Intelligent recognition of personnel intrusion into belt conveyor hazardous areas based on an improved YOLOv8 model
-
摘要:
针对煤矿带式输送机场景存在尘雾干扰严重、背景环境复杂、人员尺度多变且易遮挡等因素导致人员入侵危险区域识别准确率不高等问题，提出一种基于改进YOLOv8模型的井下人员入侵带式输送机危险区域智能识别系统。改进YOLOv8模型通过替换主干网络C2f模块为C2fER模块，加强模型的细节特征提取能力，提升模型对小目标人员的识别性能；通过在颈部网络引入特征强化加权双向特征金字塔网络（FE−BiFPN)结构，提高模型的特征融合能力，从而提升模型对多尺度人员目标的识别效果；通过引入分离增强注意力模块(SEAM)增强模型在复杂背景下对局部特征的关注度，提升模型对遮挡目标人员的识别能力；通过引入WIoU损失函数增强训练效果，提升模型识别准确率。消融实验结果表明：改进YOLOv8模型的准确率较基线模型YOLOv8s提升2.3%，mAP@0.5提升3.4%，识别速度为104帧/s。人员识别实验结果表明：与YOLOv10m，YOLOv8s−CA、YOLOv8s−SPDConv和YOLO8n模型相比，改进YOLOv8模型对小目标、多尺度目标、遮挡目标的识别效果均更佳，识别准确率为90.2%，mAP@0.5为87.2%。人员入侵危险区域实验结果表明：井下人员入侵带式输送机危险区域智能识别系统判别人员入侵危险区域的平均准确率为93.25%，满足识别需求。
Abstract:To address challenges such as severe dust and fog interference, complex background environments, and variable personnel scales with frequent occlusions in coal mine belt conveyor scenarios, which resulted in low accuracy in recognizing personnel intrusions into hazardous areas, an intelligent recognition system based on an improved YOLOv8 model was proposed. The improved YOLOv8 model enhanced detailed feature extraction by replacing the C2f module in the backbone network with the C2fER module, which improved recognition performance for small targets. The Feature Enhancement Weighted Bi-Directional Feature Pyramid Network (FE-BiFPN) structure was introduced into the neck network to strengthen feature fusion capabilities, thereby enhancing recognition of multi-scale personnel targets. The Separated and Enhancement Attention Module (SEAM) was incorporated to improve the model's attention to local features in complex backgrounds, which boosted its ability to recognize occluded personnel targets. Furthermore, the WIoU loss function was applied to enhance training outcomes, improving recognition accuracy. Ablation experiment results showed that the improved YOLOv8 model achieved a 2.3% increase in accuracy and a 3.4% improvement in mAP@0.5 compared to the baseline YOLOv8s model, with a recognition speed of 104 frames per second. Personnel recognition experiments demonstrated that, compared to YOLOv10m, YOLOv8s-CA, YOLOv8s-SPDConv, and YOLOv8n models, the improved YOLOv8 model delivered superior recognition performance for small, multi-scale, and occluded targets, achieving a recognition accuracy of 90.2% and an mAP@0.5 of 87.2%. Personnel intrusion experiments revealed that the intelligent recognition system achieved an average accuracy of 93.25% in identifying personnel intrusions into belt conveyor hazardous areas, satisfying recognition requirements.
-
-
![]()
图 1 井下人员入侵带式输送机危险区域智能识别系统结构
Figure 1. Structure of intelligent recognition system for personnel intrusion into belt conveyors hazardous areas
![]()
图 8 基于射线法的人员入侵危险区域判别原理
Figure 8. Principle of personnel intrusion into hazardous areas detection based on ray casting method
![]()
图 10 图像去雾处理前后效果对比
Figure 10. Comparison of dehazing effects before and after processing
表 1 模型训练参数
Table 1 Model training parameters
参数 值 训练轮次（epochs） 200 批次（batch） 32 学习率（lr） 0.01 优化器（optimizer） SGD 正则化系数（weight_decay） 0.000 5 
下载: 导出CSV
表 2 改进YOLOv8模型消融实验结果
Table 2 Ablation experiment results for improved YOLOv8 model
C2fER FE−
BiFPNSEAM WIoU p/% R/% mAP@
0.5/%帧率/
（帧·s−1）× × × × 87.9 76.3 83.8 178 √ × × × 90.3 77.2 85.4 133 × √ × × 88.6 78.6 86.6 178 × × √ × 90.2 77.3 86.1 151 × × × √ 89.2 78.5 86.0 196 √ √ × × 90.4 77.8 85.9 116 √ √ √ × 89.9 79.1 86.4 90 √ √ √ √ 90.2 79.9 87.2 104
下载: 导出CSV
表 3 改进YOLOv8模型与其他模型对比实验结果
Table 3 Comparison of experimental results between improved YOLOv8 model and other models
模型 p/% mAP@0.5/% 帧率/（帧·s−1） YOLOv10m 88.6 82.5 125 YOLOv8s−CA 87.4 86.1 172 YOLOv8s−SPDConv 88.5 84.2 185 YOLOv8n 86.9 83.3 208 改进YOLOv8 90.2 87.2 104
下载: 导出CSV
表 4 人员入侵检测结果统计
Table 4 Statistics of personnel intrusion detection results
组别 入侵人员数量 识别入侵人员数量 准确率/% 组1 2 967 2 882 97.14 组2 1 598 1 519 95.05 组3 201 176 87.56
下载: 导出CSV
-
[1] 王国法,张良,李首滨,等. 煤矿无人化智能开采系统理论与技术研发进展[J]. 煤炭学报,2023,48(1):34-53. WANG Guofa,ZHANG Liang,LI Shoubin,et al. Progresses in theory and technological development of unmanned smart mining system[J]. Journal of China Coal Society,2023,48(1):34-53.
