JPH0670138A - Method and device for image processing - Google Patents

Abstract

PURPOSE:To acquire a halftone image close to an original image by deciding the halftone data on the picture elements of the halftone image corresponding to a picture element matrix based on the number of black or white picture elements. CONSTITUTION:In a conversion state, a binary image 100 is vertically and horizontally divided into the picture element matrices every 3 (=n) picture elements. This matrix has a 4X4 (mXM) size equal to a dither matrix. Therefore the image 100 is divided into 16 picture element matrices. Then the number of black picture elements included in each divided picture element matrix are counted and the halftone data are decided after conversion based on the counted number of black picture elements. In other words, a white picture element is assigned to the corresponding picture element of a halftone image with the number of black picture elements included in a picture element matrix is equal to 0 with a black picture element assigned with 16 black picture elements and an intermediate level accordant with the number of black picture elements assigned with 1-15 black picture elements respectively. Thus the image 100 is converted into a halftone image 101 of 17 (=4<2>+1) gradations at most.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to image processing for converting a binary image into a reduced halftone (multivalued) image.

[0002]

2. Description of the Related Art This type of image processing is required when a monochrome binary image is reduced and displayed on a display capable of displaying halftones (multivalues).

Conventionally, in such a case, a binary image is divided into a two-dimensional pixel matrix having a size corresponding to a reduction ratio,
There is known a method of determining halftone data of a pixel on a reduced image corresponding to a pixel matrix according to the number of black pixels (or white pixels) in the pixel matrix (for example, Japanese Patent Laid-Open No. 1-296772).

An example of vertically and horizontally reducing a binary image to one-third will be described with reference to FIG. In FIG. 6, 600 indicates a binary image, and 601 indicates a halftone image after conversion. In addition,
A small grid shows one pixel.

First, an 18 × 18 binary image 601 is converted into 3 ×
The pixel matrix of 3 is divided vertically and horizontally into every 3 pixels. Two
Regions of the value image 601 separated by thick lines indicate a 3 × 3 pixel matrix. It should be noted that the numbers 1 to 36 are the pixel numbers of the pixel matrix and the pixels of the reduced image 601, and the pixel matrix and the pixels having the same number correspond to each other on a one-to-one basis.

Next, the number of black pixels (or white pixels) in each pixel matrix is counted, and the halftone (multi-value) data of the corresponding pixel of the reduced image is determined according to the count value. For example, among 9 pixels in a 3 × 3 pixel matrix, white is used when there are no black pixels, black is used when all pixels are black pixels, and an intermediate value corresponding to the number of black pixels is used when there are 1 to 8 black pixels. level,
Thus, up to 10 gradations are assigned.

As described above, conventionally, when a binary image is converted into a halftone image reduced to 1 / n (where n is a positive integer), the binary image is vertically and horizontally n by an n × n pixel matrix. It is divided for each pixel, and halftone data is assigned according to the number of black or white pixels in each image matrix. As a result, a halftone reduced image having a maximum [n ² +1] gradation can be obtained. In general, the larger the number of gradations to be assigned, the better the image quality of the reduced image.

[0008]

At present, the systematic dither method is used as a method for reproducing a grayscale original image such as a photograph in binary with black and white because of its good image quality, processing principle and simple circuit configuration. It is often done. However, when a binary image (pseudo-halftone image) obtained by such a systematic dither method is converted by the above method, a problem may occur in the converted halftone image. This will be described below.

FIG. 7A shows an example of a halftone dot type dither matrix (threshold matrix) used in the systematic dither method. In this dither matrix, the gradation (level) of the original input image is set to 16 gradations of 0 to 15, and 0
The thresholds up to 15 are distributed according to a rule in a 4 × 4 matrix.

This dither matrix is repeated in the vertical and horizontal directions of the original image to determine the threshold value for all images. Then, the halftone image is converted into a binary image by comparing the level of the pixel in the original image with the threshold value for the pixel and determining whether to make it white or black.

FIG. 7B shows the dither matrix of FIG. 7A when all 4 × 4 pixels of the original image are at the same level (each level of 0, 3, 6, 9, 12, 16). 7 shows a pattern (dither pattern) of a binary image obtained by the systematic dither processing using.

FIG. 8 shows the binary image obtained for the larger level 9 region of the original image. This binary image 80
As can be seen from 0, in the same level portion of the original image, the same pattern appears repeatedly in a cycle of 4 × 4 pixels corresponding to the dither matrix. The 4 × 4 area delimited by the solid line is the area corresponding to the dither matrix.

Now, consider the case where the binary image (pseudo-halftone image) 800 shown in FIG. 8 is converted into a halftone image reduced to one-third by the above-described conversion method. FIG. 9 is an explanatory diagram thereof.

