JP2002252756A - Image processing unit, recording medium and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology that can separate a dot image area from a character area in a digital image. SOLUTION: The image processing unit obtains 4 vectors V1-V4 respectively tying a prescribed mesh point D0 with each of 4 adjacent dots D1-D4 on the basis of dot information. Then 4 neighboring pixels obtained from a target pixel shifted by a vector quantity corresponding to each of the 4 vectors V1-V4 are selected, the result of sum summing absolute values of each difference between the target pixel and each of the selected four neighboring pixels is obtained, and whether the target pixel is a pixel in the dot image area or a pixel in the character area is discriminated on the basis of the result of sum. Employing the result of the discrimination processing can apply different processing between the dot image area and the character area.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an image processing technique for processing a digital image, and more particularly to an image processing technique for processing a digital image having a halftone image area and a character area.

[0002]

2. Related Art There is a case where an input image is further output using an image reading device such as a scanner.

In such a case, when the resolution of an input digital image is relatively large and the resolution of an output device is relatively small (for example, a halftone image having a line number of 133 lines or more is reduced to a smaller size). Number of lines (for example, 106
When the image is output as an image having a line), the gradation may be lost or the tone jump may occur.

In order to avoid such a situation, a process of converting the number of halftone dots (hereinafter also referred to as a "screen ruling conversion process") after subjecting the input image to a blurring process is performed. Conceivable. If only the process of converting the halftone frequency is performed without performing the blurring process, moire occurs because the screen patterns having the new and old halftone frequencies produce a periodic interference state. Therefore, in order to avoid this, it is necessary to perform a blurring process in the screen ruling changing process.

[0005]

However, when a digital image having both a halftone dot image area and a character area is subjected to the screen ruling conversion processing including the above-mentioned blur processing, the characters in the character area become unclear. A situation that blurs occurs. FIG. 28 is a diagram illustrating an example of a processing target image having a halftone image region SR and a character region CR.
FIG. 29 is an enlarged view of the upper left character portion of FIG. 28. FIG. 30 is an image obtained by performing the above-described screen ruling conversion processing on the processing target image. FIG.
Shows a situation in which characters are blurred as a result of the screen ruling conversion process.

This phenomenon is a problem caused by performing the same processing on both areas (halftone image area and character area) having different properties. That is, it is required to separate a halftone image area and a character area in a digital image.

Separating the halftone image area and the character area as described above is not only performed when performing the screen ruling conversion processing, but also when performing various processing in digital image processing having both areas. This is a required technology.

In view of the above problems, it is an object of the present invention to provide a technique capable of separating a halftone image area and a character area in a digital image.

[0009]

In order to achieve the above object, an image processing apparatus according to claim 1 is an image processing apparatus for processing a digital image, wherein halftone dot information obtaining means for obtaining halftone dot information. Means for obtaining four vectors connecting a predetermined halftone dot and each of four adjacent halftone dots based on the halftone dot information, and moving the target pixel by a vector amount corresponding to each of the four vectors from the pixel of interest. Determining whether the pixel of interest is a pixel in a halftone dot image area or a pixel in a character area using the means for selecting the four neighboring pixels and the pixel of interest and the four neighboring pixels. Determining means for performing processing.

According to a second aspect of the present invention, in the image processing apparatus of the first aspect, the determination unit adds an absolute value of a difference between the target pixel and each of the four neighboring pixels. An addition result is obtained, and the determination process is performed based on a magnitude relationship between the addition result and a predetermined value.

According to a third aspect of the present invention, in the image processing apparatus according to the first or second aspect, the image processing apparatus further includes means for performing a smoothing process on the digital image. The above-described determination processing is performed on the image that has been subjected to the smoothing processing.

According to a fourth aspect of the present invention, in the image processing apparatus according to any one of the first to third aspects, the determining means includes a median filter for an area including the target pixel and its peripheral pixels. The processing is used to correct the determination result in the determination processing.

According to a fifth aspect of the present invention, in the image processing apparatus according to any one of the first to fourth aspects, the determining means expands an area determined as a character area. Is used to correct the determination result.

According to a sixth aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, the halftone information obtaining means is configured to input the halftone information. An input unit is provided, and information input using the dot information input unit is obtained as the dot information.

According to a seventh aspect of the present invention, in the image processing apparatus according to any one of the first to fifth aspects, the halftone dot information obtaining means extracts the halftone dot information from the digital image. And extracting information extracted by the dot information extracting means as the dot information.

An image processing apparatus according to an eighth aspect of the present invention is the image processing apparatus according to any one of the first to seventh aspects, further comprising an area designating unit for designating an area in which the determination process is to be performed. It is characterized by.

According to a ninth aspect of the present invention, in the image processing apparatus of the eighth aspect, the halftone dot information obtaining means obtains the halftone dot information for each area specified by the area specifying means. It is characterized by doing.

According to a tenth aspect of the present invention, in the image processing apparatus according to any one of the first to seventh aspects, the halftone dot information extracting means converts the halftone dot information into predetermined block units. It is characterized by extracting.

According to an eleventh aspect of the present invention, in the image processing device of the first aspect, a pixel determined to be a pixel in the halftone dot image area is blurred. A result of performing the processing and the re-dotting process, and a result of not performing the blurring process and the re-dotting process for a pixel determined to be a pixel in the character area. .

A recording medium according to a twelfth aspect is a computer-readable recording medium that records a program for causing a computer to function as the image processing apparatus according to any one of the first to eleventh aspects. It is characterized by.

A program according to a thirteenth aspect is a program for causing a computer to function as the image processing apparatus according to any one of the first to eleventh aspects.

