曲线回归分析在织物染色计算机配色中的应用研究  

Research on Application of Cure Regression Analysis in Computer Color Matching for Textile Dyeing

作　　者：武洋洋[1] 张秉森[1] 李敏[1] 

机构地区：[1]青岛大学信息工程学院,山东青岛266071

出　　处：《青岛大学学报（工程技术版）》2013年第1期22-26,共5页Journal of Qingdao University(Engineering & Technology Edition)

基　　金：国家自然科学基金资助项目(60973158;60743004)

摘　　要：针对三刺激值C,M,Y与浓度之间的关系,本文在Kubelka-Munk理论的基础上,应用数学建模的方法对织物染色计算机配色进行研究,通过对实验样本数据的选取与分析,建立了在相同染料浓度下三刺激值C,M,Y与染料浓度的曲线回归模型,并对该模型进行回归分析,分析结果表明,实验数据的误差绝大部分在0.65%以内,说明该回归方程的拟合效果比较好,与C,M,Y值相对应的所求解的染料浓度比实验数据的误差小,该研究精度高,拟合效果好,能满足实际织物染色配色的需要,具有一定的实用价值。Curve regression model of the value of C, M, Y and the concentration of dye in same dye concentration is established by the way of mathematical modeling and analysis of experiment sample data in this paper based on the fabric dyeing color matching theory, which is called Kubelka-Munk theory, in order to get the relationship of concentration of dye and its three stimulus values. Then curve regression equation can be got. With the equation being put the experiment data experiment concentration can be obtained by solving the equation. The experiment results show that the error of the experiment data is less than 0.65 % and the fitting result of the experiment curve regression equation is good enough. The error of the solution of C, M, Y value and its corresponding concentration of dye is smaller than experiment data, which can meet the requirements of actual color dyeing matching.

关 键 词：织物染色 计算机配色 数学建模 曲线回归 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]