航拍图像中绝缘子缺陷检测和定位.PDFVIP

Vo 1. 43 No. 6A 第 43 卷第6A 期 计算机科学 2016 年 6 月 Computer Science June 2016 航拍图像中绝缘子缺陷的检测与定位 方挺韩家明 (安徽工业大学电气与信息工程学院 马鞍山 243000) 摘 要 长期暴露在自然环境下的绝缘子易产生自爆缺失故障，对此提出了一种绝缘子缺陷检测和定位的图像处理 方法。依据绝缘子在航拍图像中的型态特征，采用最大类问方差法以及中值滤泼进行图像预处理，提出基于粒子群优 化参数的蚁群算法来检测其中的绝缘子轮廓及其数目，最终实现在原图中标记出缺陷绝缘子所在位置的目的。该方 法针对简单背景下绝缘子缺陷的标记具有良好的效果，并为进一步实现复杂背景下绝缘子缺陷检测和定位提供了预 研基础。 关键词 最大类间方差法，粒子群算法，蚁群算法，绝缘子缺陷检测和定位 中固法分类号 TP39 1. 4 文献标识码 A Detection and Localization of Insulator Defects in Aerial hnages FANG Ting HAN Jia-ming (School of Electrical and Information Engineering ,Anhui University of Technology ,Maanshan 243000 ,China) Abstract The insulator is easy to result in self-shattering fault if exposed in the natural environrnent for long time. For this situation ,this paper presented an image processing method for insulator defect detection and location. Based on the pattern characteristics of insulator in aerial image ,the between-cluster variance method and median filtering were used to preprocess the image. The ant colony algorithm with optimized parameters obtained by particle swa口n optimization algorithm was proposed to detect the contour and nurnber of insulator in the aerial image. At last ,the location of defec tive insulator was marked in the original image. This method has good effect for defective insulator detection under sim ple background ,and establishes pre-research foundation to realize insulator defect detection and localization under com plex background. Keywords Between-cluster variance method , Particle swa口n optimization algorithm , Ant