JP2000331174A - Edge processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the edge connectivity by controlling a parameter in the direction where the output image data are thinned if the primary differential value of an attentional pixel is larger than the threshold with the secondary differential value set smaller than zero and then controlling the parameter in the direction where the output image data are thickened if the primary differential value of the attentional pixel is larger than the threshold with the secondary differential value set larger than zero respectively. SOLUTION: The primary differential value of an attentional pixel is calculated (S101) and the secondary differential value of the pixel is calculated (S102). An edge decision result is calculated by comparing the minimum differential value with the threshold and deciding the positive or negative secondary differential value (S103). The image data correction value is set and outputted as a parameter (S104) and then binarized by the similar gradation processing to generate an output image (S105). Then the parameter is controlled in the direction where the output image data are thinned if the primary differential value is larger than the threshold with the secondary differential value set smaller than zero and in the direction where the output image data are thickened if the primary differential value is larger than the threshold with the secondary differential value set larger than zero respectively.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an edge processing method for detecting and correcting character edges from a grayscale image read by a reading device such as an image scanner or a digital copying machine.

[0002]

2. Description of the Related Art Conventionally, digital image data read by a reading device such as an image scanner or a digital copying machine is subjected to processing such as detection and enhancement of an edge portion of the image, and then processed by a printer or a display device. By outputting to an output device, high-quality printing and a display image with high clarity can be expected. Therefore, a demand for a more advanced edge processing method is increasing. In addition, there is a demand for a binarization method for modulating and outputting multi-tone data into binary data while satisfying good reproducibility of an image edge portion, particularly for a binary printer and a multi-level printer in a binary mode. Is growing.

The configuration and operation of a conventional edge processing system will be described below.

FIG. 9 is a flowchart showing a conventional edge processing method, FIG. 10 is a block diagram showing a configuration of an edge detecting section in a conventional edge processing technique, and FIG. 11 is an explanatory diagram showing a primary differential operator in a conventional edge processing technique. ,
FIG. 12 is a block diagram showing a configuration of an edge correction unit in a conventional edge processing technique.

First, a processing flow for one pixel in the conventional edge processing method will be described with reference to FIG.

Here, image data D is input (S10
0), the maximum value Emax of the difference values in the four directions of up, down, left, right, right skew, and left skew is calculated (S108), Emax is compared with the threshold value TH_E (S109), and the result of the determination in S109 is obtained. The image data D is corrected according to Corr, and an edge correction output Q is output (S110).

Here, step S108 is a step of calculating a difference value in a total of four directions of a difference value Ea in the vertical direction, a difference value Eb in the left and right direction, a difference value Ec in the right oblique direction, and a difference value Ed in the left oblique direction. Step (SS1), Ea, Eb, Ec, E
| Ea |, | Eb |, | Ec |, |
Substep (SS2) for obtaining Ed |, | Ea |, |
And a sub-step (SS3) of calculating the maximum value Emax among Eb |, | Ec | and | Ed |.

In step S109, the judgment formula Em
When ax <TH, the determination result Corr is 0 (the pixel of interest is not the edge of the image), and when the determination formula Emax <TH is not 1, the determination result Corr is 1 (the pixel of interest is the edge of the image). Then, referring to this determination result Corr, step S110 is a sub-step (SS) where Q = D when the target pixel is not an edge (Corr = 0).
16) and when the pixel of interest is an edge (Corr =
1) includes a substep (S = Q + D + Emax / 4)
S17).

By repeating the flow of FIG. 9 in the main scanning direction and the sub-scanning direction, edge processing of the entire image is performed.

Next, the configuration of an edge detecting section in a conventional edge processing technique will be described with reference to FIG.

The edge detecting section has a first derivative calculating means 10 to which the image data D is input and an image storing means 12.

The primary differential calculating means 10 comprises a differential value calculating means 14, an absolute value calculating means 15, and a maximum value calculating means 16, and the image data D is converted to the primary differential calculating means 10.
Is input to the difference value calculating means 14. Also, the control signal D
A control unit 13 for sending L_CTR to the image storage unit 12 is provided.
_CTR is written, and DL, which is a line delay signal of the image data D, is read out to the primary differential calculation means 10 in accordance with the control signal DL_CTR.

