:工学版, 2020, 50(4): 1495-1502.
	WANG Song, DANG Jianwu, WANG Yangping, et al. Action recognition based on 3D motion history image and multi-task learning[J]. Journal of Jilin University: Engineering and Technology Edition, 2020, 50(4): 1495-1502.
[8]	PATEL V, SINGH R. Real-time detection of unsafe behaviors using LSTM[J]. Journal of Intelligent Manufacturing, 2020, 31: 593-607.
[9]	郭如意, 金杰, 刘高华, 等. 基于双流快速区域卷积神经网络改进的人体动作识别算法[J]. 激光与光电子学进展, 2020, 57(24): 346-350.
	GUO Ruyi, JIN Jie, LIU Gaohua, et al. Improved human action recognition algorithm based on two-stream faster region convolutional neural network[J]. Laser & Optoelectronics Progress, 2020, 57(24): 346-350.
[10]	刘锁兰, 顾嘉晖, 王洪元, 等. 基于关联分区和ST-GCN的人体行为识别[J]. 计算机工程与应用, 2021, 57(13): 168-175. doi: 10.3778/j.issn.1002-8331.2003-0342
	LIU Suolan, GU Jiahui, WANG Hongyuan, et al. Human behavior recognition based on associative partition and ST-GCN[J]. Computer Engineering and Applications, 2021, 57(13): 168-175. doi: 10.3778/j.issn.1002-8331.2003-0342

类型	识别结果	类型	识别结果	类型	识别结果	类型	识别结果
未戴安全帽		区域入侵		跨越传送带		叉车操作
消防区域		攀爬笼车		搬运车操作		攀爬货架

模型	数据集	P/%	R/%	RP/%	F₁/%	检测时间/ms
单次多框目标检测算法(Single Shot multibox Detector,SSD)	COCO数据集	81.78	91.49	74.82	86.36	42
单次多框目标检测算法(Single Shot multibox Detector,SSD)	自建数据集	83.66	97.61	81.66	90.10	43
YOLOv8s	COCO数据集	90.33	95.35	86.13	92.77	25
YOLOv8s	自建数据集	81.18	97.82	79.41	88.73	28
OpenPose-MobileNetV1	COCO数据集	77.48	84.96	65.83	81.05	91
OpenPose-MobileNetV1	自建数据集	80.01	97.47	77.99	87.88	95
Faster RCNN	COCO数据集	90.39	90.2	81.53	90.29	126
Faster RCNN	自建数据集	82.88	99.36	82.35	90.37	117
YOLOv11x	COCO数据集	90.48	89.51	80.99	89.99	50
YOLOv11x	自建数据集	88.51	87.76	77.68	88.13	48
改进YOLOv11	COCO数据集	88.69	98.32	87.20	93.26	52
改进YOLOv11	自建数据集	94.91	95.55	90.69	95.23	56

[1]	JIANG Yuanliang, REN Qingying, REN Yuan, LIU Haipeng, DONG Shaohua. High-consequence area indentation of remote sensing images of China-Myanmar oil and gas pipeline based on improved YOLO model [J]. China Safety Science Journal, 2025, 35(1): 103-111.
[2]	WANG Junwu, HE Juanjuan, SONG Yinghui, LIU Yipeng, CHEN Zhao, GUO Jingyi. Research on early warning for prefabricated building workers' unsafe behaviors of working at height based on RF-SFLA-SVM [J]. China Safety Science Journal, 2024, 34(3): 1-8.
[3]	LI Hujun, YANG Ke, LUO Xun, CHEN Huihua, ZHANG Jianshe, YUAN Dongliang. Effects of tunnel workers' cognitive bias on their unsafe behaviors: mediating role of risk perception [J]. China Safety Science Journal, 2024, 34(10): 24-29.
[4]	SHANG Shuhong, CAO Liang, LU Wei. Fault diagnosis algorithm for medium draining screen based on wavelet transform feature extraction [J]. China Safety Science Journal, 2023, 33(S2): 233-237.
[5]	JIRI Gele, LIU Yao, SHANG Shuhong. Research on fault detection algorithm of rollers of coal conveyer belts [J]. China Safety Science Journal, 2023, 33(S2): 195-201.
[6]	GUO Jie, JING Qinghe, YAN Shouqing, WANG Xin, XIE Miao, WU Yibing. Roller fault diagnosis based on MFCC sound feature signal extraction [J]. China Safety Science Journal, 2023, 33(S2): 116-121.
[7]	JIA Shangfeng, WANG Zhiqiang, HAO Xinyu. Video detection algorithm based on small sample for foreign objects in conveyor belt [J]. China Safety Science Journal, 2023, 33(S1): 134-137.
[8]	ZHAO Dan, SHEN Zhiyuan, SONG Zihao, XIE Li'na, LIU Baichen. Mine airflow speed sensor data cleaning model for intelligent ventilation [J]. China Safety Science Journal, 2023, 33(9): 56-62.
[9]	MA Qinglu, SUN Xiao, TANG Xiaoyao, LU Jiaping, DUAN Xuefeng. Optimization method for tunnel initial fire detection based on YOLOv5s algorithm [J]. China Safety Science Journal, 2023, 33(10): 214-223.
[10]	KANG Mengyue, QI Shenjun, ZHANG Yunbo, CHEN Mei. Intervention mechanism of emotional support on negative emotions of construction workers [J]. China Safety Science Journal, 2023, 33(1): 56-63.
[11]	WANG Lulu, CHEN Xingbang, YUAN Jiacong, WU Song, TONG Ruipeng. Review of research on miners' unsafe behaviors in perspective of occupational psychology [J]. China Safety Science Journal, 2023, 33(1): 48-55.
[12]	MENG Qian. Monitoring system of locomotive disconnector in railway [J]. China Safety Science Journal, 2022, 32(S2): 88-93.
[13]	LI Lingchao, KOU Yuzhuo. Cause analysis and preventive measures of coal mine safety accidents [J]. China Safety Science Journal, 2022, 32(S2): 206-210.
[14]	WANG Rubing, WAN Xin, MAO Peng, LI Dezhi, WANG Xiang. Relevance study between unsafe behaviors of passengers and metro accidents based on fsQCA [J]. China Safety Science Journal, 2020, 30(7): 152-158.
[15]	XIE Bo, SHI Fuqiang, LIAO Xueyan, MA Sheng, YANG Wei, LU Xiangxiang. Slope displacement prediction method based on EEMD-PSO-ELM model [J]. China Safety Science Journal, 2020, 30(3): 157-162.