陶瓷砖表面缺陷视觉检测算法研究  

The Study on Visual Inspection Algorithm of Ceramic Tiles for Defects Detection

作　　者：李庆利[1] 王永强[1] 张帆[1] 

机构地区：[1]唐山学院,唐山063000

出　　处：《中国陶瓷》2015年第3期44-47,共4页China Ceramics

基　　金：唐山市科技计划项目(13130250z)

摘　　要：提出了一种基于灰色关联分析的特征颜色提取算法,在此基础上,完成了陶瓷砖表面缺陷检测软件设计。应用灰色关联分析的方法来确定像素点间的隶属度,对图像进行分块处理来求解初始特征颜色,使用灰色聚类进行图像分割。系统以ARM为工作平台,工作在RGB颜色空间。实验表明:算法可快速有效地完成陶瓷砖表面缺陷的检测。In this paper a feature color extracted algorithm is proposed based on the gray incidence analysis for defects detection of ceramic tiles. Firstly, the number of clusters and the gray relational grade of pixel is calculated. Then, the feature color is extracted with the improved gray clustering algorithm. The system utilizes the embedded technology and machine vision technology, using the ARM as the hardware platform and. Experimental results show that the algorithm could effectively detect the defect of ceramic tiles.

关 键 词：陶瓷砖 缺陷检测 特征颜色 灰色关联度 灰色聚类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]