Here, the difference value calculation means 14 forms a window (not shown) with the image data D and the line delay signal DL read from the image storage means 12,
11 (a), (b), (c),
(D) By performing the operator operation of the first derivative operators 17, 18, 19, and 20, the vertical difference value Ea,
The difference value Eb in the left-right direction, the difference value Ec in the rightward slant direction, and the difference value Ed in the leftward slant direction are calculated. The absolute value calculating means 15 converts the difference values Ea, Eb, Ec, and Ed into absolute values to calculate | Ea |, | Eb |, | Ec |, | Ed |
The maximum value calculating means 16 calculates | Ea |, | Eb |, | Ec |,
The maximum value Emax of | Ed | is calculated.

Next, the configuration of the edge correction unit in the conventional edge processing technique will be described with reference to FIG.

[0015] The edge correction unit calculates the Emax and the threshold value TH_E.
And a comparison means 22 which outputs 0 when the judgment result Corr is smaller than Emax and 1 when Emax <TH is not larger than the judgment result Corr.
rr is input and Emax / 4 is output as an edge correction value when Corr = 1, and 0 (no correction) is output as an edge correction value when Corr = 0 is output.
3 and the edge correction value of the correction value calculation means 23 are input,
An adder 21 is provided for adding the image data D and the edge correction value to generate an edge correction output Q.

[0016]

However, the above technique has a problem that when the threshold value TH to be compared with the edge amount is large, many pixels for which no edge is detected are generated, and the connectivity of the edge is deteriorated. are doing. And
Because of the pixels whose edges are not detected, the smoothness of the distribution of the detected edges is reduced.

On the other hand, when the threshold value TH to be compared with the edge amount is small, there is a problem that the edge tends to be thick.

Accordingly, an object of the present invention is to provide an edge processing method in which the thickness of an edge hardly depends on the MTF of a reading device.

[0019]

In order to solve this problem, an edge processing method according to the present invention is an edge processing method for detecting and correcting an edge portion of an image by inputting multi-gradation image data. Inputting image data, obtaining a first differential value of the pixel of interest from the image data, obtaining a second differential value of the pixel of interest from the image data, a first edge determination result in which the first differential value does not exceed a threshold, a first differential Three edge discrimination results of a second edge discrimination result whose value is equal to or more than a threshold value and whose secondary differential value does not exceed zero, and a third edge discrimination result whose primary differential value is not less than a threshold value and whose secondary differential value is not less than zero , And a parameter is set from the three edge determination results. Output image data is generated using the parameters. In the generation of the output image data, the parameter is set when the pixel of interest is the first edge determination result. When the target pixel is the second edge determination result, the parameter is controlled so that the output image data becomes thinner than the default value. When the target pixel is the third edge determination result, the output image data is set higher than the default value. The parameter is controlled in such a direction that the density becomes higher.

Thus, the connectivity of the edges is improved,
The edge thickness can be highly independent of the MTF of the reader.

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The invention according to claim 1 of the present invention is an edge processing method for detecting and correcting an edge portion of an image by inputting multi-gradation image data, wherein the image data is input. Then, the first differential value of the pixel of interest is obtained from the image data, the second differential value of the pixel of interest is obtained from the image data, the first edge determination result in which the first differential value does not exceed the threshold value, the first differential value is not less than the threshold value, and The second edge determination result in which the secondary differential value does not exceed zero, the third edge determination result in which the primary differential value is equal to or more than the threshold value and the secondary differential value is equal to or more than zero,
One edge determination result is obtained, a parameter is set from the three edge determination results, output image data is generated by the parameter, and in the generation of the output image data, the parameter is set to a default value when the pixel of interest is the first edge determination result. When the target pixel is the second edge determination result, the parameter is controlled so that the output image data becomes lighter than the default value. When the target pixel is the third edge determination result, the output image data becomes darker than the default value. This is an edge processing method for controlling parameters in the direction, and has the effect that the connectivity of edges is improved, and the thickness of the edge can be highly independent of the MTF of the reading device.

According to a second aspect of the present invention, in the first aspect, the output image data is an edge processing method generated by binarizing an image by a pseudo halftone process. Has an effect that the thickness of the edge can be highly independent of the MTF of the reading device.

