基于计算机视觉的织物疵点检测的近期进展  

Recent Progress of Computer-vision-based Fabric Defects Detection

作　　者：步红刚[1] 黄秀宝[1] 

机构地区：[1]东华大学纺织学院,上海201620

出　　处：《东华大学学报（自然科学版）》2006年第3期128-133,共6页Journal of Donghua University(Natural Science)

摘　　要：按照模式识别的一般顺序并根据当前的研究中所使用的方法,简要综述了近年来计算机视觉在织物疵点检测领域的应用和发展情况。首先分析了将计算机视觉应用于织物疵点检测的现实意义。考虑到特征提取与选择模块以及检测模块是瑕疵检测系统中最关键的两个部分,着重讨论了这两个模块近期所涉及的各种方法及其在疵点检测应用上的效果。最后给出了结论,说明了开发国产化自动验布机的意义。This paper briefly summarizes computer vision＇s applications to and progress in the research area of fabric defects detection according to the general process of pattern recognition as well as the research approaches adopted in recent years. The realistic significance of applying computer vision to fabric defect detection is analyzed firstly. Considering that the feature extraction and selection module as well as the detection module are especially the two most crucial components of a defect detection system, we lay emphasis on the discussion of the various methods and their application effects on fabric defect detection involved in these two modules. Conclusion is reached lastly in which the meaning of developing homemade automatic fabric inspection system is explained.

关 键 词：织物疵点 计算机砚觉 检测 

分 类 号：TS101.9[轻工技术与工程—纺织工程]