JP2005278209A - Image processing apparatus and method therefor, and printer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem of image quality deterioration, due to a periodical dot forming pattern, when a multi-level processing is performed to an image data by arranging a dither matrix in quadrille. <P>SOLUTION: The multi-level processing is performed to the image data by arranging a dither matrix in stepwise. That is to say, the dither matrix is arranged, while sequentially shifting it in a main scanning direction or in a sub-scanning direction. It is possible to shift it in either direction or in both directions. Thus, the dot forming pattern being periodically repeated in the main scanning direction or in the sub-scanning direction is averted. Thereby, generation of bandings is suppressed, and an image quality can be improved. It can be applied to an error diffusion method, in which a matrix data is added as a noise. A dot dispersion type matrix, designed on the premise that it is arranged in stepwise manner, can be used as a matrix other than the existing matrixes. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  According to the present invention, image data having a gradation value for each pixel constituting an image within a predetermined gradation number range is converted into a gradation value at an output gradation number that is a gradation number lower than the gradation number. Multi-value image processing apparatus, image processing method, and recording medium recording program for the same, as well as a printing apparatus for printing an image using the image processing technology, and a matrix used for the image processing It relates to the design method.

  Conventionally, as an output device of a computer, an ink jet printer for recording an image by forming dots with several colors of ink ejected from a plurality of nozzles provided in a head has been proposed. Widely used for printing with multiple colors and multiple gradations. In such a printer, normally, only two gradations of dot on / off can be taken for each pixel. Therefore, an image is printed after performing image processing for expressing the gradation of the original image data by the dispersibility of dots, so-called halftone processing.

  In recent years, in order to enrich gradation expression, an ink jet printer capable of expressing gradation of two or more values, on / off for each dot, a so-called multi-value printer has been proposed. For example, a printer that can express three or more types of density for each pixel by changing the dot diameter and ink density, and a multi-tone can be expressed by forming a plurality of dots overlapping each pixel. It is a printer. Even in such a printer, halftone processing is necessary because the gradation of the original image data cannot be expressed sufficiently in each pixel unit.

  Various methods have been proposed for such halftone processing, and a dither method is one of representative methods. In the dither method, a dither matrix having a preset threshold value as each component is used. As an example of the dither matrix, a matrix called a Bayer type is shown in FIG. As shown in FIG. 25, a 4 × 4 Bayer matrix has 16 threshold values from 0 to 15. In the dither method, this dither matrix is associated with image data to determine whether each pixel dot is on or off. FIG. 26 shows the correspondence between the dither matrix and the image data. In FIG. 26, the components shown as (a, b) represent the components of the dither matrix shown in FIG. Naturally, the image data has pixels that are many times larger than the dither matrix size in the vertical and horizontal directions. Therefore, in the dither method, the dither matrix shown in FIG. 25 is repeatedly used by arranging it in a grid pattern in the vertical and horizontal directions.

  In a so-called dot-concentrated halftone dither method or the like, in order to generate a screen angle simulating halftone printing, the repetition direction of the matrix is set to have a certain angle with respect to the two-dimensional pixel arrangement direction. Although a dither matrix may be arranged, in the dot dispersion type dither method, since it is not related to halftone dot printing, the dither matrix is usually arranged in a grid pattern.

  The concept of dot on / off determination in the dither method is shown in FIG. FIG. 27 shows the result of dot on / off determination for an image having a constant gradation value 8 for each pixel. As shown in the figure, the gradation value of the image data is compared with the threshold value of the dither matrix corresponding to each pixel, and it is determined that a dot is formed at a pixel having a larger gradation value of the image data. For the gradation value 8, dots are turned on in a checkered pattern as shown in FIG. Of course, the pixel for which the dot is turned on changes in accordance with the gradation value of the image data and the dither matrix used in the multi-value conversion.

  FIG. 28 shows a state in which the on / off determination result of FIG. 27 is developed in the image area. FIG. 28 shows the result of dot on / off determination obtained by using the dither matrix with the correspondence shown in FIG. The hatched portions in FIG. 28 are pixels for which dots should be turned on. As shown in the figure, pixels on which dots are to be formed are widely distributed in a checkered pattern. If the dither matrix used for multi-value conversion is changed, dots are turned on in a pattern different from that in FIG. In the ink jet printer, for example, dots are formed in each pixel according to the result of the multi-value quantization means such as the dither method described above.

  However, when multi-value processing is performed using the dither method, as is apparent from the correspondence between the dither matrix and the image data (FIG. 26), the image area has a substantially constant size in units of the dither matrix. The dot formation pattern appears in a grid pattern. In the conventional printing apparatus, the image quality may be deteriorated due to such a fixed pattern for the following reason.

  In an inkjet printer, if there is a mechanical manufacturing error or the like in a nozzle that ejects ink, variations in the amount of ink to be ejected and the positions of dots to be formed occur. When the amount of ink discharged from a specific nozzle is large or the interval between adjacent rasters varies due to a shift in the dot formation position, density unevenness occurs in the printed image at that portion. In particular, in a printer that forms dots while performing main scanning with the head, the same unevenness tends to occur continuously in the main scanning direction, and therefore, streaky density unevenness called banding may appear in the main scanning direction.

  When the generation of dots is determined by the dither matrix, the dot generation rate may be biased for each raster depending on the threshold value of the dither matrix and the gradation value of the image data. When the occurrence of dots in the entire image is determined by arranging the dither matrix in a grid pattern, such a biased pattern is repeatedly generated, and the bias in the dot generation rate for each raster is promoted. In such a case, if a raster having a large number of dots is formed by nozzles formed by shifting the positions of the dots, remarkable banding may occur and image quality may be degraded.

  Such a problem occurs not only when a Bayer type dither matrix is used, but also when a so-called blue noise mask type dither matrix having a random dot generation pattern having no specific regularity is used. This is because even if the random dot generation pattern without regularity generated by the blue noise mask type dither matrix has a large variation in the dot generation rate for each raster.

  On the other hand, as a technique for improving the image quality by suppressing the occurrence of banding, so-called overlap recording is proposed. This is a technique for forming each raster using two or more different nozzles. As an example, consider a case where each raster is formed by two main scans using two nozzles. In the first main scan, only odd-numbered dots of each raster are formed intermittently. Thereafter, sub-scanning is performed, and in the second main scanning, only even-numbered dots are intermittently formed using nozzles different from the first. If such a recording method is adopted, each raster is formed by two or more different nozzles, so that the deviation of dot formation positions caused by the characteristics of the nozzles can be dispersed on each raster, and the image quality can be improved. Can be improved.

  However, when multi-value processing is performed using the dither method, the above-described overlap effect may not be sufficiently obtained. Consider a case where the result shown in FIG. 28 is obtained by multi-value processing by the dither method. At this time, the pixels whose dots should be turned on are arranged in a checkered pattern. This means that the dots of either odd-numbered or even-numbered pixels are turned on when viewed for each raster. In the above-described example of printing by the overlap method, dots of odd-numbered pixels of each raster are formed in the first main scan, and dots of even-numbered pixels of each raster are formed in the second main scan. When recording by the overlap method is executed for the multi-value quantization result of FIG. 28, each raster is eventually formed only by either the first or second main scan. In such a case, the advantage of the overlap method obtained by forming dots on each raster using different nozzles cannot be utilized, and the advantage of improving the image quality may not be sufficiently obtained.

  The Bayer-type dither matrix used in the above example is a matrix that is relatively frequently used in multi-value processing by the dither method. When such a matrix is used, it has been found that there is a significant problem that the advantage of the overlap method cannot be obtained sufficiently.

  In the above description, a Bayer-type dither matrix is taken as an example. However, even when other dither matrices are used, the dot generation rate for each raster and the dot generation rates of even pixels and odd pixels are set for every gradation value. Since it is difficult to make them substantially equal, the same problem has arisen. If a dither matrix different from the Bayer type is used, the dots are formed in a pattern different from the pattern shown in FIG. However, even in this case, the same pattern is repeatedly displayed in a grid pattern in the vertical and horizontal directions of the dither matrix. Therefore, if the number of dots formed for each raster is uneven, banding may occur based on variations in the ink ejection direction. Further, if there is a raster in which a relatively small threshold is biased in odd-numbered pixels, the effect of recording by the overlap method cannot be sufficiently obtained.

  The same problem occurs in the multi-value quantization means other than the dither method. As a multilevel method other than the dither method, for example, there is a method called an error diffusion method. This method is originally multi-valued without using a matrix, but multi-valued after intentionally adding predetermined noise to image data for the purpose of ensuring the continuity of the multi-valued result. May be performed. When a matrix having predetermined noise data is used as the noise data in the correspondence relationship shown in FIG. 26, the above-described problem described by taking the dither method as an example due to the addition of noise with a fixed repetitive pattern The same problem may occur.

  The present invention has been made to solve the above-described problems, and in a multi-value quantization means using a predetermined matrix, a pattern in which dots are formed periodically appears in the vertical and horizontal directions within an image area. It is an object of the present invention to provide a technique for avoiding this and improving the image quality.

In order to solve at least a part of the above-described problems, the present invention employs the following configuration.
The image processing apparatus of the present invention
Depending on the magnitude relationship between the gradation value and a predetermined threshold value for each pixel, the image data is composed of a two-dimensional array of pixels and has gradation values in a predetermined range for each pixel. An image processing apparatus that determines whether dots are on or off and multi-values the image data,
Storage means for storing a two-dimensional matrix having a predetermined value as a component in a size smaller than the size of the image data constituted by the two-dimensional pixel array;
A multi-value conversion is performed by reflecting the matrix component in the gradation value or the threshold value of the image data in a correspondence relationship in which the matrix is arranged with respect to the image data in a state shifted from a grid. The gist is to provide a valuation means.

The image processing method of the present invention includes:
Depending on the magnitude relationship between the gradation value and a predetermined threshold value for each pixel, the image data is composed of a two-dimensional array of pixels and has gradation values in a predetermined range for each pixel. An image processing method for determining on / off of dots and performing multi-value conversion of the image data,
A two-dimensional matrix having a preset value as a component is reflected in the gradation value or the threshold value of the image data in a correspondence relationship in which the two-dimensional matrix is arranged with respect to the image data in a state shifted from a grid. Thus, the gist is to perform the multi-value processing.