According to a third aspect of the present invention, in the first or second aspect, the image data correction value is set as a parameter, and the generation of the output image data includes
The distribution error is calculated, the sum of the image data of the pixel of interest, the parameter, and the distribution error is binarized to generate the output image data, and the quantization error at the time of binarization of the pixel of interest is calculated and stored. Edge connection method, the edge connectivity is improved, and the thickness of the edge is reduced by the MTF of the reading device.
Has the effect of realizing high independence on

According to a fourth aspect of the present invention, in the first or second aspect, a binarization threshold is set as a parameter, and in generating output image data, a distribution error is calculated. An edge processing method that binarizes the sum of image data of a pixel and a distribution error with a parameter to generate output image data, calculates a quantization error at the time of binarization of a target pixel, and stores the quantization error. Has an effect that the thickness of the edge can be highly independent of the MTF of the reading device.

Hereinafter, an embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG. In these drawings, the same members are denoted by the same reference numerals, and duplicate description is omitted.

(Embodiment 1) FIG. 1 is a flowchart showing an edge processing method according to Embodiment 1 of the present invention.
FIG. 3 is a block diagram showing a configuration of an edge detection unit in which the edge processing method of FIG. 1 is executed, FIG. 3 is an explanatory diagram showing a second derivative operator in the edge processing method of FIG. 1, and FIG. FIG. 5 is a graph showing an edge detection result.
5 is a block diagram showing a configuration of an error diffusion unit in which the edge processing method of FIG. 1 is executed, and FIG. 6 is an explanatory diagram showing an error matrix in the error diffusion unit of FIG.

First, a processing flow of one pixel by the edge processing method of the present embodiment will be described with reference to FIG.

In this embodiment, image data D, which is a density value, is input (image input step S100), and up, down,
The maximum value Em of the difference values in a total of four directions: left, right, right and left slopes
ax is calculated to obtain a first derivative of the pixel of interest from the image data D (first derivative calculation step S101);
The secondary differential value R of the target pixel is calculated (secondary differential value calculation step S102), and an edge determination result is calculated by comparing Emax with the threshold value TH_E and determining whether the secondary differential value R is positive or negative (edge determination step S103), the image data correction value D according to the edge determination result of step S103
corr is set and output as a parameter (parameter setting step S104), and the image data correction value Dcor is set.
An output image is generated by performing binarization processing by pseudo gradation processing based on error diffusion processing using r as a parameter (output image generation step S105).

Here, step S101 is a subroutine for calculating a difference value in a total of four directions of a difference value Ea in the vertical direction, a difference value Eb in the left and right direction, a difference value Ec in the right oblique direction, and a difference value Ed in the left oblique direction. Step (SS1) and difference values Ea, Eb, E
c and Ed are converted to absolute values, and | Ea |, | Eb |
substep (SS2) for obtaining c |, | Ed |
a |, | Eb |, | Ec |, | Ed |
ax calculation sub-step (SS3).

In step S103, the judgment formula Ema
When x <TH, the edge determination result is 1 (the pixel of interest is not the edge of the image), and when the determination formula (Emax> = TH) and (R <0), the edge determination result is 2 (the pixel of interest is the image edge). The edge determination result is calculated as 3 (the target pixel is on the peak side of the image edge) when the valley side) and the determination formula (Emax> = TH) and (R> = 0) are satisfied.

Step S104 sets the image data correction value Dcorr as a parameter based on the edge determination result, and sets a parameter when the edge determination result is 1 (SS6). Sub-step of setting parameters (SS8);
And a sub-step (SS7) for performing parameter setting when the edge determination result is 3.

Then, step S105 is a substep (SS9) of calculating a distribution error Etot, which is a convolution of the element value of the error matrix and the quantization error around the pixel of interest, and the image data and image data of the pixel of interest. The sum of the parameter that is the correction value Dcoor and the distribution error Etot is binarized with a binarization threshold to obtain output image data that is the binary image Q of the pixel of interest, and the quantization error E at the time of binarization of the image of interest. (SS10)
And a sub-step of storing the quantization error E (SS11)
It is composed of

The edge processing of the entire image is performed by repeating the processing shown in FIG. 1 in the main scanning direction and the sub-scanning direction.

Next, the configuration of the edge detector in the edge processing method of the present embodiment will be described with reference to FIG.

The edge detecting section has a first derivative calculating means 10 to which image data D is input, a second derivative calculating means 11, and an image storing means 12.