[0022]

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

<A. Configuration> FIG. 1 is a conceptual diagram showing a hardware configuration of an image processing apparatus 1 according to an embodiment of the present invention. As shown in FIG. 1, the image processing apparatus 1 has a CPU
2, a storage unit 3 including a semiconductor memory and a hard disk, a media drive 4 for reading information from various recording media, a display unit 5 including a monitor, etc., an input unit 6 including a keyboard and a mouse, and communication with other devices. It is configured by a computer system (hereinafter, also simply referred to as “computer”) including a communication unit 7 that performs the communication. C
The PU 2 is connected to the storage unit 3, the media drive 4, the display unit 5, the input unit 6, the communication unit 7, and the like via the bus line BL and the input / output interface IF. In addition, the media drive 4 includes a CD-ROM, a DVD (Di-
The information recorded therein is read from a portable recording medium 9 such as a digital versatile disk or a flexible disk.

This computer reads a software program (hereinafter, also simply referred to as a “program”) recorded on a recording medium 9 and stores the program in a CPU.
When the image processing apparatus 1 is executed by using the image processing apparatus 2 or the like, it functions as the image processing apparatus 1 that realizes various operations described below. Note that the program having each function is not limited to being supplied (or distributed) via the recording medium 9, and may be transmitted to the computer via a network (communication line) such as a LAN or the Internet and the communication unit 7. May be supplied (or distributed).

As described above, the image processing apparatus 1 is constructed by software in a computer.

FIG. 2 is a functional block diagram of the image processing apparatus 1. As shown in FIG. 2, the image processing apparatus 1 performs image processing on a digital image input from an image reading device 70 such as a scanner, and outputs the processed image to an image output device 80 such as a print output device. Is output.

The image processing apparatus 1 includes an area designating section 12 for designating an area to be processed, a halftone information input section 20 for inputting halftone information, and a predetermined halftone area to four adjacent halftone areas. Four-direction vector calculation unit 4 for obtaining four heading vectors
0, a character halftone separation processing section 50, and a screen ruling conversion processing section 60. Further, the character halftone dot separation processing unit 50 has a smoothing processing unit 51, a difference value addition processing unit 52, a binarization processing unit 53, a median filter processing unit 54, and a fattening processing unit 55. The line number conversion processing unit 60 includes a blur processing unit 61 and a synthesis processing unit 6.
Two. Each of these units is constructed in software on a computer by executing a predetermined program.

The image processing apparatus 1 determines whether each pixel of interest is a pixel in a halftone image area or a pixel in a character area by using a character halftone dot separation processing section 50 or the like. It is also possible to perform the screen ruling conversion processing using the screen ruling conversion processing unit 60 or the like using the determination result. The detailed operation of each processing unit will be described later.

<B. Operation><SchematicOperation> FIG. 3 is a flowchart showing the operation of the screen ruling conversion process in the image processing apparatus 1. The screen ruling conversion processing will be described with reference to FIG.

First, in step SP10, the image processing device 1 acquires a digital image from the image reading device 70 or the like. As shown in FIG. 4, the digital image includes a halftone image area SR and a character area CR. This digital image is acquired, for example, as a binary image. Here, since it is preferable to increase the number of gradations in the following processing, at this point, the gradation values 0 and 1 of each pixel of the binary image are respectively set to 0 and 2 for convenience.
55 in advance. This binary image is actually converted into a multi-valued image having 256 gradation values by the processes described below.

Further, the operator (operator) of the image processing apparatus 1 can execute steps SP30 and SP40 as necessary.
An area to be processed (described later) may be manually specified. In other words, the processing target area can be designated by trimming. By limiting the processing target area in this way, it is possible to increase the processing speed.

Specifically, as shown in FIG. 4, in the digital image P displayed on the screen, an area DR including the halftone image area SR can be designated as a processing target area. Further, the designated area DR may include not only the halftone image area SR but also a character area CR, or may include a plurality of halftone image areas SR having the same halftone information. Even in such a case, the subsequent processing can be performed well. However, when the plurality of dot image regions SR have different dot information from each other, it is preferable to designate each dot image region SR as a separate region.

This designation processing is performed using the area designation section 12. More specifically, the designated area DR can be designated as a rectangular area having an arbitrary size by a mouse operation or the like.

Next, in step SP20, dot information about the dot image region SR included in the processing target region is obtained. The halftone information is information including “halftone frequency” and “halftone angle” of the screen pattern used at the time of generating the halftone image area SR. The screen frequency is the screen frequency per inch of the screen pattern, and the screen angle is the angle between the screen pattern and the image.

Here, it is assumed that the operator inputs the dot information. Specifically, the operator uses the halftone dot information input unit 20 (FIG. 2) to display the halftone dot information (more specifically, the halftone frequency and the halftone angle) on the halftone image region SR. Information). Thereby, the image processing apparatus 1
Obtains dot information.

When a plurality of areas are designated as processing target areas DR as shown in FIG. 4, it is preferable to obtain dot information for each processing target area (designated area). This makes it possible to accurately extract halftone information even when halftone information such as halftone frequency differs for each designated area.

By using the halftone dot information extracting section 30, the halftone dot information of the halftone dot image region SR is converted to the halftone dot image region SR.
It is also possible to automatically extract based on the information, but the operation and the like will be described later.

In this step SP20,
Based on the dot information, a process of obtaining four different vectors connecting a predetermined dot and an adjacent dot is further performed. This process is performed by the four-direction vector calculation unit 40.

FIG. 5 is a diagram for explaining the principle. For example, as shown in FIG. 5, when the halftone dot angle の of the halftone dot image region SR is 45 degrees, the halftone dots D1 to D4 adjacent to the predetermined halftone dot D0 are separated from the predetermined halftone dot D0. As you can see, they are located in four directions: lower right, lower left, upper left, and upper right. Further, when the halftone frequency (lines per inch) of the halftone image area SR is 150 lines and the resolution (pixels per inch) of the processing target image is 600 (dpi). The distance corresponding to 600/150 = 4 pixels corresponds to the distance D from the halftone dot D0 to the adjacent halftone dot. Therefore, in this case, the horizontal distance dX from the halftone dot D0 to the adjacent halftone dot D1 is calculated by adding the value D (= 4) to cosΘ
(Cos 45 degrees) (ie, about 2.8), and the vertical distance dY is obtained by adding sin 値 to the value D (= 4).
(Sin 45 degrees) (that is, about 2.8). Therefore, the halftone dot D0 is the starting point and the adjacent halftone dot D1
Is obtained as a vector V1 = (dX, dY).

