JPH04212857A - Abstract pattern plate making system for printing building material - Google Patents

Abstract

PURPOSE:To perform a retouch work in a short time by a method wherein mask data is produced from pattern data in the given position of pattern data produced by a first means, and a second means to remove an undesired pattern in a way that picture element data is in a desired position is copied in other different desired position based on the mask data is provided. CONSTITUTION:A unit raw material on a basis of which an abstract pattern is formed is decomposed into picture elements by means of an input scanner 1 to effect reading. Longitudinal and lateral watermark synthesis is effected on picture data of the unit raw material read by the scanner the given number of times by means of a first means equivalent to an endless processing part 9 to produce pattern data having a given size. When an undesired pattern equivalent to a pattern curl is produced at pattern data, a second means produces mask data based on pattern data in the given position of pattern data produced by the first means. Further, based on mask data, picture element data in a first desired position is copied in other different second desired position.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plate-making system for printing building materials, and more particularly to a plate-making system for printing building materials having abstract patterns.

0002

[Prior Art] In recent years, there has been a shortage of natural materials such as wood, and various building materials such as plywood and gypsum board have been developed as substitutes. The role of printing for building materials is becoming increasingly important.

[0003] Patterns for printing on building materials include wood grain patterns that imitate patterns of natural materials such as wood grain or stone grain, geometric patterns, sand grain patterns,
Abstract patterns created by humans, such as background patterns and floral patterns, are known, but since the present invention relates to the latter abstract patterns, the plate-making of abstract patterns will be explained below.

FIG. 18 shows an outline of a conventional plate-making process for abstract patterns.

[0005] As shown in FIG. 18A, first, when a material sample with an abstract pattern consisting of a design drawing, photograph, etc. is submitted, the material sample is photographed with a prepress camera to create a separation film. In order to eliminate the difference in level between adjacent material samples, a suitable mask is applied and the plates are repeatedly printed one after another by film composition to create a plate for proof printing (hereinafter, this plate is referred to as a baby plate). The film on which the material sample is photographed is usually about 3 to 5 cm square in size, but by repeating the film 101 nine times vertically and horizontally as shown in Figure 18B, a plate with a size of about 1 meter square is created. be able to. This is the baby version 102.

[0006] Next, gravure engraving is performed based on the baby plate to create a printing plate for gravure printing, a proof is made, and if the desired abstract pattern is obtained, the baby plate is repeated and the book is printed. A printing plate is created, a printing plate is created based on it, proof printing is performed again, and if the desired abstract pattern is obtained, printing is performed using this machine.

[0007] The above is the process for creating a baby version by film composition, and the whole process is done manually by the operator, but in recent years, by using a layout scanner, for example, it is possible to create a baby version by photographing a material sample. Attempts have also been made to digitize part of the above process by using a layout scanner to read the image data, repeat the image data, output it on film, and create a baby version using a layout scanner. However, even in this case, the process of repeating the baby version to create the main version is carried out manually by photographically composing the baby version.

[0008]

[Problem to be Solved by the Invention] However, as mentioned above, in the past, most of the plate-making process for printing abstract patterns on building materials was done manually, which not only caused the problem of extremely high costs. , there were also the following problems. That is, the material sample may have uneven density or may have a lopsided pattern. These density unevenness and pattern deviations are not a problem with the material sample itself, but when the baby version is created by repeating the process, these density unevenness and pattern deviations are emphasized and the creation becomes difficult. It appears as a repeating pattern that was not intended by the creator or designer, and the repeating pattern stands out more than the original abstract pattern, which is a big problem. ). For example, suppose that, as shown in FIG. 19A, density unevenness as indicated by 104 has occurred in a material sample 103 consisting of a fine pattern (not shown), and as shown in FIG. Assuming that a baby plate 105 is created by repeating the process four times in the vertical direction, if a printing plate is created based on this baby plate 105 and gravure printing is performed, density unevenness will occur as shown in Figure C. Repeated patterns are emphasized and pattern curls occur. On the other hand, when joints are added as shown in Figure D, the grid pattern of the joints stands out and the pattern curls become relatively inconspicuous. No, but it's not a problem.

[0009] As described above, the occurrence of pattern curls is a fatal defect for abstract pattern building materials, so work is carried out to confirm the presence or absence of pattern curls during the printing process of the main version. Conventionally, a method of checking by proof printing was adopted. Of course, it is possible to check for pattern curls using plate-making film, but since the occurrence of pattern curls changes depending on the color, it is difficult to reliably determine the presence or absence of pattern curls using black-and-white plate-making film. very difficult,
In the end, we had no choice but to confirm the proof by using a baby plate, although it was costly.

[0010] Also, the work of removing pattern quirks is called retouching work, which involves re-creating the mask, rearranging the pattern, or rotating the pattern, compositing it with film, and repeating it to create a baby version. The retouching work not only requires great skill, but also requires a large workload, and since everything is done manually, it takes time, and as a result, it takes a long time to deliver. This resulted in high plate-making costs.

[0011] Moreover, there are many patterns that cannot be removed by manual retouching work, and in such cases, the project itself may be canceled and the plate making process may end up in vain.

