JP2003309725A - Image processing method, image processing apparatus, and image forming apparatus provided therewith - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image forming apparatus capable of preventing occurrence of color missing, pseudo contour and color blur caused around black letters and color letters attendant on emphasis processing by an emphasis filter and avoiding void characters from being deformed. <P>SOLUTION: Whether or not a target pixel indicates a character edge is discriminated on the basis of pixel information in a block (S1), when the character edge is discriminated (S2), emphasis discrimination filter processing is applied to the target pixel and whether or not the emphasis filter processing is applied to the target pixel is discriminated on the basis of the result (S3). When it is discriminated that no emphasis processing by the emphasis filter is performed as the result of applying the emphasis discrimination filter processing to the target pixel, a difference between the density of the target pixel and the density of the target pixel subjected to the emphasis processing by the emphasis filter is obtained and the difference is distributed pixels around the target pixel (S10, S11). <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method and an image processing apparatus used in an image forming apparatus such as a digital color copying machine or a facsimile, and more particularly to detecting black character and color character areas of an image. The present invention relates to an image processing method, an image processing apparatus, and an image forming apparatus including the same, which performs appropriate processing based on the detection result.

[0002]

2. Description of the Related Art In recent years, digitalization of OA equipment has progressed rapidly, and the demand for color image output has increased, so that image forming apparatuses such as digital color copying machines and color laser printers have become widely popular. There is.

In these image forming apparatuses, the image processing technology occupies an important position in realizing high image quality and high functionality. In the digital color copying machine, color originals are R (red), G (green), and B (blue).
Colors are separated and read, and various image processing is applied to C (cyan), M (magenta), Y (yellow) and K (black).
4 color image data is converted to form a color image on the recording medium.

Here, the reason why the K signal is added is that, when reproducing an achromatic color such as black, when three colors C, M, and Y are superimposed and expressed, color turbidity and color shift occur.

In order to solve the above problems of color turbidity and color shift, black generation / background removal is performed. However, even if such a process is performed, 100% of the background is not removed, so that black characters are not reproduced as black and there is a problem that a color blur occurs around the black characters.

In order to improve this, there has been proposed a method of detecting a black character area existing in a document and performing processing such as an emphasis filter processing and a black emphasis processing to enhance the reproducibility of the character to improve the image quality. .

For example, Japanese Laid-Open Patent Publication No. 9-139843 (published on May 27, 1997) discloses a first-order differential filter and a second-order differential filter from image data of a document composed of chromatic color components and achromatic color components. The edge part of the black character is detected using the filter, the outer part and the inner part of the edge part are identified, and the chromatic / achromatic color data of the outer part and the chromatic color data of the inner part are respectively detected and the peripheral pixel. An image processing device is disclosed which replaces the smallest value among pixels and performs edge enhancement processing on the achromatic color data of the inner portion.

Further, as another approach, Japanese Patent Laid-Open No.
No. 0-316097 (Published November 14, 2000)
In Japanese), multiple color components in the color image input data
An edge area is detected by the edge part detecting means, emphasis determination processing is performed on a plurality of color components of the data portion determined to be the edge area, the processing result is compared with a predetermined threshold value, and the color image input is performed. It is determined whether or not to perform emphasis processing on the multiple color components in each pixel of the data,
Pixels that are determined to be subjected to the enhancement process are directly subjected to the enhancement process, and for pixels that are determined not to be subjected to the enhancement process, the target pixel is replaced as an achromatic color by using the average value of the multiple color components of the color image input data. An image processing method is disclosed.

[0009]

However, in the technique proposed in the above-mentioned Japanese Laid-Open Patent Publication No. 9-139843, the outside and inside of the edge portion are identified and the outside portion is replaced with a predetermined minimum value. By performing the edge enhancement processing only on the inner portion, it is possible to prevent the occurrence of blank areas around the characters and prevent the black characters from becoming thick.

However, if the data of the pixel of interest is replaced with a predetermined minimum value in this way, there is a large possibility that color blur around the black characters will remain, and further, small characters and thin lines may be deleted. . Further, in this conventional technique, the inside and outside of the edge portion are discriminated by the positive / negative of the coefficient of the second-order differential filter. Therefore, even when the inside portion of the edge portion is discriminated, the emphasis filter processing is performed. As a result, the density becomes lower than the background density, and a contour may occur around the character.

Therefore, in the method of distinguishing the outer part and the inner part of the edge portion and switching the processing, small characters and thin lines are deleted, and color fringing and contours occur around the characters, resulting in image quality. It has a problem of decreasing.

Further, in the technique proposed in Japanese Patent Application Laid-Open No. 2000-316097, emphasis processing for emphasis judgment is performed within a mask in a predetermined range centering on a target pixel, and the processing result and a predetermined reference value are set. Threshold comparison is performed. Here, for example, by setting the reference value based on the background of the image, the emphasis processing result becomes a value lower than that of the background, and the emphasis processing is performed on the pixels in which white voids around the characters may occur. It suggests to prevent the occurrence of color fringing by switching the processing so that it does not exist, calculating the average value of the multiple color components of the target pixel, and replacing it with an achromatic color using the average value of the multiple color components of the target pixel. ing.

However, according to the present technique, even a portion determined to be in the same edge area based on the result of the enhancement determination processing is replaced with the portion subjected to the enhancement processing and the achromatic color without performing the enhancement processing. Because of the divergence into parts, there may be a gap in emphasis between the two parts.

Further, in the edge, in addition to the black character, the color character, and the continuous tone and the contour line of the photograph portion which are arranged on the usual background or background, the edge is compared with the background or background of high density. There are also white letters and color letters that are hollowed out at low density. When such a blank character is emphasized, particularly in the case of a thin blank character, the character may be crushed by being dragged by the surrounding high-density background or background.

The present invention has been made in view of the above-mentioned conventional problems, and an object of the present invention is to eliminate color loss, pseudo contours, and color blurring around black characters and colored characters associated with the emphasis processing by the emphasis filter. It is an object of the present invention to provide an image processing method, an image processing apparatus, and an image forming apparatus including the image processing method, which can prevent the occurrence of the occurrence of characters and the collapse of a blank character.

[0016]

In order to solve the above-mentioned problems, an image processing method of the present invention extracts a block consisting of a plurality of pixels including a target pixel from input image data consisting of a plurality of color components, If the pixel of interest is a character edge based on the pixel information in the above, and if it is determined to be a character edge, the pixel of interest is subjected to emphasis determination filter processing, and based on the result, the pixel of interest is emphasized. In the image processing method for determining whether or not to apply the filter processing, when the result of performing the emphasis determination filter processing on the target pixel is determined not to perform the emphasis processing by the emphasis filter, the density value of the target pixel and the Find the difference from the density value when the pixel of interest is enhanced by the enhancement filter,
The feature is that this difference is distributed to pixels around the pixel of interest.

