JPH05191632A - Binary picture processor - Google Patents

Abstract

PURPOSE:To obtain a magnified binary picture with high quality by storing or receiving simultaneously a binary picture subject to adaptive processing and an image separation result with respect to a multi-value processing picture and utilizing the separated result when the binary picture is subject to picture element density conversion. CONSTITUTION:An image area separation section 4 outputs the result of discrimination for each picture element as to whether an image is a character or not a character. A selection circuit 5 selects a binary output from a character processing section 2 or a binary output from a pattern processing section 3 for each picture element according to the discrimination result from the image area separation section 4. A memory 6 stores an output from the image area separation section 4 and an adaptive processing picture from the selection circuit 5. A picture element density conversion section 7 receives an image area separation picture from the memory 6 and the adaptive processing picture to implement picture element density conversion.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus,
In particular, the present invention relates to a binary image processing device that performs a scaling process on a binarized input image.

[0002]

2. Description of the Related Art Conventionally, in a device such as a facsimile machine for processing binarized image information, a scaling process or a pixel density conversion process is performed on a binary image. For example, in a communication method called memory communication in facsimile, binarized image data is temporarily stored in a memory and communication is started after a predetermined time has elapsed. Then, after the communication is started, the specifications of the other party are known, and if the size of the other party's printer is smaller than the size of the original to be transmitted, the reduction / magnification processing is performed on the binarized data stored in the memory.

By the way, binary images are roughly classified into two types, that is, a character image composed of thin lines and a picture image expressed in pseudo gradation, and a scaling method for these two types of images is required. The nature is different. That is, in a character image, fine line gaps and blurring become a problem, and it is important to preserve the structure of the fine line, whereas in the picture image, the arrangement of fine pixels has little meaning,
Preservation of average density and texture within the neighborhood is important.

As described above, with respect to a binary image original image in which characters and a pattern are mixed in one original, it is determined whether it is a character region or a pattern region by measuring the input binary image. , It is necessary to perform a scaling process suitable for each area (if the target original image is only full-face characters or patterns, it is recognized by some means, for example, by key input, each original image is It is possible to switch the scaling process to, but 2
Such processing is not possible for the original value image).

Therefore, in order to perform high-quality scaling processing on an image in which characters and patterns are mixed, image area separation processing is performed on the input binary image, and the pixel of interest is part of the character. If it is a part of the character, scaling processing is performed to retain the resolution. If it is not a part of the character,
Several adaptive scaling processing methods have been proposed in which a scaling processing is performed in which gradation and texture are preserved (for example, see Japanese Patent Laid-Open No. 3-11877).

[0006]

In the technique described in the above publication, the image area separation process is performed using binary original image information. That is, a binary image is subjected to multi-value estimation (multi-valued), scaling is performed, and then the binary image is subjected to local character / pattern determination processing by 4 × 4 pattern matching. However, since the above-described image area separation is performed on the image data that has already been binarized, it cannot be said that the accuracy of the separation is good, and rather, it is more likely that the image data that is the source of the image data is more than the image data that has already been binarized. There is a problem in that the accuracy of separation is significantly improved when the image area separation is performed on the value image data.

On the other hand, the image filing apparatus has the following problems. That is, conventionally, in order to store a large amount of image data, the image data is binarized, or the binarized data is encoded and stored. Then, the accumulated binary image data is displayed on the CRT.
When displaying on a display device such as the CRT, since the CRT can represent multi-values, the binary image data is multi-value estimated (multi-valued) and displayed. Generally, a pixel counting method is used as this multivalued estimation method. That is, a mask of a predetermined size (for example, 8 ×
If it is 8, 64 gradations are possible), and the number of black pixels is counted, and the counted value is used as the value of the central pixel of the mask. Then, when pixel density conversion (that is, resolution conversion) is required at this time, thinning (reduction processing) or double writing (enlargement processing) is performed on the obtained multivalued image data. However, the above-mentioned method has a problem that characters are blurred because the smoothing process acts on the character image.

