摘要
概述了验布技术的现状 ,对传统人工验布与自动验布进行了对比 ,介绍了自动验布系统的四个组成模块 ,并分析了人工神经网络技术在自动验布中的必要性、可行性和实现的一般模式 ,详细介绍了疵点样本图像的采集、特征值的提取、神经网络的训练、网络的移植等具体实施过程。
The present situation of cloth inspecting technology is summarized, the four modules of automatic cloth inspecting system are introduced by comparing the traditional manual cloth inspection and automatic cloth inspection. Then the necessity, feasibility and usual modular of artificial neural network are analysed, too. The sampling of defects picture, the extraction of characteristic value, the training and grafting of artificial neural network are detailed from this comparison.
出处
《棉纺织技术》
CAS
CSCD
北大核心
2004年第1期30-33,共4页
Cotton Textile Technology
关键词
神经网络
自动化
疵点
特征提取
小波分析
自动验布系统
图像采集
Artificial Neural Network, Cloth Inspection, Automation, Defects, Characteristic Extraction, Wavelet Analysis
