JP2001298617A - Threshold matrix and method and device for gradation reproduction using it - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a threshold matrix, which can obtain a visually preferable output image even in the case that a very high image quality is required by a simple processing method of an organized dither method, and a method and device for gradation reproduction using it. SOLUTION: The threshold matrix is generated by performing processing which giving a repulsion potential to each dot of a dot pattern of each gradation to determine the dot position for determination of a dot pattern representing an adjacent gradations and when this threshold matrix is used to perform gradation conversion processing of an input image having a uniform density corresponding to the size of the threshold, it has an isotropic and acyclic dot pattern of less low frequency components and has an anisotropic and cyclic or pseudo- periodic dot pattern at least in the vicinity of the lowest gradation or the highest gradation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a threshold matrix,
The present invention relates to a tone reproduction method and apparatus using the same, and more particularly to a threshold matrix for performing gradation processing of input image data into binary or multi-value data, and a tone reproduction method and an apparatus using the same. It is.

[0002]

2. Description of the Related Art One of the most commonly used gradation reproduction methods is an error diffusion method. The error diffusion method converts an input image into a binary or multi-valued image by comparing the input image with a threshold for each pixel,
An image density is preserved by weighting and diffusing an error (quantization error) generated between an output value and an input value to a predetermined group of neighboring pixels. Although the output image obtained by this method has good image quality and good resolution, the processing speed is slow because the processing is complicated. Also, in general, the dot pattern generated by the error diffusion method differs in the position at which dots are formed depending on the input image, so it is difficult to predict the degree of color mixing, and the color reproducibility is poor compared to the systematic dither method described later. There is also a disadvantage.

As a tone reproduction method which is simple in processing and has a high speed, an independent decision type dither method is known. In the independent decision dither method, the output value is determined by comparing the pixel value of the input image with the threshold point by point. In this method, only the pixel of interest is processed independently, and processing for surrounding pixels is not performed, so that the processing speed is high.

[0004] This method is classified into two types, a random dither method and an organized dither method, depending on the difference in the way of giving a threshold.

[0005] The random dither method is a method of randomly changing a threshold value for each pixel. The dot pattern generated by this method has the advantage of having a white noise characteristic and no moiré, but the image quality is not good because of the remarkable graininess, and it is hardly used at present.

On the other hand, the systematic dither method is a method using a threshold matrix (also called a dither matrix, a mask, etc.) in which thresholds are arranged. Depending on the arrangement of the thresholds in the threshold matrix, the system is largely divided into a dot concentration type and a dot dispersion type. Divided.

[0007] In the dot concentration type, as the number of gradations increases,
The dots increase closely at a position corresponding to the center of the threshold value matrix. Since the spatial frequency of the dot pattern generated by this method is low, an output device having relatively low definition has poor image quality and is used in a printing field with high definition.

In the dot dispersion type, a threshold matrix is designed so that the dot arrangement of an output pattern is spatially dispersed, and a Bayer-type systematic dither method is conventionally known as a typical example ( References: An Optical m
ethod for two-level rendition of continuous-tone p
ictures, Bayer, Proc. IEEE Int. Conf. Commun., Con
ference Rec. p.26-11,1973). Because the Bayer-type systematic dither method has a very regular threshold array, processing an input image with a uniform gray level to halftone processing produces an extremely regular output pattern for all grayscale input images. Is done. Therefore, although the dot pattern uniformity is good, if the definition of the output device is low, the size of the threshold matrix (25
There is a problem that an unsightly texture (dither pattern) is perceived at a period of 16 × 16 for 6 gradations, and moire may occur in the output image if the input image contains a periodic pattern. . Therefore, an output device with a relatively low definition of several hundred dpi has inferior image quality as compared with the error diffusion method, and is not used for obtaining a high-quality halftone image.

On the other hand, in recent years, it has been known that good image quality can be obtained when the dot pattern of an output image is a blue noise pattern (RLUlichney, Di.
thering with Blue Noise, Proc.IEEE, vol.76, No.1,
p.56). The blue noise pattern is characterized in that it has a power spectrum (blue noise characteristic) which is aperiodic, isotropic, and has a low-frequency noise component. Ulichney devised a perturbed error diffusion method that introduced irregularities into the conventional error diffusion method, and realized a blue noise pattern.

A method proposed to realize this blue noise pattern by using a systematic dither method is a blue noise mask method. (Patent Gazette No. 2622429, USP5,
111,310, T. Mitsa and KJParker, Digital halftonin
g technique using a blue-noise mask, J.Opt.Soc.Am,
vol.9, No.11, pp.1920-1929 (1992)) When an input image with a uniform gray level is processed using the blue noise mask method, the output dot pattern has blue noise characteristics. . Therefore, moire does not occur as in the case of Ulichney's perturbation error diffusion method, and the granularity is smaller than that of the random dither method. Also, the operation is simpler and the processing speed is faster than the perturbation error diffusion method.

However, the dot pattern formed by the blue noise mask method has poor uniformity of dot distribution depending on the number of gradations, and unevenness appears on the output screen. In particular, the dot patterns near the lowest gradation and the highest gradation have a large sense of noise.

