JP3212339B2 - Image magnification device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image scaling device for enlarging or reducing a binary image at an arbitrary magnification.

[0002]

2. Description of the Related Art Document images handled by facsimile apparatuses and various image processing apparatuses include black-and-white images such as characters and line drawings (hereinafter collectively referred to as character images) and halftone images such as pictures and photographs. is there.

Normally, image information of a character image is image information obtained by reading a document image with a scanner and simply binarizing the obtained image signal with a certain threshold level. On the other hand, image information of a halftone image is image information obtained by binarizing an image signal obtained by reading a document image by pseudo halftone processing such as dither processing or error diffusion processing.

In the above-described apparatus, the number of pixels of the image information is increased or decreased by a linear interpolation method or the like, so that the image size is frequently changed.

[0005] In general, the quality of a character image deteriorates little due to such scaling, but the quality of a halftone image greatly deteriorates due to scaling. For example, when a halftone image is greatly enlarged due to an increase in the number of pixels, moiré occurs in the image or a density change is conspicuous, so that the image quality deteriorates.

In order to reduce such image quality deterioration, for example, as disclosed in Japanese Patent Application Laid-Open No. 2-239380, binary image information is once multi-valued and scaled, and then binarized again. Things have been suggested. Thus, by scaling the multi-valued image information, in the case of a halftone image,
Moire and unnatural sudden changes in density are eliminated, and image quality is improved.

Conventionally, however, binary image information is multi-valued by a constant calculation method regardless of the type of image to be processed and the zoom ratio.

In this case, when the image information of the character image is multi-valued, pixels where white density or black density is supposed to have an intermediate density increase, and the outline of the character becomes unclear. Rather, the image quality was degraded. Also, as in the case of the character image, the edge portion of the halftone image has an unclear outline and the image quality is deteriorated.

[0009]

As described above, conventionally,
When the binary image information is once multi-valued and scaled, there is a problem that the image quality is degraded at the edge portion of the character image or the halftone image.

SUMMARY OF THE INVENTION It is an object of the present invention to solve the above-mentioned problems and to provide an image scaling device capable of always obtaining good image quality for both character images and halftone images.

[0011]

For this purpose, each invention of the present application focuses on each pixel of binary image information and multi-values each pixel by an operation using the pixel of interest and surrounding pixels as parameters. According to the first aspect of the present invention, in the calculation, when the magnification ratio of the image size is small, the number of surrounding pixels as parameters is reduced, and when the magnification ratio is large, the number is reduced. I try to make more.

In the second invention, it is determined whether each pixel is an edge portion or a non-edge portion of an image. In the above calculation, when the pixel of the edge portion is multi-valued, a parameter is determined. When the number of surrounding pixels to be used is small, and the number of pixels in the non-edge portion is multi-valued, the number is increased.

In the third aspect, when each pixel is an edge portion, the edge direction is further determined, and the position of a peripheral pixel as a parameter is changed according to the edge direction.

According to a fourth aspect of the present invention, in the above calculation, in the case of image information obtained by simple binarization, the number of surrounding pixels as parameters is reduced, and the image information is obtained by pseudo halftone processing. In the case of image information, the number is increased.

In the fifth invention, when the image information obtained by the simple binarization and the image information obtained by the pseudo halftone processing are mixed, the two are discriminated. In the case of pixels, the number of surrounding pixels as parameters is reduced, and in the case of pixels in the pseudo halftone processing portion, the number is increased.

[0016]

When the number of surrounding pixels used as a parameter is increased in the above calculation for multi-value conversion, the gradation difference between the pixels is smoothed, so that the image quality of a halftone image is improved. On the other hand, in a character image or an edge portion of an image, the image becomes clearer when the number of surrounding pixels serving as parameters is smaller.

In the first aspect of the invention, when the zoom ratio is small, the number of surrounding pixels used as the above parameters is reduced, so that the image quality of the character image can be prevented from deteriorating. Further, when the magnification ratio is large and the deterioration in the halftone image becomes conspicuous, the number is increased, so that the deterioration of the halftone image can be reduced.

In the second invention, at the edge portion of the image,
Since the number of the surrounding pixels is reduced and the number of the surrounding pixels is increased in the non-edge portion, the image quality deterioration of each image portion is reduced.

