JP2017011569A - Imaging processing device, image processing method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize proper tone conversion over an entire image when half-tone processing is applied to image data.SOLUTION: An image processing device includes: first holding means that has plural elements in which the growth order of dots of each pixel constituting a predetermined pixel region in image data is set, and holds cells arranged so as to have a predetermined screen angle; and second holding means for holding a noise matrix in which plural noise values are arranged in a matrix form in a predetermined spatial frequency characteristic. The image processing device further includes: setting means for setting corresponding noise values for plural cells corresponding to each element of the noise matrix arranged so as to have a screen angle; and deriving means for deriving a threshold value to be compared with a pixel of interest in the half-tone processing on the basis of the growth order of dots corresponding to the pixel of interest and the noise value set for the cell corresponding to the pixel region containing the pixel of interest by the setting means.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an image processing apparatus, an image processing method, and a program for applying halftone processing to image data.

Conventionally, when printing multi-gradation image data handled by a digital camera or the like, halftone processing for converting multi-gradation image data into a small number of gradations that can be expressed by the image forming apparatus is applied. . The halftone process includes a dither method and an error diffusion method. For example, in an electrophotographic printer, a dither method using a dot concentration type threshold matrix is often applied in consideration of dot formation stability and processing speed. In the dither method, a threshold matrix in which a plurality of threshold values are arranged is regularly arranged and superimposed on the input image, and a halftone image is generated by comparing the gradation value of the processing target pixel with the threshold value corresponding to the processing target pixel. To do.
However, in the dither method using the dot concentration type threshold matrix, since the output value of the pixel to be processed is determined by comparison with a periodically varying threshold value, a periodic striped pattern called moire tends to occur. Further, if each threshold value matrix to be superimposed on the input image has the same threshold value group, only gradations corresponding to the number of threshold values (number of pixels) included in one threshold value matrix can be expressed.

  In view of this, Japanese Patent Laid-Open No. 2004-228620 proposes a method for improving gradation reproducibility by performing halftone processing using a blue noise mask. In Patent Document 1, first, a plurality of cells in which the dot output order (growth order) is determined for each position corresponding to each pixel of the input image data are arranged side by side so as to have a screen angle, thereby creating a super cell. To do. Then, according to the screen angle of the supercell, two axes indicating the cell arrangement direction are set, and these axes are made to coincide with two orthogonal axes of the blue noise mask so that each cell in the supercell corresponds to each blue cell. Noise data is set to create a threshold matrix.

Japanese Patent No. 4987833

However, in the technique described in Patent Document 1, a part of the blue noise mask is cut out and used in accordance with the shape of the supercell. For this reason, the gradation can be improved by taking advantage of the blue noise characteristics in the supercell, but the continuity of the blue noise characteristics is not considered at the boundary between the supercells, and low frequency components are generated in the entire image. Resulting in.
Therefore, an object of the present invention is to realize appropriate gradation conversion over the entire image when halftone processing is applied to image data.

  In order to solve the above-described problem, an aspect of the image processing apparatus according to the present invention includes a plurality of elements in which the growth order of dots of each pixel constituting a predetermined pixel region in image data is set. A first holding means for holding cells arranged to have a screen angle; a second holding means for holding a noise matrix in which a plurality of noise values are arranged in a matrix with a predetermined spatial frequency characteristic; and For each of the cells corresponding to each element of the noise matrix arranged so as to have a screen angle, setting means for setting the corresponding noise value, the growth order of the dots corresponding to the pixel of interest, Based on the noise value set for the cell corresponding to the pixel region including the target pixel by the setting unit, the target image is obtained by halftone processing. And a deriving means for deriving a threshold value to be compared with the gray scale value.

  According to the present invention, when halftone processing is applied to image data, appropriate gradation conversion can be realized over the entire image.

It is a figure which shows the structural example of the printing apparatus in 1st embodiment. It is a functional block diagram which shows the structural example of the image process part of 1st embodiment. It is a functional block diagram which shows the structural example of a halftone process part. It is a figure which shows an example of a cell of 10 pixels. It is a figure which shows an example of a blue noise matrix. It is a flowchart which shows a threshold value calculation process procedure. It is a figure which shows the example which has arrange | positioned the cell on the image. It is a figure which shows the example which shrunk the cell in the row direction. It is a figure for demonstrating a normalization process. It is a functional block diagram which shows the structural example of the image process part of 2nd embodiment. It is a figure which shows an example of the threshold value table for noises. It is a figure which shows an example of distribution of the threshold value for every dot growth order. It is a figure which shows an example of distribution of the threshold value for every dot growth order.

DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the accompanying drawings. The embodiment described below is an example as means for realizing the present invention, and should be appropriately modified or changed depending on the configuration and various conditions of the apparatus to which the present invention is applied. It is not limited to the embodiment.
(First embodiment)
FIG. 1 is a diagram illustrating a configuration example of a printing apparatus according to the first embodiment.
The printing apparatus 100 includes an image input unit 10, an image processing unit 20, and an image output unit 30. The image input unit 10 and the image processing unit 20 are connected by a communication interface or circuit, and the image processing unit 20 and the image output unit 30 are connected by a communication interface or circuit.

