JP7336255B2 - Image processing device and its control method and program - Google Patents

Description

The present invention relates to a technique for converting a multi-tone image into an image with a lower number of gradations than the image.

2. Description of the Related Art Conventionally, an inkjet recording apparatus is known as an output device for computers. An inkjet recording apparatus has a recording head in which a plurality of ink ejection openings (nozzles) are arranged. The recording head is moved relative to a recording medium, and ink droplets (dots) are ejected from the nozzles. A desired image is formed on a recording medium.

In particular, inkjet printers for commercial printing are required to have high printing productivity, and a single-pass drawing method (also called a full-line method or a full-multi method) is often adopted. In the single-pass drawing method, drawing is performed in the recording medium width direction (hereinafter referred to as the "recording medium width direction" or "y direction") perpendicular to the recording medium conveyance direction (hereinafter referred to as the "conveyance direction" or "x direction"). A configuration is adopted in which a long line head having a nozzle array covering the entire area is used, and the line head extends in the direction perpendicular to the conveying direction of the recording medium. Since it is sufficient to move the recording medium relative to the line head only once, the single-pass drawing method is characterized in that the printing speed is faster than that of the multi-pass method in which an image is completed by a plurality of scans.

When printing with such a printer, so-called halftone processing, ie, image processing that expresses the gradation values of the original image data to be printed by turning dots on and off, is performed. Then, the printer executes print processing according to the image data after this halftone processing.

Various methods have been proposed for halftone processing, and one of the representative methods is the dither method. The dither method uses a dither matrix of a predetermined size in which different threshold values are arranged. Specifically, the dither matrix is repeatedly developed in tiles on the image data, and the gradation values (pixel values) of the input image are compared with the corresponding threshold values. Then, if the gradation value is greater than the threshold, dots are turned on, and if the gradation value is equal to or less than the threshold, the dots are turned off to perform gradation expression.

With such a dither method, the number of dots ejected from each nozzle (ejection frequency) may be uneven in the dot pattern generated by the dither matrix. There is a problem that the life of the device is shortened.

Japanese Patent Application Laid-Open No. 2002-200002 proposes a method of developing a dither matrix by shifting it in the width direction of a recording medium, instead of developing it in a tile shape as in the conventional art, as an image recording apparatus for solving this problem. As a result, it is possible to suppress unevenness in the number of ejected dots for each nozzle.

JP 2005-161817 A

However, in the case of using a so-called dot-dispersed dither matrix in which the spatial frequency distribution is biased to a certain frequency or more that is difficult for humans to perceive to achieve high image quality, the method of Patent Document 1 does not allow sufficient dots to be discharged from each nozzle. In some cases, it may not be possible to suppress the bias in numbers.

Examples of dot-dispersed dither matrices include a blue noise dither matrix having blue noise characteristics and a green noise dither matrix having green noise characteristics. The blue noise dither matrix is a dither matrix in which threshold storage positions are adjusted so that dots are dispersed in consideration of human visual characteristics. As a result, the spatial frequency of the dot arrangement is biased toward high frequencies, and graininess can be suppressed. Also, the green noise dither matrix is a dither matrix in which threshold storage positions are adjusted so that dots are formed adjacent to each other in units of several dots, but clusters of dots are dispersed as a whole. Printers that have difficulty in stably forming fine dots of about one pixel can suppress the occurrence of isolated dots by referring to such a green noise matrix to determine the presence or absence of dot formation.

A dither matrix of a relatively large size is used in such a dot-dispersed dither matrix. For example, when using a dither matrix with a size of 256 pixels in the transport direction of the print medium and 256 pixels in the direction perpendicular to the transport direction of the print medium, a width of 65536 pixels in the transport direction is required to completely uniform the nozzle usage rate. (Development by shifting a matrix of 256-pixel width in the transport direction 256 times in the direction of the recording medium) is required. On the other hand, when the image object of the original image data to be drawn is, for example, 20 mm (1200 dpi) in the transport direction, the size of the image object is about 945 pixels. As a result, dots are generated with a partial threshold value of the developed dither matrix, and it is not possible to suppress the bias in the nozzle usage rate. That is, when the image object of the original image data is relatively small with respect to the dither matrix developed by being shifted in the width direction of the recording medium, there is a problem that the number of dots ejected from each nozzle cannot be made sufficiently uniform. Ta.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a technique for suppressing unevenness in the number of ejected dots for each printing element such as a nozzle in halftone processing using a dither matrix.

In order to solve this problem, for example, the image processing apparatus of the present invention has the following configuration. i.e.
a recording head having a plurality of recording elements arranged in a first direction, the recording head directed toward a recording medium relatively conveyed along a second direction perpendicular to the first direction; an image forming apparatus for forming an image on a recording medium by means of recording elements of an image processing apparatus for acquiring correction information for correcting density unevenness on the image caused by characteristics of each of the recording elements; ,
acquisition means for acquiring image data having the same gradation value at each position in the first direction and for acquiring the correction information ;
A distributed dither matrix having a size Sx (Sx is a natural number) in the second direction is repeatedly developed in the second direction , and the gradation value of the image data and the corresponding threshold value are compared in size. a converting means for converting into halftone image data representing a halftone pattern by
generating means for generating correction information for correcting density unevenness caused by the characteristics of each of the recording elements by reading a chart image formed based on the halftone image data ;
In the converted halftone image data, the difference in sum of pixel values included in the range of Sx in the second direction between each position in the first direction is calculated by the image forming apparatus. The number of tones that can be expressed is less than or equal to
the size Tx in the second direction of the area having the same gradation value in the image data for acquiring the correction information is greater than or equal to Sx;
The generating means generates measurement data obtained by accumulating read signal values of a print width on the recording medium corresponding to a positive integral multiple of the size Sx of the dither matrix in the second direction in the read image of the chart image. is characterized by generating a correction table based on

According to the present invention, it is possible to suppress unevenness in the number of ejected dots for each printing element such as a nozzle.