  According to the image processing apparatus and the image processing method, it is possible to suppress the dot formation pattern from being repeatedly generated in a square pattern in units of the matrix, and the image quality can be improved. The grid shape means that, for example, when the matrix is captured as one unit as shown in FIG. 26, the arrangement on the image data is a grid shape. In the image processing apparatus and the image processing method of the present invention, since the matrix is associated in a state deviated from the grid shape, it is possible to avoid repeated formation of dots in a constant pattern in the x direction and the y direction.

  When dots are repeatedly formed in a fixed pattern, the number of dots formed tends to be biased for each arrangement in the x direction or y direction. For example, when there is a deviation in the number of dots formed for each arrangement in the x direction, if there is a deviation in the recording position of the arrangement with a large number of dots, noticeable banding occurs and the image quality deteriorates. In the image processing apparatus and the image processing method of the present invention, it is possible to avoid the occurrence of bias in the number of dots formed for each arrangement by avoiding dots being repeatedly formed in a fixed pattern. Therefore, it is possible to suppress the occurrence of banding and to perform multi-value conversion with excellent image quality.

  From the standpoint of avoiding the repeated formation of dots in a fixed pattern, it is possible to set the matrix size to a very large size. According to such means, the memory for storing the matrix data becomes enormous. In addition, it takes a long time to read data from the memory. According to the image processing apparatus and the image processing method of the present invention, the above effects can be obtained without increasing the size of the matrix. That is, according to the image processing apparatus and the image processing method of the present invention, there is an advantage that the image quality can be improved without increasing the memory and the processing time.

  An example of a state deviating from the grid is shown in FIG. FIG. 1 is an explanatory diagram showing the correspondence between image data and a matrix. Each square in FIG. 1 represents a pixel. Moreover, the thick line in FIG. 1 shows the area | region matched by a matrix. Conventionally, as shown in FIG. 26, the matrix is associated in a grid pattern. However, in the image processing apparatus of the present invention, for example, as shown in FIG. is there. Although FIG. 1 shows an example of shifting in the y direction, it may be handled while shifting in the x direction, or may be shifted in both directions. Also, the amount of shifting in each direction can be set to various values. Further, it is not necessary to shift the matrix in all areas of the image data, and it is possible to shift the matrix from some squares to correspond.

In the image processing apparatus of the present invention described above, if the correspondence between the image data and the matrix is described more accurately using mathematical expressions,
Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, for each pixel, an image processing apparatus that performs multi-value conversion of two or more values for determining dot on / off according to the magnitude relationship between the gradation value and a predetermined threshold value,
Storage means for storing a matrix of mx × my (an integer satisfying 1 ≦ mx <nx, 1 ≦ my <ny) having a preset value as a component;
Multi-value conversion means for performing multi-value conversion by reflecting the components of the matrix in the gradation value or the threshold value for each pixel,
The matrix component to be reflected in the kxth pixel in the x direction (kx is an integer satisfying 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer satisfying 0 ≦ ky ≦ ny) is expressed by the following equation (1). The image processing apparatus is given by (xt, yt) (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my).

xt = (rx · dx1 + ry · dx2 + kx)% mx;
yt = (rx · dy1 + ry · dy2 + ky)% my;
rx = kx div mx;
ry = ky div my; (1)
here,
a div b is an operator for calculating the quotient of a / b by an integer value;
a% b is a remainder operator for calculating the remainder of a / b,
dx1, dx2, dy1, dy2 are integers of 0 or more, and at least one of dx1, dx2, dy1, dy2 is a value other than 0.

  According to such an image processing apparatus, as in the above-described image processing apparatus, it is possible to reduce the occurrence of repeated dot formation patterns in one direction in units of the matrix, and the image quality can be improved. The correspondence between the image data and the matrix in the above invention will be specifically described with reference to FIG.

  As shown in FIG. 1, the image data is composed of pixels two-dimensionally arranged in the x direction and the y direction, and nx pixels from 0 to nx-1 in the x direction and 0 in the y direction. It has ny pixels from No. to ny−1. An arbitrary pixel of this image data is represented as (kx, ky) using numbers in the x direction and y direction (hereinafter referred to as pixel numbers). The matrix is also composed of data arranged two-dimensionally, mx components from 0 to mx-1 in the x direction, and my components from 0 to my-1 in the y direction. have. An arbitrary component of the matrix is expressed as (xt, yt).

The correspondence between the matrix and the image data is obtained by the above equation (1). Consider a region where kx <mx and ky <my. This is an area A1 shown in FIG. At this time,
rx = kx div mx = 0;
ry = ky div my = 0;
Therefore, the above equation (1) becomes
xt = kx% mx = kx;
yt = ky% my = ky;
It becomes. In other words, in this area, the upper left pixel in FIG. 1 of the image data corresponds to the upper left component of the matrix.

Next, consider a region where m1 ≦ kx <2mx and ky <my. It is area | region A2 in FIG. In such areas,
rx = kx div mx = 1;
ry = ky div my = 0;
Therefore, the above equation (1) is
xt = (dx1 + kx)% mx;
yt = (dy1 + ky)% my;
It becomes. If dx1 = dy1 = 0, xt = kx% mx, yt = ky% my, and kx = mx, mx + 1, mx + 2,..., kx% mx = 0, 1, 2,. The same correspondence as that in which the matrix is associated with the area A1 is given between the image data of the area A2 and the matrix.

  On the other hand, when dy1 = 1, since yt = (1 + ky)% my = 1, 2, 3... With respect to ky = 0, 1, 2,. On the other hand, the matrix is shifted by a value of 1 in the negative direction of the y direction. FIG. 1 shows the relationship between the matrix and the image data in such a state. In FIG. 1, the matrix is associated with the portion indicated by the bold line and the broken line.

  Similarly, a matrix can be associated with image data in other image regions. When dy1 = 1, the matrix is associated as shown by a thick line in FIG. That is, the corresponding matrix is gradually shifted in the negative y direction as the pixel number in the x direction increases. If dy1 is a value larger than 1, the amount by which the matrix shifts is even larger.

  FIG. 1 shows a case where dx1 = dx2 = dy2 = 0 and dy1 = 1. If dx1 ≠ 0, the corresponding matrix shifts in the x direction as the pixel number in the x direction increases. If dx2 ≠ 0, the corresponding matrix shifts in the x direction as the pixel number in the y direction increases. If dy2 ≠ 0, the corresponding matrix shifts in the y direction as the pixel number in the y direction increases. In the present invention, at least one of dx1, dx2, dy1, dy2 is set to a value other than zero. Therefore, in the present invention, as the pixel number increases, the corresponding matrix shifts in at least one of the x direction and the y direction.

  Although FIG. 1 shows a case where dy1 is a constant value, for example, dy1 may be changed every time the value of rx or ry changes. If dy1 is set in this way, how the matrix is displaced in the y direction changes variously. Similarly, the other values dx1, dx2, dy2 may take different values each time the value of rx or ry changes.

  For example, as shown in FIG. 1, consider a case where multi-value conversion is performed on pixels of kx = 0 to nx−1 and ky = 0 with respect to image data having a corresponding gradation value as shown in FIG. The multi-value quantization result of kx = 0 to mx-1 (corresponding to the area A1 in FIG. 1) and the multi-value quantization result of kx = mx to 2mx-1 (corresponding to the area A2) have matrix values corresponding to each other. Since they are different, the multi-value conversion result is naturally different. As a result, it is possible to avoid the multi-value conversion result in the pixel of kx = 0 to nx−1, that is, the dot formation pattern from appearing repeatedly in the x direction. In the image processing apparatus of the present invention, multi-value conversion with excellent image quality can be performed by suppressing repetition of dot formation patterns. Of course, such an effect is not limited to image data having a constant gradation value.

  In the above description, the case where there is one type of matrix used for multi-value processing has been described. However, the present invention is also applicable to a case where a plurality of types of such matrices are provided. For example, a different matrix may be used for the other area including the area A2 in FIG. 1 and the area A1.

Various means can be considered as the multi-value conversion means in the image processing apparatus.
For example, the multi-value conversion means may be a means for performing multi-value conversion after reflecting the matrix components as noise data in the image data.
Specifically, for example, a case where multi-value conversion is performed by an error diffusion method can be given.

On the other hand, as a case where a matrix component is reflected in a threshold value used for multi-leveling,
The multi-value conversion means may be means for performing multi-value conversion by a dither method using any component of the matrix as the threshold value.

  In the dither method, multi-value processing is performed according to the magnitude relationship between the tone value of the image data and the threshold value of the matrix. Therefore, a certain dot formation pattern is likely to be repeatedly generated according to the correspondence between the image data and the matrix. According to the image processing apparatus, it is possible to avoid such repetition and improve the image quality by shifting the matrix to correspond. In addition, the image quality can be improved without losing the advantage of the dither method in which the processing time required for multi-value processing is short.

Furthermore, in the image processing apparatus of the present invention,
The matrix is preferably a dot-dispersed dither matrix that can perform multi-valued dot dispersion with high dot dispersion within the image area.

  The dot dispersion type dither matrix is a matrix in which thresholds are set so that the positions of dots formed according to gradation values are dispersed as much as possible in the matrix. Such a matrix is often used for the purpose of reducing dot visibility by forming dots sparsely. The image processing apparatus of the present invention avoids the occurrence of uneven dot dispersibility by avoiding repetition of dot formation patterns, thereby obtaining the effect of suppressing density unevenness. If a matrix is used, the effect can be exhibited effectively and the image quality can be improved.

As such a dot dispersion type dither matrix, various matrices are known. For example,
The matrix may be a Bayer type dither matrix.

  FIG. 2 shows an example of a 4 × 4 matrix as a Bayer type dither matrix. It is a matrix having 16 types of threshold values from 0 to 15. Looking at the 3 × 3 portion shown in the region P1 of FIG. 2, the components at the four corners are assigned values 0 to 3 in order. Similarly, values 4 to 7 are assigned to the four corners of the 3 × 3 portion shown in the region P4, values 8 to 11 are assigned to the four corners of the region P2, and values 12 to 15 are assigned to the four corners of the region P3. It has been. A dither matrix called a Bayer type is created by preparing a plurality of small matrices to which four threshold values are assigned in a predetermined pattern for four corners of a predetermined size smaller than the size of the dither matrix. Matrix. In FIG. 2, the case of 4 × 4 has been described as an example, but other size matrices are similarly defined.

