| 李俊峰,张之祥,沈军民.基于掩模图像的磁环表面缺陷提取研究[J].光电子激光,2017,28(7):732~741 |
| 基于掩模图像的磁环表面缺陷提取研究 |
| Study on surface defect extraction of magnetic ring based on masking technology |
| 投稿时间：2016-09-18 |
| DOI： |
| 中文关键词: 掩模图像 表面缺陷 磁环 边缘检测 |
| 英文关键词:masking image surface defect magnetic ring edge detection |
| 基金项目:国家自然科学基金(61374022)、浙江省公益性技术应用研究计划(2014C33109)和浙江省新型 网络标准及其应用技术重点实验室开放课题(2013E10012)资助项目 (1.浙江理工大学 自动化系,浙江 杭州 310018; 2.浙江理工大学 电子信息工程系,浙江 杭州 310018) |
|
| 摘要点击次数: 1302 |
| 全文下载次数: 0 |
| 中文摘要: |
| 磁环表面缺陷图像具有对比度低、纹理背景复杂 、缺陷种类多和亮度不均匀等特点,提出了一种 基于掩模图像的磁环表面缺陷提取方法。首先,通过分析磁环表面图像中不同缺陷区域与正 常区域的灰度 差别,将磁环表面缺陷划分为两类,第1类为缺陷灰度值与正常区域差别大易分割,第2类为 缺陷灰度 值与正常区域灰度值差别小不易分割；其次,根据两类表面缺陷的成像特点以及与背景的 关系,利用掩 模技术分别设计了相应的缺陷提取方法；最后,在不同的光照、规格、缺陷类型及大小等方 面,利用开发 的样机进行了大量的在线实验。实验结果表明,本文缺陷提取算法稳定性好,鲁棒性强,能 够准确、快速 地提取出磁环表面图像各区域的缺陷,表面缺陷检测的准确率为95.3%。 |
| 英文摘要: |
| The magnetic ring surface defects image has low contrast,complex bac kground texture,many kinds of defects,uneven brightness and so on,so a surface defect extraction method of magnetic ring is proposed based on masking technology in this paper.First,the surface defects of magnetic ring are classified into two types through the analysis of the gray difference between different defect regions and normal regions in magnetic surface image. The first one is the difference between the gray value and the normal area,which is easy to segment.The second one is the difference between the gray value of the defect and that of the normal area.Secondly,according to the characteristics of the two kinds of surface defects and the relationship wit h the background,the corresponding defect extraction methods are designed using the masking technology.Finally ,a large number of online experiments are carried out using the developed prototype in different illum ination,sizes,defect types and sizes, etc.Experimental results show that the defect extraction algorithm has good st ability,strong robustness,and can accurately and quickly extract the defects of various regions in a magnetic ring surface image.And the surface defect detection accuracy rate is 95.3%. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|