纺织学报 ›› 2005, Vol. 26 ›› Issue (2): 69-71.

• 分析探讨 • 上一篇    下一篇

基于最大多符号信息熵的织物图像匹配

包晓敏,汪亚明,罗一平,许洲   

  1. 浙江理工大学计算机视觉与模式识别中心 浙江杭州310018
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2005-04-15 发布日期:2005-04-15

Matching of textile image based on the biggest multi-symbol information entropy

BAO Xiao-min;WANG Ya-ming;LUO Yi-ping;XU Zhou   

  1. Computer Vision and Pattern Recognition Research Center;Zhejiang University of Science and Technology;Hangzhou;Zhejiang 310018;China
  • Received:1900-01-01 Revised:1900-01-01 Online:2005-04-15 Published:2005-04-15

摘要: 为了实现利用机器视觉技术识别织物图案的组成 ,对织物图案进行研究。依据熵函数是一个连续函数 ,且具有极值性 ,将其运用到图像检测与匹配中。把织物图像中的每一点作为单符号信源 ,整幅图像作为多符号信源 ,在织物成品图像上选取和基准样图尺寸大小相同的图像计算信息熵 ,若与样品图的信息熵相等 ,则说明在织物成品图中检测到样图 ,否则没有。实验结果表明该方法在织物图案匹配中是一种实用和成功的方法。

Abstract: Detecting the composing of the textile image by using machine vision technology.The matching way of textile image was researched.The entropy function is a continuous function and has extreme property.The biggest multi-symbol information entropy was applied into the detection and matching of fabric images,each point in an image was considered as a one-symbol source,the whole image was considered as multi-symbol source.The information entropy of the selected sub-image in the textile image of finished product which has the same size as the standard sub-image was calculated.If the information entropy of the selected sub-image was equal to that of the standard image,the standard sub-image was detected from the textile image,otherwise the standard sub-image was not detected.The experimental results indicate that method is efficient and practicable.

No related articles found!
Viewed
Full text
478
HTML PDF
Just accepted Online first Issue Just accepted Online first Issue
0 0 0 0 0 478

  From Others local
  Times 362 116
  Rate 76% 24%

Abstract
956
Just accepted Online first Issue
0 0 956
  From Others local
  Times 892 65
  Rate 93% 7%

Cited

Web of Science  Crossref   ScienceDirect  Search for Citations in Google Scholar >>
 
This page requires you have already subscribed to WoS.
  Shared   
  Discussed   
No Suggested Reading articles found!