Geometrically reducing mask error diffusion

• US 6,108,452 A
• Filed: 01/06/1997
• Issued: 08/22/2000
• Est. Priority Date: 01/05/1996
• Status: Expired due to Fees

• Current Assignee: Canon Finetech Incorporated (Canon Inc.)
• Original Assignee: Canon Aptex, Inc. (Canon Inc.), Canon Information Systems Research Australia Pty Limited (Canon Inc.)
• Inventors: Ilbery, Peter William Mitchell
• Primary Examiner(s): Lee, Thomas D.
• Application Number:
  US08/779,357
• Time in Patent Office: 1,324 Days
• Field of Search: 382/251, 382/252, 382/232, 382/237, 358/448, 358/455, 358/456, 358/457, 358/458, 358/447
• US Class Current: 382/252
• CPC Class Codes: H04N 1/4052 by error diffusion, i.e. transferring the binarising error to neighbouring dot decisions

Abstract

An image processing system in which a computer image is halftoned by halftoning a current pixel using error diffusion, with a halftoned output value of the current pixel being determined using errors derived from substantially all previously halftoned pixels of the image. Output error values are determined for each pixel, and are acted upon by an error diffusion mask of geometrically reducing values. In certain instances, previously halftoned pixels are divided into two or more groups to obtain a desired contribution from previously halftoned pixels.

Claims

1. A method of error diffusing an image, the image having a plurality of scan lines each having a plurality of pixels, said method comprising, for each of said pixels, the steps of:

applying an error diffusion mask to an output error value of previously processed pixels to determine an input error value for said pixel wherein the error diffusion mask comprises error diffusion coefficients corresponding to all previously error diffused pixels of said image; and

determining an output value and an output error value of said pixel, both said output pixel value and said output error value being dependent upon the input error value and an input pixel value of said pixel.

2. A method as claimed in claim 1, wherein said error diffusion mask comprises at least one two-dimensional array of error diffusion coefficients geometrically decreasing by a first predetermined ratio for each element of the array away from a first predetermined error diffusion coefficient in one dimension of the array and geometrically decreasing by a second predetermined ratio for each element of the array away from the first predetermined error diffusion coefficient in the other dimension of the array.

3. A method as claimed in claim 1, wherein each pixel is sequentially processed in consecutive scan lines in a raster-scan manner.

4. A method as claimed in claim 1, wherein each pixel is sequentially processed on each alternate scan line in one direction and sequentially processed on each intermediate alternate scan line in the opposite direction.

5. A method of halftoning an image, said image comprising a plurality of pixels arranged in scan lines, said method comprising the step of error diffusing each pixel of said image characterized in that a halftoned output value and an output error value of said pixel are determined from error values obtained from all previously halftoned pixels of said image, both said output pixel value and said output error value being defendant upon an input error value and an input pixel value of said pixel.

6. A method as claimed in claim 5, wherein said method comprises, for a current pixel of said image:

(a) receiving an input pixel value of said current pixel;

(b) determining a derived input value of said current pixel by summing said input pixel value with an input error value, wherein the input error value is determined from output error values of all pixels of said image that preceded said current pixel;

(c) using said derived input value to determine a halftoned output value of said current pixel; and

(d) determining an output error value of said current pixel by subtracting said halftoned output value from said derived input value.

7. A method as claimed in claim 6, wherein said step (c) comprises selecting a closest one of a plurality of output values to said derived input value and outputting for display the selected value as said halftoned output value.

8. A method as claimed in claim 6, wherein said input error value is determined by applying an error diffusion mask to output error values of said substantially all of said preceding pixels, said mask having error diffusion coefficients corresponding to each of said substantially all preceding pixels.

9. A method as claimed in claim 8, wherein said error dimension mask comprises at least one two dimensional array of error diffusion coefficients geometrically decreasing by a first predetermined ratio for each element of the array away from a first predetermined error diffusion coefficient in one dimension of the array and geometrically decreasing by a second predetermined ratio for each element of the array away from the first predetermined error diffusion coefficient in the other dimension of the array.

10. A method as claimed in claim 8, wherein said substantially all preceding pixels are divided into a plurality of groups of pixels and a corresponding mask is applied to each of said groups to provide a corresponding plurality of group error values, said group error values being summed to provide said input error value.

11. A method as claimed in claim 8, wherein one or more of said groups comprise wedges of pixels that extend substantially radially from said current pixel.

12. Apparatus for halftoning an image, said apparatus comprising:

receiving means for receiving input pixels of said image intended for display in raster scan order, and

error diffusing means for error diffusing said input pixels to provide error diffused halftoned output pixels for display and an output error value associated with each output pixel, wherein a halftoned output value of a current pixel of said image is determined from input error values obtained from all previously halftoned pixels of said image, and further wherein both said output pixel value and said output error value are dependent upon the input error values and an output pixel value of said pixel.

13. A computer software system for executing instructions on a general purpose computer, wherein said instructions comprise a plurality of process modules for halftoning an image, said image comprising a plurality of pixels arranged in a plurality of scan lines, said modules comprising:

a first module for determining a derived input value of a current pixel of said image by summing an input pixel value of said current pixel with an input error value;

a second module for determining said input error value from output error values of all pixels of said image preceding said current pixel;

a third module for determining a halftoned output value for said current pixel from said derived input value; and

a fourth module for determining an output error value, both said output value and said output error value being dependent upon the input error value and an input pixel value of said current pixel.

14. A system as claimed in claim 13, wherein said instructions are configured for operation via a computer readable medium, said medium being one of a memory device, a computer, and a computer network.