Screening of Zero-Value Insulators Infrared Thermal Image Features Based on Binary Logistic Regression Analysis

The zero-value insulator is one of the important reasons causing line failures. The key part of detecting zero-value insulators by infrared thermal imaging is to pick up the sensitive infrared thermal image character parameters. A method based on binary logistic regression analysis is proposed to screen the infrared thermal image characteristics of zero-value insulators in the paper. First of all, the infrared thermal image was denoised with the combination of wavelet transform and mean filtering method, and the contrast of the image was enhanced by histogram equalization method. Then the image was segmented to binary image by maximum variance method. After the minimum enclosing rectangle around the disc surface of insulator string was intercepted automatically, 13 texture features of rectangle area were extracted. Finally, 14 characteristic parameters consisting of pollution class and texture features were screened by binary logistic regression analysis and 7 parameters among them having significant influence on the classification result was gotten. The experiment shows that the proposed selection method is simple, and can effectively screen characteristic parameters.