自适应特征引流管故障智能识别方法

黄博; 江慎旺; 张增; 张静; 张巍; 许廷发

doi:10.3788/CO.20171003.0340

自适应特征引流管故障智能识别方法

doi: 10.3788/CO.20171003.0340

黄博^1,,
江慎旺¹,
张增²,
张静²,
张巍³,
许廷发^1, ,

1.
北京理工大学光电学院光电成像系统与技术教育部重点实验室, 北京 100081
2.
天津航天中为数据系统科技有限公司, 天津 300301
3.
南方电网科学研究院有限责任公司, 广州 510080

基金项目:

南方电网直升机重大专项资助项目 K-KY2014-500

详细信息

作者简介:
黄博(1993-), 男, 重庆长寿人, 硕士研究生, 主要从事光电成像探测与识别、目标跟踪、光电测量和可见光通信方面的研究。E-mail:a1039377853@163.com

许廷发(1968—)，男，黑龙江肇东人，博士，教授，博士生导师，1992年、2000年于东北师范大学分别获得学士、硕士学位，2004年于中国科学院长春光学精密机械与物理研究所获得博士学位，2006年于华南理工大学博士后出站，主要从事光电成像探测与识别等方面的研究。E-mail:xutingfa@163.com

通讯作者:
许廷发, E-mail:xutingfa@163.com

中图分类号: TP317.4;TN219
计量
- 文章访问数: 1314
- HTML全文浏览量: 325
- PDF下载量: 628
- 被引次数: 0
出版历程
- 收稿日期: 2017-01-19
- 修回日期: 2017-03-27
- 刊出日期: 2017-06-01

Intelligent identification algorithm of adaptive feature drainage tube fault

1.
Key Laboratory of Optoelectronic Imaging System and Technology, Ministry of Education, School of Optoelectronics, Beijing Institute of Technology, Beijing 100081, China
2.
Tianjin Aerospace Science and Technology Co., Ltd. for the Data System, Tianjin 300301, China
3.
Southern Power Grid Science Research Institute Co., Ltd, Guangzhou 510080, China

Funds:

Helicopter Major Project of China Southern Power Grid K-KY2014-500

摘要

摘要: 为了实现对高压输电线存在的故障隐患进行自动检测，本文提出了一种自适应特征引流管故障隐患智能识别算法。首先，分析了故障引流子的红外热图像特征，把故障分为两类：明显发热和微弱发热；其次，针对引流管所引起的明显发热，采用改进的Otsu阈值分割法对红外图像进行分割，运用改进的Sobel算子提取轮廓；第三，用种子填充算法分离连通域，通过Thread特征判断是否为故障引流管；最后，进入引流管所引起的微弱小区域发热识别，运用高压输电线平行特征寻找主干线区域，在主干线区域检测Harris角点，通过STWN特征判断是否为故障引流子。实验结果表明，发热隐患的识别率为94.6%，漏检率为2.2%，误识别率为5.5%。
- 红外热图像 /
- 边界拓展 /
- 形态特征 /
- 智能识别
Abstract: In this paper, an intelligent recognition algorithm for hidden danger of drainage tube is presented in order to realize the automatic detection of the faults of the high voltage transmission line. First, the thermal image feature of faults is analyzed, and the faults can be divided into two types:obvious heating and weak heating. Second in view of the obvious heating caused by the drainage tube, the improved Ostu threshold segmentation method is used to implement infrared image segmentation and the improved Sobel operator is used to implment contour extraction. Third, the seed filling algorithm separation is used to connect domains, and we can determine whether the drainage tube is fault through the thread characteristics. Finally, we check the weak heating caused by the drainage tube, applying high pressure transmission line parallel features to find the region of trunk line, and then get the Harris corner around the trunk region and determine whether it is fault drainage through the STWN characteristics. Experimental results show that the successful identification rate of hidden heat fault is 94.6%, false negative rate is 2.2%, and false recognition rate is 5.5%.
- infrared thermal image /
- boundary development /
- morphological feature /
- intelligent recognition

HTML全文

图 1 引流管引起明显发热图例

Figure 1. Obvious heating caused by drainage tube

下载: 全尺寸图片幻灯片

图 2 引流管处小区域发热图例

Figure 2. Small area heating caused by drainage tube

下载: 全尺寸图片幻灯片

图 3 阈值分割

Figure 3. Threshold segmentation

下载: 全尺寸图片幻灯片

图 4 Sobel算子边缘检测

Figure 4. Edge detection of Sobel operator

下载: 全尺寸图片幻灯片

图 5 区域划分

Figure 5. Region division

下载: 全尺寸图片幻灯片

图 6 Sobel边缘数据图像

Figure 6. Sobel edge data image

下载: 全尺寸图片幻灯片

图 7 边界拓展原理图

Figure 7. Principle diagram of boundary extension

下载: 全尺寸图片幻灯片

图 8 直接分离与边界拓展后分离对比

Figure 8. Comparison of the images obtained by direct separation and boundary extension

