粉底调色模型的建立研究

Establishment of color matching model for foundation

针对不同人种及个体对彩妆中粉底色彩的不同需求,以潘通色卡110种肤色作为调色依据,选取二氧化钛、氧化铁黄、氧化铁红和氧化铁黑等4种色粉进行调色,对调色的粉底样分别进行LAB值测试,然后通过线性回归、多项式回归及混料设计等不同方式建模。对比结果发现,线性回归结果其调整R方均大于0.914,p值均小于0.01;多项式回归结果调整R方均大于0.920,p值均小于0.01;而混料设计模拟结果其R方均大于0.989,p值均小于0.01。经验证,3种方式的建模其模拟值与样板实测值拟合效果优良,可用于配方调色时重点参考模型,具有准确的预测性。

In view of the different needs of different nationalities and individuals for make-up color,110 kinds of toners of pant color card were used as the basis of color matching,and four kinds of toners,such as titanium dioxide,iron oxide yellow,iron oxide red and iron oxide black,were selected for color matching.The LAB of the palette samples was tested separately,and then the models were modeled by means of linear regression,polynomial regression and mixture design.The comparison results showed that the adjusted R-squares of linear regression results were all greater than 0.914,and p values were all less than 0.01;the adjusted R-squares of polynomial regression results were all greater than 0.920,and p values were all less than 0.01;while the R-squares of mixture design simulation results were all greater than 0.989,and p values were all less than 0.01.The results showed that the simulation values and the measured values of the model had good fitting,which could be used as the key reference model in the color matching of the formula,and indicative of accurate prediction.