JPH06223173A - Reduction converting method - Google Patents

Abstract

PURPOSE:To provide the reduction converting method which can obtain a reduction converted image even from a character which is originally broken owing to blurring without deformation and thickening and the a breaking of a new character while reducing the influence of the break of the original character. CONSTITUTION:The reduction converting method which obtains the reduction converted image by performing a process for decreasing the number of original pixels of an original image consisting of binary pixels decides whether or not there is a thin line as to the original pixels of the original image by the converted pixels of the reduction converted image at the time of correspondence between the original image and reduction converted image, and substitutes the value of the thin line in a converted pixel of the reduction converted image when it is decided that there is the thin line and the value of the thin line is a specific value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to image processing, and more particularly to a reduction conversion method for converting an image density of binary pixels to reduce the image.

[0002]

2. Description of the Related Art Reduction conversion of an image composed of binary pixels is
This is to obtain a reduced image with high image quality, which is perceived by humans as having little deterioration. Further, the reduction conversion of the image is also necessary when the characters read by the scanner or the facsimile are reduced and the characters are recognized. A normal OCR (optical character reading device) does not require reduction conversion because it sets character recognition parameters and prepares a recognition dictionary according to the reading resolution. However, for example, when a character output from a G4 facsimile apparatus having a resolution of 400 ppi is recognized using a G3 facsimile apparatus having a resolution of 200 ppi, reduction conversion is necessary. In such a reduction conversion, it is necessary not to change the characteristics of the character so as not to reduce the recognition rate.
It is required that crushing or thickening does not occur.

As such a reduction conversion method, conventionally, a method of simply sampling a pixel (original pixel) of an original image according to a conversion ratio to determine a value of a pixel (converted pixel) of the converted image, or an original image There is known a logical sum method in which a converted pixel is preferentially set to black when there is a black original pixel in the vicinity of the converted pixel when the converted image is mapped to. However, in this sampling method, the characters are not crushed or thickened, but the characters are cut off.The logical sum method does not cause the characters to be broken, but the crushed and thickened characters are generated. However, there was a drawback as a reduction conversion for character recognition.

As a method for solving this problem and preventing character breakage without causing crushing or thickening, there is a reduction conversion method based on the preservation of thin lines (Japanese Patent Application No. 59-270414).
Japanese Patent Application No. 60-38871, Japanese Patent Application No. 61-3838.
No. 1, Japanese Patent Application No. 62-236087, Japanese Patent Application No. 1-119.
539). These differ in configuration depending on the ratio of reduction conversion and the effect on unevenness when an image is read, but basically, thin lines are distinguished, and if there are thin lines, the values of thin lines are converted. This is a method of preventing the thin line from disappearing by substituting it into a pixel and saving the thin line.

Regarding these reduction conversion methods, Japanese Patent Application No.
A description will be given based on the example of No. 236087. FIG. 7 is a block configuration diagram. In FIG.
Reference numeral 1 is an original image reading means, 12 is a thin line determination means, 1
3 is a means for substituting the value of the thin line into the conversion pixel, and 14 is a means for determining the value of the conversion pixel. The thin line determination means 1
Reference numeral 2 is a combination of two means, "fine line detecting means" and "converted pixel determining means" described in Japanese Patent Application No. 62-236087.

The original image reading means 11 is a control means 1.
By controlling from 0, original pixels required for reduction conversion are read from the original image. The thin line determination means 12 determines the presence or absence of a thin line from the original pixel read by the original image reading means 11, and when it is determined that there is a thin line, the thin line value is processed by the means 13 for substituting the thin pixel value into the converted pixel. If it is determined that there is no thin line, the conversion pixel value determination unit 14 is requested to perform processing. The means 13 for substituting the value of the thin line into the converted pixel issues a next processing request to the control means 10 when the processing is completed.