下载: 全尺寸图片幻灯片

图 9 去除噪声后输出

Figure 9. Output after removing noise

下载: 全尺寸图片幻灯片

图 10 输电线发热区域输出

Figure 10. Output of heating zone of transmission line

下载: 全尺寸图片幻灯片

图 11 高压输电线路主干线区域识别

Figure 11. Main line area identification of high voltage transmission line

下载: 全尺寸图片幻灯片

图 12 故障引流子识别

Figure 12. Fault drainage identification

下载: 全尺寸图片幻灯片

图 13 Goldeye P/CL-008 SWIR采集红外热图像

Figure 13. Infrared thermal image acquisition with Goldeye P/CL-008 SWIR camera

下载: 全尺寸图片幻灯片

图 14 故障引流子

Figure 14. Fault drainage

下载: 全尺寸图片幻灯片

图 15 发热极弱、背景复杂导致引流子故障不明显

Figure 15. Drainage fault is not obvious due to the very weak heat and the complicated background

下载: 全尺寸图片幻灯片

图 16 背景噪声造成的错检现象

Figure 16. False detection caused by background noise

下载: 全尺寸图片幻灯片

图 17 阈值选取对检测的影响

Figure 17. Effect of threshold selection on detection

下载: 全尺寸图片幻灯片

表 1 92幅红外图像处理结果

Table 1. Processing results obtained from 92 infrared images

设备总图片数量故障数正确检测到的故障数误检率漏检率

引流子 92 38 33(94.6%) 5.5% 2.2%

下载: 导出CSV

map

参考文献(20)