  Since the Bayer-type matrix is generated by the above-described method, there is a feature that the assignment of the threshold value to each component in the matrix is regular. Therefore, the dots formed according to the gradation value have a regular pattern throughout the image area, and banding is likely to occur. According to the image processing apparatus of the present invention, since the occurrence of such a regular pattern can be reduced, the image quality can be greatly improved.

The matrix may be a blue noise mask type matrix.
FIG. 3 shows an example of a 64 × 64 matrix as a blue noise mask type matrix. For the sake of illustration, only a part is shown. In this dither matrix, the threshold value is assigned so that the appearance of the threshold value (0 to 255) is not biased in any 16 × 16 region within the 64 × 64 size matrix. A matrix having such a property is called a blue noise mask type matrix. Blue noise mask type matrices can be set for various sizes and threshold ranges.

  Further, the matrix may be a matrix that can perform multi-valued dot dispersion with high image dispersibility in an image area wider than the area corresponding to the size of the matrix.

By using the image processing apparatus of the invention described above, the invention of the following printing apparatus can be realized.
The printing apparatus of the present invention includes:
Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, the image is formed on the print medium by driving the head to form dots according to the multi-value quantization result determined according to the magnitude relationship between the gradation value and the predetermined threshold value for each pixel. A printing device for printing
Storage means for storing a matrix of mx × my (an integer satisfying 1 ≦ mx <nx, 1 ≦ my <ny) having a preset value as a component;
Multi-value conversion means for performing multi-value conversion by reflecting any component of the matrix in either one of the gradation value and the threshold value for each pixel,
The matrix component to be reflected in the kxth pixel in the x direction (kx is an integer satisfying 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer satisfying 0 ≦ ky ≦ ny) is expressed by the above equation (1). The gist is that (xt, yt) given (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my).

  According to such a printing apparatus, as in the image processing apparatus described above, multi-value conversion is performed after shifting the matrix, and dots are formed according to the result of the multi-value conversion. Printing is possible. In such a printing apparatus, as in the image processing apparatus described above, the matrix and the image data can be variously handled.

Furthermore, in the printing apparatus,
The head includes a plurality of nozzles arranged in the y direction;
Main scanning means for reciprocating the head in the x direction relative to the print medium;
Sub-scanning means for relatively moving the head and the print medium in the y direction;
Drive control means for controlling the driving of the main scanning means, the sub-scanning means and the head to form each dot row arranged in the x direction using two or more nozzles,
It is desirable that at least dy1 ≠ 0.

  In such a printing apparatus, each dot row arranged in the x direction is formed using two or more nozzles. By forming the dot rows using different nozzles, it is possible to disperse the deviations in dot formation positions caused by the characteristics of the nozzles and improve the image quality. In the printing apparatus, by setting dy1 ≠ 0, dots are formed after multi-value conversion is performed by sequentially shifting the matrix in the y direction with respect to the image data. Therefore, in the printing apparatus, the effect of forming each dot row using different nozzles can be sufficiently exhibited, and high-quality printing is possible. This effect will be described with a specific example.

  For example, consider a case where a dot row is recorded using two nozzles A and B. As a dot row recording method in this case, a method of recording odd-numbered dots in the x direction with the nozzle A and recording even-numbered dots with the nozzle B can be considered. When the matrix is associated with the image data in a grid pattern, the pixels on which dots are formed may be biased to odd-numbered pixels in the x direction depending on the relationship between the gradation value of the image data and the matrix threshold value. . When such a deviation occurs, most of the dot rows are formed by the nozzles A, and the effect of forming the dot rows using different nozzles cannot be sufficiently obtained. If multi-value conversion is performed by shifting the matrix in the y direction, the image data and the threshold value of the matrix are associated with various changes, so that the pixels in which dots are formed are biased to the odd-numbered pixels. Can be suppressed. As a result, it is possible to sufficiently obtain the effect of forming each dot row with different nozzles. Here, the case where the dot row is formed by two nozzles has been described as an example, but the same effect can be obtained even when the number of nozzles used for forming each dot row is different.

  In the printing apparatus, “at least the dy1 ≠ 0” is set. This is because the effect of improving the image quality is great when the matrix is shifted in the y direction. Of course, the matrix may be shifted and corresponded only in the x direction, or may be shifted and corresponded in both the x direction and the y direction.

  As the matrix used in the image processing apparatus of the present invention, an existing matrix such as the Bayer type matrix or the blue noise mask type matrix described above may be used, but the matrix designed by the design method shown below. Can also be used.

The matrix designing method of the present invention includes:
Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, when performing multi-value determination that determines dot on / off according to the magnitude relationship between the gradation value and a predetermined threshold value for each pixel, the gradation value or the value to be reflected in the threshold value Mx × my (1 ≦ mx <nx, 1 ≦ my <ny integer) matrix design method,
(A) setting a correspondence between each pixel in the multi-value quantization and the matrix component;
(B) n values (n is an integer of 1 or more) from the top when values assigned to the matrix are arranged in accordance with the magnitude relationship, as an arbitrary component of the matrix in consideration of threshold dispersion A setting process;
(C) sequentially setting values remaining as values to be assigned to the matrix to components obtained by a predetermined calculation,
The step (c)
(C-1) The distance between each component of the matrix to which no value is assigned and the component of the matrix to which a value has already been set, is set for each of the matrices associated with each pixel constituting the image data. A process of evaluating the relationship between the components across a plurality of matrices;
(C-2) From the step of setting the first value when the remaining values are arranged according to the magnitude relationship to the component evaluated to be the longest distance from the component for which the value has already been set It becomes the summary.

  The matrix designed by the design method includes a component for which a threshold is set by the intention of the designer and a component for which a threshold is set by calculation. In other words, the step (a) in the design method is a step of setting the arrangement of the matrix in the image data. As the arrangement, various arrangements such as a square arrangement as shown in FIG. 26 and an arrangement in which the matrix is sequentially shifted in the y direction as shown in FIG. 1 can be considered.

  In step (b), n threshold values are assigned to the components of the matrix. FIG. 4 shows an example in which n threshold values are assigned in ascending order when the threshold values are arranged in ascending order. In FIG. 4, n threshold values 0, 1, 2,..., N−1 are assigned to components such as (0, 0) and (2, 0), respectively. This assignment is made in consideration of dot dispersibility, but can basically be assigned to any component. The n threshold values may be one. In the case of n = 1, only the value 0 in FIG. 4 is assigned.

  In the step (c-1), the distance is evaluated over a plurality of matrices. The meaning of “over a plurality of matrices” will be described. The example of evaluation concerning FIG. 5 was shown. FIG. 5 shows a state in which a matrix is associated with image data according to the correspondence set in step (a). A portion indicated by a thick line in FIG. 5 corresponds to a matrix. If threshold values from 0 to n-1 are set for the components shown in FIG. 4, threshold values from 0 to n-1 are set at the locations shown in FIG. . Here, consider a case where the distance is evaluated for a component of (2, my−1), for example, as a component whose value is undefined. In FIG. 5, mp, which is one of the pixels corresponding to this component, is shown with hatching. At this time, there are some pixels around the pixel mp corresponding to the components for which the threshold 0 has already been set. As the distances from these pixels, various values are calculated as indicated by d1 to d4 in FIG. 5, for example. Among them, d2 to d4 are distances from pixels corresponding to a matrix different from the pixel mp. “Across a plurality of matrices” means that distances to pixels on different matrices are also evaluated. One of the distances d1 to d4 calculated in this way is selected as an evaluation value for the threshold 0. In FIG. 5, the distance is calculated for a part of the pixels to which the threshold value 0 is assigned, but naturally all the pixels may be the target. Similarly, evaluation values for all threshold values from 0 to n−1 that have been set are calculated. Further, such an evaluation value is calculated not only for the pixel mp but also for all components for which the threshold is undefined.

  In step (c-2), a component that is evaluated to be the farthest from the component for which a threshold has already been set is obtained based on the evaluation value calculated in step (c-1). Such a component corresponds to the component having the highest dispersibility. Then, an nth threshold value is set for the component. In the example of FIG. 5, since the threshold values are given in ascending order, the set threshold value is the nth smallest value.

  According to the matrix design method of the present invention, the steps (c-1) and (c-2) described above are sequentially repeated, so that the dispersibility of the dots is emphasized in consideration of the arrangement of the matrix in the image data. The threshold can be set. Therefore, if multi-value processing is performed using a matrix designed by the design method of the present invention, high-quality multi-value processing can be executed.

  In contrast to the example shown in FIG. 5, the threshold values may be set in descending order. For example, when it is desired to sufficiently secure the dispersibility of white dots generated in a so-called solid area, it may be desirable to set in order from a larger threshold value. In addition, the matrix threshold value can be set by evaluating the distance in consideration of elements other than the dispersibility of dots.

In the above matrix design method,
The correspondence in the step (a) is as follows:
The matrix component to be reflected on the x1th pixel in the x direction (x1 is an integer satisfying 0 ≦ x1 ≦ nx) and the y1th pixel in the y direction (y1 is an integer satisfying 0 ≦ y1 ≦ ny) is expressed by the above equation (1). It is also possible to correspond to given (xt, yt) (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my).

  Such a correspondence relationship corresponds to a case where the matrix is made to correspond to the image data in a state shifted from a grid shape as described in the image processing apparatus. That is, according to the above design method, it is possible to generate a matrix in which the dispersibility of dots is sufficiently ensured over the entire image area, on the premise that the matrix is made to correspond in a state of being deviated from the grid. Therefore, high-quality image processing can be performed by using the matrix designed by the above design method.

  Since the image device of the present invention described above can also be configured by realizing the above multi-value processing by a computer, the present invention can also take an aspect as a recording medium recording such a program.

