首页 | 本学科首页   官方微博 | 高级检索  
     

混纺纱截面图象的识别和分析
引用本文:许鹤群,章纪红,黄健. 混纺纱截面图象的识别和分析[J]. 东华大学学报(自然科学版), 1993, 0(1)
作者姓名:许鹤群  章纪红  黄健
作者单位:中国纺织大学图象处理中心,中国纺织大学图象处理中心,中国纺织大学图象处理中心
摘    要:选取由三种具有典型横截面形状的纤维(兔毛、晴纶和粘胶)组成的混纺纱作为研究对象。经过高通滤波、二值化后,对输入图象进行细化。为了消除存在的孤立点或孤立团,本文对 Hilditch 细化算法作了改进。纤维的轮廓通过以最小夹角原则跟踪获得,然后对轮廓进行平滑并计算其曲率,根据曲率变化以及轮廓内是否有孔洞对纤维进行分类、识别.最后根据识别的结果进行定量分析。

关 键 词:图象分析  图象处理  混纺纱  用计算机分析

PATTERN RECOGNITION AND ANALYSIS ON BLENDED YARN CROSS-SECTION
Xu Hequn Zhang Jihong Huang Jian. PATTERN RECOGNITION AND ANALYSIS ON BLENDED YARN CROSS-SECTION[J]. Journal of Donghua University, 1993, 0(1)
Authors:Xu Hequn Zhang Jihong Huang Jian
Affiliation:image processing center
Abstract:A blended yarn,spun with rabbit hair,acrylic and viscose fibres,is chosen for the research.TheHilditch algorithm is improved and adopted for image refinement after high-pass-filtering and bina-ryzation.The contour of its cross-section is obtained by tracing its pixels based on the rule of least in-ner angle.The curvature of the contour is then cal culated after smoothing.Classification and recogni-tion of fibres is performed according to variation of their contour curvature and the inner hole of therabbit hair.Results are used for quantitative analysis.
Keywords:image analysis  image processing  blended yarn  computerized analysis  
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号