[0012] This is because, in the blue noise mask method, a dot pattern of an intermediate gradation is first created by optimizing, and then a dot pattern of a high gradation and a low gradation is successively created. This is because the degree of freedom in dot selection is small near the highest gradation, and it is difficult to obtain an ideal dot pattern. In particular, when used in an ink jet printer, the unevenness of the dot pattern in the low gradation portion where the dot arrangement is sparse and visually conspicuous needs to be improved.

Further, there is another problem that the method of repeatedly using the two-dimensional discrete Fourier transform and the two-dimensional inverse discrete Fourier transform is complicated and time-consuming.

[0014]

As a systematic dither method with improved dot pattern uniformity, there is a method disclosed in JP-A-2000-59626. This method maintains the uniformity of the dot pattern by using the regularity of the dot pattern generated by the Bayer-type systematic dither method, and at the same time, introduces the irregularity into the dot pattern to achieve the dither pattern and the moiré pattern. The problem has been alleviated.

Since the threshold matrix of this method has a plurality of identical threshold structures therein, it is anisotropic for all gradations,
Generates a periodic dot pattern and has a characteristic opposite to the blue noise characteristic.

In addition, by eliminating redundancy of information in portions having the same threshold structure, the memory capacity for storing the threshold matrix can be reduced.

On the other hand, in recent years, with the spread of ink jet printers, momentum has been generated not only for general users but also for business purposes, that is, for creating forms such as banks and advertisement pamphlets. In the case where such a very high image quality is required, the above-mentioned method uses a texture (dither pattern) in which a plurality of identical structures are arranged in a vertical or horizontal direction or an oblique direction on an output screen, though not so noticeable depending on the gradation. ) May be seen slightly, and improvement was required.

The required halftone processing method is a simple processing method of the systematic dither method, in which no unsightly pattern or moiré occurs in all the grayscales, and the uniformity of the dot distribution. Is a method that does not cause unevenness in the output image.

In order to prevent the occurrence of moire on the output screen, a random dither method, a blue noise mask method, etc.
It is effective to make the dot pattern isotropic and aperiodic.

In order to prevent the output screen from becoming uneven, it is effective to adopt a periodicity in the dot pattern to make the dot pattern uniform, and to achieve the above object. Must take both properties well.

The threshold matrix creation method disclosed in Japanese Patent Application Laid-Open No. 2000-59626 is a method that incorporates both properties. However, since the threshold matrix has a plurality of identical threshold structures inside the threshold matrix, depending on the gradation, Texture may appear slightly, and there is room for improvement when very high image quality is required.

The present invention eliminates the above-mentioned disadvantages of the prior art, eliminates a plurality of identical threshold structures, and increases the memory capacity for storing the threshold matrix, but solves the texture problem and uses the threshold matrix. It is an object of the present invention to provide a gradation reproduction method and an apparatus therefor.

[0023]

According to the present invention, there is provided a gradation reproducing method for achieving the above object, wherein each pixel of an original image and each element of a threshold matrix correspond to each other on a one-to-one basis. In a tone reproduction method in which the density of a pixel is represented by a binary or multi-valued dot, a dot for defining a dot pattern representing an adjacent tone is obtained by applying a repulsive potential to each dot of the dot pattern representing a certain tone. When the threshold value matrix is created by performing the process of determining the position at each gradation, and using the threshold value matrix, when an input image having a uniform density corresponding to the size of the threshold value matrix is subjected to gradation conversion processing, Many gradations in the intermediate gradation part, with a dot pattern that is isotropic, non-periodic, and low in low frequency components, and at least one of near the lowest gradation or near the highest gradation , Non-isotropic, characterized by having a periodic or quasi-periodic dot pattern.

Further, the tone reproduction device of the present invention expresses the density of each pixel of the output image in binary or multi-level by making each pixel of the original image correspond to each element of the threshold matrix on a one-to-one basis. In the tone reproduction device, a means for storing a threshold matrix, a comparing means for comparing the density of each pixel of the original image with each pixel using the value of the threshold matrix as a threshold, and a comparing means based on a comparison result of the comparing means. Output means for outputting a value or a multi-valued dot pattern, wherein the threshold value matrix is a dot representing an adjacent gradation by applying a repulsive potential to each dot of the dot pattern representing a certain gradation. It is created by performing a process of determining a dot position for determining a pattern at each gradation, and corresponds to the size of the threshold matrix using the threshold matrix. That an input image of uniform density when gradation conversion processing, in many gradations of halftone portion, isotropic, has a small dot patterns aperiodic and low frequency components,
At least one of the vicinity of the lowest gradation and the vicinity of the highest gradation has an anisotropic, periodic or pseudo-periodic dot pattern.