An object of the present invention is to store or receive an adaptively processed binary image and an image area separation result for a multivalued image at the same time, and use the separation result when converting the binary image into pixel density. It is therefore to provide a binary image processing device capable of obtaining a high-quality scaled binary image.

Another object of the present invention is to simultaneously store or receive the adaptively processed binary image and the image area separation result for the multivalued image, and when the multivalued image is obtained from the binary image, the separation is performed. An object of the present invention is to provide a binary image processing device that can obtain a high-quality multivalued estimated image by utilizing the result.

[0010]

In order to achieve the above object, in the invention according to claim 1, a first binary image obtained by binarizing a multi-tone digital signal is converted into a second binary image having a different pixel density. In a binary image processing device for performing a scaling process on a binary image, based on at least two or more pixel density converting means, a result of image area separation for the multi-tone digital signal, or a result of image area separation and a scaling ratio. And a means for switching the pixel density conversion means for each pixel area.

According to a second aspect of the present invention, there is provided a binary image processing device for performing a scaling process on a first binary image obtained by binarizing a multi-tone digital signal into a second binary image having different pixel densities. , At least two or more pixel density converting means, means for separating the image area of the multi-gradation digital signal, means for storing the image area separation data together with the first binary image data, and It is characterized in that a means for switching the pixel density conversion means for each pixel area is provided based on the result, or the result of image area separation and the scaling factor.

According to a third aspect of the present invention, in a binary image processing apparatus for converting a binary image obtained by binarizing a multi-tone digital signal into a multi-valued image, at least two or more multi-valued means and the multi-valued means are provided. It is characterized in that a means for switching the multi-valued means for each pixel area is provided based on the result of the image area separation for the gradation digital signal.

According to a fourth aspect of the present invention, in a binary image processing device for converting a binary image obtained by binarizing a multi-gradation digital signal into a multi-valued image, at least two or more multi-valued conversion means and the multi-valued conversion means are provided. Means for separating the gradation digital signal from the image area,
It is characterized by further comprising means for storing the image area separation data together with the binary image data, and means for switching the multi-value conversion means for each pixel area based on the result of the image area separation.

According to a fifth aspect of the invention, means for binarizing the multi-gradation digital signal, means for separating the multi-gradation digital signal into image areas, binarized image data and the image area separation. It is characterized in that it is provided with means for transmitting data.

According to a sixth aspect of the invention, the image area separation is performed by determining whether or not the pixel of interest is a part of a character edge.

[0016]

The image area separating section is composed of an edge area detecting circuit, a white background detecting circuit and a judging circuit, and outputs a judgment result of whether each pixel is a character or a non-character. The selection circuit selects the output from the character processing unit and the output from the pattern processing unit for each pixel according to the determination result from the image area separation unit. In the memory of this embodiment, the output from the image area separation unit and the adaptive processed image from the selection circuit are stored. In the pixel density converter,
The image area separation image and the adaptive processing image from the memory are input, and the pixel density conversion means is selected for each pixel area based on the image area separation result. As a result, it is possible to perform a scaling process while maintaining high image quality even for an image in which a picture and characters are mixed.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram of a binary image processing apparatus (for example, a facsimile machine) according to an embodiment of the present invention. In FIG. 1, a scanner 1 has a photoelectric conversion element such as a CCD camera, reads a document, outputs a grayscale signal in black and white, and converts the grayscale signal into a 6-bit digital signal for output. As will be described later, the image area separation unit 4 is composed of an edge area detection circuit, a white background detection circuit, and a determination circuit, and outputs the determination result for each pixel from the original image.

The binary image processing apparatus of this embodiment further includes
A character processing unit 2 including an MTF correction circuit 21 and a fixed threshold binarization circuit 22 and a pattern processing unit 3 including a binarization circuit 31 using an error diffusion method are provided. The selection circuit 5 for selecting an image signal selects binary output from the character processing unit 2 and binary output from the pattern processing unit 3 for each pixel according to the determination signal from the image area separation unit 4. The memory 6 of this embodiment is
Output from the image area separation unit 4 and character processing from the selection circuit 5 /
The image-processed adaptive processed image is stored, and the pixel density conversion unit 7 receives the image area separation image and the adaptive processed image from the memory 6 and performs pixel density conversion.

