Image analysis – Image compression or coding – Quantization
Reexamination Certificate
1999-01-05
2002-08-27
Lee, Thomas D. (Department: 2624)
Image analysis
Image compression or coding
Quantization
C382S237000
Reexamination Certificate
active
06442300
ABSTRACT:
BACKGROUND OF THE INVENTION
A method of generating one dot for each three pixels on one line of an image to be printed to lower the frequency of the image while keeping the detail, the dot having either an all-white central portion or two all-black end portions.
For maintaining sharp edges for printed text and line graphics it is necessary to have a large number of pixels per inch, and yet there is an upper limit to the number of dots per inch that can be used to print xerographic images that are received as grayscale. A common example is the technique of clustering where an 8 by 8 pixel screen is used to reduce sixty-four 8-bit pixels to one halftone dot.
One prior art method of improving the apparent amount of detail in a dot image without changing the dot frequency is high addressability. Here, the printing system may have a number of dots per inch that is so low that diagonal lines appear jagged to the eye. Increasing the number of dots per inch is clearly an improvement, but is expensive. A lower cost alternative is to divide up each dot at the printer into a number of parts, and to control each part separately, to result in dots that are larger or smaller than the originals. When correctly done, it enhances image quality without incurring the cost of increasing the number of dots per inch.
An additional prior art technique to improve image quality is error diffusion. If a highly precise quantity must be approximated by an image dot of lower precision, the error term is carried to the next dot and to the dots on the next raster. In this way, the precision is not lost. This process is described in detail in U.S. Pat. No. 5,317,653, which is incorporated by reference herein.
SUMMARY OF THE INVENTION
This invention combines high addressability, error diffusion and clustering to maintain as much high-frequency image quality as possible while reducing a number of pixels to one dot. It is thus applicable for scanned images containing much detail, as well as text, line art and previously generated halftones.
Any number of pixels can be combined using this method, but three pixels will be used as a numerical example for this discussion. This clustering process first adds the three pixels to form one number which indicates the total amount of black If the total amount of black is more than ⅔ of maximum, then each end pixel is set to black and the central pixel will be set to the appropriate shade of gray. Next, high addressability is used to convert the gray central pixel to one having white central sub pixels and black end sub pixels. The three generated pixels are then connected, which forms one dot for the three-pixel line having a central white portion and black end portions.
If the total amount of black is less than ⅔ of maximum, then the central pixel is set to white and the end pixels are set to shades of gray. Next, the left gray pixel is converted into a set of sub pixels where there are white sub pixels to the right of the black, and the right gray pixel is converted into a set of sub pixels where there are white sub pixels to the left of the black. The three generated pixels are then connected, which again forms a dot for the three-pixel line having a central white portion and black end portions.
In this way, dot density is reduced by a factor of three. One advantageous characteristic of this process is that the left and right ends of each dot need not be equal. For example, if the original left pixel of the three is darker than the right, then the left black portion of the dot can be set larger than the right black portion. This increases the preservation of original detail. However, the white portion is always centered so that the white portion can not preserve this detail. The question then arises, whether the center of the dot should be white or black. It was decided that the ends should be black. This way, the greater accuracy is in the highlights, and less in the shadows.
REFERENCES:
patent: 5317653 (1994-05-01), Eschbach et al.
patent: 5438431 (1995-08-01), Ostromoukhov
patent: 5444551 (1995-08-01), Miller et al.
patent: 5489991 (1996-02-01), McMurray et al.
patent: 5625716 (1997-04-01), Borg
patent: 5808755 (1998-09-01), Delabastita
patent: 6122407 (2000-09-01), Peters
patent: 6130661 (2000-10-01), Ilbery
Brinich Stephen
Lee Thomas D.
Xerox Corporation
LandOfFree
Three-pixel line screen with high addressability and error... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Three-pixel line screen with high addressability and error..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Three-pixel line screen with high addressability and error... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2915574