A threshold matrix according to the present invention is a threshold matrix used when converting the density of each pixel of an original image into binary or multi-value data, wherein the threshold matrix is a dot pattern of a certain gradation. By applying a repulsive potential to each dot, a process of determining a dot position for defining a dot pattern representing an adjacent tone is performed at each tone, and is created by using the threshold matrix. When an input image of uniform density corresponding to the size is subjected to gradation conversion processing, it has an isotropic, non-periodic, low-frequency component dot pattern with many gradations in the intermediate gradation part,
At least one of the vicinity of the lowest gradation and the vicinity of the highest gradation has an anisotropic, periodic or pseudo-periodic dot pattern.

Further, the storage medium according to the present invention provides a gradation which expresses the density of each pixel of the output image in binary or multi-level by making each pixel of the original image correspond to each element of the threshold matrix on a one-to-one basis. A storage medium for storing a control program for controlling a reproduction process in a computer-readable manner, in which a dot pattern representing an adjacent gradation is determined by applying a repulsive potential to each dot of the dot pattern representing a certain gradation. A threshold matrix created by performing the process of determining the dot position at each gradation, and comparing the density of each pixel of the original image with each pixel using the value of the threshold matrix as a threshold, according to the comparison result ,
Including a module for controlling the output of a binary or multi-valued dot pattern, using the threshold matrix, when the input image of uniform density corresponding to the size of the threshold matrix is subjected to gradation conversion processing Has a dot pattern that is isotropic, non-periodic, and low in low frequency components at many gray levels in the intermediate gray level portion, and is anisotropic in at least one of the vicinity of the lowest gray level or the highest gray level; It has a periodic or pseudo-periodic dot pattern.

[0027]

DETAILED DESCRIPTION OF THE INVENTION In the simplest case, such as a conventional ink jet printer or a bubble jet (BJ) printer, the present invention applies ink droplets to paper for each pixel of an output image. It can be applied to express a halftone in a device that forms an image by determining whether to hit or not, and similarly, a halftone can be applied to a liquid crystal display device or the like that displays an image in bright or dark for each pixel. Is preferably used to express

More generally, in a device such as a laser beam printer, a facsimile or a printing machine, such as an ink jet printer for converting a continuous tone black-and-white or color image into a binary or multi-valued image with respect to the density, and outputting the converted image. It is used to preferably express the key.

FIG. 2 shows a basic system for processing an image in the present embodiment. In FIG. 1, reference numeral 10 denotes an image input device such as a scanner for scanning an input image 11. This apparatus is provided with a pre-processing unit 12 that digitizes the number of tones to, for example, 256 tones, performs γ correction, and performs various color conversions on an input image 11 having continuous tones.

Reference numeral 13 denotes a gradation processing device, which compares the number of gradations of each pixel of the input image with the value (threshold) of the corresponding mask, and outputs an output value accordingly. Includes a comparator 16 for determining.

Reference numeral 17 denotes an apparatus for outputting an output image 18 formed on the basis of an output value from the comparator 16 in a form such as display or printing.

<Outline of this Embodiment> A basic mask making method of this embodiment will be described below with reference to the accompanying drawings. FIG. 1 shows a flowchart of the creation. First, in step S1, a basic structure of a mask, that is, a mask size, a mask shape, and the like are determined. When the input image is larger than the mask size, the masks are two-dimensionally arranged and used, and the arrangement of the masks is also determined.

In the present embodiment, the dot pattern representing each tone is determined sequentially from the low tone, but the step S2 of artificially determining the dot pattern of the low tone and the dot pattern of the higher tone are determined. Is largely divided into step S3 and subsequent steps in which is determined using the repulsive potential.

By determining a dot pattern from low gradations in which the dot arrangement is sparse, a uniform and less uneven dot pattern can be created. Here, for the sake of explanation, the mask is used for 256 gradations, and steps are performed until the second gradation.
It will be decided in S2.

The dot pattern at the low gradation determined in step S2 is periodic or pseudo-periodic with randomness. FIG. 3 shows the dot pattern of the first gradation determined in step S2. However, only the upper left part of the dot pattern is shown for convenience of the drawing. Here, a periodic dot pattern generated by the Bayer-type systematic dither method is used for the first gradation.

FIG. 4 shows the dot pattern of the second gradation determined in step S2. However, only the upper left part of the dot pattern is shown for convenience of the drawing. It was created by adding a pseudo-periodic dot pattern as indicated by 30 to 36 to the dot pattern of the first gradation. The dots indicated by 30 to 36 were randomly selected from the frames indicated by 37 to 43 with a uniform probability distribution. The randomness can be controlled by changing the size of the frame and the probability distribution. By giving the period / pseudo-periodicity, an extremely uniform low gradation dot pattern can be obtained, and the unevenness at the low gradation, which has been a problem of the blue noise mask method, can be solved.

In general, a periodic dot pattern may cause an unsightly pattern or moiré, but does not pose a problem at a low image density and a very low gradation.

Next, step S3 will be described. Hereinafter, a method of determining the dot pattern of the third gradation will be described. First, a repulsive potential as shown in FIG. 5 is applied to a pixel where a dot has already been formed in the second gradation. The sum of the repulsive potentials applied to all the dots is calculated, and a new dot is formed on a pixel having the lowest sum in a group of pixels where no dot is formed. However, when the lowest point of the sum of the repulsive potentials is two or more pixels, one pixel is randomly selected from the two points. Next, a repulsive potential is applied centering on the pixel, the sum of the repulsive potentials is calculated again, and a new dot is formed on the pixel having the lowest sum.

