| 摘要 |
Tone reproduction of an image is improved by the use of error diffusion or dithering in the mapping of input pixel values on an input values scale to printer pixel values on a printer values scale, where the input values scale and the printer values scale both have more than two levels, to generate an image where a spatial average of printer values approximates a spatial average of the corresponding input values, as measured on the printer values scale. To improve the tone reproduction of reduced-resolution images, the printer pixel values are grouped into multi-pixel cells, printer pixel values within each cell are mapped to a font index value, and a font cell corresponding to the index level is printed at the location of the cell. The font set is chosen such that differences in total deposited ink between consecutive font cells is minimized. In the font cell having a font index value of zero all printer pixel values are zero, in the font cell having a font index value of one all printer pixel values are zero except for a single pixel having an isolated-single-pixel cutoff value, and as the font index value of subsequent font cells is incremented a printer pixel value is incremented such that all printer pixel values are zero except for pixels in a line along the scan direction. The final font cell consists of a line of pixels along the scan direction having a full-cell-width printer pixel value, and pixels of zero value elsewhere. The dynamic range of the font set is extended by a linear remapping of font index values.
|
