文献类型：期刊文献

中文题名：图像形态特征提取在铜箔基板疵点分析中的应用

英文题名：Shape Character Abstraction Method for the Flaw Analysis of CCL Image

作者：朱煜[1];张伟利[1];姚传俊[2]

机构：[1]华东理工大学信息学院,575信箱上海200237;[2]上海贝尔-阿尔卡特股份有限公司,mnd上海200230

年份：2004

卷号：34

期号：6

起止页码：490

中文期刊名：激光与红外

外文期刊名：Laser & Infrared

收录：CSTPCD;;Scopus;北大核心:【北大核心2000】；CSCD:【CSCD2011_2012】；

语种：中文

中文关键词：铜箔基板;边缘提取;链码;形态特征

外文关键词：CCL;edge detection;chain code;shape characteristic

摘要：对铜箔基板生产过程中产生的疵点进行了有效的识别与分类。采用图像二值化分析及边缘检测方法进行疵点图像的预处理 ,并应用链码跟踪方法得到链码表及边缘点坐标表 ,从而计算出疵点的各类形态特征参数。根据该参数值可将疵点进行合理的分类统计。该方法可靠便捷 。
The recognition and category of the flaw of Copper Clad Laminates were achieved in this paper. Prewitt edge detection algorithm and histogram analysis were used to pre-process of the images. Edge pixels were tracked by Freeman chain code. According to chain code table, the coordinate of edge pixel can be used to calculate some parameters of the shape character, such as the length, width and area of the aim, especially the circle alike parameter. These methods were effective and credible.