| 摘要 |
Improved processes for quantizing a set of mean-preserving Tone Transfer Functions (TTFs) for a multi-level halftoning system which minimize or at least significantly reduce the overall quantization error in the system as a function of a prespecified system halftone cell size, P, and the number of micro output levels, Ly, supported by the multi-level halftoning system. The processes call for dividing the range of possible input contone representative intensity values into T subintervals, where T=PLy; determining the values of the centers of the T-subintervals; and quantizing each of the normalized continuous mean preserving TTFs as a function of the values of the centers of the T-subintervals. According to a preferred embodiment, the quantizing is performed such that the quantized TTFs approximate the normalized continuous mean preserving TTFs and ideally satisfy the criteria that the maximum quantization error between any quantized and unquantized TTF is minimized. Further aspects of the invention are directed to methods for determining threshold levels for assignment to threshold matrices, and methods for specifying look-up tables, either of which (the matrices and look-up tables) may be used for converting input contone representative intensity values into a predetermined number of available micro output levels for a multilevel halftone system having a predetermined halftone cell size.
|
