计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (12): 8-18.DOI: 10.3778/j.issn.1002-8331.2002-0169

• 热点与综述 • 上一篇    下一篇

织物瑕疵检测方法研究进展

田宸玮,王雪纯,杨嘉能,钱育蓉   

  1. 1.新疆大学 软件学院,乌鲁木齐 830000
    2.中国石油吐哈油田分公司 勘探开发研究院,新疆 哈密 839009
  • 出版日期:2020-06-15 发布日期:2020-06-09

Research Progress on Fabric Defect Detection Methods

TIAN Chenwei, WANG Xuechun, YANG Jianeng, QIAN Yurong   

  1. 1.School of Software, Xinjiang University, Urumqi 830000, China
    2.Research Institute of Exploration and Development, Tuha Oil Field Branch Company Ltd. of PetroChina, Hami, Xinjiang 839009, China
  • Online:2020-06-15 Published:2020-06-09

摘要:

计算机视觉在纺织物瑕疵检测方面已经有了较为广泛的应用,织物瑕疵检测是纺织行业质量控制的必要步骤。为了能够及时了解织物瑕疵检测的最新研究进展,把握织物瑕疵检测方法的研究热点和方向,介绍了织物图案的放置规则以及织物瑕疵检测的相关指标,针对织物瑕疵检测方法较多的问题,将这些方法分为四类(基于结构、基于统计、基于频谱、基于学习)并归纳分析了这些织物瑕疵检测方法在应用中的特点,且对比了四类检测方法中各个方法的特性以及优点和缺点,目的是为了找到如何提高织物瑕疵检测效率的方法,实现实时在线检测,展望了织物瑕疵检测方法的未来研究方向。

关键词: 织物瑕疵, 纺织物, 计算机视觉, 图像处理

Abstract:

Computer vision has been widely used in the detection of textile defects. Fabric defect detection is a necessary procedure in the quality control of the textile industry. In order to timely understand the latest research progress of fabric defect detection and to grasp the research hotspots and directions of fabric defect detection methods. In view of the problems that fabric defect detection methods, this paper introduces the combination method of fabric pattern and related indicators of fabric defect detection. These methods are divided into four categories(structure-based, statistical-based, spectrum-based, learning-based) and the characteristics of these fabric defect detection methods are summarized and analyzed. The characteristics, advantages and disadvantages of the four types of detection methods are analyzed. And the purpose, in order to find a way to improve the detection efficiency of fabric defects, to achieve the purpose of real-time online detection. The future research direction of fabric defect detection methods are prospect.

Key words: defect detection, textiles, computer vision, image processing