基于机器视觉的织物疵点检测方法综述  

Survey on fabric defect detection based on machine vision

作　　者：邹超[1] 汪秉文[1] 孙志刚[1] 

机构地区：[1]华中科技大学控制科学与工程系,武汉430074

出　　处：《天津工业大学学报》2009年第2期78-82,85,共6页Journal of Tiangong University

基　　金：图像信息处理与智能控制教育部重点实验室开放基金(200702)

摘　　要：对基于机器视觉的各种织物疵点检测方法进行综述,将检测系统分成3个层次:疵点判别、疵点分割和疵点分类,给出了对各个层次的要求和完成的目标,并将已经提出的大多数疵点检测方法按其特点归类到3个层次中,具体描述了各类方法的优缺点,分析其实用性,最后总结了该领域的发展趋势.Various methods of fabric defect detection based on machine vision are surveyed. The whole defect detection system is divided into 3 levels： defect judgment, defect segmentation and defect classification. The requirements and objects of each level are introduced respectively. Most of the proposed methods on this topic are classified to the three levels according to their characterization. The advantages and disadvantages of each method are introduced and feasibilities of each method are analyzed. Finally, the prospective of this research area is summarized.

关 键 词：疵点检测 机器视觉 疵点判别 疵点分割 疵点分类 

分 类 号：TS101.97[轻工技术与工程—纺织工程] TU352.110.4[轻工技术与工程—纺织科学与工程]