The above process is repeated, and when the number of dots necessary to represent the third gradation is reached, a dot pattern of the third gradation is obtained.

The repulsive potential has the effect of striking the dots as far apart as possible, and the dot pattern in the halftone portion of this embodiment using this is aperiodic, isotropic and uniform. High quality and less unevenness. The shape of the repulsive potential greatly affects the uniformity of the dot pattern, and a specific shape will be described later.

When this repulsive potential is applied to a dot-dotted pixel, the portion that protrudes from the boundary of the mask is determined by taking into account the two-dimensional arrangement of the mask determined in step S1, and setting the periodic boundary condition. Process using
In other words, when masks are arranged as shown in FIG.
The portions where the repulsive potential applied to the portion protrudes from the boundary of the mask are applied as if the repulsive potential were applied to the pixels at 51 to 53 as 54 to 56. If this method is used, an unsightly dot arrangement cannot be formed near the boundary of the mask when the masks are arranged and used.

The same process is repeated for dot patterns of the third and subsequent gradations, and the dot patterns are sequentially determined. In other words, the process of hitting a dot at the pixel position where the sum of the repulsive potential is the lowest and applying the repulsive potential around that pixel is repeated, and when the number of dots is sufficient to express the gradation, It is stored as a gradation dot pattern.

If all the dot patterns up to the dot pattern of the highest gradation are determined, they are accumulated to output these dot arrangements when an input image of a uniform gray level is subjected to halftone processing. A simple mask can be formed.

That is, the dot pattern d (g; x, y) of the g-th gradation is defined by using (x, y) as coordinates representing the pixel position.

[0045]

[Outside 1] Then the mask m (x, y) is

[0046]

[Outside 2] (Step S4).

As described above, the basic procedure for creating a mask according to the present embodiment has been described. The determination of the basic structure of the mask in step S1, the determination of the dot pattern at low gradation in step S2, and the repulsive force in step S3 The tendency of the dot arrangement can be freely and widely controlled depending on the method of determining the shape of the potential, and it is possible to realize a dot pattern that matches the characteristics of the output device.

Hereinafter, various embodiments of the present invention will be described.

<First Embodiment> A procedure for creating one of the masks having the features of the present embodiment will be described with reference to the flowchart of FIG.

First, in step S1, the basic structure of the mask is determined. The mask in this embodiment is a square matrix of 256 × 256, and is for 256 gradations.

FIG. 7 shows a mask arranging method when the input image is larger than the size of the mask. In the figure, 256 × 256 pixels painted in gray are the size of one mask. When the output device is a printer, the arrow direction indicated by is the main scanning direction such as the direction of the ink ejection head.
The arrow direction indicated by is a sub-scanning direction such as paper feeding.

In this embodiment, up to the dot pattern of the second gradation is created in step S2. In the present embodiment, the first gradation as a result of executing step S2 in FIG. 1 is as shown in FIG. However, for convenience of the drawing, only a part of the dot pattern is shown. The positions of x pixels in the main scanning direction and y pixels in the sub-scanning direction from the upper left pixel of the mask
It is represented by coordinates (x, y). Upper left pixel is (0,0)
It is. The dot position of the first gradation is (16m, 16n) (where m = 0,
1, ..., 15, n = 0,1, ..., 15).

The method of determining the dot pattern of the second gradation in this embodiment will be described with reference to FIG. However, for convenience of the drawing, only a part of the dot pattern is shown. In the figure, a small block of 7 × 7 pixels indicated by a thick solid line has the upper left coordinate of (16m + 5, 16n + 5) (where m, n = 0, 1,..., 15)
It is represented by coordinates. One dot is randomly assigned to each of the small sections with a uniform probability distribution. In this manner, a dot array of a pseudo-period second gradation is completed.

In the present embodiment, the repulsive potential used to execute step S3 in FIG.

[0055]

[Outside 3] And Here, r _max is 10, r is the distance from the pixel to which the repulsive potential is applied,

[0056]

[Outside 4] It is.

As described above, this repulsive potential is applied to the pixels of all the dots hit in the second gradation, and one new dot is hit at the pixel position where the sum of the repulsive potentials is lowest. A repulsive potential is given to the newly hit dot, and the next dot is hit at a pixel position where the sum of the repulsive potentials becomes minimum. This process is repeated, and when the number of dots necessary to represent the third gradation is completed, the dot pattern is stored as a dot pattern of the third gradation.

The above process is simplified, and a repulsive potential is applied to all pixels where dots are formed in the second gradation, and the sum of the repulsive potentials is equal to the number of dots required to represent the third gradation. Pixels may be selected in ascending order and dots may be uniformly printed on the pixels to form a dot pattern of the third gradation.

When a repulsive potential is applied to a dot-dotted pixel, a portion which is outside the boundary of the mask is processed using the periodic boundary condition as described above.

