Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/11917
Title: Investigation on the classification of weave pattern based on an active grid model
Authors: Xin, B
Hu, J 
Baciu, G 
Yu, X
Issue Date: 2009
Publisher: SAGE Publications
Source: Textile research journal, 2009, v. 79, no. 12, p. 1123-1134 How to cite?
Journal: Textile research journal 
Abstract: In this paper, a new method based on the active grid model (AGM) is used to identify the weave pattern of woven fabrics. The two-dimensional geometrical weaving structure of the woven fabrics could be described mathematically using the concept of active grid alignment, so that the analysis of the fabric weave pattern could be implemented in the field of an AGM model of a fabric. This proposed method utilizes dual-side scanning technology to merge the dual-side images of a fabric at the yarn level. It contains a four-step method to construct an AGM. First, a yarn-detecting algorithm is applied on the dual-side scan images to initialize the AGM. Second, the AGM self-adjustment scheme is used to adjust the AGM accurately. Then, the types of the yarn interlacing are classified based on the edge map and the result is refined using the neighboring information of yarns. Finally, the color pattern is determined by using color clustering and matching; error correction is also made based on the color configuration. Some preliminary experiments show that the AGM is effective for the classification of fabric weaving patterns.
URI: http://hdl.handle.net/10397/11917
ISSN: 0040-5175
EISSN: 1746-7748
DOI: 10.1177/0040517508101459
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

37
Last Week
0
Last month
0
Citations as of Sep 8, 2017

WEB OF SCIENCETM
Citations

33
Last Week
0
Last month
0
Citations as of Sep 22, 2017

Page view(s)

44
Last Week
2
Last month
Checked on Sep 17, 2017

Google ScholarTM

Check

Altmetric



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.