轻量化煤矸目标检测方法研究

杜京义; 史志芒; 郝乐; 陈瑞

doi:10.13272/j.issn.1671-251x.2021040029

轻量化煤矸目标检测方法研究

doi: 10.13272/j.issn.1671-251x.2021040029

1. 西安科技大学电气与控制工程学院, 陕西西安 710054;
2. 西安科技大学安全科学与工程学院, 陕西西安 710054

基金项目:

工信部物联网集成创新与融合应用项目（工信部科函〔2018〕470号）。

详细信息

作者简介:
杜京义(1965-),男,山东淄博人,教授,研究方向为模式识别与神经网络,E-mail:517571853@qq.com。

中图分类号: TD67
计量
- 文章访问数: 118
- HTML全文浏览量: 8
- PDF下载量: 26
- 被引次数: 0
出版历程
- 收稿日期: 2021-04-11
- 修回日期: 2021-11-07
- 刊出日期: 2021-11-20

Research on lightweight coal and gangue target detection method

1. College of Electrical and Control Engineering, Xi'an University of Science and Technology, Xi'an 710054, China;
2. College of Safety Science and Engineering, Xi'an University of Science and Technology, Xi'an 710054, China

摘要

摘要: 针对目前基于深度学习的煤矸目标检测方法精度低、实时性差、小目标易漏检等问题，采用轻量化网络、自注意力机制、锚框优化方法对SSD模型进行改进，构建Ghost-SSD模型，进而提出一种轻量化煤矸目标检测方法。Ghost-SSD模型以SSD模型为基础框架，采用GhostNet轻量化特征提取网络代替主体网络层VGG16，以提高煤矸目标检测速度；针对浅层特征图中包含较多背景噪声及语义信息不足问题，引入自注意力模块对浅层特征图进行特征增强，提高对前景区域的关注度，并采用扩张卷积增大浅层特征图的感受野，丰富浅层特征图的语义信息；采用K-means算法对锚框进行聚类，优化锚框尺寸设置，进一步提高煤矸目标检测精度。实验结果表明，基于Ghost-SSD模型进行煤矸目标检测时，平均精度均值较SSD模型提高3.6%，检测速度提高75帧/s，且检测精度与速度均优于Faster-RCNN，Yolov3模型，同时对煤矸小目标具有较好的检测效果。
- 煤矸分选 /
- 煤矸识别 /
- 煤矸目标检测 /
- 自注意力机制 /
- SSD模型 /
- GhostNet /
- 锚框聚类优化
Abstract: In order to solve the problems of low precision, poor real-time performance and easy missing detection of small targets in the current deep learning-based coal and gangue target detection methods, the SSD model is improved by using lightweight network, self-attention mechanism and anchor frame optimization method to construct Ghost-SSD model, and then a lightweight coal and gangue target detection method is proposed.The Ghost-SSD model is based on the SSD model, and the GhostNet lightweight characteristic extraction network is used to replace the main network layer VGG16 so as to improve the detection speed of coal and gangue targets.In order to solve the problem that the shallow characteristic map contains more background noise and insufficient semantic information, the self-attention module is introduced to enhance the characteristics of the shallow characteristic map and increase the focus on the foreground region.Moreover, the dilated convolution is applied to increase the receptive field of the shallow characteristic maps and enrich the semantic information of the shallow characteristic maps.The K-means algorithm is used to cluster the anchor frames, optimize the size of the anchor frame, and further improve the precision of coal and gangue target detection.The experimental results show that when the Ghost-SSD model is applied in coal and gangue target detection, the mean average precision is 3.6% higher than that of the SSD model, the detection speed is increased by 75 frames/s, and the detection precision and speed are better than that of the Faster-RCNN and Yolov3 models.Moreover, the model has a good detection effect on small coal and gangue targets.
- coal and gangue separation /
- coal and gangue identification /
- coal and gangue target detection /
- self-attention mechanism /
- SSD model /
- GhostNet /
- anchor frame clustering optimization

HTML全文

参考文献(22)

