基于机器视觉的刚性接触网绝缘子病害检测系统  

Machine Vision-Based Detection System for Insulator Defects in Rigid Catenary Systems

作　　者：曾光 黄健盛 武文星 佟景泉[1] 黄杨灵 ZENG Guang;HUANG Jiansheng;WU Wenxing;TONG Jingquan;HUANG Yangling(Guangdong Communication Polytechnic,Guangzhou,Guangdong 510630,China)

机构地区：[1]广东交通职业技术学院,广东广州510630

出　　处：《自动化应用》2024年第2期151-155,共5页Automation Application

基　　金：2021年度广东省普通高校特色创新项目(2021KTSCX223);2021年广东省科技创新战略专项资金(“攀登计划”专项资金)项目(pdjh2022b0855,pdjh2022b0854);2023年广东省科技创新战略专项资金(大学生科技创新培育)项目(pdjh202360852)。

摘　　要：针对刚性接触网绝缘子病害造成的城市轨道交通弓网故障,以及在雨雾环境下对绝缘子裂纹病害检测的工程需求,设计了基于机器视觉的刚性接触网绝缘子病害检测系统。基于深度摄像头和激光雷达技术,采用暗通道先验算法预处理在雨雾环境下的图像;采用SURF算法实现绝缘子的快速特征识别;采用小波变换和维纳滤波技术进一步优化图像;采用改进型Canny算法提取绝缘子裂纹病害边缘,使得系统对绝缘子病害检测的准确率达到92.5%,误判率仅为5%。A machine vision-based detection system for insulator defects in rigid catenary system is designed to address urban rail transit pantograph faults caused by rigid contact line insulator diseases,as well as the engineering requirements for detecting insulator crack diseases in rainy and foggy environments.Based on depth camera and LiDAR technology,a dark channel prior algorithm is used to preprocess images in rainy and foggy environments.Using SURF algorithm to achieve fast feature recognition of insulators,and further optimize the image using wavelet transform and Wiener filtering techniques.The improved Canny algorithm was used to extract the edges of insulator crack defects,resulting in an accuracy rate of 92.5%and a misjudgment rate of only 5%for insulator defect detection in the system.

关 键 词：机器视觉 刚性接触网 绝缘子 裂纹病害 图像处理 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]