Image Processing Based Insulator Fault Detection Method

Abstract

This paper describes blob detector based burn-mark detection method of the high voltage insulators. The insulator is one of main element of the high voltage transmission line. The PS-70 type insulator is main object of our research. We divided our burn detection method into several steps. In first step image processing focuses on recognition of insulators from giving image. Main image processing algorithm for this step is symmetry detection algorithm and grabcut algorithm. After recognition of insulators from giving image, have to define the region of interest, on what the burn-mark detection algorithm could run. In the end image processing finds the damages on the insulator surface. In this paper, we discuss only about last step of burn detection method. Our current/ongoing detection rate of burn-mark detection algorithm is 80.43%.