总变差模型在书画打印宣纸印刷质量预测中的应用

Application of Total Variation Model in Prediction of Xuan Paper Printing Quality

中文关键词:  书画打印宣纸  总变差模型  帘纹  预测

中文摘要:

      为了建立适用于书画打印宣纸印刷质量的预测模型，本研究测量了14种书画打印宣纸的粗糙度、白度、不透明度、定量、光泽度和针对宣纸特别设定的帘纹深浅以及帘纹疏密度等表面物理参量，并在相同条件下，使用喷墨打印设备输出并测量印品色度值，利用总变差模型构建去除帘纹色差的测定方法，得到与人眼视觉特征相符的色差。运用GRNN广义回归神经网络结合书画打印宣纸表面物理参量与宣纸去帘纹后的色差值，建立预测模型。结果表明，该模型能够在仅测量书画打印宣纸表面物理参量的情况下，便能较为准确地预测书画打印宣纸印刷质量，为书画打印宣纸印刷前的选纸工作提供指导依据。

Abstract:

      In order to establish a paper printing quality prediction model suitable for Xuan paper, 14 kinds of Xuan paper's surface parameters such as roughness, whiteness, opacity, basis weight, gloss, were tested. According to the characteristics of Xuan paper, the parameters of depth and density of paper grains were specially set and measured. Chroma of the printed matter was measured under the same printing condition, total variation model was used to construct a measuring method which eliminated the influence of paper grains, and the chromatic aberration values in accord with human visual characteristics were obtained. And a prediction model was established by utilizing generalized regression neural network combined with Xuan paper's surface parameters and chromatic aberration value without the influence of paper grains. The results showed that the model could accurately predict the paper printing quality by measuring the surface physical properties of Xuan paper only. This method provided guidance for paper selection before printing by measuring the specific surface parameters of Xuan paper.