JP3917634B2 - How to create a mesh screen - Google Patents

Description

  The present invention relates to a method for creating a mesh screen comprising a threshold matrix for converting a multi-value image signal into a binary image signal.

  A so-called halftone dot conversion method for converting a multi-value image signal into a binary image signal is an amplitude modulation method (also referred to as “AM method”) in which the size of a dot (pixel) is modulated in accordance with the multi-value image signal. The frequency can be roughly divided into two types: a frequency modulation method (also referred to as “FM method”) in which the dot size is the same and the density is modulated in accordance with the multi-value image signal. In this case, in the frequency modulation method, when a multi-value image signal is converted into a binary image signal using a predetermined threshold, an error that occurs between the multi-value image signal and the binary image signal is surrounded by the error. An error diffusion method for generating a binary image signal having a desired density by diffusing to pixels, a threshold method for converting a multi-value image signal into a binary image signal using a threshold set as a dither matrix, and the like. is there.

  By the way, when a conventional threshold method using a distributed dither matrix (Bayer dither matrix) corresponding to each region of an image divided into a plurality of blocks, an unsightly regular pattern is visually recognized. There is. In addition, the conventional threshold method does not consider the influence of dot gain due to an output device or an output medium that outputs an image. Therefore, when the dot gain is large, the image quality deteriorates. .

  On the other hand, among the arrangements in the row direction and the column direction of the dither matrix, the arrangement in one direction is set larger than the arrangement in the other direction, and the arrangement in the other direction is randomized for each block of the divided image. There is a method of reducing the memory capacity required for the dither matrix while reducing the visual recognition of the regular pattern generated by the repetition of the dither matrix by repeatedly rearranging (see Patent Document 1).

  However, even in this method, the regular pattern cannot be completely eliminated, and as in the case of using a Bayer dither matrix, deterioration in image quality due to the influence of dot gain depending on the output device is avoided. It is not possible. Further, since it is necessary to repeat the rearrangement of the dither matrix arrangement in the other direction, more processing time is required than when a Bayer dither matrix is used.

  In addition, when an error diffusion method is used, when converting a multi-value image signal to a binary image signal, a value corresponding to a dot generation pattern (binarization pattern) already determined in peripheral pixels of the pixel is 2 There is a method of correcting an error due to the digitization using a correction value corresponding to the dot generation pattern stored in the pattern table memory, and performing error diffusion processing using the corrected error as a new error (Patent Document 2). reference).

  In this case, it is possible to accurately set a correction value for an ideal dot generation pattern that can be patterned, for example, a circular dot generation pattern, and thus a good image can be reproduced. However, it cannot be applied to all output devices and output media having various dot gain characteristics. Moreover, since it is an error diffusion method, there is also a problem that requires a considerable processing time.

Japanese Patent Application Laid-Open No. 6-313152 JP-A-5-3540

  The present invention has been made to solve the above-described problems, and provides a method for creating a mesh screen capable of generating a high-quality image in consideration of visual characteristics, an output device, and dot gain of an output medium. For the purpose.

The present invention relates to a method for creating a mesh screen comprising a threshold matrix for converting a multi-value image signal into a binary image signal.
When determining threshold values in ascending order or descending order, a plurality of different weighting values are provided in a predetermined range centered on the position of each already determined threshold value, including other threshold value matrices having the same configuration arranged around the threshold value matrix. The threshold value matrix and the threshold value matrix and the threshold value matrix and the other threshold value matrixes arranged around the threshold value matrix are actuated by adding or subtracting the weight values for each position. By obtaining a weighting distribution by the weighting filter in another threshold value matrix arranged around the threshold value, and setting the position corresponding to the minimum value or the maximum value of the weighting distribution as a new threshold value position, the position of the threshold value sequentially It is characterized by determining.

  In the present invention, the spatial frequency components of the binarization error by this threshold value matrix can be set to the high frequency side by determining the threshold positions in the order of the most distance from each other. In addition, in the determination, the visibility of periodicity due to repetition of the threshold matrix can be reduced by taking into account the positions of the thresholds constituting the surrounding threshold matrix.

