JPH09146262A - Halftone method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a high-definition halftone image without making the diameter of a laser beam for exposure smaller. SOLUTION: By reading one picture element value at of a digital multilevel image and comparing it with the threshold of a randam number generated by a random number generator 11 by a comparator 12 after adding a correction value E thereto, the picture element is binarized. A quantizing error E, in the case of binarizing the picture element is used to generate the correction value of the picture element read next in an error arithmetic operation part 15. Since the halftone image is obtained by binarizing every picture element of the digital multilevel image, the high-definition halftone image is obtained without making the diameter of the laser beam for exposure smaller.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a halftone method used for area-modulating a multi-valued image, and more particularly to a halftone method suitable for high-definition printing.

[0002]

2. Description of the Related Art The halftone method is a method of expressing a binary image by area-modulating a multivalued image, and an image obtained by the halftone method is called a halftone image. Therefore, the halftone image used for printing is also a halftone image that reproduces gradation depending on the size of the halftone dot. It is assumed here that each pixel of the digital multi-valued image is represented by 256 gradations. Also, for full-color printing, cyan,
The four-color digital multi-valued image data of magenta, yellow, and black are each halftoned. Since halftoning is basically performed by the same method for all four-color digital multi-valued image data, In the following, the halftoning of a digital multi-valued image of one color will be described.

Therefore, a conventional method for forming a halftone image from a digital multi-valued image by halftone dots will be described as follows.

When forming a halftone image by halftone dots, it is usual to form one halftone dot by a plurality of pixels of a digital multi-valued image, and now, one halftone dot is formed by 4 pixels of 2 × 2. Assuming that dots are formed, the relationship between the pixels of the digital multi-valued image and the halftone dots is as shown in FIG. FIG.
In, the white rectangle shows one pixel of the digital multi-valued image to be halftoned, and the circle filled with black shows one halftone dot. Therefore, it can be seen that one halftone dot is formed by four 2 × 2 pixels.

At this time, each pixel area of the digital multi-valued image is developed into binary dots of about 6 × 6 or 8 × 8. FIG. 6 shows an example in which a 1-pixel area of a digital multi-valued image is developed into 6 × 6 dots.

When forming halftone dots, 2 × 2
The area of the upper left pixel of the four pixel areas of FIG.
7B, the area of the pixel on the upper right is filled with the value of the dot from the lower right, as shown in FIG. 7B, the area of the pixel of the lower left is filled with the value of the dot on the upper right as shown in FIG. 7C. From the upper left, the area of the lower right pixel is filled with the value of the halftone dot from the upper left as shown in FIG. 7D.

Such a halftone method can be performed by the method shown in FIG. Assuming that the circled pixels of 128 gradations in the digital multi-valued image 1 are to be converted into halftone dots, the gradation value of the pixel is read and the comparator 3
In, each value in the threshold matrix 2 is compared.

The threshold matrix is 1 for a digital multi-valued image.
This is a table for converting pixels into halftone dots, and in the figure, the threshold matrix is 6 × 6. In this case, therefore, one pixel area of the digital multi-valued image is developed into 6 × 6 dots. .

Then, the comparator 3 compares each value of the threshold value matrix with the gradation value of the pixel, and when the value of the threshold value matrix is less than or equal to the gradation value of the pixel, the value of the halftone dot, for example, Prints 255, 0 otherwise.
As a result, for the pixel concerned, 2 as shown in FIG.
Value data is created. This is the halftone dot data.

Binarized halftone dot data is obtained by performing the above processing for all the pixels of the digital multi-valued image. The exposure laser beam is scanned based on the halftone dot data and the list is formed. Irradiate the mold film. At this time, for example, when the halftone dot data value is 255, the exposure laser beam is turned on and the film is irradiated, and when the halftone dot data value is 0, the exposure laser beam is turned off and the film is irradiated. Otherwise, a halftone image will be obtained by developing the film.

Since the pixel to be halftone dot in FIG. 8 is the pixel corresponding to the lower left quarter of the halftone dot, the pixel is 6 × 6 as shown in FIG. B9 by the above processing.
It becomes a halftone image developed into binary dots.