In this manner, the dot patterns of the 255th gradation from the fourth gradation onward are sequentially determined, and the dot patterns are accumulated to complete the mask (step S).
Four).

FIG. 10 shows a dot pattern in the vicinity of the lowest gradation (fourth gradation) created by the present embodiment.
256 pixels). It can be seen that a very uniform and uniform dot pattern is obtained due to the periodicity and pseudo-periodicity introduced in the first and second gradations.

FIG. 11 shows a dot pattern of the intermediate gradation portion (64th gradation) created by the present embodiment (256 × 2
56 pixels). It can be seen that an isotropic and non-periodic dot pattern is obtained, which is uniform and has no unevenness (that is, low in low frequency components).

As described above, in this embodiment, no unsightly pattern or moiré occurs at all gradations.
In addition, dot distribution is uniform and a dot pattern without unevenness is obtained.

Although the shape of the repulsive potential has a great effect on the image quality, it has been found that good image quality can be selected even when the following repulsive potential is used.

[0065]

[Outside 5] However, r _max = 10.

Or

[0067]

[Outside 6] However, r _max = 10.

In addition to the artificial periodicity or pseudo-periodicity not only near the lowest gradation but also near the highest gradation, it is possible to improve the uniformity of the dot pattern. In that case, for example, the pixel on the right of each dot hit at the second gradation may be held without hitting a dot until dots are hit on all other pixels. .

In this way, the dot pattern of the 253rd gradation becomes a pseudo-periodic dot pattern in which the dot pattern of the second gradation is inverted in black and white (dot on / off is reversed) and shifted by one pixel.

<Second Embodiment> A procedure for creating one of the masks having the features of the present embodiment will be described with reference to the flowchart of FIG.

First, in step S1, the basic structure of the mask is determined. The mask in the present embodiment is a square matrix of 128 × 256, and is for 256 gradations.

FIG. 12 shows a mask arrangement method when the input image is larger than the size of the mask. In the figure, 128 × 256 pixels painted in gray are the size of one mask. When the output device is a printer, the arrow direction indicated by is the main scanning direction such as the direction of the ink ejection head.
The arrow direction indicated by is a sub-scanning direction such as paper feeding.

In the present embodiment, a dot pattern of the first gradation is created in step S2. The positions of x pixels in the main scanning direction and y pixels in the sub-scanning direction from the upper left pixel of the mask are represented by coordinates (x, y). The upper left pixel
(0,0).

In the present embodiment, the first gradation as a result of executing step S2 in FIG. 1 is as shown in FIG. However,
For convenience of the drawing, only a part of the dot pattern is shown. In the figure, a small block of 6 × 6 pixels indicated by a bold solid line has an upper left coordinate of (16m, 16n) (where m = 0, 1,..., 7, n
= 0,1, ..., 15). One dot is randomly assigned to each of the small sections with a uniform probability distribution. In this way, the pseudo-periodic first
The gradation dot array is completed.

The creation of the dot pattern for the second and subsequent gradations in the present embodiment at step S3 in FIG. 1 is performed in the same manner as in the first embodiment. That is, the repulsive potential P (r) described above is applied to all the pixels hit in the first gradation.
And a new dot is placed at the pixel position where the sum is the lowest. A repulsive potential is applied centering on the pixel position of the newly hit dot, the sum of repulsive potentials is recalculated, and a dot is hit at the pixel position with the lowest sum. This is sequentially repeated, and when the number of dots required to represent the second gradation is hit, the dot pattern is stored as a dot pattern of the second gradation.

When the repulsive potential is applied to a dot-dotted pixel, a portion that protrudes from the boundary of the mask uses the periodic boundary condition in consideration of the two-dimensional arrangement of the mask determined in step S1. Process. In other words, the portion of the repulsive potential applied to the pixel 60 in FIG. 14 that protrudes from the boundary of the mask is considered as if the repulsive potential was applied to the pixel at 61, and
Is given as follows. (However, in FIG. 14, for the sake of explanation,
Using this method (rmax is drawn to be larger than the actual value) does not allow an unsightly dot arrangement near the boundary of the mask when the masks are arranged and used.

In this manner, the dot patterns of the 255th gradation from the third gradation onward are sequentially determined, and the dot patterns are accumulated to complete the mask. (Step S
Four)

FIG. 15 shows a dot pattern of the fourth gradation created by this embodiment (128 × 256 pixels). It can be seen that even in this embodiment, a very uniform and uniform dot pattern is obtained.

FIG. 16 shows a dot pattern (128 × 2) of the intermediate gradation part (64th gradation) created by the present embodiment.
56 pixels). It can be seen that an isotropic and non-periodic dot pattern is obtained which is uniform and has no unevenness (that is, low in low frequency components).

The method of arranging masks that are not parallel to the main scanning direction as in the present embodiment has an advantage that mechanical stripe unevenness that occurs in the main scanning direction in an ink jet printer or the like can be reduced.

