一种帘子布疵点检测新方法

帘子布作为一种特殊纺织品,其外形的不确定性给疵点的自动识别带来了困难。针对帘子布的纹理特点,提出了一种参数优化的第三代人工神经网络(PCNN)模型,利用此模型对帘子布疵点特征值进行提取,然后用BP神经网络实现帘子布疵点检测。实验表明该方法可准确检测帘子布中劈缝、浆斑、经线粘连、颜色不均、稀经、断经和断纬七种疵点,具有分类准确、识别速度快的优点。

A new method for cord fabrics defect detection

Cord fabric is a kind of special textile and its indefinite shape brings new difficulties to defect automatic detection.According to the characteristics of cord fabric,an improved template is used to extract the PCNN that its parameters are optimized.It is used BP neural network to implement the cord fabric's defect detection after the histogram statistics of the point directional image.Experiment results show that the method accurately identifies 7 defects such as broken end,split slot,starch lump,warp adhesion and broken warp etc,which has the advantages of accurate classification and fast identification speed.