  The image input unit 10 is, for example, a scanner unit including a lens, a CCD (Charge Coupled Device) sensor, and an analog / digital conversion unit (A / D conversion unit). The CCD sensor photoelectrically converts an image of a subject (original) imaged by a lens and generates analog electric signals of R (Red), G (Green), and B (Blue). The generated analog electric signal is input to the A / D conversion unit, and after correction for each of the R, G, and B colors, A / D conversion is performed. The digitized R, G, B full color signals (hereinafter referred to as “digital image signals”) are output to the image processing unit 20. The image input unit 10 is not limited to the scanner unit, and may be an image capturing unit that captures an image of a subject and generates a digital image signal, or an image capturing unit that captures a digital image signal from an external device. .

The image processing unit 20 performs various image processing described later on the multi-tone digital image signal input from the image input unit 10. Then, the image processing unit 20 outputs the digital image signal subjected to the image processing to the image output unit 30.
The image output unit 103 includes a print output unit such as a raster plotter using an inkjet head or a thermal head, and prints an image on a print medium using a digital image signal input from the image processing unit 20.
In the present embodiment, the case where the image input unit 10, the image processing unit 20, and the image output unit 30 are mounted on one apparatus (printing apparatus 100) will be described, but each is mounted on a separate apparatus. Also good.

(Configuration of the image processing unit 20)
FIG. 2 is a functional block diagram illustrating a configuration example of the image processing unit 20. The image processing unit 20 includes an input correction unit 21, a color space conversion unit 22, a density correction unit 23, and a halftone processing unit 24. For the digital image signal input from the image input unit 10, The image processing described below is performed.
The input correction unit 21 performs correction processing on the input digital image signal. The digital image signal input to the input correction unit 21 is composed of R, G, and B luminance signals, on which the sensitivity characteristics of the CCD sensor and the light distribution characteristics of the document illumination lamp are superimposed. Therefore, the input correction unit 21 performs a correction process for these characteristics. Digital image signals (luminance signals R, G, B) output from the input correction unit 21 are input to the color space conversion unit 22.

  The color space conversion unit 202 converts the luminance signals R, G, and B of the digital image signal into density signals C (Cyan), M (Magenta), Y (Yellow), and K (blacK). Digital image signals (density signals C, M, Y, and K) output from the color space conversion unit 202 are input to the density correction unit 23. The density correction unit 23 performs halftone processing by the halftone processing unit 24 (hereinafter referred to as “half-tone processing”) so that density fluctuation does not occur when the digital image signal is converted into a binary digital image signal by the subsequent halftone processing unit 24. In consideration of the characteristics of “tone processing”), density correction is performed in advance. The digital image signal (density signals C, M, Y, K) output from the density correction unit 23 is input to the halftone processing unit 24.

The halftone processing unit 24 performs halftone processing by a dither method using a threshold matrix on the multi-tone digital image signal (image data) input from the density correction unit 23 and converts the halftone processing into a halftone expression. A binary digital image signal (print signals C, M, Y, K) is generated. The generated binary digital image signal is output to the image output unit 30. 2 can be realized by a CPU (not shown) included in the printing apparatus 100 executing a predetermined program.
FIG. 3 is a functional block diagram illustrating a configuration example of the halftone processing unit 24. The halftone processing unit 24 includes a cell information holding unit 241, a noise matrix holding unit 242, a threshold value calculation unit 243, and a comparison unit 244.

  The cell information holding unit 241 holds a cell in which a plurality of elements corresponding to each pixel constituting a predetermined pixel region are arranged. For each element constituting the cell, the dot growth order of the corresponding pixel is set. FIG. 4 is a diagram illustrating an example of a cell. FIG. 4 shows a cell having 10 elements corresponding to 10 pixels. In FIG. 4, one square corresponds to one pixel. Therefore, in the following description, a cell in a cell is also referred to as a “pixel”. The numerical value assigned to each pixel in the cell indicates the dot growth order. The dot growth order is the order of pixels in which writing is performed as the image density increases, and is set so that dots grow in order from the lowest number. Further, T (j, i) in FIG. 4 is a coordinate on the cell of each pixel, and a predetermined pixel in the cell is a reference pixel, and a coordinate of the reference pixel is T (0, 0). The coordinates T (j, i) indicate how far each pixel in the cell is from the reference pixel. In the example shown in FIG. 4, a pixel whose dot growth order is 0 is used as a reference pixel. In addition, this reference pixel is assumed to be aligned with the x-axis and y-axis origin on the image.

  Returning to FIG. 3, the noise matrix holding unit 242 holds a noise matrix in which a plurality of noise values are arranged in a matrix with predetermined noise characteristics. In the present embodiment, as the noise characteristic, a characteristic that a spatial frequency is high and periodicity is hardly visually perceived is adopted. In the present embodiment, blue noise characteristics are used as noise characteristics. FIG. 5 shows an example of a noise matrix having a blue noise characteristic of 32 × 32 pixels. In this noise matrix, a plurality (32 × 32 = 1024) of noise values from 0 to 1023 are arranged in order so that the spatial frequency becomes higher. In FIG. 5, N (m, n) is a coordinate on the noise matrix, and the coordinate of the upper left element (pixel) of the noise matrix is N (0,0).

