煤炭作为我国的主要能源，不断地为工业生产提供动力。但是在煤炭开采和运输的过程中常伴有较多矸石，降低煤质，污染环境。因此，煤矸识别的研究具有十分重要的理论意义和实用价值。本文依据煤和矸石的视觉差异，对煤矸图像识别算法进行研究。

(1) 提出基于迁移学习的煤矸图像识别算法。利用VGG16卷积基提取煤矸图像特征，并与机器学习算法结合，验证了VGG16卷积基提取特征的有效性。然后分别通过特征提取和模型微调方式实现VGG16的迁移，并构建自定义密集连接分类器，形成两种识别模型。对单目标煤矸图像的仿真结果是：两种模型的准确率分别为96.30%和98.15%，都具有良好性能，可以快速准确识别单目标煤矸图像。

(2) 提出基于YOLOv5的轻量化煤矸图像识别算法。对于CPU和GPU设备，分别利用轻量型网络修改YOLOv5s网络结构，并引入卷积注意力机制，通过权重参数去除冗余特征，最终形成YOLOv5s_MC、YOLOv5s_SC、YOLOv5s_GC三种识别模型。实验结果显示，针对CPU和GPU设备，改进后的模型在保证检测精度基本不变的同时，参数量分别减少了54.66%、75.62%和44.46%，计算量分别降低了62.58%、73%和38%。结果表明：改进后的模型大大降低了模型参数量和计算量，提升了检测实时性，为解决多目标煤矸图像检测算法模型复杂、计算量大的问题提供了参考。

Coal can continuously provide power for industrial production. However, it is often accompanied by gangue, which pollutes the environment. Therefore, the recognition of coal gangue is of great significance. This paper proposes image-based coal gangue recognition algorithms. The main contents of the study are as follows:

(1) Build a coal gangue recognition model based on the idea of transfer learning. First, the VGG16 convolution base is used to extract coal gangue image features, combined with three machine learning algorithms to verify the effectiveness of the features. Then, the migration of VGG16 is realized through feature extraction and model fine-tuning respectively, and a custom classifier is constructed to form two recognition models. The simulation results show that the accuracy rates of the two models are 96.30% and 98.15% respectively.

(2) Build a lightweight coal gangue recognition model based on YOLOv5s. This paper uses lightweight networks to modify YOLOv5s network to make the model light for CPU and GPU devices respectively; and introduces a convolution attention mechanism, improves the feature extraction capabilities of model. Compared with the original network, the number of parameters of the improved models is reduced by 54.66%, 75.62% and 44.46%, and the amount of calculation is reduced by 62.58%, 73% and 38% respectively. The results show that the improved models greatly reduce the amount of model parameters and calculation, improve the real-time detection, and provide a reference for solving the problem of complex multi-object coal gangue image detection algorithm and large amount of calculation.

[1] 曹现刚, 李莹, 王鹏, 等. 煤矸石识别方法研究现状与展望[J]. 工矿自动化, 2020, 46(01): 38-43.

[2] 饶中钰, 吴景涛, 李明. 煤矸石图像分类方法[J]. 工矿自动化, 2020, 46(03): 69-73.

[3] 邢伟, 宁玉伟. 基于γ射线探测技术的煤矸石分选系统的设计[J]. 河南农业大学学报, 2007(04): 455-457.

[4] 潘越, 曾哲, 张恩瑜. 基于MATLAB和图像灰度值对X射线探测煤矸识别的研究[J]. 煤炭技术, 2017, 36(11): 307-309.

[5] Kehong Z, Changlong D, Daolong Y. Orthogonal Test and Support Vector Machine Applied to Influence Factors of Coal and Gangue Separation[J]. International Journal of Coal Preparation & Utilization, 2014, 34(2): 75-84.

[6] Yang D, Li J, Zheng K, et al. Impact-Crush Separation Characteristics of Coal and Gangue[J]. International Journal of Coal Preparation & Utilization, 2016: 1-8.

[7] 王卫东, 张晨, 马中良,等. 煤矸光电分选系统设计[J]. 工矿自动化, 2013(12): 5-8.

[8] 宋亮, 刘善军, 虞茉莉,等. 基于可见-近红外和热红外光谱联合分析的煤和矸石分类方法研究[J]. 光谱学与光谱分析, 2017, 37(02): 416-422.