  Further, in the present invention, for example, a point spread function considering visual characteristics is used as a weighting filter, and the variation of the weighting distribution obtained by operating this weighting filter around a predetermined threshold position is minimized. By sequentially determining the threshold position, it is possible to obtain a threshold matrix that can reduce the visibility of annoying graininess in the image. Further, by determining the threshold value in consideration of the surrounding threshold value matrix, it is possible to reduce the visibility of the periodicity of the image caused by the discontinuity of the surrounding threshold value.

  In the present invention, by setting the threshold values in the order in which the distances between the threshold values constituting the threshold value matrix are separated, it is possible to obtain a good binary image in which the dot pattern is not offset. Further, since the threshold values are also considered in determining the threshold value, periodicity due to repetition of the threshold matrix does not appear on the created binary image.

  In the present invention, when the threshold value is determined, for example, a point spread function obtained from visual characteristics is used as a weighting filter, and this weighting filter is operated centering on the position of the predetermined threshold value. The threshold value is sequentially determined from the portion where the weight distribution obtained by subtraction is minimized. In this case, it is possible to obtain a good binary image in which the graininess and the periodic pattern are not conspicuous.

  Further, a correction amount for the dot gain is obtained by creating a test pattern using a desired output device or a desired output medium, and after correcting the point spread function with the correction amount, the threshold matrix is set as described above. As a result, it is possible to correct the dot gain and create a more accurate binary image.

  FIG. 1 shows a system to which a method for creating a mesh screen according to the present invention is applied. In this system, first, a procedure for creating the printed matter P will be described.

  The multi-value image signal I supplied from the external device to the raster image processor (RIP) 10 is selected by the selection unit 14 according to the output conditions such as the number of screen lines and the output medium, and output from the threshold matrix storage unit 16. Is compared with the threshold signal TH to be converted into a binary image signal B. The threshold value signal TH is created in the threshold value matrix creating device 22 as described later, but the threshold value matrix creating device 22 may be integrated with the RIP 10. Next, the binary image signal B is supplied to the output device 20, and for example, a halftone dot image to be a film original is formed on the film F by on / off control of the laser beam. A printing plate is created from the film original plate, and the printing plate is set in a printing machine, whereby a desired printed matter P is formed on a predetermined printing paper as a recording medium.

  Next, each embodiment of a method for creating a threshold value matrix (mesh screen) used when creating the film F or printed matter P as described above will be described. In each of the following embodiments, it is described that a threshold value of 1 to 25 is set. However, it is needless to say that the threshold value can be extended when a threshold value having an arbitrary range is set.

Example 1
2 to 4 are flowcharts illustrating the processing procedure of the first embodiment. In this embodiment, by setting a threshold value matrix so that threshold values close to each other are not arranged close to each other, the dot pattern of the image formed by the threshold value matrix is not locally offset, thereby reducing the graininess. In addition, by arranging the threshold values in consideration of another threshold matrix set adjacent to the threshold matrix, appearance of a regular pattern due to repetition of the threshold matrix can be reduced.

First, as shown in FIG. 5, threshold doors with array T _5, the threshold phased array T ₅ the surrounding of the same configuration as the threshold phased array T ₁ through T ₄ is the desired threshold value matrix, T ₆ to define and ~T _9. Each of the 25 threshold values t (P _kl ) (k = 1 to 9, l = 1 to 25, t = 1 to 25) constituting each of the threshold arrays T _k (k = 1 to 9). Pixel point _Pkl is defined (step S100). The x and y coordinates of each pixel point P _kl are assumed to be x _kl and y _kl .

Therefore, after setting all threshold values t (P _kl ) of each threshold array T _k to 0 (step S102), two pixel points P _k1 and P _k2 that are appropriately separated from each threshold array T _k are obtained. The threshold values t (P _k1 ) = 1 and t (P _k2 ) = 2 are set as initial values (step S104, see FIG. 5).