[1]	商德勇,章林,牛艳奇,等.煤矸分拣机器人设计与关键技术分析[J/OL].煤炭科学技术:1-7[2021-04-08].http://kns.cnki.net/kcms/detail/11.2402.TD.20200616.1022.004.html. SHANG Deyong,ZHANG Lin,NIU Yanqi,et al.Design and key technology analysis of coal and gangue sorting robot[J/OL].Coal Science and Technology:1-7[2021-04-08].http://kns.cnki.net/kcms/detail/11.2402.TD.20200616.1022.004.html.
[2]	郭永存,何磊,刘普壮,等.煤矸双能X射线图像多维度分析识别方法[J].煤炭学报,2021,46(1):300-309. GUO Yongcun,HE Lei,LIU Puzhuang,et al.Multi-dimensional analysis and recognition method of coal and gangue dual-energy X-ray images[J].Journal of China Coal Society,2021,46(1):300-309.
[3]	ROBBEN C,CONDORI P,PINTO A,et al.X-ray-transmission based ore sorting at the San Rafael tin mine[J].Minerals Engineering,2020,145:105870.
[4]	杨慧刚,乔志敏,高绘彦,等.煤与矸石分选系统设计[J].工矿自动化,2018,44(8):91-95. YANG Huigang,QIAO Zhimin,GAO Huiyan,et al.Design of separation system of coal and gangue[J].Industry and Mine Automation,2018,44(8):91-95.
[5]	李曼,段雍,曹现刚,等.煤矸分选机器人图像识别方法和系统[J].煤炭学报,2020,45(10):3636-3644. LI Man,DUAN Yong,CAO Xiangang,et al.Image identification method and system for coal and gangue sorting robot[J].Journal of China Coal Society,2020,45(10):3636-3644.
[6]	薛光辉,李秀莹,钱孝玲,等.基于随机森林的综放工作面煤矸图像识别[J].工矿自动化,2020,46(5):57-62. XUE Guanghui,LI Xiuying,QIAN Xiaoling,et al.Coal-gangue image recognition in fully-mechanized caving face based on random forest[J].Industry and Mine Automation,2020,46(5):57-62.
[7]	DOU Dongyang,ZHOU Deyang,YANG Jianguo,et al.Coal and gangue recognition under four operating conditions by using image analysis and Relief-SVM[J].International Journal of Coal Preparation and Utilization,2018,40(7):473-482.
[8]	HOU Wei.Identification of coal and gangue by feed-forward neural network based on data analysis[J].International Journal of Coal Preparation and Utilization,2019,39(1):33-43.
[9]	DHILLON A,VERMA G K.Convolutional neural network: a review of models,methodologies and applications to object detection[J].Progress in Artificial Intelligence,2020,9(2):85-112.
[10]	徐志强,吕子奇,王卫东,等.煤矸智能分选的机器视觉识别方法与优化[J].煤炭学报,2020,45(6):2207-2216. XU Zhiqiang,LYU Ziqi,WANG Weidong,et al.Machine vision recognition method and optimization for intelligent separation of coal and gangue[J].Journal of China Coal Society,2020,45(6):2207-2216.
[11]	PU Yuanyuan,APEL D B,SZMIGIEL A,et al.Image recognition of coal and coal gangue using a convolutional neural network and transfer learning[J].Energies,2019,12(9):1-11.
[12]	LI Dongjun,ZHANG Zhenxin,XU Zhihua,et al.An image-based hierarchical deep learning framework for coal and gangue detection[J].IEEE Access,2019,7:184686-184699.
[13]	王鹏,曹现刚,夏晶,等.基于机器视觉的多机械臂煤矸石分拣机器人系统研究[J].工矿自动化,2019,45(9):47-53. WANG Peng,CAO Xiangang,XIA Jing,et al.Research on multi-manipulator coal and gangue sorting robot system based on machine vision[J].Industry and Mine Automation,2019,45(9):47-53.
[14]	LYU Ziqi,WANG Weidong,XU Zhiqiang,et al.Cascade network for detection of coal and gangue in the production context[J].Powder Technology,2021,377:361-371.
[15]	LIU Wei,ANGUELOV D,ERHAN D,et al.SSD:single shot multibox detector[C]//European Conference on Computer Vision,Amsterdam,2016:21-37.
[16]	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,2016,39(6):1137-1149.
[17]	REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once: unified,real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,Las Vegas,2016:779-788.
[18]	LUO Wenjie,LI Yujia,URTASUN R,et al.Understanding the effective receptive field in deep convolutional neural networks[Z/OL].arXiv Preprint,arXiv:1701.04128,2017.
[19]	HAN Kai,WANG Yunhe,TIAN Qi,et al.Ghostnet:more features from cheap operations[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,Seattle,2020:1580-1589.
[20]	HE Kaiming,ZHANG Xiangyu,REN Shaoqing,et al.Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas,2016:770-778.
[21]	ZHANG Han,GOODFELLOW I,METAXAS D,et al.Self-attention generative adversarial networks[C]//International Conference on Machine Learning,Long Beach,2019:7354-7363.
[22]	SELVARAJU R R,COGSWELL M,DAS A,et al.Grad-cam:visual explanations from deep networks via gradient-based localization[C]//Proceedings of the IEEE International Conference on Computer Vision,Long Beach,2017:618-626.