By the way, when evaluating the resolving power of a printed matter, a scale called screen ruling is usually used. This scale shows the number of halftone dots per inch when the halftone dots arranged periodically are viewed one-dimensionally. Therefore, assuming that the number of screen lines is L in FIG. 5, the pitch P of the halftone dots is 1 / L (inch), which is the pixel length of the pixel of the digital multivalued image, that is, 1 of the digital multivalued image.
The area S occupied by the pixels in the halftone image is 1/2
It becomes L (inch).

From the above, when the density of the exposure laser beam is viewed one-dimensionally, when one pixel of the digital multi-valued image is expanded to 6 × 6 dots, it is 12 times as large as one halftone dot, and the density of the digital multi-valued image is increased. When one pixel of the value image is expanded to 8 × 8 dots, it becomes 16 times as large as one halftone dot, and the density of the exposure laser beam becomes large. Conversely, the pixel density of a digital multi-valued image is
It is only 1/6 to 1/8.

[0014]

By the way, the number of screen lines is usually about 175 for general printed matter, but in recent years, the demand for high-definition printing has been increasing.
In printed matter called high-definition, halftone dots are usually made as fine as 400 to 600 lines or more.

Therefore, when performing halftoning for high-definition printing, first, it is conceivable to increase the input resolution of a digital multi-valued image and increase the number of screen lines.

However, in this case, the relationship between the input resolution in the number of screen lines and the beam diameter of the exposure laser beam is as shown in FIG. In the figure, “6 beams” indicates a case where one pixel of a digital multi-valued image is expanded to 6 × 6 dots, and “8 beams” indicates one pixel of a digital multi-valued image is 8 × 8. It shows the case of expanding to dots. Hereinafter, the same applies.

Therefore, according to the figure, when the screen frequency is 175 as in a general printed matter, the input resolution is 300.
(Dot / inch) is enough. In this case, the exposure laser beam diameter is 14 μm in the case of 6 beams and also in the case of 8 beams
Although it is 10 μm, if the screen frequency is 600, an input resolution of approximately 1200 (dot / inch) is required.
In this case, it can be seen that the exposure laser beam diameter is as small as 4 μm for 6 beams and 3 μm for 8 beams.

If the number of screen lines is increased in this way,
Since it is necessary to increase the pixel density in proportion to this and the exposure laser beam diameter is also small, a high-resolution output scanner or imagesetter must be used to perform output for high-definition printed matter. However, since such a high-resolution output scanner and imagesetter are expensive, there is a problem in that the cost of printing eventually increases.

FIG. 11 is a diagram showing the relationship between the screen frequency and the diameter of the halftone dots on the highlight side when the halftone dots have a circular shape. It can be seen that the diameter of the halftone dot becomes smaller in the highlight region.

However, the limit of the diameter of the halftone dots that can be reproduced in the ordinary plate printing process is about 20 μm, and the lower limit of the gradation to be reproduced is 3% or more, which is good in the ordinary plate printing process. It is possible to create a printing plate on the screen and print well. The screen frequency is 250 lines or less and the input resolution is 500 (dot / inch).
In the following case, so the screen frequency is 300 lines,
When the input resolution is 600 (dot / inch) or more, each dot exposed by the exposure laser beam cannot be reproduced well in the normal printing plate printing process, so in order to realize such high-definition printing However, there is also a problem in that a dedicated printing plate printing process that enables stable reproduction from finer dots is required.

Further, when attempting high-definition printing, the number of dots per unit area increases in proportion to the square of the number of screen lines, so that the processing time for halftoning also increases. is there.

The present invention is intended to solve the above problems, and an object thereof is to provide a halftone method capable of obtaining a high-definition halftone image without reducing the beam diameter of the exposure laser beam. It is what

[0023]

By the way, it is generally accepted that high-definition printing has a large number of screen lines, but the present inventor has found that the resolution of printed matter is digital multi-valued rather than the number of screen lines. It was found that it depends on the pixel density of the image.

Therefore, paying attention to this, the pixel density of the digital multi-valued image is increased to the dot density of the exposure laser beam to realize a high-definition printing halftone image.