Next, a pixel point P _kl for setting a threshold value t (P _kl ) from 3 to 25 is obtained.

First, z = 3,2 one maximum value Rmax = 0 of the distance between pixel points P _kl, and a random number RND = 0, i = 1 (step S106, S108, S112, S114, S116), the threshold value t (P _5i) Is 0 (step S120), the minimum value Rmin (i) of the distance between the pixel point P _5i and the surrounding pixel point P _kl for which the threshold value t (P _kl ) has been set is obtained ( Step S122).

The minimum value Rmin (i) can be obtained according to the flowchart shown in FIG. That is, after setting the initial value of the minimum value Rmin (i) to 99999 (step S150), a threshold value t (P _kl ) other than 0 is set through steps S152, S154, S156, S158, S160, and S162. extracting pixel points P _kl which are (step S164), obtains the distance RR between the pixel point P _kl and the pixel point P _5i according to the following equation (1) (step S166).
RR = √ ((x _kl −x _5i ) ² + (y _kl −y _5i ) ² ) (1)

Then, the distance RR obtained as described above is compared with the minimum value Rmin (i) (step S168), and if RR <Rmin (i), Rmin (i) = RR is set (step S170). By performing this process for all pixel points P _kl for which the threshold value t (P _kl ) of all threshold arrays T _k is not 0, the pixel point P _5i and the surrounding threshold value t (P _kl ) are already set. The minimum value Rmin (i) of the distance between the pixel point P _kl thus obtained can be obtained.

Next, returning to FIG. 3, the minimum value Rmin (i) obtained as described above is compared with the maximum value Rmax (steps S124 and S130). In this case, when i = 1, the maximum value Rmax is 0 (step S112). Therefore, Rmax = Rmin (i) is set (step S126), and x _5z = x _5i and y _5z = y _5i ( Step S128). Then, i is updated (step S116), the minimum value Rmin (i) is obtained again (step S122), and the minimum value Rmin (i) is compared with the maximum value Rmax updated in step S126 (step S124). , S130).

Here, as long as Rmin (i)> Rmax (step S124), the coordinates x _5z and y _5z for setting the threshold value t (P _5i ) = 3 (z = 3) are sequentially updated. That is, by this processing, coordinates x _5z and y _5z that are separated from any pixel point P _kl for which the threshold value t (P _kl ) has already been set can be obtained.

If it is determined in step S130 that Rmin (i) = Rmax, as shown in FIG. 6, pixels having the same distance from the pixel point _Pkl for which the threshold value t ( _Pkl ) has already been set are shown. There are two or more points P _5i (for example, when the pixel point P _{5i at} which the threshold value t (P _5i ) = 3 is two points 3A and 3B). In this case, a random number NEWRND having a value of 0 or 1 is obtained (step S132), the random number NEWRND is compared with the random number RND set in step S112 (step S134), and if NEWWRND> RND, RND = a NEWRND (step S136), the coordinates x _5i determined later, threshold _{y 5i t (P 5i) =} 3 (z = 3) coordinates to set the x _5z, and y _5z (step S138). If NEWWRND ≦ RND, the coordinates x _5i and y _5i obtained first are set as coordinates x _5z and y _5z for setting the threshold t (P _5i ) = 3 (z = 3). As a result, any one of 3A and 3B shown in FIG. 5 is randomly selected as coordinates x _5z and y _{5z at} which the threshold value t (P _5i ) = 3 (z = 3) should be set.

In this way, among all the pixel points P _5l of the threshold array T ₅ , the pixel point P that is located farthest from the other pixel points P _kl for which the threshold value t (P _kl ) has already been set. _5Z can be extracted. Therefore, a threshold value t (P _kz ) = 3 (z = 3) is set for the pixel point P _{kz of} each threshold array T _k including the pixel point P _5Z (steps S118 and S140).

Similarly, by repeating the processing from step S108 and setting threshold value t (P _kl ) = 4 to 25, threshold value t (P) for all pixel points P _kl in threshold _array T _k . _kl ) is set (step S110).