As described above, in the present embodiment, since the output from the image area separation unit 4 and the adaptively processed image are stored in the memory 6, it is possible to perform high-quality scaling processing during memory communication, and The double-processed image data is transmitted to another facsimile. It should be noted that at the time of immediate communication, the image is transmitted without going through the memory 6.

The character processing unit 2 uses the MTF correction circuit 21 to transfer the image data from the scanner 1 to the MT shown in FIG.
After applying the F correction filter, the fixed binarization circuit 22 performs binarization processing with a predetermined threshold value.

The picture (non-character) processing section 3 binarizes by the error diffusion method. In the error diffusion method, as shown in FIG. 3, the density x (i, j) of the pixel of interest in the input image corresponds to each pixel y of the binary image of which the pixel value has already been determined near the pixel of interest.
Error e (i, j) caused by thresholding (i, j)
Method for obtaining a binary image by taking a correction value f (i, j) obtained by adding a weighted sum E (i, j) of the weight w (i, j) of Is. That is, y (i, j) = th {f (i, j)} th (f) = 1 (when f> th), 0 (when other) f (i, j) = x (i, j) ) + E (i, j) E (i, j) = Σw (n, m) e (n, m) e (n, m) = f (n, m) -y (n, m).

Other binarization methods can be adopted depending on the required image quality or transmission efficiency. For example,
After smoothing with a filter as shown in FIG.
It may be binarized by the spiral dither shown in FIG.

FIG. 6 is a view showing the arrangement of the image area separating section 4,
An edge area detection circuit 41, a white background detection circuit 42,
It is composed of a determination circuit 43. The determination process in the image area separation unit 4 is as follows. The determination result is output in 1 bit for each pixel. That is, (1) When the pixel of interest is determined to be an edge region and a white background region is present within a predetermined horizontal range with respect to the pixel of interest, it is determined to be a character region and a 1-bit output signal “1” is output. Output. (2) When the pixel of interest is other than that, it is determined to be a non-character area, and a 1-bit output signal “0” is output.

The details of the edge area detection circuit, the white background detection circuit, and the determination circuit will be described below.

As a concrete example of the edge area detection circuit 41, the applicant / applicant previously applied for a character / picture determination method (Japanese Patent Application No. 11-132242) or the applicant of the present invention, the Technical Report of IEICE (Vol. .90 No. 168 I
E90-29 to 38, July 27, 1990). FIG. 7 is a diagram showing a configuration of the edge area detection circuit 41 that outputs an edge area (bit 1) and a non-edge area (bit 0) for each pixel. The edge area is detected as follows. That is, after performing the MTF correction with the filter shown in FIG. 2, in the binarization circuit,
The image signal is converted into a binary signal of black / non-black and white / non-white by a predetermined threshold value.

The connected black pixel detection circuit detects whether or not each black pixel in the image signal of the black pixel group from the binarization circuit is a black pixel connected to another surrounding black pixel. FIG. 8 shows a 3 × 3 matrix pattern for detecting the connected black pixels. In the figure, black circled pixels indicate black pixels, and blank pixels may be either black / white pixels. That is, in the pattern shown in the figure, the connected black pixels are detected by utilizing the fact that the black lines are connected in any of the vertical, horizontal, and diagonal directions in the black edge region. When the pattern of black pixels matches any of the patterns in the figure, the central pixel of the matrix is detected as a black pixel connected to other black pixels in the surroundings.

The connected white pixel detection circuit detects whether or not each white pixel in the image signal of the white pixel group from the binarization circuit is a white pixel connected to other surrounding white pixels.
FIG. 9 shows a 3 × 3 matrix pattern for detecting the white connected pixels. In the figure, white portions indicate white pixels, and blank pixels may be either black / white pixels. In other words, the pattern in the figure detects white connected pixels by utilizing the fact that white lines are connected in any of the up, down, left, and right directions in the white edge area, and the pattern of white pixels is one of the patterns in the figure. When it matches with, the central pixel of the matrix is detected as a white pixel connected to other surrounding white pixels.