That is, the halftone method of the present invention is characterized in that the digital multivalued image data to be halftoned is binarized for each pixel.

Therefore, according to the present invention, the pixel length of one pixel of the digital multi-valued image becomes the same as the beam diameter of the exposure laser beam, which is larger than that of the conventional high-definition printing. Since the diameter can be made relatively large, it is possible to obtain a high-definition halftone image for printing without using a high-performance imagesetter, and to perform stable dot reproduction in the plate printing process. Can provide a halftone image for the user.

[0027]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration example for realizing a halftone method according to the present invention.
10 is an adder, 11 is a random number generator, 12 is a comparator, 13
Is a subtracter, 14 is an error data buffer, and 15 is an error calculator. In the following, one pixel of a digital multi-valued image has 8 bits and 256 gradations, and the pixel level of the digital multi-valued image on the memory where the coordinate in the main scanning direction X is m and the coordinate in the sub-scanning direction Y is n. The key value is represented by P _mn .

The random number generator 11 generates one random number each time one pixel of a digital multi-valued image is read, and generates a random number.
Feed to 2. This random number has a value of 0 to 255, and its average value is preferably 128.

Now, assuming that P _mn is binarized and halftoned, the adder 1 is added to the read P _mn.
At 0, the correction value E is added. The correction value will be described later.

[0030] 'When _mn, the P' those obtained by adding the correction value E in P _mn P _mn is input to the comparator 12 and compared with the generated random number R _mn random number generator 11 at this time It is also supplied to the subtractor 13.

[0031] comparator 12 threshold value supplied from the random number generator 11, P _'mn outputs an H _mn = 0 in the case of less than or less R _mn,, otherwise H _mn =
255 is output. That is, H _mn takes only a value of 0 or 255, and thus the pixel P _mn is binarized.

The 'difference between _mn (H _mn -P' H _mn and P _mn) are calculated in the subtracter 13, the result is the quantization error E
It is stored in the error data buffer 14 as _mn . That is, the quantization error E _mn is written in the error data buffer 14 at a position where the coordinate in the X direction is m and the coordinate in the Y direction is n. Therefore, the error data buffer 14 has the same size as the digital multi-valued image, and the vicinity of E _mn is as shown in FIG.

The error calculator 15 calculates the correction value E based on the past quantization error stored in the error data buffer 14 and the preset weighting coefficient. Now, assuming that the weighting coefficient is a matrix as shown in FIG. 3A, the error calculator 15 calculates the correction value E by the following equation.

E = E _{m-1, n-1} / 16 + 5 * E _{m, n-1} / 16 + 3 * E _{m + 1, n-1} / 16 + 7 * E _{m-1, n} / 16 That is, the error calculator 15 Is to calculate the sum of products of the corresponding positions of the matrix of weighting factors and the past quantization error data to obtain the correction value E.

The matrix of weighting factors shown in FIG. 3A is of the so-called Floyd type, but a matrix of the Jarvis type shown in FIG. 3B may be used. It's good. When a Javies-type matrix is used, the correction value E is calculated by the following formula.

E = E _{m-2, n-2} / 48 + 3 * E _{m-1, n-2} / 48 + 5 * E _{m, n-2} / 48 + 3 * E _{m + 1, n-2} / 48 + E _{m + 2 , n-2} / 48 + 3 * E _{m-2, n-1} / 48 + 5 * E _{m-1, n-1} / 48 + 7 * E _{m, n-1} / 48 + 5 * E _{m + 1, n-1} / 48 + 3 * E _{m + 2, n-1} / 48 + 7 * E _{m-2, n} / 48 + 5 * E _{m-1, n} / 48 The correction value E thus obtained is added to P _mn by the adder 10 and The added value _P'mn is binarized as described above. Therefore, now P _mn =
If 125, E = -10, R _mn = 139, then _P'mn
= 125-10 = 115, which is smaller than R _mn ,
H _mn = 0.

By performing the above processing for all the pixels of the digital multi-valued image, it is possible to obtain binarized image data for each pixel of the digital multi-valued image. The error data buffer 14 is reset at the start of operation and
You can set it to 0.