That is, when the pixel of interest is the character edge, the emphasizing processing by the emphasizing filter may make the density lower than the background density, and the contour may occur around the character. Therefore, in the method of switching the processing by identifying the outer portion and the inner portion of the edge portion, small characters and thin lines are deleted, and color blurring and contours occur around the characters, which deteriorates the image quality. I have a problem.

Further, even in a portion determined to be the same edge area based on the result of the enhancement determination processing, a portion subjected to the enhancement processing by the enhancement filter and a portion replaced with an achromatic color without performing the enhancement processing. Due to the branching into and, there may be a gap in emphasis between the two parts.

With respect to this point, in the present invention, when it is determined that the enhancement processing by the enhancement filter is not performed as a result of performing the enhancement determination filter process on the target pixel, the density value of the target pixel and the enhancement filter are applied to the target pixel. The difference with the density value when the enhancement processing of is performed is calculated, and this difference is distributed to the pixels around the pixel of interest.

Therefore, first, as a result of performing the emphasis determination filter processing on the pixel of interest, when the emphasis processing by the emphasis filter is performed, color loss, pseudo contour, color blurring and the like that occur around the above black characters and color characters occur. Considering that there is a possibility that this will occur, it is determined that the enhancement processing by the enhancement filter is not performed.

However, in this case, since the emphasis processing by the emphasis filter is not performed, the characters cannot be emphasized.

Therefore, in the present invention, when it is determined that the enhancement processing by the enhancement filter is not performed, the difference between the density value of the pixel of interest and the density value when the enhancement processing by the enhancement filter is performed on the pixel of interest. And the difference is distributed to pixels around the pixel of interest. By this method, characters and the like can be substantially emphasized with respect to surrounding pixels.

As a result, it is possible to prevent white spots and color bleeding around the black character edge caused by the highlighting process by the highlighting filter, and to prevent small blank characters from being crushed, and to properly select black letters, colored letters, blank letters, and blank letters. The edge part can be appropriately emphasized.

Therefore, a color dropout or a pseudo contour generated around a black character or a color character due to the emphasis processing by the emphasis filter,
Also, it is possible to provide an image processing method capable of preventing the occurrence of color fringing.

Further, in the image processing method of the present invention, in the above image processing method, when it is determined that the pixel of interest is a black character edge, an average value of color components other than the black component in the input image data is obtained. And replaces the density value of the pixel of interest with an achromatic color, and for image data representing the black component, obtains the difference between the density value of the pixel of interest and the density value of when the pixel of interest is emphasized, and the difference is calculated. Is distributed to pixels around the pixel of interest.

According to the above invention, when it is determined that the pixel of interest is a black character edge, the density value of the pixel of interest and the emphasis process are applied to the image data representing the black component. The difference from the density value of is calculated, and this difference is distributed to the pixels around the pixel of interest. As a result, as described above, the black character can be substantially emphasized with respect to the peripheral pixels.

Further, in the present invention, at this time, the average value of the color components other than the black component in the input image data is obtained and the density value of the pixel of interest is replaced with an achromatic color. Here, the reason why the average value is replaced is to eliminate the density difference between the respective color components such as C, M, and Y, and bring it closer to an achromatic color (black). As a result, black characters can be emphasized while preventing color fringing around the characters.

Therefore, when it is determined that the pixel of interest is a black character edge as a result of the determination, even if the enhancement process is not performed by the enhancement filter, the effect of the enhancement process and the factor of image quality deterioration associated with the enhancement process Both reduction can be achieved.

In the image processing method of the present invention, in the above image processing method, when it is determined that the enhancement processing by the enhancement filter is not performed as a result of performing the enhancement determination filter processing on the target pixel, the attention processing is performed. It is determined whether or not the pixel is the edge of the extracted character. When it is determined that the pixel is the edge of the extracted character, the density value of the pixel of interest and the density when the emphasis process is performed on the pixel of interest by the emphasis filter. The difference from the value is distributed by subtracting it from the density values of the peripheral pixels of the pixel of interest, while the difference is added to the density values of the peripheral pixels of the pixel of interest when it is determined that it is not the edge of the extracted character. It is characterized by allocating by doing so.

That is, in the edge, in addition to the black character, the color character, and the continuous gradation and the outline of the photo area which are arranged on the normal background or the background, the background is compared with the high density background or the background. There are white letters and color letters that are hollowed out at low density. When such an extraction character is emphasized by an emphasis filter, particularly in the case of a thin extraction character, the character may be crushed by being dragged by the surrounding high density background or background.

With respect to this point, in the present invention, when it is determined that the edge is an outline character, the difference between the density value of the target pixel and the density value when the target pixel is subjected to the enhancement processing by the enhancement filter is noted. The pixel is distributed by subtracting it from the density values of surrounding pixels.

As described above, at the edge of the extracted character, the effect of the enhancement filter at the edge is always exerted in the decreasing direction with respect to the density of the peripheral pixels, so that the density of the edge is higher than that of the background pixel in the peripheral (high density It is possible to prevent the characters from being crushed due to an unnecessarily high height compared to the background pixels) and to substantially emphasize the extracted characters.

On the other hand, when it is judged that the edge is not the edge of the extracted character, the difference is distributed by adding it to the density values of the peripheral pixels of the target pixel.

As described above, when the edge is not the outline character, that is, in the case of the normal edge area, the effect of the emphasis filter on the edge is always exerted on the peripheral pixels in the increasing direction, so that the edge density is increased. It is possible to prevent white spots and color fringing from occurring due to the density being lower than the density of background pixels in the vicinity (background pixels having a low density), and substantially normal edges can be emphasized.

Therefore, by performing the processing suitable for each of the processing method in the normal edge area where white spots and color bleeding may occur and the non-extracting character where the character collapse may occur. , It is possible to further minimize the influence of the image quality deterioration caused by the enhancement filter.

As a result, there is provided an image processing method capable of preventing the occurrence of color dropouts, pseudo contours, and color bleeding around black characters and color characters associated with the emphasis processing by the emphasis filter, and preventing the collapse of the omitted characters. Can be provided.

