登录    |    注册

您好,欢迎来到中国测试科技资讯平台!

首页> 《中国测试》期刊 >本期导读>蚁群和粒子群混合优化SVM的钢板表面缺陷分类研究

蚁群和粒子群混合优化SVM的钢板表面缺陷分类研究

266    2020-01-19

免费

全文售价

作者:李爱莲1, 郭志斌1, 解韶峰2, 赵多祯1, 张帅1

作者单位:1. 内蒙古科技大学信息工程学院, 内蒙古 包头 014010;
2. 内蒙古科技大学基建处, 内蒙古 包头 014010


关键词:钢板表面缺陷分类;图像特征融合;蚁群算法;粒子群算法;支持向量机


摘要:

热轧带钢表面的温度高、生产速度快,辐射光强,并且存在着水、氧化铁皮、光照不均等现象,难以通过人工进行表面质量在线检测。针对当前国内某钢厂热轧钢板表面缺陷检测仍由人工离线完成、缺陷识别准确率低的生产问题,充分利用大量图像信息,提出一种图像处理与蚁群和粒子群混合优化支持向量机结合的缺陷分类方法。首先,融合局部二值模式和局部相位量化两种特征提取方式的优点,进行钢板缺陷图片的特征提取,采用蚁群和粒子群优化出支持向量机的惩罚参数和核函数参数进行钢板表面的缺陷分类。最后采用Matlab仿真平台,将蚁群和粒子群混合优化的支持向量机分类模型与传统的支持向量机分类模型进行仿真对比分析。试验结果表明,采用蚁群和粒子群混合优化的支持向量机分类模型的分类精度高于传统的支持向量机模型。


Research on surface defect classification of steel plate based on ant colony and particle swarm optimization SVM
LI Ailian1, GUO Zhibin1, XIE Shaofeng2, ZHAO Duozhen1, ZHANG Shuai1
1. Information Engineering Institute, Inner Mongolia University of Science and Technology, Baotou 014010, China;
2. Capital Construction Department, Inner Mongolia University of Science and Technology, Baotou 014010, China
Abstract: The temperature of hot-rolled steel plate surface is high, production speed is fast, radiation is strong, and there are water, iron oxide scale, light uneven phenomenon, it is difficult to carry out surface quality on-line detection by manual. In view of the problem that the defect detection of hot-rolled steel plate surface in a domestic steel plant is still completed by manual off-line and the accuracy of defect recognition is low, a defect classification method combining image processing with ant colony optimization and particle swarm optimization support vector machine was proposed by making full use of a large number of image information on the spot. Firstly, the advantages of two feature extraction methods, local binary pattern (LBP) and local phase quantization (LPQ), were combined to extract the features of steel plate defect images.The penalty parameters and kernel function parameters of support vector machine (SVM) were optimized by ant colony optimization and particle swarm optimization (PSO) to classify the surface defects of steel plate. Finally, the support vector machine classification model based on ant colony optimization and particle swarm optimization was compared with the traditional support vector machine classification model by using Matlab simulation platform. The experimental result shows that the classification accuracy of the support vector machine classification model based on ant colony optimization and particle swarm optimization is higher than that of the traditional support vector machine model.
Keywords: surface defect classification of steel plate;image feature fusion;ant colony optimization;particle swarm optimization;support vector machine
2020, 46(1):110-116  收稿日期: 2019-05-29;收到修改稿日期: 2019-07-09
基金项目: 内蒙古自治区自然科学基金项目资助(2016MS0610,2014MS0612);内蒙古科技大学产学研合作培育基金资助项目(PY-201512)
作者简介: 李爱莲(1973-),女,河北唐山市人,副教授,硕士,主要从事复杂过程建模、优化控制研究
参考文献
[1] 胡超. 基于机器视觉的热轧钢板表面缺陷检测系统关键技术研究[D]. 沈阳:东北大学, 2012.
[2] 刘钟, 吴杰, 张华. 热轧带钢表面质量检测系统的工程设计与实践[J]. 宝钢技术, 2005(6):57-61
[3] 胡联亭. 基于机器视觉的钢板表面缺陷识别研究[D]. 武汉:武汉科技大学, 2018.
[4] 刘海博. 基于机器视觉的钢板表面缺陷检测系统研究[D]. 长春:吉林大学, 2008.
[5] 徐科, 徐金梧, 陈雨来. 冷轧带钢表面缺陷在线监测系统[J]. 北京科技大学学报, 2002(3):329-332
[6] 胡亮, 段发阶, 丁克勤, 等. 钢板表面缺陷计算机视觉在线检测系统的研制[J]. 钢铁, 2005(2):59-61
[7] 徐科, 杨朝霖, 周鹏. 热轧带钢表面缺陷在线检测的方法与工业应用[J]. 机械工程学报, 2009, 45(4):111-114+124
[8] 周诗洋. 基于视觉显著性和稀疏表示的钢板表面缺陷图像检测方法研究[D]. 武汉:华中科技大学, 2017.
[9] JAYAPRAKASH A, KEZISEL V C. Feature selection using ant colony optimization (ACO) and road sign detection and recognition (RSDR) system[J]. Cognitive Systems Research, 2019, 58:123-133
[10] WANG X K, GUAN S Y, HUA L, et al. Classification of spot-welded joint strength using ultrasonic signal time-frequency features and PSO-SVM method[J]. Ultrasonics, 2018, 91:161-169
[11] 韩顺杰, 齐冀樊, 姜玉莲, 等. 基于主成分分析与遗传算法-支持向量机的喷溅预测方法[J]. 钢铁研究学报, 2016, 28(12):21-26
[12] 李春祥, 丁晓达, 叶继红. 基于混合蚁群和粒子群优化LSSVM的脉动风速预测[J]. 振动与冲击, 2016, 35(21):131-136
[13] 李培玉, 张志欣, 董月. 基于图像特征的板坯表面裂纹检测方法[J]. 钢铁研究学报, 2013, 25(8):58-62
[14] 李哲, 于梦茹. 基于多种LBP特征集成学习的车标识别[J/OL]. 计算机工程与应用:1-10[2019-05-25]. http://kns.cnki.net/kcms/detail/11.2127.TP.20181025.1740.039.html.
[15] 王宪, 秦磊, 孙子文, 等. 基于局部Shearlet相位量化特征的人脸识别算法[J]. 光电工程, 2013, 40(4):106-112
[16] 张弢, 柳雨农, 任帅, 等. 基于LBP人脸纹理特征的差分直方图移位无损信息隐藏算法[J/OL]. 计算机应用研究:1-6[2019-05-25]. https://doi.org/10.19734/j.issn.1001-3695.2018.11.0889.
[17] 刘帅师, 田彦涛, 万川. 基于Gabor多方向特征融合与分块直方图的人脸表情识别方法[J]. 自动化学报, 2011, 37(12):1455-1463
[18] PRIYADARSHI B K, DAYAL R P, CHINMAYA S. An approach to optimize the path of humanoid robots using a hybridized regression-adaptive particle swarm optimization-adaptive ant colony optimization method[J]. Industrial Robot:the international journal of robotics research and application, 2019, 46(1):104-117
[19] 田野, 高晓华, 汤建民, 等. 基于粒子群遗传优化的热轧带钢头尾宽度控制[J]. 钢铁, 2011, 46(1):55-59
[20] JUN H T, ZHI C O, ZUBAIDAH I, et al. A new hybrid GA-ACO-PSO algorithm for solving various engineering design problems[J]. International Journal of Computer Mathematics, 2019, 96(5):292-305