From the threshold array T _k determined as described above, only the central threshold array T ₅ is extracted and set as a threshold matrix in the threshold matrix storage unit 16 shown in FIG.

Therefore, in the RIP 10, the selection means 14 selects the threshold matrix (threshold array T ₅ ) from the threshold matrix storage unit 16, and supplies the threshold signal TH (corresponding to the threshold t (P _5l )) to the comparator 12. The binary image signal B is generated by being supplied and compared with the multi-value image signal I. The binary image signal is output as a halftone dot image on the film F by the output device 20, or a printed matter P is created from the film F.

In this case, since the threshold value matrix is arranged so that the threshold values close to each other are not close to each other, the dot pattern of the image formed by the threshold value matrix is not shifted, and a good network with no noticeable graininess is obtained. A point image can be obtained. In addition, the threshold matrix is arranged in consideration of other threshold matrices arranged around it, such as the threshold arrays T _{1 to} T ₄ and T _{6 to} T ₉ shown in FIG. Therefore, a regular pattern due to repetition of the threshold matrix does not appear on the halftone dot image, and a better image can be obtained.

Example 2
In the first embodiment, the threshold value t (P _kl ) is set in ascending order such as 1, 2,..., But in this case, when setting a large value threshold value t (P _kl ), most of the pixels are already set. since the threshold t (P _kl) is set for the point P _kl, there is a concern that becomes the pixel points P _kl close to each other and selectively forced state.

Therefore, in the second embodiment, as shown in FIG. 7, four pixel points P _k1 , P _k2 , P _k24 , and P _k25 that are appropriately separated from each threshold _array T _k are selected, and a threshold value t is set as an initial value. (P _k1 ) = 1, t (P _k2 ) = 2, t (P _k24 ) = 24, and t (P _k25 ) = 25 are set. Next, in the same manner as in the first embodiment, after _obtaining the pixel point P _k3 for setting the threshold value t (P _k3 ) = 3, the pixel point P _k23 for setting the threshold value t (P _k23 ) = 23 is obtained. In this manner, the threshold array T ₅ is obtained by alternately setting the threshold value t (P _kl ) one by one from the highlight side and the shadow side.

In the threshold array T ₅ set as described above, the threshold value t (P _kl ) is sufficiently separated on the highlight side and the shadow side. Therefore, when forming a halftone dot image using the threshold matrix of the threshold phased array T _5, it is possible to obtain good halftone images without roughness in both the highlight side and the shadow side.

Example 3
8 and 9 are flowcharts showing the processing procedure of the third embodiment. In this embodiment, the point spread function shown in FIG. 10 is obtained by two-dimensional Fourier transform of the visual spatial frequency characteristic (MTF) shown in FIG. 15, and the point spread function is applied to the position corresponding to the threshold position. By adding or subtracting cumulatively, the virtual density distribution of the image obtained from the threshold value is obtained, and the threshold value matrix is set so as to minimize the variation, thereby suppressing annoying graininess to a minimum. In addition, by arranging threshold values in consideration of another threshold matrix set adjacent to the threshold matrix, appearance of a regular pattern due to discontinuous repetition of the threshold matrix can be reduced.

First, as in the first embodiment, a threshold array T _k and pixel points P _kl (k = 1 to 9, l = 1 to 25) are defined (step S200), and all threshold values t (P _kl ) are defined. After setting to 0 (step S202), two pixel points P _k1 and P _k2 that are appropriately separated from each threshold array T _k are selected, and threshold values t (P _k1 ) = 1 and t (P _k2 ) are set as initial values. ) = 2 is set (see step S204, FIG. 4).

Next, z = 2 (step S206), the weighting filter f _kl (P _kz ) shown in FIG. 11 is applied to the threshold array T _k shown in FIG. 5, and the weighted threshold array T shown in FIG. _{A k} density distribution _Dkl is obtained (step S208).

