Insulator pollution level detection method based on DAG-SVM

The on-line detection of pollution level of transmission line insulators is of great significance for pollution flashover prevention. Therefore, a method of insulator pollution level detection based on hyperspectral image is proposed in this paper. Initially, the original hyperspectral images are corrected using black and white correction and multiple scattering correction to enhance the characteristics of the spectral data for different contamination levels. A method for detecting insulator contamination levels based on DAG-SVM is proposed, achieving accurate detection of insulator contamination levels. The experimental results show that the method proposed in this paper achieves a detection accuracy rate of over 94% for insulators made of different substrates, meeting the needs of practical engineering applications.