[9] 马宪民. 煤矸石在线识别与自动分选系统的研究[J]. 西安科技学院学报, 2003(01): 66-68.

[10] 陈立, 杜文华, 曾志强,等. 基于小波变换的煤矸石自动分选方法[J]. 工矿自动化, 2018, 44(12): 60-64.

[11] 马宪民. 一种基于小波变换的煤矸石图像边缘检测方法[J]. 仪器仪表学报, 2006(S3): 2130-2131.

[12] 张世杰, 陈泽华. 基于梯度下降算法的煤矸石分割识别方法[J]. 煤矿机械, 2014, 35(08): 128-130.

[13] 于腾飞. 基于相位一致性的煤矸边缘检测方法[J]. 电子器件, 2016, 39(06): 1349-1352.

[14] 胡倍倍. 基于DSP的煤和煤矸石图像识别研究[D]. 开封: 河南大学硕士论文, 2017.

[15] 何敏, 王培培, 蒋慧慧. 基于SVM和纹理的煤和煤矸石自动识别[J]. 计算机工程与设计, 2012, 33(03): 1117-1121.

[16] 余乐, 郑力新, 杜永兆,等. 采用部分灰度压缩扩阶共生矩阵的煤和煤矸石图像识别[J]. 华侨大学学报(自然科学版), 2018, 39(06): 906-912.

[17] 郭永存, 于中山, 卢熠昌. 基于PSO优化NP-FSVM的煤矸光电智能分选技术研究[J]. 煤炭科学技术, 2019, 47(04): 13-19.

[18] 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].煤炭科学技术,2020: 1-7.

[19] Li M, Duan Y, He X, et al. Image positioning and identification method and system for coal and gangue sorting robot[J]. International Journal of Coal Preparation and Utilization, 2020(04): 1-19.

[20] 庞尚钟, 李博, 王学文,等. 基于机器视觉的煤矸识别系统设计及试验研究[J]. 煤炭工程, 2021, 53(02): 141-146.

[21] Hou, Wei. Identification of Coal and Gangue by Feed-forward Neural Network Based on Data Analysis[J]. International Journal of Coal Preparation & Utilization, 2017: 1-11.

[22] 田冬艳, 丁苏凡, 郭星歌. 基于图像处理的煤矸识别方法[J]. 煤炭技术, 2022, 41(03): 201-204.

[23] Eshaq R M A, Hu E, Li M, et al. Separation Between Coal and Gangue Based on Infrared Radiation and Visual Extraction of the YCbCr Color Space[J]. IEEE Access, 2020, 8: 55204-55220.

[24] Hu F, Zhou M, Yan P, et al. Multispectral Imaging: A New Solution for Identification of Coal and Gangue[J]. IEEE Access, 2019, 7: 169697-169704.

[25] Kai L, Xi Z, Chen Y Q. Extraction of Coal and Gangue Geometric Features with Multifractal Detrending Fluctuation Analysis[J]. Applied Sciences, 2018, 8(3): 463.

[26] 曹现刚, 费佳浩, 王鹏,等. 基于多机械臂协同的煤矸分拣方法研究[J]. 煤炭科学技术, 2019, 47(04): 7-12.

[27] Alfarzaeai M S, Niu Q, Zhao J, et al. Coal/Gangue Recognition Using Convolutional Neural Networks and Thermal Images[J]. IEEE Access, 2020, 8: 76780-76789.

[28] Gao R, Sun Z, Li W, et al. Automatic Coal and Gangue Segmentation Using U-Net Based Fully Convolutional Networks[J]. Energies, 2020, 13(4): 1-13.

[29] Xing Jichuan et al. Coal and Gangue Identification Method Based on the Intensity Image of Lidar and DenseNet.[J]. Applied Optics, 2021, 60(22) : 6566-6572.

[30] 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(06): 2207-2216.

[31] Pu Y, 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(09): 1-11.

[32] 雷世威, 肖兴美, 张明. 基于改进YOLOv3的煤矸识别方法研究[J]. 矿业安全与环保, 2021, 48(03): 50-55.

[33] Li D, Wang G, Zhang Y, et al. Coal Gangue Detection and Recognition Algorithm Based on Deformable Convolution YOLOv3[J]. IET Image Processing, 2022, 16(1): 134-144.

