Method and device for compressing image data

• US 20070013953A1
• Filed: 07/10/2006
• Published: 01/18/2007
• Est. Priority Date: 07/13/2005
• Status: Active Grant

• Current Assignee: Konica Minolta Business Technologies Incorporated (Konica Minolta Inc.)
• Original Assignee: Konica Minolta Business Technologies Incorporated (Konica Minolta Inc.)
• Inventors: Yamamoto, Toshitsugu
• Granted Patent: US 7,746,501 B2
• Time in Patent Office: Days
• US Class Current: 358/3.130
• CPC Class Codes: H04N 1/41 Bandwidth or redundancy reduction by scanning H04N1/17 ; methods or arrangements for coding, decoding, compressing or decompressing digital video signals H04N19/00

Abstract

A method for compressing image data includes a first step for dividing image data into predetermined areas and determining a representative value about density of pixels in each area, a second step for determining density of pixels included in each area by using the representative value of the area and an interpolation rule, a third step for determining reproducibility of an image obtained, on the basis of the density of pixels determined in the second step, by using the image data and a dither pattern that is used for a pseudo gradation process performed on the image data, and a fourth step for encoding, on the basis of the representative value determined in the first step, for areas that are determined to have a good reproducibility in the third step.

Claims

1. A method for compressing image data having a gradation property, the method comprising:

a first step for dividing the image data into predetermined areas, so as to determine a representative value of densities of pixels in each area;

a second step for determining a density of each pixel included in the area using the representative value of the area and an interpolation rule;

a third step for testing reproducibility of an image obtained on the basis of the density of each pixel determined in the second step, by using the image data and a dither pattern that is used for a pseudo gradation step performed on the image data; and

a fourth step for performing an encoding process on the basis of the representative value determined in the first step for areas where the reproducibility was good in the test of the third step.

2. The method according to claim 1, wherein the image that is obtained on the basis of the density of each pixel is a virtual gradation image on the basis of the density of each pixel.

3. The method according to claim 1, wherein the image obtained on the basis of the density of each pixel is a pseudo gradation image that is obtained by a pseudo gradation process on the virtual gradation image on the basis of the density of each pixel using the dither pattern.

4. The method according to claim 1, wherein the representative value is a density value of a specific pixel in the area.

5. The method according to claim 1, wherein the representative value is an intermediate value of density values of a plurality of pixels in the area.

6. The method according to claim 1, wherein the representative value is an intermediate value of density values of a specific pixel and pixels surrounding the specific pixel in the area.

7. The method according to claim 1, wherein the representative value is a value between a minimum value K of a threshold value in the case where a threshold value of the dither pattern is larger than a density value of the image data and a maximum value J of a threshold value in the case where a threshold value of the dither pattern is smaller than a density value of the image data.

8. The method according to claim 1, wherein the interpolation rule is a process in which the density value of each pixel included in the area is equal to the representative value of the area.

9. The method according to claim 1, wherein the interpolation rule is a process of obtaining the density value of each pixel included in the area by interpolation using a representative value of the area and representative values of one or more neighboring areas of the area.

10. The method according to claim 9, wherein the interpolation is a linear interpolation.

11. The method according to claim 9, wherein the interpolation is a spline interpolation.

12. The method according to claim 1, wherein the compressing process is performed by determining difference information between representative values of neighboring areas and encoding the determined difference information.

13. The method according to claim 1, wherein the encoding process is performed on the basis of the image data in an area where reproducibility was not good in the test of the third step.

14. The method according to claim 1, wherein, in the third step, an error range in a case where the image data are character image data is set smaller than an error range in a case where the image data are picture image data when the reproducibility is tested.

15. A device for compressing image data having a gradation property, the device comprising:

a first portion for dividing the image data into predetermined areas, so as to determine a representative value of densities of pixels in each area;

a second portion for determining, for each area, a density of each pixel included in the area using the representative value of the area and an interpolation rule;

a third portion for testing reproducibility of an image obtained on the basis of the density of each pixel determined by the second portion, by using the image data and a dither pattern that is used for a pseudo gradation step performed on the image data; and

a fourth portion for performing an encoding process on the basis of the representative value determined by the first portion for areas where the reproducibility was good in the test of the third portion.

16. The device according to claim 15, wherein the fourth portion performs the encoding process on the basis of the image data in an area where reproducibility was not good in the test of the third portion.