1 is a block diagram showing configurations of an image processing apparatus and an image forming apparatus according to a first embodiment; FIG. 4 is a block diagram showing a halftone processing unit according to the first embodiment; FIG. 4A and 4B are diagrams for explaining a print head according to the first embodiment; FIG. 4A and 4B are diagrams for explaining halftone processing according to the first embodiment; FIG. 4 is a flowchart showing a procedure for creating a ditherrix applied to the first embodiment; 4 is a flow chart showing the procedure of an initial pattern in the creation of dithertrix. 4A and 4B are diagrams showing characteristics of dot patterns generated by dithertrix applied to the first embodiment; FIG. FIG. 2 is a block diagram showing the configurations of an image processing apparatus and an image forming apparatus according to a second embodiment; FIG. 11 is a diagram showing an example of a density measurement test chart according to the second embodiment; 9 is a flowchart showing a procedure for creating density unevenness correction parameters according to the second embodiment;

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Note that the configurations in the embodiments shown below are merely examples, and the present invention is not limited to the illustrated configurations.

[First embodiment]
FIG. 1 is a block diagram showing configurations of an image processing apparatus and an image forming apparatus applicable to the first embodiment. In FIG. 1, an image processing apparatus 10 and an image forming apparatus 20 are connected by an interface or circuit. The image processing apparatus 10 is, for example, a general personal computer in which a printer driver is installed. In that case, each unit in the image processing apparatus 10 described below is implemented by a computer executing a predetermined program. However, the image forming apparatus 20 may include the image processing apparatus 10 .

The image processing apparatus 10 stores image data to be output input from the image data input unit 101 in the input image buffer 102 . The format of the image data input to the image data input unit 101 is not particularly limited. is a gradation image. For example, if the image forming apparatus 20 is an inkjet printing system that achieves an output resolution of 1200 dpi using four color inks of cyan (C), magenta (M), yellow (Y), and black (K), the image data is CMYK. is multivalued image data having 8 bits (256 gradations) for each color. There may be various formats of image data of an image to be printed. When printing image data specified by a combination of colors and a format of resolution different from the type of ink color and resolution used in the image forming apparatus 20, a preprocessing unit (not shown) precedes the image data input unit 101. , color conversion, resolution conversion, and the like to convert the image data into image data having the ink color and resolution used in the print image forming apparatus 20, and then input from the image data input unit 101. FIG.

The halftone processing unit 103 uses a dither matrix for the image data stored in the input image buffer 102 to perform conversion (halftone) processing to the number of gradations that can be directly expressed by the image forming apparatus 20 and nozzle group processing. determines the dot arrangement to be formed, and creates halftone image data. The dither matrix has threshold groups arranged according to the number of gradations that can be simulated. The number of values used as thresholds in the dither matrix determines the number of gradations that can be simulated. A detailed method of generating the dither matrix will be described later. The halftone processing unit 103 outputs the created halftone image data to the halftone image buffer 104 . The stored halftone image data is output from the output terminal 105 to the image forming apparatus 20 .

Based on the halftone image data received from the image processing apparatus 10 via the input terminal 201 , the image forming apparatus 20 moves a recording medium such as a recording sheet relative to the recording head 203 while moving the recording medium. An image is formed on a recording medium by creating dots thereon. Here, the print head 203 is of an ink jet printing type, and has a print element array in which a plurality of print elements capable of ejecting ink are arranged. FIG. 3 is a diagram showing a configuration example of the print head 203. As shown in FIG. The recording head typically has nozzles for four types of ink: cyan (C), magenta (M), yellow (Y), and black (K). are shown. The nozzle array is a long line head that covers the entire drawing area in the recording medium width direction (x direction) perpendicular to the recording medium conveying direction (y direction). A printed image is formed by ejecting the . A head driving unit 202 generates a driving signal for controlling the recording head 203 based on the halftone image data. The print head 203 actually prints each ink dot on the print medium based on the drive signal. The print elements of the cyan (C), magenta (M), and yellow (Y) print heads are also arranged in parallel with the black head.

The halftone processing unit 103 in this embodiment will be described in detail below.

FIG. 2 is a block diagram showing the detailed configuration of the halftone processing unit 103. As shown in FIG. The halftone processing unit 103 converts the image data stored in the input image buffer 102 into binary (1-bit) data for each nozzle group and outputs the data. For this processing, the halftone processing unit 103 has a comparator 1031 , a dither matrix 1032 (composed of ROM or the like), and an output data generator 1033 .

Here, the input image data stored in the input buffer 102 is defined as Iin, and the number of pixels in the x direction (here, the print medium conveying direction) is W, and the y direction (with respect to the print medium conveying direction) is Let H be the number of pixels in the orthogonal direction (hereinafter simply referred to as the width direction), and coordinates (x, y) (0≤x≤W-1, 0≤y≤H-1) of the input image Iin. The pixel value of is denoted by Iin (x, y), and the dither matrix 1032 is assumed to have Sx rows (size in the x direction) and Sy columns (size in the y direction), and coordinates in the dither matrix The threshold value of (i, j) is expressed as M(i, j), the output image data (binary image data) generated by the output data generator 1033 is defined as Iout, and the coordinates (x, y) are defined as Assume that the output value is Iout(x, y), and the remainder obtained by dividing the integer "a" by the integer "b" is represented as "a % b".

The comparator 1031 determines the magnitude of the pixel value Iin (x, y) of the coordinates (x, y) of the input image data Iin from the input image buffer 102 and the dither matrix threshold value M (x % Sx, y % Sy). and outputs the determination result (comparison result) to the output data generator 1033 . The output data generator 1033 generates, as the output value Iout (x, y), "BLACK" indicating that a black dot is to be printed when Iin (x, y) is greater than the threshold value M (x % Sx, y % Sy). do. In addition, the output data generator 1033 outputs "WHITE" indicating that dots are not printed when Iin(x, y) is equal to or less than the threshold value M(x%Sx, y%Sy) as the output value Iout(x,y). ). Note that, in the following description, dot printing/non-printing is also expressed as "turning dots ON/OFF".

The processing of the halftone processing unit 103 is summarized as follows.
Iout(x,y)=BLACK if Iin(x,y) > M(x % Sx, y % Sy)
Iout(x,y)=WHITE if Iin(x,y) ≤ M(x % Sx, y % Sy)
Note that the output data generator 1033 actually generates a binary image in which the output value “BLACK” is set to, for example, one bit “1” and “WHITE” is set to “0”.

