纺织学报 ›› 2010, Vol. 31 ›› Issue (6): 43-47.

• 纺织工程 • 上一篇    下一篇

机织物组织自动识别技术

潘如如;高卫东;刘基宏;王鸿博   

  1. 生态纺织教育部重点实验室(江南大学)
  • 收稿日期:2009-07-17 修回日期:2009-12-02 出版日期:2010-06-15 发布日期:2010-06-15
  • 通讯作者: 高卫东

Automatic identification of woven fabric weave

PAN Ruru; GAO Weidong; LIU Jihong; WANG Hongbo

  

  1. Key Laboratory of Eco-Textiles(Jiangnan University), Ministry of Education
  • Received:2009-07-17 Revised:2009-12-02 Online:2010-06-15 Published:2010-06-15
  • Contact: GAO Weidong

摘要:

为实现机织物组织的自动识别,对织物表面反射图像进行纠偏和直方图均衡预处理,并按图像的经纬向灰度投影曲线实现织物组织点的定位。根据经纬向组织点灰度的相关性,求出织物组织的经纬纱循环数。通过比较组织循环范围内组织点的灰度均值来确定组织点的属性,并以组织点的灰度方差对组织点的属性进行验证。实现单循环组织识别后,再对多个组织循环内的对应组织点进行验证,以进一步提高识别的准确性。织物的测试结果表明,所提出的算法能够实现简单的机织物组织的准确识别,并能输出对应的织物组织图。

Abstract:

To realize automatic identification of woven fabric weave, the reflection image of woven fabric is pre-processed by skew rectification and histogram equalization. Then the interlacing points are located in the image by the warp and weft gray projection curves. The warp and weft repeats are found according to the correlation of the interlacing points in warp and weft directions. The types of interlacing points are judged by comparing the average gray value of interlacing point in a weave repeat, and the results are examined with the gray variance of the interlacing points. Finally,the accuracy of identification is further improved by comparing the recognized results in several weave repeats. The experiment for actual fabrics shows that the algorithm can identify the three basic weaves correctly and output the diagram of the identified fabric weave.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!