The primary differential calculating means 10 comprises a differential value calculating means 14, an absolute value calculating means 15, and a maximum value calculating means 16, and the image data D is converted to the primary differential calculating means 10.
Is input to the difference value calculating means 14. Also, the control signal D
A control unit 13 for sending L_CTR to the image storage unit 12 is provided.
The image data D is written in accordance with _CTR, and the line delay signal DL of the image data D is read out to the primary differential calculation means 10 and the secondary differential calculation means 11 in accordance with the control signal DL_CTR.

Here, the difference value calculating means 14 forms a window (not shown) with the image data D and the line delay signal DL read from the image storing means 12,
With respect to this window, FIG.
(B), (c), (d) first derivative operator 17,1
By performing 8, 19, and 20 operator calculations, a vertical difference value Ea, a horizontal difference value Eb, a right diagonal difference value Ec, and a left diagonal difference value Ed are calculated, respectively.
The absolute value calculating means 15 calculates the difference values Ea, Eb, Ec,
| Ea |, | Eb |, | Ec |, |
Ed |, and the maximum value calculating means 16 calculates | Ea |, | E
The maximum value Emax of b |, | Ec |, | Ed | is calculated. Then, the second derivative calculator 11 forms a window (not shown) with the image data D and the line delay signal DL read from the image storage 12, and performs the second derivative operator 1 of FIG. Thus, the second derivative R is calculated.

Next, the result of the edge judgment will be described with reference to FIG. FIG. 4 is a graph showing a typical level change in the main scanning direction (X) with respect to the image data D (density data) near the image edge.

In FIG. 4, the edge judgment result = 1 (when the judgment formula Emax <TH) corresponds to the portion shown in FIG. 4, and means that the target pixel is not the edge of the image as described above. Edge judgment result = 2 (judgment expression (Emax
> = TH) and (R <0)) correspond to the portion of FIG. 4 and mean that the pixel of interest is on the valley side of the image edge, that is, on the portion with low density, as described above. And
The edge judgment result = 3 (judgment expression (Emax> = TH) and (R> = 0)) corresponds to the portion shown in FIG. 4. As described above, the target pixel is on the hill side of the image edge, that is, the density is high. Means in part.

Next, an error diffusion unit in which the edge processing method of the present embodiment is executed will be described with reference to FIG.

The error diffusion unit outputs an edge discrimination result based on the maximum value Emax and the secondary differential value R. The error diffusion unit receives the edge discrimination result from the edge discrimination unit 2 and receives the image data correction value as a parameter. Dcorr
(Calculated values in sub-steps SS6, SS7, SS8 in FIG. 1), a parameter setting means 3, an image data D, a distribution error Etot around the target pixel from the error control means 7, and an image from the parameter setting means 3. An adding means 8 for outputting an added value to the data correction value Dcorr, and an added value of the adding means 8 are input, and the added value is set to a fixed threshold TH (=
128) When larger than binary image Q = 1, fixed threshold T
When H or less, the binary image Q is binarized to 0 and the binary image Q
And an error generating means 5 which receives the binary image Q and the fixed threshold value TH of the binarizing means and outputs D−Q × 255 as a quantization error E at the time of binarization. Storage means 6 for storing the quantization error E, and a control signal EL.
_CTR controls the writing of the quantization error E to the error storage means 6 and EL which is the line delay data of the quantization error E
And an error control means 7 for calculating a distribution error Etot around the target pixel from the EL based on the error matrix 9 (FIG. 6).

Here, when the edge determination result = 1, for example, Dcorr = 0 (default value) when the edge determination result = 1, and when the edge determination result = 2, for example, Dcorr = -Emax / 4 to emphasize the pixel.
(The value at which the output image data becomes lighter)
In the case of 3, for example, Dcorr
= Emax / 4 (a value at which the output image data becomes darker) is used as a parameter to set the image data correction value Dcorr.

As described above, by adding the image data correction value Dcorr, which is a parameter, to the adding means 8, the image data is corrected to be lower on the valley side of the image edge, and the image data is corrected on the hill side of the image edge. Is corrected to the higher side, and no correction value can be added in other places, so that the edge of the image can be emphasized.

(Embodiment 2) FIG. 7 is a flowchart showing an edge processing method according to Embodiment 2 of the present invention.
FIG. 8 is a block diagram showing a configuration of an error diffusion unit in which the edge processing method of FIG. 7 is executed. Note that, in the present embodiment, the configuration of the edge detection unit and the edge determination result are the same as those in the first embodiment.
Overlapping illustrations and descriptions have been omitted.