An example of the above halftone processing will now be described with reference to FIG. Reference numeral 401 in FIG. 4 indicates part of the input image data, reference numeral 402 indicates part of the dither matrix M, and reference numeral 403 indicates part of the output data. In the input image data 401, one pixel is represented by, for example, 8 bits, and the range of possible values of each pixel is 0-255.

The dither matrix 402 has a total of 65536 thresholds, 256 in the x direction and 256 in the y direction, and these thresholds are uniformly selected from the range of values 0 to 255. .

Here, when the pixel value of the input image data exceeds the corresponding threshold value, BLACK (black dot is printed) is determined as the output value, and when the pixel value is below the corresponding threshold value, WHITE (dot is not printed) is determined. ) is determined as the output value. Therefore, when the illustrated dither matrix 402 is used, a dot arrangement of 256 gradations can be obtained. That is, when halftone processing is performed using the dither matrix 402, 256 levels of gradation from 0 to 255 can be simulated as output data in a range of 256 pixels×256 pixels. In the illustration, an input image 401 is an image having a single pixel value of "33", and when a dither matrix 402 is used, output data 403 as shown is obtained.

As can be seen from the above, it can be said that the pixel positions with a small threshold value of the dither matrix have a high probability of turning on the dots, and conversely, the pixel positions with a large threshold value have a high probability of turning off the dots. Therefore, if the relatively small threshold value of the dither matrix is biased toward a specific nozzle row, the nozzle will frequently eject ink, shortening the life of the nozzle.

In order to make the print head usable as long as possible, it is preferable to suppress the occurrence of nozzles ejecting an extremely large number of ink by eliminating bias in the number of ink dots ejected from each nozzle and making the number of dots uniform. For this purpose, it can be said that it is effective to adjust the threshold value arrangement of the dither matrix so that the frequency of generating dots formed by each nozzle with respect to the same input tone value becomes uniform.

Therefore, a method of creating a dot dispersion type dither matrix that is adjusted so that the frequency of generating dots formed by each nozzle will be uniform will be described below. In the following description, it is assumed that M is a dither matrix in the process of generation or after generation. The dither matrix M is a two-dimensional array in which Sx is the number of thresholds (size) arranged in the x-direction (printing medium transport direction), and Sy is the number of thresholds arranged in the y-direction (width direction of the printing medium). , Sy are natural numbers. Although the size (Sx, Sy) of the dither matrix is arbitrary, it is preferable that it is a quadrangle having a side length that is a power of 2 and each side is 256 or more (for example, 256×256 or 512×512). be. In this embodiment, Sx is 256 and Sy is 256. The dither matrix M created in this embodiment stores threshold values of 0 to 65535. FIG. As a result, halftone processing using the dither matrix M can reproduce 256 gradations in a pseudo manner in the area of Sx×Sy.

A Void & Cluster method is known as a method capable of generating a dot-dispersed dither matrix. The Void & Cluster method applies a low-pass filter to obtain a smoothed density image, and determines the placement of dots to be added/deleted so as to suppress local density fluctuations. In the present embodiment, a similar method is used to generate a dither matrix having dot dispersion type blue noise characteristics.

Let d(x, y) be the dot pattern generated in this process. d(x, y) is a two-dimensional array with the same size as M(x, y). The value of each pixel of d(x, y) is a binary representation of "1" when a dot exists and "0" when a dot does not exist.

The dot pattern d(x,y) changes in the iterative process. According to the iterative process, Sx×Sy+1 dot patterns are generated from a dot pattern with 0 dots to Sx×Sy dots. Therefore, if g is the number of dots in the dot pattern d(x, y), it is possible to identify a point in the iterative process by using g. In the following description, the number of dots g will be referred to as the gradation value g. In the following description, the dot pattern d(x, y) when the gradation value is g is also written as d(g, x, y), or d(g) by omitting x and y. . As described above, in this embodiment, in order to create a dither matrix capable of pseudo-reproducing 256 gradations, all natural numbers from 0 to Sx×Sy−1 are set for the gradation g in the following processing. be.

The details of this embodiment will be described below with reference to FIG. This figure is a flowchart showing the overall flow of the dither matrix generation method, and shows the dither matrix generation processing used by the halftone processing unit 103 by the CPU (not shown) of the image processing apparatus 10 of FIG. Note that if the dither matrix is stored in a non-volatile memory (not shown) so that it can be reused in the next printing process, it should be executed only at the first time. When the image processing apparatus 10 is an information processing apparatus such as a personal computer, the program related to the illustrated flowchart indicates part of an application or printer driver.

At S100, the CPU generates an initial dot pattern d(g0) having a predetermined gradation value g0. In this embodiment, the number of gradations g0 of the initial dot pattern is Sx*Sx*0.5=256*256*0.5=32768. A method of generating the initial dot pattern will be described later. Note that the initial dot pattern d(g0) can also be an image with the number of gradations g0=0 in which there is no dot. In that case, S100 is not performed.

Steps S101 to S107 are processes for repeating dot addition starting from the initial dot pattern. Further, steps S201 to S207 are processes for repeatedly deleting dots starting from the initial dot pattern. Dot deletion is not performed when dot addition is repeated. Conversely, when repeating dot deletion, dot addition is not performed.

S101 and S107 indicate loop ends, and indicate that the processing from S102 to S106 surrounded by the loop ends is repeated until the gradation value g reaches gMAX. The processes from S102 to S106 are processes for adding one dot to the dot pattern with the gradation value g, thereby generating a dot pattern with continuous gradation values, that is, with the gradation value g+1. Also, gMAX that determines the processing termination condition is gMAX≧g0, and gMAX=Sx×Sy−1=65535 in this embodiment.

Similarly, S201 and S207 indicate loop ends, and indicate that the processing from S202 to S206 surrounded by the loop ends is repeated until the gradation value g reaches gMIN. The processing from S202 to S206 is processing for deleting one dot from the dot pattern with the gradation value g to generate a dot pattern with continuous gradation values, that is, with the gradation value g−1. gMIN≦g0, and gMIN=0 in this embodiment. By repeating dot addition and dot deletion in this way, a dot pattern of all gradations is generated.

