煤炭在开采过程中，往往混有大量的煤矸石，不但会影响煤的质量，还会对环境造成严重的污染。因此，对煤和矸石进行正确的分离就显得尤为重要。传统的选煤方法往往采用人工的方法，这种方法不仅效率低，还会危害工人的身体健康。因此，本文将采用机器视觉的方法实现煤和矸石的识别。本文的主要研究工作如下：
    （1）传统的图像分割方法对图像进行分割时往往会存在过分割或欠分割等问题，因此，本文引入一种新的改进分水岭分割方法，通过对煤和矸石图像进行形态学重构，以及对局部极大值区域的调整和标记来完成分割，并将其与传统方法进行对比，结果表明改进的分割方法可以将图像完整的分割出来，且边缘轮廓等细节信息都比较清晰。
    （2）本文分别研究了煤和矸石图像的灰度特征和纹理特征，并将灰度特征、纹理特征以及灰度-纹理联合特征与支持向量机模型和K-近邻模型分类器相结合分别对煤和矸石图像分类，通过实验对比可知，采用灰度-纹理联合特征与支持向量机相结合对煤和矸石识别的准确率更高，达到了94.0%。
    （3）针对灰狼算法（GWO）存在的种群多样性差、求解精度低、收敛速度慢等问题，本文结合Chebyshev混沌映射，对控制参数a进行非线性变化以及改进位置更新公式对灰狼算法进行改进，并通过10个常用的测试函数进行寻优验证，结果表明，改进后的算法有效地平衡了全局搜索和局部寻优能力，从而使求解精度更高、收敛速度更快。
    （4）为了进一步提高支持向量机的分类准确率，本文通过改进的GWO算法来优化SVM，并与Cancer数据集相结合，对改进后的算法进行分类性能的评估。分别采用GWO-SVM、IGWO-SVM、MGWO-SVM以及本文分类算法对煤和矸石图像进行分类，结果显示本文提出的算法识别的准确率最高，达到了96.5%。
      基于MATLAB2018来完成煤和矸石图像的分选，通过实验对比，验证了改进灰狼算法优化支持向量机分类模型的可靠性和有效性，具有一定的理论意义和实用价值。

         In the process of coal mining, there is often a great quantity of coal gangue, which will not only affect the quality of coal, but also cause serious pollution to the environment. Therefore, the correct separation of coal and gangue is particularly important. The traditional coal preparation method often adopts manual method, which is not only inefficient, but also imperils the health of workers. Consequently, this paper will use the method of machine vision to realize the recognition of coal and gangue. The major research contributions are as follows：
         (1) Traditional image segmentation methods often have problems such as over segmentation or under segmentation when segmenting images. Therefore, this paper introduces an improved watershed segmentation method, which completes the segmentation by morphological reconstruction of coal and gangue images and marking the local maximum region. Compared with the traditional methods, the results show that the improved segmentation algorithm can segment the image completely, and the detail contour is relatively clear.
         (2) In this paper, the gray characteristics and texture characteristics of coal and gangue are studied respectively, and classifies coal and gangue by combining gray features, texture features and gray texture joint features with SVM model and K-nearest neighbor model classifier. Through experimental comparison, the combination of support vector machine and joint features has a higher accuracy of coal and gangue recognition, reaching 94.0%.
         (3) Aiming at the problems of poor population diversity, low accuracy and easy to fall into local optimization in Grey Wolf Optimizer (GWO), combined with Chebyshev chaotic map, this paper improves the GWO algorithm by nonlinear change of control parameter a and improving the position update formula, and optimizes and verifies 10 commonly used test functions. The results show that the Improved GWO algorithm successfully balances the all-round search capabilities and local optimization capabilities, and the convergence speed is enhanced.
         (4) To further improve the classification accuracy of SVM, this paper optimizes the SVM by improved GWO and evaluates the classification performance of the improved algorithm by combining it with the Cancer data set. GWO-SVM, IGWO-SVM, MGWO-SVM and the classification algorithm in this paper were used to classify coal and gangue images, respectively, and the results showed that the proposed algorithm in this paper achieved the highest accuracy of 96.5% in recognition.
         Coal and gangue images are classified and recognized based on matlab2018. Through experimental comparison, the reliability and effectiveness of the improved gray wolf algorithm to optimize the support vector machine classification model are verified, which has certain theoretical significance and practical value.

