[1]吴文海,孙 磊,王国志,等.基于近义词分配的铁路接触网绝缘子识别与分类[J].电瓷避雷器,2020,(01):156-160.[doi:10.16188/j.isa.1003-8337.2020.01.026]
 WU Wenhai,SUN Lei,WANG Guozhi,et al.Insulator Identification and Classification of Railway Catenary Based on Homoionym-Assignment[J].,2020,(01):156-160.[doi:10.16188/j.isa.1003-8337.2020.01.026]
点击复制

基于近义词分配的铁路接触网绝缘子识别与分类()
分享到:

《电瓷避雷器》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2020年01期
页码:
156-160
栏目:
绝缘子
出版日期:
2020-02-20

文章信息/Info

Title:
Insulator Identification and Classification of Railway Catenary Based on Homoionym-Assignment
作者:
吴文海 孙 磊 王国志 张 霆
(西南交通大学机械工程学院, 成都 610031)
Author(s):
WU Wenhai SUN Lei WANG Guozhi ZHANG Ting
(School of Mechanical Engineering Southwest Jiaotong University, Chengdu 610031, China)
关键词:
绝缘子 视觉词包模 型潜在狄利克雷分布模型 相对熵 近义词
Keywords:
insulator bag of words model latent dirichlet allocation the relative entropy homoionym
DOI:
10.16188/j.isa.1003-8337.2020.01.026
摘要:
为提高铁路接触网绝缘子检测中的图像识别精度并准确区分绝缘子类型,提出一种建立在“视觉词包模型”基础上,计算视觉单词语义距离并合理分配的绝缘子识别分类模型。首先应用潜在狄利克雷分布模型与相对熵计算视觉单词的语义相关性,然后根据底层特征与各单词的欧氏距离大小分配单词数目,最后采用支持向量机实现绝缘子的识别与分类。实验结果表明,该方法可以提高识别精度与分类效果。
Abstract:
In order to improve the accuracy of image recognition in the detection of insulators on railway catenary, and also to distinguish insulator types accurately, an insulator classification model based onBag of words modeland calculating semantic distance between visual words and assigning them reasonably is proposed. Firstly, Latent Dirichlet Allocation and the relative entropy are introduced to calculate the semantic relevance of visual words. Then the number of words is allocated according to the Euclidean metric between low-level features and each word. Finally, the recognition and classification of insulators are achieved by using a support vector machine. Experimental results show that this method can improve the recognition accuracy and classification effect.

参考文献/References:

[1] 贺博,林辉.绝缘子污秽闪络的研究现状及思索[J].电瓷避雷器,2006(2):7-11. HE Bo, LIN Hui. Present status and future prospects for flashover research of contaminated insulators[J]. Insulators and Surge Arresters, 2006(2): 7-11.
[2] 梁志豪.绝缘子污闪的形成机理及预防措施[J].广东电力,2013, 26(6):87-92. LIANG Zhihao. Formation mechanism of insulator pollution flashover and preventive measures[J]. Guangdong Electric Power, 2013, 26(6): 87-92.
[3] 律方成,刘杰,石鹏,汪佛池,熊俊,李聪.复合绝缘子故障检测方法的有效性分析[J].高压电器,2013, 49(5):95-100. LV Fangcheng, LIU Jie, DAN Peng, WANG Fochi, XIONG Jun, LI Cong. Experimental research on composite insulator fault detection[J]. High Voltage Apparatus, 2013, 49(5): 95-100.
[4] 姜浩然,金立军,闫书佳.航拍图像中绝缘子的识别与故障诊断[J].机电工程,2015(2):274-278. JIANG Haoran, JIN Lijun, YAN Shujia. Recognition and fault diagnosis of insulator string in aerial images[J]. Mechanical & Electrical Engineering Magazine, 2015(2): 274-278.
[5] 于兰英,姚波,吴文海,刘桓龙,谭万秋.一种基于多特征的绝缘子识别方法[J].电瓷避雷器,2016(3):79-83. YU Lanying, YAO Bo, WU Wenhai, LIU Huanlong, TAN Wanqiu. Insulator identification method based on multi-feature[J]. Insulators and Surge Arresters, 2016(3): 79-83.
[6] 李卫国,叶高生,黄锋,王世坤,常文治.基于改进MPEG-7纹理特征的绝缘子图像识别[J].高压电器,2010, 46(10):65-68. LI Weiguo, YE Gaosheng, HUANG Feng, WANG Shikun, CHANG Wenzhi. Recognition of insulator based on developed MPEG-7 texture feature[J]. High Voltage Apparatus, 2010, 46(10): 65-68.
[7] 葛玉敏,李宝树,赵书涛,梁爽.基于航拍图像的绝缘子表面状态检测[J].高压电器,2010, 46(4):65-68, 73. GE Yumin, LI Baoshu, ZHAO Shutao, LIANG Shuang. A method baced on aerial images to detect the surface state of insulators[J]. High Voltage Apparatus, 2010, 46(4): 65-68, 73.
[8] 林文文.基于双目视觉的接触网绝缘子识别定位[D].成都:西南交通大学,2017.
[9] MEI Xin, LU Tiecheng, WU Xiaoyun, ZHANG Bo. Insulator surface dirt image detection technology based on improved watershed algorithm[C]2012 Asia-Pacific Power And Energy Engineering Conference(APPEEC).[S.l.]:[s.n.], 2012: 1-5.
[10] 姚春羽,金立军,闫书佳.电网巡检图像中绝缘子的识别[J].系统仿真学报,2012, 24(9):1818-1822. YAO Chunyu, JIN Lijun, YAN Shujia. Recognition of insulator string in power grid patrol images[J]. Journal of System Simulation, 2012, 24(9): 1818-1822.
[11] WEINSHALL D, HANUKAEV D, LEVI G. LDA topic model with soft assignment of descriptors to words[C]// International Conference on Machine Learning. 2013.
[12] 赵永威,周苑,李弼程,柯圣财.基于近义词自适应软分配和卡方模型的图像目标分类方法[J].电子学报,2016, 44(9):2181-2188. ZHAO Yongwei, ZHOU Yuan, LI Bicheng, KE Shengcai. Image object classification method with homoionym based adaptive Soft-Assignment and Chi-Square model[J]. Acta Electronica Sinica, 2016, 44(9): 2181-2188.
[13] 石晶,李万龙.基于LDA模型的主题词抽取方法[J].计算机工程,2010, 36(19):81-83. DAN Jing, LI Wanlong. Topic words extraction method based on LDA model[J]. Computer Engineering, 2010, 36(19): 81-83.
[14] 赵萌,邱菀华,刘北上.基于相对熵的多属性决策排序方法[J].控制与决策,2010, 25(7):1098-1100, 1104. ZHAO Meng, QIU Wanhua, LIU Beishang.Relative entropy evaluation method for multiple attribute decision Making[J].Control and Decision[J]. Control and Decision, 2010, 25(7): 1098-1100, 1104.
[15] CHANG C C, LIN C J. LIBSVM: A library for support vector machines[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3, SI): 27.
[16] UsiholoIruansi, Jules R Tapamo, Innocent Davidson. Insulator condition using local binary patterns combined with support vector machines[C].IEEE Africon 2017 Proceedings. 2017:21-25.

备注/Memo

备注/Memo:
收稿日期:2018-05-21作者简介:吴文海(1979—),男,博士,主要从事机电液一体化研究与设计。
更新日期/Last Update: 2020-02-20