The connected black pixel counting circuit counts the number of connected black pixels in, for example, a 3 × 3 matrix centered on the target pixel, and outputs the target pixel as active when the count value is a certain value or more. To do. Similarly, the connected white pixel counting circuit counts the number of connected white pixels in, for example, a 3 × 3 matrix centered on the target pixel, and outputs the target pixel as active when the count value is equal to or more than a certain value. To do.

The AND circuit outputs a logical product signal of the black active / non-active signal from the connected black pixel counting circuit and the white active / inactive signal from the connected white pixel counting circuit. That is, when black active and white active are present at the same time in a 3 × 3 matrix centered on the pixel of interest, the AND circuit outputs “1”, and at this time, this pixel of interest is an edge area pixel.

Further, expansion processing is performed on a 5 × 5 mask. The expansion processing is, for example, processing all pixels in the 5 × 5 matrix as edge area pixels if there is even one edge area pixel in the 5 × 5 matrix centered on the pixel of interest. The edge area detection circuit 41 outputs an edge area signal “1” for each pixel and a non-edge area signal “0”.

FIG. 10 shows the white background detection circuit 42 of this embodiment.
Shown in. The detection circuit 42 detects whether or not a group of white pixels of a predetermined size (1 × 5 or 5 × 1) exists near the pixel of interest. This is because there is a large white background area in the background of the character, but the above-mentioned edge area is also detected in the pattern, but the edge area with no white background in the vicinity is judged as a non-character area. Is represented.

The determination result of the white background detection is as follows. (1) When there is a white background area within a predetermined horizontal range across the pixel of interest, a 1-bit output signal "1" is output. (2) When there is no white background area within a predetermined horizontal range across the pixel of interest, a 1-bit output signal “0” is output.

The detection of the white background is performed as follows. That is, the MTF correction circuit is applied to the image signal as shown in FIG.
The image signal is binarized into white / non-white by a predetermined threshold value (white pixel is made active and bit 1) by a binarization circuit after sharpening processing by applying the MTF correction filter shown in FIG. Input to the pixel detection circuit.

FIG. 11 shows a pattern of white pixels, that is, a white background pattern in which 1 × 5 or 5 × 1 are all white pixels. The white background pixel detection circuit detects a group of white pixels in the vicinity of the target pixel by pattern matching, and sets the target pixel as an active pixel (bit 1) when matching the pattern shown in FIG.

In the expansion circuit, the output result of the white background pixel detection circuit is expanded using the 3 × 9 mask shown in FIG. That is, if there is even one active pixel in the mask shown in FIG. 12, the target pixel in the mask is set as a white background pixel. Such expansion processing is repeated for each pixel, and 1 is output if the pixel of interest is active, and 0 is output otherwise.

Judgment circuit 43 in the image area separation section 4 of FIG.
Then, the edge area detection circuit 41 and the white background detection circuit 4
The following signals are output based on the output signal from 2. That is, for the pixel of interest, (1) the output signal from the edge area detection circuit 41 is "1".
When the output signal from the white background detection circuit 42 is "1", the character area signal ("1") is output. (2) For combinations other than the above, a non-character area signal (“0”) is output.

The selection circuit 5 of FIG. 1 responds to the output (character = 1 / non-character = 0) from the image area separation section 4 by the character processing section 2
The binary output (1/0) from (1) and the binary output (1/0) from the pattern (non-character) processing unit 3 are selected for each pixel.

The memory 6 stores the output (character = 1 / non-character = 0) from the image area separation unit 4 and the adaptive processed image (1/0) from the selection circuit 5. At this time, in order to improve the compression rate, an appropriate encoding (for example, MH
(Encoding, MR encoding) may be performed.

The pixel density conversion section 7 receives the two image signals from the memory 6, that is, the image area separation image signal and the adaptive processed image signal, and performs pixel density conversion (reduction / magnification processing in a facsimile). At this time, when encoded and stored in the memory, a decoding process is performed. FIG.
Shows a configuration of the pixel density conversion unit 7, and includes a pixel density conversion unit 71 for an image area separation image and a pixel density conversion unit 72 for a character.
And a pixel density conversion unit 73 for the design, and conversion units 72 and 73.
It is composed of a selection circuit 74 for selecting the output from.