  Returning to FIG. 3, the threshold value calculation unit 243 acquires the dot growth order corresponding to the processing target pixel that is the target pixel in the image from the cell information holding unit 241, and the processing target pixel from the noise matrix holding unit 242. Get the noise value corresponding to. Then, the threshold value calculation unit 243 calculates a threshold value corresponding to the processing target pixel based on the acquired dot growth order and the noise value. The threshold value calculation process will be described in detail later. The threshold value calculation unit 243 can create the threshold value matrix used for the halftone process as a result by calculating the above threshold value for all the pixels in the image. The comparison unit 244 compares the gradation value of the processing target pixel with the threshold corresponding to the processing target pixel calculated by the threshold calculation unit 243. Then, the comparison unit 244 turns on the print signal of the pixel to be processed when the gradation value is greater than or equal to the threshold value, and turns off the print signal when the gradation value is less than the threshold value. The above processing is applied to all pixels in the image.

(Threshold calculation processing procedure)
FIG. 6 is a flowchart illustrating an example of a threshold calculation processing procedure executed in the threshold calculation unit 243. The process of FIG. 6 is started when a digital image signal is input from the image input unit 10 to the image processing unit 20, for example. However, the start timing of the process of FIG. 6 is not limited to the above timing.
First, in S1, the threshold value calculation unit 243 acquires the dot growth order Dn corresponding to the processing target pixel. In the case of the 10-pixel cell shown in FIG. 4, each cell is arranged on the image as shown in FIG. As described above, the cells arranged on the image are regularly arranged in a predetermined direction, thereby forming a screen angle indicating the direction in which the dot pattern is arranged. The screen angle is represented by an angle with respect to the horizontal axis (x axis) on the image, and can be represented by vectors Rm and Rn shown in FIG. The vector Rm is a vector directed to the reference pixel P1 in another cell that is closest to the reference pixel P0 in a certain cell. Further, the vector Rn is a vector directed to the reference pixel P2 located closest to the reference pixel P0 on a straight line that is not parallel to the vector Rm. In the example shown in FIG. 7, it is Rm (3, -1), Rn (1, 3).

In S <b> 1 of FIG. 6, the threshold value calculation unit 243 performs processing for obtaining the dot growth order Dn corresponding to the processing target pixel based on the coordinates I (x, y) on the processing target pixel image. The dot growth order of each pixel set by the 10-pixel cell arranged on the image is such that the basic matrix shown in FIG. 8 is shifted by 7 pixels in the x-axis direction every time one pixel moves in the y-axis direction. It can also be expressed by placing them in The numerical values assigned to each pixel of the basic matrix shown in FIG. 8 indicate the dot growth order. Further, t (k) in FIG. 8 is a pixel number on the basic matrix, and can be expressed by k = (x + y × 7)% 10 (% is a remainder calculation).
Therefore, the dot growth order Dn of the processing target pixel can be derived by obtaining the above t (k) based on the coordinates I (x, y) of the processing target pixel. For example, when the coordinates of the pixel to be processed are I (6, 3) shown in FIG. 7, since t (k) = t (7), the dot growth order Dn = 2.

In S <b> 2 of FIG. 6, the threshold value calculation unit 243 acquires a noise value Nz corresponding to the processing target pixel. In S2, the threshold value calculation unit 243 normalizes the noise matrix based on the screen angle, and among the cells arranged with the screen angle, the cell including the pixel corresponding to the processing target pixel is included in the noise matrix. A noise value Nz is assigned. Then, the noise value Nz assigned to the cell is set as the noise value Nz corresponding to the processing target pixel. That is, the same noise value Nz is assigned to each of a plurality of pixels (10 pixels in this embodiment) in the pixel region corresponding to one cell. First, the threshold value calculation unit 243 obtains coordinates N (m, n) on the noise matrix of the cell corresponding to the pixel region including the processing target pixel, based on the coordinates I (x, y) of the processing target pixel. The threshold calculation unit 243 obtains coordinates N ′ (x, y) obtained by translating the processing target pixel I (x, y) to the reference pixel on the corresponding cell.
N '= IT ......... (1)

Next, the threshold calculation unit 243 obtains an angle θ1 of the vector Rm with respect to the x-axis direction of the image, an angle θ2 of the vector Rn with respect to the x-axis direction, and an angle θ3 of the vector Rn with respect to the vertical direction of the vector Rm. In the case of the example shown in FIG. 7, since the vector Rn matches the vertical direction of the vector Rm, the angle θ3 = 0. An example in which the angle θ3 is not 0 degrees is shown in FIG. In FIG. 9, a straight line A is a straight line that matches the x-axis direction, and a straight line B is a straight line that matches the vertical direction of the vector Rm.
The angles θ1 to θ3 can be calculated based on the following expression based on the coordinates (a1, b1) of the reference pixel P1 and the coordinates (a2, b2) of the reference pixel P2.
θ1 = tan ⁻¹ (b1 / a1),
θ2 = tan ⁻¹ (b2 / a2),
θ3 = π / 2− (| θ1 | + | θ2 |) (2)
From the above equation (2), the angle θ1 is π / 2 when a1 = 0, and the angle θ2 is π / 2 when a2 = 0. Here, the reference pixel P1 is a reference pixel located closest to the reference pixel P0 of a certain cell. That is, the reference pixel P0 and the reference pixel P1 are located on the same straight line in the screen angle direction. The reference pixel P2 is a reference pixel located closest to the reference pixel P0 on a straight line that intersects with a straight line connecting the reference pixel P0 and the reference pixel P1.

