基于改进YOLOv8n的煤矿井下钻杆计数方法

姜媛媛; 刘宋波

doi:10.13272/j.issn.1671-251x.2024040073

基于改进YOLOv8n的煤矿井下钻杆计数方法

doi: 10.13272/j.issn.1671-251x.2024040073

姜媛媛^{1, 2,},
刘宋波^1, ,

1.
安徽理工大学电气与信息工程学院，安徽淮南　232001
2.
安徽理工大学环境友好材料与职业健康研究院（芜湖），安徽芜湖　241003

基金项目: 安徽省重点研究与开发计划项目（202104g01020012）；安徽理工大学环境友好材料与职业健康研究院研发专项基金资助项目（ALW2020YF18）。

详细信息

作者简介:
姜媛媛（1982—），女，安徽颍上人，教授，博士，主要研究方向为人工智能、机器学习，E-mail:jyyll672@163.com

通讯作者:
刘宋波（1999—），男，安徽六安人，硕士研究生，主要研究方向为计算机视觉图像处理，E-mail：1186784769@qq.com。

中图分类号: TD67
计量
- 文章访问数: 104
- HTML全文浏览量: 32
- PDF下载量: 4
- 被引次数: 0
出版历程
- 收稿日期: 2024-04-22
- 修回日期: 2024-08-30
- 网络出版日期: 2024-08-22

A coal mine underground drill pipes counting method based on improved YOLOv8n

JIANG Yuanyuan^{1, 2
,},
LIU Songbo^{1
, ,}

1.
School of Electrical and Information Engineering, Anhui University of Science and Technology, Huainan 232001, China
2.
Institute of Environment-friendly Meterials and Occupational Health, Anhui University of Science and Technology, Wuhu 241003, China

摘要

摘要: 为提高煤矿井下钻杆计数的效率和精度，提出了一种基于改进YOLOv8n模型的煤矿井下钻杆计数方法。建立了YOLOv8n−TBiD模型，该模型可准确检测矿井钻机工作视频中的钻杆并进行有效分割：为有效捕获钻杆的边界信息，提高模型对钻杆形状识别的精度，使用加权双向特征金字塔网络（BiFPN）替换路径聚合网络（PANet）；针对钻杆易与昏暗的矿井环境混淆的问题，在Backbone网络的SPPF模块后添加三分支注意力（Triplet Attention），以增强模型抑制背景干扰的能力；针对钻杆在图像中占比小、背景信息繁杂的问题，采用Dice损失函数替换 CIoU损失函数来优化模型对目标钻杆的分割处理。利用YOLOv8n−TBiD模型分割出的钻杆及其掩码信息，根据打钻过程中钻杆掩码面积变小而装新钻杆时钻杆掩码面积突然增大的规律，设计了一种钻杆计数算法。选取综采工作面实际采集的钻机工作视频对基于YOLOv8n−TBiD模型的钻杆计数方法进行了实验验证，结果表明：① YOLOv8n−TBiD模型检测钻杆的平均精度均值达94.9%，与对比模型GCI−YOLOv4，ECO−HC，P−MobileNetV2，YOLOv5，YOLOX相比，检测准确率分别提升了4.3%，7.5%，2.1%，6.3%，5.8%，检测速度较原始YOLOv8n模型提升了17.8%。② 所提钻杆计数算法在不同煤矿井下环境的视频数据集上实现了99.3%的钻杆计数精度。
- 矿井钻机 /
- 钻杆计数 /
- YOLOv8n−TBiD /
- BiFPN /
- Triplet Attention /
- Dice损失函数 /
- 钻杆掩码 /
- 图像分割
Abstract: In order to improve the efficiency and precision of underground drill pipe counting in coal mines, a coal mine underground drill pipe counting method based on the improved YOLOv8n model is proposed. The YOLOv8n-TbiD is established.The model can accurately detects and segments drill pipes in mine drilling rig working videos. The main improvements include the following points. In order to effectively capture the boundary information of drill rods and improve the precision of the model in recognizing drill rod shapes, the weighted bidirectional feature pyramid network (BiFPN) is used instead of the path aggregation network (PANet). To address the issue of drill pipe objects being easily confused with dim mine environments, Triplet Attention is added to the SPPF module of the Backbone network to enhance the model's capability to suppress background interference. In response to the small proportion of drill pipes in the image and the complexity of background information, the Dice loss function is used to replace CIoU loss function to optimize the segmentation processing of drill pipe objects in the model. The method uses the YOLOv8n-TBiD model to segment the drill pipe and its mask information. A drill pipe counting algorithm is designed based on the rule that the mask area of the drill pipe decreases during drilling and suddenly increases when a new drill pipe is installed. The working video of the drilling rig in the fully mechanized working face is selected, in order to conduct experimental verification of drill pipes counting method based on YOLOv8n-TBiD model. The experimental results show that the mean average precision of the YOLOv8n-TBiD model for detecting drill pipes reaches 94.9%. Compared with the comparative experimental models GCI-YOLOv4, ECO-HC, P-MobileNetV2, YOLOv5, and YOLOX, the accuracy increases by 4.3%, 7.5%, 2.1%, 6.3%, and 5.8%, respectively, and the detection speed increases by 17.8% compared to the original YOLOv8n model. The proposed drill pipe counting algorithm achieves precision of 99.3% on video datasets from different underground coal mine environments.
- mine drilling rig /
- drill pipe counting /
- YOLOv8n-TBiD /
- BiFPN /
- Triplet Attention /
- Dice loss function /
- mask of drill pipe /
- image segmentation