Since the processing in S102 and S202 have the same processing content, S102 will be described below. In S102, the CPU applies a low-pass filter to the dot pattern d(g) corresponding to the gradation value g to calculate a density variation map n(g). In the following description, the density variation map n(g) is abbreviated as density variation n(g).

Density variation n(g) is used to evaluate dot density in dot pattern d(g) with tone value g. The smaller the value of density variation n(g), the lower the smoothed density and the sparse dots are evaluated. evaluated. In S104, which will be described later, dots are added to positions where the density variation n(g) is small and the dots are sparse. Also, in S204, dots are deleted from positions where the density variation n(g) is large and dots are dense. As a result, local variations in density are suppressed and low graininess is achieved.

The density variation n(g) is a two-dimensional array of the same size as the dot pattern d(g), and the array values change according to the tone value g, similar to the dot pattern d(g). In the following description, n(g) is also written as n(g, x, y). It is assumed that the dither matrix is applied periodically to the input image. Therefore, in the filtering process of S103, the CPU generates the density variation n(g) by cyclic convolution of the dot pattern d(g) and the filter coefficient. A circular convolution operation is an operation in which a normal convolution operation is performed between a dot pattern d(g) in which periodic boundary conditions are set and filter coefficients. Details of the filter coefficients used in S103 will be described later.

Since the processing in S103 and S203 have the same processing content, S103 will be described below. In S103, the CPU extracts the x-direction total dot number map s(g) for each position in the y-direction for the dot pattern d(g) corresponding to the gradation value g. The total dot number map s(g) is a one-dimensional array, and s(g) is also written as s(g, y). It is used as a placement constraint to equalize the number of dots for each position (for each nozzle) in the y direction. In S104, which will be described later, dots are added to the dot pattern d(g, x, y) at the y position where the value of s(g, y) is small and the total number of dots is small. In S204, dots are deleted from the dot pattern d(g, x, y) at the y position where the value of s(y) is large and the total number of dots is large. As a result, the difference in dot generation frequency between nozzles can be adjusted to be small.

In S104 and S204, the CPU adds or deletes dots based on density variation n(g) and total dot number map s(g, y). In S104, the CPU creates a dot pattern at the position (xMIN, yMIN) where the density variation n(g, x, y) is the smallest among the y position group where the total dot number map s(g, y) is the smallest. Add a dot to d(g, xMIN, yMIN). On the other hand, in S204, the CPU selects the dot at the position (xMAX, yMAX) where the density variation n(g, x, y) is the largest among the y positions where the total dot number map s(g, y) is the largest. Delete dots in pattern d(g, xMAX, yMAX). Details of the processing will be described later.

In S105 and S205, the CPU determines the value of the dither matrix M(x, y) based on the dot addition or deletion positions (xMIN, yMIN, or xMAX, yMAX). In S105, the CPU sets the value of M(xMIN, yMIN) to the gradation value g in the dither matrix M(x, y). In S205, the CPU sets the value of M(xMAX, yMAX) to the gradation value g−1 in the dither matrix M(x, y).

In S106, the CPU increments the tone value g to g+1. Similarly, in S206, the CPU decrements the tone value g to g-1.

After exiting the loop of S101-107 and the loop of S201-S207 as described above, the CPU adjusts the pixel value range of the input image in accordance with the value range of the dither matrix in S208. Before performing S208, the dither matrix M(x, y) stores values from gMIN to gMAX. In this embodiment, values from 0 to 65535 (mMIN to mMAX) are stored. When the input image for dither processing is 8-bit, the range of the input image is 0 to 255 (thMIN to thMAX). Therefore, using a dither matrix that stores values from 0 to 65535 will not produce a proper halftoning result. In S208, the CPU expands the number of bits of the pixel values of the input image so as to match the value range of the dither matrix M(x, y). For example, when the range of dither matrix values is 16 bits from 0 to 65535 and the input image is 8 bits, the input image is 0-padded to the lower order to make it 16 bits for comparison.

On the other hand, the values of the dither matrix may be adjusted according to the range of pixel values of the input image. For example, the lower 8 bits of a 16-bit dither matrix storing values from 0 to 65535 are truncated, and the upper 8 bits are used as the dither matrix values. Further, when it is desired to change from thMIN to thMAX, the CPU obtains the integer part of the following equation as the value of the dither matrix after adjustment.
a×M(x,y)+b
however,
a = (thMAX-thMIN)/(mMAX-mMIN),
b=thMIN-a×mMIN
is.

According to the processing flow of FIG. 5 described above, it is possible to generate a dither matrix that can equalize the ink ejection frequency for each nozzle while maintaining low graininess. Details of each process in FIG. 1 will be described below.

First, the filter coefficient f used in S102 and S202 will be described. This filter coefficient is used to extract the density variation map n(g). f is a two-dimensional array and is also written as f(x, y). In this embodiment, the array size of f(x, y) is assumed to be the same as the dot pattern d(g). That is, when the filter size in the x direction is Sfx and the filter size in the y direction is Sfy, these values are "256". As the value of f(x, y), the value of the following equation (1) is used.
r={(x−x ₀ ) ² +(y−y ₀ ) ² } ^1/2
f(x,y)=1/(r+1) (1)
In Equation (1), x0 and y0 are the center positions of the filter coefficients, and x0=Sfx/2 and y0=Sfy/2.

In S104 and S204, the CPU adds or deletes dots so as to moderate the density variation n(g), thereby reducing the density of dots, thereby realizing low graininess. In order to achieve this, it is necessary to extract the density between dots. To extract the density between dots, the filter f should be a function of the reciprocal of the distance r from the dot as shown in equation (1). In this embodiment, "1" is added to the denominator to avoid division by zero at the distance r=0.

Note that the filter f used in S104 and S204 is not limited to this, and a low-pass filter capable of extracting frequency components perceived as graininess may be used.

Next, the dot arrangement method in S104 and S204 will be explained. The CPU adds or deletes dots based on the density variation n(g) and the total dot number map s(g, y). S104 is a process of adding dots. In S104, the CPU finds the smallest density variation n(g, x, y) in the y position where no dots exist and where the total dot number map s(g, y) is the smallest. Add a dot at location (xMIN, yMIN). Specifically, the pixels in the y-position group with the smallest number of dots in the total dot number map s(g, y) and the pixel value of "0" in the dot pattern d(g, x, y) The position (xMIN, yMIN) at which the value of the density variation n(g, x, y) is the minimum is searched from among the pixels having the . A dot is added by setting the pixel value of d(g, xMIN, yMIN) to "1". Note that if there are a plurality of positions where the values of the density variation map n(g, x, y) are the smallest, the CPU randomly selects a dot-adding position from among them.

