Clustered halftone dot design process

Claim 1

A printing system suitable for forming marks on a substrate at one of c possible levels of colorant, and receiving image signals represented at d possible levels, where d>c, and having a halftone processor to reduce the number of levels at which the image is represented from d levels to c levels in order to replicate gray level printing, including: a memory, storing a set of halftone threshold level signals, each threshold signal corresponding to a unique location in a halftone cell; a comparator, receiving said image signal and one of the halftone threshold signals from the memory, and producing an output signal at one of c possible levels, varying according to the comparison of said halftone threshold signal to said image signal to cause marking at a selected location on a substrate; said set of threshold level signals together forming a screen matrix arranged for use with respect to the image to generate multiple clusters of printed spots within a single repetition of the screen and generated by a dot clustering optimization process which optimizes the screen by approximating a condition where: A. At any level of the input, all clusters have the same shape and the same size; and B. All adjacent clusters are equal-distance separated.

Claim 3

A method of designing a halftone cell, for converting an image received at d levels, for reproduction at two levels, said cell having elements each corresponding to a pixel in a discrete area of an image the method including the steps: a) choosing a desired cluster pattern for said halftone cell; b) assigning threshold values to the halftone cell; c) generating halftone screens; d) calculating a quality metric, by measuring divergence from the conditions: 1. At any level of the input, all clusters have the same shape and the same size; and 2. All adjacent clusters are equal-distance separated; e) randomly choosing a limited set of threshold values and swapping the threshold values; f) recalculating the quality metric and determining whether the quality metric has improved; g) determining whether to keep the swap or returning the screen to its initial state based on the determination of whether the quality metric has improved.

Claim 4

A method of designing a halftone cell, for converting an image received at d levels, for reproduction at two levels, said cell having elements each corresponding to a pixel in a discrete area of an image the method including the steps: a) choosing a desired cluster pattern for said halftone cell; b) assigning threshold values to the halftone cell; c) generating halftone screens; d) calculating a quality metric, by measuring divergence from the conditions: 1. At any level of the input, all clusters have the same shape and the same size; and 2. All adjacent clusters are equal-distance separated; e) randomly choosing a limited set of threshold values and swapping the threshold values; f) recalculating the quality metric and determining whether the quality metric has improved; g) determining whether to keep the swap or returning the screen to its initial state based on the determination of whether the quality metric has improved.

Claim 12

A method of generating threshold values appropriately positioned in a halftone cell, for converting an image received at d levels, for reproduction at c levels, where c is less than d, said cell having N elements each corresponding to a pixel in a discrete area of an image the method including the steps: selecting a number of clusters for a halftone cell; a determining a threshold values between 1 and N corresponding to each elements; measuring a quality metric for the halftone cell, as a function of uniformity of shape and size of clusters, and distance between clusters; making at least one swap of threshold values between elements; re-measuring the quality metric; reversing the swap of threshold values if the quality metric change is not appropriate.

Claim 13

A method of generating threshold values appropriately positioned in a halftone cell, for converting an image received at d levels, for reproduction at c levels, where c is less than d, said cell having N elements each corresponding to a pixel in a discrete area of an image the method including the steps: selecting a number of clusters for a halftone cell; a determining a threshold values between 1 and N corresponding to each elements; measuring a quality metric for the halftone cell, as a function of uniformity of shape and size of clusters, and distance between clusters; making at least one swap of threshold values between elements; re-measuring the quality metric; reversing the swap of threshold values if the quality metric change is not appropriate.