输电线路绝缘子缺陷目标图像视觉检测方法研究  

Research on Target Image Visual Detection Method of Insulator Defects on Transmission Lines

作　　者：杨腾 张继伟 蔡澍雨 YANG Teng;ZHANG Jiwei;CAI Shuyu(Ransmission Branch,Yunnan Power Grid Co.,Ltd.,Kunming 650000,China)

机构地区：[1]云南电网有限责任公司输电分公司,云南昆明650000

出　　处：《自动化仪表》2024年第12期19-23,共5页Process Automation Instrumentation

摘　　要：由于输电线路周围环境较为复杂,且绝缘子安装在户外容易受到恶劣天气条件的影响,会导致绝缘子表面覆盖污垢,从而干扰缺陷目标检测的精度。因此,结合图像分割和角点匹配,提出输电线路绝缘子缺陷目标图像视觉检测方法。获取绝缘子图像,通过中值滤波和双线性插值等方法去除噪声和补全图像失真部分,从而提高图像质量。利用角点匹配技术凸显绝缘子缺陷特征,并根据绝缘子长度和水平夹角进行缺陷目标定位。在图像中标定缺陷范围并输出检测结果,从而实现对绝缘子缺陷的准确识别。试验结果表明:所提方法在检测精度上表现卓越;损失函数曲线下降空间短且收敛值低;损失值波动幅度平缓、收敛情况稳定。所提方法在实践应用中的有效性和高精度,为输电线路绝缘子缺陷检测提供了可靠的技术支持。As the environment around transmission lines is more complex,and insulators installed outdoors are susceptible to harsh weather conditions,which can lead to the insulator surface being covered with dirt,thus interfering with the accuracy of defect target detection.Therefore,combining image segmentation and corner point matching,a target image visual detection method of insulator defects on transmission lines is proposed.The insulator image is acquired,and the image quality is improved by removing noise and complementing the distorted part of the image through median filtering and bilinear interpolation and other methods.Corner point matching technique is utilized to highlight the defective features of insulators,and the defective targets are localized according to the length and horizontal angle of insulators.The defect range is marked in the image and the detection results are output,thus realizing the accurate identification of insulator defects.The test results show that the proposed method is excellent in detection accuracy;the loss function curve has a short descent space and a low convergence value;the loss value fluctuation is smooth,and the convergence is stable.The effectiveness and high accuracy of the proposed method in practical application provide reliable technical support for the detection of insulator defects on transmission lines.

关 键 词：输电线路 绝缘子 缺陷目标 图像分割 角点匹配 缺陷检测 

分 类 号：TH862.72[机械工程—仪器科学与技术]