HTML全文

图 1 YOLOv8n−TBiD网络结构

Figure 1. YOLOv8n-TBiD network structure

下载: 全尺寸图片幻灯片

图 2 PANet与BiFPN结构对比

Figure 2. Comparison of structure of path aggregation network（PANet） and bi-directional feature pyramid network（BiFPN）

下载: 全尺寸图片幻灯片

图 3 Triplet Attention网络结构

Figure 3. Triplet attention network structure

下载: 全尺寸图片幻灯片

图 4 Triplet Attention分支网络结构

Figure 4. Triplet attention branch network structure

下载: 全尺寸图片幻灯片

图 5 掩码面积曲线二值化滤波

Figure 5. Binarization filtering of mask area curve

下载: 全尺寸图片幻灯片

图 6 数据集中部分钻杆标注

Figure 6. Part of drill pipe annotation in data set

下载: 全尺寸图片幻灯片

图 7 不同模型在训练集的损失曲线对比

Figure 7. Comparison of loss curves of different models in training sets

下载: 全尺寸图片幻灯片

图 8 不同模型在验证集的损失曲线对比

Figure 8. Comparison of loss curves of different models in validation sets

下载: 全尺寸图片幻灯片

图 9 不同场景下钻杆检测结果对比

Figure 9. Comparison of drill pipe detection results in different scenarios

下载: 全尺寸图片幻灯片

表 1 消融实验结果

Table 1. Ablation experiment results

模型	BiFPN	Triplet Attention	Dice	mPA/%	mIoU/%	参数量/10⁶个	浮点运算数/10⁹	权重大小/MiB	帧率/（帧·s⁻¹）
YOLOv8n	×	×	×	89.2	81.1	3.4	12.8	6.4	90
YOLOv8n−Bi	√	×	×	92.5	85.3	2.3	11.7	4.4	108
YOLOv8n−T	×	√	×	91.9	84.7	3.4	12.8	6.5	87
YOLOv8n−D	×	×	√	90.2	83.6	3.4	12.8	6.5	88
YOLOv8n−TBiD	√	√	√	94.9	87.3	2.3	11.7	4.5	106

下载: 导出CSV

表 2 不同模型钻杆检测结果对比

Table 2. Comparison of drill pipe detection results by different models

模型	mAP/%
GCI−YOLOv4	90.6
ECO−HC	87.4
P−MobileNetV2	92.8
YOLOv5	88.6
YOLOX	89.1
YOLOv8n−TBiD	94.9

下载: 导出CSV

表 3 不同计数方法实验结果

Table 3. Experimental results of different counting methods

方法	实际钻杆数量/个	检测钻杆数量/个	准确率/%
人工计数方法	420	410	97.6
文献[4]方法	420	411	97.8
文献[8]方法	420	410	97.6
文献[9]方法	420	413	98.1
本文计数方法	420	417	99.3

下载: 导出CSV

参考文献(21)

