Claims
- 1. An image processing method for converting an N-value image to an M-value image, wherein N<M, said method comprising the steps of:
- obtaining an appearance frequency distribution for each of a plurality of different N-value pixel patterns, each appearance frequency distribution representing frequencies of occurrence of a plurality of different M-values; and
- estimating M-value image data from the N-value image in accordance with the appearance frequency distributions obtained in said obtaining step,
- wherein said estimating step uses values based on a plurality of peaks when the plurality of peaks exist in a respective appearance frequency distribution.
- 2. A method according to claim 1, wherein the step of estimating the M-value image data includes the step of estimating M-value image data at a barycenter of a histogram of the frequency distribution.
- 3. A method according to claim 1, wherein said estimating step includes estimating M-value image data at a maximum frequency of an appearance frequency distribution.
- 4. A method according to claim 1, wherein the, step of estimating the M-value image data includes the step of estimating M-value image data located at a center of all the frequencies of the frequency distribution.
- 5. An image processing method for converting an N-value image to an M-value image, wherein N<M, said method comprising the steps of:
- discriminating which N-value pixel pattern, among a plurality of different N-value pixel patterns, is disposed around an objective pixel of the N-value image;
- obtaining an appearance frequency distribution for each of a plurality of different N-value pixel patterns, each appearance frequency distribution representing frequencies of occurrence of a plurality of different M-values; and
- estimating M-value image data, in accordance with (i) the N-value image and (ii) the appearance frequency distribution corresponding to the N-value pixel pattern discriminated in said discriminating step,
- wherein said estimating step uses values based on a plurality of peaks when the plurality of peaks exist in a respective appearance frequency distribution.
- 6. A method according to claim 5, wherein the feature of the objective pixel indicates whether the objective pixel belongs to an edge or non-edge section.
- 7. A method according to claim 5, wherein the step of estimating the M-value image data includes the step of estimating M-value image data at a barycenter of a histogram of the frequency distribution.
- 8. A method according to claim 5, wherein said estimating step includes estimating M-value image data at a maximum frequency of the respective appearance frequency distribution.
- 9. A method according to claim 5, wherein the step of estimating the M-value image data includes the step of estimating M-value image data located at a center of all the frequencies of the frequency distribution.
- 10. A method according to claim 5, wherein multivalue conversion is performed based on data estimated in units of features.
- 11. An image processing method for converting an N-value image to an M-value image, wherein N<M, said method comprising the steps of:
- discriminating which N-value pixel pattern, among a plurality of different N-value pixel patterns, is disposed around an objective pixel of the N-value image;
- obtaining an appearance frequency distribution for each of the plurality of different N-value pixel patterns, each appearance frequency distribution representing frequencies of occurrence of a plurality of different M-values;
- estimating an M-value from each respective appearance frequency distribution; and
- estimating M-value image data in accordance with the N-value image and the M-value estimated in said step of estimating an M-value,
- wherein said estimating step uses values based on a plurality of peaks when the plurality of peaks exist in a respective appearance frequency distribution.
- 12. A method according to claim 11, wherein the nature of the objective pixel indicates whether the objective pixel belongs to an edge or non-edge section.
- 13. A method according to claim 12, wherein if M-value image decoded values estimated for the edge and non-edge sections are represented by ave1 and ave2, (ave1+ave2)/2 is determined as the M-value image decoded value.
- 14. A method according to claim 12, wherein if M-value image data estimated values of the edge and non-edge sections are represented by ave1 and ave2, and M-value image total appearance frequencies sum1 and sum2 of the edge and non-edge sections in a given N-value pattern m are represented by sum1 and sum2, an M-value image decoded value for the N-value pattern m is given by (ave1.times.sum1+ave2.times.sum2).div.(sum1+sum2).
- 15. A method according to claim 11, wherein the step of estimating the M-value image decoded values in units of features includes the step of estimating an M-value image value at a barycenter of a histogram of the frequency distribution for each of the features.
- 16. A method according to claim 11, wherein said estimating step comprises a step of determining an M-value corresponding to a maximum frequency of an appearance frequency distribution.
- 17. A method according to claim 11, wherein the step of estimating the M-value image values in units of features includes the step of estimating an M-value image value located at a center of all the frequencies of the frequency distribution for each of the features.
Priority Claims (5)
Number |
Date |
Country |
Kind |
2-126371 |
May 1990 |
JPX |
|
2-126372 |
May 1990 |
JPX |
|
2-126373 |
May 1990 |
JPX |
|
2-126374 |
May 1990 |
JPX |
|
2-126375 |
May 1990 |
JPX |
|
Parent Case Info
This application is a continuation of application Ser. No. 07/700,762, filed May 15, 1991, now abandoned.
US Referenced Citations (10)
Continuations (1)
|
Number |
Date |
Country |
Parent |
700762 |
May 1991 |
|