The recording medium of the present invention is
Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, the present invention is a recording medium on which a computer readable recording program for performing multi-value determination for determining on / off of dots according to the magnitude relationship between the gradation value and a predetermined threshold value for each pixel is provided. And
A matrix of mx × my (an integer satisfying 1 ≦ mx <nx, 1 ≦ my <ny) in which preset values are stored as components;
Of the matrix, the gradation value or the threshold value of the kxth pixel in the x direction (kx is an integer satisfying 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer satisfying 0 ≦ ky ≦ ny) A recording in which a program for realizing the function of multi-value conversion reflecting the (xt, yt) (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my) component given by the above equation (1) is recorded It is a medium.

  The above-described image processing apparatus of the present invention can be realized by executing the program recorded in each recording medium on the computer. Storage media include flexible disks, CD-ROMs, magneto-optical disks, IC cards, ROM cartridges, punch cards, printed materials printed with codes such as bar codes, and computer internal storage devices (memory such as RAM and ROM). ) And external storage devices can be used. In addition, an aspect as a program supply apparatus that supplies a computer program for realizing a multi-valued function of the image processing apparatus to a computer via a communication path is also included.

Hereinafter, embodiments of the present invention will be described based on examples.
(1) Device configuration:
FIG. 6 is a block diagram showing the configuration of an image processing apparatus and a printing apparatus as an embodiment of the present invention. As shown in the figure, a scanner 12 and a color printer 22 are connected to a computer 90. In addition to functioning as an image processing apparatus by loading and executing a predetermined program on the computer 90, the computer 90 functions as a printing apparatus together with the printer 22. The computer 90 includes the following units interconnected by a bus 80 with a CPU 81 that executes various arithmetic processes for controlling operations related to image processing according to a program. The ROM 82 stores programs and data necessary for executing various arithmetic processes by the CPU 81 in advance, and the RAM 83 temporarily reads and writes various programs and data necessary for the CPU 81 to execute various arithmetic processes. Memory. The input interface 84 controls input of signals from the scanner 12 and the keyboard 14, and the output interface 85 controls output of data to the printer 22. The CRTC 86 controls signal output to the CRT 21 capable of color display, and the disk controller (DDC) 87 controls data exchange with the hard disk 16, the flexible drive 15, or a CD-ROM drive (not shown). The hard disk 16 stores various programs loaded in the RAM 83 and executed, various programs provided in the form of device drivers, and the like.

  In addition, a serial input / output interface (SIO) 88 is connected to the bus 80. The SIO 88 is connected to the modem 18 and is connected to the public telephone line PNT via the modem 18. The computer 90 is connected to an external network via the SIO 88 and the modem 18, and a program necessary for image processing can be downloaded to the hard disk 16 by connecting to a specific server SV. It is also possible to load a necessary program from the flexible disk FD or CD-ROM and cause the computer 90 to execute it.

  FIG. 7 is a block diagram illustrating a software configuration of the printing apparatus. In the computer 90, an application program 95 operates under a predetermined operating system. A video driver 91 and a printer driver 96 are incorporated in the operating system, and image data FNL to be transferred to the printer 22 is output from the application program 95 via these drivers. An application program 95 that performs image retouching or the like reads an image from the scanner 12 and displays the image on the CRT display 21 via the video driver 91 while performing predetermined processing on the image. Data ORG supplied from the scanner 12 is original color image data ORG that is read from a color original and includes three color components of red (R), green (G), and blue (B).

  When the application program 95 issues a print command, the printer driver 96 of the computer 90 receives image information from the application program 95, and signals that can be processed by the printer 22 (here, cyan, magenta, yellow, and black colors). Multi-valued signal). In the example illustrated in FIG. 7, the printer driver 96 includes a resolution conversion module 97, a color correction module 98, a color correction table LUT, a halftone module 99, and a rasterizer 100.

  The resolution conversion module 97 serves to convert the resolution of the color image data handled by the application program 95, that is, the number of pixels per unit length into a resolution that can be handled by the printer driver 96. Since the image data thus converted in resolution is still image information consisting of three colors of RGB, the color correction module 98 refers to the color correction table LUT, and cyan (C) and magenta used by the printer 22 for each pixel. (M), yellow (Y), and black (K) data are converted into data. The color-corrected data in this way has gradation values with a width of, for example, 256 gradations. The halftone module 99 executes halftone processing for expressing such gradation values by the printer 22 by forming dots dispersedly. The halftone module 99 in this embodiment is included in at least the image processing apparatus in the present invention. The processed image data is rearranged in the order of data to be transferred to the printer 22 by the rasterizer 100 and output as final image data FNL. In this embodiment, the printer 22 only serves to form dots according to the image data FNL and does not perform image processing, but of course, the printer 22 may perform these processes.

  Next, a schematic configuration of the printer 22 will be described with reference to FIG. As shown in the figure, the printer 22 includes a mechanism for transporting the paper P by the paper feed motor 23, a mechanism for reciprocating the carriage 31 in the axial direction of the platen 26 by the carriage motor 24, and a print head mounted on the carriage 31. And a control circuit 40 for exchanging signals with the paper feed motor 23, the carriage motor 24, the print head 28 and the operation panel 32. .

  The mechanism for reciprocating the carriage 31 in the axial direction of the platen 26 is an endless drive belt between the carriage motor 24 and a slide shaft 34 that is installed in parallel with the axis of the platen 26 and slidably holds the carriage 31. 36, a pulley 38 for extending 36, a position detection sensor 39 for detecting the origin position of the carriage 31, and the like.

  The carriage 31 can be mounted with a black ink (Bk) cartridge 71 and a color ink cartridge 72 containing cyan (C), magenta (M), and yellow (Y) inks. . A total of four ink ejection heads 61 to 64 are formed on the print head 28 below the carriage 31. An inlet pipe 67 (which guides ink from the ink tank to the color heads) is provided at the bottom of the carriage 31. 9) is erected. When the black (Bk) ink cartridge 71 and the color ink cartridge 72 are mounted on the carriage 31 from above, the introduction pipe 67 is inserted into the connection hole provided in each cartridge, and the ejection heads 61 to 64 are ejected from each ink cartridge. Ink can be supplied to the printer.

  A mechanism for ejecting ink and forming dots will be described. FIG. 9 is an explanatory diagram showing a schematic configuration inside the ink ejection head 28. When the ink cartridges 71 and 72 are mounted on the carriage 31, the ink in the ink cartridge is sucked out via the introduction pipe 67 using the capillary phenomenon as shown in FIG. The print head 28 is guided to each color head 61 to 64. When the ink cartridge is first installed, an operation of sucking ink to the respective color heads 61 to 64 is performed by a dedicated pump. In this embodiment, a pump for sucking and a cap that covers the print head 28 at the time of sucking are performed. The illustration and description of such a configuration is omitted.

  As will be described later, the heads 61 to 64 of each color are provided with 48 nozzles Nz for each color (see FIG. 12), and each of the nozzles is one of electrostrictive elements and is responsive. An excellent piezo element PE is arranged. FIG. 10 shows the structure of the piezo element PE and the nozzle Nz in detail. As illustrated in the upper part of FIG. 10, the piezo element PE is installed at a position in contact with the ink passage 68 that guides ink to the nozzle Nz. As is well known, the piezo element PE is an element that transforms electro-mechanical energy at a very high speed because the crystal structure is distorted by application of a voltage. In this embodiment, by applying a voltage having a predetermined time width between the electrodes provided at both ends of the piezo element PE, the piezo element PE extends for the voltage application time as shown in the lower part of FIG. One side wall of 68 is deformed. As a result, the volume of the ink passage 68 contracts according to the expansion of the piezo element PE, and the ink corresponding to the contraction becomes particles Ip and is ejected from the tip of the nozzle Nz at high speed. Printing is performed by the ink particles Ip soaking into the paper P mounted on the platen 26.

  Next, an internal configuration of the control circuit 40 of the printer 22 will be described, and a method of driving the head 28 including a plurality of nozzles Nz provided in the head will be described. FIG. 11 is an explanatory diagram showing the internal configuration of the control circuit 40. As shown in FIG. 11, in addition to the CPU 81, PROM 42, and RAM 43, the control circuit 40 includes a PC interface 44 for exchanging data with the computer 90, a paper feed motor 23, a carriage motor 24, an operation panel 32, and the like. A peripheral input / output unit (PIO) 45 for exchanging signals with the timer, a timer 46 for timing, a driving buffer 47 for outputting dot on / off signals to the heads 61 to 66, and the like. These elements and circuits are connected to each other via a bus 48. The control circuit 40 also includes a transmitter 51 that outputs a drive waveform for driving the piezoelectric element at a predetermined frequency, and a distributor 55 that distributes the output from the transmitter 51 to the heads 61 to 64 at a predetermined timing. Is provided. The control circuit 40 receives the dot data processed by the computer 90, temporarily stores it in the RAM 43, and outputs it to the driving buffer 47 at a predetermined timing.

  A mode in which the control circuit 40 outputs signals to the heads 61 to 64 will be described. FIG. 12 is an explanatory diagram showing the arrangement and connection of one nozzle row of the heads 61 to 64 as an example. In these nozzles, 48 nozzles Nz are arranged in a staggered manner at a constant nozzle pitch k. The state of dots formed by the nozzle row of this embodiment is shown on the right side of FIG. A circle indicated by a solid line is a dot that can be formed by one main scanning. The broken line is shown as a standard indicating the dot interval. As shown in FIG. 12, in this embodiment, the nozzle pitch k corresponds to 2 dots. Note that the 48 nozzles Nz included in each nozzle array need not be arranged in a staggered manner, and may be arranged on a straight line. However, when arranged in a zigzag pattern as shown in FIG. 12, there is an advantage that the nozzle pitch k can be easily set small.