Further, based on the coordinates (a1, b1) of the reference pixel P1 and the coordinates (a2, b2) of the reference pixel P2, the threshold calculation unit 243 calculates the length Rx of the vector Rm and the vector Rn based on the following equations: The length Ry is obtained.
Rx = √ (a1 ² + b1 ² ),
Ry = √ (a2 ² + b2 ² ) (3)
Then, the threshold value calculation unit 243 performs processing for converting the coordinates N ′ (x, y) into the coordinates N (m, n) on the noise matrix using the parameters θ1, θ3, Rx, and Ry. Specifically, the rotational deformation at the angle (−θ1), the shear deformation at the angle (−θ3), and the enlargement / reduction deformation of the lengths Rx and Ry are performed on the coordinates N ′ (x, y). However, when the angle θ3 = 0, shear deformation is not necessary.

Through the above processing, the coordinates N (m, n) on the noise matrix corresponding to the processing target pixel can be obtained, and as a result, the noise value Nz corresponding to the processing target pixel can be obtained. For example, when the coordinate of the processing target pixel is I (6, 3), the coordinate on the noise matrix corresponding to the processing target pixel is N (2, 0), and the noise value Nz corresponding to the processing target pixel is 394. It becomes.
The above equation (4) is based on the fact that the noise matrix can be normalized based on the screen angle by performing deformation processing on the noise matrix in the order of enlargement / reduction of Rx and Ry, shear of θ3, and rotation of θ1. . Here, normalization of the noise matrix means performing the following two processes. First, the first processing consists of two intersecting axes that indicate the cell arrangement direction set based on the screen angle, with the spacing between elements in the direction of two orthogonal axes (m-axis, n-axis) of the noise matrix. This is a process of matching the cell arrangement interval in the direction of. This first process corresponds to the deformation process of enlargement / reduction of Rx and Ry. The second process is a process of matching two orthogonal axes of the noise matrix with two intersecting axes indicating the cell arrangement direction. This second process corresponds to the deformation process of the shear of θ3 and the rotation of θ1.

  As described above, the processing of the above equation (4) is performed by normalizing the noise matrix based on the screen angle and corresponding to each cell corresponding to each element of the noise matrix arranged with the screen angle. This is equivalent to the process of assigning the noise value Nz. At this time, all noise values on the noise matrix are assigned to the corresponding cells. Therefore, noise can be set in all cells arranged on the image while maintaining the continuity of the blue noise characteristics.

  Further, in the above normalization, one of the two intersecting axes of the cell is set to a predetermined dot growth order (in the above example, dot growth order = 0) in the cells arranged with a screen angle. Among the plurality of points (pixels), the axis passes through the two closest points. Further, the other axis of the two intersecting axes is positioned closest to the one point on the straight line that intersects one of the two points and the axis passing through the two points among the plurality of points. The axis passing through the point to be That is, two axes having a relationship close to vertical are set as the two axes indicating the cell arrangement direction. Thus, when normalizing the noise matrix, the coordinate axis of the noise matrix can be converted into a coordinate axis having a relationship close to vertical determined according to the screen angle. Since the noise matrix is designed with two coordinate axes that are orthogonal to each other, it is easier to exhibit noise characteristics when converted to a coordinate axis that is nearly perpendicular.

  In S3 of FIG. 6, the threshold calculation unit 243 calculates a threshold corresponding to the processing target pixel based on the dot growth order Dn acquired in S1 and the noise value Nz acquired in S2. Hereinafter, a case where the gradation value of the digital image signal input to the halftone processing unit 24 is expressed by 256 gradations from 0 to 255 will be described as an example. As described above, the comparison unit 244 outputs a print signal for forming dots when the gradation value ≧ the threshold value. Therefore, when processing an image with uniform gradation values, if the threshold values 1 to 255 are uniformly distributed, the average gradation value of the entire image after halftone processing is the input image. Gradation value. Therefore, in this embodiment, the threshold value calculation unit 243 calculates the threshold value corresponding to the processing target pixel as follows in order to uniformly distribute the threshold values corresponding to the respective gradation values while maintaining the screen angle.

First, the threshold range (1 to 255) is divided by the number of thresholds set by one cell (the number of pixels in the cell = 10), and the inter-threshold distance THW is obtained.
THW = (255−1) / 10 (5)
Next, the noise value Nz acquired in S2 is normalized to a noise value that vibrates by the distance between thresholds THW, and the result is set as a variation amount THN.
THN = (Nz / (1024-1)) × THW (6)
As represented by the above equation (6), the fluctuation amount THN is obtained by dividing the noise value Nz by the number of noises and multiplying by the inter-threshold distance THW. The fluctuation amount THN becomes a value between 0 and THW, and takes a larger value as the noise value Nz is larger.

Finally, the threshold value TH corresponding to the processing target pixel is calculated using the inter-threshold distance THW, the fluctuation amount THN, and the dot growth order Dn. In the following expression, the threshold value TH is obtained by rounding off after the decimal point.
TH = THW × Dn + THN + 1 (7)
In the above equation (7), (THW × Dn) is a basic threshold set based on the dot growth order Dn. The basic threshold is a basic value of a threshold at which dots start to grow, and takes a larger value as the dot growth order Dn increases. Thus, the threshold value TH is calculated by adding the fluctuation amount THN determined based on the noise value Nz to the basic threshold value. Here, the maximum value of the variation THN is an inter-threshold distance THW that is a difference value between the basic threshold corresponding to the processing target pixel and the basic threshold corresponding to the pixel for which the next dot growth order of the processing target pixel is set. is there.

