基于纹理滤波与局部熵的织物疵点检测  

The Defect Detect of Fabric Based on Texture Filtering and Local Entropy

作　　者：陈利珍[1] 邓中民[1] 

机构地区：[1]武汉纺织大学,湖北武汉430073

出　　处：《纺织科技进展》2011年第5期42-44,共3页Progress in Textile Science & Technology

基　　金：湖北省重大项目;纺织面料及服装数字化设计与制造成套技术研究(C2010036)

摘　　要：针对织物疵点检测中织物纹理常被误认为噪声而给疵点自动检测造成很大干扰的问题,以织物纹理和疵点频谱的不同分布为依据,提出了一种结合纹理滤波和局部熵的织物疵点检测方法。通过频域滤波降低织物纹理的干扰,然后再结合局部熵与领域操作实现疵点的自动检测,并提取相关特征参数。In view of that the fabric texture in defect detect is usually considered as noise to cause great interference to the automatic detect of defect,according to the different distribution of fabric texture and defects spectrum,it proposed a detection method for fabric defect by combining with texture filtering and local entropy.By filtering in the frequency domain,it reduced the interference of fabric texture,then combined with local entropy and field operations to realize automatic detection of defects and extract related features parameters.

关 键 词：织物疵点 自动检测 纹理域波 局部熵 领域操作 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]