[1]王粟,李庚,曾亮.高斯差分空间的多尺度改进CLBP对带钢表面缺陷的分类[J].华侨大学学报(自然科学版),2020,41(4):534-540.[doi:10.11830/ISSN.1000-5013.201908006]
　WANG Su,LI Geng,ZENG Liang.Classification of Strip Surface Defects by Multi-Scale Improved CLBP Based on Gaussian Difference Space[J].Journal of Huaqiao University(Natural Science),2020,41(4):534-540.[doi:10.11830/ISSN.1000-5013.201908006]
点击复制

高斯差分空间的多尺度改进CLBP对带钢表面缺陷的分类()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第41卷
期数:
2020年第4期
页码:
534-540
栏目:
出版日期:
2020-07-20

文章信息/Info

Title:
Classification of Strip Surface Defects by Multi-Scale Improved CLBP Based on Gaussian Difference Space
文章编号:
1000-5013(2020)04-0534-07
作者:
王粟1; 2;  李庚1; 2;  曾亮1; 2
1. 湖北工业大学 太阳能高效利用及储能运行控制湖北省重点实验室, 湖北 武汉 430068;2. 湖北工业大学 电气与电子工程学院, 湖北 武汉 430068
Author(s):
WANG Su1; 2;  LI Geng1; 2;  ZENG Liang1; 2
1. Hubei Key Laboratory for High-Efficiency Utilization of Solar Energy and Operation Control of Energy Storage System, Hubei University of Technology, Wuhan 430068, China; 2. College of Electrical and Electronic Engineering, Hubei Univeraity of Technolog
关键词:
带钢表面;  缺陷分类;  多尺度完全局部二值模式;  高斯差分空间;  特征提取;  非线性流行学习
Keywords:
strip surface;  defect classification;  multi-scale completed local binary mode;  Gaussian difference space;  feature extraction;  nonlinear popular learning
分类号:
TG356.21;TP391.41
DOI:
10.11830/ISSN.1000-5013.201908006
文献标志码:
A
摘要:
提出一种引入高斯差分空间的改进多尺度完全局部二值模式对带钢表面进行分类,解决由于带钢表面缺陷纹理存在复杂性和多样性,导致对带钢表面缺陷进行分类难度大的问题.首先,根据人类的视觉注意机制,采用高斯差分空间对带钢表面缺陷进行预处理.然后,采用多尺度改进的完全局部二值模式对预处理之后的图片进行特征提取.最后,采用非线性流行学习的方式对特征进行降维,并导入分类器中进行分类.实验结果表明:该方法具有较好的区分性;针对常见的冲孔、污渍、刮边、黑氧化条、结疤等带钢表面缺陷,其最终的分类精度能达到95.7%,优于目前传统的方式.
Abstract:
An improved multi-scale complete local binary model with Gauss difference space is proposed to classify the surface defects of strip steel, which solves the problem that the classification of surface defects is difficult due to the complexity and diversity of surface defect texture. Firstly, according to human visual attention mechanism, Gaussian differential space is used to preprocess the surface defects of strip steel. Then, the multi-scale improved complete local binary model is used to extract the features of the preprocessed image. Finally, the nonlinear popular learning method is used to reduce the dimension of features and import them into the classifier for classification. The experimental results show that the method has good regional characteristics: for common surface defects such as punching, stains, scraping, black oxide strip, scab and other surface defects, the final classification accuracy can reach 95.7%, which is better than the traditional method.

参考文献/References:

[1] 刘坤,张阿龙,屈尔庆,等.基于局部特征加权编码的带钢表面缺陷分类[J].控制工程,2018,25(12):127-132.DOI:10.14107/j.cnki.kzgc.160531.
[2] XIANG Yuming,WANG Feng,YOU Hongjian.OS-SIFT: A robust SIFT-like algorithm for high-resolution optical-to-SAR image registration in suburban areas[J].IEEE Transactions on Geoscience and Remote Sensing,2018,56(6):3078-3090.DOI:10.1109/TGRS.2018.2790483.
[3] 贾银江, 徐哲男,苏中滨,等.基于优化SIFT算法的无人机遥感作物影像拼接[J].农业工程学报,2017,33(10):131-137.DOI:10.11975/j.issn.1002-6819.2017.10.016.
[4] LI Yanshan,YANG Congzhu,ZHANG Li,et al.A novel SURF based on a unified model of appearance and motion-variation[J].IEEE Access,2018,6:31065-31076.DOI:10.1109/ACCESS.2018.2832290.
[5] CHAO Jianshu,HUITL R,STEINBACH E,et al.A novel rate control framework for SIFT/SURF feature preservation in H.264/AVC video compression[J].IEEE Transactions on Circuits and Systems for Video Technology,2015,25(6):958-972.DOI:10.1109/tcsvt.2014.2367354.
[6] 许良凤,刘泳海,胡敏,等.语谱图改进完全局部二值模式的语音情感识别[J].电子测量与仪器学报,2018,32(5):25-32.DOI:10.13382/j.jemi.2018.05.004.
[7] 周宇旋,吴秦,梁久祯,等.判别性完全局部二值模式人脸表情识别[J].计算机工程与应用,2017,53(4):163-169,194.DOI:10.3778/j.issn.1002-8331.1507-0048.
[8] 管声启,洪奔奔,梁洪,等.高斯差分滤波显著性的刀具磨损检测[J].机械科学与技术,2018,37(2):276-279.DOI:10.13433/j.cnki.1003-8728.2018.0218.
[9] 闫保中,雷雯静.基于改进SIFT算法的目标识别[J].应用科技,2018,45(5):76-81.DOI:10.11991/yykj.201803010.
[10] KHADHRAOUI T,BORGI M A,BENZARTI F,et al.Local generic representation for patch uLBP-based face recognition with single training sample per subject[J].Multimedia Tools and Applications,2018,67(3):24203-24222.DOI:10.1007/s11042-018-5679-0.
[11] TABATABAEI S M,CHALECHALE A,MOGHIMI S.Facial expression recognition using high order directional derivative local binary patterns[C]//2nd International Conference on Pattern Recognition and Image Analysis.Rasht:IEEE Press,2015:1-5.DOI:10.1109/PRIA.2015.7161619.
[12] GUO Yimo,ZHAO Guoying,PIETIK?INEN M.Discriminative features for texture description[J].Pattern Recognition,2012,45(10):3834-3843.DOI:10.1016/j.patcog.2012.04.003.
[13] LIU Zhe,QIU Chengjian,SONG Yuqing,et al.Texture feature extraction from thyroid MR imaging using high-order derived mean CLBP[J].Journal of Computer Science and Technology,2019,34(1):35-46.DOI:10.1007/s11390-019-1897-9.
[14] 刘金培,郭艺,陈华友,等.基于非结构数据流行学习的碳价格多尺度组合预测[J].控制与决策,2019,34(2):279-286.DOI:10.13195/j.kzyjc.2018.0130.
[15] 王振亚,戚晓利,吴保林.基于ASL-Isomap流形学习的滚动轴承故障诊断方法[J].噪声与振动控制,2019,39(2):167-174.DOI:10.3969/j.issn.1006-1355.2019.02.032.

备注/Memo

备注/Memo:
收稿日期: 2019-08-02
通信作者: 曾亮(1980-),男,教授,博士,主要从事机器视觉与人工智能、优化计算方法、调度与优化的研究.E-mail:zengliang@hbut.edu.cn.
基金项目: 国家自然科学基金资助项目(61601176, 41601394); 湖北工业大学博士科研启动基金资助项目(BSQD2017008)
更新日期/Last Update: 2020-07-20