摘要
针对人工染污的悬式复合绝缘子,获取不同污秽度下的复合绝缘子表面污秽图像,提出识别表面污秽度的图像处理技术流程。首先将图像灰度化,再进行滤波处理,采用最大类间方差法(Otsu)进行分割得到伞裙间积污区域。接着提取积污区域的RGB、HSV颜色空间的36种特征量,筛选得到与绝缘子盘面污秽程度相关性较高的特征量S均值和S中值。最后利用这两种特征量作为联合判据建立复合绝缘子污秽度识别模型。对40张不同污秽度下的图像样品进行识别,结果表明所提方法的正确识别率高达97.5%,可有效识别复合绝缘子的表面污秽度。
In this paper, surface contamination images of composite insulators under different contamination levels are obtained by artificial contamination test. After image graying and image filtering processing, the maximum inter class variance (Otsu) method is used to segment the contaminated area. After extracting 36 characteristic quantities of RGB and HSV color spaces in contaminated area, then the mean value and mean value of S component which are highly correlated with the contamination level of insulators are obtained. At last, using these two characteristic quantities as the joint criterion, the composite insulator contamination recognition model is established. The recognition results of 40 image samples with different contamination levels show that the correct recognition rate of the method is as high as 97.5%. The method proposed in this paper can effectively identify the surface contamination of composite insulators.
作者
史训涛
毕继凯
阳林
SHI Xuntao;BI Jikai;YANG Lin(Guangzhou Power Supply Bureau Co. ,Ltd. ,Guangzhou 510000,China;School of Electric Power,South China University of Technology,Guangzhou 510640,China)
出处
《电工技术》
2019年第3期1-4,共4页
Electric Engineering
基金
国家自然科学基金资助项目(编号51507067)
关键词
复合绝缘子
污秽度
图像
特征量
HSV
composite insulators
contamination level
images
characteristic quantities
HSV