On the other hand, S204 is a process of deleting dots. In S204, the CPU maximizes the value of the density variation map n(g, x, y) in the y position where dots exist and where the total dot number map s(g, y) is maximum. Delete a dot from position (xMAX, yMAX). Specifically, the pixels in the y-position group having the maximum dot number in the total dot number map s(g, y) and the pixel value of "1" in the dot pattern d(g, x, y) A position (xMAX, yMAX) at which the value of the density variation map n(g, x, y) is maximized is searched from among the pixels having the . Then, the CPU deletes the dot by setting the pixel value of d(g, xMAX, yMAX) to "0". Note that if there are a plurality of positions where the values of the density variation map n(g, x, y) are maximum, the CPU randomly selects positions from which dots are to be deleted.

Next, the details of the initial dot pattern generation method in S100 will be described with reference to FIG. Briefly, the CPU first generates a random dot pattern. Starting from this random dot pattern, the CPU repeats movement of the CPU dots to generate an initial dot pattern d(g0) with low graininess and inconspicuous density unevenness.

In S301, the CPU generates a random dot pattern d_r with the number of gradations g0. d_r is a two-dimensional array and is also written as d_r(x, y). The size is the same as the dither matrix M(x, y), with the size in the x direction being Sx and the size in the y direction being Sy. As described above, the values of Sx and Sy are set to "256" in this embodiment. As with d(x, y) described above, the pixel value is "1" when a dot exists, and the pixel value is "0" when no dot exists.

In S301, the CPU repeats dot addition starting from an image in which no dots exist until the number of gradations reaches g0, thereby generating a random dot pattern with the number of gradations g0.

The CPU sequentially increments the y-direction position yr from 1 to Sy to determine the dot addition position. Then, when it reaches the extreme end Sy, it returns to 1 and repeats. On the other hand, the CPU uses random numbers to determine candidates for positions in the x direction where dots are to be added. Specifically, the CPU generates a random number from 1 to Sx and determines the x-direction position xr. Then, if a dot does not exist at (xr, yr), the CPU adds a dot to that position. That is, when the pixel value of d_r(xr, yr) is "0", the CPU sets the pixel value of d_r(xr, yr) to "1". However, if the pixel value of d_r(xr, yr) has already become "1", the CPU generates a new random number and changes the candidate of the position xr in the x direction where the dot is to be added. As a result, the dot pattern generated in S301 becomes a random dot pattern in which the deviation of the total number of dots in the x direction is suppressed to 1 dot or less.

S302 and S305 are loop ends, meaning that the processing from S303 to S304 is repeated a predetermined number of times. In this embodiment, the number of iterations is set to 10000, but the number is not limited to this, and a predetermined convergence condition may be set.

In S303, the CPU extracts the density variation n from the dot pattern d_r by the same method as in S103. The density variation n is a two-dimensional array and is also written as n(x, y).

At S304, the CPU moves the dots based on the density variation map n. First, the position (xMAX, yMAX) at which the value of the density variation n(x, y) is maximized is searched among the pixels in which dots exist in d_r(x, y). Next, the CPU selects the value of density variation n(x, yMAX) from among the pixels at the same y position (yMAX) as the position where the value of density variation n(x, y) is maximum and where no dot exists. Find the x-position (xMIN, yMAX) where . The CPU then moves the dot at (xMAX, yMAX) to (xMIN, yMAX). That is, the pixel value of the dot pattern d_r (xMAX, yMAX) is set to "0" and the pixel value of the dot pattern d_r (xMIN, yMAX) is set to "1". A dot at a position where density variation n is large is moved to a position where density variation n is small at the same y position. As a result, in the dot pattern generated in S304, the total number of dots for each position yr in the y direction does not change before and after the movement of the dots. This makes it possible to generate a dot pattern with good dot dispersion while suppressing the deviation of the total number of dots in the x direction to 1 dot or less.

As described above, when a dot pattern is generated using a dither matrix generated by the method described in this embodiment, the difference in the number of discharged dots for each nozzle of the print head 203 can be set within a preset range (ejection The number of dots can be almost uniform). As a result, it is possible to prevent the life of specific nozzles from being shortened and extend the life of the print head.

[Second embodiment]
Uniformity in the number of ejected dots for each nozzle also works favorably in density unevenness correction for suppressing deterioration in image quality due to variations in ink ejection characteristics for each nozzle. In the second embodiment, the effect of density unevenness correction and its implementation example will be described.

In an actual image forming apparatus, each nozzle arranged in a print head has a variation in ejection characteristics due to the manufacturing process and constituent materials of the print head. Variations in the ejection characteristics of each nozzle appear as non-uniformity in the size and ejection direction of the ink droplets ejected from each nozzle, causing density unevenness in printed images. To solve such a problem, there is a density unevenness correction technique that suppresses density unevenness by correcting input image data based on the ejection characteristics of each nozzle that constitutes a print head. Specifically, a density measurement test chart for obtaining a density conversion function for correcting density unevenness is output, the image of the output density measurement test chart is read, the recording density is measured, and the measurement is performed. A density correction parameter for image data for each nozzle is generated based on the density data. When printing an image, the input image data is corrected based on the density correction parameter, and image recording is controlled based on the corrected image data, thereby producing an image in which the occurrence of density unevenness is suppressed. Form.

If the dot positions of the dot pattern generated from the density measurement test chart are biased toward a specific nozzle, the number of ink jets will vary from nozzle to nozzle. As a result, the output density corresponding to nozzles with a large number of ejections is high, and the output density corresponding to nozzles with a small number of ejections is low. Therefore, the density measurement test chart that is output includes a density difference due to the ejection characteristics of each nozzle and a density difference due to the difference in the number of ink ejections for each nozzle. It becomes difficult to separate and measure the density variation caused by the ejection characteristics of each nozzle from the print density of the test chart, and the problem arises that the effect of density unevenness correction is reduced.