In the above description, the first step is performed in step S2.
The dot pattern for gradation was determined, but the first
A quasi-periodic dot pattern of low gradation gl, not gradation, may be determined. In that case, dot patterns of higher gradation and lower gradation than gl are sequentially determined in the same manner as in step S3. However, when determining a dot pattern with a gradation lower than gl, the dot pattern is created by erasing the dot at the pixel position where the sum of the repulsive potentials is the highest among the pixels on which dots have already been formed.

<Third Embodiment> A procedure for creating one of the masks having the features of the present embodiment will be described with reference to the flowchart of FIG.

First, in step S1, the basic structure of the mask is determined. The mask in this embodiment is a square matrix of 256 × 256, and is for 256 gradations.

The arrangement method of the mask when the input image is larger than the size of the mask is the same as in the first embodiment.
This is as shown in FIG.

In this embodiment, in order to further enhance the uniformity of the dot pattern, the following restrictions (1) and (2) on the number of dots are performed. (1) When dividing into 16 x 16 pixel sections, in all dot patterns in the process of generating a mask, the number of dots in that section group must not differ by more than 2 in all sections . (2) Also, when divided into 8 × 8 pixel subsection groups, the number of dots in the subsection group in all the dot patterns in the process of generating the mask is
Make no difference between two or more.

The above regulations (1) and (2) will be described with reference to FIG. However, some parts are omitted for convenience of the drawing. In the figure, a section of 16 × 16 pixels divided by a thick solid line is a section group used under the regulation (1). Further, a section of 8 × 8 pixels further divided by a thick broken line is a small section group used under the regulation of the above (2).

In this embodiment, the first step is performed in step S2.
A gradation dot pattern will be created. In the present embodiment, the first gradation as a result of executing step S2 in FIG. 1 is as shown in FIG. However, for convenience of the drawing, only a part of the dot pattern is shown.

In the figure, a small section of 7 × 7 pixels indicated by a bold solid line has an upper left coordinate of (16m, 16n) (where m = 0, 1,
, 15, n = 0, 1, ..., 15). One dot is selected at random from each of the small sections with a uniform probability distribution. Thus, the pseudo-periodic first
A dot array of one gradation is completed. This dot pattern satisfies the above-mentioned dot number restrictions (1) and (2).

In the present embodiment, unlike the first and second embodiments, the repulsive potential used to execute step S3 in FIG. 1 includes the number of gradations g as a parameter. P (g, r) will be used when determining the dot pattern of g gradation. The specific form is as shown below.

[0090]

[Outside 7] Here, r _max is 128 and a is 0.46. Also, r indicates the distance from the pixel (x ₀ , y ₀ ) to which the repulsive potential is given,

[0091]

[Outside 8] It is. However, in order to reduce the calculation time, the repulsive potential at the 71st gradation is used as the repulsive potential at the 70th gradation. That is, P (g> 70, r) = P (g = 70, r) This repulsive potential P (g = 2, r) is applied to all the pixels of the dots hit in the first gradation, and Regulation (1)
A new dot is formed at the pixel position where the sum of the repulsive potentials is the lowest within the range satisfying (2). A repulsion potential P (g = 2, r) is given to this newly struck dot, and the next dot is placed at the pixel position where the sum of the repulsion potentials becomes minimum within the range satisfying the above-mentioned regulations (1) and (2). Hit.

This process is repeated, and when the number of dots required to represent the third gradation is completed, the dot pattern is stored as a dot pattern of the third gradation.

When the repulsive potential is applied to a dot-dotted pixel, a portion that is outside the boundary of the mask is processed using the periodic boundary condition as described above.

When determining the dot pattern of the third gradation, the repulsive potential P (g = 2, r) given to all the dots hit in the second gradation is temporarily canceled and newly set. A repulsive potential P (g = 3, r) for the third gradation is given. Thereafter, the dot pattern of the third gradation is determined by the same process as that for determining the dot pattern of the third gradation. However, the repulsion potential used at that time is P (g = 3, r)
It is.

In this way, from the fourth gradation onward to the 255th gradation
The gradation dot patterns are also determined sequentially, and the dot patterns are accumulated to complete the mask.

FIG. 19 shows a dot pattern of the fourth gradation created by the present embodiment (256 × 256 pixels). Because of the large randomness when introducing the pseudo-periodic dot pattern at the first gradation and the use of the repulsive potential with small attenuation, compared to the first and second embodiments. Although the periodic structure is difficult to understand, it can be seen that a very uniform and non-uniform dot pattern is obtained according to the present embodiment.

FIG. 20 shows a dot pattern of the intermediate gradation portion (64th gradation) created by the present embodiment (256 × 2
56 pixels). It can be seen that the present embodiment also provides an isotropic and non-periodic dot pattern that is uniform and has no unevenness (that is, low frequency components are small).

The dot pattern obtained by the present embodiment is different from the dot patterns obtained by the first and second embodiments in that the effects obtained by incorporating the rules (1) and (2) are as follows. In addition, due to the effect of using the repulsive potential with a slower decay, the uniformity of the dot pattern in a wide range is improved, which is particularly suitable for a high-resolution output device.

In the above-described embodiment, the case where input image data is converted into binary data has been described. However, the present invention is not limited to this. It can be applied when

The case of conversion to ternary data will be described.

