International Scientific Conference (UNITECH 2017), Gabrovo, Bulgaria, 17 - 18 November 2017, pp.179-185
In this study, vented water trees in XLPE samples were initiated and grown in laboratory environment. After aging, vented type water treeing images were obtained by using light microscope. An image segmentation based on morphological image processing and color image processing has been applied to clean dye residues and/or other spots and insignificant regions in the image. Component analysis was performed for RGB and HSV color spaces of images before morphological image processing. Appropriate components were selected to recognize the areas with the vented water treeing. Thresholding has been applied for selected components. Finally, thresholded images were segmented using morphological image processing.