纺织品色彩管理色表测量数据的误差优化

Error optimization for measuring color chart data in textile color management

针对目前纺织品色彩管理中,测量误差引起呈色规律跳变和反演等问题,提出一种基于多方向的曲线拟合算法,并结合每一方向曲线对应的权重系数,对测量数据中误差过大的坏点进行优化。通过客观评价和主观评价对该优化方法的应用效果进行分析。结果表明:多方向曲线拟合优化算法对误差大的数据优化显著,对其他正常数据影响小,且可有效降低色表测量数据误差带来的影响,优化后生成的国际色彩联盟(ICC)色彩特性文件整体平均色差降低12.30%,难打色平均色差降低16.67%,中性灰色平均色差降低16.74%;在软打样过程中,优化后生成的ICC色彩特性文件的打样色差也小于优化前的色差。

In current textile color management,measurement errors are found to cause problems such as the color shade jump and inversion.In this research,a multi-directional curve fitting algorithm combined with the weight coefficient corresponding to each directional curve was proposed to optimize the bad points with excessive error in the measurement data.By analyzing the application effect of the optimization method through objective and subjective evaluation,it is found that the multi-directional curve fitting optimization algorithm works well with data having large errors,and it has little impact on other normal data,therefore effectively reducing the impact of the error from color chart measurement data.The overall average color difference of the International Color Consortium(ICC)Profile generated by the optimization is reduced by 12.30%.The average color difference of the difficult colors is reduced by 16.67%,and the average color difference of the neutral gray is reduced by 16.74%.During the soft proofing process,the proof color difference of the ICC Profile generated after optimization is also smaller than that before optimization.