[0012] This situation is the same when a layout scanner is used, and although the creation of baby version patterns is automated, retouching work still relies on manual work, so the above problems still occur. has not been resolved in any way. Of course, it is possible to display the pattern of the material sample read by the input scanner on the monitor screen, rotate it, or repeat it using an appropriate mask pattern, but this work is basically manual retouching. The work was the same, and the effort was exactly the same, except that instead of optical work using film, operations were performed on a monitor using a pointing device such as a mouse or stylus pen. In addition, as mentioned above, it is not possible to display the entire pattern of the baby version, which has a size of about 1 meter square, on the screen, so even if the retouching work was done on the display screen, it is not possible to remove the pattern irregularities. It was not possible to confirm. In other words, it is not possible to know whether pattern curl will occur or not just from the pattern of the material sample; it is only possible to confirm the occurrence of pattern curl by repeating the material sample and creating a baby version. Therefore, in order to confirm the presence or absence of pattern curl, it is necessary to be able to display the entire baby version pattern on the monitor screen, but conventional layout scanner monitors have not been able to display this.

Furthermore, even if the entire baby version pattern could be displayed, the pattern pattern could not be confirmed when the baby version was repeated to create the main version. Abstract pattern plate making systems using conventional layout scanners have not been able to deal with such situations.

The present invention solves the above-mentioned problems and provides an abstract pattern plate making system for printing building materials that can easily remove pattern curls, especially in abstract patterns having a relatively large area. The purpose is to

[0015]

[Means for Solving the Problems] In order to achieve the above object, the abstract pattern plate making system for printing building materials of the present invention performs watermark synthesis for a predetermined number of times using an input scanner and unit material image data read by the input scanner. a first means for generating pattern data of a predetermined size; and creating mask data from the pattern data at a predetermined position of the pattern data created by the first means, and creating mask data at a desired position based on the mask data. and a second means for removing the undesired pattern by copying the pixel data of the pixel data to another different desired position.

[0016]

[Operation] The input scanner decomposes the unit materials that form the basis of the abstract pattern into pixels of a predetermined density and reads them. The image data of the unit material read by the input scanner is processed by the first processing unit, which corresponds to the endless processing unit in the embodiment below.
The watermark is synthesized a predetermined number of times in the vertical and horizontal directions using the above means, and pattern data of a predetermined size is created. If an undesired pattern corresponding to the pattern pattern in the embodiment described below occurs in the pattern data, the second means moves the pattern data to a predetermined position created by the first method. Mask data is created based on the pattern data, and further based on the mask data, the pixel data at the first desired position is copied to a different second desired position.

By repeating the above operations, undesired patterns can be removed.

[0018]

Embodiments Hereinafter, embodiments will be described with reference to the drawings.

FIG. 1 is a diagram showing the configuration of an embodiment of an abstract pattern plate-making system for printing building materials according to the present invention, in which 1 is an input scanner, 2 is an output scanner, 3 is a control means, and 4 is an input device. means, 5 is a storage means, 6 is a gravure engraving machine, 7 is a color hard copy, 8 is an abstract pattern plate making processing means, 9 is an endless processing section, 10 is a pattern pattern confirmation processing section, 11 is a pattern pattern removal processing section, 12 is a repeat processing section, 13 is a display processing means, and 14 is a monitor.

In FIG. 1, an input scanner 1 scans a color film obtained by photographing a material sample to classify each pixel into cyan (C), magenta (M), and yellow (Y).
), black (K), and a predetermined number of bits,
For example, it outputs 1-byte digital image data, and has the same configuration as an input scanner used in a conventional layout scanner.

The output scanner 2 outputs C, M, Y, and K color separation plates on film, and has the same configuration as an output scanner used in a conventional layout scanner.

The control means 3 is for controlling the operation of the abstract pattern plate-making system for printing building materials, and is composed of a microcomputer, ROM, RAM, etc.

The input means 4 is composed of input devices such as a keyboard and a pointing device.

The storage means 5 is composed of a storage device such as a RAM and/or a hard disk.

The gravure engraving machine 6 is used to engrave baby plates and book plates.

The color hard copy 7 is composed of a sublimation transfer printer, an inkjet printer, etc., and is capable of outputting at least a color hard copy of the baby version size, and more preferably a color hard copy of the main version size. What is possible is good.

The abstract pattern plate making processing means 8 includes an endless processing section 9, a pattern curl confirmation processing section 10, and a pattern curl removal processing section 11.
and a repeat processing section 12. The operation of each of these parts will be described later.

The display processing means 12 controls the display of the monitor 14, and includes a video RAM having a predetermined capacity corresponding to the number of pixels of the monitor 14.

The monitor 14 is composed of a display device such as a color CRT, but it is desirable that the monitor 14 can display images with high definition.

