基于改进YOLOv8的煤矿输送带异物检测

洪炎; 汪磊; 苏静明; 汪瀚涛; 李木石

doi:10.13272/j.issn.1671-251x.2024050006

基于改进YOLOv8的煤矿输送带异物检测

doi: 10.13272/j.issn.1671-251x.2024050006

安徽理工大学电气与信息工程学院，安徽淮南　232001

基金项目: 国家重点研发计划项目（2021YFD2000204）；国家自然科学基金项目（12304236，32301688，52174141）；安徽数字农业工程技术研究中心开放项目（AHSZNYGCZXKF021）；大学生创新创业基金项目（202210361053，202310361037）；安徽理工大学研究生创新基金项目（2024cx2067）。

详细信息

作者简介:
洪炎（1979—），男，重庆万州人，教授，博士，主要研究方向为图像处理与物联网，E-mail：hong5212724@163.com

通讯作者:
汪磊（1997—），男，安徽六安人，硕士研究生，主要研究方向为深度学习与嵌入式系统、目标检测，E-mail：wljy2023@163.com。

中图分类号: TD634.1
计量
- 文章访问数: 220
- HTML全文浏览量: 32
- PDF下载量: 44
- 被引次数: 0
出版历程
- 收稿日期: 2024-05-06
- 修回日期: 2024-05-25
- 网络出版日期: 2024-07-10

Foreign object detection of coal mine conveyor belt based on improved YOLOv8

School of Electrical and Information Engineering, Anhui University of Science and Technology, Huainan 232001, China

摘要

摘要: 现有基于深度学习的输送带异物检测模型较大，难以在边缘设备部署，且对不同尺寸异物和小目标异物存在错检、漏检情况。针对上述问题，提出一种基于改进YOLOv8的煤矿输送带异物检测方法。采用深度可分离卷积、压缩和激励(SE)网络将YOLOv8主干网络中C2f模块的Bottleneck重新构建为DSBlock，在保持模型轻量化的同时提升检测性能；为增强对不同尺寸目标物体信息的获取能力，引入高效通道注意力(ECA) 机制，并对ECA的输入层进行自适应平均池化和自适应最大池化操作，得到跨通道交互MECA模块，以增强模块的全局视觉信息，进一步提升异物识别精度；将YOLOv8的3个检测头修改为4个轻量化小目标检测头，以增强对小目标的敏感性，有效降低小目标异物的漏检率和错检率。实验结果表明：改进YOLOv8的精确度达91.69%，mAP@50达92.27%，较YOLOv8分别提升了3.09%和4.07%；改进YOLOv8的检测速度达73.92帧/s，可充分满足煤矿输送带异物实时检测的需求；改进YOLOv8的精确度、mAP@50、参数量、权重大小和每秒浮点运算数均优于SSD，Faster-RCNN，YOLOv5，YOLOv7−tiny等主流目标检测算法。
- 输送带异物检测 /
- YOLOv8 /
- SE网络 /
- 高效通道注意力机制 /
- 轻量化 /
- 小目标检测 /
- 自适应平均池化 /
- 自适应最大池化
Abstract: The existing deep learning based foreign object detection models for conveyor belts are relatively large and difficult to deploy on edge devices. There are errors and omissions in detecting foreign objects of different sizes and small objects. In order to solve the above problems, a foreign object detection method for coal mine conveyor belts based on improved YOLOv8 is proposed. The depthwise separable convolution, squeeze-and-excitation (SE) networks are used to reconstruct the Bottleneck of the C2f module in the YOLOv8 backbone network as a DSBlock, which improves the detection performance while keeping the model lightweight. To enhance the capability to obtain information from objects of different sizes, an efficient channel attention (ECA) mechanism is introduced. The input layer of ECA is subjected to adaptive average pooling and adaptive maximum pooling operations to obtain a cross channel interactive MECA module, which enhances the global visual information of the module and further improves the precision of foreign object recognition. The method modifies the 3 detection heads of YOLOv8 to 4 lightweight small object detection heads to enhance sensitivity to small objects and effectively reduce the missed and false detection rates of small object foreign objects. The experimental results show that the improved YOLOv8 achieves a precision of 91.69%, mAP@50 reached 92.27%, an increase of 3.09% and 4.07% respectively compared to YOLOv8. The detection speed of improved YOLOv8 reaches 73.92 frames/s, which can fully meet the demand for real-time detection of foreign objects on conveyor belts in coal mines. The improved YOLOv8 outperforms mainstream object detection algorithms such as SSD, Faster-RCNN, YOLOv5, and YOLOv7-tiny in terms of precision, mAP@50, number of parameters, weight size, and number of floating point operations.
- foreign object detection on conveyor belts /
- YOLOv8 /
- SE network /
- efficient channel attention mechanism /
- lightweight /
- small object detection /
- adaptive average pooling /
- adaptive maximum pooling