The other three adjacent halftone dots D2
Similar distances dX and dY are obtained for all of D4. That is, a vector V2 starting from the halftone dot D0 and ending at the adjacent halftone dot D2 is obtained, and further, a vector V3 starting from the halftone dot D0 and ending at the adjacent halftone dot D3,
A vector V4 starting from the halftone dot D0 and ending at the adjacent halftone dot D4 is determined. Thus, four different vectors V1 to V4 connecting the predetermined halftone dot D0 and each of the adjacent halftone dots D1 to D4 are obtained. These four vectors V1 to V1
V4 is used in step SP30 described later.

Further, in step SP30 (FIG. 3), processing for generating a mask image for separating the character area CR (FIG. 4) and the halftone image area SR for the specified area DR is performed. In SP40,
A screen ruling conversion process is performed using the mask image. In the following, referring to FIGS. 6 and 7,
The operation in step SP30 and step SP40 will be described in detail. FIG. 6 is a flowchart showing the detailed operation of step SP30, and FIG. 7 is a flowchart showing the detailed operation of step SP40.

<Mask Image Generation Operation> Next, a mask image generation operation (step SP30) will be described.

FIGS. 8 to 13 are diagrams for explaining this operation. FIG. 8 is an enlarged view showing a partial region of the processing target image. FIG. 8A shows a partial region in the halftone dot image region SR, and FIG. 8B shows a partial region in the character region CR. Of FIG. Each unit area of the square in FIG.
One pixel is represented, and the pixel indicated by oblique lines represents a black pixel in the processing target image. FIG. 9 is a diagram showing the pixel values of the respective regions. FIG. 9 (a) corresponds to FIG. 8 (a), and FIG. 9 (b) corresponds to FIG. 8 (b). are doing. Further, FIGS. 10 to 13 are diagrams showing the pixel values of each area according to the progress of each of the following steps, and FIG. 10 (a) shows the processing result for FIG. 9 (a). (B), FIG. 11, and FIG. 13 show processing results for FIG. 9 (b).

Hereinafter, the processing of each step will be described with reference to FIG.

First, as shown in FIG. 6, step SP3
In 1, the smoothing processing section 51 performs a smoothing process on the digital image. This smoothing process can be realized, for example, by using a smoothing filter as shown in FIG. This smoothing filter is a filter having a size of 3 × 3 pixels. The weighting coefficient of the center is 4, and the weighting coefficients of the surrounding 8 pixels are 1 and 2. By using this smoothing filter, it is possible to calculate the weighted average value of the pixel value of the target pixel (center pixel) in consideration of the pixel values of the peripheral pixels.

Here, the processing of step SP31 (smoothing processing) is not necessarily performed, but in order to suppress noise at the time of determination, the following steps SP31 are required.
It is preferable to perform the smoothing process of step SP31 before 33 to SP38. In other words, the following steps SP33 to SP33 are performed on the image subjected to the smoothing process.
It is preferable to perform a determination process such as SP37.

When the image to be processed is a binary image, by performing this smoothing process, the process of multi-valued pixel values of each pixel can be performed at the same time. For example, when converting this binary image into a multi-valued image having 256 levels of gradation values, as described above, each of the gradation values 0 and 1 of each pixel of the binary image is conveniently used. This smoothing process is performed after conversion to 0,255 in advance. Each pixel of the image after the smoothing process is 0, 2
It also has a gradation value other than 55, and is subjected to weighted averaging processing in consideration of the pixel values of the peripheral pixels, and then multi-valued.

Next, in steps SP33 to SP37 (FIG. 6), a determination is made as to whether the pixel of interest is a pixel in the halftone dot image area or a pixel in the character area. This determination process is roughly performed by the following five processes, namely, (1) a process of selecting four neighboring pixels moved by a vector amount corresponding to each of the four vectors from the target pixel (step SP33), and ( 2) a process of obtaining the absolute value of the difference between the pixel of interest and each of the four neighboring pixels (step SP34); and (3) a process of obtaining an addition result obtained by adding these four differential absolute values (step SP34).
35), (4) a determination process based on the magnitude relationship between the addition result and a predetermined threshold (step SP36), and (5) a process of applying a median filter to the determination result (step SP37). Contains. Hereinafter, each process will be described.

First, in step SP33,
(1) A process of selecting four neighboring pixels moved by a vector amount corresponding to each of the four vectors from the target pixel is performed. In FIG. 8A, the pixel of interest PE
, Four neighboring pixels E1 to E4 existing at positions shifted by the vectors V1 to V4, respectively, are shown. Here, the vectors obtained in step SP20 described above are used as the vectors V1 to V4. Therefore, when it is assumed that the pixel of interest PE exists at the position of the predetermined halftone dot D0, the four pixels existing at the respective positions of the adjacent halftone dots D1 to D4 become four neighboring pixels E1.
To E4.

Next, in step SP34,
(2) A process of calculating the absolute value of the difference between the target pixel PE and each of the four neighboring pixels E1 to E4 is performed. Specifically, the absolute value A1 of the difference between the pixel value EV of the pixel of interest PE and the pixel value EV1 of the neighboring pixel E1 is obtained, and the difference between the pixel value EV of the pixel of interest PE and the pixel value EV2 of the neighboring pixel E2 is obtained. Find the absolute value of. Similarly, the pixel value E of the pixel of interest PE
The absolute value of the difference between V and the pixel value EV3 of the neighboring pixel E3, and the absolute value of the difference between the pixel value EV of the target pixel PE and the pixel value EV4 of the neighboring pixel E4 are obtained.