  Each nozzle row of the heads 61 to 64 is interposed in a circuit having the drive buffer 47 as the source side and the distribution output device 55 as the sink side. Each piezo element PE constituting the nozzle row is connected to its electrode. One is connected to each output terminal of the drive buffer 47, and the other is connected to the output terminal of the distribution output device 55 in a lump. As shown in FIG. 12, a drive waveform of the transmitter 51 is output from the distribution output device 55. When the CPU 81 determines ON / OFF for each nozzle and outputs a signal to each terminal of the drive buffer 47, only the piezo element PE that has received the ON signal from the drive buffer 47 side is driven according to the drive waveform. The As a result, the ink particles Ip are simultaneously ejected from the nozzles of the piezo element PE that has received the ON signal from the transfer buffer 47. As shown in FIG. 12, since the nozzle rows are formed in a staggered pattern, when dots are formed while transporting the carriage 31, in order to form dots arranged in a row in the main scanning direction, It is necessary to shift the ejection timing of the ink in the nozzle row. Similarly, it is necessary to shift the ink ejection timing for each of the heads 61 to 64. The CPU 81 considers such a timing shift and outputs an ON / OFF signal for each dot via the driving buffer 47 to form dots for each color.

  With the hardware configuration described above, the printer 22 of this embodiment transports the paper P by the paper feed motor 23 (hereinafter referred to as sub-scanning), and reciprocates the carriage 31 by the carriage motor 24 (hereinafter referred to as main power). At the same time, the piezo elements PE of the respective color heads 61 to 64 of the print head 28 are driven to discharge the respective color inks to form dots and form a multicolor image on the paper P.

  In this embodiment, as described above, the printer 22 including the head that discharges ink using the piezo element PE is used. However, a printer that discharges ink by another method may be used. For example, the present invention may be applied to a printer of a type in which electricity is supplied to a heater arranged in the ink passage and ink is ejected by bubbles generated in the ink passage.

(2) Dot formation control:
Next, the dot formation control process in this embodiment will be described. The flow of the dot formation control processing routine is shown in FIG. This is a process executed by the CPU 81 of the computer 90.

  When this process is started, the CPU 81 inputs image data (step S100). This image data is data delivered from the application program 95 shown in FIG. 2, and 16 levels of gradation of values 0 to 15 for each color of R, G, B for each pixel constituting the image. Data having a value. The resolution of the image data changes according to the resolution of the original image data ORG.

  The CPU 81 converts the resolution of the input image data into a resolution for the printer 22 to print (step S105). When the image data is lower than the printing resolution, resolution conversion is performed by generating new data between adjacent original image data by linear interpolation. Conversely, when the image data is higher than the print resolution, resolution conversion is performed by thinning out the data at a constant rate. Note that the resolution conversion processing is not essential in the present embodiment, and printing may be executed without performing such processing.

  Next, the CPU 81 performs color correction processing (step S110). The color correction process is a process of converting image data composed of R, G, B gradation values into data of gradation values of C, M, Y, K colors used in the printer 22. This processing is performed by using a color correction table LUT (see FIG. 7) that stores combinations of C, M, Y, and K for expressing the color composed of each combination of R, G, and B by the printer 22. . Various known techniques can be applied to the process of performing color correction using the color correction table LUT. For example, a process based on an interpolation operation (such as the technique described in Japanese Patent Laid-Open No. Hei 4-144482) can be applied.

  The CPU 81 performs multi-value processing on the image data that has been color corrected in this way (step S200). Multi-leveling means that the gradation value (16 gradations in this embodiment) of the original image data is converted into gradation values that can be expressed by the printer 22 for each pixel. As will be described later, in this embodiment, multi-value conversion to two gradation levels of dot on / off is performed, but multi-value conversion to more gradations may be performed. In the printing apparatus of this embodiment, multilevel processing is performed by the dither method.

  FIG. 14 shows the flow of multilevel processing by the dither method. When this process is started, the CPU 81 inputs the image data CD (step S202). The image data CD input here is data that has been subjected to color correction processing (step S110 in FIG. 13) and has 16 gradations for each color of C, M, Y, and K. Also, initialization is performed by substituting a value of 0 for each pixel number (kx, ky) constituting the image data (step S204). The image data is composed of nx pixels arranged two-dimensionally in the main scanning direction and ny pixels in the sub-scanning direction. The relationship between image data and pixels is shown in FIG. Each pixel has a pixel number kx given in the main scanning direction (x direction in FIG. 1) from the left side in FIG. 1, and a pixel number given in the sub scanning direction (y direction in FIG. 1) from the upper side in FIG. It is expressed using ky. With the above initialization, the upper left pixel shown in FIG. 1 is set as a pixel to start processing.

  Next, the CPU 81 calculates the component number (xt, yt) of the dither matrix used for multi-value conversion (step S206). The dither matrix is made up of mx two-dimensionally arranged components in the main scanning direction and my subordinate in the sub-scanning direction. Each component is represented by using integer component numbers (xt, yt) such that 0 ≦ xt <mx and 0 ≦ yt <my. As the dither matrix, various matrices can be applied. In this embodiment, a 4 × 4 Bayer-type dither matrix (see FIG. 2) is used. Therefore, mx = my = 4. In addition, for example, a blue noise mask type matrix (see FIG. 3) may be used. Various matrix sizes can be used.

In step S206, each component number is calculated by the following equation (2).
xt = kx% mx;
yt = ((kx div mx) × dy + ky)% my; (2)
Here,% is a remainder operator. For example, kx% mx means the remainder of kx / mx.
Further, (kx div mx) means a quotient of kx / mx.
dy is an integer of 1 or more and can be set to an arbitrary value. In this embodiment, dy = 1 was set.

  As will be described later, a threshold value corresponding to the component number obtained by the calculation of the above equation (2) is used for multi-valued pixel (kx, ky). FIG. 15 shows the correspondence between the dither matrix components and each pixel. For the sake of illustration, only a part of the image data is shown. In FIG. 15, the dither matrix components corresponding to the respective pixels are represented in the form of (xt, yt). The values from 0 to 9 given in the upper part of FIG. 15 mean pixel numbers in the main scanning direction, and the values from 0 to 5 given on the left side mean pixel numbers in the sub-scanning direction. .

For example, for the pixel (0, 0), the calculation result of the above equation (2) is
xt = 0% 4 = 0;
yt = ((0 div 4) × 1 + 0)% 4 = 0;
It becomes. Accordingly, the dither matrix component (0, 0) corresponds to the pixel (0, 0). For each pixel of kx ≦ 3 and ky ≦ 3, the components specified by xt = kx and yt = ky correspond to the same calculation, respectively.

Next, for the pixel (4, 0), the calculation result of the above equation (2) is
xt = 4% 4 = 0;
yt = ((4 div 4) × 1 + 0)% 4 = 1;
It becomes. Accordingly, the dither matrix component (0, 1) corresponds to the pixel (4, 0). Similarly, the following correspondence is obtained for the pixels and the dither matrix components. The right component corresponds to the left pixel of “→”.
Pixels (5,0) to (7,0) → components (1,1) to (3,1);
Pixels (4,1) to (7,1) → components (0,2) to (3,2);
Pixels (4, 2) to (7, 2) → components (0, 3) to (3, 3);

  Similarly, the correspondence between the image data and the dither matrix can be obtained for the other pixels. As shown in FIG. 15, these correspondences correspond to a state in which the dither matrix is made to correspond to the image data by shifting it upward by one pixel in a staircase pattern. The thick line in FIG. 15 shows the arrangement when the dither matrix is captured as one unit. In the present embodiment, since dy in the above equation (2) is set to a value 1, the correspondence is shifted upward by one pixel. That is, dy means the shift amount of the matrix. In this embodiment, the matrix is shifted only in the sub-scanning direction. However, the matrix may be shifted in the main scanning direction, or may be shifted in both directions. In this embodiment, the shift amount dy in the sub-scanning direction is also a constant value in the entire image area. However, the shift amount may be changed each time the matrix is repeatedly used in the main scanning direction or the sub-scanning direction. it can.

  When the component number of the dither matrix is set in this way, the CPU 81 compares the gradation value CD (kx, ky) of the image data with the threshold value DM (xt, yt) of the dither matrix (step S208). As a result, when the gradation value is larger than the threshold value, that is, when CD (kx, ky)> DM (xt, yt), it is determined that the dot should be turned on for the pixel, and the result value CDR ( A value indicating dot ON is substituted for kx, ky) (step S212). In the opposite case, it is determined that the dot should be turned off, and a value indicating dot OFF is substituted into the result value CDR (kx, ky) (step S210). The result value CDR is transferred to the drive buffer 47 (see FIG. 12) through a process such as rasterization described later, and determines on / off of each nozzle.

  With the above processing, dot on / off determination is performed for one pixel. Next, the CPU 81 increases the pixel number kx by a value 1 (step S214). That is, the pixel to be processed is shifted by one in the main scanning direction. Further, it is determined whether or not the pixel number kx set in this way is greater than or equal to the value nx (step S216). As described above, the image data has nx pixels in the main scanning direction, and the pixel number kx can take values from 0 to nx-1. If the pixel number kx is greater than or equal to the value nx, it means that the processing for one column of pixels arranged in the main scanning direction (hereinafter referred to as a raster) has been completed. Accordingly, the CPU 81 initializes the pixel number kx in the main scanning direction by substituting the value 0, and increases the pixel number ky in the sub scanning direction by the value 1 (step S218). This means that processing of the next raster is started.

  As described above, the image data has ny pixels in the sub-scanning direction, and the pixel number ky can take values from 0 to ny-1. If the pixel number ky is greater than or equal to the value ny, it means that the processing has been completed for all the pixels of the image data. Therefore, after increasing the pixel number ky in step S218, the CPU 81 compares the pixel number ky with the value ny (step S220). If ky is greater than or equal to the value ny, the multivalue processing is terminated. To do.

  On the other hand, if it is determined in step S216 that the pixel number kx is smaller than the value nx, and if it is determined in step S220 that the pixel number ky is smaller than the value ny, unprocessed image data remains. Therefore, the process returns to step S206 to execute dot on / off determination.

  An example of the result of the above processing is shown in FIG. FIG. 16 shows the result of multi-value processing for image data having a constant gradation value 8 using the Bayer matrix shown in FIG. The correspondence between the matrix and the image data is as shown in FIG. When the gradation value 8 is compared with the Bayer matrix shown in FIG. 2, (0,0), (2,0), (1,1), (3,1), (0,2), (2 , 2), (1, 3), (3, 3), the gradation value is larger than the threshold value set as the component. Accordingly, dots are turned on at pixels corresponding to such components. In FIG. 16, a hatched pixel means a pixel in which a dot is turned on. When viewed in a matrix for each matrix, the dots are formed in a checkered pattern, but the matrix does not correspond to the checkered pattern in the entire image data because the matrix is correlated with the image data in the sub-scanning direction.

