Segmentation Algorithm of Hydrophobic Image Based on Improved Canny Operator
Wang Fo-ch · Gaoya dianqi · 2014
Hydrophobicity measurement is a vital way to evaluate the performance of insulators, which is important for safe operation of an electric power system. However, the traditional segmentation algorithm is unsatisfactory in classification accuracy of hydrophobicity. Therefore, a hydrophobic image segmentation algorithm based on the improved Canny operator is proposed in this paper. Considering the complexity of hydrophobic image, the image is enhanced by adaptive histogram equalization, then the improved Canny operator is used to extract the edge of water drops(or water marks). Finally,some methods in mathematical morphology, such as region filling, expansion, corrosion and removing, are used to process the image. Experiments show that the proposed segmentation algorithm is suitable for all kinds of hydrophobic images, and it may provide a criterion for hydrophobicity grade of an insulator.