Search

This paper presents results of a technique developed to measure the volume and the surface area of flaws in pin-type polymer insulators used in electrical energy distribution networks by means of computed tomography (CT). The tomography images were reconstructed using 180 digital radiographs (projections) acquired from a high-resolution system (a 50 μm × 50 μm, a-Si pixel detector). An acrylic phantom containing 12 cylindrical holes of known volume was used to test the tomography measurement system. In order to improve the quality of tomography images, two procedures were used, namely, ‘subtraction of dark image’ and ‘correction of light field’, as well as a sequence of filters such as the median and the Gaussian. Image processing techniques were used for the segmentation of flaws in the CT images. The Marching Cubes algorithm was used in the visualisation process and the Decimation algorithm was used to obtain volume and surface area measurements of flaws. The programs were developed in C++ and Java, using the VTK – Visualisation Toolkit library. The results obtained showed to be satisfactory for quantitatively and qualitatively analysing the effects of electrical discharges in insulation materials such as insulators and cables. [ABSTRACT FROM AUTHOR]