基于红外技术的高压绝缘子缺陷智能检测方法
Intelligent detection method of high voltage insulator defect based on infrared technology
宋岸峰 1白杨 2代鑫波 1张志伟1
作者信息
- 1. 郑州祥和集团有限公司,郑州 450006
- 2. 河南工业大学,郑州 450001
- 折叠
摘要
为提高对高压绝缘子缺陷智能检测的智能化和可视化水平,提出基于红外技术的高压绝缘子缺陷智能检测方法.采用红外视觉图像检测技术提取高压绝缘子运行过程中的视觉特征信息.在获取的高压绝缘子图像中,采用图像中存在的模式、图像纹理以及设备异常状态信息进行分析和融合,设计缺陷检测方法.通过将地理信息特征量与检测系统相结合,实现高压绝缘子缺陷定位.同时,通过红外图像的视觉特征点和异常特征点标定技术,标定高压绝缘子存在的故障、缺陷等位置信息;在图像导入及管理、缺陷识别审核、人工标注等各模块的基础上,实现基于红外技术视觉特征识别的高压绝缘子缺陷智能检测.测试结果表明,设计的高压绝缘子缺陷智能检测方法能准确标定高压绝缘子异常状态信息特征量,对缺陷部位检测的均方根误差最低为0.061,对高压绝缘子缺陷目标检测的可靠性和准确性较好.
Abstract
In order to improve the intelligence and visualization level of high voltage insulator defect intelligent de-tection,an intelligent detection method of high voltage insulator defect based on infrared technology is proposed.Infra-red visual image detection technology is used to extract the visual characteristics of high voltage insulators during opera-tion.In the image of the high voltage insulator,the pattern,image texture and equipment abnormal state information in the image are analyzed and fused,and the defect detection method is designed.By combining the characteristic quanti-ty of geographic information with the detection system,the defect location of high voltage insulators is realized.At the same time,through the infrared image visual feature point and abnormal feature point calibration technology,the fault,defect and other location information of the high-voltage insulator are calibrated.On the basis of image import and management,defect identification audit,manual labeling and other modules,the intelligent detection of high voltage insulator defect based on infrared technology visual feature recognition is realized.The test results show that the de-signed intelligent detection method for high voltage insulator defects can accurately calibrate the characteristic quantity of abnormal state information of high voltage insulator,and the root-mean-square error of the detection of defect parts is the lowest 0.061.The reliability and accuracy of the detection of high voltage insulator defects are good.
关键词
红外技术/高压绝缘子/特征点标定/视觉特征/缺陷检测Key words
infrared technology/high voltage insu-lator/feature point calibration/visual features/defect detection引用本文复制引用
基金项目
郑州祥和集团有限公司郑州管城回族区金岱变南高压输电线路迁改工程(ZD-2020-0070)
出版年
2024
NETL
NSTL
万方数据