The conversion pixel value determining means 14 determines the conversion pixel value by some method, substitutes the value into the conversion pixel, and then issues the next processing request to the control means 10. There are several methods for determining the value of the converted pixel in the determining unit 14, such as a method of selecting the majority value of the four original pixels in the vicinity of the converted pixel, but a method of selecting the value of the original pixel in the vicinity of the converted pixel. Is easy. With this configuration, reduction conversion is performed on the entire screen under the control of the control unit 10.

Next, a specific example of thin line determination by the thin line determination means 12 will be described with reference to FIG. FIG. 3 shows an example in which the vertical / horizontal co-conversion ratio is larger than 1/3 and smaller than 1/2, and FIG. 3A shows the original pixel obtained by the original image reading unit 11. (B) represents a conversion pixel of the reduced conversion image. The conversion pixel of (b) is one in which pixels of vertical 2 × horizontal 2 × vertical 3 × horizontal 3 of the original pixel are one converted pixel.

In the figure, circles and circles are pixels, and their coordinates are represented by Pn, m. □ is the conversion pixel 41 of interest. In the illustrated example, the original pixel closest to the target conversion pixel 41 is P0,0. Determination of Thin Line For the thin line, the original pixel surrounded by the windows X and Y is used. The window X includes three original pixels on the left and right of the target conversion pixel 41, and the window Y
Includes three original pixels above and below the target conversion pixel 41. For each original pixel surrounded by the windows X and Y, thin line determination is performed under the determination conditions shown in FIG.

As shown in FIG. 3A, when the pixels P1, -2 to 1,2 are black and the surroundings are white, that is, when there is a black thin line, ((b) in FIG. The condition 2) is satisfied, and it is determined that the conversion pixel 41 of interest has a thin line. When the determination is made in this way, the means 13 for substituting the value of the thin line into the conversion pixel in FIG. 7 substitutes black as the value of the thin line into the conversion pixel. If the conversion pixel of interest is at the position 42, the above method is performed for the conversion pixel of interest 42, it is determined that there is no thin line, and the conversion pixel of interest value determination unit 14 of FIG. The value of the pixel in the vicinity of 42, that is, white is determined as the value of the conversion pixel. By performing this determination work on all screens, the result shown in FIG.
The converted pixel shown in is obtained.

[0011]

In the above reduction conversion method, for example, when a character obtained from a facsimile apparatus is used as an original image, the density of the described character varies so that the character becomes faint and cut. There is a case. Such an example is shown in FIG. In the figure, P-4,0-
When the thin line of P5,0 is cut at the point P1,0, the pixels of P1, -2 to P1,3 are regarded as white thin lines in the above-described conventional reduction conversion method, and (2) in FIG. ) Is satisfied, and white is assigned to the converted pixel as the value of the thin line. Therefore, the reduced converted image is as shown in FIG. 5B, and the black thin lines corresponding to P-4,0 to P5,0 are cut off.

In the above-mentioned Japanese Patent Application No. 1-119539, there is provided means for determining a thin line when the length of the candidate for the thin line is not less than the specified length. Since lines such as P1, -2 to P1,3 in) are sufficiently long, white lines that cannot be cut are saved as shown in Fig. 5 (b).
Results in.

FIG. 6B shows an example of a reduced conversion image obtained by the above-described conventional reduction conversion method. In this example, the horizontal line of the katakana "U" is cut.

As described above, in the conventional reduction conversion method for preserving thin lines, white lines and black lines are equally stored, so that a character that is originally distorted is faithfully reproduced. However, there was a defect as a reduction conversion method for character recognition.

The present invention does not cause crushing or thickening even for a character that originally has a cut due to fading,
An object of the present invention is to provide a reduction conversion method capable of obtaining a reduced conversion image in which the effect of the original character breakage is reduced without causing new character breaks.