In the third invention, the positions of the surrounding pixels as the above parameters are further changed depending on the edge direction.
The pixel density is corrected according to the image content, and the image quality is further improved.

According to the fourth aspect of the invention, in the case of image information of a character image which is usually simply binarized, the number of surrounding pixels is reduced, so that image quality deterioration of the character image is reduced, and the pseudo halftone is normally used. In the case of processed image information, the number of processed image information is increased, so that deterioration of image quality of a halftone image is also reduced.

According to the fifth aspect of the invention, when the image information by the simple binarization and the image information subjected to the halftone processing are mixed, the same operation and effect as described above can be obtained.

[0022]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing an image scaling device according to a first embodiment of the present invention. In FIG. 1, a binary image input unit 1 inputs binary image information, and is, for example, an image receiving unit of a scanner device or a facsimile device. The multi-level conversion unit 2 focuses on each pixel of the input binary image information, and multi-values each pixel by an operation using the pixel of interest and surrounding pixels as parameters.
The calculation method switching means 2a switches the number of surrounding pixels and the calculation formula used as parameters in the calculation.

The scaling unit 3 enlarges or reduces the image size by increasing or decreasing the number of pixels according to a known technique according to the scaling factor of the image size. The binarizing unit 4 binarizes the multi-valued image information by an error diffusion method. The image output unit 5 outputs the processed image information, and is, for example, an image transmission unit of an image recording device or a facsimile device.

The operation method switching means 2a in the multi-value conversion section 2
Further, a signal for instructing a scaling ratio is input to the scaling unit 3.

With the above configuration, assuming that the image scaling device of this embodiment is activated, a signal indicating a scaling factor r is input from outside. The magnification r is set by an operator or, in the case of a facsimile machine, determined by communication with the destination machine.

The binary image input unit 1 sequentially inputs binary image information from an external device (not shown). The multi-level conversion unit 2 inputs the image information one pixel at a time and extracts eight surrounding pixels centered on the target pixel. Here, as shown in FIG. 2, each value of the surrounding pixels is A to H, the value of the target pixel is X, and the value obtained by multi-valued the target pixel is X ′. In the present embodiment, the number of gradations for the multi-value quantization is eight gradations of the density “0” to “7”.

As shown in FIG. 3, when the scaling factor r is set to 95% or more and less than 105%, the multivalue conversion section 2 sets the value X of the target pixel to a multivalued value X 'according to the following equation. Become

X ′ = 7 · X (1) Further, the magnification r is 80% or more and less than 95% or 10%.
When it is 5% or more and less than 120%, multi-value is obtained by the following equation.

X ′ = 3 · X + B + D + E + G (2) Further, when the magnification r is less than 80% or more than 120%, multi-value is obtained by the following equation.

X ′ = X + B + D + E + G + (A + C + F + H) / 2 (3) The multi-value conversion unit 2 sequentially outputs image information multi-valued into eight gradations by such calculation.

The scaling unit 3 enlarges or reduces the image size by increasing or decreasing the number of pixels in the horizontal / vertical direction in the multi-valued image information by a predetermined method. For example, in the case of enlargement, the number of pixels is increased by a linear interpolation method according to the zoom ratio.

The binarizing section 4 binarizes the multi-valued image information subjected to the scaling process by a known error diffusion method. The image output unit 5 outputs the binarized image information by a predetermined method.

As described above, in the present embodiment, when multi-valued binary image information is used, if the zoom ratio is small, surrounding pixels are not included in the parameters of the arithmetic expression, and the target pixel is set to a predetermined value. The gray level is converted to white or black. This prevents each pixel of the character image from changing to the intermediate density and deteriorating the image quality.

The larger the zoom ratio, the larger the number of surrounding pixels used as parameters of the arithmetic expression. In the case of a halftone image, when the zoom ratio is large, a sharp change in density between pixels is conspicuous, and the image quality is likely to deteriorate, but by increasing the number of surrounding pixels as a parameter as described above, Since it is smoothed according to the density of surrounding pixels,
The deterioration of the image quality is reduced.

