| 摘要 |
There is disclosed a color image processing method and apparatus therefor, in which the input image data are converted into a color coordinate system represented by luminosity and hue, for example L*a*b* space. The colors reproducible on the output device, for example a printer, are also converted into this coordinate system. An average value is calculated over plural pixels in the vicinity of an object pixel to be digitized, and plural average values are then obtained by including the object pixel selected at each of the reproducible colors. The input image signal is digitized to the closest one of the plural average signals. The digitization error generated upon digitization is distributed to the succeeding pixels with weighted ratios. This method achieves faithful color reproduction, matching the colors reproducible on the recording or display device.
|