The pixel density conversion unit 71 for the image area separation image is
Depending on the scaling factor, the scaling is performed by the logical sum method. Here, the logical sum method means four pixels a, b, c, around the point P whose brightness is unknown.
If at least one pixel of d is black, the point P
Is a method of making the value of black be black, and this method has an advantage that the fading of characters can be prevented because the processing is simple and the preservation of the image data at the time of image reduction is good. .. Further, the pixel pixel density conversion unit 72 for characters is
Scale with the logical sum method described above.

FIG. 1 shows the configuration of the pixel density conversion unit 73 for a picture.
4 shows. The binary-multivalue conversion unit converts the input binary image data into multivalued data having the same pixel density. In this embodiment, it is configured by filtering using a 3 × 3 size mask shown in FIG. It This filter is a low-pass filter and is a process for improving the image quality of the scaling process. That is, the pseudo gradation image is
The gradation is pseudo-expressed by the average density of the value pixels, and a fine shape within the range does not have much meaning, and it causes image quality deterioration such as moiré in the scaling process. By removing such high frequency components in advance, the image quality of the pseudo gradation image that has been subjected to the scaling process can be improved.

The multi-value scaling unit is a scaling unit for converting multi-valued data into multi-valued data having different pixel densities.
By performing linear interpolation using the three pixels near the conversion pixel shown in (3), multi-valued data having different pixel densities are converted. FIG. 17 shows the target pixel A and the original pixel P in the vicinity of the target pixel A when the converted target pixel A is mapped onto the original image.
It is a figure which shows Q, R, and S. In the present embodiment, three of these four points are linearly interpolated to obtain the density value of the pixel of interest A after conversion (this is what the applicant of the present application filed as Japanese Patent Application No. 3-108728). Have already applied).

That is, the coordinate value (x, y) of the target pixel A
Is x> y (type 1), the four original pixels P, Q,
When three points P, Q, and S are selected from R and S, and the coordinate value (x, y) of the target pixel A is x ≦ y (type 2), four original pixels P, Q, and R , S, three points P, R, S are selected, and the density values f (A) of the target pixel A are determined by linearly interpolating the density values of the three points P, Q, S or P, R, S. calculate. That is, to explain the case of type 1,
In the triangular surface PQSP shown in FIG. 17, three points P, Q, S
Letting f (P), f (Q), and f (S) be the density values of, respectively, the pixel value (density value) f (A) of the target pixel A is f (A) = (f (Q) − f (P)) x + (f (S) -f (Q)) y + f (P). Then, the multi-value binary conversion unit binarizes by the error diffusion method shown in FIG.

As described above, according to the binary image processing apparatus of this embodiment shown in FIG. 1, the pixel density conversion means is selected for each pixel area based on the image area separation result for the multi-tone digital signal. Therefore, it is possible to perform a scaling process while maintaining high image quality even for an image in which a picture and characters are mixed.

In this embodiment, it is also possible to select the pixel density conversion means in consideration of the scaling factor. That is, for example, the low-pass filter of the binary-to-multi-value conversion means is composed of an anti-alias filter, and the required cutoff frequency differs depending on the magnification. Therefore, FIG. , The coefficient matrix shown in FIG. 16 is used. Then, the processing using each filter is performed by the high-magnification-pixel density conversion processing for the picture,
This is called low-magnification-pixel density conversion processing for patterns, and the processing is switched as shown in FIG. 18 depending on the image area separation result and the magnification change.

<Embodiment 2> In a facsimile, relaying between facsimiles is a case where pixel density conversion processing is performed on a binary image. In this case, as shown in FIG. 19, when the facsimile B receives the binary image from the facsimile A and transmits it to the facsimile C again, reduction scaling is required. In this case, in the facsimile A of FIG. 19, if the adaptive processed image output from the selection circuit of FIG. 1 and the image area separated image are mixed and transmitted,
In the facsimile B of FIG. 19, adaptive pixel density conversion is possible.

