Defect Detection of Jacquard Fabrics Using Multiple Color-channel Analysis

Xiaobo Yu, Jinlian Hu, George Baciu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

1 Citation (Scopus)

Abstract

The automated defect detection of jacquard fabrics is intuitively appealing to the textile industry but as yet there has been little in-depth research in this field. In this paper, we introduce a defect detection method for use on jacquard fabrics that is based on multiple color-channel analysis. According to the number of color yarns employed, a number of color channels can be extracted from an arbitrary jacquard fabric. Images of each color channel are patterned. By first separating the color channel of an input test fabric image, and then eliminating noise and applying a pattern extraction process, it is possible to produce a set of channel patterns. To characterize each defined defect, we introduce a pattern comparison method which makes use of Fourier transform and frequency spectrum analysis. In experiments, this method efficiently and precisely detected defects in test fabrics and provided relevant information about defects such as the defect category and which color yarns are involved in the defect.

Original languageEnglish
Pages (from-to)21-29
Number of pages9
JournalResearch Journal of Textile and Apparel
Volume9
Issue number1
DOIs
Publication statusPublished - 1 Feb 2005

Keywords

  • Jacquard
  • Defect
  • Color channel

ASJC Scopus subject areas

  • Business and International Management
  • Materials Science (miscellaneous)
  • Industrial and Manufacturing Engineering
  • Management of Technology and Innovation

Cite this