Assuming that the output device is, for example, an ink jet printer, and the printer has two inks of different shades, the values that can be expressed are three values.

Assuming that the input data is data of 256 gradations of 8 bits per pixel, the value of the input data up to the 128th gradation is doubled, and the binary data is obtained by the mask created in the above embodiment. Become If the result is 1, light ink is output, and the input data is binarized by the mask created in the embodiment as it is until the 256th to 129th gradations, and the result is 1. In this case, dark ink is output. Further, as another method, up to the 128th gradation, the individual threshold value of the mask created in the above embodiment is set to 1/2.
A mask (cut off in the case of a small number) may be separately prepared and used as a mask for light ink. In this way, 12
Since the number of dots to be printed is twice as high for low gradations as the 8th gradation and lower than when only dark ink is output, the part where the input image changes slowly at low gradations can be reproduced smoothly. .

Therefore, in order to enhance the reproducibility of a gradation change in a human skin portion or the like, such a multi-value conversion technique is important. At that time, the mask created in the present embodiment is applied. As a result, an output image with higher image quality can be obtained.

When the present invention is applied to color image processing, binary or multi-level conversion is performed using a mask created in the above embodiment for each color (for example, Y / M / C / K). It should be processed.

The present invention can be applied to a system constituted by a plurality of devices such as a host computer, an interface device, a reader, and a printer.
Further, for example, the present invention can be applied to a single apparatus such as a copying machine and a facsimile machine.

Further, according to the present invention, a storage medium storing a program code of software for realizing the functions of the above-described embodiments is supplied to a system or an apparatus, and the computer (or CPU or MPU) of the system or apparatus stores the storage medium. The present invention is also applicable to reading and executing a program code stored in a medium.

In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention.

The storage medium also stores the threshold value matrix created in the above-described embodiment.

As the storage medium for supplying the program code, for example, a floppy disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, magnetic tape, nonvolatile memory card, ROM, etc. are used. be able to.

When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also the OS (Operating System) running on the computer based on the instruction of the program code. ) And the like perform part or all of the actual processing, and the processing may realize the functions of the above-described embodiments.

Further, after the program code read from the storage medium is written into the memory provided in the function expansion board inserted into the computer or the function expansion unit connected to the computer, the program code is read based on the instruction of the program code. It goes without saying that a CPU or the like provided in the function expansion board or function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

[0112]

As described above in detail, according to the present invention, an extremely high quality image is required by a simple processing method of the systematic dithering method without causing any unsightly pattern or moiré. In this case, it is possible to provide a threshold matrix capable of obtaining a visually preferable output image, a gradation reproduction method using the threshold matrix, and a device therefor.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a procedure for creating a mask according to the present embodiment.

FIG. 2 is a diagram illustrating a configuration example of a basic system for processing an image according to the present embodiment.

FIG. 3 is a diagram for explaining a method for creating a dot pattern of the first gradation according to the present embodiment.

FIG. 4 is a diagram for explaining a method of creating a second gradation dot pattern according to the present embodiment.

FIG. 5 is a diagram schematically showing a shape of a repulsive potential.

FIG. 6 is a diagram illustrating a periodic boundary condition of a repulsive potential.

FIG. 7 is a diagram showing a mask shape and a mask arrangement method in the first embodiment.

FIG. 8 is a diagram for explaining a method of creating a dot pattern of a first gradation in the first embodiment.

FIG. 9 is a diagram for explaining a method of creating a dot pattern of a second gradation in the first embodiment.

FIG. 10 shows 256 × 2 of the fourth gradation in the first embodiment.
It is a figure showing a dot pattern of 56 pixels.

FIG. 11 is a diagram showing 256 × 256 × gradation in the first embodiment;
It is a figure showing a dot pattern of 256 pixels.

FIG. 12 is a diagram showing a mask shape and a mask arrangement method according to the second embodiment.

FIG. 13 is a diagram for explaining a method of creating a dot pattern of the first gradation in the second embodiment.

FIG. 14 is a diagram illustrating a periodic boundary condition of a repulsive potential in the second embodiment.

FIG. 15 shows 128 × 2 of the fourth gradation in the second embodiment.
It is a figure showing a dot pattern of 56 pixels.

FIG. 16 shows 128 × 64 × gradation in the second embodiment.
It is a figure showing a dot pattern of 256 pixels.

FIG. 17 is a diagram for explaining the regulation (1) and (2) of the number of dots in the third embodiment.

FIG. 18 is a diagram for explaining a method of creating a dot pattern of a first gradation in the third embodiment.

FIG. 19 shows 256 × 2 of the fourth gradation in the third embodiment.
It is a figure showing a dot pattern of 56 pixels.

FIG. 20 is a diagram illustrating 256 × 256 × gradation in the third embodiment.
It is a figure showing a dot pattern of 256 pixels.