In order to solve the above problems, the image processing apparatus of the present invention extracts a block composed of a plurality of pixels including a target pixel from input image data composed of a plurality of color components,
A character edge determination unit that determines whether or not the pixel of interest is a character edge, and whether or not to perform emphasis determination filter processing on the pixel of interest and whether or not to perform emphasis processing by the emphasis filter on the pixel of interest based on the result. In an image processing apparatus including a character edge determination unit that includes an enhancement processing switching determination unit that performs determination, as a result of performing enhancement determination filtering processing on a pixel of interest by the enhancement processing switching determination unit, the enhancement processing by the enhancement filter is performed. When it is determined that there is no such difference, the difference between the density value of the target pixel and the density value when the target pixel is subjected to the enhancement processing by the enhancement filter is calculated, and the difference is distributed to the pixels around the target pixel. It is characterized in that means are provided.

According to the above invention, when the emphasis processing switching judgment unit judges that the emphasis processing by the emphasis filter is not performed as a result of performing the emphasis judgment filter processing on the pixel of interest, the spatial filter means The difference between the density value of the target pixel and the density value when the target pixel is subjected to the enhancement processing by the enhancement filter is calculated, and this difference is distributed to the pixels around the target pixel.

Therefore, it is possible to prevent black spots and color bleeding around the black character edge caused by the highlighting processing by the highlighting filter, and to prevent the small blank characters from being crushed, and to appropriately select black letters, colored letters, blank letters, and blank letters. The edge part can be appropriately emphasized.

Further, in the image processing apparatus of the present invention, in the above-mentioned image processing apparatus, the character edge determination means is provided with an extraction edge determination section for determining whether or not the pixel of interest is an edge of an extraction character. It is characterized by being.

That is, in the normal edge, the target pixel often has a higher density than the peripheral pixels, and when the enhancement processing is normally performed by the enhancement filter, the peripheral color loss and the color blurring occur, while the blank edge is The target pixel often has a lower density than the peripheral pixels, and the problem of collapsing in small letters is likely to occur.

In this respect, according to the present invention, the character edge determining means is provided with the extracted edge determining section, and the extracted edge determining section determines whether or not the pixel of interest is the edge of the extracted character. To do.

Therefore, by dividing the character edge into a normal edge and a blank character, the respective edge portions can be appropriately emphasized.

An image forming apparatus of the present invention is characterized by including the above image processing apparatus in order to solve the above problems.

According to the above-mentioned invention, it is possible to prevent white spots and color bleeding around the edges of black characters due to the emphasis processing by the emphasis filter, and to prevent the collapse of small blank characters.
By dividing the case with blank characters and white characters, you can properly emphasize each edge part,
An image forming apparatus that can output a high quality image can be provided.

[0046]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to FIGS. In the present embodiment, as an image forming apparatus to which the present image processing method is applied and which includes the present image processing apparatus,
Although the digital color copying machine will be described, the image forming apparatus is not necessarily limited to this, and can be applied to an image forming apparatus such as a facsimile.

The image processing apparatus used in the digital color copying machine as the image forming apparatus of this embodiment is shown in FIG.
As shown in FIG. 3, an A / D (analog / digital) conversion unit 2, a shading correction unit 3, an input gradation correction unit 4, a color correction unit 5, a black generation / undercolor removal unit 6, and characters as character edge determination means. An edge determination unit 10, a spatial filter unit 7 as a spatial filter unit, and a halftone generation unit 8 are provided. Then, the image processing apparatus is connected to the image input apparatus 1 and the image output apparatus 9 which are, for example, a scanner, thereby forming a digital color copying machine.

In the image processing apparatus, a color image signal (RGB analog signal) input by scanning an original with the image input apparatus 1 is converted into a digital signal by the A / D (analog / digital) conversion section 2. To be converted.
Next, the shading correction unit 3 removes various distortions generated in the illumination system, imaging system, and imaging system of the image input apparatus 1.

In the input gradation correction section 4, the color balance is adjusted, and at the same time, the input color image signal is a reflectance signal, that is, a signal proportional to the amount of reflected light reflected from the original by the exposure lamp. It is converted into a density signal suitable for processing.

Next, the color correction unit 5 performs a process for removing color turbidity in order to realize faithful color reproduction, and the black generation / undercolor removal unit 6 reproduces a color image with four CMYK colors. Processing is done. At this time, in order to improve the reproducibility of black characters / colored characters in a mixed character / photo document, the character edge determination unit 10 detects the character edge portion, and the number of prepared ones is further determined based on a predetermined determination result. The enhancement process considered to be most appropriate is selected from those enhancement processes. Details will be described later.

The result is output to the black generation / undercolor removal unit 6, the spatial filter unit 7, and the halftone generation unit 8 and appropriate emphasis processing is applied to all character edge portions.

Here, as shown in FIG. 3, the character edge determination unit 10 is roughly classified into a character edge detection / color determination unit 20, a feature amount calculation / determination unit 30, an emphasis process switching determination unit 40, and The removal edge determination unit 50 is configured. The character edge detection / color determination unit 20 and the feature amount calculation / determination unit 30 constitute a character edge determination unit of the present invention.

In the character edge determination unit 10, first, 4
As shown in FIG. 4, a 5 × 5 mask centered on the pixel of interest P is formed from the input image data (CMY signal) that has been subjected to the color correction processing using the line memory of the book, and is input to each unit. . Although a 5 × 5 mask is formed in this embodiment, the size of the mask is not limited to this.

Next, in the character edge detection / color determination unit 20, the edge is detected by the Sobel filter and the color determination is performed, and the black edge determination signal Bk or the color edge determination signal Col is output. The feature amount calculation / judgment unit 20 calculates the feature amount in the mask, determines whether or not it is a character edge, and outputs a character edge determination signal EDGE. And
AND signals 11 and 12 take the logical product of both signals,
The black character edge determination signal BkEDGE and the color character edge determination signal ColEDGE are output.

Further, in the emphasis processing switching determination section 40,
By the emphasis determination process, it is determined whether or not there is a possibility that a blank area or a contour will occur around the character, and an emphasis switching determination signal EMP is output. In addition, the removal edge determination unit 5
At 0, it is determined whether the edge in the mask is a normal edge that exists as a relatively high density on a light background or a cut edge that is in a dug state with a relatively low density on a dark background. Is performed, and the extraction edge determination signal B
GC is output.

The four determination processes will be described in detail below.

First, with respect to the character edge detection / color determination unit 20, the flow of processing of edge detection and color determination will be described with reference to FIG.

The character edge detection / color determination unit 20 includes an edge detection circuit 21 and a color determination circuit 2 as shown in FIG.
2 and AND circuits 23 and 24. First,
The edge determination circuit 21 will be described.