Therefore, in order to accurately measure density fluctuations caused by nozzle ejection characteristics, a dither matrix adjusted so that the number of dots formed by each nozzle to be corrected for density unevenness is uniform is used. It is effective to generate a dot pattern for the test chart for measurement.

In addition, if there is a difference in the number of output dots for the input gradation for each nozzle position, even if the density fluctuation caused by the ejection characteristics of each nozzle can be correctly corrected, the effectiveness of the correction (the number of output dots due to correction) for each nozzle change) is different, and the effect of correcting density unevenness may be reduced. Specifically, for example, it is assumed that there are two nozzles whose ejection volume per dot is approximately the same as the ideal ejection volume. At this time, in the input image, the pixel lines corresponding to the two nozzles to be corrected are similarly corrected so that the pixel values are increased. However, if the number of output dots differs for each nozzle according to the gradation, different numbers of output dots will be ejected even if the pixel values before and after correction are the same. As a result, the density unevenness remains even after performing density unevenness correction processing.

Therefore, in order to improve the density unevenness correction effect of the recorded image, adjustment is made so that the number of dots corresponding to each gradation formed by each nozzle is uniform for the input image that has undergone density unevenness correction for each nozzle. It is useful to use a modified dither matrix.

FIG. 8 is a block diagram showing configurations of an image processing apparatus 10 and an image forming apparatus 20 applicable to the second embodiment. In the following description, the description of the parts common to the first embodiment is simplified or omitted.

In FIG. 8, the components other than the image reading device 30, the density unevenness correction parameter calculation unit 105, the density unevenness correction parameter storage unit 106, the density unevenness correction processing unit 107, and the test chart generation unit 108 are the same as those in FIG. .

The image reading device 30 reads an image recorded on a recording medium by the recording head 203 and converts it into electronic image data (read image data). For example, a CCD line sensor can be used as the image reading device 30 . The image reading device 30 of this embodiment is an in-line sensor installed in the middle of the medium conveying path, and reads an image recorded by the recording head 203 during conveyance before discharging the paper. The image reading device 30 can read the output result of a test chart for density measurement, which will be described later.

The density unevenness correction parameter calculation unit 105 performs density measurement based on the read image of the density measurement test chart acquired from the image reading device 30, and generates density unevenness correction value data (density unevenness correction parameter) for each nozzle. do. The density non-uniformity correction parameter generated by the density non-uniformity correction parameter calculation unit 105 is stored in the density non-uniformity correction parameter storage unit 106 .

A density unevenness correction processing unit 107 performs image signal correction for suppressing density unevenness of a print image on a print medium caused by variations in ejection characteristics of nozzles in the print head 203 or the like. In this embodiment, a density unevenness correction LUT, which is a lookup table for density unevenness correction that describes the conversion relationship between an input signal value and an output signal value, is prepared for each nozzle of the print head 203, and this density unevenness correction LUT is prepared. is used to convert the signal value.

The test chart generation unit 108 generates density measurement test chart data for obtaining density measurement data necessary for calculating density unevenness correction parameters, and provides the data to the input image buffer 102 . FIG. 9 is an example of a test chart for density measurement, in which gradation values D1 to D10 are determined in advance in a rectangular area having a width Ty in the width direction of the recording medium of the recording head 203 and a height Tx in the conveying direction. It has gradation patches of uniform density in a belt shape in the transport direction. As shown in FIG. 9, in this embodiment, there is chart data for the nozzle width so that the ejection characteristics of all noises can be derived. Further, in the present embodiment, all gradations from 0 to 255 are targeted for density unevenness correction. Therefore, the test chart for density measurement is generated so as to include patches for each gradation corresponding to densities that are discrete from 0 to 255 but are equally spaced from low density to high density.

In order to obtain density fluctuations caused by ejection characteristics, it is possible to generate a dot pattern of a test chart for density measurement using a dither matrix adjusted so that the number of dots formed by each nozzle is uniform. It is valid. Therefore, the x-direction size Tx of the density area of each gradation in the density measurement test chart must be equal to or larger than the x-direction size Sx pixels of the dither matrix (Sx is 256 in the embodiment). In the second embodiment, Tx=800 pixels.

FIG. 10 is a flow chart showing an outline of a procedure for creating density unevenness correction parameters executed by a CPU (not shown) in the image processing apparatus 10 of the second embodiment. It should be noted that the program shown in the figure is stored in a predetermined memory (RAM, etc.).

In S401, the CPU sends the image data of the test chart for density measurement generated by the test chart generation unit 108 to the input image buffer 102, thereby generating an image of the test chart for density measurement on the recording medium by the recording head 203. output. At this time, the CPU causes the density unevenness correction processing unit 107 not to perform density unevenness correction, and outputs a gradation patch with a uniform density. Further, the halftone processing unit 103 is caused to generate a dot pattern of a test chart for density measurement using a dither matrix adjusted so that the number of dots formed by each nozzle is uniform.

In S402, the CPU controls the image reading device 30 to read the output result of the test chart for density measurement and to measure the test chart.

In S403, the CPU integrates signals in the conveying direction (x direction) for each band-shaped density patch corresponding to each gradation by image processing from the captured image, and adds 1 in the recording medium width direction (y direction). Create dimensional measurement data. At this time, in order to obtain density fluctuations caused by the ejection characteristics of each nozzle, it is desirable to integrate signals in the transport direction (x direction) within a range in which the number of dots for each nozzle is uniform. That is, it is desirable that the integration range Mx in the transport direction (x direction) is an integral multiple of the print width (mm) of the dither matrix in the x direction. When the x-direction size Sx of the dither matrix is 256 (pixels), the print width (mm) is 256 (pixels)/1200 (pixels/inch)*25.4 (mm/inch)≈5.4 (mm). . On the other hand, the print width (mm) of the x-direction size Tx of the density area of each gradation in the density measurement test chart is 800 (pixel)/1200 (pixel/inch)*25.4 (mm/inch)≈16.9. (mm). In this embodiment, the integration range Mx in the transport direction (x direction) is set to be equal to or less than the print width (16.9 mm) of the x-direction size of the density area and the print width (5.4 mm) of the x-direction size Sx of the dither matrix. 16.3 mm, which is an integral multiple (three times) of

Based on the measured one-dimensional data at the reading resolution of the image reading device 30 (for example, 600 dpi in the width direction of the recording medium), the CPU performs interpolation in units of the resolution (recording resolution) of the nozzles of the recording head 203, and performs interpolation for each nozzle position. Create densitometry data.