Next, the operation of the configuration shown in FIG. 1 will be explained with reference to FIG. 2. FIG. 2 is a flowchart showing an example of the plate-making process when using the abstract pattern plate-making system for printing building materials configured as shown in FIG. A unit material is set on the input scanner 1 and read (step S1). This unit material can be read in the same way as a conventional layout scanner. If the unit material is film, it can be loaded directly into the input scanner 1, but if the unit material is a design image other than film, it may be taken on color film and read in, or it may be loaded into the prepress camera as necessary. , create a separation film, and load it.

The image data of the unit material read in step S1 is stored in the storage means 5.

Next, when the endless processing is instructed from the input means 4, the control means 3 activates the endless processing section 9 of the abstract pattern plate making processing means 8 to execute the endless processing (step S2). .

Endless processing in the present invention refers to pattern data (hereinafter referred to as The image data for the baby version is created by simply repeating the endless pattern data a predetermined number of times. That is,
Endless patterns are located between unit materials and baby plates, and can significantly reduce the operator's workload. In other words, conventionally, a baby version was created directly by repeating mask processing on a unit material, but as mentioned above, a unit material is about 3 to 5 cm square, and a baby version is about 1 meter square. Because of its size, it takes a lot of effort to create a baby version, whereas the endless pattern of the present invention is
Although mask processing is applied, it is only necessary to synthesize the unit materials nine times at most, and the synthesis is performed automatically.The baby version can simply combine the endless patterns created in this way without applying mask processing. Since the process is simply repeated, the workload is significantly reduced compared to the conventional method.

The details of the endless processing will be explained below with reference to FIG. Now, if an instruction is given to create an endless pattern by combining unit materials three times in both the horizontal and vertical directions, the endless processing section 9 will combine the unit materials three times in the horizontal direction and three times in the vertical direction. However, at this time, Figure 3
As shown in A, when compositing in the horizontal direction, overlap is performed with a width of WY, and when performing compositing in the vertical direction, the width is WT.
overlap with a width of , and the second row is horizontally W
Shift only O. Then, watermark synthesis is performed in the horizontal overlapping area and the vertical overlapping area at the ratio shown in area R2 in FIG. 3B. For example, if we take the composition of unit material 151 and unit material 152 as an example, in region R1 of FIG. 3B, the pattern of unit material 151 is 100%, but in region R2, the pattern of unit material 151 is straight down to 0%. While decreasing, the pattern of the unit material 152 linearly increases from 0% to 100%, and in region R3, the pattern of the unit material 152 becomes 100%. The same applies to vertical composition.

Unit materials can be automatically synthesized as described above, but if it is predicted that there will be a step in the overlapping part of the unit materials, or if there is a pattern with a relatively large area in the overlapping part. In cases where it is predicted that the pattern will be destroyed by the combination, the automatic combination described above is not appropriate. Therefore, in addition to the automatic compositing mode described above, the endless processing unit 9 also has a mode in which a mask for compositing can be arbitrarily set, and in this mode, for example, as shown in FIG. 3C, Two masks M1 and M2 are set in the overlapping portion of , and watermark synthesis can be performed in the area sandwiched by these two masks at the ratio shown in area R2 in FIG. 3B. The same applies to the vertical direction. However, it goes without saying that a mask is set separately when compositing in the vertical direction.

Note that the horizontal overlap width WY, the vertical overlap width WT, and the second stage shift amount WO may be set in advance as fixed values in the endless processing section 9, or
The operator may be able to set it each time using the input means 4. Also, the settings for the two mask patterns are as follows:
For example, a horizontally overlapping pattern and a vertically overlapping pattern are displayed on the monitor 14, a mask line is traced on the screen using a mouse or a stylus pen, and the traced pattern data is converted into a mask pattern. All you have to do is import it as data. Masks set for horizontal compositing are commonly used for horizontal compositing, and masks set for vertical compositing are commonly used for vertical compositing. used.

When the composition of the unit materials is completed as described above, the endless processing section 9 displays the composition result on the screen in mode 14. Therefore, for example, if an arbitrary point P1 (FIG. 3A) on the unit material 151 located at the upper left corner is specified, the endless processing unit 9 For the unit material 157 located and the unit material 159 located in the lower right corner, the coordinate values on the composite pattern of points P2, P3, and P4 having the same address as the indicated point P1 on each unit material are determined, and these four points P1 are calculated. , P2, P3, and P4 are cut out and registered as an endless pattern. The reason for cutting out in this way is to prevent differences in level from occurring when the baby version is created. That is, as will be described later, the baby version is created by simply repeating an endless pattern, but if the above cutting is performed, the patterns on both sides of the straight line connecting the two points P1 and P2 in Fig. 3A will be 2. Point P
Since the patterns on both sides of the straight line connecting 3 and P4 are exactly the same,
Even if the endless pattern is repeated vertically, no steps will occur. The same applies to the horizontal direction.

The above is the process of step S2 in FIG.
In this way, by creating a pattern that is intermediate between a unit material called an endless pattern and a baby version, the workload can be significantly reduced compared to the conventional method.

[0039] When the endless patterning process is completed, it is next determined whether or not the endless pattern has a pattern curl (
Step S3). Conventionally, as mentioned above, proof printing was performed to check the presence or absence of pattern curl, but the abstract pattern plate making system for printing building materials shown in FIG. 1 is equipped with a pattern curl confirmation processing section 10. This makes it possible to quickly and easily confirm the presence or absence of pattern curl.