  Therefore, the threshold value TH calculated as described above is an integer from 1 to 255. Then, as a result of calculating the threshold value TH for all the pixels on the image and creating the threshold value matrix, the threshold values (1 to 255) are uniformly distributed over the entire image and appear with equal probability. Therefore, it is possible to faithfully convert each gradation included in the input image into an area gradation. That is, each gradation value of the input image can be converted into an area gradation while maintaining the density distribution. Furthermore, while this threshold value is set in a cell while maintaining the distance between the threshold values in the order of dot growth, the threshold value changes with a blue noise characteristic for each cell. Therefore, it is possible to obtain a threshold value matrix capable of realizing good dot dispersion and high gradation reproducibility over the entire image.

As described above, in this embodiment, the cell information holding unit 241 holds a cell, and the noise matrix holding unit 242 holds a noise matrix. Here, the cell has a plurality of elements in which the dot growth order of each pixel in a predetermined pixel area in the image data is set, and is arranged so as to have a predetermined screen angle. The noise matrix is a matrix in which a plurality of noise values are arranged in a matrix with predetermined spatial frequency characteristics.
Then, the threshold calculation unit 243 sets a corresponding noise value for each cell corresponding to each element of the noise matrix arranged so as to have a screen angle. The threshold calculation unit 243 also performs halftone processing based on the growth order of dots corresponding to the processing target pixel (target pixel) and the noise value set for the cell corresponding to the pixel region including the processing target pixel. A threshold to be compared with the gradation value of the processing target pixel is calculated (derived) in the processing.

  Therefore, it is possible to create a threshold matrix to which high tone reproducibility can be realized, to which noise having a predetermined spatial frequency characteristic is added. In particular, it is possible to reproduce low-density thin lines and improve image quality. In addition, the noise value set for each element of the noise matrix can be assigned to each number of cells equal to the number of elements of the noise matrix arranged with a screen angle, making full use of noise characteristics. Gradation reproducibility can be improved. Furthermore, since the continuity of noise can be maintained over the entire image, it is possible to prevent the occurrence of low frequency components and to prevent deterioration in image quality.

  Further, since the threshold value is calculated based on the dot growth order and the noise value, it is possible to adjust the appearance probability of each threshold value corresponding to each gradation value that can be taken by the pixel of the image data while utilizing the characteristics of noise. Therefore, the gradation can be appropriately reproduced. Furthermore, the threshold value can be calculated simply by retaining only the dot growth order and the noise matrix. Thus, since it is not necessary to hold a threshold matrix having a large size of 256 × 256 such as a blue noise mask, the amount of data to be held can be suppressed.

  Further, the threshold calculation unit 243 calculates (derived) a basic threshold that increases as the dot growth order increases, based on the dot growth order corresponding to the processing target pixel. Further, the threshold value calculation unit 243 calculates (derived) a variation amount with respect to the basic threshold value corresponding to the processing target pixel based on the noise value set for the cell corresponding to the pixel region including the processing target pixel. Then, the threshold calculation unit 243 adds (calculates) a threshold corresponding to the processing target pixel by adding the calculated basic threshold and the calculated variation. Therefore, the fluctuation range (noise amplitude) of the basic threshold can be easily adjusted, and a threshold having an appropriate appearance probability can be easily calculated.

In addition, the threshold value calculation unit 243 determines the maximum value of the variation amount with respect to the basic threshold value as the difference between the basic threshold value corresponding to the processing target pixel and the basic threshold value corresponding to the pixel in which the growth order of the next dot of the processing target pixel is set. Value. That is, the range of fluctuation of the basic threshold is set to the distance between thresholds (THW). Therefore, the threshold value calculation unit 243 can calculate a threshold value for appropriately reproducing each gradation value that the image data can take.
Further, the threshold value calculation unit 243 calculates the threshold value so that the threshold value corresponding to each gradation value appears with equal probability in the threshold value matrix. Therefore, each gradation value of the input image can be converted into area gradation while maintaining the density distribution.

  The threshold calculation unit 243 normalizes the noise matrix based on the screen angle, and assigns each noise value on the noise matrix to each cell arranged to have the screen angle. Specifically, the spacing between the elements in the direction of the two orthogonal axes orthogonal to each other in the noise matrix matches the array spacing in the direction of the two intersecting axes indicating the cell arrangement direction set based on the screen angle. Let Further, the threshold value calculation unit 243 makes the directions of the two orthogonal axes of the noise matrix coincide with the directions of the two intersecting axes of the cell. As a result, the deformation process can be applied to the noise matrix in the order of the above-described enlargement / reduction of Rx and Ry, shear at the angle θ3, and rotation at the angle θ1, and can be appropriately normalized to the screen angle.

Further, the threshold value calculation unit 243 has one axis of two intersecting axes as a closest position among a plurality of points set with a predetermined dot growth order in a plurality of cells arranged with a screen angle. An axis passing through two points in relation. Further, the threshold calculation unit 243 sets the other axis of the two intersecting axes as one of the two points and the two points among the plurality of points set with the predetermined dot growth order. Let it be an axis passing through a point closest to the one point on the straight line intersecting the passing axis. Therefore, two orthogonal axes of the noise matrix can be converted into two axes that are close to the vertical among the axes indicating the cell arrangement direction. Since the noise matrix is designed with orthogonal coordinate axes, it is possible to easily exhibit noise characteristics by converting to a coordinate axis close to vertical.
The noise matrix holding unit 242 holds a noise matrix having a spatial frequency characteristic in which periodicity is difficult to visually perceive. For example, the noise matrix holding unit 242 holds a noise matrix having blue noise characteristics. Therefore, it is possible to obtain good dot dispersibility and appropriate gradation reproducibility while appropriately suppressing noise.
Appropriate gradation conversion can be realized over the entire image by performing halftone processing on input multi-gradation image data using the threshold value calculated by the method described above.

