目的 研究印刷品质量检测中的颜色转换模型。方法 基于色靶的测量数据， 采用三维查找表法和多项式回归法， 分别建立RGB与CIE L*a*b*颜色空间转换模型， 并实验比较了2种算法转换模型的精度。结果 三维查找表法转换模型的最大色差为3， 四面体插值算法精度稍高于三线性、 三棱柱和金字塔插值算法， 平均色差为 0.64； 多项式回归算法的精度随着项数的增多而提高， 20 项的平均色差为 2.58。结论 三维查找表法转换模型精度高于多项式回归法， 效果理想， 能判断出 CCD 采集到的印刷图像是否存在色偏， 且随着RGB颜色空间划分越细小， 转换精度会越高。

This paper studied the color transformation model of printing quality Inspection. The color space conversion models between RGB and CIEL*a*b* based on the three-dimensional look-up table and the polynomial regression algorithm were built using the test data of target color． Then the experiment was designed to compare the accuracy of the two algorithms. The maximum average color difference of the three-dimensional look-up table model reached 3，and the tetrahedron algorithm accuracy was slightly higher than the other three methods, the average color difference reached 0.64． Polynomial regression algorithm accuracy improved with the increase of the number of items, and the average color difference of the 20 items was 2.58. The accuracy of the three-dimensional look-up table model was higher than that of the polynomial regression model, and it had high precision and good effect. It was proved that the three-dimensional look-up table method could determine whether there was color cast in the printing image acquired by CCD, and the conversion accuracy of this method increased as the RGB color space was divided finer.