17. A method for compressing image data having a gradation property, the method comprising:

a first step for dividing the image data into predetermined areas, so as to determine a representative value of densities of pixels in each area;

a second step for determining, for each area, a density of each pixel included in the area using the representative value of the area and an interpolation rule;

a third step for comparing a first dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on the image data using a dither pattern with a second dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on a density of each pixel determined in the second step using the dither pattern so as to determine a mismatch value indicating a difference between the first dot pattern and the second dot pattern, and for determining whether or not to use the representative value obtained in the first step for a process for generating compressed data on the basis of the mismatch value and a predetermined criterion value; and

a fourth step for performing the process for generating the compressed data by using the representative value if it was determined that the representative value should be used for generating the compressed data.

19. The method according to claim 17, wherein the process for generating the compressed data is performed by using the first dot pattern in the fourth step if it was determined that the representative value should not be used for generating the compressed data in the third step.

20. The method according to claim 17, wherein it is determined in the third step that the representative value obtained in the first step should be used for generating the compressed data if the mismatch value is smaller than the predetermined criterion value.

21. The method according to claim 17, wherein the fourth step includes performing the process for generating the compressed data by encoding the first dot pattern or the representative value.

18. A method for compressing image data having a gradation property, the method comprising:

a first step for dividing the image data into predetermined areas, so as to determine a representative value of densities of pixels in each area;

a second step for determining, for each area, a density of each pixel included in the area using the representative value of the area and an interpolation rule;

a third step for comparing a first dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on the image data using a dither pattern with a second dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on a density of each pixel determined in the second step using the dither pattern so as to determine a mismatch value indicating a difference between the first dot pattern and the second dot pattern, and for determining for each area whether or not to use the representative value obtained in the first step for a process for generating compressed data on the basis of the mismatch value and a predetermined criterion value; and

a fourth step for performing the process for generating the compressed data by using the representative value of an area where it was determined to use the representative value for generating the compressed data.

22. A method for compressing image data having a gradation property, the method comprising:

a first step for dividing the image data into predetermined areas, so as to determine a representative value of densities of pixels in each area;

a second step for determining, for each area, a density of each pixel included in the area using the representative value of the area and an interpolation rule;

a third step for comparing a first dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on the image data using a dither pattern with a second dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on a density of each pixel determined in the second step using the dither pattern so as to determine a mismatch value indicating a difference between the first dot pattern and the second dot pattern, and for determining a criterion value on the basis of the mismatch value;

a fourth step for determining, on the basis of the mismatch value and the criterion value, whether or not to use the representative value obtained in the first step for a process for generating compressed data; and

a fifth step for performing the process for generating the compressed data by using the representative value if it was determined that the representative value should be used for generating the compressed data.

23. A method for compressing image data having a gradation property, the method comprising:

a first step for dividing the image data into predetermined areas, so as to determine a representative value of densities of pixels in each area;

a second step for determining, for each area, a density of each pixel included in the area using the representative value of the area and an interpolation rule;

a third step for comparing a first dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on the image data using a dither pattern with a second dot pattern that is an arrangement pattern of dots obtained by performing a pseudo gradation process on a density of each pixel determined in the second step using the dither pattern so as to determine a mismatch value indicating a difference between the first dot pattern and the second dot pattern, and for determining a criterion value on the basis of the mismatch value;

a fourth step for determining, on the basis of the mismatch value and the criterion value, for each area whether or not to use the representative value obtained in the first step for a process for generating compressed data; and

a fifth step for performing the process for generating the compressed data by using the representative value of an area where it was determined to use the representative value for generating the compressed data.

24. A method for generating compressed data of image data having a gradation property, the method comprising the steps of:

dividing the image data into predetermined areas;

generating model data from the image data for each of the areas by performing a modeling process for extracting characteristics of the image data;

calculating in trial a quantity of data of thinned data generated by a thinning process for thinning data from the model data on the basis of a threshold level to determine a thinning threshold level that is a threshold level such that the quantity of data becomes a predetermined target quantity of data or less;

generating the thinned data by performing the thinning process on the basis of the determined thinning threshold level;

outputting the generated thinned data as the compressed data;

in the modeling process,

in the case where a noted area that is one of the areas is concerned,

determining a representative value of the noted area;

determining an interpolation density that is a density of each pixel included in the noted area by using the representative value of the noted area and an interpolation rule;

comparing a reproduced dot pattern that is a dot pattern of the noted area obtained by performing a pseudo gradation process on the interpolation density using a dither pattern with a pseudo gradation image dot pattern that is a dot pattern obtained by performing a pseudo gradation process on the image data in the noted area using the dither pattern;

generating the model data that can determine the representative value of the noted area if the reproduced dot pattern matches the pseudo gradation image dot pattern, while generating the model data that can determine the number of matching dots or the number of mismatching dots between the reproduced dot pattern and the pseudo gradation image dot pattern, a representative value of the noted area and the pseudo gradation image dot pattern if the reproduced dot pattern does not match the pseudo gradation image dot pattern;

in the thinning process,

in the case where the noted area is concerned,

determining which should be used for generating the thinned data, a value for deriving the representative value for determining the representative value that can be obtained from the model data or the pseudo gradation image dot pattern that can be obtained from the model data, on the basis of a comparison between the number of matching dots or the number of mismatching dots that is obtained from the model data and the thinning threshold level; and

generating the thinned data for the noted area by using the value for deriving the representative value if it was determined that the value for deriving the representative value should be used for generating the thinned data, while generating the thinned data for the noted area by using the pseudo gradation image dot pattern if it was determined that the pseudo gradation image dot pattern should be used for generating the thinned data.