[2] 王国法,杜毅博,陈晓晶,等. 从煤矿机械化到自动化和智能化的发展与创新实践——纪念《工矿自动化》创刊50周年[J]. 工矿自动化,2023,49(6):1-18. WANG Guofa,DU Yibo,CHEN Xiaojing,et al. Development and innovative practice from coal mine mechanization to automation and intelligence:commemorating the 50th anniversary of the founding of Journal of Mine Automation[J]. Journal of Mine Automation,2023,49(6):1-18.
[3] 毛清华,郭文瑾,翟姣,等. 煤矿带式输送机异常状态视频AI识别技术研究[J]. 工矿自动化,2023,49(9):36-46. MAO Qinghua,GUO Wenjin,ZHAI Jiao,et al. Research on video AI recognition technology for abnormal state of coal mine belt conveyors[J]. Journal of Mine Automation,2023,49(9):36-46.
[4] CHAN K,LOUIS J,ALBERT A. Incorporating worker awareness in the generation of hazard proximity warnings[J]. Sensors,2020,20(3). DOI: 10.3390/s20030806.
[5] HUANG Yusheng,HAMMAD A,ZHU Z. Providing proximity alerts to workers on construction site-s using Bluetooth Low Energy RTLS[J]. Automation in Construction,2021,132. DOI: 10.1016/J.AUTCON.2021.103928.
[6] TANG Qiang. Engineering design of electronic fence system based on intelligent monitoring and wireless local area network[J]. Alexandria Engineering Journal,2022,61(4):2959-2969. DOI: 10.1016/j.aej.2021.08.019
[7] 吴浩楠,史宏,王瑞,等. 基于改进YOLOv8的铁路人员入侵检测方法研究[J/OL]. 铁道科学与工程学报:1-12[2024-09-02]. https://doi.org/10.19713/j.cnki.43-1423/u.T20241118. WU Haonan,SHI Hong,WANG Rui,et al. Research on railway intrusion detection method based on improved YOLOv8[J/OL]. Journal of Railway Science and Engineering:1-12[2024-09-02]. https://doi.org/10.19713/j.cnki.43-1423/u.T20241118.
[8] MEI Xinyu,ZHOU Xiaojing,XU Feng,et al. Human intrusion detection in static hazardous areas at construction sites:deep learning–based method[J]. Journal of Construction Engineering and Management,2023,149(1). DOI: 10.1061/(ASCE)CO.1943-7862.0002409.
[9] PARK J,CHEN Jingdao,CHO Y K,et al. CNN-based person detection using infrared images for night-time intrusion warning systems[J]. Sensors,2019,20(1). DOI: 10.3390/s20010034.
[10] CAO Zhiwei,QIN Yong,XIE Zhengyu,et al. An effective railway intrusion detection method using dynamic intrusion region and lightweight neural network[J]. Measurement,2022,191. DOI: 10.1016/J.MEASUREMENT.2021.110564.
[11] 邵小强,李鑫,杨涛,等. 改进YOLOv5s和DeepSORT的井下人员检测及跟踪算法[J]. 煤炭科学技术,2023,51(10):291-301. SHAO Xiaoqiang,LI Xin,YANG Tao,et al. Underground personnel detection and tracking based on improved YOLOv5s and DeepSORT[J]. Coal Science and Technology,2023,51(10):291-301.
[12] ZHU Ji,YANG Hua,WANG Jia,et al. Description-based person search with multi-grained matching networks[J]. Displays,2021,69. DOI: 10.1016/J.DISPLA.2021.102039.
