Method for Generating Threshold Arrays for Clustered-Dot Halftone Screening
Publication Date: 2003-Jan-29
The IP.com Prior Art Database
AbstractThis report describes a method for generating threshold arrays for clustered-dot halftone screening.
Author: Mike Slone-Murphy
Document No. R321
Date: 5 July 2001
FUJIFILM Electronic Imaging Ltd, 5 July 2001
This report describes a method for generating threshold arrays for clustered-dot halftone screening.
Clustered-dot ordered dithering is the conventional method of halftoning that has been widely used throughout the printing industry in one form or another since the mid 19th century . Despite the introduction of more exotic forms of halftoning made possible by electronic and computing advances, it remains the most durable and flexible option available for the industry, particularly where high-volume output is a priority.
Recent years have seen stiff competition everywhere, and the ever-increasing demand for quality and productivity together with the ubiquitous availability of computing power, has taken clustered-dot halftoning well into the realms of mathematics and computing. In this report, it will be shown how this form of halftoning involving the use of a threshold array, now part of a standard paradigm, can be controlled, measured and optimised before being used with ink and paper.
The method of constructing a threshold array in two or three principal stages is presented here. First the basic array is written using a spot function, then a second, optional stage performs histogram equalisation and organises the array data spatially to ensure minimal processing in the third stage. Finally, the third stage is a multi-pass process that successively reorganises the numerical order of the spots for optimal spatial frequency properties.
In this report, a threshold array is an area of computer memory organised as a 2-dimensional square of size MxM, where M can be typically between 32 and 1024. Values in the array define the shape and size of the halftone dots for each image signal value across the image plane. The array contains integers typically in the range 0-1022, which can then be processed further for calibration to 1-255. Later versions of PostScript™ support higher ranges than this, but the basic principles under discussion remain the same. Fig. 1 is a hidden line representation of a typical threshold array, this one containing the information needed to generate several dots within a region called a supercell, which is tiled across the image plane.  in particular and  refer to the supercell,  describing this device as a square box of dot cells. In particular,  and  generalise the supercell by transposing two triangular areas in the plane so as to be adjacent to the x and y axes for convenience of scanning. The result is an L-shape region consisting of two distinct squares. Exam...