The Fault Diagnosis of Infrared Bushing Images Based on Infrared Thermography

摘要

The aim of this paper is to diagnose the bushing faults by making use of the infrared images. Before diagnosing the bushing faults, the preprocessing and segmentation should be processed. The preprocessing, which includes adaptive median filtering, K-means, opening operation, is applied to filter the noise and remove the background. To segment the bushing from the image, simple linear iterative clustering (SLIC) is applied which allows the bushing segmented according to the shape of it and makes it works better. After analyzing the thermal characteristics of the fault image of bushings from the DL/T-664-2016 Application rules of infrared diagnosis for live electrical equipment, it can be concluded that diagnose the fault area's relative position can be a great way to diagnose the bushing, so it's regarded as a principle to classify the faults. In this paper, we selected some typical infrared bushing images with faults. After the processing, the results have been verified that the method is able to segment the bushing and diagnose the fault preliminarily.