The pattern pattern confirmation processing section 10 is activated by the control means 3 when a predetermined menu is selected by the input means 4, calls up the endless pattern created and registered in step S2, and displays an image of the endless pattern. Thin the data to a predetermined magnification, e.g. 1/4 of the original size.
Reduce to. For example, if the endless pattern is as shown in FIG. 4A, the pattern pattern confirmation processing unit 10 thins out the pixels of the endless pattern 16 at predetermined intervals, and
At B, it is reduced to 1/4 as shown at 16'. Next, the pattern pattern confirmation processing unit 10 performs a repeat process of arranging the reduced endless pattern pattern vertically and horizontally without gaps. As a result, a pattern in which endless patterns are repeated as shown in FIG. 4C is obtained, and the pattern pattern confirmation processing unit 10 passes this pattern to the display processing means 13 to display it on the screen of the monitor 14.

The reason why the endless pattern is reduced and displayed repeatedly is as follows. As mentioned above, pattern curls may occur at the baby version stage,
Although it cannot be confirmed in the baby version, it may occur in the main version, so it is best to check the pattern as close to the main version as possible. Therefore, instead of displaying only the endless pattern as described above, the endless pattern is reduced in size and displayed repeatedly, thereby making it easier to confirm the pattern pattern.

[0042] How much to reduce the endless pattern and how many times to repeat it in the vertical and horizontal directions is arbitrary, but if the amount of thinning of the endless pattern is increased, the number of repeats can be increased, but the image data It is possible that the state of the pattern pattern changes due to thinning, so the pattern pattern shown in Figure 4C
It is best to repeat the pattern only two or three times in both the vertical and horizontal directions, as shown in the figure, and it has been confirmed that this is actually a good way to check the pattern pattern.

If the occurrence of pattern curl is not confirmed on the screen shown in FIG. 4C, baby version data is created (step S5); however, if pattern curl is confirmed to occur, the pattern curl is Perform removal. To this end, the abstract pattern plate making system for printing building materials according to the present invention removes pattern quirks by replacing the pattern located at a predetermined address of the endless pattern with a pattern located at another address based on a predetermined mask pattern. (hereinafter referred to as the scramble method), and a method in which the pattern is removed by moving or rotating a predetermined pattern to another desired position (hereinafter referred to as the pixel copy method) ,
and a method for removing pattern quirks by removing density unevenness of a unit material (hereinafter referred to as a Fourier transform method).

[0044] First, the removal of pattern curl by the scrambling method will be explained. The scramble method is suitable for removing pattern curls from abstract patterns that have fine patterns such as grains and background patterns. It also has a batch mode for removal.

Now, the pattern curl removal processing section 11 controls the control means 3 when a pattern curl removal process is instructed by the input means 4.
However, if the scramble method is instructed to remove pattern curls, the scramble method display mode is first activated. In display mode,
First, in order to finish the simulation in a short time,
Endless pattern image data with pixel count of 2K bits x 2K
Thinned to less than a bit, image memory 2
0, and also sends it to the display processing means 13 and displays it on the monitor 14. Then, the pixels of the endless pattern from which the image data has been thinned out are divided into blocks of a predetermined size, for example, about 200 pixels x 200 pixels.

The random address generation unit 21 is instructed from the input means 4 the number of times the pixel replacement is to be executed as instructed by the operator, and by generating random numbers, the number of times the pair of addresses specifying two different pixel blocks is executed is determined. is generated and sent to the pixel block transfer unit 22. Furthermore, various mask patterns as shown in FIGS. Sent. Then, the pixel block transfer unit 22 takes in the pixel data of the block located in the pixel block address pair AD1, AD2 generated by the random address generation unit 21, and transfers the pixels in these blocks to the pixel data generated by the mask data unit 23. The data is replaced according to the mask shape and written to the original address location. As a result, the pixels inside the mask pattern of the block with address AD2 are written inside the mask pattern of the block with address AD1, and the pixels inside the mask pattern of the block with address AD1 are written inside the mask pattern of the block with address AD2. Pixel is written. Repeat the above operations the number of times instructed. For example, as shown in FIG. 6, the thinned endless pattern 24
is divided into 9 (horizontal) x 8 (vertical) blocks, and a mask pattern having the shape shown in FIG. 7A is selected. Then, if the random address generation unit 21 generates the address pair BL1 and BL2 as the first replacement, the pixel block transfer unit 22 generates the block with the address BL1 and the block with the address B
The pixels of the L2 block are taken in and the pixels included in the circular mask are replaced. Next, if the address pair BL3, BL4 is generated as the second replacement, the pixel block transfer unit 22 transfers the block with the address BL3 and the block with the address BL4.
Takes the pixels of the block and replaces the pixels included in the circular mask. Thereafter, this operation is performed the same number of times as the execution number.