  The CPU 81 rasterizes the result value CDR that has been multivalued by the above processing (step S300 in FIG. 13). This means that data for one raster is rearranged in the order of transfer to the head of the printer 22. There are various modes for the recording method in which the printer 22 forms a raster. The simplest is a mode in which all dots of each raster are formed by one forward movement of the head. In this case, data for one raster may be output to the head in the processed order. Another mode is so-called overlap. For example, in the first main scanning, for example, every other dot of each raster is formed, and the remaining dots are formed in the second main scanning. In this case, each raster is formed by two main scans. When such a recording method is employed, data obtained by picking up every other dot of each raster needs to be transferred to the head. Another recording mode is so-called bidirectional recording. This forms dots not only when the head moves forward but also during backward movement. When such a recording mode is employed, it is necessary to reverse the transfer order of the data for the forward movement and the data for the backward movement. The user can specify which mode is used for recording. The processing in step S240 described above creates data to be transferred to the head according to the recording method performed by the printer 22. When data that can be printed by the printer 22 is generated in this way, the CPU 81 outputs the data and transfers it to the printer 22 (step S310).

  As an example, FIG. 17 shows an example of dot recording by the overlap method. For convenience of illustration, the number of nozzles is reduced to six. The nozzle pitch in the sub-scanning direction is 2 dots. On the left side of FIG. 17, the position of the head in the sub-scanning direction is shown corresponding to the first to fifth main scans. Each nozzle position is indicated by “◯”. The numbers enclosed in circles indicate the nozzle numbers. The right side of FIG. 17 shows how dots are formed in each main scan.

  As shown in FIG. 17, in the first main scanning, odd-numbered pixels in the main scanning direction are formed intermittently. At this time, dots are formed only by the 5th and 6th nozzles, and the 1st to 4th nozzles do not form dots. In FIG. 17, dots actually formed are indicated by solid lines, and dots corresponding to the first to fourth nozzles are indicated by broken lines. The reason why these dots are not formed is that each raster cannot be formed completely due to the relationship with the sub-scan feed amount.

  When the first main scan is completed, sub-scanning is performed with a feed amount corresponding to 3 dots, and then dots are formed while performing the second main scanning. In this case as well, dots are formed using only the 4th to 6th nozzles for the reason described above. Further, after performing the sub-scan corresponding to 3 dots, dots are formed while performing the third main scan. The position of the No. 2 nozzle in the third main scan coincides with the position of the No. 5 nozzle in the first main scan. Accordingly, the second nozzle records pixels in which dots are not formed in the first main scan, that is, even-numbered pixels. Similarly, if main scanning is performed while performing sub-scanning for three dots, dots can be formed without gaps in the area indicated by PA in FIG. At this time, in each raster, odd-numbered pixels and even-numbered pixels are formed by different nozzles. By performing such recording, it is possible to disperse the dot formation position caused by the characteristics of each nozzle on each raster, thereby improving the image quality.

  According to the printing apparatus described above, it is possible to suppress the formation of dots in a regular pattern in units of matrix by causing the dither matrix to correspond to the image data by shifting from a grid shape to a step shape. it can. When the same dither matrix as that of the present embodiment is made to correspond to a grid pattern, dots are completely formed in a checkered pattern as shown in FIG. 28 for image data having a constant gradation value 8. On the other hand, according to the printing apparatus of the present embodiment, it can be seen that the regularity of dots is broken as shown in FIG.

  According to the printing apparatus of this embodiment, it is possible to suppress the deviation of the number of dots for each raster due to the loss of the regularity of dots. Therefore, when a nozzle that causes a shift in the dot formation position based on a mechanical manufacturing error is included, an increase in the number of dots formed by such a nozzle can be avoided. As a result, it is possible to avoid significant banding from occurring in the raster portion formed by the nozzle, and to improve the image quality. This effect can be obtained in the same manner in both cases where each raster is completed in one main scan when printing is performed by the overlap method.

  Further, according to the printing apparatus of this embodiment, the image quality can be improved even when recording is performed by the overlap method for the following reason. When dots are formed in a regular pattern as shown in FIG. 28, the pixels in which dots are formed tend to be biased to odd-numbered pixels or even-numbered pixels when viewed from each raster. When such a bias occurs, even if the recording by the overlap method is adopted, most of each raster is formed by a single nozzle, so that the effect of the overlap method is sufficiently obtained. Can't get. On the other hand, in the printing apparatus of the present embodiment, as a result of suppressing the regular pattern, it is possible to suppress the pixels on which dots are formed from being biased to odd or even numbers. Therefore, it is possible to sufficiently obtain the effect of improving the image quality by the overlap method.

  The above effect has been described by taking image data having a constant gradation value as an example. The above-described effects obtained by the printing apparatus according to the present embodiment are not limited to image data having a constant gradation value, but are similarly obtained for image data having various gradation values.

(3) Second embodiment:
Next, a printing apparatus as a second embodiment of the present invention will be described. The hardware configuration of the printing apparatus of the second embodiment is the same as that of the printing apparatus of the first embodiment (see FIGS. 6 to 12). The dot formation control process is the same as that in the first embodiment (see FIG. 13). In the second embodiment, the contents of the multilevel processing (step S200 in FIG. 13) are different from those in the first embodiment. In the second embodiment, multilevel conversion is performed using an error diffusion method.

  FIG. 18 shows the contents of the multi-value processing in the second embodiment, that is, the multi-value processing by the error diffusion method. When this process is started, the CPU 81 reads the image data CD (step S250), and initializes by substituting the value 0 for both pixel numbers (kx, ky) (step S252). These processing contents are the same as those in the multi-value processing (steps S202 and S204 in FIG. 14) in the first embodiment.

  The CPU 81 generates diffusion error correction data CDX based on the initialized image data of the pixel number (0, 0) (step S254). In the error diffusion process, an error in gradation expression generated for a processed pixel is distributed in advance with a predetermined weight applied to pixels around the pixel. In step S254, the corresponding error is read out and this error is now read. This is reflected in the pixel that is focused on to be processed. FIG. 19 exemplifies how much weight is distributed to which pixels in the periphery with respect to the pixel PP of interest. With respect to the pixel PP of interest, the density error is a predetermined weight (1/4, 1/8) for several pixels in the scanning direction of the carriage 31 and several adjacent pixels on the rear side in the transport direction of the paper P. , 1/16). The error diffusion process will be described in detail later.

  Next, the CPU 81 calculates a matrix component number (xt, yt) (step S256). The component number (xt, yt) is calculated using the equation (2) described in the first embodiment. The matrix size and shift amount dy used in this embodiment are the same as those in the first embodiment. Accordingly, the correspondence between the matrix and the image data is the same as that in the first embodiment. In other words, the correspondence is obtained by shifting the matrix from the grid to the sub-scanning direction by one pixel.

The matrix component DM (xt, yt) thus set is added as noise to the diffusion error correction data CDX (kx, ky) obtained in step S254. Assuming that the image data after the addition of noise is the noise addition data CDN (kx, ky), the following equation (3) is calculated. The reason for adding noise in this way will be described later.
CDN (kx, ky) = CDX (kx, ky) + DM (xt, yt) (3)

  The noise added data CDN thus generated is compared with a predetermined threshold value TH (step S260). If the data CDN is larger than the threshold value TH, the dot value is turned on in the result value CDR (kx, ky). Is substituted, and if the value is equal to or smaller than the threshold value TH, a value meaning dot off is substituted for the result value CDR (kx, ky) (step S240). The threshold value TH is a reference value for determining whether dots are turned on or off. The threshold value TH can be set to any value, but in this embodiment, the threshold value TH is set to a value 7 that is an intermediate value of 16 gradation values from 0 to 15 that the image data CD can take.

  With the above processing, the dot on / off state is determined for one pixel. Next, the CPU 81 calculates an error caused by the multi-value conversion, and executes a process of diffusing the error to surrounding pixels (step S266). The error means a value obtained by subtracting the gradation value of the original image data from the evaluation value of the density expressed by each dot after multi-value conversion. For example, a pixel having a gradation value of 15 in the original image data is considered, and an evaluation value of density due to dot formation is assumed to be equivalent to a gradation value of 15. The density evaluation value when no dots are formed corresponds to a gradation value of zero. If it is determined that a large dot is to be formed for this pixel, the density evaluation value expressed as the gradation value of the original image data is equal to the value 15, so no error occurs. On the other hand, if it is determined that no dot is formed, an error of Err = 0-15 = -15 is generated.

  The error thus calculated is diffused to surrounding pixels at the rate shown in FIG. For example, when an error corresponding to the gradation value 4 is calculated in the pixel PP of interest, an error corresponding to the gradation value 1 that is ¼ of the error is diffused to the adjacent pixel P1. It will be. Similarly, the error is diffused for other pixels at the ratio shown in FIG. The diffused error is reflected on the image data CD in step S254 described above, and diffusion error correction data CDX is generated.

  When the above processing is completed, the CPU 81 increases the pixel number kx by 1 (step S268), and compares the magnitude relationship with the number of pixels nx in the main scanning direction (step S270). If the pixel number kx is greater than or equal to the number of pixels nx, after substituting the value 0 into kx, the pixel number kx in the sub-scanning direction is increased by the value 1 (step S272), and the pixel number kx in the sub-scanning direction The magnitude relation is compared (step S274). The pixels to be processed are sequentially moved until the processing is completed for all the pixels by these processes. Such processing is the same as in the first embodiment. When the processing is completed for all the pixels in this way, the multi-value processing routine is terminated and the process returns to the dot formation control processing routine (FIG. 13). The subsequent processing is the same as in the first embodiment.