That is, as shown in FIG. 20, the image transmitting apparatus for synthesizing and transmitting the adaptively processed image and the image area separated image is the facsimile A in FIG. 19, and the facsimile B which is its receiving apparatus is in the figure. Pixel density conversion is performed with the configuration shown in FIG. 13, and the pixel density converted image is transmitted to the facsimile C. The format of the image transmitted by the image transmission apparatus of FIG. 20 is 2 bits per pixel (image area separation image information is 1 bit, adaptive processing image information is 1 bit), and the image is in the format as shown in FIG. Is transmitted.

<Embodiment 3> FIG. 22 is a diagram showing the structure of a filing apparatus of this embodiment. In the configuration of the apparatus, elements other than the multi-value estimating section and the CRT are the same as those of the embodiment shown in FIG. 1, so the multi-value estimating section will be described here.

The multi-valued estimation unit receives two image signals from the memory, that is, the image area separation image and the adaptively processed image, and performs multi-valued estimation. FIG. 23 is a diagram showing the configuration of the multi-value estimating unit, so that the output from the character multi-value estimating process and the output from the pattern multi-value estimating process are selected based on the image area separation result. It is configured.

That is, in the character multi-value estimation processing, 1 /
The data of 0 is a through. Here, 1 to 6
3 is assigned to 0, and 0 is assigned to 0. In addition, in the multi-value estimation process for a picture (non-character), the number of black pixels in a matrix of a predetermined size is set as the gradation value of the pixel of interest. Here, the matrix is 8 × 8, and the maximum gradation value is 63. As a result, the multi-valued conversion is adaptively performed, and the data is displayed and output on the CRT of the filing apparatus of this embodiment.

By the way, in an apparatus having an adaptive multi-level estimation section for adaptively multi-valued as shown in FIG. 23, an image area separation image and an adaptive processing image are received from another apparatus, and multi-valued A binary-to-multivalue converter for estimating the value is also conceivable.
That is, this is a case where the receiving-side apparatus shown in FIG. 23 is equipped with a CRT, which makes it possible to configure a facsimile having a soft copy function that does not output unnecessary information to paper.

[0052]

As described above, according to the first aspect of the present invention, the pixel density converting means is selected for each pixel area based on the result of the image area separation. Magnification processing (that is, reduction processing) during relay
In the case of applying, it is possible to perform the scaling processing while maintaining high image quality for a binary image in which a picture and characters are mixed. According to the second aspect of the present invention, at least two or more pixel density converting means, means for separating the multi-tone digital signal from the image area, image data obtained by binarizing the multi-tone digital signal, and image area separation data. And a means for switching the pixel density conversion means for each pixel area based on the result of the image area separation or the result of the image area separation and the scaling ratio. Therefore, the scaling processing is performed during the memory communication. When applying
It is possible to perform a scaling process while maintaining high image quality for a binary image in which a picture and characters are mixed. According to the third aspect of the invention, at least two or more multi-valued means and means for switching the multi-valued means for each pixel area based on the result of the image area separation for the multi-tone digital signal are provided. Therefore, when the transmitted binary image is multivalued, it is possible to multivalue the binary image in which a picture and characters are mixed while maintaining high image quality. According to the invention described in claim 4, at least two or more multi-value converting means, means for separating image areas of multi-gradation digital signals, means for storing image area separation data together with binary image data, and image area Since the means for switching the multi-valued means for each pixel area based on the result of the separation is provided, when the binary image stored in the memory or the like is multi-valued, a binary image in which a pattern and characters are mixed is formed. On the other hand, it becomes possible to multi-value while maintaining high image quality. Claim 5
According to the described invention, means for binarizing the multi-tone digital signal, means for separating the multi-tone digital signal into image areas, and means for transmitting the binarized image data and the image area separation data In the apparatus according to claim 1, which receives the transmitted data, high image quality is maintained for a binary image in which a picture and a character are mixed when the scaling processing is performed at the time of relaying. 4. The apparatus according to claim 3, which is capable of performing scaling processing while receiving the transmitted data, when the transmitted binary image is multi-valued,
It becomes possible to multi-value a binary image in which a picture and characters are mixed while maintaining high image quality. According to the invention described in claim 6, since the image area separation is performed on the multi-valued digital image by utilizing the continuity of the character pixels, the binary image in which the picture and the character are mixed is highly accurately. Image area separation processing is performed. Therefore, in the device according to any one of claims 1 to 4, while performing high-resolution processing and multi-value conversion while maintaining high image quality for a binary image in which patterns and characters are mixed. Is possible.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of a binary image processing apparatus of this embodiment.