When the pixels in the block have been replaced the number of times instructed, the image data on the image memory 20 is sent to the display processing means 13 and displayed on the monitor 14. This allows the operator to check on the monitor 14 whether or not the handle quirk has been removed. If the handle quirk has been removed, the display mode ends and the batch mode is executed, but if the handle quirk has not been removed, If not, continue to input the number of executions and repeat the above operation.

The above is the operation in the display mode, and the number of executions and mask shape data at the end of the display mode are saved in a predetermined file.

Next, the operation in batch mode will be explained. If the removal of the pattern curl is confirmed in the display mode, the batch mode is instructed. In the batch mode, original endless pattern image data is written to the image memory 20, and the random address generator 2
1 and the mask data section 23 are loaded with the number of executions and the mask shape in the display mode, respectively, and the pixels in the block described above are replaced.

The image data of the endless pattern from which the pattern pattern has been removed in this way is registered in a predetermined file in the storage means 5. In addition, when the image data of the endless pattern is color-coded into four colors, Y, M, C, and K, the capacity of the image memory 20 is determined by the capacity of all In order to perform the above scrambling process for colors, approximately 8K x 8K x 8 x 4 (bits) are required.

Although the above explanation deals with the case where only one mask pattern is used, the mask pattern used may be changed for each replacement by, for example, generating random numbers. Alternatively, addresses of two mask patterns to be replaced may be generated without forming blocks as described above. Furthermore, in the above embodiment, the pixel replacement is displayed on the monitor 14 after the specified number of executions, but the number of executions is not specified, and for example, an image is displayed every time pixel replacement is performed 100 times. However, it is also possible to forcibly stop the replacement operation when the operator determines that the handle pattern has been removed.

[0052] In the manner described above, the work of removing pattern curls, which has conventionally been done manually, can be done automatically.

The above is the removal of pattern curl using the scramble method.Next, the removal of pattern curl using the pixel copy method will be explained. The pixel copy method is applied to remove pattern curls from abstract patterns that have relatively large characteristic patterns such as stone grains. That is, as shown in FIG. 8, there are relatively large patterns 261 to 2 in the unit material 25.
64 are lined up in an archipelago pattern, such an archipelago-like pattern may cause pattern curls, but if the resulting pattern curls are removed by a scrambling method, the scrambling method is an endless process. Since the image of the pattern is divided into blocks, there is a possibility that these characteristic patterns may be destroyed. Therefore, a pixel copy method is provided to remove pattern curl from abstract patterns having such characteristic patterns.

The pixel copy method is performed with the configuration shown in FIG. 9, and includes an interactive mode in which pixel copy is simulated interactively on the screen of the monitor 14, and a batch mode in which the same processing as performed in the interactive mode is automatically performed. There is a mode.

Now, when pattern pattern removal using the pixel copy method is instructed, an interactive mode is activated. As a result, the number of pixels of the endless pattern image data is 2K bits x 2
The data is thinned out to about K bits or less and developed in the image memory 30. The image data on the image memory 30 is also sent to the display processing means 13 and displayed on the monitor 14.

The operator observes the screen in mode 14 and specifies a pattern to be moved (hereinafter, this pattern will be referred to as a source pattern). The source pattern is specified by specifying the diagonal vertices of the area surrounding the source pattern using a mouse or the like, but the mouse movement at this time is sent to the command control unit 35, and the cursor generation unit 36 generates the cursor pattern. The cursors 381 to 384 as shown in FIG. 10 are displayed on the screen of the monitor 14, and the cursor addresses are recorded in the command recording section 37. be done. Note that 39 in FIG. 10 indicates a source pattern.

When a rectangular area is specified in this way,
The pixel block transfer unit 31 transfers the pixel data within the rectangular area to the source buffer unit 32. The contents of the source buffer section 32 are further transferred to the mask automatic generation section 34, and a source pattern mask is generated by the following processing. That is, first, the automatic mask generation section 34 performs the calculation of 0.3R+0.59G+0.11B on the image data of the three primary colors of R, G, and B transferred from the source buffer section 32 to convert the image into monochrome. Next, the monochrome image data is subjected to smoothing filter processing a predetermined number of times, for example, about 20 times. As a result, fine patterns are erased and only patterns with a large area remain. Note that the smoothing filter process can employ a well-known method such as taking the average value of nine neighboring pixels.

After the smoothing filter processing, the automatic mask generation unit 34 calculates a histogram of the density values of the pixels, and calculates the density value at which the cumulative frequency from density 0 is about 50% and the frequency is minimum as a threshold value, The binarization direction is determined from the representative values of the periphery and center of the rectangular area, and the binarization process is performed to generate a mask of the source pattern. Specifically, for example, the frequency and cumulative frequency of the level value of each pixel are shown in Figure 1.
1, 40 and 41, ND0 is calculated as the concentration value at which the cumulative frequency is about 50% and the frequency is minimal, and this is determined as the threshold value. Assuming that the contour shown by 42 in FIG. 12 is obtained using the threshold value,
A mask of the source pattern is generated by binarizing the inside of the outline 42 as "1" and the peripheral part as "0". As a result, the periphery of the contour 42 is always masked. Note that the threshold value may be set arbitrarily by the input means 4.