[1]	陈韬. 直升机在输电线路运检工作中的技术研究[D]. 北京: 华北电力大学, 2010. CHEN T. Research on the technology of helicopter in the operation of transmission line inspection[D]. Beijing:North China Electric Power University, 2010.(in Chinese)
[2]	TANG W Y, HANG Y, QIU Q R, et al.. Overhead Power Line Detection from UAV Video Images[C]. 19th international conference on Mechatronics and Machine Vision in Practice. Auckland, New Zealand, 2012:74-79.
[3]	DU SH ZH, TU CH L. Power line inspection using segment measurement based on HT butterfly[C]. International Conference on Signal Processing, Communications and Computing. Xi'an, 2011:1-4.
[4]	张来明, 郭劲, 杨贵龙, 等.中波红外金宝搏188软件怎么用技术最新进展[J].中国光学, 2013, 6(4):501-512. http://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201304010.htm ZHANG L M, GUO J, YANG G L, et al.. The latest progress in mid infrared laser technology[J]. Chinese Optics, 2013, 6(4):501-512.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-ZGGA201304010.htm
[5]	赵建川, 王弟男, 陈长青, 等.红外金宝搏188软件怎么用主动成像和识别[J].中国光学, 2013, 6(5):795-802. //www.illord.com/CN/abstract/abstract9066.shtml ZHAO J CH, WANG D N, CHEN C Q, et al.. Infrared laser active imaging and recognition[J]. Chinese Optics, 2013, 6(5):795-802.(in Chinese) //www.illord.com/CN/abstract/abstract9066.shtml
[6]	唐攀龙. 高压输电线路绝缘子红外监测方法的研究[D]. 长沙: 长沙理工大学, 2009: 8-11. TANG P L. Research on infrared monitoring method of high voltage transmission line insulator[D]. Changsha:Changsha University of Science and Technology, 2009:8-11.(in Chinese)
[7]	陈豪, 陈原, 赵雪松, 等.红外测温技术在复合绝缘子检测中的应用[J].电力设备, 2006, 7(9):42-43. http://www.cnki.com.cn/Article/CJFDTOTAL-ZXDB201408029.htm CHEN H, CHEN Y, ZHAO X S, et al.. Application of infrared temperature measurement technology in the detection of composite insulators[J]. Electric Power Equipment, 2006, 7(9):42-43.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-ZXDB201408029.htm
[8]	王灿进, 石宁宁, 孙涛.同态非局部滤波在金宝搏188软件怎么用主动成像散斑抑制中的应用研究[J].液晶与显示, 2016, 31(2):193-200. http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201602011.htm WANG C J, SHI N N, SUN T. Application of homomorphic non-local filters in speckle noise suppression for laser active imaging[J]. Chinese J. Liquid Crystals and Displays, 2016, 31(2):193-200.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201602011.htm
[9]	黄富瑜, 沈学举, 刘旭敏, 等.基于空时域融合处理检测超大视场红外目标[J].光学精密工程, 2015, 23(8):2328-2338. http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201508027.htm HUANG F Y, SHEN X J, LIU X M, et al.. Detection of super wide-field infrared target based on spatial-temporal fusion processing[J]. Opt. Precision Eng., 2015, 23(8):2328-2338.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201508027.htm
[10]	鲁剑锋.无人机光电载荷图像处理器的设计[J].中国光学, 2011, 4(5):448-452. //www.illord.com/CN/abstract/abstract8718.shtml LU J F. Design of UAV photo electric load image processor[J]. Chinese Optics, 2011, 4(5):448-452.(in Chinese) //www.illord.com/CN/abstract/abstract8718.shtml
[11]	陈冬岚, 刘京南, 余玲玲.几种图像分割阈值选取方法的比较与研究[J].机械制造与自动化, 2003(1):77-80. http://www.cnki.com.cn/Article/CJFDTOTAL-ZZHD200301025.htm CHEN D L, LIU J N, YU L L. Comparison and Research on threshold selection methods of several kinds of image segmentation[J]. Machinery Manufacturing and Automation, 2003(1):77-80.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-ZZHD200301025.htm
[12]	袁小翠, 吴禄慎, 陈华伟.基于Otsu方法的钢轨图像分割[J].光学精密工程, 2016, 24(7):1772-1781. http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201607028.htm YUAN X C, WU L SH, CHEN H W. Rail image segmentation based on Otsu method[J]. Opt. Precision Eng., 2016, 24(7):1772-1781.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201607028.htm
[13]	孟亚州, 马瑜, 白冰, 等.基于粒子群优化的Otsu肺组织分割算法[J].液晶与显示, 2015, 30(6):1000-1007. http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201506017.htm MENG Y ZH, MA Y, BAI B, et al. Improved lung segmentation algorithm based on 2 D Otsu optimized by PSO[J]. Chinese J. Liquid Crystals and Displays, 2015, 30(6):1000-1007.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201506017.htm
[14]	王智文.几种边缘检测算子的性能比较研究[J].制造业自动化, 2012, 34(11):14-16. doi: 10.3969/j.issn.1009-0134.2012.6(s).05 WANG ZH W. A comparative study on the performance of several edge detection operators[J]. Manufacturing Automation, 2012.34(11):14-16.(in Chinese) doi: 10.3969/j.issn.1009-0134.2012.6(s).05
[15]	张正峰, 马少飞, 李玮.新的种子点区域填充算法[J].计算机工程与应用, 2009(6):201-202. http://www.cnki.com.cn/Article/CJFDTOTAL-FDXB200001016.htm ZHANG ZH F, MA SH F, LI W. New seed point region filling algorithm[J]. Computer Engineering and Applications, 2009(6):201-202.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-FDXB200001016.htm
[16]	MUKHOPADHYAY P, CHAUDHURI B B. A survey of hough transform[J]. Pattern Recognition, 2015, 48(3):993-1010. doi: 10.1016/j.patcog.2014.08.027
[17]	王崴, 张宇红, 洪军, 等.一种改进的Harris角点提取算法[J].光学精密工程, 2008, 16(10):1995-2001. doi: 10.3321/j.issn:1004-924X.2008.10.034 WANG W, ZHANG Y H, HONG J, et al.. An improved Harris corner detection algorithm[J]. Opt. Precision Eng., 2008, 16(10):1995-2001.(in Chinese) doi: 10.3321/j.issn:1004-924X.2008.10.034
[18]	李海, 张宪民, 陈忠.基于直线检测的棋盘格角点自动提取[J].光学精密工程, 2015, 23(12):3480-3489. http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201512026.htm LI H, ZHANG X M, CHEN ZH. Automatic extraction of corner points of a chess board based on line detection[J]. Opt. Precision Eng., 2015, 23(12):3480-3489.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201512026.htm
[19]	刘博超, 赵建, 孙强.基于边缘改进的Harris角点检测方法[J].液晶与显示, 2013, 28(6):939-942. http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201306031.htm LIU B CH, ZHAO J, SUN Q. Improved Harris corner detection method based on edge[J]. Chinese J. Liquid Crystals and Displays, 2013, 28(6):939-942.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201306031.htm
[20]	杨云, 岳柱.基于融合图像轮廓矩和Harris角点方法的遮挡人体目标识别研究[J].液晶与显示, 2013, 28(2):273-277. http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201302029.htm YANG Y, YUE ZH. Human body target recognition under occlusion based on fusion of image contour moment and Harris angular points[J]. Chinese J. Liquid Crystals and Displays, 2013, 28(2):273-277. (in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-YJYS201302029.htm