In the edge determination circuit 21, as shown in FIG. 6A, a 5 × 5 mask centered on the target pixel P read from the line memory is changed to a 3 × 3 pixel mask centered on the target pixel P. It is extracted and input to the main scanning direction Sobel filter unit 21a and the sub scanning direction Sobel filter unit 21b. In the main scanning direction Sobel filter portion 21a, the 3 × 3 pixel mask is formed, for example, as shown in FIG.
As shown in (b), the main scanning direction Sobel filter MF
While being folded, the Sobel filter unit 2 in the sub-scanning direction
In 1b, as shown in FIG. 6C, the sub-scanning direction Sobel filter SF is convoluted. The absolute values of the results obtained by convoluting the Sobel filters MF and SF with respect to the pixel of interest P are obtained by absolute value calculation units 21c and 21c, and the results are added by an adder (ADD) 21d. To be done. Here, the absolute value is obtained because there is a negative coefficient in each Sobel filter MF · SF, and the addition is to take into consideration the results in both the main scanning direction and the sub scanning direction. . The addition results in both directions are input to the next comparator 21e and compared with the threshold edgeTH. If the addition result is equal to or greater than the threshold edgeTH, "1" is output, and "0" is output otherwise.

The above processing is applied to each of C, M and Y colors. Then, the output signals of the three colors for the pixel of interest P are added up by the adder (ADD) 21f, and then the comparator 21g.
Is compared with the threshold value addTH, and an edge detection signal of "1" is output if it is an edge and "0" if it is not an edge. For example, when the target pixel P is determined to be an edge only when the edges of the three colors overlap, the threshold edgeTH =
If an edge is detected even in 3 or 1 color and the pixel of interest P is determined to be an edge, a threshold value edgeTH = 1 is set.

Next, in the color judgment circuit 22, the maximum value detection unit 22a and the minimum value detection unit 22b detect the maximum value and the minimum value of the C, M, Y signals in the target pixel P, and the difference unit
The difference is obtained at (SUB) 22c. Next, the result is compared with the threshold value subTH by the comparator 22d to determine whether it is a chromatic color or an achromatic color. That is, the difference between C, M, and Y in the pixel of interest P is within a predetermined range (threshold value subT
If it is larger than H), it is determined to be a chromatic color, and if it is within a predetermined range, it is determined to be an achromatic color. By this processing, a color determination signal of "1" is output for an achromatic color, that is, black, and "0" otherwise.

The AND circuit 23 obtains the logical product of the edge detection signal and the color determination signal output by the above-described processing, and the black edge determination signal Bk is obtained. That is, only pixels that are detected as edges and whose color determination result is black are determined as black edge pixels. At the same time,
The logical product of the edge detection signal and the negative signal of the color determination signal is also A
The color edge determination signal Col obtained by the ND circuit 24 is obtained. Similar to the case of the black edge determination signal, only pixels which are detected as edges and whose color determination result is not black, that is, chromatic color are determined as color edge pixels.

Next, the characteristic amount calculation / determination unit 30 will be described.

As shown in FIG. 7, the feature amount calculation / determination unit 3
0 is composed of a first characteristic amount calculation circuit 31, a second characteristic amount calculation circuit 32, a determination circuit 33, and an AND circuit 34. In this embodiment, the processing is performed using only the M signal, but the reference signal here is not limited to the M signal.

In the first feature amount calculating circuit 31, first, in the 5 × 5 mask read from the line memory, the maximum value detecting unit 31a calculates the maximum density in the mask, while the minimum value detecting unit 31b. Then, the minimum density is calculated, and the difference (maximum density difference) is calculated by the difference unit (SUB) 31c. These values, that is, the maximum density / minimum density / maximum density difference, are determined by the threshold value THma set for each of the comparators 33a, 33b, 33c in the determination circuit in the subsequent stage.
x.THmin.THdiff are compared with each other, and the comparison result is ANDed by the AND circuit 34. And
A signal "1" is output for a character edge, and a signal "0" is output otherwise. This is the character edge determination signal EDGE.

Three threshold values THmax used at this time
THmin · THdiff is determined by the following feature amount calculated from the 5 × 5 mask in the second feature amount calculating circuit 32.

Generally, even if characters on a document have the same density, the density value read by a CCD (Charge Coupled Device) and given as an input varies depending on the size of the character. In consideration of this, the thresholds THmax, THmin, and THdiff are switched based on the size of the character.
As the feature amount for switching the threshold values THmax, THmin, and THdiff, the area and the number of inversions obtained when binarizing the inside of the 5 × 5 mask are used.

First, the 5 × 5 mask is binarized by the binarizing unit 32a. For example, a binary mask is obtained by setting "1" if the density of each pixel is higher than the average density in the mask and "0" otherwise. When there is a character on a low-density background, the character area corresponds to "1" and the background area corresponds to "0". When there is a blank character on the high-density background, the character area corresponds to "0" and the background area corresponds to "1".

Next, the area and the maximum number of inversions are calculated from this binary mask. At this time, when the number of pixels of "1" is larger than the number of pixels of "0" and when the number of pixels of "0" is larger than the number of pixels of "1", that is, whether the character is a normal character or a blank character, the determination procedure is Although different, the description will be made below assuming a normal character.

That is, the area is calculated by the area counting unit 32.
In b, the area area1 is output by counting the number of "1" pixels in the binary mask, and the area area0 is output by counting the number of "0" pixels. Calculate the maximum number of inversions
The maximum inversion counter 32c counts the number of times "1" and "0" are inverted in each line in the main scanning and sub-scanning directions, and outputs the maximum inversion number value change. For example, in the case of the binary mask shown in FIG. 8, the area area1 = 1
5, area area0 = 10, and maximum inversion count value change = 3 are obtained. Area area1 and area, which are these three features
Based on area0 and the maximum inversion count value change, the threshold selection unit 35
At the threshold value THmax / THmin / THdiff is switched.

Next, the threshold selection section 35 will be described with reference to FIG.

The threshold selection unit 35 has strict thresholds maxH for maximum density, minimum density, and maximum density difference.
-MinL-diffH and a loose threshold value maxL-minH-diffL are predetermined, and each is [maxH, maxL]-
A combination of [minL, minH] and [diffH, diffL] is set for each selector. Further, in the preceding stage of each selector, the area area1, the area area0, and the area area1, which is the number of pixels of "1" in the binary mask, are set to the maximum inversion count value chan.
The value divided by ge is calculated by the divider A / B 35a, and the threshold values a1TH, a0TH, and chT are set in advance corresponding to the respective values.
Comparators 35b, 35c and 35d for comparing with H are provided.

The method of selecting the thresholds THmax, THmin and THdiff will be described below.

