摘要
在分析输电线路部件图像样本的特征的基础上,提出了一种基于熵值的自动阈值图像分割方法.该方法利用图像的信息熵值来自动选择图像的分割阈值,将由输电线路主要部件构成的前景图像分割出来.实验结果表明,该算法具有实现简单、计算开销小、运算速度快的优点,能满足以天空为主要背景的输电线路自动巡视的高分辨率图像实时预处理的需要.
This paper presents a method of image segmentation based on entropy for automatically calculating threshold, after analyzing features of the samples of transmission line images The method could automatically calculat threshold for image segmentation based on image＇s information entropy, and it does a good job at segmenting the foreground consisting of the main parts of the transmission line from the background. The result of experiment indicates that the algorithm can satisfy the need when doing the real-time pre-processing of the transmission line＇s high resolution images got through automatic inspection whose main part of background consists of sky, and it also proves that the algorithm is simple to achieve and has advantages of lower computational overhead and higher speed.
出处
《应用科技》
CAS
2011年第8期1-4,共4页
Applied Science and Technology
基金
国家自然科学基金资助项目(60674100)
江苏省高校自然科学基金资助项目(09KJB470002.)
关键词
输电线路
自动巡视
图像分割
熵值直方图
自动阈值
transmission line
automatic inspection
image segmentation
entropy histogram
adaptive threshold