In the above-described embodiment, the zoom ratio is divided into three ranks. However, it is obvious that the arithmetic expression may be set for each of more ranks.

FIG. 4 shows a second embodiment of the present invention.

In the figure, the same reference numerals as those in FIG. 1 denote the same parts, and in this embodiment, an edge portion judging section 6 is newly provided. The edge portion determination section 6 determines a pixel at an edge portion which is an outline of an image.

With this configuration, when the image scaling device of the present embodiment is started, the binary image input unit 1 sequentially inputs the binary image information, and the edge determination unit 6 determines that each of the input pixels has an edge portion. It is determined whether the pixel is a pixel of For this determination, for example, in the 3 × 3 pixel range shown in FIG. 2, each pixel pattern when the target pixel corresponds to an edge portion is stored in advance, and the stored pattern is compared with the actual pixel pattern. It is done by doing.

The multi-level conversion section 2 multi-levels the binary image information by the respective calculation methods according to the above determination results. That is, as shown in FIG. 5, when the pixel to be processed is a pixel in the edge portion, the value X of the target pixel is multi-valued to X ′ by the above equation (1). If the pixel is not a pixel at the edge portion, the pixel is multi-valued by the equation (3).

The scaling unit 3 scales the multi-valued image information at a predetermined scaling ratio, the binarizing unit 4 binarizes the image information, and the image output unit 5 Outputs the valued image information.

As described above, in this embodiment, when the pixel to be processed is an edge portion, the pixel of interest is multi-valued without surrounding pixels being included in the parameters of the arithmetic expression. Is prevented from changing to the intermediate density and deteriorating the image quality.

Further, in the non-edge portion, the density of the target pixel is calculated using many peripheral pixels as parameters of the arithmetic expression, so that each pixel density is smoothed in accordance with the peripheral pixels, thereby deteriorating the image quality. Is reduced.

FIG. 6 shows a third embodiment of the present invention.

4 is different from FIG. 4 in that a direction-specific edge determination unit 7 is provided in place of the edge determination unit 6. The direction-specific edge portion determination section 7 determines the pixels of the edge portion, which is the outline of the image, for each edge direction.

With this configuration, when the image scaling device of this embodiment is started, the binary image input unit 1 sequentially inputs binary image information. The direction-specific edge portion determination section 7 determines whether or not each input pixel is an edge portion of an image for each edge direction.

That is, the edge patterns P1 to P shown in FIG.
4 illustrates an example of a pixel pattern determined by the direction-specific edge portion determination unit 7. In each drawing, blank pixels indicate “white”, cross-hatched pixels indicate “black”, and broken-line hatched pixels indicate undefined pixels which may be either “white” or “black”. The edge pattern P1 is
When the target pixel is located at the left end of the white area of the image,
The edge pattern P2 is a case where the target pixel is located at the upper end of the white area. The edge pattern P3 is a case where the target pixel is located at the left end of the black area. Further, the edge pattern P4 is a case where the target pixel is located at the left end of the white region inclined at 45 degrees.

The edge determining unit 7 for each direction determines various edge patterns in which the target pixel is located in each direction of the white or black region of the image as described above. When one edge pattern is determined, the result of the determination is output. The multi-value conversion unit 2 multi-values the pixel of interest by a calculation method according to the determination result.

For example, when the input image information is determined to be the edge pattern P1 or P3, the value X of the target pixel is set to X ′.
Multi-valued.

X '= 5.X + C + H (4) When it is determined that the edge pattern is P2, multi-value is obtained by the following equation. X ′ = 5 · X + F + H (5) Further, when it is determined that the edge pattern is P4, the value is multi-valued by the following equation. X ′ = 5 · X + E + G (6) As described above, the direction-specific edge portion determination section 7 performs the calculation using the indeterminate pixel which may be either “white” or “black” and the pixel of interest at the time of the edge pattern determination. Thus, the target pixel is multi-valued. On the other hand, when it is not determined that any of the edge patterns is used, the multi-value is obtained by the calculation method for non-edge pixels described with reference to FIG. The multivalued image information is processed in the same manner as in the above-described embodiment.