That is, the weighting filter f _kl (P _kz ) shown in FIG. 11 represents the distribution characteristic of the point spread function shown in FIG. 10 as a two-dimensional density distribution consisting of a 5 × 5 matrix, and this weighting filter f _{By causing kl} (P _kz ) to act on the pixel point P _{kl for} which the threshold value t (P _kl ) of the threshold array T _k is set, and adding them, the density distribution D _kl shown in FIG. 12 is obtained. Obtainable. In this case, by setting the threshold value t (P _kl ) sequentially so that the variation in the density distribution D _kl is minimized (in this case, it corresponds to the minimum value), it is possible to suppress annoying graininess to a minimum. it can.

First, z = 3, the minimum value Dmin = 99999 density distribution D _kl, and a random number RND = 0, i = 1 (step S210, S214, S216, S218) , if the threshold t (P _5i) is 0 (step S222), and compares the density distribution D _5i and the minimum value Dmin of the pixel point P _5i (step S224, S230). In this case, when i = 1, the minimum value Dmin is 99999 (step S214), so Dmin = _D5i is set (step S226), and _x5z = _x5i and _y5z = _y5i are set (step S228). ). Then, i is updated (step S218), and the density distribution _D5i and the minimum value Dmin are compared again (steps S224 and S230).

Thereafter, processing is performed in the same manner as steps S124 to S138 of the first embodiment (steps S224 to S238), and the pixel point P _5Z having the lowest density in the density distribution D _kl shown in FIG. (Corresponding to the position) can be extracted. Therefore, a threshold value t (P _kz ) = 3 (z = 3) is set for the pixel point P _{kz of} each threshold array T _k including the pixel point P _5Z (steps S220 and S240).

Similarly, by repeating the processing from step S208 and setting threshold value t (P _kl ) = 4 to 25, threshold value t (P) for all pixel points P _kl in threshold _array T _k . _kl ) is set (step S212).

As in the case of the first embodiment, only the threshold threshold array T ₅ is extracted from the threshold array T _k set as described above, and the threshold value matrix storage unit shown in FIG. 16 is set.

  Since the threshold value matrix set in this way has threshold values arranged so as to minimize annoying graininess, a good dot image can be obtained. In addition, since threshold values are arranged in consideration of other threshold matrixes arranged around, a regular pattern by repeating the threshold matrix is displayed on the halftone image as in the case of the first embodiment. A better image can be obtained without appearing.

In the third embodiment, as in the second embodiment, by setting the threshold values alternately one by one from the highlight side and the shadow side, it is possible to form a better halftone image. Is possible. Further, weighting filters f _kl a (P _kz) instead of obtaining the density distribution D _kl by cumulatively adding, may be obtained density distribution D _kl cumulative subtraction.

Example 4
By the way, in the third embodiment, the weighting filter f _kl (P _kz ) corresponding to vision has a finite size, and the threshold is separated from the range where the weighting filter f _kl (P _kz ) acts. / In the shadow area, even if the density distribution _Dkl is obtained, the minimum value and the maximum value are not uniquely determined, and there is a possibility that an appropriate threshold position cannot be determined.

Therefore, in the fourth embodiment, the highlight side (0 <t (P _kl ) ≦ TH _W ) of the threshold value t (P _kl ) is set using the method of the first embodiment (FIGS. 2 and 3). The intermediate value (TH _W <t (P _kl ) ≦ TH _B ) of the threshold value t (P _kl ) is set using the method of the third embodiment (see FIGS. 8 and 9). Further, for the shadow side (TH _B <t (P _kl ) ≦ TH _MAX ) of the threshold value t (P _kl ), the setting of the first embodiment is used again. Note that 0 <TH _W <TH _B <TH _MAX , and TH _MAX represents the maximum value of the threshold value t (P _kl ).

  In this case, it is possible to obtain a threshold matrix capable of forming an image with no noticeable graininess or periodic pattern in the entire range from highlight to shadow.

  In the fourth embodiment, as in the second embodiment, by setting the threshold values alternately one by one from the highlight side and the shadow side, a better halftone image can be formed. Is possible.

