Evaluating the Thermal Condition of Electrical Equipment via IRT Image Analysis

The integrity of electrical power equipment is of paramount importance when it supplies electricity throughout a facility. However, the reliability of the equipments will degraded after sometime, and appropriate maintenance has to be taken accordingly to avoid future faults. Infrared thermography (IRT) image analysis is a commonly used technique for diagnosing the reliability of electrical equipments. Conventionally, the analysis of infrared image is done manually and takes very long time for further analysis. This paper proposes an automatic thermal fault detection and classification system for evaluating the condition of electrical equipment by analyzing its infrared image. First, the image is segmented to find the target region of interest (ROI). The detected regions which have the same region properties are grouped together in order to remove the unwanted regions. Finally, statistical features from each detected region are extracted and classified using the support vector machine (SVM) algorithm. The thermal condition of electrical equipments is evaluated based on qualitative measurement technique. The experimental result shows that the proposed system can detect and classify the thermal condition of electrical equipments.