Next, in S404, the CPU calculates a density unevenness correction parameter for each nozzle, that is, a density unevenness correction LUT, from the obtained density measurement data for each nozzle position and the target gradation characteristics. By interpolating density measurement data for discrete tone values, density data for tone values other than measurement points can be obtained.

Next, with reference to FIGS. 7A and 7B, gradation characteristics of dot patterns generated by halftone processing using the dither matrix of the present invention will be described.

In the graphs of FIGS. 7A and 7B, the horizontal axis represents the input gradation value, and the vertical axis represents the input image in of uniform gradation having the same size (Sx, Sy) as the dither matrix and the dither matrix M. It is the total number of dots xSum(y) in the generated dot pattern in the x direction (hereinafter referred to as the total number of dots). If the y-direction size Sy of the dither matrix is 256, 256 graphs can be generated, but note that the correspondence at some y-positions is shown here for the sake of easy viewing of the graph. want to be

FIG. 7A shows an example of dot patterns generated using the dither matrix created in this embodiment. For comparison, FIG. 7B shows an example of a dot pattern generated using a general blue noise dither matrix created without considering that the dot generation frequency for each nozzle is uniform. show. Here, a general blue noise dither matrix is created by a method that differs only in S103, S104, S203, and S204 from the matrix creation method using FIG. Assume that a general blue noise dither matrix is a matrix generated without S103 and S203 and also referring to density variation n(g, x, y) in S104 and S204.

FIG. 7A shows the total number of dots for input tone values at multiple positions in the y direction, but the graphs overlap, so it looks like one line. In other words, the total number of dots in the dot pattern generated using the dither matrix for input tone values of uniform density is approximately uniform regardless of the tone, and the number of dots increased between tones is constant (linear ).

On the other hand, when a dot pattern is generated using a general blue noise dither matrix as shown in FIG. This indicates that the number of dots increased between gradations is non-linear.

The total number of dots for each nozzle shown in FIG. 7A can be completely matched if the total number of dots is divisible by the y-direction size Sy of the dither matrix (if it is an integral multiple of Sy). . This holds because the number of gradations (the number of g) that can be simulated by the dither matrix is Sy. That is, if the total number of dots in the dot pattern generated using the dither matrix is d, and d % Sy=0, the total number of dots for each nozzle can be d/Sy. That is, by adjusting the threshold range of the dither matrix M so that d % Sy=0, the total number of dots for each nozzle can be made uniform. When density unevenness correction for each nozzle targets all gradations as in the present embodiment, it is desirable to equalize the total number of dots for each nozzle in all gradations.

Note that the total number of dots for each nozzle may not match completely, for example, when the size of the dither matrix is limited. For example, depending on the gradation, the total number of dots may not be divisible by the size Sy of the dither matrix in the y direction, and the total number of dots for each y position may not be uniform. In such cases, the difference in the total number of dots for each y-position is preferably 1 or less.

For the test chart for density measurement output in this embodiment, a dot pattern for the test chart for density measurement is generated using a dither matrix adjusted so that the number of dots formed by each nozzle is uniform. Therefore, there is no bias in the number of ink ejections for each nozzle, and it is possible to accurately measure density fluctuations caused by the ejection characteristics of each nozzle from the recording density of the test chart. As a result, it is possible to improve the effect of density unevenness correction based on the density measurement data.

On the other hand, a dither matrix that is adjusted so that the number of dots formed by each nozzle is uniform is effective even when density unevenness correction for each nozzle is performed on an input image by referring to density unevenness correction parameters for each nozzle. is. When actually printing, the density unevenness correction processing unit 107 refers to density unevenness correction parameters and performs density unevenness correction for each nozzle on the input image. Further, the halftone processing unit 103 generates a dot pattern using a dither matrix that is adjusted so that the number of output dots for each nozzle position is uniform with respect to the input tone. To improve the density unevenness correction effect of a printed image by generating a dither pattern with an appropriate number of dots for the density unevenness correction amount for each nozzle since the number of dots increased between gradations is constant (linear). can be done.

Therefore, the measurement of the density unevenness correction parameter for each nozzle is not limited to the density measurement test chart described above, and the density unevenness correction for each nozzle is performed using the density unevenness correction parameter created by other measurement methods. Also, it is possible to improve the density unevenness correction effect of the recorded image.

[Third Embodiment]
In the second embodiment, all gradations are targeted for density unevenness correction. In the third embodiment, an example will be described in which density unevenness correction is performed on only some tones. For example, when there is almost no density unevenness in patches of low densities D9 and D10 in FIG. 9, only patches higher than density D9 may be subjected to density unevenness correction. In this case, the range of densities in which the density unevenness correction effect should be improved is the range of densities D8 to D1 (where the density D1 is assumed to be the maximum density). The density unevenness correction processing unit 107 determines whether or not the pixel value of the pixel in the input image is equal to or greater than the density D8. If the pixel value is equal to or greater than the density D8, the density unevenness correction processing unit 107 refers to the density unevenness correction parameter and executes density unevenness correction. On the other hand, if the pixel value is less than the density D8, the density unevenness correction is not executed, and the pixel value is output as it is.

On the other hand, in the low-density area, there is a case where the dispersibility of the dots is emphasized even though the density unevenness is inconspicuous. When a small number of dots are ejected onto the recording medium, a plurality of dots are ejected close to each other, so that the appearance of clusters of dots is visually perceived and the graininess is reduced. Therefore, in the third embodiment, when halftone processing is performed on an input image, dither matrices are switched and applied. When the pixel values of the pixels to be processed have densities D8 to D1, the same dither matrix (M1) as in the first embodiment is used. When the pixel value of the pixel to be processed is in a low-density portion with a density of D9 or higher, a dither matrix (M2) created without considering uniform dot generation frequency for each nozzle is used.

