基于人工智能的色差分类技术  

Techniques of chromatism classification based on artificial intelligence

作　　者：李伟青[1] 陈广社 

机构地区：[1]中国地质大学(北京)工程技术学院,北京100083 [2]北京地特思科技有限公司,北京100083

出　　处：《微型机与应用》2007年第S1期106-108,112,共4页Microcomputer & Its Applications

基　　金：中国地质大学(北京)科学技术基金(200520)

摘　　要：利用人工智能和图像处理技术,提出了一种新的瓷砖色差自动分类方法。在传统的图像颜色直方图方法基础上,充分考虑了颜色的空间分布对人类视觉的影响,将图像主要颜色的H(色调)、S(饱和度)、P(颜色的概率)及D(颜色的空间分布)作为神经网络的输入,克服了仅仅利用颜色直方图进行颜色分类带来的因颜色空间信息丢失而无法精确分类的问题,采用模糊技术对瓷砖表面色差进行了精确分类。A new method of tile chromatism classification based on artificial intelligence and image processing was presented. Color spatial distribution information of color was applied to the old method. Hue, saturation, probability and spatial distribution of three (or more) colors, whose appearing probability were maximum in color histogram, were selected as input parameters of artificial neural network, and fuzzy techniques was applied in the input and output of artificia neural network in order to make the resul...

关 键 词：颜色直方图 颜色空间分布 神经网络 色差分类 模糊技术 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]