(Second embodiment)
Next, a second embodiment will be described. In the first embodiment described above, the halftone processing method in which the halftone processing unit 24 faithfully reproduces the color of the image output from the density correction unit 23 with the area gradation has been described. In the second embodiment, a method for simultaneously performing density correction and halftone processing will be described.
FIG. 10 is a diagram illustrating a configuration example of the image processing unit 20 in the second embodiment. 10, parts having the same configurations as those of the image processing unit 20 in the first embodiment described above are denoted by the same reference numerals as those in FIG. 2, and the following description will focus on parts having different configurations.
The density / halftone processing unit 25 inputs the digital image signal (density signals C, M, Y, K) output from the color space conversion unit 22. The density / halftone processing unit 25 performs a density correction process and a halftone process on the input digital image signal (image data) at the same time, and converts the digital image signal into a halftone expression (print signal C). , M, Y, K). The density / halftone processing unit 25 calculates the threshold value by the same procedure as in the first embodiment shown in FIG. However, the threshold value calculation process corresponding to the process of S3 is different from that of the first embodiment.

First, the concept of performing density correction processing and halftone processing simultaneously will be briefly described. For example, if halftone processing is performed using a threshold matrix in which threshold values 1 to 255 appear with equal probability, each gradation value of the input image can be converted to area gradation while maintaining the density distribution. On the other hand, when halftone processing is performed using a threshold value matrix in which each threshold value is distributed so that the appearance rate decreases as the threshold value is closer to 1 or 255, the probability that the threshold value is lower than the threshold value decreases as the gradation value decreases. The higher the value, the higher the probability that the threshold value is exceeded. That is, fewer dots are generated with low gradation and more dots are generated with high gradation. Therefore, it is possible to generate an image with increased contrast, in other words, an image with density correction.
Therefore, in the present embodiment, density correction is performed simultaneously with the halftone process by setting each threshold value to appear with a predetermined probability that is not equal. When density correction is performed to increase the contrast as described above, the threshold value TH is calculated based on the following equation.

In the above equation, the threshold value TH is obtained by rounding off after the decimal point.
The threshold TH of the processing target pixel calculated in this way is less likely to appear as the threshold value is lower and the threshold value is higher, and the intermediate threshold value is likely to appear. Therefore, the density correction for increasing the contrast can be performed simultaneously with the halftone process. In addition, while changing the threshold value for each cell, the change can be changed with blue noise characteristics, so it is possible to achieve density correction and halftone processing with high dot dispersion while maintaining the screen angle. It is.
In the above example, the case of increasing the contrast as the density correction has been described. However, it is also possible to reduce the contrast similarly.

As described above, in the present embodiment, the threshold value calculation unit 243 derives a threshold value so that the threshold value corresponding to each gradation value appears with a predetermined probability in the threshold value matrix. Thereby, the density correction can be performed simultaneously with the halftone process.
For example, the threshold value calculation unit 243 derives the threshold value so that the threshold value corresponding to the gradation value closer to the minimum gradation value (0) or the maximum gradation value (255) appears with a lower probability in the threshold value matrix. . As a result, it is possible to simultaneously realize density correction and halftone processing for increasing the contrast. It is also possible to derive the threshold value so that the threshold value corresponding to the gradation value closer to the minimum gradation value (0) or the maximum gradation value (255) appears with higher probability in the threshold value matrix. Thereby, density correction and halftone processing for reducing contrast can be realized simultaneously.

(Third embodiment)
Next, a third embodiment will be described. In the first and second embodiments described above, the method of calculating the threshold value using a calculation formula including division has been described. In the third embodiment, a method for calculating a threshold value using a table without using division will be described.
The image processing unit 20 in the third embodiment has the same configuration as the image processing unit 20 in the second embodiment shown in FIG. 10 described above. However, the threshold value calculation process corresponding to S3 in FIG. 6 is different from the second embodiment. Therefore, the following description will focus on the different parts of the process.
In the threshold value calculation process in the present embodiment, the basic threshold value Ti determined according to the dot growth order Dn and the variation amount Mt to be added to the basic threshold value Ti are derived using a preset threshold value table. Then, by adding the derived basic threshold value Ti and the variation amount Mt, a threshold value TH corresponding to the processing target pixel is calculated (TH = Ti + Mt).

  First, the density / halftone processing unit 25 refers to the threshold table (a) shown in FIG. 11 and derives a basic threshold Ti corresponding to the dot growth order Dn of the processing target pixel. The basic threshold Ti represents a basic threshold value at which dots start to grow. In order to simplify the explanation, FIG. 11 shows an example in the case where the density is uniformly converted into area gradation. That is, in the threshold value table (a), Ti = 1 when the dot growth order Dn = 0, Ti = 26 when the dot growth order Dn = 1, and Ti = 52 when the dot growth order Dn = 2. The basic threshold value Ti is uniformly distributed within a threshold value range of 1 to 255.