As shown in FIG. 9 separated by thick lines, the same binary image 900 as that of FIG. 8 is divided vertically and horizontally into every 3 pixels by a 3 × 3 pixel matrix (1st to 16th). The number of black pixels in each pixel matrix is 4 for the 8th, 11th, and 12th pixel matrices, 6 for the 3rd, 5th, 6th, and 15th pixel matrices, and 5 for the other pixel matrices. Becomes Therefore, the halftone image 9 is obtained as the conversion result.
01 is obtained.

As described above, although the original image of the binary image has a uniform level (level 9 in this example), the halftone image after the reduction conversion becomes an image having a level difference for each pixel. . Then, by converting a wider binary image having the same pattern as the binary image 900 of FIG. 9, a halftone image 1000 shown in FIG. 10 is obtained, which has a regular pattern (moiré) which was not present in the original image. ) Has occurred.

This is a case where a binary image composed of only dither patterns of the same level is converted. However, even when a binary image including a plurality of types of dither patterns of different levels is converted, the converted halftone image. Gradation may deteriorate compared to a binary image.

In order to deal with such a problem, when allocating halftone data from the number of black pixels in the pixel matrix, the same data will be used even if the number of black pixels is close, and the pixel level after conversion will be the same. Although it is possible to make them the same, it is clear that deterioration of gradation is inevitable compared to the original image.

The present invention has been made in view of the above-mentioned problems, and it is based on the systematic dither method for halftone images such as photographs.
An object of the present invention is to provide a method and apparatus for reducing and converting a value image into a high-quality halftone image having excellent gradation.

[0019]

In order to achieve the above object, the present invention provides a binary image by a tissue dither method using a threshold matrix of m × m (where m is a positive integer),
When reducing to 1 / n (where n is a positive integer) and converting to a halftone image, a binary image is vertically arranged with a pixel matrix of km × km (where k is a positive integer). And a configuration in which halftone data of pixels of a halftone image corresponding to the pixel matrix is determined in accordance with the number of black pixels or white pixels in the pixel matrix in the horizontal direction. It is a thing.

[0020]

According to the present invention, as described above, the size of the pixel matrix is set to be an integral multiple of the threshold matrix.
Since the number of black pixels or white pixels included in the pixel matrix is uniform in the uniform level portion of the original image of the value image, the level difference does not occur in the pixels of the corresponding portion of the converted halftone image, and as a result, Generation of noise such as moire can be suppressed. Also, a maximum [m ² +1] gradation can be assigned according to the number of black pixels or white pixels in the pixel matrix. Therefore, it is possible to obtain a high-quality halftone image having a gradation similar to the original image before the processing by the systematic dither method.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A specific example of conversion by the image processing method of the present invention will be described with reference to FIGS.

FIG. 1 is an explanatory diagram in the case of converting a binary image by a systematic dither method using a 4 × 4 dither matrix into a halftone image reduced to one-third. 100
Shows an example of a binary image to be converted. This binary image is obtained by processing the level 9 part of the original image by the systematic dither method using the dither matrix shown in FIG. 7 (a). It corresponds to the one given. Reference numeral 101 denotes a halftone image after conversion.

In the conversion, first, the binary image 100 is divided in the vertical and horizontal directions by a pixel matrix every 3 (= n) pixels. However, the pixel matrix has the same size of 4 × 4 (= m × m) as the dither matrix (here, k = 1). Therefore, the binary image 100
Are divided into 16 pixel matrices in such a manner that adjacent pixels overlap each other in the vertical and horizontal directions, as indicated by a bold line.

The numeral 1 shown in the binary image 100
16 to 16 are pixel matrix numbers. Also, the numbers 1 to 16 attached to each pixel in the halftone image 101 indicate the numbers of the corresponding pixel matrix.

The number of black pixels included in each pixel matrix thus divided is counted, and the converted halftone data is determined according to the number of black pixels. That is, if the number of black pixels in the pixel matrix is 0, the number of white and black pixels is 16
Then, if black and the number of black pixels are 1 to 15, the intermediate level corresponding to the number of black pixels is assigned to the corresponding pixel of the halftone image. That is, the maximum value of the binary image 100 is 17 (= 4 ² +
1) Converted to a halftone image 101 of gradation.

In the case of the binary image 100 shown in FIG. 1, since the number of black pixels included in all pixel matrices is 9, level 9 is assigned to all the pixels of the converted halftone image 101. This level is the same as the level of the original image, and there is no level difference between the images.

FIG. 2 shows an example of a halftone image obtained when a wider binary image having the same pattern as the binary image 100 shown in FIG. 1 is processed in the same manner. This halftone image 200
However, no pixel level difference occurs and the level matches the level of the original image.

As can be seen from the example described here,
According to the method of the present invention, it is possible to obtain a halftone image which does not cause moire, which has been a problem in the past, and which has a gradation property close to that of the original image.