[13] 戚玲珑,高建瓴. 基于改进YOLOv7的小目标检测[J]. 计算机工程,2023,49(1):41-48. QI Linglong,GAO Jianling. Small object detection based on improved YOLOv7[J]. Computer Engineering,2023,49(1):41-48.
[14] 张秀再,邱野,沈涛. 基于改进SSD算法的地铁场景小行人目标检测[J/OL]. 计算机研究与发展:1-13[2024-09-02]. http://kns.cnki.net/kcms/detail/11.1777.TP.20240603.1111.011.html. ZHANG Xiuzai,QIU Ye,SHEN Tao,Small pedestrian target detection in subway scene based on improved SSD algorithm[J/OL]. Journal of Computer Research and Development:1-13[2024-09-02]. http://kns.cnki.net/kcms/detail/11.1777.TP.20240603.1111.011.html.
[15] HAN Zhixiong,MA Bingpeng. Enhancing identification for person search with multi-scale multi-grained representation learning[J]. Pattern Recognition,2024,150. DOI:10.1016/J.PATCOG. 2024.110361.
[16] 陈伟,江志成,田子建,等. 基于YOLOv8的煤矿井下人员不安全动作检测算法[J/OL]. 煤炭科学技术:1-19[2024-09-02]. http://kns.cnki.net/kcms/detail/11.2402.td.20240322.1343.003.html. CHEN Wei,JIANG Zhicheng,TIAN Zijian,et al. Unsafe action detection algorithm of underground personnel in coal mine based on YOLOv8[J/OL]. Coal Science and Technology:1-19[2024-09-02]. http://kns.cnki.net/kcms/detail/11.2402.td.20240322.1343.003.html.
[17] MAO Qinghua,WANG Yufei,ZHANG Xuhui,et al. Clarity method of fog and dust image in fully mechanized mining face[J]. Machine Vision and Applications,2022,33(2). DOI: 10.1007/S00138-022-01282-1.
[18] VARGHESE R,SAMBATH M. YOLOv8:a novel object detection algorithm with enhanced performance and robustness[C]. 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems ,2024. DOI: 10.1109/ADICS58448.2024.10533619.
[19] ZHANG Xin,LIU Chen,YANG Degang,et al. RFAConv:innovating spatial attention and standard convolutional operation[EB/OL]. [2024-10-20]. https://arxiv.org/abs/2304.03198.
[20] OUYANG Daliang,HE Su,ZHANG Guozhong,et al. Efficient multi-scale attention module with cross-spatial learning[C]. 2023 IEEE International Conference on Acoustics,Speech and Signal Processing,Rhodes Island,2023. DOI: 10.1109/ICASSP49357.2023.10096516.
[21] LIU Shu,QI Lu,QIN Haifang,et al. Path aggregation network for instance segmentation[C]. IEEE Conference on Computer Vision and Pattern Recognition,Salt Lake,2018:8759-8768.
[22] TAN Mingxing,PANG Ruoming,LE Q V. EfficientDet:scalable and efficient object detection[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition,Seattle,2020:10781-10790.
[23] YU Ziping,HUANG Hongbo,CHEN Weijun,et al. YOLO-FaceV2:a scale and occlusion aware face detector[J]. Pattern Recognition,2024,155. DOI: 10.1016/J.PATCOG.2024.110714.
[24] TONG Zanjia,CHEN Yuhang,XU Zewei,et al. Wise-IoU:bounding box regression loss with dynamic focusing mechanism[EB/OL]. [2024-10-20]. https://arxiv.org/abs/2301.10051v3.
[25] 王锴欣,黄丹,于永兴,等. 基于CRS−YOLO算法的高密度柔性封装基板缺陷检测方法[J]. 激光杂志,2024,45(5):41-47. WANG Kaixin,HUANG Dan,YU Yongxing,et al. High density flexible packaging substrate defect detection method based on CRS-YOLO algorithm[J]. Laser Journal,2024,45(5):41-47.
[26] 赵小虎,张狄,谢礼逊,等. 基于改进YOLOv8的煤矿皮带异物检测方法[J/OL]. 工程科学与技术:1-16[2024-09-02]. http://kns.cnki.net/kcms/detail/51.1773.tb.20250114.1313.001.html. ZHAO Xiaohu,ZHANG Di,XIE Lixun,et al. A foreign body detection method for coal mine belt based on improved YOLOv8[J/OL]. Advanced Engineering Sciences:1-16[2024-09-02]. http://kns.cnki.net/kcms/detail/51.1773.tb.20250114.1313.001.html.
计量
- 文章访问数: 143
- HTML全文浏览量: 27
- PDF下载量: 31
