基于欧氏距离方法的缺陷识别技术研究  

Research of Defect Identification Technology Based on Euclidean Distance Method

作　　者：吕晶[1] 唐万有[1] 蒋瑞雪[1] 郝健强[1] 易端阳[1] 王慧芳[1] 

机构地区：[1]天津科技大学,天津300222

出　　处：《包装工程》2013年第7期90-93,109,共5页Packaging Engineering

基　　金：广东省重大科技专项(2010A080402010)

摘　　要：通过分析常见的印刷缺陷特征,提出了基于缺陷几何形状和墨量变化共同判断缺陷类型,即通过欧氏距离方法检测缺陷的形状、位置等特征,并设计了一种色度和密度转化算法,来判断印刷缺陷的墨量变化,成功实现了对印刷品缺陷的提取、分类及反馈。Characters of common printing defects were analyzed. The method to determine the type of defect based on defected geometric shape and printing ink quantity changes was put forward, which used Euclidean distance method to detect shape, position and other parameters of defect. A kind of color density conversion algorithm was designed to judge ink quantity change of printing defect. Printing defect extraction, classification, and feedback was realized.

关 键 词：在线检测 欧氏距离 色密转换 故障反馈 

分 类 号：TS807[轻工技术与工程]