摘要
绝缘子作为输电线路中最重要的基础设施之一,对其准确识别是实现输电线路运行状态的自行监测与故障诊断的重要前提。为了能够对无人机航拍巡检中的绝缘子进行准确识别,提出基于红蓝色差和改进K-means算法的航拍绝缘子分类识别方法。首先,结合红蓝色差灰度化和加权灰度化,采用改进K-means算法对灰度图像进行聚类分割;其次,通过形态学滤波弥补分割缺陷;最后,根据绝缘子目标区域的红蓝色差均值,将绝缘子的分类问题简化为一维数据分类问题,从而实现分类识别。实验结果表明,该方法对复杂背景及不同拍摄角度下的绝缘子均能快速进行准确的分类识别,总识别率可达94.4%,为无人机巡检中输电线路绝缘子的分类识别提供了新的思路。
Insulator is a part of the most significant infrastructure in transmission lines. The accurate identification is the essential premise to realize the monitor of running states and fault diagnosis of transmission lines. To improve the accuracy of insulator recognition in the Unmanned Aerial Vehicles（UAV） aerial inspection, we put forward a kind of aerial insulator classification recognition method based on the red-blue difference and developed K-means algorithm. First, the original images were processed by the red-blue difference graying combined with the weighted graying, then the improved K-means algorithm was used for the clustering segmentation of gray images. Second, morphological filtering was used to make up the defect of the segmentation results. Finally, according to the average of the red-blue difference of insulator target area, the classification of insulators is simplified to one-dimensional data classification. The results show that the method can accurately recognize and classify the insulators under complex background and in different shooting angles, its recognition rate can reach 94.4% and recognition speed is fast, which provides a new way for the insulator recognition and classification in the UAV inspection of the transmission lines.
作者
黄新波
刘新慧
张烨
李菊清
张慧莹
邢晓强
HUANG Xinbo1, LIU Xinhui1, ZHANG Ye2, LI Juqing1, ZHANG Huiying1, XING Xiaoqiang1(1. College of Electronics and Information, Xi＇an Polytechnic University, Xi＇an 710048, China; 2. College of Mechanical and Electrical Engineering, Xi＇an Electronic and Science University, Xi＇an 710071, Chin)
出处
《高电压技术》
EI
CAS
CSCD
北大核心
2018年第5期1528-1534,共7页
High Voltage Engineering
基金
陕西省重点科技创新团队计划(2014KCT-16)
陕西省协同创新计划项目(2014XT-07)
陕西省工业科技攻关项目(2016GY-052)
西安市科技计划项目(2017074CG/RC037(XAGC003))
陕西省教育厅专项科研计划(17JK0322)
陕西省自然科学基础研究计划(2017JQ6054)~~
关键词
绝缘子
红蓝色差
灰度化
形态学滤波
K-MEANS
分类识别
insulator
the red-blue difference
graying
morphological filtering
K-means
classification recognition
