基于分水岭算法的绝缘子串红外图像分割方法  

Infrared Image Segmentation of Insulator Strings Based on Watershed Algorithm

作　　者：刘正庭 尹骏刚 李凯迪 汪宵飞 王欣 万勋 姚建刚[1] LIU Zhengting;YIN Jungang;LI Kaidi;WANG Xiaofei;WANG Xin;WAN Xun;YAO Jiangang(College of Electrical and Information Engineering,Hunan University,Changsha 410082,China;SGCC Hunan Electric Power Research Institute,Changsha 410007,China)

机构地区：[1]湖南大学电气与信息工程学院,长沙410082 [2]国网湖南省电力科学研究院,长沙410007

出　　处：《电瓷避雷器》2020年第2期216-221,228,共7页Insulators and Surge Arresters

基　　金：国网湖南省电力公司科研基金(编号:5216A517000T)。

摘　　要：为解决500 kV瓷绝缘子串红外图像分割问题,提出了一种基于分水岭算法的图像分割方法。通过调整红外图像G分量消除云雾干扰,图像预处理后采用灰度拉伸法进行图像增强;然后利用Otsu算法将灰度图像二值化,通过Hough变换确定绝缘子串基准线,并据此使用Floodfill算法提取绝缘子串目标区域;角度校正后做欧式距离变换,以局部最小值为标记符避免过分割,再采用分水岭算法进行绝缘子片分割。据此对湖南省某座500 kV变电站700张红外图像进行分析,结果表明,本文所提方法具有良好的自适应性和实用性,可显著提高绝缘子红外图像智能分割准确率。A new segmentation method based on watershed algorithm is proposed for the infrared image of500 kV insulator strings.The interference of cloud and fog is eliminated by adjusting the G component of the infrared image,and the image quality is enhanced by gray processing,median filtering and gray scale stretching method.The gray image is segmented to binary image based on Otsu threshold.The baseline of an insulator string is found by Hough transformation,which is based on the Flood Fill algorithm to extract the insulator string target area.After correcting angle of insulator string,the Euclidean distance transformation is done.The local minimum value is used as a marker to avoid over segmentation.Based on the Watershed algorithm,insulators area is extracted.700 infrared images of a 500 kV substation were analyzed for validation.The results show that the proposed method is adaptive and practical,and it can significantly improve the accuracy of intelligent segmentation of insulator strings.

关 键 词：瓷质绝缘子 红外图像分割 分水岭算法 泛洪算法 霍夫变换 

分 类 号：TM216[一般工业技术—材料科学与工程] TP391.41[电气工程—电工理论与新技术]