JPH11191845A - Error diffusing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent the generation of jaggy at the edge of a character by performing error diffusion processing while changing the degree of error influence corresponding to the sharpness degree of an image. SOLUTION: Maximum and minimum values are found from the density data of pixels around an object pixel to be binarized, and the sharpness degree of the image is found. The degree of error influence corresponding to that degree is calculated. Bias density data V of the object pixel are calculated by density data A of the object pixel, error data and a calculated degree αof error influence. If the V is <=255, the binarized data of the object pixel are made into '0' and a value to be stored in an error buffer is made into V. When the V is not <=255, the binarized data of the object pixel are made into '1', and the value to be stored in the error buffer is made into (V-255). When the degree α of error influence is equal with '0', the density data V become (A+V) and the error diffusion is performed to the binarized data. In the case of α=1, the density data V become (2A) and the error diffusion is not performed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error diffusion method for binarizing image data, and more particularly, to changing the degree of the influence of an error depending on the degree of sharpness of an image, thereby enabling a character in an image to be displayed. Eliminates jaggies that occur at the edges

[0002]

2. Description of the Related Art Methods for binarizing image data include a simple slice, a floating slice, and a dither method. The simple slice and the floating slice have a good reproduction image of the text image, but have a poor reproduction image of the photographic image. On the other hand, in the dither method, a reproduced image of a photographic image is good, but a reproduced image of a text image is not good. Then, a moiré phenomenon also occurs.

For this reason, a technique such as image area separation is used, but it is difficult to separate a photographic image from a text image, and especially "characters in a photograph" cannot be reproduced well.

As a means for solving this, "error diffusion method"
It has been known. Unlike the dither method, the error diffusion method does not perform a fixed slice with respect to the space, but determines a slice in the area from the density distribution. That is, the weight of the target pixel to be binarized is set to １ ／, and the remaining 1
The binarization process is performed by referring to the surrounding pixels at / 2.

[0005] For this reason, the error diffusion method improves the reproduced image of a text image and a photographic image having a halftone.

FIG. 4 shows a flowchart of a conventional error diffusion process. FIG. 5 is an explanatory diagram of a target pixel to be binarized, surrounding pixels, and coefficients applied to error data.

Hereinafter, the operation will be described with reference to the flow chart of FIG. Although this flow shows the error diffusion processing for one pixel of the image image, the processing of the entire image image is performed only by sequentially changing the target pixel.

Step S401: The density data A of the target pixel, the error data of the peripheral pixel with respect to the target pixel, and the error data addition value B obtained by multiplying the error data by a coefficient provided corresponding to the position of each peripheral pixel are added. To calculate bias density data V.

That is, the bias density data V is obtained as follows. Bias density data V = A + B

At this time, the error data of the surrounding pixels with respect to the target pixel and their positions are represented by Ea1, Ea2, Ea in FIG.
a3 and Eb1. Then, the coefficients provided corresponding to the positions of the error data of the surrounding pixels corresponding to the target pixel are represented by (1/8), (1/4), (1/8), (1) in FIG.
/ 2), the error data addition value B is obtained as follows. Error data addition value B = Ea1 × (1/8) + Ea2 ×
(1/4) + Ea3 × (1/8) + Eb1 × (1 /
2)

Step S402: bias density data V
Is 255 or less. If less than 255, step S
Proceed to 403; if not less than 255, step S405
Proceed to.

Step S403: The binary data of the target pixel is set to 0 (black).

Step S404: The value stored in the error buffer is set to V. Then, the process ends.

Step S405: The binary data of the target pixel is set to 1 (white).

Step S406: The value stored in the error buffer is set to (V-255). Then, the process ends.

[0016]

However, in the prior art shown in FIG. 4, since the weight of the target pixel to be binarized is fixed to い か な る in any case, an intermediate pixel such as a character edge is used. In the case of output, the result of binarization alternates between "white" or "black" due to the surroundings, resulting in jaggies.

FIG. 6 is an enlarged view of a character edge of image data binarized by a conventional error diffusion method. This figure is an enlarged view of a part of the character “/”, and the part surrounded by a circle is a jaggy.

As described above, even in the error diffusion method, there is a problem that a text image is jagged around a character and the image quality is not as high as that of a simple slice.