Then, in step SP35, a process of adding the four absolute differences obtained in step SP34 and obtaining an addition result RV is performed. Here, a value obtained by dividing the addition result by 4 (that is, an average value) is obtained as a new addition result RV, and the new addition result RV is compared with the threshold value TH. FIG. 10 is a diagram showing the addition result RV. This FIG.
9 illustrates a processing result for FIG. 9, and illustrates a case where step SP31 is not performed for simplification. As described above, FIG. 10A shows the processing result of FIG. 9A (that is, the processing result for the halftone image region SR), and FIG. 10B shows the processing result of FIG. 9B. The processing result (that is, the processing result for the character area CR) is shown. Note that the processing in steps SP33 to SP35 is performed by the difference value addition processing unit 52.

In the next step SP36, this addition result RV is compared with a predetermined threshold value TH, and a binarization process is performed based on the magnitude relation. This binarization processing is performed by the binarization processing unit 53. Specifically, the addition result R
When V is greater than or equal to the threshold value TH, the pixel value of the target pixel PE is set to “1”. When the addition result RV is smaller than the threshold value TH, the pixel value of the target pixel PE is set to “0”. Pixel values can be binarized. FIG.
FIG. 9 is a diagram showing a processing result of step SP36, and FIG.
The processing result regarding (b) is shown. As shown in FIG. 11, assuming that the threshold value TH is set to 130, only the pixel value of the pixel having “255” in FIG. 10 is converted to “1”, and the pixel values of the other pixels are set to “0”. Is converted to

It should be noted that the predetermined threshold value TH is determined by the type of halftoning method used when the halftone image region SR is formed (for example,
A different value may be determined according to the error diffusion method or the simple binarization method). Specifically, if the type of the halftoning method is known, the threshold value TH according to the type can be set. For example, the threshold value TH for the halftone dot image region SR by the error diffusion method can be set to 80, and the threshold value TH for the halftone dot image region SR by the simple binarization method can be set to 100. By setting the threshold value TH as a value corresponding to the type of halftoning method, the binarization process can be performed more appropriately. Note that, as the information on the type of the halftoning method, information input by the operator using the halftone information input unit 20 as halftone information can be used.

Here, a value obtained by dividing the addition result by 4 (that is, an average value) is obtained as a new addition result RV, and the new addition result RV is compared with the threshold value TH. The addition result RV of the absolute value is directly used as the threshold value TH.
It may be compared with. The threshold value TH in this case may be determined as having a value four times the threshold value TH when using the averaged addition result.

As described above, the image to be processed is binarized based on the magnitude relationship between the addition result RV and the threshold value TH. Then, as a result of the binarization processing, the portion having the value “0” is determined to be highly likely to be a pixel in the halftone image region, and the portion having the value “1” is determined to be a pixel in the character region. Is determined to be highly likely.

This determination operation is based on the following principle.

As shown in the upper part of FIG. 15, in the halftone dot image area SR, when halftone dots expressing the same level of halftone value are continuous, the target pixel PE and its four neighboring pixels E
1 to E4 have substantially equal pixel values. Therefore, as shown in the lower part of FIG. 15, the absolute value of the difference becomes small (ideally, the absolute value of the difference = 0), and the addition result RV becomes a small value. Therefore, it is highly possible that the portion having the value “0” is a pixel in the halftone dot image area. In addition,
The lower part of FIG. 15 shows a state corresponding to the result of FIG.

On the other hand, as shown in FIG.
Since the pixel value distribution is not uniform, the pixel value of the pixel of interest PE and its four neighboring pixels E1 to E4 often differ greatly, so that the absolute value of the difference becomes large and the addition result RV becomes This is a large value. Therefore, the portion having the value “1” is likely to be a pixel in the character area. The lower part of FIG. 16 shows a state corresponding to the result of FIG. Also, in FIG.
The white square area indicates an area having a pixel value between the black area and the white area.

The lower part of FIG. 16 and FIG.
The results obtained by applying the processing of steps SP33 to SP36 when a thin line having a width of about one pixel or several pixels (two pixels in FIG. 10B) are continuous in one direction. As shown in these figures, the absolute difference value remains as a relatively large value.

Based on such a principle, a pixel having a value “0” is determined to be a pixel in the halftone image area, and a portion having a value “1” is determined to be a pixel in the character area. Can be determined.

Here, since the absolute values of the differences from all the neighboring pixels in the four directions toward the adjacent halftone dot are considered,
Misjudgment due to direction dependency can be prevented. In other words, an accurate determination operation can be realized as compared with a case where only two of the four directions are considered.

Further, in step SP37, a median filter process is performed on the above determination result. This median filter is a filter having a predetermined size (for example, 3 × 3 pixels), and has a predetermined number (here, nine) including a target pixel and its peripheral pixels.
Is a filter that uses the central value of the pixels of the pixel of interest as the pixel value of the pixel of interest. For example, when a median filter having a size of 3 × 3 pixels is used, the fifth value of the nine pixels including the target pixel and its surrounding pixels can be set as the pixel value of the target pixel. By applying this median filter, a pixel having a pixel value equal to or greater than the threshold value TH exists alone without having a pixel with a high gradation value in the periphery (in short, when the pixel exists as an “enclave”). ), The pixel value of the pixel can be corrected (more specifically, corrected to zero). As described above, by correcting the determination result using the median filter processing, it is possible to suppress noise due to the presence of the “enclave”. When the median filter is applied to the processing result of FIG. 11, the result is the same as that of FIG.