HTML全文

图 1 YOLOv8网络结构

Figure 1. YOLOv8 network structure

下载: 全尺寸图片幻灯片

图 2 CED−YOLO网络结构

Figure 2. CED-YOLO network structure

下载: 全尺寸图片幻灯片

图 3 数据增强流程

Figure 3. Data enhancement process

下载: 全尺寸图片幻灯片

图 4 SC2f模块结构

Figure 4. SC2f module structure

下载: 全尺寸图片幻灯片

图 5 DSBlock模块结构

Figure 5. DSBlock module structure

下载: 全尺寸图片幻灯片

图 6 MECA模块结构

Figure 6. MECA module structure

下载: 全尺寸图片幻灯片

图 7 改进后检测层

Figure 7. Improved detection layer

下载: 全尺寸图片幻灯片

图 8 改进后轻量化检测头

Figure 8. Improved lightweight detection head

下载: 全尺寸图片幻灯片

图 9 数据集图像

Figure 9. Dataset images

下载: 全尺寸图片幻灯片

图 10 模型性能指标

Figure 10. Model performance indicators

下载: 全尺寸图片幻灯片

图 11 不同模型的识别结果

Figure 11. Recognition results of different models

下载: 全尺寸图片幻灯片

图 12 模型改进前后热力图对比

Figure 12. Comparison of heat maps before and after model improvement

下载: 全尺寸图片幻灯片

表 1 实验硬件配置

Table 1. Experimental hardware configuration

实验环境	配置
操作系统	Windows 10
CPU	Intel（R） Core（TM）i5−13490F CPU@2.50 GHz
GPU	NVIDIA GeForce GTX 4060（8 G）
深度学习框架	PyTorch 1.9.1+CUDA 11.1+CUDNN 8.0.5
编译器	Python 3.8.18
内存	32 GiB

下载: 导出CSV

表 2 消融实验结果

Table 2. Ablation experiment results

序号	A	B	C	D	E	精确度/%	召回率/%	mAP@50/%	mAP@50∶95/%	参数量/ 10⁶个	权重大小/ MiB	每秒浮点运算数/10⁹	速度/ （帧·s⁻¹）
1	×	×	×	×	×	88.60	80.19	88.20	56.70	3.00	6.3	8.1	162.55
2	√	×	×	×	×	89.25	83.09	89.36	58.62	3.00	6.3	8.1	163.23
3	√	√	×	×	×	89.45	87.98	92.32	59.94	2.68	5.5	6.9	156.68
4	√	√	√	×	×	89.02	86.26	92.96	62.21	2.79	5.9	11.7	94.49
5	√	√	√	√	×	92.03	84.30	91.92	60.89	2.79	5.9	11.7	111.26
6	√	√	√	√	√	91.69	83.25	92.27	61.59	2.34	5.0	6.2	73.92