[1]姜大霖, 程浩. 中长期中国煤炭消费预测和展望[J]. 煤炭经济研究, 2020, 40(7): 16-21.

[2] 董继友. 我国煤炭行业新旧动能转换的思考及应对措施[J]. 中国集体经济, 2019, (32): 26-27.

[3] 赵开功, 李彦平. 我国煤炭资源安全现状分析及发展研究[J]. 煤炭工程, 2018, 10: 185-189.

[4] Huang Y L, Li J M, Song T Q, et al. Microstructure of coal gangue and precipitation of heavy metal elements[J]. Journal of Spectroscopy, 2017.

[5] Qiu J S, Zhu M Y, Zhou Y X, et al. Effect and mechanism of coal gangue modification by fla ash[J]. Construction and Building Materials, 2021, 294.

[6] 张园园. 煤矸石燃烧特性及影响机制研究[D]. 山西: 山西大学, 2016.

[7] 杨慧刚, 乔志敏. 基于X射线和机器视觉的煤与矸石分选系统设计[J]. 工矿自动化, 2017, 43(3): 85-89.

[8] 康利, 黄金辉, 刘畅. GDRT型γ射线智能干法分选系统在六家煤矿的应用[J]. 煤炭加工与综合利用, 2017, (3): 22-24.

[9] 王锐, 桂志国, 刘祎, 等. 基于X射线和结构光相机的煤矸石分拣方法研究[J/OL]. 中北大学学报(自然科学版), 2021, (2): 123-128+134.

[10]王志刚, 庞卫东, 常建勇, 等. 新型大处理能力双侧进风跳汰分选技术[J]. 采矿与安全工程学报, 2020, 37(1): 202-206.

[11]Vithu P, Moses J A. Machine vision system for food grain quality evaluation: A review[J], Trends in Food Science & Technology, 2016, 01: 13-20.

[12]饶晓慧. 机器视觉的农业图像识别系统分析[J]. 电子世界, 2021, (16): 49-50.

[13]张鹏远. 机械制造自动化中关于机器视觉技术的实践应用[J]. 流体测量与控制, 2022, 03(2): 1-4.

[14]郭磊. 机器视觉在商用车产线的应用与研究[J]. 汽车与驾驶维修(维修版), 2022, (3): 21-23.

[15]Fookes R A, Gravitis V L, Watt J S, et al. Online determination of the ash content of coal using a ”siroash” gauge based on the transmission of low and high energy rays[J]. The International Journal of Applied Radiation and Isotopes, 1993, 34(1): 63-69.

[16]丰建荣. 煤和矸石井下破碎分选理论及实验研究[D]. 太原: 太原理工大学, 2006.

[17]徐龙江. 井下鼠笼式选择性煤矸分离装备关键技术研究[D]. 徐州: 中国矿业大学, 2012.

[18]David M H, Robert M C, Yong Y. Differentiation between coal and stone through Image analysis of texture features[J]. IEEE International Workshop on Imaging Systems and Techniques, 2007, 1-4.

[19]Jayachitra S, Prasanth A. Multi-feature analysis for automated brain stroke classification using weighted gaussian naive bayes classifier[J]. Journal of Circuits, Systems and Computers. 2021, 30(10): 2150178.

[20]Tripathy D P, Guru Raghavendra Reddy K. Novel methods for separation of gangue from Limestone and coal using multispectral and joint color-texture features[J]. Journal of the Institution of Engineer, 2017: 109-117.

