【摘要】 针对函数转换方法速度快但精度差,查表插值方法精度高但速度慢的特点,提出将函数转换和查表修正相结合的色彩匹配方法.先建立基本的转换关系,再通过查表的方法进行修正.基本转换关系和修正表的建立过程,都通过机器学习方法完成.其特点是不依赖于特定的打印环境,即在不同的打印环境下,可通过对样本数据的学习,自动建立基本转换关系和修正表.更多还原

【Abstract】 Function-based transformation is speedy but poor in precision, while look-up table method is slow but precise. Considering these characteristics of the two color matching ways, a new method which combines these two ways for color matching is proposed in this paper. The basic transformation - function is set up at first, and then the correction-table is built. Both steps are processed by machine learning independent to printing conditions, that is to say, the transformation-function and correction-table can be set up by learning from sample data automatically under different printing conditions.更多还原