为了获得高精度的配色样品, 提出了一种配色方法—成分分析配色法。 该方法将化学分析法与计算机配色相结合, 其核心是选取与目标色成分最接近的油墨进行配色, 从而实现高精度的色彩匹配, 为计算机配色发展提供新的思路。 与目标色成分相近的油墨配色效果验证: 使用某种油墨印制目标色, 并用相同的油墨进行配色, 以实现目标色成分与配色色样成分的一致性。 使用三个不同品牌的油墨对目标色进行配色, 比较配色精度及效率。 使用泗联牌三种颜色油墨以任意比例通过印刷适性仪IGT-CI（荷兰）印制目标色, 这些目标色包括间色和复色, 各3个色样; 使用配色软件X-Rite color master（美国）建立泗联、 东洋、 牡丹三个品牌油墨的配色基础数据库, 并对不同目标色进行配色。 结果表明使用与目标色相同的泗联油墨的配色精度远高于东洋、 牡丹两个品牌的油墨, 配色色差整体都很小, 校正1～2次就能得到小于1.0的色差, 最小达到0.36, 几乎实现了目标色的同色同谱匹配。 实验验证了成分分析配色法的核心“选取与目标色成分最接近的油墨进行配色, 可以实现高精度色彩匹配”的可行性。 判别目标色与配色油墨在成分上区别的化学分析工具探讨: 为了判别目标色色料与配色油墨在成分上有区别, 尝试使用“红外光谱相似度”作为判别的分析工具。 使用红外光谱仪Thermo Nicolet 6700（美国）测出泗联、 东洋、 牡丹三个品牌的三种颜色油墨的红外光谱图, 使用OMNIC软件中的相关性算法得到它们与目标色油墨的红外光谱相似度, 并计算出平均相似度; 将各品牌油墨的红外光谱相似度与其配色实验的精度进行对比分析, 评价红外光谱相似度作为化学分析判别工具的有效性。 结果表明泗联牌油墨与目标色的平均红外光谱相似度为100%, 东洋的为86.53%, 牡丹的为64.63%。 当校正次数相同时, 泗联油墨配色色差最小, 配色精度最高; 东洋次之, 是泗联油墨配色色差的2倍左右; 牡丹最差, 是泗联油墨配色色差的3倍以上。 配色结果与其红外光谱相似度的规律是一致的, 目标色油墨与配色油墨之间的红外光谱相似度越高, 越容易得到高精度的配色样品。 实验证明了用成分分析配色法获得高精度的色彩匹配是可行的, 使用红外光谱相似度作为目标色与配色油墨在成分上的分析工具对判别配色精度是有效的。 今后的工作将探讨红外光谱相似度与配色精度间的相关性数值关系, 以及进一步寻求更为有效的化学分析方法来判断目标色色料与配色油墨之间的成分关系。

To acquire the high-accuracy of samples from the computer color matching, a novel color matching method was proposed in this paper which combines chemical component analysis with computer color matching. The kernel of this method is to select the inks that are most similar to the printing inks of target colors, thus the high-accuracy color matching is achieved. This method can provide a new referential direction for future development of computer color matching. The verification of color matching effects was conducted using inks that were similar to target colors in composition. The target colors were firstly printed by printing inks, and then the color matching was carried out using the same printing inks to maintain the consistency between target colors and test colors of printing inks. In this paper, three different brands of printing inks were used to match colors with target colors, and the accuracy and efficiency of computer color matching from those printing inks were compared intuitively. The target colors were printed using three different colors Silian ink in arbitrarily proportions and volume, and they were obtained by using IGT-C1 printability tester (IGT Inc.,Netherlands), and these target colorsranges included secondary and tertiary colors, with each color range having three samples, respectively. Three primary printing inks Cyan, Magenta and Yellow of three brands of printing inks—Silian, Dongyang and Mudan were used as basic inks to establish the fundamental database by X-Rite color matching software (X-Rite Inc., America), and then different target colors were matched with foundational database of three brands of printing inks respectively. The results showed that the color matching accuracy of the Silian ink outperformed the other two brands of printing inks because of the fact that Silian ink was used as the printing inks of target colors, and the holistic color differences of Silian ink were the smallest among the three brands of printing inks and the color differences less than 1.0 were achieved just after one or tw-ice corrections. The smallest color difference was acquired even 0.36 and from the spectrum matching, which implied that Silian ink almost achieved isomeric match with the target color. This experiment has verified the feasibility of the emphasis of component analysis—computer color matching method, which picks the most similar inks to the printing inks of target colors so that we can achieve a high-precision color matching. The chemical analytical tool assessing the difference between printing inks of target colors and color-matching test colors. To distinguish the printing inks of target colors and color-matching test colors from the component levels, the infrared spectral similarity was used as an analytical tool in this paper. The spectra of printing inks of three colors of the three brands were measured by the Thermo Nicolet 6700 Fourier transform infrared spectrometer (Thermo Fisher Scientific Inc., Waltham, USA), and the infrared spectral similarities of the printing inks of target colors and color-matching were all obtained and then their average similarities were also calculated by the OMNIC software. Through the comparison between the infrared spectral similarities of different brands of printing inks and the precision of computer color matching experiment, the rationality and validity of the infrared spectra as a chemical analytical discriminant tool to evaluate similarity between inks of target colors and color-matching colors was verified. The results indicated that the similarity between spectra of Silian inks and printing inks of target colors was the highest and even reached 100%, while Dongyang inks offered a high similarity of 86.53%, and Mudan inks provided the lowest similarity of 64.63%. The results showed that when the number of correction was the same, taking color difference as the criteria of judgment, the color differences of color matching for Silian ink were the smallest and it meant this ink provided the highest accuracy of color -matching; and Dongyang ink took the second place, and its color differences were about twice as large as Silian ink; and the Mudan ink showed the highest color differences, and its color differences were three times more than those of the Silian ink. The result of computer color matching experiment was consistent with infrared spectral similarities results, and the principle suggested that the higher the infrared spectral similarity between printing inks of target colors and test colors was, the easier high precision color matching sample could be received. Conclusion and Prospect: The feasibility of the new color matching method which combines the component analysis and computer color matching to gain the high-accuracy color-matching samples was proved by experiments and result analysis. Use the infrared spectral similarity as an analysis tool to distinguish the difference in components between the printing inks of target colors and test colors is feasible, which can be an effective criterion for determining the color matching accuracy. Future research will focus on probing into the correlative numerical relationship between infrared spectral similarity and color-matching precision, and further seeking more effective chemical analytical method to estimate the componential relationship between printing inks of target colors and color-matching test colors.