Then, the exposure laser beam is scanned on the film by using the binarized image data thus obtained, and when H _mn = 0, the laser beam is turned off and H
_{When mn} = 255, a halftone image can be obtained by exposing with the laser beam turned on and developing the film.

The halftone method described above has various characteristics. One of the major features is that random numbers are used as threshold values for binarization. In this way, the random number is used as the threshold for binarization in the halftone method of the present invention. One pixel of a digital multi-valued image having a predetermined number of bits, for example, 8 bits, has 1 bit of information. 1
This is because gradation is not reproducible by using a fixed threshold value as a threshold value for binarization because it is converted into dots, but this results in overlapping moire and rosette. The effect is that the periodic structure of the binary image that causes moire can be removed.

However, if a random number is simply used as a threshold value for binarization, the graininess of the obtained halftone image may deteriorate. Therefore, the subtractor 1 subtracts the quantization error E _mn generated when the comparator 12 binarizes.
In order to realize a so-called error diffusion method, a correction value E is created by a circuit including the error data buffer 14, the error calculator 15 and the adder 10, and the correction value E is added to the read pixel value P _mn. This means is used to propagate to neighboring pixels. This is another major feature.

As a synergistic effect of these, it is possible to increase the pixel density of a digital multi-valued image and obtain a high-definition halftone image without reducing the beam diameter of the exposure laser beam. Confirmed by experiment.

Actually, the relationship between the input resolution of the digital multi-valued image and the beam diameter of the exposure laser beam when the halftone image is obtained by the above method is as shown in FIG. According to the method, an exposure laser beam having a beam diameter of 21 μm can be used even when the input resolution is 1200 (dot / inch), and as described above, the exposure laser beam has a beam diameter of 21 μm.
If it is μm, a normal plate printing process can be used. Therefore, the halftone image obtained by the halftone method of the present invention is good in each dot exposed by the laser beam for exposure in the conventional normal plate printing process. It will be clear that this can be reproduced.

As described above, compared with the conventional high-definition printing,
By using relatively large dots, it is possible to provide a high-definition halftone image for printing without the need for a high-performance imagesetter, and a stable halftone image for printing that enables stable dot reproduction in the plate printing process. Can be provided.

Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various modifications can be made. For example, the matrix of weighting factors used to obtain the correction value E is not limited to that shown in FIG.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration example for realizing a halftone method according to the present invention.

FIG. 2 is a diagram for explaining the structure of an error data buffer 14.

FIG. 3 is a diagram showing an example of a matrix of weighting factors used to obtain a correction value E.

FIG. 4 is a diagram for explaining the effect of the halftone method of the present invention.

FIG. 5 is a diagram for explaining the relationship between pixels of a digital multi-valued image and halftone dots when a halftone image is formed by halftone dots in the related art.

FIG. 6 shows a case where a pixel area of a digital multi-valued image is 6 × when forming a halftone image using conventional halftone dots.
It is a figure which shows the example at the time of developing into 6 dots.

FIG. 7 is a diagram for explaining formation of one halftone dot with four 2 × 2 pixels in the case of forming a halftone image using conventional halftone dots.

FIG. 8 is a diagram showing a configuration example for explaining a method for realizing a conventional halftone method using halftone dots.

9 is a diagram for explaining the operation of the configuration shown in FIG.

FIG. 10 is a diagram for explaining the relationship between the input resolution and the beam diameter of the exposure laser beam at each screen ruling in the conventional halftone method.

FIG. 11 is a diagram showing the relationship between the number of screen lines and the diameter of a halftone dot on the highlight side when the halftone dot shape is circular in the conventional halftone method.

[Explanation of symbols]

10 ... Adder, 11 ... Random number generator, 12 ... Comparator, 13
... subtractor, 14 ... error data buffer, 15 ... error calculator.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification number Internal reference number FI Technical indication H04N 1/405 H04N 1/40 104

Claims (1)

[Claims] 1. A halftone method comprising binarizing each pixel of digital multivalued image data to be halftoned.