基于嵌入式的复合绝缘子憎水性检测方法研究  

Research of Hydrophobicity Measuring Method for Composite Insulators based on Embedded

作　　者：董政呈 方彦军[1] 王先培[2] 王泉德[2] 

机构地区：[1]武汉大学动力与机械学院,武汉430072 [2]武汉大学电子信息学院,武汉430072

出　　处：《绝缘材料》2015年第7期67-72,共6页Insulating Materials

基　　金：国家自然科学基金资助项目(50677047);湖北省自然科学基金资助项目(2010CDZ051)

摘　　要：为了客观、快速、准确地测量在线或离线复合绝缘子表面的憎水性,依据憎水性分级法(HC),设计了一套便携式绝缘子憎水性检测装置。该装置通过一系列简单、有效的图像处理和分类方法,实现了嵌入式平台上的绝缘子憎水性判断。首先采用一种基于边缘的图像分割方法提取水珠目标,然后采用一种综合决策树实现目标的训练和分类。结果表明:该装置能在无辅助光源的自然光条件下判定复合绝缘子的憎水性,且处理速度快、实现方便,避免了人工观测的主观性和不确定性,具有较高的精度和实用性。In order to measure the surface hydrophobicity of online or offline composite insulators objectively, quickly, and accurately, we proposed a portable hydrophobicity dedection device according to the hydrophobicity classification method（HC）. This device can realize hydrophobicity judgment through a series of simple and effective image processing and classification methods that can be implemented on embedded platform. First, a novel image segmentation method based on edge detection is applied to extract droplet, and then a comprehensive decision tree is used to realize training and classifying of target. The results show that the device can measure the hydrophobicity of composite insulators under natural light without auxiliary spotlights. It has fast processing speed and easy implementation. This device has higher precision and practicability, avoiding the subjectivity and uncertainty of manual operation.

关 键 词：憎水性 自适应阈值分割 边缘检测 数学形态学 特征描述 决策树 

分 类 号：TM391[电气工程—电机]