FIG. 2 is a diagram showing an example of an MTF correction filter.

FIG. 3 is a diagram illustrating an error diffusion method.

FIG. 4 is a diagram showing an example of a smoothing filter.

FIG. 5 is a diagram showing an example of a spiral dither.

FIG. 6 is a diagram showing a configuration of an image area separation unit.

FIG. 7 is a diagram showing a configuration of an edge area detection circuit.

FIG. 8 is a diagram showing a pattern for detecting connected black pixels.

FIG. 9 is a diagram showing a pattern for detecting connected white pixels.

FIG. 10 is a diagram showing a configuration of a white background detection circuit.

FIG. 11 is a diagram showing a pattern for white background detection.

FIG. 12 is a diagram showing an example of an expansion mask.

FIG. 13 is a diagram showing a configuration of a pixel density conversion unit.

FIG. 14 is a diagram showing a configuration of a pixel density conversion unit for a picture.

FIG. 15 is a diagram showing a high-magnification low-pass filter.

FIG. 16 is a diagram showing a low-magnification low-pass filter.

FIG. 17 is a diagram illustrating linear interpolation.

FIG. 18 is a diagram showing a method of selecting a scaling unit.

FIG. 19 is a diagram illustrating relay between facsimiles.

FIG. 20 is a diagram showing a configuration of an image transmission device of the present embodiment.

FIG. 21 is a diagram showing a format for transmitting an image.

FIG. 22 is a diagram showing a configuration of a filing apparatus of this example.

FIG. 23 is a diagram showing a configuration of a multi-value estimation unit.

[Explanation of symbols]

 1 Scanner 2 Character Processing Section 3 Picture Processing Section 4 Image Area Separation Section 5 Selection Circuit 6 Memory 7 Pixel Density Conversion Section

Claims (6)

[Claims] 1. A first binarized multi-tone digital signal.
A binary image processing apparatus for performing a scaling process on the binary image of No. 2 into a second binary image having different pixel densities, at least two or more pixel density converting means, and a result of image area separation for the multi-tone digital signal. Alternatively, the binary image processing device is provided with means for switching the pixel density conversion means for each pixel area based on the result of image area separation and the scaling factor. 2. A first binarized multi-tone digital signal
In the binary image processing device for scaling the binary image of No. 2 into the second binary image having different pixel densities, at least two or more pixel density converting means and means for separating the multi-tone digital signal from the image area. And means for storing the image area separation data together with the first binary image data, and the pixel density for each pixel area based on the result of the image area separation or the result of the image area separation and the scaling factor. A binary image processing apparatus comprising means for switching conversion means. 3. A binary image processing apparatus for converting a binary image obtained by binarizing a multi-tone digital signal into a multi-valued image,
Binary image processing characterized by comprising at least two or more multivalued means and means for switching the multivalued means for each pixel region based on the result of image area separation for the multi-tone digital signal. apparatus. 4. A binary image processing device for converting a binary image obtained by binarizing a multi-tone digital signal into a multi-valued image,
Based on at least two or more multi-value converting means, means for separating the image area of the multi-tone digital signal, means for storing the image area separation data together with the binary image data, and a result of the image area separation. A binary image processing apparatus, characterized by comprising means for switching the multi-valued means for each pixel area. 5. A unit for binarizing a multi-tone digital signal, a unit for separating an image region of the multi-tone digital signal,
A binary image processing device comprising means for transmitting the binarized image data and the image area separation data. 6. The binary according to claim 1, wherein the image area separation is performed by determining whether or not the pixel of interest is a part of a character edge. Image processing device.