| 基于纹理特征的钢丝绳图像分割方法 |
| |
| 引用本文： | 孙慧贤,张玉华,罗飞路.基于纹理特征的钢丝绳图像分割方法[J].光电工程,2009,36(4). |
| |
| 作者姓名： | 孙慧贤 张玉华 罗飞路 |
| |
| 作者单位： | 国防科技大学,机电工程与自动化学院,长沙,410073 |
| |
| 摘 要： | 针对复杂背景下钢丝绳图像难以准确分割的问题,提出一种新的基于纹理特征的图像分割方法.首先,采用局部二进制模式(Local Binary Pattern,LBP)特征直方图的一阶熵、二阶熵作为LBP特征的统计测度,降低LBP特征的维数.同时选用边缘密度作为纹理描述的特征之一,弥补LBP算子提取纹理特征不足,抗干扰能力差的缺点.然后以上述纹理特征构成特征矢量,采用模糊C-均值(Fuzzy C-Mean,FCM聚类算法进行聚类分割.在实验中,对比了该算法与灰度共生矩阵、传统LBP算子在钢丝绳图像分割中的效果.结果表明,该算法可以有效地对钢丝绳图像进行纹理分割,并能取得良好的边界定位效果,性能优于另外两种算法.
|
| 关 键 词： | 钢丝绳 纹理分割 局部二进制模式 边缘密度 模糊C-均值聚类 |
Wire Rope Image Segmentation Method Based on Texture Features |
| |
| SUN Hui-xian,ZHANG Yu-hua,LUO Fei-lu.Wire Rope Image Segmentation Method Based on Texture Features[J].Opto-Electronic Engineering,2009,36(4). |
| |
| Authors: | SUN Hui-xian ZHANG Yu-hua LUO Fei-lu |
| |
| Affiliation: | SUN Hui-xian,ZHANG Yu-hua,LUO Fei-lu(College of Mechatronics Engineering and Automation,National University of Defense Technology,Changsha 410073,China) |
| |
| Abstract: | A novel texture image segmentation method is proposed to segment the wire rope in complex background.Firstly,the Local Binary Pattern(LBP) operator is used to extract local texture feature histogram of sub windows in an image.Then,these histograms are measured statistically by the first-order entropy and second-order entropy to reduce the dimension of LBP characteristic.At the same time,edge density is combined to describe the texture feature.Based on the three features,the Fuzzy C-Mean clustering algorithm... |
| |
| Keywords: | wire rope texture segmentation local binary pattern edge density fuzzy C-mean clustering |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