Next, the density / halftone processing unit 25 refers to the threshold tables (b0) to (b9) illustrated in FIG. 11 and derives the variation amount Mt corresponding to the noise value Nz of the processing target pixel. In the threshold tables (b0) to (b9), a noise threshold value Nt for deriving the fluctuation amount Mt based on the noise value Nz is set. Here, in order to simplify the description, an example in which the noise threshold value Nt is uniformly distributed within the range of 0 to 1023 noise values is shown. For example, when the dot growth order Dn = 0, the threshold value table (b0) is referred to. When the noise value Nz is less than 40, the fluctuation amount Mt = 0, and when the noise value Nz is 40 or more and less than 81, the fluctuation amount. Set Mt = 2. That is, in the threshold value tables (b0) to (b9) illustrated in FIG. 11, the noise threshold value Nt is set by dividing the number of noises into equal intervals by the distance THW between the threshold values.
In this way, in the example shown in FIG. 11, the threshold values TH are uniformly distributed by arranging the basic threshold value Ti and the noise threshold value Nt at equal intervals. When the distribution of the noise threshold value Nt in the threshold value table shown in FIG. 11 is expressed in another form, it is as shown in FIG. As shown in FIG. 12, by setting the noise threshold value Nt to be uniformly distributed with respect to each gradation value, each threshold value is set to appear with an equal probability in the image. As a result, it is possible to perform processing for faithfully converting each gradation value of the input image into an area gradation.

On the other hand, by changing the distribution of the threshold value TH obtained by adding the basic threshold value Ti and the fluctuation amount Mt, halftone processing can be performed while performing density correction. When increasing the contrast, many threshold values are assigned to the lower dot growth order Dn so that it is difficult to generate dots at low density. Then, many threshold values are assigned to the higher dot growth order Dn so that dots are easily generated at a high density. The distribution of the noise threshold value Nt in this case is shown in FIG. As shown in FIG. 13, by setting the threshold value table so that each threshold value appears with a predetermined probability, density correction and halftone processing can be performed simultaneously.
Similarly, the contrast can be lowered by changing the assignment of the threshold value. When the contrast is lowered, a smaller threshold is assigned to the lower dot growth order Dn so that dots are more easily generated at a low density. Also, a smaller threshold is assigned to the higher dot growth order Dn so that dots are less likely to be generated at high densities.

As described above, by changing the setting of the basic threshold value Ti and the noise threshold value Nt, density correction and halftone processing can be performed simultaneously, and a desired tone curve can be set. In the above example, the threshold table (a) for deriving the basic threshold Ti based on the dot growth order Dn and the threshold tables (b0) to (b9) for deriving the variation Mt based on the noise value Nz. ) Has been described as a separate table. However, the form of the threshold table is not limited to the form shown in FIG. 11 and can be set as appropriate.
The threshold value tables (b0) to (b9) shown in FIG. 11 are set so as to derive the fluctuation amount Mt of 0 to 24 according to the noise value Nz. That is, in the example shown in FIG. 11, the maximum value of the variation Mt with respect to each basic threshold Ti is set equal. However, when the maximum value of the threshold value TH is not divisible by the number of basic threshold values Ti (the number of pixels in the cell), the maximum value of the variation amount Mt for each basic threshold value Ti may be set to a different value.

As in the example shown in FIG. 11, when the maximum value of the threshold TH is 255 and the number of basic thresholds Ti (the number of pixels in the cell) is 10, the maximum value of the threshold TH is the number of basic thresholds Ti (cells The number of pixels is not divisible. In this case, the distance between the thresholds of each basic threshold Ti is not strictly equal. For example, as shown in the threshold value table (a0), the difference (inter-threshold distance) between the basic threshold value Ti corresponding to the dot growth order Dn = 0 and the basic threshold value Ti corresponding to the dot growth order Dn = 1 is 25. . On the other hand, the difference (inter-threshold distance) between the basic threshold value Ti corresponding to the dot growth order Dn = 1 and the basic threshold value Ti corresponding to the dot growth order Dn = 2 is 26. Therefore, in the above case, the threshold value table (b0) referred to when the dot growth order Dn = 1 may be used as a table for deriving the fluctuation amount Mt from 0 to 25.
As described above, in the present embodiment, the density / halftone processing unit 25 includes the threshold table indicating the correspondence between the dot growth order Dn and the basic threshold Ti, and the threshold indicating the correspondence between the noise value Nz and the variation Mt. Hold the table. Then, the threshold value calculation unit 243 derives the basic threshold value Ti and the fluctuation amount Mt with reference to the above threshold value table. Therefore, the threshold value can be derived without requiring a complicated calculation. It is also easy to adjust the threshold appearance probability.

(Modification)
In each of the above embodiments, the case where the blue noise characteristic is adopted as the noise characteristic has been described, but the noise characteristic is not limited to the blue noise characteristic. The noise characteristic may be a noise characteristic having a spatial frequency characteristic in which periodicity is hardly visually perceived, and a green noise characteristic or the like may be adopted. Further, the noise characteristics may be switched according to the density. For example, a halftone image showing blue noise characteristics on the highlight side and white noise characteristics on the shadow side may be generated. As a result, graininess in the highlight area and streaky unevenness (banding unevenness) in the shadow area can be suppressed.

