摘要
针对织物表面质量检测仍处于人工检测的现状,根据织物疵点破坏织物固有纹理组织结构,体现为灰度值纵向或横向频率变异的特点,提出一种疵点自动检测方法。该方法先由织物组织循环结构确定大小两个检测窗口的尺寸,然后通过合适的离散小波基把试样分解成含经向细节和纬向细节的两个子图,分别统计两者能量和后分离出低频的疵点信息。该方法简单易行,漏检率低,适应于大多数方向性、块状类疵点的检测。
Considering that fabric surface quality inspection still depends on manual work,and that fabric defect damages the texture structure of fabric is featured by vertical or horizontal frequency departure of grey level,an auto inspection method proposed,according that.Firstly,the size of the big and small inspection windows were determined by the loop structure of fabric weave.The the sample is decomposed into two sub-images containing warp and weft details by proper discrete wavelet basis.Finally the energy sum and the defect information of the low frequency were calculated respectively.This method is simple and easy to operate,good in real time,and suitable for detection of most directional and lumpy defects.
作者
吴军
WU Jun(College of Information&Business,Zhongyuan University of Technology,Zhengzhou 451191,China)
出处
《现代纺织技术》
2020年第1期47-51,共5页
Advanced Textile Technology
基金
2014年度河南省重点科技攻关(142102210524)
关键词
织物疵点检测
双窗口设计
小波分解
能量统计
fabric defect
double-windows design
wavelet decomposition
energy sum