[1]	梁运培,郑梦浩,李全贵,等. 我国煤与瓦斯突出预测与预警研究现状[J]. 煤炭学报,2023,48(8):2976-2994. LIANG Yunpei,ZHENG Menghao,LI Quangui,et al. A review on prediction and early warning methods of coal and gas outburst[J]. Journal of China Coal Society,2023,48(8):2976-2994.
[2]	PAN Xiaokang,CHENG Hao,CHEN Jie,et al. An experimental study of the mechanism of coal and gas outbursts in the tectonic regions[J]. Engineering Geology,2020,279. DOI: 10.1016/j.enggeo.2020.105883.
[3]	姚超修,胡亚磊. 基于视频识别的煤矿井下钻杆计数算法[J]. 煤炭技术,2023,42(8):203-206. YAO Chaoxiu,HU Yalei. Drilling pipe counting algorithm based on video analysis in coal mine[J]. Coal Technology,2023,42(8):203-206.
[4]	张栋,姜媛媛. 融合注意力机制与逆残差结构的轻量级钻机目标检测方法[J]. 电子测量与仪器学报,2022,36(11):201-210. ZHANG Dong,JIANG Yuanyuan. Lightweight target detection method of drilling rig based on attention mechanism and inverse residual structure[J]. Journal of Electronic Measurement and Instrumentation,2022,36(11):201-210.
[5]	胡少兵,罗明璋,程峰,等. 基于应力波频谱图的护栏金属立柱埋深检测法[J]. 公路,2022,67(6):336-341. HU Shaobing,LUO Mingzhang,CHENG Feng,et al. Method of detecting the buried depth of guardrail metal column based on stress wave spectrum image[J]. Highway,2022,67(6):336-341.
[6]	徐钊,房咪咪,周红伟,等. 基于电驻波的锚杆长度无损测量方法[J]. 工矿自动化,2013,39(9):112-115. XU Zhao,FANG Mimi,ZHOU Hongwei,et al. Non-destructive measurement method of anchor stock length based on electricity standing wave[J]. Industry and Mine Automation,2013,39(9):112-115.
[7]	李泽鹏. 煤矿视频监控系统智能化升级及应用[J]. 自动化应用,2024(3):226-228. LI Zepeng. Intelligent upgrade and application of coal mine video monitoring system[J]. Automation Application,2024(3):226-228.
[8]	方杰,李振璧,夏亮. 基于ECO−HC的钻杆计数方法[J]. 煤炭技术,2021,40(11):186-189. FANG Jie,LI Zhenbi,XIA Liang. Drill pipe counting method based on ECO-HC[J]. Coal Technology,2021,40(11):186-189.
[9]	张栋,姜媛媛. 基于改进MobileNetV2的钻杆计数方法[J]. 工矿自动化,2022,48(10):69-75. ZHANG Dong,JIANG Yuanyuan. Drill pipe counting method based on improved MobileNetV2[J]. Journal of Mine Automation,2022,48(10):69-75.
[10]	杜京义,党梦珂,乔磊,等. 基于改进时空图卷积神经网络的钻杆计数方法[J]. 工矿自动化,2023,49(1):90-98. DU Jingyi,DANG Mengke,QIAO Lei,et al. Drill pipe counting method based on improved spatial-temporal graph convolution neural network[J]. Journal of Mine Automation,2023,49(1):90-98.
[11]	TAN Mingxing,PANG Ruoming,LE Q V. EfficientDet:scalable and efficient object detection[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition,Seattle,2020:10778-10787.
[12]	卢子册,刘小芳,王德伟. 基于改进YOLOv8的PCB焊点语义分割方法[J]. 无线电工程,2024,54(7):1614-1621. LU Zice,LIU Xiaofang,WANG Dewei. Semantic segmentation method for PCB solder joint based on improved YOLOv8[J]. Radio Engineering,2024,54(7):1614-1621.
[13]	熊恩杰,张荣芬,刘宇红,等. 面向交通标志的Ghost−YOLOv8检测算法[J]. 计算机工程与应用,2023,59(20):200-207. XIONG Enjie,ZHANG Rongfen,LIU Yuhong,et al. Ghost-YOLOv8 detection algorithm for traffic signs[J]. Computer Engineering and Applications,2023,59(20):200-207.
[14]	LI Taiguo,ZHANG Yingzhi,LI Quanqin,et al. AB-DLM:an improved deep learning model based on attention mechanism and BiFPN for driver distraction behavior detection[J]. IEEE Access,2022,10:83138-83151. doi: 10.1109/ACCESS.2022.3197146
[15]	吴慧海,沈文忠. 基于TA−YOLO的电力设备红外图像检测方法[J]. 信息技术与信息化,2022(3):17-20. WU Huihai,SHEN Wenzhong. Infrared image detection method of power equipment based on TA-YOLO[J]. Information Technology and Informatization,2022(3):17-20.
[16]	郑兆晖. 基于几何因子的目标检测与实例分割的研究[D]. 天津:天津大学,2021. ZHENG Zhaohui. Research on object detection and instance segmentation based on geometric factors[D]. Tianjin:Tianjin University,2021.
[17]	黄文博,屈超凡,燕杨. 融合注意力机制的TransGLnet脉络膜自动分割[J]. 光学精密工程,2023,31(23):3482-3489. HUANG Wenbo,QU Chaofan,YAN Yang. Automatic segmentation of choroid by TransGLnet integrating attention mechanism[J]. Optics and Precision Engineering,2023,31(23):3482-3489.
[18]	于营,王春平,付强,等. 语义分割评价指标和评价方法综述[J]. 计算机工程与应用,2023,59(6):57-69. YU Ying,WANG Chunping,FU Qiang,et al. Survey of evaluation metrics and methods for semantic segmentation[J]. Computer Engineering and Applications,2023,59(6):57-69.
[19]	崔多,王秋生. 基于深度学习的无人机引导线识别模型[J/OL]. 计算机应用:1-7[2024-04-26]. https://kns.cnki.net/kcms/detail/51.1307.TP.20240424.1452.004.html. CUI Duo,WANG Qiusheng. Drone guide line recognition model based on deep learning[J/OL]. Journal of Computer Applications:1-7[2024-04-26]. https://kns.cnki.net/kcms/detail/51.1307.TP.20240424.1452.004.html.
[20]	马超伟,张浩,马新明,等. 基于改进YOLOv8的轻量化小麦病害检测方法[J]. 农业工程学报,2024,40(5):187-195. MA Chaowei,ZHANG Hao,MA Xinming,et al. Method for the lightweight detection of wheat disease using improved YOLOv8[J]. Transactions of the Chinese Society of Agricultural Engineering,2024,40(5):187-195.
[21]	GE Zheng,LIU Songtao,WANG Feng,et al. YOLOx:exceeding YOLO series in 2021[EB/OL]. [2024-03-20]. https://arxiv.org/abs/2107.08430.