FIG. 12 is a diagram showing another example of the median filter processing in step SP37.
(A) shows the processing result up to step SP36, and FIG. 12 (b) shows the result of applying the median filter processing to the processing result of FIG. 12 (a). This figure 1
Referring to FIG. 2B, “enclave” existing in FIG. 12A (in this case, a pixel having only three or less pixels having the same pixel value around it) is shown. You can see that it has disappeared. As described above, by performing the binarization process of step SP36 after the process of step SP37, it is possible to eliminate the above-mentioned enclave. This median filter processing is performed by the median filter processing unit 54.

Then, in the next step SP38, a fattening process (specifically, a process for enlarging a portion determined as a character area) is performed on the binary image that is the result of this determination. For example, a maximum value selection filter having a size of 8 × 8 pixels, 16 × 16 pixels, or the like can be used. The maximum value selection filter is a filter that sets the maximum value of the pixel values in a predetermined area including the target pixel to the pixel value of the target pixel. As a result, it is possible to enlarge the area of the pixel value “1” (to make it thicker). FIG. 13 corresponds to FIG. 11 (or FIG. 12 (b)).
This shows the result of performing the fattening process in step SP38 on the result of the above using a maximum value selection filter having a size of 8 × 8 pixels. As shown in FIG. 13, the pixel value is “1” not only for the region corresponding to the thin line surrounded by the thick line in the center but also for the pixels in the peripheral region. That is, the determination result is corrected so that the area determined as the character area CR is enlarged. Thus, if the fattening process of step SP38 is used,
In separating the character region CR from the halftone image region SR, the determination as the character region CR is prioritized in the boundary region (that is, the boundary region between the character region CR and the halftone image region SR). Therefore, in the screen ruling changing process described later, it is possible to convert the ruling of the halftone dot image region SR while keeping the characters clear. This process is performed by the fattening processing unit 55.

By the processing as described above, step SP
An image after the processing up to 38 is generated as a mask image. By using this mask image, it is possible to determine whether each pixel of interest is a pixel in the halftone dot image area or a pixel in the character area. That is, it is possible to separate the halftone dot image area and the character area in the digital image.

Next, the screen ruling conversion processing in step SP40 will be described with reference to FIG. For example, a halftone dot image region SR having a halftone frequency of 150 lines can be re-converted into a halftone image having another halftone frequency (here, 106 lines).

For this purpose, first, in step SP41, an averaging process (blur process) is performed. Specifically, 6
A blur filter (averaging filter) that performs a simple averaging process with a size of pixels × 6 pixels can be used.

In the next step SP42, a halftone process (re-RIP) is performed again using a screen pattern having a different number of lines from the halftone image region SR. This allows
A re-RIP image QA (see FIG. 2) can be obtained. The processing in steps SP41 and SP42 is performed by the blurring processing unit 61.

Further, in step SP43, a synthesizing process is performed. Specifically, the processing result image of step SP30 is used as a mask image QM (see FIG. 2). That is, the pixel value is “0” in the mask image QM.
The corresponding pixel of the re-RIP image QA generated in steps SP41 and SP42 is output for the pixel of, and the pixel having a pixel value of "1" in the mask image QM corresponds to the original image QB (see FIG. 2). The pixel is output as it is.

That is, steps SP41 and SP42 are performed for pixels determined to be pixels in the halftone dot image area.
Are applied, and the pixels determined to be pixels in the character area are determined in steps SP41 and SP4.
By adopting the result of not performing the processing of 2, the screen ruling conversion processing can be performed while separating the character area and the halftone image area in the designated area DR and suppressing the deterioration of the image.

Here, the case where the processing of steps SP41 and SP42 is performed for all the pixels in the designated area DR has been described. However, only the pixels of which the pixel value of the mask image QM is "0" are subjected to steps SP41 and SP42.
After performing the above processing, the images may be combined. For a pixel having a pixel value of “1” in the mask image QM,
Since the processing in steps SP41 and SP42 is not performed, the processing can be performed efficiently. That is, the processing can be speeded up.

FIG. 26 is a diagram showing an example of a mask image QM generated for the processing target image of FIG. FIG.
In 6, the black area represents the pixel value “1”,
A state in which a black area exists in an area corresponding to a character and an area adjacent thereto is shown. In other words, it can be seen that the character area CR is accurately separated from the halftone image area SR.

FIG. 27 shows the mask image QM of FIG.
FIG. 31 is a diagram showing a result of performing a screen ruling conversion process in step SP40 by using, and is a partially enlarged view corresponding to FIG. 30. In this way, it is possible to eliminate the blurring of the characters seen in the prior art and output the characters clearly.

<C. Modified Example><Automatic Extraction of Halftone Information> In the above description, the case has been described in which halftone information of the halftone image area SR is obtained by input from the operator. However, as described above, it is also possible to automatically extract the dot information of the dot image region SR based on the dot image region SR. Hereinafter, this automatic extraction operation will be described. This automatic extraction operation is performed by the dot information extraction unit 30.

When there are a plurality of areas designated using the area designation section 12, it is preferable to obtain dot information for each of the plurality of designated areas. Thereby, the halftone image area S included in each designated area
Even when the dot information of R is different from each other, the dot information of each dot image region SR can be accurately extracted.

FIG. 17 is a flowchart showing the outline of the dot information extraction processing, and FIG. 18 is an explanatory diagram relating to the dot information extraction processing.

As shown in FIG. 17, the intermediate value area extraction processing (step SP21), the autocorrelation data calculation processing (step SP22), and the dot information determination processing (step SP2)
Each process of 3) is performed.

First, in step SP21 of FIG. 17, an intermediate value area extraction process is performed.

Specifically, this processing is performed in the designated area DR
(Or the entire image to be processed) by using a reduced averaging process. For example, as shown in FIG. 18A, by adding and averaging the pixel values of 64 × 64 pixels in the designated area DR of the original image, the converted image (in other words, the reduced image) DS The pixel value of one pixel is calculated once. Then, as shown in FIG. 18B, a pixel ES whose pixel value is an intermediate value (for example, a gradation value of 10% to 90% of the maximum gradation value) among the pixels in the reduced image DS. Selected and each selected pixel E in the reduced image DS
A process of extracting each area ER corresponding to S from the designated area DR of the original image is performed.