As described above, in the present embodiment, when a pixel to be processed is an edge portion, pixels around the image area where the pixel is located are set as parameters of an arithmetic expression. Is also corrected to a density corresponding to the image area, and the deterioration of the image quality is further reduced as compared with the embodiment of FIG.

FIG. 8 shows a facsimile apparatus according to a fourth embodiment of the present invention.

In the figure, the same symbols as those in FIG. 6 indicate the same parts. In FIG. 1, a photoelectric conversion unit 8 is a scanner device that reads an original image and obtains an analog image signal. A /
The D converter 9 converts the analog image signal into a digital signal. The binarizing unit 10 binarizes the image signal by simple binarization at a fixed threshold level or by halftone processing such as an error diffusion method. The memory 11 stores the binary image information. The transmitting unit 12 transmits an image to a communication partner in a known procedure by making and receiving a call.

With this configuration, the facsimile apparatus according to the present embodiment first performs an image information accumulation operation during transmission processing. In this case, the operator operates the operation unit (not shown) to
Set the value conversion method. The set binarization method is instructed to the binarization unit 10 and the multi-value unit 2.

When the facsimile apparatus starts reading an original image, the analog image signal output from the photoelectric conversion unit 8 is converted into a digital signal of a constant gradation by an A / D converter 9, Valued. The binarized image information is temporarily stored in the memory 11.

Thereafter, the transmitting unit 12 transmits the call to the other party and executes a known transmission control procedure. At this time, the recording paper size of the destination apparatus is determined, and the magnification is determined as necessary.

After that, the binary image information is sequentially read out from the memory 11 and is multi-valued by the multi-value converting section 2. in this case,
As shown in FIG. 9, when the binary image information is obtained by simple binarization, the multi-level quantization unit 2 multiplies the value X of the target pixel by X ′ according to the equation (1). Value. If the data is obtained by a pseudo halftone process such as an error diffusion method, it is converted into multi-valued data by the above equation (3).

The multivalued image information scaling unit 3
To zoom. Then, after being binarized again by the binarization unit 4, the data is transmitted to the destination by the transmission unit 12.

By the way, when reading the original of the character image, the operator usually sets the simple binarization. Also,
When reading a halftone image document such as a photograph, the halftone processing is set.

In the present embodiment, image information obtained by simple binarization is binarized without using surrounding pixels as parameters, whereas image information obtained by halftone processing is binarized using surrounding pixels as parameters. ing. As a result, similarly to the above-described embodiments, deterioration in image quality of both the character image and the halftone image is reduced.

FIG. 11 shows an image scaling device according to a fifth embodiment of the present invention.

6, the same reference numerals as those in FIG. 6 denote the same parts. The difference from FIG. 6 is that the halftone part determining part 13 and the character part determining part are replaced with the edge part determining part 7 for each direction. 1
4 is provided. The halftone part determination unit 13 determines a pixel in a halftone region in an image in which a character image and a halftone image are mixed in one page. Is determined.

With this configuration, when the image scaling device of this embodiment is started, the binary image input unit 1 sequentially inputs binary image information. The halftone portion determination section 13 determines whether each of the input pixels is a pixel in the intermediate density area, and outputs the determination result. Further, the character part determination unit 14 determines whether the pixel is a pixel in the character image area, and outputs the determination result. In determining these image areas, for example, in the 3 × 3 pixel range shown in FIG. 2, each pixel pattern in various image areas is stored in advance, and the stored pattern is compared with an actual pixel pattern. It does by doing.

By the way, in this case, the halftone portion judging section 13
May be inconsistent with the determination result of the character part determination unit 14. One is a case where the halftone portion determination section 13 determines that the pixel of interest is a halftone area and the character portion determination section 14 determines that the pixel of interest is a character image area. The other is a case where neither of these areas is determined. In the present embodiment, when one of the halftone portion determination portion 13 and the character portion determination portion 14 outputs a predetermined determination result, the determination result is made valid. Then, if the determination result is output from both, or if the determination result is not output from either, the area is determined to be unknown.

As shown in FIG. 11, when it is determined that the area is a character image area, the multi-value converting section 2 multi-values the value X of the pixel of interest to X 'according to the above equation (1). If it is determined that the area is a halftone image area, the image is multi-valued by the equation (3). Further, when the region is unknown, the value is multi-valued by the equation (2).