The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

[0017]

In order to achieve the above object, a representative means of the present invention is to obtain a reduced converted image by performing processing for reducing the number of original pixels of an original image consisting of binary pixels. In the conversion method, the presence or absence of a thin line is determined for the original pixel of the original image for each conversion pixel of the reduced conversion image when the original image and the reduced conversion image are associated, and it is determined that there is a thin line and the value of the thin line. Is a specific value, the value of the thin line is substituted into the conversion pixel of the reduced conversion image.

[0018]

According to the above-mentioned means, the white thin line due to the break is not stored. Moreover, since the black thin line is preserved, no new cut occurs. Therefore, it is possible to obtain a reduced converted image in which the influence of the original character breakage is reduced.

The structure of the present invention will be described below with reference to an embodiment in which an image formed of normal pixels in black is used as an image composed of binary pixels.

In all the drawings for explaining the embodiments, parts having the same functions are designated by the same reference numerals, and the repeated description thereof will be omitted.

[0021]

【Example】

[Embodiment 1] FIG. 1 is a block diagram for explaining Embodiment 1 of the present invention. In the figure, 10 is a control unit, 11 is an original image reading unit, 12 is a thin line determination unit, 21 is a thin line value determination unit, 22 is a processing request synthesizing unit, and 13 is a thin line value assigned to a conversion pixel. Means, 14
Is a means for determining the value of the converted pixel. The reference numerals 10 to 14 are the same as those in the case of the conventional reduction conversion method for preserving thin lines in FIG. Further, the thin line determination condition used by the thin line determination means 12 is as shown in FIG.

The original image reading means 11 is a control means 1.
The original pixel required for the reduction conversion is read from the original image by control from 0. The thin line determination unit 12 determines the presence or absence of a thin line from the original pixel read by the original image reading unit 11.
When it is determined that there is a thin line, the thin line value determination unit 21
A processing request is issued to the thin line value, and if it is black, the processing request is sent to the means 13 for substituting the value of the thin line into the conversion pixel. Here, the value of the thin line is substituted as the value of the converted pixel.

On the other hand, when the thin line determination means 12 determines that there is no thin line, and when the thin line value determination means 21 determines that the thin line value is white, both are passed through the processing request combining means 22. A processing request is issued to the conversion pixel value determining means 14, and the conversion pixel value determining means 14 determines the conversion pixel value as a value of a nearest original pixel in a predetermined method. This process is performed for all screens under the control means 10.

The detailed description of the thin line determination by the thin line determination means 12 is made in the same manner as the method in FIG. 3 which has already been used in the description of the conventional example. The resulting converted image is similar to that of FIG. Therefore, a further description here is omitted.

Next, this example will be described with respect to an example in which the original image shown in FIG. Figure 5
When the thin line determination unit 12 determines the conversion pixel 41 of interest in the original image of (a), P1, -2 to P1,
Since 3 pixels satisfy the condition of (2) in FIG. 4A, it is considered that there is a white thin line. At the same time, the pixels of P-4,0 to P5,0 satisfy the condition (10) of FIG. 4A, and are therefore considered to have a black thin line. Next, the result of the determination of the two thin lines is output to the determination unit 21 for the value of the thin lines.
Here, of the two thin lines, there is a thin line for which the value of the thin line is determined to be black. Therefore, as the value of the thin line, black is substituted into the conversion pixel by the means 13 for substituting the value of the thin line into the conversion pixel. By such an operation, only the black thin line is saved, and the white thin line due to the break is not saved. Therefore,
In the example of FIG. 5 (a) used for the description of the conventional example described above, P1,
Since the values of the thin lines from -2 to P1,3 are determined to be white, they are not saved, and the reduced converted image is as shown in FIG. 5 (c).

As described above, according to this example, a reduced conversion image can be obtained without causing crushing or thickening even for a character originally having a discontinuity due to fading. Figure 6
An example of a reduced conversion image according to this example is shown in (a). As is clear from comparison with the conventional example (b), it can be understood that the horizontal line of "c" does not have a break.

