In respect that the high risk of flashover may cause power outage

it is urgent to realize the safe and accurate monitoring of insulator contamination severity to prevent flashover and ensure operation of power system. In this paper

a method based on visible image features of contaminated insulators and support vector machine is put forward by establishing a mapping between the contamination grade and the color features for the identification purpose of insulator contamination grade. Firstly

the improved image segmentation method based on the seed region growing method is adopted to achieve the discal surfaces of the red porcelain insulators with different contamination grades in the transformer substations in Shenzhen City. Then

thirty-six contamination features are extracted in RGB and HSV color space of the contaminated insulator images and the mean and median of S are selected as the feature values for the insulator contamination grade according to the Fisher criterion. Finally

a multi-class SVM for classification decision is designed. Experimental results show that the identification accuracy of the proposed method reaches 96.67%.