Here, the converted pixel value of the solid black area or blank area in the designated area DR becomes a very large value or a very small value, while the designated area D
The pixel value of the converted pixel ES corresponding to the character area portion and the halftone image area portion in R is an intermediate value (intermediate value). Therefore, the portion of the halftone image region SR of the original image is
It is represented as a pixel having an intermediate value in the reduced image. Here, by utilizing this property, some pixels ES among a plurality of pixels having an intermediate value in the reduced image DS are used.
Is extracted, and each area of the original image corresponding to each extracted pixel ES is extracted as an intermediate value area ER. That is, these intermediate value areas ER are extracted as candidates for the dot image area SR.

Next, in step SP22, a process of calculating autocorrelation data described below is performed on each of these intermediate value regions ER.

FIG. 19 shows a target pixel in each intermediate value area ER.
FIG. 9 is a diagram illustrating a method for obtaining autocorrelation data of (xc, yc). As the pixel of interest, for example, the intermediate value area E
A pixel located at the center of R can be employed.

The periodic data for the pixel of interest (xc, yc) is calculated by examining the repetition of a periodic image pattern in the processing target image.

Specifically, first, the autocorrelation data S (a, b) of the peripheral area of the target pixel (xc, yc) is obtained by the following equation (1).

[0085]

(Equation 1)

Where ABS ｛｝ is a function for obtaining an absolute value,
P (x, y) is the gradation value of the pixel (x, y) of the image before processing, m and n are constants that determine the difference integration area E, a and b are shift amounts for comparing autocorrelation, and wx and wy are It is a constant that determines a range W for examining the autocorrelation characteristic for one central pixel (xc, yc).

(Xc, yc) = (4, 4), m = n = 1 (difference accumulation area: 3
× 3), wx = xy = 2 (a = b = -2 to +2), a = b =
The calculation form of the autocorrelation data S (a, b) at +2 is shown in FIG.
Shown in

Note that m, n, wx, and wy may be processed as fixed values set in advance, or may be configured to be appropriately changed by the operator from the input unit 6.

In the above equation (1), the autocorrelation data is obtained two-dimensionally.
The extraction principle will be described using one-dimensional autocorrelation data.

More specifically, in the two-dimensional pre-processed image, the existence of a repetition of a periodic image pattern in each of two pixel row directions orthogonal to each other, that is, the x direction and the y direction, is examined to determine the periodic image area. The description will be made assuming a case of extracting Specifically, the one-dimensional autocorrelation data H (a) and V (b) along the x direction and the y direction are respectively expressed by the following equations (2) and (3).
Ask by.

[0091]

(Equation 2)

[0092]

(Equation 3)

(Xc, yc) = (4, 4), m = 1 (difference integration area: 3 ×
1), when wx = 2 (a = −2 to +2), the calculation form of the autocorrelation data H (a) along the x direction when a = + 2 is shown in FIG. xc, yc) = (4,4), n = 1 (difference integration area: 1
× 3), wy = 2 (b = −2 to +2), and FIG. 21 shows the calculation form of the autocorrelation data V (b) along the y direction when b = + 2.

Next, in step SP23, the autocorrelation data (S (a, b) or H (a), V (
Based on (b)), a process of determining halftone dot information is performed by checking whether or not a periodic image pattern is repeated in the image.

In a portion where a periodic image pattern is repeated, such as a portion having the same intermediate gradation value in a halftone dot image, the autocorrelation increases in each cycle of the image pattern. , And the autocorrelation data obtained by the equations (3) become regularly smaller. Therefore, first, (A)
The minimum value of the autocorrelation data is searched, and it is checked that (B) those minimum values are equal to or lower than a predetermined level and (C) that these minimum values are regularly present. Note that the process (C) can be said to be a process in which symmetry with respect to the target pixel is considered.

In the following, the principle of the extraction will be described using the autocorrelation data H (a).

FIG. 22 shows (xc, yc) = (7, 3), m = 1, wx = 5 (a = -5
FIG. 11 is a diagram showing data indicating an example of P (x, y) and H (a) when the value is set to +5) and a graph of the H (a). That is, FIG. 11 shows the correlation characteristics between the attention area including the attention pixel and the surrounding area other than the attention area for the attention pixel (xc, yc) = (7, 3). Since a = 0 is the autocorrelation between the same pixels, H (0) = 0, which is the minimum value.

For the autocorrelation data H (a), the process of (A) is performed as follows: [(H (k-1)> H (k)) and (H (k) <H (k +
1)) k that satisfies the condition of [+] is obtained on the + side and-side. The value of H (k) for k that satisfies this condition is the minimum value.

The process (B) determines whether H (k) that satisfies the condition (A), that is, H (k), which is the minimum value, is equal to or less than a predetermined threshold value. A plurality of the thresholds are set in advance by the input unit 6 or the like. For example, in FIG. 22, “SL1 = 7.5” is set. If the minimum value H (k) is equal to or less than the threshold value SL1, it is determined in the process (B) that the value is equal to or less than the predetermined level.

In the processing of (C), for example, if k on the + side and k on the − side satisfying the condition of (A) are kP and km on the − side, whether or not (ABS {kp + km} ≦ 1) is satisfied To determine the presence or absence of regularity. Even if there is a minimum value in H (a), it is hard to say that there is periodicity if the value is somewhat large (exceeds threshold value SL1) or if these minimum values are present irregularly. . In this processing, since the level determination can be performed by the above (B), the presence or absence of the periodicity can be determined more reliably. Further, by determining the regularity of the minimum value using the above (C), the presence or absence of the periodicity can be more reliably determined.