First, when the area area1 of the character area is large to some extent (area area1> threshold a1TH), the maximum density is easy to be sensed, so the strict threshold maxH is selected as the threshold THmax by the selector 35e, otherwise it is loose. Threshold
maxL is selected. When the area area0 of the background region is large to some extent (area area0> threshold a0TH), the minimum density is easy to be sensed, so the strict threshold minL is selected by the selector 35f as the threshold THmin, and otherwise the loose threshold minH is selected. . Also, small letters are area
1 becomes smaller or the maximum inversion number value change becomes larger. Conversely, for large characters, the area area1 increases or the maximum inversion count value change decreases.

Therefore, the area area1 is set to the maximum inversion number value chan.
When the value divided by ge is large to some extent ((area area1 / maximum inversion count value change)> threshold chTH), it is easy to detect the maximum density difference, so the strict threshold diffH is selected as the threshold THdiff by the selector 35g, and other than that. , A loose threshold value diffL is selected.

The above description is for a normal edge,
In the case of a punched edge, the handling of the area area0 and the area area1 is reversed, but the determination procedure is the same and therefore omitted.

As described above, the three thresholds THmax, THmin and THdiff are switched using the feature amount of the binary mask. In the present embodiment, two stages of a strict threshold and a loose threshold are provided, but it goes without saying that thresholds of several stages can be provided for selection.

Next, the cut edge determining section 50 shown in FIG. 3 will be described in detail with reference to FIG.

As shown in the figure, the above-mentioned cut edge determining section 50 includes a chromatic / achromatic determining section 51 and a density determining section 52.
And an AND circuit 53 and an adder circuit (ADD) 54.

In the chromatic / achromatic judging section 51, the absolute difference value | C− for the set of color components C, M, and Y at any pixel in the 5 × 5 mask read from the line memory is first calculated.
M |, absolute difference value | MY |, and absolute difference value | Y-C
| Is calculated, a pair having the largest absolute value is obtained from the pair of difference absolute values, the difference absolute value at that time is compared with the threshold value BkTH, and when the difference absolute value is smaller than the threshold value BkTH. "1" as the pixel is achromatic
Is output, and when the absolute difference value is equal to or greater than the threshold value BkTH, the pixel is determined to be a chromatic color, and the determination signal of “0” is output.

In the density determining section 52, the color components C, M,
The density value of any one of Y (here, M) and the threshold value whT
H is compared, and when the M density value is larger than the threshold value whTH, "1" is output as the density determination signal, and when the M density value is within the threshold value whTH, "0" is output as the density determination signal. Then, a logical product of the chromatic / achromatic judgment signal and the density judgment signal is obtained by the AND circuit 53, and a signal of "1" is output in the case of high density and achromatic color, that is, black, and otherwise " The signal "0" is output.

The above determination is applied to all the pixels in the 5 × 5 mask, and the adder (ADD) 54 calculates the number of pixels determined to be black in the 5 × 5 mask. The number of black pixels is compared with the threshold value bgTH, and a signal of "1" is output when the number of black pixels is larger than the threshold value bgTH, and a signal of "0" is output when the number of black pixels is smaller than the threshold value bgTH. This is the extracted edge determination signal BGC. When the extracted edge determination signal BGC is "1", it is determined that the background or background is black, and when the target pixel is a character edge, the edge may be an extracted edge. It can be said that it is highly effective.

Next, the emphasis process switching determination section 40 shown in FIG. 3 will be described with reference to FIGS. 11 and 12.

In general, when there is a black character on a low-density background as shown in FIG. 12A, if emphasis filter processing is performed on the entire black character edge, as shown in FIG. White spots and outlines of lower density than the background (background) occur. Further, as shown in FIG. 12C, when there is a white (blank) character on a high-density background, if emphasis processing is performed on the entire white character edge, especially when the blank character is a thin line, As shown in FIG. 12 (d), there is a possibility that the characters will be crushed. The emphasis switching determination is performed for the purpose of preventing these.

At the black character edge and the white character edge,
Since the flow of the determination process is the same, the following description will be given in detail for the black character edge.

First, as shown in FIG. 11, the 5 × 5 masks of C, M, and Y colors read from the line memory are input to the emphasis determination filter section 41 ... The emphasis determination filter EJF shown in 13 is convoluted. The emphasis determination filter EJF used at this time is a kind of emphasis filter, and is not necessarily limited to the one shown in FIG.
If the actual enhancement filter parameter to be performed in the subsequent processing has already been determined, it is more suitable to use it as it is.

For example, when the emphasizing judgment filter EJF is convoluted into the character area having the density distribution of the character area shown in FIG. 14A, as shown in FIG. 14B, a portion showing a negative value in the character edge portion. May appear. In this way, the part that shows a negative value when applying one type of emphasis filter is
It may be a factor of white spots around the characters. Therefore, by detecting this portion in advance, it is possible to prevent the occurrence of blank areas. Furthermore, since color bleeding often exists around black characters, color bleeding around black characters can be prevented by replacing the data in this portion with data close to black. Also, when replaced with data close to black, small characters and thin lines will not disappear. Strictly speaking, color fringing also occurs around the color letters, but it is generally impossible to determine whether the color fringing around the color letters is a color fringing or a part of the color letters. This is not a problem here because it is difficult and does not cause visual deterioration of image quality as compared with black characters.

As shown in FIG. 11, the value of the 5 × 5 mask result of each of C, M, and Y colors obtained by convolving the above-described emphasis determination filter EJF is predetermined by each comparator 42 as a comparison means. It is compared with the reference value (reference value C / reference value M / reference value Y). If it is larger than the reference value, "1" is output, and if it is smaller, "0" is output. Here, as the reference value, for example, the output “0”, that is, the positive / negative sign of the result obtained by convolving the emphasis determination filter EJF is used. By making a determination based on the positive and negative signs, it is possible to prevent blank areas around the characters.

On the other hand, it is also possible to use the background density value as the reference value. As the background density, for example, the minimum value of each color in the 5 × 5 mask is used. If the background density value is used as the reference value, not only white spots around the character but also
As shown in (b), it is possible to prevent a contour having a density value lower than that of the background. As a result of performing the emphasis determination filter process, even if the sign of the determination result is positive, if the value of that part is lower than the background density, if the emphasis process is uniformly applied, it occurs as a contour around the character. This is because it can be a factor.

Next, as shown in FIG. 11, the comparison results of the respective colors are added by the adder (ADD) 43 to obtain the feature amount of the pixel of interest P, and then the comparator 44 as the signal generating means. Is compared with the threshold value empTH, and a signal of "0" or "1" is output as a reference value determination signal. The threshold value empTH is set to "3" because emphasis processing is performed only when it is determined that all three colors are larger than the reference value.