[Explanation of symbols]

S1, S2, S3, S4 Steps for creating a mask 10 Image input device 11 Input image 12 Preprocessing 14 Tone processing device 15 Memory 16 Comparator 17 Output device 18 Output image 30, 31, 32, 33, 34, 35, 36 Dots printed at the second gradation 37, 38, 39, 40, 41, 42, 43 Frames used to determine the dot position of the second gradation

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Takashi Suzuki 3-30-2 Shimomaruko, Ota-ku, Tokyo F-term in Canon Inc. (reference) 5B057 AA11 CA01 CA02 CA08 CB01 CB02 CB07 CE13 5C006 AA02 AA12 AF23 BB11 BF02 FA22 5C077 LL03 MP01 MP08 NN09 PP48 PQ12 PQ17 RR02 RR09 RR11 TT05 5C079 LA02 LA12 LC05 MA01 MA11 NA02 NA03 NA05 NA06 5C080 BB05 DD05 EE29 GG09 JJ01 JJ02 JJ05 JJ07

Claims (9)

[Claims] 1. A gradation reproducing method for expressing the density of each pixel of an output image in binary or multi-value by associating each pixel of an original image with each element of a threshold matrix on a one-to-one basis. By applying a repulsive potential to each dot of the dot pattern representing each dot pattern, a process of determining a dot position for defining a dot pattern representing an adjacent tone is performed at each tone, thereby creating the threshold matrix. When an input image having a uniform density corresponding to the size of the threshold matrix is subjected to gradation conversion processing using a matrix, isotropic, aperiodic, and low-frequency Having a dot pattern with few components and having an anisotropic, periodic or pseudo-periodic dot pattern at least in the vicinity of the lowest gradation or the highest gradation Tone reproduction wherein. 2. When defining the dot pattern, the dot pattern is divided into sections smaller than the size of the threshold matrix, and the number of dots in the sections is equal or substantially equal in all sections. 2. The gradation reproducing method according to claim 1, wherein the number of dots is increased or decreased. 3. The gradation reproducing method according to claim 1, wherein when the dot pattern is determined, the dot pattern is determined sequentially from a low gradation dot pattern. 4. The method according to claim 1, wherein when the threshold matrix is repeatedly used two-dimensionally and regularly, a repetition direction is shifted in a vertical or horizontal direction. The tone reproduction method according to any one of the above. 5. The gradation reproducing method according to claim 1, wherein the threshold matrix has a shape different from a square. 6. A color image obtained by decomposing a color image into a plurality of color components, using an original image of at least one color component as an input image, and applying the tone reproduction method according to any one of claims 1 to 5. Gradation reproduction method. 7. A gradation reproducing apparatus for expressing the density of each pixel of an output image in binary or multi-level by associating each pixel of an original image with each element of a threshold matrix on a one-to-one basis. Storing means; comparing means for comparing the density of each pixel of the original image with each pixel using the value of the threshold matrix as a threshold; and a binary or multi-valued dot pattern according to the comparison result of the comparing means. Output means for outputting a dot pattern.The threshold matrix determines a dot position for defining a dot pattern representing an adjacent gradation by applying a repulsive potential to each dot of the dot pattern representing a certain gradation. The threshold value matrix is used to generate an input image having a uniform density corresponding to the size of the threshold value matrix. When the conversion processing is performed, many gradations in the intermediate gradation part have an isotropic, non-periodic, and a dot pattern with few low-frequency components, and at least one of the vicinity of the lowest gradation or the vicinity of the highest gradation. A tone reproduction device characterized by having a dot pattern that is isotropic, periodic or pseudo-periodic. 8. A threshold matrix used when converting the density of each pixel of an original image into binary or multi-valued data, wherein the threshold matrix applies a repulsive potential to each dot of a dot pattern representing a certain gradation. By performing a process of determining a dot position for determining a dot pattern representing an adjacent tone for each tone, the threshold value matrix is used to create a uniform pattern corresponding to the size of the threshold matrix. When an input image with density is subjected to gradation conversion processing, it has a dot pattern that is isotropic, non-periodic, and has few low-frequency components in many gradations in the middle gradation part, and is near the lowest gradation or the highest gradation A threshold matrix having at least one of its neighbors having an anisotropic, periodic or pseudo-periodic dot pattern. 9. A control program for controlling a tone reproduction process for expressing the density of each pixel of an output image in binary or multi-level by associating each pixel of an original image with each element of a threshold matrix on a one-to-one basis. Is a computer-readable storage medium for determining a dot position for defining a dot pattern representing an adjacent gradation by applying a repulsive potential to each dot of the dot pattern representing a certain gradation. A threshold matrix created by performing each of the gradations, and using the value of the threshold matrix as a threshold, comparing the density of each pixel of the original image with each pixel, and performing binary or multi-level conversion in accordance with the comparison result. Including a module that controls to output the obtained dot pattern, and using the threshold matrix, corresponding to the size of the threshold matrix. When an input image with uniform density is subjected to gradation conversion processing, it has an isotropic, aperiodic, low-frequency component dot pattern with many gradations in the middle gradation part, and is near the lowest gradation or the highest gradation. A storage medium characterized by having an anisotropic, periodic or pseudo-periodic dot pattern in at least one of the tonality vicinity.