  In the multilevel processing described above, noise is added to the diffusion error correction data CDX in step S258. The error diffusion method can be multi-valued without adding noise. However, depending on the relationship between the threshold value TH and the dot density evaluation value, the dot generation rate may change abruptly in accordance with the change in gradation value. For example, in the case of image data having a uniform gradation value of 8, a large error occurs regardless of whether the dots are turned on or off, so that the determination of dot on / off is very unstable. It is easy to become. If the dot generation rate changes abruptly, a pseudo contour is generated there, and the image quality may deteriorate. In this embodiment, noise is added to avoid such a phenomenon. If noise is added to the image data, it is possible to avoid the existence of the above-described unstable gradation values over a wide range, and thus it is possible to suppress a sudden change in the dot generation rate.

  In the present embodiment, a matrix set so that the average value of each component value is 0 is used. This is because when the average value is a value other than 0, the density expressed as a whole changes. In this example, a matrix was used in which each component of the Bayer matrix was shifted so as to be within the range of -7 to 7. Of course, various matrices such as a blue noise mask type matrix can be used by shifting so that the average value of the thresholds becomes zero.

  According to the printing apparatus of the second embodiment described above, it is possible to perform multi-value conversion and printing with high image quality by adding noise based on the matrix. At this time, when the matrix is made to correspond to the image data in a grid pattern, the added noise is a repeating regular pattern, so that dots may be formed in a regular pattern. According to the printing apparatus of the second embodiment, since the matrix is shifted in correspondence with the image data in a stepwise manner, it is possible to avoid the formation of such a regular pattern and to perform printing with higher image quality. it can. Such an effect can be obtained both in the case of performing recording in which each raster is completed by one main scan and in the case of performing recording by an overlap method.

  In the printing apparatus of the second embodiment, the matrix is made to correspond to the sub-scanning direction while being shifted stepwise. However, the matrix may be shifted to the main scanning direction, or may be shifted to correspond to both. .

  Since the image processing apparatus and the printing apparatus described above include processing by a computer, it is possible to adopt an embodiment as a recording medium that records a program for realizing such processing. Such recording media include flexible disks, CD-ROMs, magneto-optical disks, IC cards, ROM cartridges, punch cards, printed matter on which codes such as bar codes are printed, computer internal storage devices (such as RAM and ROM). A variety of computer-readable media can be used, such as memory) and external storage devices. In addition, an aspect as a program supply apparatus that supplies a computer program for performing the above-described image processing or the like to a computer via a communication path is possible.

(4) Matrix design method:
In the image processing apparatus and the printing apparatus of each of the above-described embodiments, a Bayer-type matrix that is generally well-known is used for multi-value processing. Of course, it is also possible to use a blue noise mask type matrix. In addition to the known matrix, a matrix designed for the image processing apparatus and the printing apparatus can also be used.

  FIG. 20 is a flowchart showing the procedure of a matrix design method as an embodiment of the present invention. 21 to 24 show specific design examples. The matrix design method in this embodiment will be described with reference to these drawings.

  First, the matrix size is set, and the threshold range stored in the matrix is set (step S10). The matrix size is set according to the gradation value to be expressed by multi-value conversion and the memory capacity for storing the matrix. Further, the threshold value is set according to the gradation value to be expressed by multi-value. In this embodiment, the matrix size is set to 4 × 4, and the threshold value is set to a range from 0 to 15. In this embodiment, the threshold value is a continuous integer in the range of 0 to 15, but a discrete threshold value may be stored. In addition, overlapping threshold values may exist.

  Next, the arrangement of the matrix is set (step S12). Matrix arrangement refers to the correspondence of the matrix to the image data. The matrix may be arranged in a grid pattern with respect to the image data, or may be arranged in a staircase pattern. In this embodiment, as will be described later, it is arranged in a staircase pattern.

  Next, the threshold values 0 and 1 are assigned to the components in the matrix (step S14). FIG. 21 shows assignment of threshold values in the present embodiment. The value 0 was assigned to the component (0,0), and the value 1 was assigned to the component (2,1). Two threshold values are assigned from the smaller of all the threshold values arranged in the matrix. Three or more threshold values may be assigned from the smallest one, or only the smallest threshold value may be assigned. The threshold value can be assigned to an arbitrary component, but is preferably assigned in consideration of dot dispersibility.

  After setting some threshold values in this way, the following processing is repeatedly executed, and the remaining threshold values are set by calculation. The threshold is set in order from the smallest. As described above, since the threshold values 0 and 1 are already set, the threshold value 2 is set next. FIG. 20 shows a step of substituting the value 2 for the threshold value nc to be set in this sense.

  Next, one component for which a threshold is not yet set in the matrix is selected (step S18), and an evaluation value of the distance from this component to the component for which a threshold has already been set is calculated (step S20). In this example, the component (3, 0) was first selected. The position of this component on the image data is shown in FIG. FIG. 22 shows the correspondence between the matrix and the image data in this embodiment. The component (3, 0) corresponds to the pixel DD in FIG. Of course, since the matrix is arranged stepwise in FIG. 22, there are many pixels corresponding to the component (3, 0), but one of them may be selected.

  Since there are already two types of threshold values 0 and 1, a distance evaluation value is calculated for each. The distance evaluation value is calculated through a step of calculating a distance from a pixel for which a threshold has already been set as described later, and a step of obtaining a distance evaluation value based on the distance.

  The distance is calculated as follows. Around the pixel DD whose threshold value is not defined, there are many pixels to which the threshold value 0 is assigned according to the arrangement of the matrix. Thus, the distance is calculated by selecting the pixel with the shortest distance from the large number of pixels. In FIG. 22, the pixel adjacent to the upper right of the pixel DD is the closest pixel. The distance between the two is indicated by di0 in FIG. 22, and is obtained by √ (dx · dx + dy · dy) using the distance dx in the main scanning direction and the distance dy in the sub scanning direction. In the case of FIG. 22, di0 = √2. Similarly, the distance between the pixel to which the threshold value 1 is assigned and the pixel DD is also calculated. In this embodiment, the pixel adjacent to the lower left of the pixel DD is the closest pixel. If the distance to this pixel is di1, then di1 = √2. At this stage, since only two threshold values 0 and 1 are set, two distances di0 and di1 are calculated and the process proceeds to the next step. When the set threshold value increases, the distance corresponding to each threshold value is calculated. As described above, the distance from the pixel to which the threshold is assigned is calculated in consideration of the arrangement of the matrix in the image data. The reason for this will be described later.

Based on the distance thus calculated, an evaluation value evl of the distance is obtained by the following equation (4). Here, “^” is a power operator.
evl = 1 / di0 ^ 2 + 1 / di1 ^ 2 (4)
That is, the sum of 1 / square of each distance is set as the evaluation value evl. When the set threshold value increases and the distance is calculated as di2, di3,..., Values obtained by adding 1 / di2 ^ 2, 1 / di3 ^ 2. As described above, since the distance di0 = di1 = √2 for the component (3, 0), evl = 1. The above calculation is performed for all components to which no threshold is assigned (step S22). The evaluation values thus calculated are shown in FIG.

  Next, the component having the smallest evaluation value is selected (step S24). As is clear from FIG. 23, in this example, the component (2, 3) is the minimum with an evaluation value of 0.45. Accordingly, a threshold nc is set for this component. Since nc = 2 at this stage, the threshold value 2 is assigned to the component (2, 3). As is clear from the above equation (4), the component with the smallest evaluation value evl corresponds to a pixel that is far from any pixel for which a threshold has already been set. Therefore, if multi-value processing is performed using a matrix designed by assigning threshold values to such components, it is possible to ensure the dispersibility of dots in the entire image data. In this embodiment, the distance between a pixel for which a threshold is already set and an undefined pixel is calculated in consideration of the arrangement of the matrix on the image data. The reason for this is to ensure the dispersibility of dots in the entire image data.

  Since the threshold values 0, 1, and 2 are assigned to the components by the above processing, the next threshold value is set as a setting target. In the present embodiment, the threshold value nc is increased by 1 (step S26). The processing from step S18 to S24 described above is performed for this threshold value, and the threshold value nc is assigned to any component of the matrix. This process is repeatedly executed until threshold values are assigned to all components (step S28).

  FIG. 24 shows the result of assigning the threshold values in this way. This is an example of a matrix designed by the matrix designing method of the present invention. The matrix shown in FIG. 24 is a different matrix from the Bayer type and the blue noise mask type. The matrix obtained in this way may be corrected according to the relationship with the image data to be processed. For example, when the image data has gradation values from 0 to 255, correction may be performed by multiplying each threshold value by a factor so that the matrix in FIG. Further, when used as a noise matrix in the multi-value processing shown in FIG. 18, correction may be performed to shift the value of each component so that the average value becomes zero. In addition, various corrections can be made according to the relationship with image data.

  According to the matrix design method described above, it is possible to obtain a matrix with high dot dispersibility in consideration of the arrangement of the matrix with respect to the image data. If multi-value conversion is performed using the matrix thus designed, high-quality multi-value conversion with high dot dispersibility in the entire image data can be achieved. Further, according to the above design method, a matrix can be obtained on the premise that the matrix is arranged in a staircase pattern as shown in FIG. As a result, it is possible to obtain a matrix most suitable for the image processing apparatus and printing apparatus of the present embodiment described above.

  Although various embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various embodiments can be implemented without departing from the spirit of the present invention. For example, some or all of the various control processes described in the above embodiments may be realized by hardware.

It is explanatory drawing which shows the correspondence of image data and a matrix. It is explanatory drawing which shows the content of a Bayer type | mold matrix. It is explanatory drawing which shows the example of a blue noise mask type | mold matrix. It is explanatory drawing which shows arrangement | positioning of the threshold value to a matrix. It is explanatory drawing shown about calculation of the distance with each component. It is a schematic block diagram of the printing apparatus of this invention. It is explanatory drawing which shows the structure of software. 1 is a schematic configuration diagram of a printer of the present invention. It is explanatory drawing which shows schematic structure of the dot recording head of the printer of this invention. It is explanatory drawing which shows the dot formation principle in the printer of this invention. FIG. 3 is an explanatory diagram illustrating an internal configuration of a printer control device. It is explanatory drawing which shows the mode of the dot formed with the drive waveform of the nozzle in the printer of this invention, and this drive waveform. It is a flowchart which shows the flow of a dot formation control routine. It is a flowchart which shows the flow of the multi-value process by a dither method. It is explanatory drawing which shows a response | compatibility with the matrix and image data in an Example. It is explanatory drawing which shows the formation result of the dot by an Example. It is explanatory drawing which shows the mode of the formation of the dot by an overlap system. It is a flowchart which shows the flow of the multi-value process by an error diffusion method. It is explanatory drawing which shows the weight at the time of diffusing an error. It is a flowchart which shows the procedure of the design method of an Example. It is explanatory drawing shown about the setting of the threshold value in the design method of an Example. It is explanatory drawing shown about calculation of the distance in the design method of an Example. It is explanatory drawing which shows the evaluation value of the distance in the design method of an Example. It is explanatory drawing which shows the matrix designed by the design method of the Example. It is explanatory drawing which shows a Bayer type matrix. It is explanatory drawing which shows a response | compatibility with the matrix and image data in a prior art. It is explanatory drawing which shows the view of the dot on / off determination by the dither method. It is explanatory drawing which shows the formation result of the dot in a prior art.