[Embodiment 2] In the method of Embodiment 1, when it is determined that there is no thin line, the value of the nearest original pixel is substituted into the converted pixel, and the value of the converted pixel is substituted. The value of the nearest original pixel is substituted for the converted pixel even if the determination is not made when the original pixel of
The thin line does not disappear, and the processing can be simplified. That is, in FIG. 4A showing the determination conditions for the thin lines, the determinations (4) to (6) regarding the vertical thin lines and the determinations (10) to (12) regarding the horizontal thin lines can be omitted. In this example,
A thin line determination condition that does not determine when the original pixel closest to the converted pixel itself is a candidate for a thin line and the value of the original pixel closest to the converted pixel when it is determined that there is no thin line. This is an effective method for substituting.

FIG. 2 is a block diagram for explaining the present example. In the figure, the control means 10; original image reading means 11; thin line determination means 12; thin line value substitution means 13 for converting pixels; belongs to. However, the determination of the thin line determination unit 12 adopts the determination condition of FIG. 4B, which is a determination condition that does not include the determination in the case where the original pixel itself closest to the converted pixel is a candidate for the thin line.

In this example, a determining unit 31 for determining the value of the nearest original pixel is inserted between the reading unit 11 for the original image and the determining unit 12 for the thin line. The means 31 determines the value of the original pixel closest to the converted pixel, and when the original pixel is white, a processing request is made to the thin line determination means 12, and when the original pixel is black,
A processing request is made to the processing request combining means 22. Further, the output of the processing request synthesis means 22 is the conversion pixel value determination means 1
Instead of 4, the value of the nearest original pixel is input to the means 32 for substituting into the converted pixel. Here, the processing request combining unit 22
In response to the processing request from, the value of the nearest original pixel is substituted into the conversion pixel, and the next processing request is issued to the control means 10.

With the above configuration, under the control of the control means 10, the original pixel is read by the original image reading means 11 and the judgment is made by the nearest original pixel value judging means 31. Here, if the value of the nearest original pixel is determined to be white, a processing request is issued to the thin line determination means 12, and the thin line determination is performed here. If there is a thin line, the value of the thin line is converted into a pixel. A processing request is issued to the substituting means 13 for substituting the value of the thin line as the value of the converted pixel.

On the other hand, if the nearest neighbor original pixel value is judged to be black by the nearest neighbor original pixel value judging means 31 and if the thin line judging means 12 judges that there is no thin line, the processing request composition is performed. A processing request is issued via the means 22 to the means 32 for substituting the value of the nearest original pixel into the converted pixel, where the value of the original pixel nearest to the converted pixel is substituted into the converted pixel. This process is performed for all screens under the control of the control means 10. Since the operation is performed in this way, in this example, the thin line is stored only when the original image closest to the converted pixel is the white thin line.

A detailed description of the thin line determination by the thin line determination means 12 will be given. In the example of the original pixel of FIG. 3, the nearest original pixel of the conversion pixel 41 of interest is P0,0, and its value is white. Is. Therefore, the processing request is issued from the determination means 31 of the value of the nearest original pixel to the determination means 12 of the thin line. Since the condition (2) of the conditional expression in FIG. 4B is satisfied, the thin line determination unit 12 determines that there is a thin line and makes a processing request to the unit 13 that substitutes the value of the thin line into the conversion pixel. Fine line determination means 1
Since the value of the thin line determined in 2 is black, the means 13 for substituting the value of the thin line into the conversion pixel substitutes black into the conversion pixel.
Therefore, the converted image obtained as a result is as shown in FIG. 3B, and the fine line is not broken.

Next, the operation of this example will be described with respect to an example in which the original image shown in FIG.
The nearest original pixel of the conversion pixel 41 of interest is P0,0, and its value is black. Therefore, a processing request is issued from the closest original pixel value determining means 31 to the processing request combining means 22, and the means 32 for substituting the nearest original pixel value for the converted pixel substitutes black for the converted pixel. Therefore, the resulting converted image is as shown in FIG. 5C, and the fine line is not broken.

