RTV绝缘子憎水性图像增强新方法及应用  

A New Algorithm for RTV Insulator Hydrophobic Image Enhancement and Application

作　　者：廖平军 LIAO Pingjun(State Grid Zhejiang Tongxiang City Power Supply Company, Jiaxing 314000, China)

机构地区：[1]国网浙江桐乡市供电公司,浙江嘉兴314000

出　　处：《电瓷避雷器》2018年第3期139-143,共5页Insulators and Surge Arresters

基　　金：高等学校博士学科点专项科研基金课题资助(编号:20120161110009)

摘　　要：在研究RTV绝缘子憎水性等级过程中,低照度的RTV绝缘子憎水性图像对等级的判定有一定影响,故需对其增强处理。然而,目前传统增强法以及Retinex增强法在处理该类图像都存在一定的问题。为此,提出一种基于改进双边滤波Retinex算法,该算法首先对经典双边滤波进行改进,对其空间邻度因子最优化、在水平和垂直方向上降维滤波以及自适应求解σr;然后将改进双边滤波对RTV绝缘子憎水性图像照度估计,与Retinex算法相结合,最终达到增强的目的。实验表明,该算法对RTV绝缘子憎水性图像增强效果明显,在RTV绝缘子憎水性等级判定应用中提高了其准确率。In the study of the RTV insulator hydrophobic sizing process, it＇s necessary to enhance the image of the low illuminated RTV insulator hydrophobic. However, at present there are some problems in processing of the image with the traditional algorithms and the Retinex algorithm for image en- hancement. Therefore , a Retinex algorithm based on improved bilateral filtering is proposed, Firstly, this algorithm improves the classic bilateral filtering, optimizes the space adjacent factor, make a dimension reduction both in horizontal and vertical direction filter and solve σr by self-adaption; Then this algorithm uses the improved bilateral filter to estimate the RTV insulator hydrophobic image illuminance, and combines with the Retinex algorithm, Finally it achieves the purpose of enhancing. The experimental results show that the algorithm has a good effect of enhancement, reduces the filtering computation and plays an important role in improving the RTV insulator hydrophobic classification accuracy.

关 键 词：双边滤波 RETINEX 图像增强 憎水性 

分 类 号：TM216[一般工业技术—材料科学与工程]