Further, the difference between the processing result obtained by convoluting the emphasis judgment filter EJF of each color and the reference value (hereinafter, "reference difference value").
Is written. ) Is calculated by the difference unit (SUB) 45 ..., The reference difference value of each color is added by the adder (ADD) 46, and then the sum of the reference difference values and the threshold value ovTH are compared.
7, when the total of the reference difference values is less than the threshold ovTH, "1" is output as the reference difference value determination signal, and when the total of the reference difference values is greater than or equal to the threshold ovTH, "0" is the reference difference value. It is output as a determination signal. This shows how much the degree of change in density occurs when emphasis processing is performed, and even if the result of the reference value determination is negative,
When the reference difference value is small, it is considered that the effect of color loss or color fringing is visually small. At this time, the threshold value ovTH is 20 to 20 as a density at which a clear difference is visually felt.
For example, 60 to 90 are set as 30 × 3 colors.

Next, the reference value determination signal and the reference difference value determination signal are bit-combined, and finally the emphasis switching determination signal EMP is obtained. Emphasis switching determination signal EMP =
Pixels of "11" or "10" can be subjected to normal enhancement processing. Pixels with the emphasis switching determination signal EMP = "01" may have problems such as white spots, outlines, and color bleeding, but they are considered to occur at a level at which there is no problem visually. Good. In addition, since the pixels of the emphasis switching determination signal EMP = “00” indicate that white spots, outlines, and color bleeding occur to a visually problematic level, the emphasis process is not performed.

In the case of a blank character, the flow of processing is the same as in the case of a normal character, but the branching of the judgment result is different. That is, the emphasis switching determination signal EMP = “00” or “0”
The pixel of 1 ”can be subjected to the emphasis process. The pixel of the emphasis switching determination signal EMP =“ 11 ”may cause character collapse, but it is considered to be a level that does not have a visual problem. The pixel of the emphasis switching determination signal EMP = “10” may not be emphasized because the character collapse may occur up to a visually problematic level.

As for the pixels determined not to perform the enhancement processing based on the enhancement switching determination signal, the change in the density value before the enhancement processing and the change in the density value after the enhancement processing are performed around the target pixel. It will be allocated to existing pixels. At this time, when the target pixel is a normal edge, subtraction with respect to peripheral pixels causes white spots (see FIG. 1).
2 (b)), the density change amount is allocated in the addition (+) direction, and conversely, when the target pixel is the extraction edge, addition to peripheral pixels causes a character collapse (FIG. 12). Because of (d)), the density change amount is allocated in the subtraction (-) direction.

Next, the flow of the entire processing in the above image processing method will be described based on the flowchart shown in FIG.

First, in the character edge detection / color judgment unit 20 and the feature amount calculation / judgment unit 30 of the character edge judgment unit 10, a character edge judgment process is performed on the pixel of interest P (S1, S2), and a black character edge judgment is made. The signal BkEDGE or the color character edge determination signal ColEDGE is generated. With respect to the pixel determined to be the edge portion (black character edge or color character edge), the emphasis processing switching determination unit 40 performs the emphasis switching determination on the target pixel P (S3, S4).

The pixel of interest P is selected by the emphasis process switching determination unit 4
At 0, the emphasis determination filter processing is performed and compared with the reference value determined for each color. The reference difference value which is the difference between the processing result value of each color and the reference value is compared with the threshold value ovTH to obtain the reference difference value determination signal, and the comparison result of each color is added and then compared with the threshold value empTH. The reference value determination signal is obtained. From the reference difference value determination signal and the reference value determination signal thus obtained, the final emphasis switching determination signal EMP
Is generated. In the extraction edge determination unit 50, the target pixel P
The extraction edge determination signal BGC is generated from the frequency of occurrence of the number of black pixels in the peripheral region including the.

These black character edge determination signals BkEDG
Based on E, the emphasis switching determination signal EMP, and the extraction edge determination signal BGC, black generation / undercolor removal processing (S5,
S6), emphasis filter process (S7), process of replacing chromatic color with achromatic color (S8), through process (S9), distribution of density change amount addition direction in spatial filter section (S10),
Sorting in the subtraction direction by the density change in the spatial filter (S1
1), halftone generation processing for areas other than edges (S12),
The edge area halftone generation processing (S13) is selected and processed. The double line below S6 and S8 and S1
Double lines below 0 and S11 indicate parallel processing.

Detailed contents of the above processing will be described with reference to FIGS.
It will be described below using 0.

First, the black generation process is a process for reproducing CMY colors with four CMYK colors for the purpose of improving the reproducibility of black and saving the total amount of toner and ink used. For example, FIGS. ), The density of the color having the lowest density (C in the figure) of C, M, and Y forming an arbitrary pixel is replaced with K, and the original C, M, and
The concentration obtained by subtracting the concentration of K from the concentration of Y is newly calculated as C.
C, M, Y, and K four colors are generated as the density of M and Y. This time, simply set the minimum density of the C, M and Y pairs to 100%.
Although it is replaced with K, in reality, if 100% of all is replaced with K, the saturation decreases, so it is common to replace some of the minimum density with K.

For example, as shown in FIGS. 16 (b) to 16 (e), a pixel region of 3 × 3 size and the enhancement filter shown in FIG. 16 (a) are assumed. The central pixel (hatched pixel) shown in FIGS. 16B to 16E is the pixel to be processed, and the emphasis processing shown in FIG. 16A is performed on the pixel area centered on this pixel. (1) When it is determined that the target pixel is not a character In this case, black generation / undercolor removal and enhancement filter processing are not performed on the target pixel, and the target pixel is sent to the halftone generation unit in the subsequent stage, Binarization or multi-value conversion processing by a halftone screen that emphasizes the key is performed. (2) When the target pixel is black characters and it is determined that enhancement processing is possible In this case, black generation / undercolor removal is performed on the target pixel, and the minimum density is obtained from the C, M, Y signals after color correction. Detect
The K (black) signal is generated with this minimum density, and the original C
This K signal is subtracted from the M ・ Y signal to create a new C ・ M
Generate a Y signal. Of these signals, the K signal
The emphasis process is performed using the filter coefficient illustrated in FIG. 17 (a) and 17 (b) show how the density of the target pixel changes before and after these processes. The processing result is sent to the halftone generating unit in the subsequent stage, and binarized or multivalued processing is performed by a halftone screen that emphasizes resolution. (3) When the target pixel is a black character and it is determined that the enhancement processing is not possible In this case, as shown in FIG. 18A, the same black generation / undercolor as the case of (2) is performed on the target pixel. A new C, M, Y, K signal is generated by the removal, and for these C, M, Y signals, a process of replacing the chromatic color with an achromatic color as shown in FIG. 18B. Is done. This is a process of calculating an average value of three C, M, and Y signals and using this average value as a new C, M, and Y signal. The reason why the average value is replaced here is to eliminate the density difference between the C, M, and Y colors and bring it closer to an achromatic color (black). As a result, black characters can be emphasized while preventing color fringing around the characters.