In each of the above embodiments, the case where the noise matrix is 32 × 32 pixels has been described. However, the size of the noise matrix can be set as appropriate. Furthermore, although the case where the number of pixels in the cell is 10 has been described, the size of the cell can also be set as appropriate.
(Other embodiments)
The present invention supplies a program that realizes one or more functions of the above-described embodiments to a system or apparatus via a network or a storage medium, and one or more processors in a computer of the system or apparatus read and execute the program This process can be realized. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

  DESCRIPTION OF SYMBOLS 10 ... Image input part, 20 ... Image processing part, 21 ... Input correction part, 22 ... Color space conversion part, 23 ... Density correction part, 24 ... Halftone processing part, 25 ... Density / halftone processing part, 30 ... Image Output unit, 100 ... printing device, 241 ... cell information holding unit, 242 ... noise matrix holding unit, 243 ... threshold value calculating unit, 244 ... comparison unit

Claims (14)

A first holding means for holding cells arranged to have a predetermined screen angle, having a plurality of elements in which the growth order of dots of each pixel constituting a predetermined pixel area in image data is set;
A second holding means for holding a noise matrix in which a plurality of noise values are arranged in a matrix with predetermined spatial frequency characteristics;
Setting means for setting the corresponding noise values for a plurality of the cells respectively corresponding to the elements of the noise matrix, arranged so as to have the screen angle;
Based on the growth order of the dots corresponding to the pixel of interest and the noise value set for the cell corresponding to the pixel region including the pixel of interest by the setting means, the pixel of interest by halftone processing Deriving means for deriving a threshold value to be compared with the gradation value of
An image processing apparatus comprising: The derivation means includes
First deriving means for deriving a basic threshold value that increases as the growth order of the dots increases, based on the growth order of the dots corresponding to the target pixel;
Second derivation means for deriving a variation amount with respect to the basic threshold corresponding to the target pixel based on the noise value set for the cell corresponding to the pixel region including the target pixel by the setting means;
A third deriving unit for deriving the threshold by adding the basic threshold derived by the first deriving unit and the fluctuation amount derived by the second deriving unit. The image processing apparatus according to claim 1, wherein: The second derivation means includes
The maximum value of the variation amount is a difference value between the basic threshold value corresponding to the target pixel and the basic threshold value corresponding to a pixel in which the growth order of the next dot of the target pixel is set. The image processing apparatus according to claim 2. A third holding means for holding a first table indicating a correspondence between the growth order of the dots and the basic threshold, and a second table indicating a correspondence between the noise value and the variation amount;
The first deriving unit derives the basic threshold with reference to a first table held by the third holding unit,
The image processing apparatus according to claim 2, wherein the second deriving unit derives the variation amount with reference to a second table held by the third holding unit.   The derivation means derives the threshold value so that the threshold value corresponding to each gradation value that can be taken by the pixel of the image data appears with a predetermined probability. The image processing apparatus according to item 1.   6. The image processing apparatus according to claim 5, wherein the derivation unit derives the threshold value so that the threshold value corresponding to each gradation value that can be taken by a pixel of the image data appears with an equal probability. .   The deriving means derives the threshold value so that the threshold value corresponding to the minimum gradation value or the gradation value close to the maximum gradation value that can be taken by the pixel of the image data appears with a lower probability. The image processing apparatus according to claim 5.   The derivation means derives the threshold value so that the threshold value corresponding to the minimum gradation value or the gradation value close to the maximum gradation value that can be taken by the pixel of the image data appears with higher probability. The image processing apparatus according to claim 5. The setting means includes
The spacing between each element of the noise matrix in the direction of two orthogonal axes orthogonal to each other of the noise matrix is set in the direction of two intersecting axes indicating the arrangement direction of the cells set based on the screen angle. Match the cell spacing,
By matching the direction of the two orthogonal axes with the direction of the two intersecting axes,
The image processing apparatus according to claim 1, further comprising a normalizing unit that normalizes the noise matrix based on the screen angle. The normalizing means includes
Two points having the closest positional relationship among a plurality of points in which the growth order of a predetermined dot in the plurality of cells arranged with the screen angle is set as one axis of the two intersecting axes. The axis passing through
The other axis of the two intersecting axes is positioned closest to the one point on the straight line intersecting one of the two points and an axis passing through the two points among the plurality of points. The image processing apparatus according to claim 9, wherein the image processing apparatus has an axis passing through the point. The second holding means is
The image processing apparatus according to claim 1, wherein the noise matrix having a spatial frequency characteristic in which periodicity is difficult to visually perceive is held. Image input means for inputting the multi-tone image data;
Image processing means for performing the halftone process on the image data input by the image input means using the threshold value derived by the derivation means;
The image processing apparatus according to claim 1, further comprising: It has a plurality of elements in which the growth order of dots of each pixel constituting a predetermined pixel area in image data is set, holds cells arranged to have a predetermined screen angle, and has a predetermined spatial frequency characteristic Holds a noise matrix in which multiple noise values are arranged in a matrix,
Setting the corresponding noise value for each of the plurality of cells corresponding to each element of the noise matrix arranged to have the screen angle;
Based on the growth order of the dots corresponding to the pixel of interest and the noise value set for the cell corresponding to the pixel region including the pixel of interest by the setting means, the pixel of interest by halftone processing Deriving a threshold value to be compared with the tone value of
An image processing method comprising:   The program for functioning a computer as each means of the image processing apparatus of any one of Claims 1-12.

Images