The halftone processing unit 103 executes the following processing.
Iin(x,y)≧D8
Iout(x,y)=BLACK if Iin(x,y) > M1(x % Sx, y % Sy)
Iout(x,y)=WHITE if Iin(x,y) ≤ M1(x % Sx, y % Sy)
Iin(x,y)<D8
Iout(x,y)=BLACK if Iin(x,y) > M2(x % Sx, y % Sy)
Iout(x,y)=WHITE if Iin(x,y) ≤ M2(x % Sx, y % Sy)
Since the dimatrix M2 refers only to the density variation n when adding dots or deleting dots, it is more restrictive in determining dot placement than the dot creation method shown in FIG. few. Therefore, the dither matrix M2 is better than the dither matrix M1 only in terms of dot dispersion. to run. When the pixel value is within the density range in which density unevenness correction processing is executed, the effect of density unevenness correction is improved by using a dither matrix created so that the number of dots for each nozzle is uniform. On the other hand, in a density range in which pixel values do not require density unevenness correction processing, a dither matrix that increases dispersion is used.

In this embodiment, the dither matrix referred to by the halftone process is switched according to the pixel value. However, there is also a method of achieving the same effect as in the third embodiment with one dither matrix. This is a method of determining whether or not to refer to the total dot number map according to the gradation value g when creating the dither matrix. For example, in the flowchart shown in FIG. 5, in S201 to S207, if the gradation g is equal to or less than a predetermined value (for example, the density D9), S203 is not executed, and in S204 the density change n (g, x, y ) to determine where to remove the dot. Thus, in the dither matrix, when the gradation g is equal to or greater than a predetermined value, the total number of dots for each nozzle is uniform within the range of integral multiples of the size Sy in the y direction. By performing halftone processing using the dither matrix created in this way, the same effect as in the third embodiment can be obtained.

(Other examples)
In the above-described embodiment, the case where the dither matrix simulates 256 gradations has been described as an example. However, the number of gradations to be pseudo-reproduced by the dither matrix may be only discrete gradations among the number of gradations of the input image. For example, 0, 15, 30, 45, 60, 75, 90, 105, 120, 135, 150, 165, 180, 195, 210, 225, 240, and 255 of 0 to 255 are simulated. , the constraint on the total number of dots for each nozzle shown in FIG. In this case, the size of the dither matrix in the y direction should be set to be a multiple of 18 tones to be simulated.

Further, in the above-described embodiments, the method in which the halftone processing unit performs comparison processing using a dither matrix has been described as an example. However, the dither matrix created in the first embodiment can also be applied to the method of determining dot placement according to the dither matrix. For example, the dither matrix described above may be used in the method described in JP-A-2016-021735.

The present invention supplies a program that implements one or more functions of the above-described embodiments to a system or device via a network or a storage medium, and one or more processors in the computer of the system or device reads and executes the program. It can also be realized by processing to It can also be implemented by a circuit (for example, ASIC) that implements one or more functions.

DESCRIPTION OF SYMBOLS 10... Image processing apparatus 20... Image forming apparatus 102... Input image buffer 103... Halftone processing part 104... Halftone image buffer 202... Head driving part 203... Recording head

Claims (5)

a recording head having a plurality of recording elements arranged in a first direction, the recording head directed toward a recording medium relatively conveyed along a second direction perpendicular to the first direction; an image forming apparatus for forming an image on a recording medium by means of recording elements of an image processing apparatus for acquiring correction information for correcting density unevenness on the image caused by characteristics of each of the recording elements; ,
acquisition means for acquiring image data having the same gradation value at each position in the first direction and for acquiring the correction information ;
A distributed dither matrix having a size Sx (Sx is a natural number) in the second direction is repeatedly developed in the second direction , and the gradation value of the image data and the corresponding threshold value are compared in magnitude. a converting means for converting into halftone image data representing a halftone pattern by
generating means for generating correction information for correcting density unevenness caused by the characteristics of each of the recording elements by reading a chart image formed based on the halftone image data ;
In the converted halftone image data, the difference in sum of pixel values included in the range of Sx in the second direction between each position in the first direction is calculated by the image forming apparatus. The number of tones that can be expressed is less than or equal to
the size Tx in the second direction of the area having the same gradation value in the image data for acquiring the correction information is greater than or equal to Sx;
The generating means generates measurement data obtained by accumulating read signal values of a print width on the recording medium corresponding to a positive integral multiple of the size Sx of the dither matrix in the second direction in the read image of the chart image. Generate a correction table based on
An image processing apparatus characterized by: 2. The image processing apparatus according to claim 1, wherein the dither matrix has a size of Sy columns ( Sy is a natural number) corresponding to the first direction. 3. The image processing apparatus according to claim 2, wherein both Sx and Sy are 256 or more. a recording head having a plurality of recording elements arranged in a first direction, the recording head directed toward a recording medium relatively conveyed along a second direction perpendicular to the first direction; A method of controlling an image processing apparatus for acquiring correction information for correcting density unevenness on the image caused by the characteristics of each of the recording elements in an image forming apparatus that forms an image on the recording medium using the recording elements of and
an obtaining step in which an obtaining means obtains image data having the same gradation value at each position in the first direction and for obtaining the correction information ;
The transforming means repeatedly develops a distributed dither matrix having a size Sx (Sx is a natural number) in the second direction in the second direction, and the gradation value of the image data and the corresponding threshold value are large or small. a converting step of converting into halftone image data representing a halftone pattern by performing a comparison ;
generating means for generating correction information for correcting density unevenness caused by the characteristics of each of the recording elements by reading a chart image formed based on the halftone image data;
In the converted halftone image data, the difference in sum of pixel values included in the range of Sx in the second direction between each position in the first direction is calculated by the image forming apparatus. The number of tones that can be expressed is less than or equal to
the size Tx in the second direction of the area having the same gradation value in the image data for acquiring the correction information is greater than or equal to Sx;
In the reading image of the chart image, the generating step includes measurement data obtained by accumulating reading signal values of a print width on the recording medium corresponding to a positive integer multiple of the size Sx of the dither matrix in the second direction. Generate a correction table based on
A control method for an image processing apparatus, characterized by: A program that is read and executed by a computer to cause the computer to perform each step of the method according to claim 4 .

Images