Example 5
In the fifth embodiment, the dot gain of the film F or the printed matter P that is the output medium is taken into consideration and the dot gain is created by correcting the weighting filter f _kl (P _kz ) in the third embodiment to create a threshold matrix. The corrected halftone dot image can be obtained.

First, a method for obtaining the dot gain correction amount Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) will be described.

In this case, as shown in FIG. 1, using the output device 20, based on the test data T supplied from the test pattern data file 24 of the RIP 10, as shown in FIG. A test pattern composed of pixels is created, or a printed material P is created. Then, each of the test patterns is measured using the density measuring device 26, and from the measured values, as shown in FIG. 14, the peripheral pixels Q ₁ , Q ₂ , Q ₃ , Q ₄ are turned on (black dots) / off ( The correction amount Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) is obtained based on the density difference when the central pixel Q, which is the target pixel, is turned on from the off state with respect to the white point) state. However, among the nine pixels in FIG. 14, the on / off states of four pixels excluding the pixels Q, Q ₁ , Q ₂ , Q ₃ , and Q ₄ are constant. In FIG. 13, when the pixel Q and the peripheral pixels Q ₁ , Q ₂ , Q ₃ , and Q ₄ are off, 0 is set, and when the pixel Q is turned on, 1 is set, and [Q ₁ Q ₂ Q ₃ Q ₄ −Q] = It shall be expressed as [0000-1].

Therefore, first, the concentration D1 of [Q ₁ Q ₂ Q ₃ Q ₄ −1] and the concentration D 0 of [Q ₁ Q ₂ Q ₃ Q ₄ −0] are measured using the concentration measuring device 26. Next, in the threshold value matrix creation device 22, the difference between the densities D1 and D0 is obtained as a half-tone difference Δ (%) of each test pattern. That is, assuming that the base density before image recording of the film F or the printed matter P is Db and the maximum density when the whole is blackened is Dmax, the halftone difference Δ (%) is

Can be obtained as In this case, an ideal half-tone difference R without dot gain or the like in the test pattern shown in FIG.
R = 100/9 (%) (3)
Therefore, the correction amount Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) is
Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) = Δ / R−1 (4)
Can be obtained as

The correction amounts Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) obtained as described above are supplied to the threshold matrix creating device 22 and the threshold matrix is obtained in substantially the same manner as in the third embodiment.

That is, the correction amount Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) is the threshold value t (P _k1 ) = 1 set as the initial value in the process immediately before step S208 in FIG. It acts on the weighting filter f _kl (P _kz ) for P _k2 ) = 2. In this case, the threshold value t (P _k1), the threshold value t (P _k2), because as shown in FIG. 5, the threshold value for forming the surrounding pixels is 0, the threshold t (P _k1), the threshold value t (P _The dot gain correction amount Cor (Q ₁ , Q ₂ , Q ₃ , Q ₄ ) for the case where only the pixel corresponding to _k2 ) is blackened on the film F or the printed matter P is the test pattern [0000− 0] and [0000-1], the correction amount Cor (0, 0, 0, 0) is obtained.

Therefore, the corrected weighting filter in which the dot gain is taken into consideration is obtained by multiplying each element of the weighting filter f _kl (P _kz ) shown in FIG. 11 by 1 + correction amount Cor (0, 0, 0, 0). f _kl (P _kz ) is obtained.

Next, using the weighting filter f _kl (P _kz ) in consideration of the dot gain obtained as described above, the processing from step S208 shown in FIG. 8 and FIG. 9 is performed, and the threshold value t (P _k3 ) = A pixel point P _kl for setting 3 is obtained (step S240). Similarly, the weighting filter f _kl (P _kz ) of each pixel point P _kl for which the threshold value t (P _kl ) is set is corrected in consideration of the dot gain, and the processing in step S208 and subsequent steps is performed again to obtain the desired value. Threshold matrix is created.

  The threshold value matrix set as described above is created according to the output device 20 and the output medium, and is set in the threshold value matrix storage unit 16 shown in FIG. The selection unit 14 selects a desired threshold value matrix from the threshold value matrix storage unit 16, and a binary image signal B is generated from the multilevel image signal I using this threshold value matrix. The binary image signal is output as a halftone dot image on the film F by the output device 20, or a printed matter P is created from the film F.

