首页 | 本学科首页   官方微博 | 高级检索  
     

基于区域填充的纤维图像提取算法
引用本文:魏倬瑜,万燕,姚砺,曾培峰,吴雄英. 基于区域填充的纤维图像提取算法[J]. 东华大学学报(自然科学版), 2008, 34(5)
作者姓名:魏倬瑜  万燕  姚砺  曾培峰  吴雄英
作者单位:1. 东华大学计算机科学与技术学院,上海,201620
2. 上海出入境检验检疫局,上海,200135
基金项目:全国优秀博士学位论文作者专项资金资助项目,教育部留学回国人员科研启动基金
摘    要:对天然纤维显微图像中所有纤维对象的正确分离和提取,是棉麻纤维特征分析的必要前提.针对纤维图像中图像背景、纤维边缘和纤维内腔的灰度分布特性,提出基于区域填充的纤维图像提取算法.该算法以各个"纤维内腔"为图像分离提取的标准,采用膨胀区域填充的方法,并通过距离变换自动确定填充起始点,对经过二值化处理的纤维图像中的背景和纤维进行了区分,从而得到各个纤维对象的正确提取.该算法能够有效处理纤维图像间的黏连问题.

关 键 词:纤维  区域填充  距离变换  黏连

Fiber Image Extraction Algorithm Based on Region Filling
WEI Zhuo-yu,WAN Yan,YAO Li,ZENG Pei-feng,WU Xiong-ying. Fiber Image Extraction Algorithm Based on Region Filling[J]. Journal of Donghua University, 2008, 34(5)
Authors:WEI Zhuo-yu  WAN Yan  YAO Li  ZENG Pei-feng  WU Xiong-ying
Affiliation:WEI Zhuo-yu1,WAN Yan1,YAO Li1,ZENG Pei-feng1,WU Xiong-ying2
Abstract:To correctly separate and extract fibers from nature fiber image is the base of fiber recognition. According to the gray degree distribution of different objects in fiber image,a fiber image extraction algorithm based on region filling technology is proposed. This algorithm takes fiber lumen as the seed of fiber separation,with region filling algorithm in which the beginning points are determined automatically through distance transform,separates the fiber from background in binary processing image. The conglutinated fibers can be separated correctly,and the fiber objects can be extracted efficiently from image.
Keywords:fiber  region filling  distance transform  conglutination
本文献已被 CNKI 维普 万方数据 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