摘要
针对直升机电力巡检拍摄到的实时红外视频序列,寻找定位高温点并实时反馈给控制系统,首先对其进行Hough变换检测输电线,然后采用Otsu自适应阈值算法对红外图像中的热点区域进行分割,提取出缺陷区域,接着利用SIFT特征匹配识别红外图像中的绝缘子,最后对缺陷进行分类和分级。实验证明该算法发热点定位准确率较高,可智能识别缺陷,减轻了人工作业负荷。
For real-time infrared video sequence captured in the patrol inspection with helicopter,the high-temperature points are sought and located and then fed back to the control system in real time. First,the Hough Transformation is used to detect the transmission lines. Second,the Otsu adaptive threshold algorithm is used to segment the hot spot in the infrared image. The fault area is extracted,and then the insulator in the infrared image is identified using SIFT feature matching. Finally,the heat faults are classified and classified. The experiment shows that the proposed algorithm has a high accuracy in the hot spot location,can identify defects intelligently,and reduce the manual workload.
作者
张福
张建刚
李庭坚
罗望春
莫兵兵
余德泉
李翔
姜诚
陈佳乐
ZHANG Fu;ZHANG Jiangang;LI Tingjian;LUO Wangchun;MO Bingbing;YU Dequan;LI Xiang;JIANG Cheng;CHEN Jiale(Test Center China Southern Power Grid EHV Maintenance,Guangzhou 510663,Guangdong,China)
出处
《电网与清洁能源》
2018年第3期46-50,共5页
Power System and Clean Energy
基金
南方电网超高压输电公司科技项目(012000-KK52120008)~~
关键词
红外图像
电力巡检
HOUGH
变换
SIFT
缺陷识别
infrared image
power patrol inspection
Hough transformation
SIFT
fault recognition