25. The method according to claim 24, wherein the representative value of the noted area is the number of dots indicated in the pseudo gradation image dot pattern in the noted area.

26. The method according to claim 24, wherein the representative value of the noted area is a value “

a”

that satisfies the condition of J≧

a≧

K, when comparing a density of each pixel in the image data at a part corresponding to the noted area with a density of each pixel in the dither pattern at a part corresponding to the noted area so as to dot in a prepared pattern area if a density of a pixel of the image data is small, J denotes a minimum value of densities of pixels in the dither pattern corresponding to dotted points in the pattern area and K denotes a maximum value of densities of pixels in the dither pattern corresponding to undotted points in the pattern area.

27. The method according to claim 26, wherein the representative value of the noted area is a value having a minimum difference between itself and a representative value in a neighboring area.

28. The method according to claim 24, wherein if a dot pattern obtained by performing an interpolation process on a density of each pixel in the noted area in accordance with a predetermined interpolation rule using a tentative representative value for the noted area and a representative value of one or more areas neighboring the noted area and further by performing the pseudo gradation process using the dither pattern matches a dot pattern obtained by performing a pseudo gradation process on the image data in the noted area using the dither pattern, the tentative representative value is used as the representative value of the noted area.

29. The method according to claim 24, wherein the value for deriving the representative value in the noted area is a representative value of the noted area.

30. The method according to claim 24, wherein the value for deriving the representative value in the noted area is a difference between the representative value of the noted area and the representative value of an area just before the noted area.

31. The method according to claim 24, wherein in the thinning process if the number of mismatching dots that can be obtained from the model data is larger than the thinning threshold level, it is determined to use the pseudo gradation image dot pattern for generating the thinned data, while if the number of mismatching dots is smaller than the thinning threshold level, it is determined to use the value for deriving the representative value for generating the thinned data.

32. A device for generating compressed data of image data having a gradation property, the device comprising:

a dividing portion for dividing the image data into predetermined areas;

a modeling portion for generating model data from the image data for each of the areas by performing a modeling process for extracting characteristics of the image data;

a thinning threshold level determining portion for calculating in trial a quantity of data of thinned data generated by a thinning process for thinning data from the model data on the basis of a threshold level and for determining a thinning threshold level that is a threshold level such that the quantity of data becomes a predetermined target quantity of data or less;

a thinning portion for generating the thinned data by performing the thinning process on the basis of the determined thinning threshold level; and

a compressed data outputting portion for outputting the generated thinned data as the compressed data, wherein

the modeling portion,

in the case where a noted area that is one of the areas is concerned,

determines a representative value of the noted area,

determines an interpolation density that is a density of each pixel included in the noted area by using the representative value of the noted area and an interpolation rule,

compares a reproduced dot pattern that is a dot pattern of the noted area obtained by performing a pseudo gradation process on the interpolation density using a dither pattern with a pseudo gradation image dot pattern that is a dot pattern obtained by performing a pseudo gradation process on the image data in the noted area using the dither pattern, and

generates the model data that can determine the representative value of the noted area if the reproduced dot pattern matches the pseudo gradation image dot pattern, and generates the model data that can determine the number of matching dots or the number of mismatching dots between the reproduced dot pattern and the pseudo gradation image dot pattern, a representative value of the noted area and the pseudo gradation image dot pattern if the reproduced dot pattern does not match the pseudo gradation image dot pattern, and

the thinning portion,

in the case where the noted area is concerned,

determines which should be used for generating the thinned data, a value for deriving the representative value for determining the representative value that can be obtained from the model data or the pseudo gradation image dot pattern that can be obtained from the model data, on the basis of a comparison between the number of matching dots or the number of mismatching dots that is obtained from the model data and the thinning threshold level, and

generates the thinned data for the noted area by using the value for deriving the representative value if it was determined that the value for deriving the representative value should be used for generating the thinned data, and generates the thinned data in the noted area by using the pseudo gradation image dot pattern if it was determined that the pseudo gradation image dot pattern should be used for generating the thinned data.