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.
维普
万方数据