神经网络在地砖色差分类中的应用研究  

Research of Application of Neural Network to Classification of Floor Tile Color Difference

作　　者：秦国兴[1] 

机构地区：[1]唐山学院计算机中心,河北唐山063000

出　　处：《唐山学院学报》2009年第6期94-96,共3页Journal of Tangshan University

基　　金：唐山学院科研项目(09002B)

摘　　要：通过计算机视觉系统获取地砖图像,经过图像处理提取得到相关的图像特征数据,再将特征参数作为输入建立RBF神经网络模型,利用神经网络所具有的自组织、自学习和分布式存储信息的特点对地砖进行可控的自动分类。实验证明,此方法为地砖检测分类提供了一个有效的途径。First,floor tile images are obtained with a computer vision system,then the relevant image data are extracted by image processing and finally characteristic parameters are used as the input to establish BP neural network model,with whose features of self-organizing,self-learning and distributed storage of information floor tiles can be classified in a controlled and automatic way.Experiments show that this method provides an effective way for detecting and classifying floor tiles.

关 键 词：神经网络 色差 分类 地砖 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]