First, a processing flow for one pixel by the edge processing method of the present embodiment will be described with reference to FIG.

In this embodiment, image data D, which is a density value, is input (image input step S100),
The maximum value Em of the difference values in a total of four directions: left, right, right and left slopes
ax is calculated to obtain a first derivative of the pixel of interest from the image data D (first derivative calculation step S101);
The secondary differential value R of the target pixel is calculated (secondary differential value calculation step S102), and an edge determination result is calculated by comparing Emax with the threshold value TH_E and determining whether the secondary differential value R is positive or negative (edge determination step S103), and sets and outputs the threshold value TH as a parameter according to the edge determination result of step S103 (parameter setting step S1)
06), an output image is generated by performing a binarization process by a pseudo gradation process based on an error diffusion process using the threshold value TH as a parameter (output image generation step S107).

Here, step S101 is a step of calculating a difference value in a total of four directions of a difference value Ea in the vertical direction, a difference value Eb in the left and right direction, a difference value Ec in the right oblique direction, and a difference value Ed in the left oblique direction. Step (SS1) and difference values Ea, Eb, E
c and Ed are converted to absolute values, and | Ea |, | Eb |
substep (SS2) for obtaining c |, | Ed |
a |, | Eb |, | Ec |, | Ed |
ax calculation sub-step (SS3).

In step S103, the judgment expression Ema
When x <TH, the edge determination result is 1 (the pixel of interest is not the edge of the image), and when the determination formula (Emax> = TH) and (R <0), the edge determination result is 2 (the pixel of interest is the image edge). The edge determination result is calculated as 3 (the target pixel is on the peak side of the image edge) when the valley side) and the determination formula (Emax> = TH) and (R> = 0) are satisfied.

In step S106, the binarization threshold value TH is set as a parameter based on the edge determination result, and a sub-step (SS12) for setting parameters when the edge determination result is 1; It is composed of a sub-step for setting parameters (SS14) and a sub-step for setting parameters when the edge determination result is 3 (SS13).

Step S107 is a sub-step (SS9) of calculating a distribution error Etot, which is a convolution of the element value of the error matrix and the quantization error around the pixel of interest. Eto
The sub-step (SS1) of binarizing the sum with t with a parameter to obtain output image data that is a binary image Q of the pixel of interest and calculating a quantization error E at the time of binarizing the image of interest.
5) and a sub-step of storing the quantization error E (SS1)
1).

The edge processing of the entire image is performed by repeating the processing shown in FIG. 7 in the main scanning direction and the sub-scanning direction.

Next, an error diffusion unit in which the edge processing method of this embodiment is executed will be described with reference to FIG.

The error diffusion unit outputs an edge discrimination result based on the maximum value Emax and the secondary differential value R. The error diffusion unit receives the edge discrimination result from the edge discrimination unit 2 and receives a threshold value TH (see FIG. 7, a parameter setting means 3 for setting the values calculated in the sub-steps SS12, SS13, SS14, and an addition means 8 for outputting an added value of the image data D and the distribution error Etot around the target pixel from the error control means 7. And the addition value of the addition means 8 and the threshold value TH from the parameter setting means 3, and when the addition value is larger than the threshold value TH, the binary image Q = 1 and the threshold value T
When the value is equal to or less than H, a binarizing unit 4 for binarizing the binary image Q = 0 and the binary image Q and the threshold value TH of the binarizing unit are input, and D−Q × 255 at the time of binarization is input. An error generating unit 5 that outputs the quantization error E; an error storage unit 6 that stores the quantization error E; a control of writing the quantization error E to the error storage unit 6 by the control signal EL_CTR; There is provided an error control means 7 for controlling reading of EL which is delay data and calculating a distribution error Etot around the target pixel from the EL based on the error matrix 9 (FIG. 6).

Here, the parameter setting means 3 sets, for example, TH = 128 (default value) when the edge judgment result = 1, and sets TH = 128 + Emax / 4 (output) to emphasize the pixel when the edge judgment result = 2. When the edge discrimination result = 3, for example, TH = 128−
The threshold value TH is set using Emax / 4 (a value at which the output image data becomes dark) as a parameter.

By making the threshold value TH, which is a parameter, variable, the image data is corrected to the lower side on the valley side of the image edge and to the higher side on the ridge side of the image edge. Since no correction value can be applied to other portions, the edges of the image can be emphasized.