As described above, in this embodiment, it is determined whether each pixel is a character image area or a halftone image area, and each pixel is multi-valued by an operation suitable for each pixel. I have. As a result, the image to be scaled is 1
Even when a character image and a halftone image are mixed in a page, deterioration of the image quality of each image can be reduced.

In the above equation (1), the parameter terms of the surrounding pixels are completely eliminated. However, some parameter terms of the surrounding pixels may be set. In this case, the image quality of the halftone image is reduced. Of course, it can be improved.

In each of the embodiments described above, an example has been described in which binary image information is converted into multi-valued image information of eight gradations. However, it is needless to say that the number of gradations can be arbitrarily set.

[0069]

As described above, according to the first aspect of the present invention, when the zoom ratio is small, the number of surrounding pixels used as parameters in the multi-value calculation is reduced, so that the image quality of the character image is reduced. When the deterioration is reduced and the zoom ratio is large, the number is increased, so that the image quality deterioration of the halftone image is reduced.

In the second invention, the number of surrounding pixels as the above parameters is reduced in the case of edge pixels, and is increased in the case of non-edge pixels, so that image quality deterioration of each image portion is reduced.

In the third aspect of the present invention, the positions of the surrounding pixels as the above parameters are further changed according to the edge direction.
Since the pixel density is corrected according to the image state, the image quality is further improved.

According to the fourth aspect of the present invention, in the case of a character image that is usually simply binarized, the number of surrounding pixels used as the above parameters is reduced, so that the image quality deterioration of the character image is reduced and the pseudo halftone processing is usually performed In the case of a halftone image, since the number thereof is increased, deterioration of the image quality of the halftone image is also reduced.

According to the fifth aspect, each pixel is determined to be a character image area or a halftone image area, and multi-valued by an appropriate calculation method. Even when the toned image is mixed in one page, the image quality deterioration of each image is reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram of an image scaling device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of a pixel of interest and a pixel position of a peripheral pixel used as a parameter of an arithmetic expression;

FIG. 3 is an explanatory diagram showing correspondence between various scaling factors and calculation methods.

FIG. 4 is a block diagram of an image scaling device according to a second embodiment of the present invention.

FIG. 5 is an explanatory diagram showing the correspondence between pixel determination results and calculation methods.

FIG. 6 is a block diagram of an image scaling device according to a third embodiment of the present invention.

FIG. 7 is an explanatory diagram showing correspondence between various edge patterns and calculation methods.

FIG. 8 is a block diagram of a facsimile apparatus according to a fourth embodiment of the present invention.

FIG. 9 is an explanatory diagram showing a correspondence between each binarization method and a calculation method.

FIG. 10 is a block diagram of an image scaling device according to a fifth embodiment of the present invention.

FIG. 11 is an explanatory diagram showing correspondence between various image areas and calculation methods.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Binary image input part 2 Multi-value part 2a Calculation method switching means 3 Magnification part 4,10 Binarization part 5 Image output part 6 Edge part determination part 7 Edge part determination part for every direction 8 Photoelectric conversion part 9 A / D conversion 11 memory 12 transmission unit 13 halftone part determination unit 14 character part determination unit

Claims (2)

(57) [Claims] 1. An image scaling device which once binarizes binary image information, scales the image by increasing / decreasing the number of pixels, and binarizes the image again, determines whether each pixel of the binary image information is an edge portion of the image. Edge determining means for determining whether or not the pixel is a non-edge portion; multi-value converting means for converting each pixel to multi-value by performing an operation using the pixel of interest and surrounding pixels as parameters by focusing on each of the pixels;
An operation method switching means for reducing the number of surrounding pixels used as the parameter when the pixel to be multi-valued is the edge portion and increasing the number when the pixel to be multi-valued is the non-edge portion. An image scaling device, comprising: 2. An edge direction determining means for further determining an edge direction when each of the pixels is an edge portion, and determining the parameter based on the determined edge direction when the pixel to be multivalued is the edge portion. 2. The image scaling device according to claim 1, further comprising: a reference pixel position changing unit that changes a position of a surrounding pixel to be changed.