基于图像形态特征的复合绝缘子憎水性图像分割算法的研究

摘要

This paper mainly focuses on the research of image segmentation,which is a certain part of image processing for hydrophobic images.In the process of image segmentation,due to the difference of morphological characteristics between insulator level HC1 and other levels’images,the same algorithm may not suitable for images of all levels in practical image processing.In this paper,we separate the handling of images in HC1 level and the images in HC2-HC7 levels,and adopt different morphological operations respectively,then use the improved genetic algorithm to segment the images and finally obtain accurate image segmentation effect graph.This appraoch provides the basis for accurate judgement of the hydrophobic level of composite insulator.%对憎水性图像在图像处理中的图像分割环节进行研究。在图像分割处理过程中，因绝缘子等级HC1与其他等级图像的形态特征的不同，导致在实际处理图像时，同一算法不能适用于所有等级的图像。将HC1等级的图像同HC2－HC7等级的图像分开处理，分别采用不同的形态学操作，再利用改进的遗传算法进行图像分割，最终得到准确的图像分割效果图。为复合绝缘子憎水性等级的准确判断提供条件。