[0019]

SUMMARY OF THE INVENTION The present invention has been made in consideration of the above-described problems, and performs error diffusion processing by changing the degree of influence of an error depending on the degree of sharpness of an image. This makes it possible to eliminate jaggies that occur at the edges of characters in the image.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In an error diffusion method for binarizing image image data, the degree of sharpness of an image is determined according to the degree of sharpness of the image.
The error diffusion process is performed by changing the degree of the influence of the error. This makes it possible to eliminate jaggies that occur at the edges of characters in the image.

The sharpness of the image is set by a value obtained by subtracting the minimum value from the maximum value of the surrounding pixels. This makes it possible to calculate the degree of sharpness of the image simply and quickly.

The degree of the influence of the error is changed linearly or non-linearly according to the degree of sharpness of the image. Thereby, the image image can be adjusted to a text image style and a photographic image style. In addition, it is possible to eliminate jaggies that occur at edges of characters in the image.

[0023]

FIG. 1 is a flowchart showing an error diffusion process according to an embodiment of the present invention. FIG. 2 is an explanatory diagram of a target pixel and surrounding pixels. Hereinafter, the operation will be described according to this flow.

Step S101: The maximum value and the minimum value are obtained from the density data of the surrounding pixels for the target pixel to be binarized, and the sharpness of the image is obtained by (maximum value-minimum value).

The surrounding pixels for the target pixel are, as shown in FIG. 2, pixels Ma1 to Ma around the target pixel.
a3, Mb1, Mb3, Mc1 to Mc3.

Step S102: The degree of influence of an error corresponding to the degree of sharpness of the image is calculated from the correspondence table.

FIG. 3 shows an embodiment of a correspondence table of the degree of sharpness of an image / the degree of influence of an error. In this example, the degree of sharpness of the image, that is, the degree α of the influence of an error that changes linearly in accordance with the density data 0 to 255 of the image is set. For example, if the density data of the image is n, the degree α of the influence of the error is n / 255.

Step S103: The bias density data V of the target pixel is calculated from the density data A of the target pixel and the error data thereof and the degree of the influence α of the error calculated in step S102. FIG. 5 shows the error data for the target pixel.
The bias density data V is obtained as follows when the error data addition value is B. Bias density data V = (1 + α) A + (1-α) B

Step S104: bias density data V
Is 255 or less. If less than 255, step S
Proceed to 105, if not less than 255, step S107
Proceed to.

Step S105: The binary data of the target pixel is set to 0 (black).

Step S106: The value stored in the error buffer is set to V. Then, the process ends.

Step S107: The binary data of the target pixel is set to 1 (white).

Step S108: The value stored in the error buffer is set to (V-255). Then, the process ends.

In this processing, the bias density data obtained in step S103 is a decision value for binarization, and when the degree α of the influence of the error is α = 0, the density data V is (A +
B), the binarized image data is subjected to error diffusion, and an image suitable for a photographic image is obtained. If α = 1, the density data V is (2A), and the binarized image data is not subjected to error diffusion, and an image suitable for a text image is obtained.

[0035]

The present invention is embodied in the form described above, and has the following effects.

Even if the image data including the text image is binarized by the error diffusion method, it is possible to obtain image data in which jaggies do not occur at the edges of the characters.

[Brief description of the drawings]

FIG. 1 is a process flowchart of an embodiment of an error diffusion process of the present invention.

FIG. 2 is an explanatory diagram of a target pixel and surrounding pixels.

FIG. 3 is a diagram illustrating an example of a correspondence table of a degree of sharpness of an image / a degree of influence of an error.

FIG. 4 is a processing flowchart of a conventional error diffusion process.

FIG. 5 is an explanatory diagram of a target pixel, surrounding pixels, and coefficients applied to error data.

FIG. 6 is an enlarged view of a character edge of image data binarized by a conventional error diffusion method.

[Explanation of symbols]

 1 Error diffusion processing

Claims (3)

[Claims] 1. An error diffusion method for binarizing image image data, wherein error diffusion processing is performed by changing the degree of influence of an error depending on the degree of sharpness of an image. 2. The error diffusion method according to claim 1, wherein the degree of sharpness of the image is set by a value obtained by subtracting a minimum value from a maximum value of surrounding pixels. 3. The error diffusion method according to claim 1, wherein the degree of the influence of the error is changed linearly or non-linearly according to the degree of sharpness of the image.