Pattern quirks are removed using the mask generated as described above, and this operation is basically performed by copying the source pattern to another position, so at least The following commands are available:

■Dragging command The pixel copy method is a method of copying a source pattern by moving it to another position, but the dragging command is used to drag the source pattern and move it to the desired position. be. When a dragging command is specified and the destination of the source pattern is specified with the cursor, the pixel block transfer unit 31 sets the specified position as the target position and transfers the pixel block at the target position to the target buffer unit 33, The contents of the source buffer section 32 are copied to the target position while referring to the mask pattern created by the automatic mask generation section 34. Note that the target buffer section 33
The pixel block transferred to is used to restore the old target position to its original state when the target position changes with the next cursor indication.

Specifically, for example, in FIG. 13, if the source pattern S is to be moved to the target position T1, the operator picks the source pattern S with the mouse, and in that state moves the source pattern to the position T1. Moving. As a result, the pixel block transfer unit 31
At the position of the target T1, pixels are captured in a rectangular area of the same size as the rectangular area when the source pattern S is specified, and are written into the target buffer section 33. Next, the pixel block transfer unit 31 refers to the mask pattern generated by the automatic mask generation unit 34, and transfers the pixels in the area where the mask pattern is “1” out of the contents of the source buffer unit 32 to the target position T1. write to. As a result, the source pattern located at the position indicated by S in FIG. 13 has been copied to the target position T1.
This result is sent to the display processing means 13 and displayed on the monitor 14.

Next, when the target position changes from T1 to T2, the contents of the target buffer section 33 are returned to T1, and the same processing as above is performed with T2 as the new target position.

■Rotation command In the pixel copy method, the source pattern can of course be moved at the same angle, but it can also be rotated clockwise or counterclockwise up to 360° by a predetermined unit angle, for example, 90°. It is made so that it can be copied. This operation is performed by a rotation command, and when a rotation angle is specified by the rotation command, the pixel block transfer unit 31 copies the contents of the target buffer unit 33 to the original position of the target pattern while referring to the mask data. Then, the contents of the source buffer section 32 and the mask data created by the automatic mask generation section 34 are rotated by the specified angle, and the contents of the target position are copied to the target buffer section 33 in the rotated state. The contents of the source buffer section 32 are copied to the target position while referring to the mask data.

For example, now the target position T2 in FIG.
Assuming that the pattern indicated by the dotted line in is rotated 90° clockwise, the pixel block transfer unit 31 first transfers pixels at the target position T2 in a rectangular area of the same size as the rectangular area when the source pattern S was specified. Write from the target buffer section 33. Next, the contents of the source buffer section 32 and the mask data inside the mask automatic generation section 34 are rotated 90 degrees clockwise, and in the rotated state, the contents of the target position T2 are transferred to the target buffer section 33.
of the contents of the source buffer section 32, the pixels in the area where the mask data is "1" are transferred to the target position T.
Write in 2. As a result, the pattern at the position indicated by the dotted line at T2 in FIG. 13 has been rotated to the same position indicated by the solid line, and this result is sent to the display processing means 13 and displayed on the monitor 14. Ru.

(2) Source Pattern Erase Command When removing pattern irregularities in the pixel copy method, when moving the source pattern to another position, it is necessary to erase the pattern at the original position. The source pattern erase command performs this operation, and when this command is specified and the target position is specified, the pixel block transfer unit 31 converts the contents of the target buffer unit 32 into the source pattern while referring to the mask data. Copy to location.

For example, if the source pattern S in FIG. 13 is erased and the pattern at the target position T3 is to be copied to that position, the pixel block transfer unit 31
At the target position T3, pixels are taken in a rectangular area of the same size as the rectangular area when the source pattern S was specified by the dragging operation up to T3, and among the contents written to the target buffer section 33, the mask data is Pixels in the area that are "1" are written at the positions of the source pattern S. As a result, the source pattern S in FIG. 13 is erased, and the target position T is placed at that position.
3 will be copied, and the result will be displayed on the monitor 14.

■Pattern Swap Command This command is a command for copying pixels at a certain first target position to another second target position using the mask data of the source pattern.
When the target and the command are specified, the pixel block transfer unit 31 returns the contents of the target buffer unit 33 to the first target position and transfers the contents of the target buffer unit 33 to the first target position including the pixels to be copied.
The pixel block at the target position is transferred to the source buffer section 3.
Copy to 2. Next, when the second target position is specified, the pixel block at the second target position to be copied is transferred to the target buffer unit 33, and the contents of the source buffer unit 32 are transferred to the second target position with reference to the mask data. Copy to target location.

For example, in FIG. 13, target T
If an instruction is given to pick up a pixel at position 1 and copy it to the target position T3, the pixel block transfer unit 31 returns the contents of the target buffer unit 33 to the target position T1, and transfers the pixel block at the target position T1. In the source buffer section 32, and also in the target position T3.
The pixel blocks in the target buffer section 33 are respectively written, and then, among the contents of the source buffer section 32, the pixels in the area where the mask data is "1" are written at the position of the target T3. As a result, the pixels of the target T1 are copied to the position of the target T3 in FIG. 13, and this result is displayed on the monitor 14. Note that temporarily transferring the pixel block located at the target position to be copied to the target buffer section 33 is as follows.
This is so that the original state can be immediately restored if the copy result is undesirable.