  In this case, as in the case of the third embodiment, since the threshold value is arranged according to the visual characteristics, the threshold value matrix can obtain a good halftone image in which unobtrusive graininess is not visually recognized. it can. In addition, since threshold values are arranged in consideration of other threshold matrixes arranged around, a regular pattern by repeating the threshold matrix is displayed on the halftone image as in the case of the first embodiment. A better image can be obtained without appearing. Furthermore, since the dot gain depending on the output medium such as the film F or the printed matter P can be corrected, a halftone image with no more granularity can be obtained with higher accuracy.

  In the fifth embodiment as well, as in the second embodiment, it is possible to form a better halftone dot image by alternately setting the threshold values one by one from the highlight side and the shadow side. .

  In each of the above-described embodiments, two or more initial threshold values are set, but it is needless to say that the initial value may be one.

1 is a configuration block diagram of a system to which a mesh screen creation method according to the present invention is applied. FIG. 3 is a process flowchart of Embodiment 1. 3 is a process flowchart of Embodiment 1. 3 is a process flowchart of Embodiment 1. It is explanatory drawing of a threshold array and the initial value of the threshold value set to it. It is explanatory drawing in the case of setting the threshold value of 3 with respect to a threshold array. It is explanatory drawing in the case of Example 2 which sets a threshold value alternately with an ascending order and a descending order with respect to a threshold array. 10 is a process flowchart of Embodiment 3. 10 is a process flowchart of Embodiment 3. 10 is an explanatory diagram of a point spread function in Embodiment 3. FIG. It is explanatory drawing of the weighting filter in Example 3. FIG. In Example 3, it is explanatory drawing at the time of applying a weighting filter with respect to a threshold array. It is explanatory drawing of the test pattern in Example 5. FIG. FIG. 10 is an explanatory diagram of a relationship between a pixel of interest of a test pattern and its surrounding pixels in Example 5. It is explanatory drawing of a visual characteristic.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... RIP 12 ... Comparator 14 ... Selection means 16 ... Threshold matrix memory | storage part 20 ... Output device 22 ... Threshold matrix preparation apparatus 24 ... Test pattern data file 26 ... Concentration measuring device

Claims (5)

In a method for creating a mesh screen comprising a threshold matrix for converting a multi-value image signal into a binary image signal,
When determining threshold values in ascending or descending order, a plurality of different weighting values are included in a predetermined range centered on the positions of the already determined threshold values, including other threshold value matrices having the same configuration arranged around the threshold value matrix. The threshold value matrix and the threshold value matrix and the threshold value matrix and the other threshold value matrixes arranged around the threshold value matrix are actuated by adding or subtracting the weight values for each position. By obtaining a weighting distribution by the weighting filter in another threshold value matrix arranged around the threshold value, and setting the position corresponding to the minimum value or the maximum value of the weighting distribution as a new threshold value position, the position of the threshold value sequentially A method for creating a mesh screen, characterized by: The method of claim 1 , wherein
A plurality of combinations of test patterns in which the target pixel and its surrounding pixels are turned on / off in advance are formed on a desired output medium, and the density or halftone dot area ratio of each test pattern is measured. A method for creating a mesh screen, comprising: obtaining a correction value based on a difference between a measurement value in a state and a measurement value when the pixel of interest is in an off state; and correcting the weighting filter with the correction value . The method according to claim 1 or 2 , wherein
A method for creating a mesh screen, wherein the weighting filter is a point spread function which is a response characteristic obtained by two-dimensional Fourier transform of a spatial frequency characteristic of human vision. The method of claim 2 , wherein
The test pattern is formed as a desired printed matter, and the correction value is obtained by measuring the density or halftone dot area ratio of the printed matter. In the method in any one of Claims 1-4 ,
The method for creating a mesh screen, wherein the position is determined alternately in ascending order and descending order of threshold values.

Images