Although the image data D is a density value in the first and second embodiments, it may be a luminance signal.

In the above description, the pseudo halftone processing is performed based on the error diffusion processing. However, the pseudo halftone processing can be performed based on the average error minimum method.

[0058]

As described above, according to the present invention, the peak and valley sides of the edge are defined by the primary differential value and the secondary differential value with respect to the edge that has been blurred by the MTF of the reading apparatus, and the edge ridge side is defined. Since a certain pixel level is amplified and a pixel level on the valley side of the edge is amplified, the connectivity of the edge is improved, and the edge thickness is highly independent of the MTF of the reading device. The effective effect that it becomes possible is obtained.

Further, there is obtained an effective effect that it is possible to avoid a phenomenon that an edge, which tends to be a problem in edge reproduction, is reproduced thick.

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating an edge processing method according to a first embodiment of the present invention;

FIG. 2 is a block diagram illustrating a configuration of an edge detection unit that executes the edge processing method of FIG. 1;

FIG. 3 is an explanatory diagram showing a second derivative operator in the edge processing method of FIG. 1;

FIG. 4 is a graph showing a result of edge detection by the edge processing method of FIG. 1;

FIG. 5 is a block diagram showing a configuration of an error diffusion unit in which the edge processing method of FIG. 1 is executed.

FIG. 6 is an explanatory diagram showing an error matrix in the error diffusion unit in FIG. 5;

FIG. 7 is a flowchart showing an edge processing method according to the second embodiment of the present invention;

8 is a block diagram illustrating a configuration of an error diffusion unit in which the edge processing method of FIG. 7 is executed.

FIG. 9 is a flowchart showing a conventional edge processing method.

FIG. 10 is a block diagram showing a configuration of an edge detection unit in a conventional edge processing technique.

FIG. 11 is an explanatory diagram showing a primary differential operator in a conventional edge processing technique.

FIG. 12 is a block diagram showing a configuration of an edge correction unit in a conventional edge processing technique.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Secondary differential operator 2 Edge discriminating means 3 Parameter setting means 4 Binarizing means 5 Error generating means 6 Error storing means 7 Error controlling means 8 Adding means 10 Primary differential calculating means 11 Secondary differential calculating means 12 Image storing means 14 Difference Value calculation means 15 Absolute value calculation means 16 Maximum value calculation means

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B057 AA11 CA08 CA12 CA16 CB07 CB12 CB16 CC03 CE03 CE13 DA08 DB02 DB09 DC16 5C077 LL19 MP06 MP07 NN11 PP03 PP43 PP47 PP68 PQ08 PQ20 PQ22 RR02 RR14 TT06 5L096 AA06 GA031

Claims (4)

[Claims] An edge processing method for detecting and correcting an edge portion of an image by inputting multi-gradation image data, wherein the image data is input, and a first derivative value of a pixel of interest is obtained from the image data. Calculating a second differential value of a pixel of interest from the image data; a first edge determination result in which the first differential value does not exceed a threshold value; the first differential value is equal to or more than a threshold value; and the second differential value does not exceed zero. As a result of the second edge determination, three edge determination results of a third edge determination result in which the primary differential value is equal to or greater than a threshold value and the secondary differential value is equal to or greater than zero are set, and parameters are set from the three edge determination results. And generating output image data based on the parameter. In the generation of the output image data, when the pixel of interest is the first edge determination result, the parameter is set to a default value. When the target pixel is the second edge determination result, the parameter is controlled in a direction in which the output image data becomes thinner than the default value. When the target pixel is the third edge determination result, An edge processing method, wherein the parameter is controlled so that the output image data becomes darker than a default value. 2. The edge processing method according to claim 1, wherein said output image data is generated by binarizing an image by pseudo halftone processing. 3. An image data correction value is set in the parameter. In the generation of the output image data, a distribution error is calculated, and a sum of the image data of the pixel of interest, the parameter, and the distribution error is calculated. 3. The edge processing method according to claim 1, wherein the output image data is generated by converting the value into a value, and a quantization error at the time of binarization of the target pixel is calculated and stored. 4. A binarization threshold value is set in the parameter. In the generation of the output image data, a distribution error is calculated, and a sum of the image data of the pixel of interest and the distribution error is calculated by the parameter. The edge processing method according to claim 1, wherein the output image data is generated by performing a value conversion, and a quantization error at the time of binarization of the target pixel is calculated and stored.