Facsimile and static presentation processing – Static presentation processing – Attribute control
Reexamination Certificate
2000-02-29
2004-11-30
Lee, Thomas D. (Department: 2624)
Facsimile and static presentation processing
Static presentation processing
Attribute control
C358S003060, C358S003200
Reexamination Certificate
active
06825953
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a halftone method and more particularly to a halftone method which minimizes image distortion introduced by the halftoning process.
BACKGROUND OF THE INVENTION
Many types of printing devices use a limited variety of ink (or toner) colors for the production of printed pages (typically one color for monochrome devices and four colors for color devices). These devices are generally bistable with respect to the output of individual ink levels. In particular, output of the ink is either enabled or disabled at any given location on a page. In such devices, intermediate tones are created using a process called halftoning, in which ink output is modulated against a background defined by the color of the print media (e.g. paper) so as to create an average tone which will be visually integrated by the viewer.
Owing to historical methods for achieving the effect of halftoning, the halftone process is synonymously referred to as halftone screening or simply screening.
The process of halftoning is widely recognized, by those skilled in the art, as a potential source of image content distortion. This distortion is typically manifested as 1) a loss of sharpness, 2) introduction of moiré, and 3) exaggeration of image noise. These effects are largely due to a phenomenon known as aliasing, which in this context, is defined as the misrepresentation of image content due to inadequate sampling. This occurs because arbitrarily applied halftoning can effectively introduce its own sampling grid—which may be insufficient to yield a proper representation of the image's contents.
The phenomenon of aliasing, as it relates to the process of sampling, is known and well documented in the public literature (esp. as sampling theory). For purposes of understanding the invention to be described below, inadequate sampling can be conceptualized as sparse sampling. More particularly, there is information in the original image which does not contribute to the resulting image because it “slips through the cracks” in the (sparse) sampling grid. The standard solution to this problem is to match the sampling resolution with a (pre-sample) blurring process which “spreads out” the representation of image details to ensure that no detail remains spatially isolated (in the “cracks”) such that it can avoid being sampled.
As previously discussed, the standard way to avoid aliasing is to prefilter the data being sampled with a blurring filter. This blurring process is sometimes called smoothing and may be referred to technically as low-pass filtering (this name derives from the nature of the filter—that low frequency content is passed through unaltered while high frequency content is attenuated). Blurring an image has the disadvantage that edge and detail content become less discernible, which generally reduces the effectiveness or visual appeal of the image.
The use of a sharpening filter (sometimes called unsharp masking) can improve the visibility of edge and detail image content. Such a filter operates by increasing the localized contrast where edges and details are present, thereby increasing the localization and amplitude of such content. Application of a sharpening filter can be used to counteract the loss of sharpness which may be associated with halftoning. Unfortunately, sharpening an image has the disadvantage that it increases the visibility of image noise and strengthens the potential moiré interference between high frequency image content and the halftone pattern.
Sharpening and blurring are opposite types of filters and are effectively mutually exclusive of one another. To be effective on documents with varied content, one or both types of filtering may need to be applied dynamically—based on the results of a content analysis algorithm. The problem with this type of processing is that it tends to be complex and prone to error.
Another way to avoid halftone induced aliasing is to use very high frequency or dispersed-dot halftoning. Both of these types of halftoning utilize very small halftone dots—often placing them very close together as well. These methods can avoid introducing aliasing if their effective sampling grid is well aligned with that of the image being halftoned. This may, in fact, be the preferred solution for some types of devices—most notably printers based on ink-jet technology (which may actually require dispersed-dot halftoning due to problems with alignment and over-saturation of ink). However, due largely to output instabilities resulting from cost/performance tradeoffs, many printing devices still work best with lower frequency clustered halftones.
Accordingly, it is an object of the invention to provide an improved halftone process.
It is a further object of the invention to provide an improved halftone process that avoids creation of image artifacts.
SUMMARY OF THE INVENTION
The invention provides a computer-implemented procedure that is applicable to various halftoning methods, which minimizes image distortion introduced by the halftoning process. The procedure involves maintaining specific relationships between the halftone pattern and the phase and resolution of the input image. Specifically, that the halftone pattern is designed such that it does not bias tone reproduction with respect to the location of the input image pixels—each input pixel should be given equal weight when filtered by the halftoning process. This suggests a 1 to 1 relationship between each pixel of the input image and the corresponding tone output produced by the halftoning process. The procedure results in improved print quality, manifested by avoidance of certain types of moiré which are commonly associated with halftoning, as well as improved reproduction of edges and image details.
A preferred embodiment of the invention creates a scaled halftoned destination pixel image from a source pixel image portion that comprises plural source pixels. (1) Initially, a scaled set of matrix values is created to be used for each of the source pixels. (2) Then, depending upon a determined halftone relationship of a source pixel and each of said scaled set of matrix values, a scaled set of destination pixels is created. Each respective created destination pixel value is dependent upon a relationship of a value of the respective source pixel and a corresponding matrix value. Thereafter, step
2
is repeated for each source pixel, using the scaled set of matrix values, and arranging the scaled set of matrix values to create a determined halftone pattern over the source pixels that exhibits a desired repeating pattern. Each scaled set of destination pixels produced thus exhibits a reduced tonal error as a result of use of the same scaled set of matrix values for each source pixel.
REFERENCES:
patent: 5526445 (1996-06-01), Smutek et al.
patent: 5647026 (1997-07-01), Kwarta
patent: 5822451 (1998-10-01), Spaulding et al.
patent: 5966507 (1999-10-01), Lin
patent: 0785528 (1997-07-01), None
patent: 2097883 (1982-08-01), None
patent: 2196508 (1983-04-01), None
patent: 2141805 (1985-01-01), None
patent: 08118440 (1990-07-01), None
patent: 11005700 (1999-12-01), None
Brinich Stephen
Lee Thomas D.
LandOfFree
Halftone employing 1 to 1 screening does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Halftone employing 1 to 1 screening, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Halftone employing 1 to 1 screening will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3335305