The above four types of commands have been explained in detail, but in addition to these, there are also commands for clearing the old target position data and the contents of the target buffer section 33, returning to the initial state, and executing pixel copy, and The contents of the buffer section 33 are copied to the original target position while referring to the mask data, and the image is restored to its original state.
Exit commands and the like to escape from pixel copying are provided, and by using these commands in various combinations, it is possible to remove pattern quirks.

The above is the operation in the interactive mode,
Commands, cursor coordinates, etc. instructed in the interactive mode are stored in the command recording unit 3 in the order in which they are operated using the input means 4.
7 is recorded.

[0071] When the interactive mode is finished, the batch mode is started next. In the batch mode, original endless pattern image data is written to the image memory 30, and commands or cursor coordinate values recorded in the command recording section 37 are sequentially read out and input to the command control section 35. , executed. As a result, the same processing that the operator did in the interactive mode is performed on the original endless pattern data to remove pattern quirks.

As described above, in the pixel copy method, even the pattern irregularities that occur in abstract patterns having relatively large patterns can be easily removed by simple operations.

Next, a Fourier transform method will be explained as a third method for removing pattern curl.

As mentioned above, pattern curl also occurs when there is uneven density within the unit material. Pattern curl removal using the Fourier transform method attempts to remove density unevenness that exists in a unit material that causes pattern curl at the endless pattern stage, and is usually performed using the aforementioned scrambling method or pixel copy method. This is done prior to the removal of pattern curls.

In order to remove pattern curl using the Fourier transform method, the pattern curl removal processing section 11 has the configuration shown in FIG. 14. Then, when the pattern pattern removal using the Fourier transform method is instructed, the image data of the endless pattern is transferred to the normalization unit 44, and the image data of the endless pattern is transferred to the normalization unit 44, and the value of the nearest power of 2 is extracted as the source, exceeding the number of pixels in the vertical and horizontal directions of the endless pattern. Let the number of pixels be the number of pixels, and
Repeat to fill in the excess as shown. This is called FFT (Fast Fourier Transform), which performs Fourier transform at high speed on images with a number of pixels that is a power of 2.
This is because the algorithm can be applied.

The source image converted into image data having a number of pixels that is a power of 2 by the normalization processing unit 44 is written to the source image memory 45 as is, while the source image is thinned out by a predetermined number of pixels and written to the frame memory 46. . This is used for testing with a small number of pixels, since Fourier transform takes processing time in proportion to the number of pixels. The contents of the frame memory 46 are then sent to the display processing means 13 and displayed on the monitor 14. Note that the memory capacity of the source image memory 45 is 1 byte/pixel for each color of Y, M, C, and K, so 8K x 8K x 4 (bytes).
The frame memory 46 may be approximately 1K×1K (bytes) for each color component of R, G, and B.

The contents of the source image memory 45 and the frame memory 46 are each subjected to two-dimensional Fourier transform by a Walsh transform unit 47, and then written to the Fourier transform memory 48. As shown in FIG. 16, among the contents written in the Fourier transform memory 48, components of the horizontal frequency Fh and vertical frequency Fv below a predetermined wave number W, for example, components below a predetermined wave number W, excluding the direct current component DC, are The low frequency component is canceled by a low frequency component canceling unit 49, and a two-dimensional inverse Fourier transform is performed by a Walsh inverse transform unit 50.
The image data related to the source image memory 45 is written to the inversely transformed source image memory 51, and the thinned image data related to the frame memory 46 is written to the frame memory 52. The image data written in the frame memory 52 is displayed on the monitor 14 via the display processing means 13, and by observing the screen of the monitor 14, the operator can confirm whether or not density unevenness has been removed.

Normally, the process of outputting from the frame memory 46 to the frame memory 52 is performed first, and after checking on the monitor 14, the image in the source image memory 45 is processed.

According to the applicant's experiments, it has been confirmed that almost all density unevenness can be removed by erasing wave numbers of 10 or less both horizontally and vertically as described above. Although it is assumed that the component is removed, it is clear that the input means 4 may be used to specify the wave number component to be removed.

Conventionally, if there was density unevenness in the unit material or baby version, the unit material was photographed again, but according to the present invention, density unevenness can be easily removed from the image data. , it is possible to reduce costs and reduce workload.

Although the above three types of pattern curl removal methods have been described, most pattern curls can be removed by appropriately combining these methods.

The above is the process of step S4 in FIG.
Once the pattern quirks are removed at the endless pattern stage, baby version data is created (step S5). The baby version is shown in Figure 17.
As shown in the figure, the endless pattern is created by simply repeating it vertically and horizontally, and a repeat processing section 12 is provided for this purpose. That is, when the input means 4 instructs to create a baby version, the repeat processing unit 12 is activated and arranges the image data of the endless pattern a predetermined number of times in the vertical and horizontal directions.