下载: 导出CSV

表 3 主流算法对比结果

Table 3. Comparison results of mainstream algorithms

算法	精确度/%	mAP@50/%	参数量/ 10⁶个	权重大小/ MiB	每秒浮点运算数/10⁹
YOLOv3	87.54	89.06	12.12	24.4	18.9
YOLOv5	89.38	88.52	2.50	5.3	7.1
YOLOv7−tiny	84.40	89.70	6.01	12.3	13
YOLOv8	88.60	88.20	3.00	6.3	8.1
Faster−RCNN	66.13	55.09	136.73	108.0	401.7
SSD	74.05	65.20	23.87	91.09	274.0
文献[22]中算法	81.40	89.30	6.87	14.1	14.2
文献[24]中算法	90.60	89.60	1.92	4.1	4.7
CED−YOLO	91.69	92.27	2.34	5.0	6.2

下载: 导出CSV

参考文献(25)

[1]	中矿(北京)煤炭产业景气指数研究课题组,郭建利. 2023-2024年中国煤炭产业经济形势研究报告[J]. 中国煤炭,2024,50(3):12-20. China Mining (Beijing) Coal Industry Prosperity Index Research,GUO Jianli. Research report on the economic situation of China's coal industry from 2023 to 2024[J]. China Coal,2024,50(3):12-20.
[2]	REDMON J,DIVVALA S,GIRSHICK R,et al. You only look once:unified,real-time object detection[C]. IEEE Conference on Computer Vision and Pattern Recognition,Las Vegas,2016:779-788.
[3]	LIU Wei,ANGUELOV D,ERHAN D,et al. SSD:single shot multiBox detector[C]. The 14th European Conference on Computer Vision,Amsterdam,2016:21-37.
[4]	LIN T Y,GOYAL P,GIRSHICK R,et al. Focal loss for dense object detection [C]. IEEE International Conference on Computer Vision,Venice,2017:2999-3007.
[5]	REN Shaoqing,HE Kaiming,GIRSHICK R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149. doi: 10.1109/TPAMI.2016.2577031
[6]	HAO Zhenbang,LIN Lili,POST CHRISTOPHER J,et al. Automated tree-crown and height detection in a young forest plantation using mask region-based convolutional neural network (Mask R-CNN)[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2021,178:112-123. doi: 10.1016/j.isprsjprs.2021.06.003
[7]	刘富强,钱建生,王新红,等. 基于图像处理与识别技术的煤矿矸石自动分选[J]. 煤炭学报,2000,25(5):534-537. doi: 10.3321/j.issn:0253-9993.2000.05.020 LIU Fuqiang,QIAN Jiansheng,WANG Xinhong,et al. Automatic separation of waste rock in coal mine based on image procession and recognition[J]. Journal of China Coal Society,2000,25(5):534-537. doi: 10.3321/j.issn:0253-9993.2000.05.020
[8]	WANG Yuanbin,WANG Yujing,DANG Langfei. Video detection of foreign objects on the surface of belt conveyor underground coal mine based on improved SSD[J]. Journal of Ambient Intelligence and Humanized Computing,2020:1-10.
[9]	任国强,韩洪勇,李成江,等. 基于Fast_YOLOv3算法的煤矿胶带运输异物检测[J]. 工矿自动化,2021,47(12):128-133. REN Guoqiang,HAN Hongyong,LI Chengjiang,et al. Foreign object detection in coal mine belt transportation based on Fast_YOLOv3 algorithm[J]. Industry and Mine Automation,2021,47(12):128-133.
[10]	XIE Yehui,YU Sun,HUANG Ziyang. Foreign matter detection of coal conveying belt based on machine vision[C]. The 2nd International Conference on Computer Science and Management Technology,Shanghai,2021:293-296.
[11]	程德强,徐进洋,寇旗旗,等. 融合残差信息轻量级网络的运煤皮带异物分类[J]. 煤炭学报,2022,47(3):1361-1369. CHENG Deqiang,XU Jinyang,KOU Qiqi,et al. Lightweight network based on residual information for foreign body classification on coal conveyor belt[J]. Journal of China Coal Society,2022,47(3):1361-1369.
