Research on technology of contact insulator pollution classification and pollution level division based on machine vision

Insulator is an important guarantee for the safe operation of electrified railway, but because of working conditions, the insulator would face long-term effects of wind, snow and rain. Insulator surface would easily accumulate filth and pollution flashover occurs, which often leads to blackouts and threatens the safe operation of the railway. Therefore, detection of the insulator working state is very important. In this paper, the HSV model is applied to insulator image segmentation. Then, the features of insulator surface are extracted and the result would be used to classify pollution. Finally, cluster analysis is used to determine the pollution level of insulator.