The baby version data created in this way is registered in the storage means 5.

[0084] When the creation of the baby version is completed, it is checked again whether or not pattern curl has occurred in the baby version (step S6). In this case, the presence or absence of pattern curl may be confirmed by thinning out the baby version data and displaying it on the monitor 14, but the size of the baby version is large, and moreover, it is necessary to make a presentation to the user. Therefore, the baby version data is supplied from the storage means 5 to the color hard copy 7 and hard copied.

[0085] If the occurrence of pattern curl is confirmed in step S6, the process returns to step S4 again to remove the pattern curl described above, but if the occurrence of pattern curl is not recognized, baby version data is output. (Step S7). There are two ways to output this baby version. One is to directly supply the baby plate data to the gravure engraving machine 6 to directly create a printing plate, and the other is to output the baby plate data to a film using the output scanner 2 and print the printing plate via the film. This is how to create. Which method is adopted is arbitrary.

[0086] Once the baby printing plate is created as described above, the printing plate is then used for proof printing, and it is checked again whether or not the pattern curl has occurred (Step S
8). If pattern curl has occurred, the process returns to step S1 and starts over from reading a new unit material, or returns to step S4 to remove pattern curl at the endless pattern stage. In the case of returning to step S1, the processes up to that point will be completely wasted, but compared to the conventional process, the process is automated, the cost is low, and the workload is reduced, so the waste can be minimized. It can be fastened.

[0087] If no pattern curl occurs in the proof print using the baby version, main version data is output (step S9).
). This is done by arranging the baby version data vertically and horizontally a predetermined number of times by the repeat processing unit 12.
The created main version data is registered in the storage means 5 and is output. The main version data can be outputted in the same way as the baby version data by supplying it to the gravure engraving machine 6 to directly create the main version, or by outputting it to a film using the output scanner 2 and passing the film through the printing plate. may be created.

[0088] After outputting the main version data as described above, proof printing is performed again (step S10), and if the final check is satisfactory, the main printing is performed using the main version (step S11). , the process ends. Incidentally, if it is confirmed that pattern distortion has occurred as a result of the proof printing of the main version, the process returns to step S1 or step S4 and is repeated.

Although one embodiment of the present invention has been described above, it will be obvious to those skilled in the art that the present invention is not limited to the above embodiment, and that various modifications can be made.

[0090]

[Effects of the Invention] As is clear from the above description, according to the present invention, retouching work that was conventionally extremely time-consuming and time-consuming can be done in a short time, thereby reducing costs. Not only that, but the delivery time can be shortened.

[0091] In addition, it is possible to easily deal with pattern irregularities that could not be removed by manual retouching.
Items can be expanded significantly.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the configuration of an embodiment of an abstract pattern plate-making system for printing building materials according to the present invention.

FIG. 2 is a diagram showing an example of a plate-making process when using the abstract pattern plate-making system for printing building materials according to the present invention.

FIG. 3 is a diagram for explaining watermark synthesis when creating an endless pattern.

FIG. 4 is a diagram for explaining the operation of a pattern curl confirmation processing section.

FIG. 5 is a diagram showing an example of a configuration for removing pattern curl using a scramble method.

FIG. 6 is a diagram for explaining replacement of pixels within a block in a scrambling method.

FIG. 7 is a diagram showing an example of a mask shape used for replacing pixels within a block.

FIG. 8 is a diagram for explaining the effectiveness of the pixel copy method.

FIG. 9 is a diagram showing an example of a configuration for removing pattern curl using a pixel copy method.

FIG. 10 is a diagram showing an example of a display screen when selecting a source pattern.

FIG. 11 is a diagram for explaining how to obtain a threshold value for creating mask data in a pixel copy method.

FIG. 12 is a diagram for explaining binarization processing for creating mask data.

FIG. 13 is a diagram for explaining the operation of the pixel copy method.

FIG. 14 is a diagram showing an example of a configuration for removing pattern curl using a Fourier transform method.

FIG. 15 is a diagram for explaining the operation of the normalization processing section.

FIG. 16 is a diagram for explaining the operation of a low frequency component erasing section.

FIG. 17 is a diagram for explaining creation of a baby version.

FIG. 18 is a diagram showing an example of a conventional plate-making process for abstract patterns for printing on building materials.

FIG. 19 is a diagram for explaining pattern habit.

[Explanation of symbols]

1 Input scanner 2 Output scanner 3 Control means 4 Input means 5. Storage means 6 Gravure engraving machine 7 Color hard copy 8 Abstract pattern plate making processing means 9 Endless processing section 10 Pattern pattern confirmation processing section 11　Pattern curl removal processing section 12 Repeat processing section 13 Display processing means 14 Monitor

Claims (1)

[Claims] Claim 1: an input scanner; a first means for generating pattern data of a predetermined size by performing watermark synthesis a predetermined number of times on unit material image data read by the input scanner; A second method of removing undesired patterns by creating mask data from pattern data at a predetermined position of the pattern data and copying pixel data at a desired position to another different desired position based on the mask data. An abstract pattern plate-making system for printing building materials, comprising at least means.