[12]	MAO Qinghua,LI Shikun,HU Xin,et al. Coal mine belt conveyor foreign objects recognition method of improved YOLOv5 algorithm with defogging and deblurring[J]. Energies,2022,15(24). DOI:10.3390/ en15249504.
[13]	张旭. 带式输送机异物检测系统关键技术研究[J]. 徐州:中国矿业大学,2023. ZHANG Xu. Research on key technology of belt conveyor foreign body detection system[J]. Xuzhou:China University of Mining and Technology,2023.
[14]	LIU Jiehui,QIAO Hongchao,LIANG Lijie,et al. Improved lightweight YOLOv4 foreign object detection method for conveyor belts combined with CBAM[J]. Applied Sciences,2023,13(14). DOI: 10.3390/app13148465.
[15]	高涵,赵培培,于正,等. 基于特征增强与Transformer的煤矿输送带异物检测[J/OL]. 煤炭科学技术,1-11[2024-03-28]. http://kns.cnki.net/kcms/detail/11.2402.td.20240119.1515.012.html. GAO Han,ZHAO Peipei,YU Zheng,et al. Coal mine conveyor belt foreign object detection based on feature enhancement and Transformer[J/OL]. Coal Science and Technology,1-11[2024-03-28]. http://kns.cnki.net/kcms/detail/11.2402.td.20240119.1515.012.html.
[16]	YANG Dengjie,MIAO Changyun,LIU Yi,et al. Improved foreign object tracking algorithm in coal for belt conveyor gangue selection robot with YOLOv7 and DeepSORT[J]. Measurement,2024,228. DOI: 10.1016/j.measurement.2024.114180.
[17]	HU Jie,SHEN Li,AIBANIE S,et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,42(8):2011-2023. doi: 10.1109/TPAMI.2019.2913372
[18]	FAWZI A,SAMULOWITZ H,TURAGA D,et al. Adaptive data augmentation for image classification[C]. IEEE International Conference on Image Processing,Phoenix,2016:3688-3692.
[19]	VENKATARAMANAN S,KIJAK E,AMSALEG L,et al. AlignMixup:improving representations by interpolating aligned features[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition,New Orleans,2022:19152-19161.
[20]	WANG Niannian,ZHANG Zexi,HU Haobang,et al. Underground defects detection based on GPR by fusing simple linear iterative clustering phash (SLIC-Phash) and convolutional block attention module (CBAM)-YOLOv8[J]. IEEE Access,2024,12:25888-25905. doi: 10.1109/ACCESS.2024.3365959
[21]	PARK J,WOO S,LEE J-Y,et al. A simple and light-weight attention module for convolutional neural networks[J]. International Journal of Computer Vision,2020,128(4):783-798. doi: 10.1007/s11263-019-01283-0
[22]	郝帅,张旭,马旭,等. 基于CBAM−YOLOv5的煤矿输送带异物检测[J]. 煤炭学报,2022,47(11):4147-4156. HAO Shuai,ZHANG Xu,MA Xu,et al. Foreign object detection in coal mine conveyor belt based on CBAM-YOLOv5[J]. Journal of China Coal Society,2022,47(11):4147-4156.
[23]	CHEN Jierun,KAO S,HE Hao,et al. Run,don't walk:chasing higher FLOPS for faster neural networks[C]. IEEE/CVF Conference on Computer Vision and Pattern Recognition,Vancouver,2023:12021-12031.
[24]	HUANG Kaifeng,LI Shiyan,CAI Feng,et al. Detection of large foreign objects on coal mine belt conveyor based on improved[J]. Processes,2023,11(8). DOI: 10.3390/pr11082469.
[25]	SELVARAJU R R,COGSWELL M,DAS A,et al. Grad-CAM:visual explanations from deep networks via gradient-based localization[J]. International Journal of Computer Vision,2020,128(2):336-359. doi: 10.1007/s11263-019-01228-7