Therefore, when all of the above conditions (A), (B) and (C) are satisfied, the range W around the pixel of interest (xc, yc)
Will have a periodic image pattern repetition. On the other hand, the autocorrelation data H (a) as shown in FIG. 23 has no periodicity, and there is no periodic image pattern repetition. Note that, in the character area CR, the state shown in FIG.

Although the one-dimensional processing has been described above, the same processing as described above is preferably performed two-dimensionally. Specifically, distances dX and dY are obtained by performing similar processing using two-dimensional autocorrelation data S (a, b) instead of one-dimensional autocorrelation data H (a). Can be.

Specifically, as shown in FIG. 24, among the pixels having the minimum values satisfying the above conditions (A), (B) and (C), the pixel EM having the shortest distance from the target pixel is obtained. . Then, a distance dX according to the position of the pixel EM,
dY may be obtained. The distances dX and dY are obtained by the cycle value calculator 33. Further, the periodic direction calculation unit 32 can also calculate the periodic direction θ, which is the direction in which the image pattern is repeated, as (arctan (dY / dX)). In the case where precision is not so required, a pixel which does not satisfy the condition (C) may be adopted as the pixel EM.

As described above, the distances dX and dY are calculated as the periodicity data for each of the plurality of designated target pixels (xc, yc). It should be noted that the above pixel EM
May exist in four directions with respect to the pixel of interest, but in this step SP23, the distances dX and dY with respect to the pixel EM in one direction may be obtained.
Specifically, for each pixel of interest, the pixel EM may be selected from pixels existing in a predetermined one of four quadrants from the first quadrant to the fourth quadrant.

Further, using the distances dX and dY for a plurality of pixels of interest obtained in this manner, each distance d
An average value of X and dY is obtained. Then, the average value is obtained as a calculation result regarding the distances dX and dY regarding the dot image region SR.

Here, the combination of (distance dX, distance dY) can be expressed as a combination of (dot frequency, halftone angle), and these expressions are equivalent to each other.
That is, obtaining the distances dX and dY is equivalent to obtaining dot information. As described above, the dot information determination processing is performed.

The four-direction vector calculation unit 40 can obtain the four-direction vectors V1 to V4 by using the information of the distance dX and the distance dY. Specifically, for example, one vector (for example, V1) is obtained based on the distances dX and dY, and a vector shifted by 90 degrees with respect to the vector (V1) is used for the other three vectors.
And two vectors (V2 to V4). Other operations may be performed in the same manner as in the above embodiment.

<Others> The relationship between the magnitude of the above-mentioned gradation value (or pixel value) and the brightness (black and white) of the image is such that the larger the gradation value, the whiter the image (in other words, the smaller the gradation value, Is blackened) and
There is a case where an image is defined as being blacker as the gradation value is larger (in other words, the image is whitener as the gradation value is smaller). The present invention can be applied to both cases.

Further, in the above embodiment, when a halftone image area SR having a plurality of types of halftone information is mixed in an image, a technique for separately obtaining halftone information for each designated area DR is exemplified. did. However, the present invention is not limited to this, and it is also possible to accurately obtain dot information by extracting dot information in units of predetermined blocks.

For example, as shown in FIG. 25, dot information can be extracted for each block Bi (see FIG. 25) which is a predetermined divided area. FIG. 25 shows a case where a digital image as a processing target image is divided into a plurality of blocks (partition areas) Bi
(I = 1,..., N; where N is the number of sections, here N = 24). This figure 2
In FIG. 5, each block Bi is shown as each area divided by a broken line.

Here, a maximum of one block Bi
It is assumed that the processing target image is partitioned such that there are halftone dot image regions SR. By extracting the halftone dot information for each of the divided blocks Bi in this way, even if the halftone image region SR having a plurality of types of halftone dot information is mixed in the image, each halftone dot image region SR Different dot information can be accurately obtained for each SR.

Further, in the above embodiment, the image processing apparatus 1 is configured by realizing each of the above functions by software in a computer, but is not limited thereto.
An image processing device that realizes the same processing using hardware such as a logic circuit may be constructed.

[0113]

As described above, claims 1 to 1 are described.
According to the invention described in (3), four vectors connecting a predetermined halftone dot and each of four adjacent halftone dots are obtained based on the halftone information, and a vector amount corresponding to each of the four vectors is obtained from the target pixel. Is selected, and using the four neighboring pixels and the pixel of interest, it is determined whether the pixel of interest is a pixel in the dot image area or a pixel in the character area. Therefore, it is possible to separate the halftone dot image area and the character area in the digital image.

In particular, according to the third aspect of the present invention, there is further provided a means for performing a smoothing process on the digital image, and the determining means performs the determining process on the image subjected to the smoothing process. Noise can be suppressed.

According to the fourth aspect of the present invention, the judgment means corrects the judgment result in the judgment processing by using the median filter processing for the area including the target pixel and its surrounding pixels. Can be suppressed.

Further, according to the fifth aspect of the present invention,
Since the determination unit corrects the determination result by using a fattening process for expanding an area determined as a character area, the determination as a character area can be prioritized in a boundary area between the character area and the dot image area. . Therefore, for example, in the case where the screen ruling change processing is further performed, it is possible to convert the ruling of the halftone dot image area while keeping the characters clear.

According to the seventh aspect of the present invention, the information extracted by the dot information extracting means is automatically obtained as the dot information, so that the operability is high.

Further, according to the invention described in claim 8,
Since the area to be subjected to the determination processing is specified, the processing speed can be increased by limiting the processing target area.

Further, according to the ninth aspect of the present invention,
Since the dot information obtaining means obtains the dot information for each area specified by the area specifying means, it is possible to obtain the dot information accurately.

According to the tenth aspect of the present invention,
Since the dot information extracting means extracts the dot information in a predetermined block unit, the dot information can be accurately obtained.

