Fault localisation of electrical equipments using thermal imaging technique

Electrical Fallures caused by excess temperature, excess current or voltage, ionization radiation, mechanical shock, stress or impact needs strong attention to diagnose. In recent years, thermal inspection mechanism is widely used to find hot spots caused by defects in electrical connections and components. This paper addresses an important technique for thermal image segmentation to detect the electrical failure. Earlier works like Otsu method provide satisfactory results and optimal only for thresholding bimodal/multimodal images, fails for unimodal images. Hence in this paper, valley-emphasis method which weights the valley points in the histogram is proposed for the electrical fault analysis problem. The electrical fault has been detected and segmented using both methods.