[21]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.

[22]刘富强. 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.

[23]宋静, 杜长龙, 方树鹏. 顶水作用下井下煤与矸石分选试验[J]. 煤炭科学技术, 2009, 37(9): 122-124.

[24]刘瑜. 井下冲撞式煤矸分离中颗粒动力学行为研究[D]. 徐州: 中国矿业大学, 2011.

[25]Wang X Q, Wang S, Guo Y C, et al. Dielectric and geometric feature extraction and recognition method of coal and gangue based on VMD-SVM[J]. Powder Technology, 2021, 392: 241-250.

[26]Wang X, Hu K, Zhang L, et al. Characterization and classification of coals and rocks using terahertz time-domain spectroscopy[J]. Journal of Infrared, Millimeter, and Terahertz Waves, 2017, 38(2): 248-260.

[27]王昕, 胡克想, 喻啸, 等. 基于太赫兹时域光谱技术的煤岩界面识别[J]. 工矿自动化, 2017, 43(1): 29-34.

[28]王昕. 基于电磁波技术的煤岩识别方法研究[D]. 徐州: 中国矿业大学, 2017.

[29]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(2): 416-422.

[30]Mao Y, Ma B, Liu S, et al. Study and validation of a remote sensing model for coal extraction based on reflectance spectrum features[J]. Canadian Journal of Remote Sensing, 2014, 40(5): 327-335.

[31]Yang Y, Zeng Q. Multipoint acceleration information acquisition of the impact experiments between coal gangue and the metal plate and coal gangue recognition based on SVM and serial splicing Data[J]. Arabian Journal for Science and Engineering, 2021, 46: 2749-2768.

[32]Guo Y C, Wang X Q, Wang S, et al. Identification method of coal and coal gangue based on dielectric characteristics[J]. Institute of Electrical and Electronics Engineers Inc, 2021, 9: 9845-9854.

[33]Zhang J W, Xing H, Zheng D L. Improving coal/gangue recognition efficiency based on liquid intervention with infrared imager at low emissivity[J]. Measurement, 2022, 189: 110445.

[34]Sun Z Y, Huang L L, Jia R Q. Coal and gangue separating robot system based on computer vision[J]. Sensors, 2021, 21: 1349.

[35]Dou D Y, Wu W Z, Yang J G, et al. Classification of coal and gangue under multiple surface conditions via machine vision and relief-SVM[J]. Powder Technology, 2019, 356: 1024-1028.

[36]谭春超. 基于图像处理技术的煤矸识别与分选技术研究[D]. 山西太原: 太原理工大学, 2014.

[37]Hu J, Shen L, Sun G, Squeeze-and-excitation networks[C]// Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, 7132-7141.

[38]孔力, 郑朝阳, 钟志平. 煤矸石双能法动态识别模型研究[J]. 选煤技术, 2001(1): 9-12.

[39]陈国杰, 赵维义, 朱星. 基于单片机双能γ射线透射煤矸石在线识别仪[J]. 核电子学与探测技术, 2004, 24(2): 140-142.

[40]Zhao M, Liu H, Liu C, et al. Spatial effect analysis of coal and gangue recognition detector based on natural gamma ray method[J]. Natural Resources Research, 2022, 31: 953-969.

[41]李文斌, 杨剑瑜, 文建国. 光电选矸识别系统的研制与测试[J]. 仪器仪表学报, 2001, 22(3): 265-268.

[42]王丽, 李志宏. 煤矸石与煤的自动分选[J]. 山西煤炭, 2002, 23(2): 41-42.

[43]张宁波, 刘长友, 陈现辉, 等. 综放煤矸低水平自然射线的涨落规律及测量识别分析[J]. 煤炭学报, 2015, 40(5): 988-993.

[44]张宁波, 鲁岩, 刘长友, 等. 综放开采煤矸自动识别基础研究[J]. 采矿与安全工程学报, 2014, 31(4): 532-536.