In this example, the conversion pixel 41 of interest.
Depending on the position of, the cut of the original image may be directly output to the converted image.

For example, in FIG. 5, the nearest original pixel is P
When it becomes 1,0, the determination means 31 of the value of the nearest original pixel
Determines that the value of the nearest original pixel is white, and the thin line determination means 1
Issue a processing request to 2. The thin line determination means 12 is shown in FIG.
When the determination condition of (b) is not satisfied, the thin line is determined to be absent, and a processing request is issued to the means 32 for substituting the value of the nearest original pixel into the conversion pixel via the processing request combining means 22.
The means 32 for substituting the value of the nearest original pixel into the conversion pixel substitutes the white of the value of the nearest original pixel into the conversion pixel.

As a result, the thin lines of the converted image will be cut. However, the present embodiment 2 has an effect of performing reduction conversion by reducing existing cuts as compared with the conventional example, even when the cut of the original image may be directly output to the converted image. is there.

Although the invention made by the present inventor has been specifically described based on the embodiments, the invention is not limited to the above-mentioned embodiments and various modifications can be made without departing from the scope of the invention. Needless to say. For example, in the above example, it is explained that black characters are drawn on a white background, but if white characters are drawn on a black background, the image is reversed from white to black. On the other hand, the present invention may be applied, or the present invention may be implemented by reversing the determination of white and black.

[0038]

As described above, according to the present invention,
Never save white thin lines due to breaks. Moreover, since the black thin line is preserved, no new cut occurs. In addition, the present invention does not store white thin lines due to cuts, and thus does not cause thickening or crushing of characters. Therefore, a reduced conversion image that does not cause crushing or thickening, does not cause a new character break, and reduces the influence of the original character break even if the character originally has a break due to blur Can be obtained.

The method of the present invention can be realized by adding a very small number of functions to the conventional reduction conversion method for preserving fine lines, and is of a smaller scale than the method of connecting white lines to remove fine lines. It can be added and processed at high speed.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of a first embodiment of the present invention.

FIG. 2 is a block configuration diagram of a second embodiment of the present invention.

FIG. 3 is a diagram showing an example of thin line determination of an original image.

FIG. 4 is a diagram showing a determination condition for a thin line.

FIG. 5 is a diagram showing an example of thin line determination of an original image with a break.

FIG. 6 is a diagram showing a reduced conversion image obtained by the present invention and a conventional example.

FIG. 7 is a block diagram of a conventional example.

[Explanation of symbols]

10 ... Control means, 11 ... Original image reading means, 12 ...
Thin line determination means, 13 ... Substitution of thin line values into converted pixels, 14 ... Converted pixel value determination means, 21 ... Thin line value determination means, 22 ... Processing request combining means, 31 ... Nearest original pixel , 32 ... Means for substituting the value of the nearest original pixel into the converted pixel, 41, 42 ... Converted pixel of interest.

Claims (2)

[Claims] 1. A reduction conversion method for obtaining a reduced conversion image by performing a process of reducing the number of original pixels of an original image composed of binary pixels, in a reduced conversion image when the original image and the reduced conversion image are associated with each other. For each conversion pixel, the presence or absence of a thin line is determined for the original pixel of the original image, and when the thin line is determined to be present and the value of the thin line is a specific value, the value of the thin line is set to the conversion pixel of the reduced conversion image. A reduction conversion method characterized by substituting. 2. A reduction conversion method for obtaining a reduction conversion image by performing a process of reducing the number of original pixels of an original image composed of binary pixels, in a reduction conversion image when the original image and the reduction conversion image are associated with each other. The value of the original pixel closest to the converted pixel is determined, the thin line is determined for the original pixel of the original image for each of the converted pixels, the thin line is determined to be present, and the value of the original pixel is a specific value. A reduction conversion method, wherein the value of the thin line is substituted into the conversion pixel at times.