[34] Liu Q, Li J, Li Y, et al. Recognition Methods for Coal and Coal Gangue Based on Deep Learning[J]. IEEE Access, 2021, 9: 77599-77610.

[35] 沈科, 季亮, 张袁浩,等. 基于改进YOLOv5s模型的煤矸目标检测[J]. 工矿自动化, 2021, 47(11): 107-111+118.

[36] 杜京义, 史志芒, 郝乐, 等. 轻量化煤矸目标检测方法研究[J]. 工矿自动化, 2021, 47(11): 119-125.

[37] Sun Z, Huang L, Jia R . Coal and Gangue Separating Robot System Based on Computer Vision[J]. Sensors, 2021, 21(4): 1349.

[38] Yan P, Sun Q, Yin N, et al. Detection of Coal and Gangue Based on Improved YOLOv5. 1 Which Embedded ScSE Module[J]. Measurement, 2022, 188: 110530.

[39] Simonyan K, Zisserman A. Very Deep Convolutional Networks for Large-Scale Image Recognition[C]//International Conference on Learning Representations(ICLR).2015: 1-14.

[40] Szegedy C, Liu W, Jia Y, et al. Going Deeper with Convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2015: 1-9.

[41] He K, Zhang X, Ren S, et al. Deep Residual Learning for Image Recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016: 770-778.

[42] Sandler M, Howard A, Zhu M, et al. Mobilenetv2: Inverted Residuals and Linear Bottlenecks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 4510-4520.

[43] Zhang X, Zhou X, Lin M, et al. Shufflenet: An Extremely Efficient Convolutional Neural Network for Mobile Devices[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 6848-6856.

[44] Girshick R, Donahue J, Darrell T, et al. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2014: 580-587.

[45] Girshick R. Fast r-cnn[C]//Proceedings of the IEEE International Conference on Computer Vision. 2015: 1440-1448.

[46] Ren S, He K, Girshick R, et al. Faster r-cnn: Towards Real-time Object Detection with Region Proposal Networks[J]. Advances in Neural Information Processing Systems, 2015, 28: 91-99.

[47] 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. 2016: 779-788.

[48] Redmon J, Farhadi A. YOLO9000: Better, Faster, Stronger[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017: 7263-7271.

[49] Redmon J, Farhadi A. Yolov3: An incremental improvement[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 1-6.

[50] Bochkovskiy A, Wang C Y, Liao H. YOLOv4: Optimal Speed and Accuracy of Object Detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2020: 1-17.

[51] 曹晓杰, 么娆, 严雨灵. 应用迁移学习的卷积神经网络花卉图像识别[J]. 计算机应用与软件, 2020, 37(08): 142-148.

[52] 刘子源, 王翀, 李晓飞. 基于AdaBoost算法的直升机结构制造复杂度估算[J]. 中国电子科学研究院学报, 2020, 15(10): 977-983.

[53] 陈道争, 江倩. 基于卷积神经网络和迁移学习的肺结节检测[J]. 计算机工程与设计, 2021, 42(01): 240-247.

[54] Yu J, Jiang Y, Wang Z, et al. Unitbox: An Advanced Object Detection Network [C]//Proceedings of the 24th ACM International Conference on Multimedia. 2016: 516-520.

[55] Rezatofighi H, Tsoi N, Gwak J Y, et al. Generalized Intersection Over Union: A Metric and A Loss for Bounding Box Regression[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019: 658-666.

[56] Zheng Z, Wang P, Liu W, et al. Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2020, 34(07): 12993-13000.

[57] Howard A, Sandler M, Chu G, et al. Searching for Mobilenetv3[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2019: 1314-1324.

[58] Woo S, Park J, Lee J Y, et al. Cbam: Convolutional Block Attention Module[C]//Proceedings of the European Conference on Computer Vision (ECCV). 2018: 3-19.

[59] Ma N, Zhang X, Zheng H T, et al. Shufflenet v2: Practical Guidelines for Efficient Cnn Architecture Design[C]//Proceedings of the European Conference on Computer Vision (ECCV). 2018: 116-131.

[60] Han K, Wang Y, Tian Q, et al. Ghostnet: More Features from Cheap Operations [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020: 1580-1589.