According to the eleventh aspect of the present invention, a result of performing a blurring process and a re-dotting process on a pixel determined to be a pixel in a halftone image region is adopted, and a character region is obtained. Since the result of not performing the blurring process and the re-dotting process is adopted for the pixel determined to be a pixel within, the screen frequency conversion process can be performed by separating the halftone image region and the character region. It is possible.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a hardware configuration of an image processing apparatus 1 according to an embodiment of the present invention.

FIG. 2 is a functional block diagram of the image processing apparatus 1.

FIG. 3 is a flowchart illustrating an operation of a screen ruling conversion process in the image processing apparatus 1.

FIG. 4 is a diagram illustrating a halftone image region SR and the like.

FIG. 5 is a diagram illustrating the principle of calculating a four-direction vector.

FIG. 6 is a flowchart showing a detailed operation of step SP30.

FIG. 7 is a flowchart showing a detailed operation of step SP40.

FIG. 8 is a halftone image region SR and a character region C of an original image.
It is a figure showing R.

FIG. 9 is a diagram illustrating pixel values of respective regions SR and CR of an original image.

FIG. 10 is a diagram showing an addition result RV.

FIG. 11 is a diagram illustrating a result of a binarization process.

FIG. 12 is a diagram showing another example of the median filter processing result.

FIG. 13 is a diagram illustrating a result of a fattening process.

FIG. 14 is a diagram illustrating an example of a smoothing filter.

FIG. 15 is a conceptual diagram showing a processing result for the halftone image region SR.

FIG. 16 is a conceptual diagram showing a processing result for a character area CR.

FIG. 17 is a flowchart illustrating an outline of a dot information extraction process.

FIG. 18 is an explanatory diagram of halftone dot information extraction processing.

FIG. 19 is a diagram illustrating a method of obtaining autocorrelation data of a target pixel (xc, yc) in an intermediate value area.

FIG. 20 is a diagram showing a calculation form of autocorrelation data H (a) along the x direction.

FIG. 21 is a diagram showing a calculation form of autocorrelation data V (b) along the y direction.

FIG. 22 shows data indicating an example of P (x, y) and H (a), and H
It is the figure which plotted (a).

FIG. 23 is a diagram showing an example of autocorrelation data H (a) in which there is no periodic image pattern repetition.

FIG. 24 is a diagram showing a periodic direction θ.

FIG. 25 is a diagram showing a plurality of blocks Bi.

26 is a diagram illustrating an example of a mask image QM generated for the processing target image of FIG. 28.

FIG. 27 is a diagram illustrating an example of a screen ruling conversion processing result using the present invention.

FIG. 28 is a diagram illustrating an example of a processing target image.

FIG. 29 is an enlarged view of the upper left character portion of FIG. 28;

FIG. 30 is a diagram showing an image obtained by performing a screen ruling conversion process according to the related art on the processing target image.

[Explanation of symbols]

 Reference Signs List 1 image processing device 50 character halftone separation processing unit 60 screen ruling conversion processing unit Bi block CR character area D0 halftone dot D1 to D4 each adjacent halftone dot DR designated area E1 to E4 neighboring pixel PE attention pixel QM mask image SR halftone dot Image area V1 to V4 Vector Θ Halftone angle

Claims (13)

[Claims] 1. An image processing apparatus for processing a digital image, comprising: dot information obtaining means for obtaining dot information; and a predetermined halftone dot and four adjacent halftone dots based on the halftone information. Means for obtaining four vectors connecting the target pixel, means for selecting four neighboring pixels shifted by a vector amount corresponding to each of the four vectors from the pixel of interest, and means for selecting the pixel of interest and the four neighboring pixels A determination unit for determining whether the pixel of interest is a pixel in a halftone dot image area or a pixel in a character area. 2. The image processing apparatus according to claim 1, wherein the determination unit obtains an addition result obtained by adding an absolute value of a difference between the target pixel and each of the four neighboring pixels.
An image processing apparatus, wherein the determination process is performed based on a magnitude relationship between the addition result and a predetermined value. 3. The image processing apparatus according to claim 1, further comprising: a unit configured to perform a smoothing process on the digital image, wherein the determination unit performs processing on the image that has been subjected to the smoothing process. An image processing apparatus for performing the determination processing. 4. The image processing apparatus according to claim 1, wherein the determination unit uses a median filter process on an area including the target pixel and its surrounding pixels to perform the determination process. An image processing device for correcting a determination result. 5. The image processing apparatus according to claim 1, wherein the determination unit corrects the determination result by using a fattening process for expanding an area determined as a character area. An image processing apparatus comprising: 6. The image processing apparatus according to claim 1, wherein said halftone dot information obtaining means has halftone dot information input means for inputting said halftone dot information, and An image processing apparatus, wherein information input using information input means is obtained as the halftone dot information. 7. The image processing apparatus according to claim 1, wherein the halftone dot information obtaining unit includes a halftone dot information extracting unit that extracts the halftone dot information from the digital image. An image processing apparatus for acquiring information extracted by the dot information extracting means as the dot information. 8. The image processing apparatus according to claim 1, wherein an area specifying unit that specifies an area in which the determination process is to be performed.
An image processing apparatus, further comprising: 9. An image processing apparatus according to claim 8, wherein said halftone dot information obtaining means obtains said halftone dot information for each area designated by said area designating means. . 10. The image processing apparatus according to claim 1, wherein said halftone dot information extracting means extracts said halftone dot information in a predetermined block unit. apparatus. 11. The image processing apparatus according to claim 1, wherein a pixel determined to be a pixel in the halftone image region is subjected to a blurring process and a re-dotting process. An image processing apparatus which adopts a result obtained by applying a blurring process and a re-dotting process to a pixel determined to be a pixel in the character area. 12. A computer-readable recording medium on which a program for causing a computer to function as the image processing apparatus according to claim 1 is recorded. 13. A program for causing a computer to function as the image processing device according to claim 1. Description: