论文部分内容阅读
通过分析大量实测的打印机输入(R,G,B)与输出(L ,a ,b )数据之间的对应关系,发现在一些颜色区域存在着RGB与L a b 之间的二对一甚至多对一的关系。另一方面,一个已知的事实是,当一个神经网络已经建立后,任一个(或一组)输入只能产生一个(或一组)输出。结合以上两个事实得出的结论是,在要求打印机的输出颜色能准确地复现指定颜色的情况下,神经网络不适合作为由指定颜色的L a b 值来寻求打印机的相应输入RGB的工具。
By analyzing the correspondence between a large number of measured printer inputs (R, G, B) and the output (L, a, b) data, it is found that in some color regions two to one or even multiple pairs of RGB and L ab A relationship. On the other hand, it is a known fact that when a neural network has been established, any (or group) of inputs can produce only one (or a set) of outputs. The combination of the above two facts leads to the conclusion that the neural network is not suitable as a tool to find the corresponding input RGB of the printer from the L a b value of the specified color in case the output color of the printer is required to accurately reproduce the specified color.