For the K signal, the amount of density change before and after the processing when the enhancement processing is originally performed is distributed to the peripheral pixels. Note that this processing is performed by the spatial filter unit 7.
Originally, when the emphasis processing was performed, the density of the K signal was as shown in FIG.
As shown in FIG. 17B, it becomes “100”, and FIG.
As shown in, the density of “5
Since a difference of "100-50 = 50" occurs with respect to 0 ", this density difference is allocated to the peripheral pixels of the target pixel.

As the allocating method, in the present embodiment, the difference is evenly allocated to all the peripheral pixels, and the number of peripheral pixels is eight. Therefore, FIG.
As shown in (d), “50/8 = 6 (rounded up to the right of the decimal point)” is the assigned density.

Specifically, when the determination result of the target pixel in the extraction edge determination unit 50 is determined to be a normal edge, this allocation density is added (+) as shown in FIG. 18C. ) Direction is allocated to the peripheral pixels, and when it is determined that the pixel is a cut edge, the allocated density is allocated in the subtraction (-) direction as shown in FIG. 18D. However, since the density is not less than minus, even if the processing result after sorting is 0 or less, the result is 0.

The processing result is sent to the halftone generating section 8 in the subsequent stage, and binarized or multivalued processing is performed by a halftone screen that emphasizes resolution. (4) When it is determined that the target pixel is a color character and can be emphasized In this case, the black generation / undercolor removal processing is not performed on the target pixel, and FIG.
The enhancement process is performed using the filter coefficient shown in (a).

FIG. 19 shows how the density of the target pixel changes before and after the processing. In the present embodiment, the density range is 8 bits per color, and the maximum is 255. Therefore, even if the processing result exceeds 255, it is rounded to 255. The processing result is sent to the halftone generating unit in the subsequent stage, and binarized or multivalued processing is performed by a halftone screen that emphasizes resolution. (5) When it is determined that the target pixel is a color character and cannot be emphasized In this case, black generation / undercolor removal and emphasis filter processing are not performed on the target pixel. For the C, M, Y signals, the density change amount before and after the processing when the emphasis processing is performed is originally distributed to the peripheral pixels. Originally, when the emphasizing process is performed, the density of the C, M, Y signals is as shown in FIG.
“80, 170, 255”, which is “80-50 = 3” for the through processing result shown in FIG.
Since a density difference of 0, 170-100 = 70, 255-200 = 55 "occurs, as shown in FIGS. 20B and 20C, as in the case of (3), this difference is added to the peripheral 8 pixels in the addition direction. Or, allocate in the subtraction direction.

The processing result is sent to the halftone generating unit in the subsequent stage, and binarized or multivalued by a halftone screen that emphasizes resolution. (6) When the determination signal is taken into consideration for the allocated pixels In (3) and (5) in the above-described processing example, the change amount before and after the emphasis process is evenly allocated to all the surrounding pixels.
A method of selecting the allocation target pixel based on the area determination signal is also conceivable.

The target pixel and the peripheral pixels are modeled as shown in FIG. 22 (a), and the area determination of the target pixel and the peripheral pixels 9 pixels at this time is as shown in FIG. 21 (a). . Here, for simplification of processing, attention is paid to black characters or color characters, normal edges or blank edges as a criterion for allocation selection, and an emphasis determination signal indicating whether or not emphasis processing is possible is not considered. To do.

Of the series of processes, the black generation / undercolor removal and the process of replacing the chromatic color with the achromatic color are the processing examples (3).
Since the amount of change in density before and after the emphasis processing is 120, this is allocated to the selected pixels as shown in FIG. 22D according to the determination signal. The processing target pixel is a black character / normal edge, and the pixel to be allocated is also a black character / normal edge. As shown in FIG. 21B, the thinly shaded area is the allocation target pixel. Become.

Since there are two pixels to be allocated, "1"
20/2 = 60 ″ is added to these two pixels. If there is no pixel having a region determination to be allocated to the peripheral pixels, as shown in FIG. 22B is not performed, and the difference value is not propagated to the next block at that time.The calculation results of the above processing are shown in FIGS.

Here, the case where the processing target pixel is a black character / normal edge is considered, but the same procedure can be considered when the processing target pixel is a black character / blank edge, a color character / normal edge, and a color character / blank edge. I will omit it.

In the present embodiment, the density difference before and after the emphasizing process is equally allocated to all the peripheral pixels or to the pixels having the same area determination as the target pixel, but the invention is not limited to this. A non-uniform allocation method, such as different weighting depending on whether or not the target pixel has the same area determination as that of the target pixel, may be used.

Further, in the present embodiment, the emphasis judgment signal of the area judgment signal is not reflected for simplification of the process, but the allocation target pixel among the peripheral pixels including the emphasis judgment signal is further selected. You may.

Further, when there is no pixel to be allocated in a block, a method of propagating the density difference to be allocated to the next block or expanding the block to a wider range can be considered.

With the above processing, it is possible to apply a more appropriate emphasis filter to the character edge portion while preventing the occurrence of white spots and outlines around the character.

In the halftone generation process, in order to finally output the image, the image is divided into pixels, and a process such as an organic dither method is performed so that the gradation can be reproduced for each pixel. In the halftone generation processing, as shown in FIG. 1, regardless of the emphasis switching determination signal, by referring to the black character edge determination signal, the grayscale or resolution-oriented halftone processing is switched. By performing such processing, it is possible to perform appropriate halftone generation processing on the character area and other areas such as photographs.

Thus, the black character edge determination signal BkED
GE or color character edge determination signal ColEDGE and emphasis switching determination signal EMP are referred to, and when a suitable black generation / undercolor removal processing / spatial filter processing / halftone generation processing is performed, a series of processing ends, and image output The device outputs an image.

In the present embodiment, a series of flows are described by using edges as character edges. However, it is of course possible to process the edges as broad edges including continuous tone diagrams and photograph outlines.

As described above, in the image processing method according to the present embodiment, as a result of performing the emphasis determination filter processing on the pixel of interest P, when the emphasis processing by the emphasis filter is performed, the color generated around the black character or the color character is generated. In such a case, it is determined that the emphasis processing by the emphasis filter is not performed in consideration of the possibility that the omission, the false contour, and the color blur may occur. In this case, since the emphasis processing is not performed by the emphasis filter, the character cannot be emphasized.

