| 摘要 |
PROBLEM TO BE SOLVED: To estimate spectral distribution of basic color material by multiplying the spectral distribution by the weight of the quantity of color material and the mixing ratio and then converting it into a spectral distribution which is varied as the basic composition, quantity or mixing ratio of the color material is varied. SOLUTION: An input 1 reads in the spectral distribution of a color. material. A conversion processing 2 multiplies the spectral distribution of standard mixing quantity and mixing ratio (referred to basic spectral distribution) by the mixing quantity and mixing ratio. A conversion processing 3 comprises a plurality of neural networks (referred to network) and converts the results of conversion processing into a spectral distribution which is varied as the basic composition, mixing quantity or mixing ratio of the color material is varied. A preserving section preserves coefficients used at the conversion processings 2, 3. When it is used, parameters of the network of a color material to be estimated are read out from the preserving section and set in the network of the conversion processing 3. Basic spectral distribution of the color material having altered basic composition is then inputted and imparted to the conversion processing 2 where it is multiplied by the mixing quantity and mixing ratio. It is then converted, by the network of the conversion processing 3, into a spectral distribution corresponding to the mixing quantity and mixing ratio before being outputted 4.
|