The size of the dither matrix (m)
When the reduction ratio (n) is larger than that in (1), a similar high-quality halftone image can be obtained by setting k to 2 or more and performing conversion.

Next, an example of the image processing apparatus of the present invention will be described with reference to FIGS. FIG. 3 is a schematic block diagram of the image processing apparatus. The image processing device shown here has four
A process of converting a binary image by the systematic dither method using a × 4 dither matrix into a halftone image reduced to ⅓, that is, the conversion described with reference to FIG.

In this image processing apparatus, binary image data to be processed (bit data of "1" for black and "0" for white),
The select signal, the clock signal, and the latch signal are input from terminals 308, 309, 310, and 311 respectively, and the converted halftone data (4 bits / pixel) of the halftone image is output from the terminal 312.

This image processing apparatus has shift registers 301, 3 having a capacity capable of storing binary image data for one line.
02, 303, 304, a selector 305 for switching input data of the shift register 301, a ROM 306 storing halftone data, and a latch circuit 307 for latching output data of the ROM 306.

The selector 305 has two inputs A and B, and selects and outputs the A input signal or the B input signal according to the select signal. Binary image data from the outside is input to the A input, and the most significant bit data of the shift register 301 is input to the B input. The output data of the selector 305 is input to the shift register 301.

The most significant bit data of the shift register 301 is input to the shift register 302, and the most significant bit data of the shift register 302 is the shift register 303.
The data of the most significant bit of the shift register 303 is input to the shift register 304. The upper 4-bit data (16 bits in total) of the shift registers 301, 302, 303, 304 are input to the ROM 306 as address data. Shift registers 301, 302, 3
Each of 03 and 304 takes in the input data to the least significant bit every time a clock signal is input from the outside, and shifts the held data by one bit to the upper (right) side.

The ROM 306 has 16 addresses from 0 to 15 according to the number of "1" bits in the address data specified by the 16-bit address data.
It stores halftone data (4 bits) of gradation. The halftone data output from the ROM 306 is latched by the latch circuit 307 every time a latch signal is input from the outside, and output from the terminal 312 to the outside.

The overall operation of the image processing apparatus configured as described above will be described below. To facilitate understanding of the operation, an operation timing chart is shown in FIG. 4 and an operation flowchart is shown in FIG.

First, a select signal is set on the side where the selector 305 selects the A input (S1 in FIG. 5), and in this state, the clock signal is continuously input, and in synchronization with this, one line of binary image data is input. Minutes are continuously input (S2). The binary image data is input to the shift register 301 via the selector 305, sequentially captured by the least significant bit by the clock signal, and then shifted to the upper side. In this way, one line of binary image data (Line of FIG. 4 is
1) is held in the shift register 301. The most significant bit of the shift register 301 is the data of the first pixel of the line.

Subsequently, the subsequent binary image data for three lines (Line2, Line3, Line4 in FIG. 4) are input in synchronization with the clock signal. This image data is sequentially taken into the shift register 301, and the image data held in the shift register 301 is sequentially shifted to the lower shift registers 302, 303, 304. Thus, the first line (L
ine1) and the second line (Li
ne2), the third line (Lin
e3), the fourth line (Line
The binary image data for 4 lines in 4) is held. At this time, the shift registers 301, 302, 303, 304
The upper 4 bits of 16 bits in total, which is a total of 16 bits, is pixel data of the first 4 × 4 pixel matrix in the area of the 4 lines, and halftone data (after conversion) corresponding to the number of black pixels in the pixel matrix. The corresponding pixel level of the halftone image) is output from the ROM 306.

Next, the state of the select signal is inverted (S
4), the selector 305 comes to select the B input.
Next, one pulse of the latch signal is input (S5), and the ROM 30
The output data of No. 6 is latched by the latch circuit 307, and the halftone data of the first pixel is output from the terminal 312 to the outside.

Next, three pulses of the clock signal are input (S
6). Shift registers 301, 302, 303, 304
The binary image data therein is shifted upward by 3 bits. At this time, the shift registers 301, 302, 303
The most significant bit data of are sequentially input to the lower shift registers 302, 303, 304. Further, the most significant bit data of the shift register 1 is input again to the shift register 1 via the selector 5. Thus, the 16-bit data of the next 4 × 4 pixel matrix is stored in the ROM.
The data is input to 306, and halftone data corresponding to the number of black pixels in the pixel matrix is output from the ROM 306.

Next, one pulse of the latch signal is input (S
7), the halftone data output from the ROM 306 is latched by the latch circuit 307 and output to the outside.

After that, by repeating the operations of S6 and S7, halftone data corresponding to a 4 × 4 pixel matrix obtained by horizontally dividing the first four-line area into every three pixels is sequentially determined, and is output to the outside. It will be output. This operation continues until the shift registers 301, 302, 303, 304 have finished shifting one line of the binary image (S8).