Explanation of symbols

12 ... Scanner 14 ... Keyboard 15 ... Flexible drive 16 ... Hard disk 18 ... Modem 21 ... Color display 22 ... Color printer 23 ... Paper feed motor 24 ... Carriage Motor 26 ... Platen 28 ... Print head 31 ... Carriage 32 ... Operation panel 34 ... Sliding shaft 36 ... Drive belt 38 ... Pulley 39 ... Position detection sensor 40. ..Control circuit 41 ... CPU
42 ... Programmable ROM (PROM)
43 ... RAM
44 ... PC interface 45 ... Peripheral input / output unit (PIO)
46 ... Timer 47 ... Transfer buffer 48 ... Bus 51 ... Transmitter 55 ... Distribution output device 61, 62, 63, 64 ... Ink ejection head 67 ... Introducing pipe 68 ... Ink passage 71 ... Black ink cartridge 72 ... Color ink cartridge 80 ... Bus 81 ... CPU
82 ... ROM
83 ... RAM
84 ... Input interface 85 ... Output interface 86 ... CRTC
87 ... Disk controller (DDC)
88 ... Serial I / O interface (SIO)
90 ... Personal computer 91 ... Video driver 95 ... Application program 96 ... Printer driver 97 ... Resolution conversion module 98 ... Color correction module 99 ... Halftone module 100 ... Rasterizer

Claims (14)

Depending on the magnitude relationship between the gradation value and a predetermined threshold value for each pixel, the image data is composed of a two-dimensional array of pixels and has gradation values in a predetermined range for each pixel. An image processing apparatus that determines dot on / off and multi-values the image data,
Storage means for storing a two-dimensional matrix having a predetermined value as a component in a size smaller than the size of the image data constituted by the two-dimensional pixel array;
A multi-value conversion is performed by reflecting the matrix component in the gradation value or the threshold value of the image data in a correspondence relationship in which the matrix is arranged with respect to the image data in a state shifted from a grid. An image processing apparatus comprising a value conversion unit. Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, for each of the pixels, an image processing apparatus that determines dot on / off according to the magnitude relationship between the gradation value and a predetermined threshold value, and multi-values the image data,
Storage means for storing a matrix of mx × my (an integer satisfying 1 ≦ mx <nx, 1 ≦ my <ny) having a preset value as a component;
Multi-value conversion means for performing multi-value conversion by reflecting the components of the matrix in the gradation value or the threshold value for each pixel,
The matrix components to be reflected in the kxth pixel in the x direction (kx is an integer 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer 0 ≦ ky ≦ ny) are given by the following equations ( xt, yt) (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my).
xt = (rx · dx1 + ry · dx2 + kx)% mx;
yt = (rx · dy1 + ry · dy2 + ky)% my;
rx = kx div mx;
ry = ky div my;
here,
a div b is an operator for calculating the quotient of a / b by an integer value;
a% b is a remainder operator for calculating the remainder of a / b,
dx1, dx2, dy1, dy2 are integers of 0 or more, and at least one of dx1, dx2, dy1, dy2 is an integer other than 0. The image processing apparatus according to claim 1 or 2,
The multi-value conversion means is an image processing apparatus which performs multi-value conversion by a dither method using a component of the matrix as the threshold value. The image processing apparatus according to claim 3,
The image processing apparatus, wherein the matrix is a dot-dispersed dither matrix capable of multi-valued dot dispersion with high dispersibility within an image area in units of the matrix. The image processing apparatus according to claim 4,
The image processing apparatus is a Bayer type dither matrix. The image processing apparatus according to claim 4,
The matrix is an image processing apparatus which is a blue noise mask type matrix. The image processing apparatus according to claim 4,
The image processing apparatus is an image processing apparatus in which the matrix is an image area wider than an area corresponding to the size of the matrix and is capable of performing multi-value conversion with high dot dispersibility. The image processing apparatus according to claim 1 or 2,
The multi-value conversion means is an image processing apparatus that performs multi-value conversion after reflecting the matrix components as noise data in the image data. Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, the image is formed on the print medium by driving the head to form dots according to the multi-value quantization result determined according to the magnitude relationship between the gradation value and the predetermined threshold value for each pixel. A printing device for printing
Storage means for storing a matrix of mx × my (an integer satisfying 1 ≦ mx <nx, 1 ≦ my <ny) having a preset value as a component;
Multi-value conversion means for performing multi-value conversion by reflecting any component of the matrix in either one of the gradation value and the threshold value for each pixel,
The matrix components to be reflected in the kxth pixel in the x direction (kx is an integer 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer 0 ≦ ky <ny) are given by the following equations ( xt, yt) (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my).
xt = (rx · dx1 + ry · dx2 + kx)% mx;
yt = (rx · dy1 + ry · dy2 + ky)% my;
rx = kx div mx;
ry = ky div my;
here,
a div b is an operator for calculating the quotient of a / b by an integer value;
a% b is a remainder operator for calculating the remainder of a / b,
dx1, dx2, dy1, dy2 are integers of 0 or more, and at least one of dx1, dx2, dy1, dy2 is an integer other than 0. The printing apparatus according to claim 9, wherein
The head includes a plurality of nozzles arranged in the y direction;
Main scanning means for reciprocating the head in the x direction relative to the print medium;
Sub-scanning means for relatively moving the head and the print medium in the y direction;
Drive control means for controlling the driving of the main scanning means, the sub-scanning means and the head to form each dot row arranged in the x direction using two or more nozzles,
A printing apparatus in which at least dy1 ≠ 0. Depending on the magnitude relationship between the gradation value and a predetermined threshold value for each pixel, the image data is composed of a two-dimensional array of pixels and has gradation values in a predetermined range for each pixel. An image processing method for determining on / off of dots and performing multi-value conversion of the image data,
A two-dimensional matrix having a preset value as a component is reflected in the gradation value or the threshold value of the image data in a correspondence relationship in which the two-dimensional matrix is arranged with respect to the image data in a state shifted from a grid. An image processing method for performing multi-value processing. Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, when performing multi-value determination that determines dot on / off according to the magnitude relationship between the gradation value and a predetermined threshold value for each pixel, the gradation value or the value to be reflected in the threshold value Mx × my (1 ≦ mx <nx, 1 ≦ my <ny integer) matrix design method,
(A) setting a correspondence between each pixel in the multi-value quantization and the matrix component;
(B) n values (n is an integer of 1 or more) from the top when values assigned to the matrix are arranged in accordance with the magnitude relationship, as an arbitrary component of the matrix in consideration of threshold dispersion A setting process;
(C) sequentially setting values remaining as values to be assigned to the matrix to components obtained by a predetermined calculation,
The step (c)
(C-1) A distance between each component of the matrix to which no value is assigned and a component of the matrix to which a value has already been set is assigned to a plurality of matrices associated with each pixel constituting the image data. A process of evaluating over,
(C-2) From the step of setting the first value when the remaining values are arranged according to the magnitude relationship to the component evaluated to be the longest distance from the component for which the value has already been set The matrix design method. A method for designing a matrix according to claim 12, comprising:
The correspondence in the step (a) is as follows:
The matrix component to be reflected in the kxth pixel in the x direction (kx is an integer satisfying 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer satisfying 0 ≦ ky ≦ ny) is given by the following equation ( xt, yt) (0 ≦ xt ≦ mx, 0 ≦ yt ≦ my).
xt = (rx · dx1 + ry · dx2 + kx)% mx;
yt = (rx · dy1 + ry · dy2 + ky)% my;
rx = kx div mx;
ry = ky div my;
here,
a div b is an operator for calculating the quotient of a / b by an integer value;
a% b is a remainder operator for calculating the remainder of a / b,
dx1, dx2, dy1, dy2 are integers of 0 or more, and at least one of dx1, dx2, dy1, dy2 is an integer other than 0. Image data having a total of nx × ny pixels composed of nx pixels in the x direction and ny pixels in the y direction (nx and ny are integers of 2 or more) and having a gradation value in a predetermined range for each pixel. On the other hand, the present invention is a recording medium on which a computer readable recording program for performing multi-value determination for determining on / off of dots according to the magnitude relationship between the gradation value and a predetermined threshold value for each pixel is provided. And
A matrix of mx × my (an integer satisfying 1 ≦ mx <nx, 1 ≦ my <ny) in which preset values are stored as components;
Of the matrix, the gradation value or the threshold value of the kxth pixel in the x direction (kx is an integer satisfying 0 ≦ kx ≦ nx) and the kyth pixel in the y direction (ky is an integer satisfying 0 ≦ ky ≦ ny) A recording medium on which a program for realizing the function of performing multi-value conversion by reflecting a component (xt, yt) (an integer of 0 ≦ xt ≦ mx, 0 ≦ yt ≦ my) given by the following equation is recorded.
xt = (rx · dx1 + ry · dx2 + kx)% mx;
yt = (rx · dy1 + ry · dy2 + ky)% my;
rx = kx div mx;
ry = ky div my;
here,
a div b is an operator for calculating the quotient of a / b by an integer value;
a% b is a remainder operator for calculating the remainder of a / b,
dx1, dx2, dy1, dy2 are integers of 0 or more, and at least one of dx1, dx2, dy1, dy2 is an integer other than 0.

Images