[45]张宁波. 综放开采煤矸自然射线辐射规律及识别研究[D]. 徐州: 中国矿业大学, 2015.

[46]谭春超, 杨洁明. 煤与矸石图像灰度信息和纹理特征的提取研究[J]. 工矿自动化, 2017, 43(4): 27-31.

[47]阳振宇, 潘建平, 陈梦. 基于高低帽变换的图像迭代分割方法[J]. 测绘工程, 2018, 27(8): 60-65.

[48]冈萨雷斯. 数字图像处理[M]. 北京: 电子工业出版社, 2013.

[49]张燕红, 李瑛, 王凤芹, 等. 基于形态学重建和极大值标记的分水岭分割算法[J]. 计算技术与自动化, 2015, 34(2): 78-81.

[50]王家臣, 李良晖, 杨胜利. 不同照度下煤矸图像灰度及纹理特征提取的实验研究[J]. 煤炭学报, 2018, 43(11): 3051-3061.

[51]Mridula J K K, Patra D. Combining GLCM features and Markov random field model for colour textured image segmentation[C]// Conference on Devices and Communications of IEEE, 2011: 1-5.

[52]Huang Z K, Li P W, Hou L Y. Segmentation of textures using PCA fusion based gray-level co-occurrence matrix features[J]. International Conference on Test and Measurement, 2009, 1(9): 103-105.

[53]张晓凤, 王秀英. 灰狼优化算法研究综述[J]. 计算机科学, 2019, 46(3): 30-38.

[54]高晨峰, 陈家清, 石默涵. 融合黄金正弦和曲线自适应的多策略麻雀搜索算法[J]. 计算机应用研究, 2022, 39(2): 491-499.

[55]Lu C, Gao L, Li X, et al. Chaotic-based grey wolf optimizer for numerical and engineering optimization problems[J]. Memetic Computing, 2020, 12(4):371-398.

[56]Liu G Y, Shu C, Liang Z W, et al. A modified sparrow search algorithm with application in 3d route planning for UAV[J]. Sensors, 2021, 21 (4): 1224.

[57]罗佳, 唐斌. 基于收敛因子非线性动态变化的灰狼优化算法[J]. 中国科技论文, 2016, 11(17): 1991-1997.

[58]滕志军, 吕金玲, 郭力文, 等. 一种基于Tent映射的混合灰狼优化的改进算法[J]. 哈尔滨工业大学学报, 2018, 50(11): 40-49.

[59]龙文, 伍铁斌. 协调探索和开发能力的改进灰狼优化算法[J]. 控制与决策, 2017, 32(10): 1749-1757.

[60]Clerc M. The swarm and the queen: Towards a Deterministic and adaptive particle swarm optimization[C]// Proceedings of the Congress on Evolutionary Computation. Washington, 2002, 03: 1951.

[61]Japkowicz N, Shah M. Evaluating learning algorithms:a classification perspective[M]. Cambridge: Cambridge University Press, 2011.

[62]Jeni L A, Cohn J F, De La Torre F. Facing imbalanced data-recommendations for the use of performance metrics[C]// Humaine Association Conference on Affective Computing and Intelligent Interaction, 2013.

[63]Luque A, Carrasco A, Martin A, et al. The impact of class imbalance in classification performance metrics based on the binary confusion matrix[J]. Pattern recognition, 2019, 91: 216-231.

[64]Zhang Y D. A seven-layer convolutional neural network for chest CT based COVID-19 diagnosis using stochastic pooling[J]. IEEE Sensors Journal, 2020.

[65]王冠军, 苏婷婷, 刘文博, 等. 基于EAIDK的智能煤矸分拣系统设计[J]. 工矿自动化, 2020, 46(1): 105-108.

[66]赵明辉. 一种煤矸石优化识别方法[J]. 工矿自动化, 2020, 46(7): 113-116.

[67]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(3): 50-55.