When this is completed, the select signal is next inverted (S9), and the next three lines (Line5, Line).
6, Line 7) binary image data is input. Before the start of this input, the shift register 301 has a fourth line (L
Since the binary image data of (ine) has been accumulated, the shift register 304 of Line 4 is stored in the shift register 304.
3 for Line 5 and shift register 302 for Line 6
The binary image data of Line 7 is stored in the shift register 301.

That is, the last 1 of the 4 lines processed previously
Binary image data of the next four lines overlapping the line is accumulated in the shift registers 301, 302, 303, 304. Then, by repeating the operation of S4 to S7, the 4-line area is divided into three pixels in the horizontal direction.
The halftone data corresponding to the x4 pixel matrix is sequentially determined and output to the outside via the latch circuit 307.

The above operation is performed until the end of one page (S1
0) By continuing, conversion of one page of binary image is completed.

As can be understood from the above description, in the present image processing apparatus, at the input stage of binary image data, the binary image is divided into 4 pixels every 3 pixels (generally n pixels) in the vertical direction. Divide by a size of × 4 (generally k · m), and
Binary image data of (generally km) lines are cut out, and then shift registers 301, 302, 303, 30
By the data shift operation on 4, the image data is divided into 3 (generally m) pixels in the horizontal direction to obtain 4 × 4 (generally k · m) binary image data for each pixel matrix. It is cut out.

Such a configuration is often advantageous in terms of reducing the circuit scale and increasing the processing speed. However, the input binary image data is temporarily stored in the RAM, and the binary image data on this RAM is subjected to the same vertical and horizontal divisions and the extraction of the binary image data for each pixel matrix by software or hardware. You may.

In the image processing apparatus, the shift registers 301, 302, 303, 304 are RAM and
The same operation may be performed by substituting a combination of counters for the read and write address control. Instead of the ROM 306, a hardware or software counter or adder may be used to count the number of black pixels in the pixel matrix and output the count value as it is as halftone data.

[0049]

As is apparent from the above description, according to the present invention, when converting a binary image by a systematic dither method using a photograph or the like as an original image into a halftone image reduced to 1 / n, A binary image is divided into n pixels in the vertical and horizontal directions by a pixel matrix having a size that is an integer multiple of the threshold matrix at the time of dynamic dither processing, and the pixel is divided according to the number of black pixels or white pixels in the pixel matrix. Since the halftone data of the pixels of the halftone image corresponding to the matrix is determined, there is little noise such as moire, and it is possible to obtain a high-quality halftone image with a gradation similar to the original image. is there.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a conversion example according to an image processing method of the present invention.

FIG. 2 is a diagram showing an example of a halftone image after conversion.

FIG. 3 is a block diagram showing an example of an image processing apparatus according to the present invention.

FIG. 4 is an operation timing chart of the image processing apparatus.

FIG. 5 is an operation flowchart of the image processing apparatus.

FIG. 6 is an explanatory diagram of a conversion example according to a conventional method.

FIG. 7A is a diagram showing an example of a dither matrix. FIG. 7B is a diagram showing an example of a dither pattern.

FIG. 8 is a diagram showing an example of a binary image by the systematic dither method.

FIG. 9 is an explanatory diagram of an example of conversion of a binary image by a conventional method by the systematic dither method.

FIG. 10 is a diagram showing an example of a halftone image in which moire has occurred.

[Explanation of symbols]

 100 binary image 101 halftone image 301 shift register 302 shift register 303 shift register 304 shift register 305 selector 306 ROM 307 latch circuit 308 binary image data input terminal 309 select signal input terminal 310 clock signal input terminal 311 latch signal input terminal 312 Halftone data output terminal

Claims (2)

[Claims] 1. A binary image obtained by the systematic dither method using a threshold matrix of m × m (where m is a positive integer),
An image processing method for converting into a halftone image reduced to 1 / n (where n is a positive integer), which is a binary image with a pixel matrix of km × km (where k is a positive integer). In the vertical direction and in the horizontal direction by n pixels respectively, and determine halftone data of the pixels of the halftone image corresponding to the pixel matrix according to the number of black pixels or white pixels in each pixel matrix. An image processing method characterized by: 2. A binary image obtained by a tissue dither method using a threshold matrix of m × m (where m is a positive integer) is represented by n
An image processing apparatus for converting a halftone image reduced to 1 / (n is a positive integer), and a unit for inputting binary image data and a binary image data input by this unit. On the other hand, a binary image is represented by a pixel matrix of km × km (where k is a positive integer) in the vertical and horizontal directions.
Means for dividing each pixel matrix and cutting out binary image data for each pixel matrix, and generating halftone data according to the number of bits of black pixels or white pixels in the data cut out by this means An image processing apparatus comprising means and means for outputting the halftone data generated by this means.