Therefore, in this embodiment, when it is determined that the emphasis processing by the emphasis filter is not performed, the density value of the pixel of interest P and the density value when the emphasis process by the emphasis filter is performed on the pixel of interest. And the difference is distributed to the pixels around the pixel of interest P. By this method, characters and the like can be substantially emphasized with respect to surrounding pixels.

As a result, it is possible to prevent white spots and color bleeding around the black character edge caused by the highlighting process by the highlighting filter, and to prevent small blank characters from being crushed properly, and to appropriately select black letters, colored letters, blank letters, and blank letters. The edge part can be appropriately emphasized.

Therefore, a color dropout or a pseudo contour generated around a black character or a color character due to the emphasis processing by the emphasis filter,
Also, it is possible to provide an image processing method capable of preventing the occurrence of color fringing.

Further, according to the image processing method of the present embodiment, when it is determined that the target pixel P is a black character edge, the density value of the target pixel P and the target pixel P are determined for the image data representing the black component. The difference from the density value when P is subjected to the emphasis process is obtained, and this difference is distributed to the pixels around the pixel of interest P. As a result, as described above, the black character can be substantially emphasized with respect to the peripheral pixels.

Further, in this embodiment, at this time, the average value of the color components other than the black component in the input image data is obtained and the density value of the pixel of interest P is replaced with an achromatic color. here,
The reason why the average value is replaced is to eliminate the density difference between the respective color components such as C, M, and Y, and bring it closer to an achromatic color (black). As a result, black characters can be emphasized while preventing color fringing around the characters.

Therefore, when it is determined that the pixel of interest P is a black character edge as a result of the determination, even if the enhancement process is not performed by the enhancement filter, the effect of the enhancement process and the image quality deterioration due to the enhancement process occur. Factors can be reduced at the same time.

Further, according to the image processing method of the present embodiment, when it is determined that the edge of the extracted character, the density value of the pixel of interest P and the pixel of interest P are subjected to the emphasis process by the emphasis filter. The difference with the density value is distributed by subtracting it from the density values of the peripheral pixels in the target pixel P.

As described above, at the edge of the blank character, the effect of the enhancement filter at the edge is always exerted in the decreasing direction with respect to the density of the peripheral pixels, so that the density of the edge is higher than that of the peripheral background pixels (high density). It is possible to prevent the characters from being crushed due to an excessively higher value than the background pixels) and to substantially enhance the extracted characters.

On the other hand, when it is determined that the edge is not the edge of the omitted character, the difference is distributed by adding it to the density values of the peripheral pixels of the target pixel P.

As described above, when it is not the edge of the blank character, that is, in the case of the normal edge region, the effect of the emphasis filter at the edge is always exerted on the peripheral pixels in the increasing direction, so that the edge density is increased. It is possible to prevent white spots and color fringing from occurring due to the density being lower than the density of background pixels in the vicinity (background pixels having a low density), and substantially normal edges can be emphasized.

Therefore, by performing the processing suitable for each of the processing method in the normal edge area where white spots and color bleeding may occur and the void character where the character collapse may occur. , It is possible to further minimize the influence of the image quality deterioration caused by the enhancement filter.

As a result, there is provided an image processing method capable of preventing the occurrence of color dropouts, pseudo contours, and color bleeding around black characters and color characters, which are caused by the emphasis processing by the emphasis filter, and prevent the collapse of the omitted characters. Can be provided.

The image processing apparatus of the present embodiment extracts a block composed of a plurality of pixels including the target pixel P from the input image data composed of a plurality of color components, and determines whether or not the target pixel P is a character edge. The character edge detection / color determination unit 20 and the feature amount calculation / determination unit 30 that perform the determination, the target pixel P are subjected to the emphasis determination filter processing, and the target pixel P is subjected to the emphasis processing by the emphasis filter based on the result. The character edge determination unit 10 includes an emphasis process switching determination unit 40 that determines whether or not to perform the processing.

Then, as a result of the emphasis determination filter processing being performed on the target pixel P by the emphasis processing switching judgment unit 40,
When it is determined that the emphasis processing by the emphasis filter is not performed, the difference between the density value of the pixel of interest P and the density value when the emphasis process by the emphasis filter is performed on the pixel of interest P is calculated, and this difference is noted. A spatial filter unit 7 that distributes to pixels around the pixel P is provided.

Therefore, it is possible to prevent white spots and color bleeding around the black character edge caused by the highlighting process by the highlighting filter, and to prevent small blank characters from being crushed, and to appropriately select black letters, colored letters, blank letters, and blank letters. The edge part can be appropriately emphasized.

Further, in the image processing apparatus of this embodiment,
The character edge determination unit 10 is provided with an extraction edge determination unit 50 that determines whether or not the pixel of interest P is an edge of an extraction character.

That is, in the normal edge, the target pixel P often has a higher density than the peripheral pixels, and when the enhancement processing is normally performed by the enhancement filter, the peripheral color loss and the color blurring are caused, while the extracted edge is eliminated. In many cases, the target pixel P has a lower density than the peripheral pixels, and the problem of collapsing in small letters is likely to occur.

In this respect, according to the present embodiment, the blank edge determining unit 50 provided in the character edge determining unit 10 determines whether or not the pixel of interest P is the edge of the blank character.

Therefore, by separating the character edge into a normal edge and a blank character, each edge portion can be appropriately emphasized.

The digital color copying machine of the present embodiment is equipped with the above image processing device. Therefore, it prevents white spots and color bleeding around the black character edge due to the emphasizing process by the emphasizing filter, and prevents the collapse of a small blank character
By classifying black characters, colored characters, blank characters, and blank characters, respective edge portions can be appropriately emphasized, and an image forming apparatus capable of outputting a high quality image is provided. You can

[0140]

As described above, according to the image processing method of the present invention, when it is determined that the enhancement processing by the enhancement filter is not performed as a result of performing the enhancement determination filter processing on the attention pixel, the density value of the attention pixel is Is obtained by calculating a difference between the target pixel and the density value when the enhancement process is performed by the enhancement filter, and distributing the difference to pixels around the target pixel.

That is, in order to prevent the occurrence of color dropouts, pseudo contours, and color fringing around black characters and color characters that accompany the emphasis processing by the emphasis filter, when the emphasis processing by the emphasis filter is not performed, , Characters cannot be emphasized.

Therefore, in the present invention, when it is determined that the enhancement processing by the enhancement filter is not performed, the difference between the density value of the target pixel and the density value when the enhancement processing by the enhancement filter is performed on the target pixel. And the difference is distributed to pixels around the pixel of interest. By this method, characters and the like can be substantially emphasized with respect to surrounding pixels.

Therefore, there are missing colors and pseudo contours around black characters and colored characters that accompany the emphasizing processing by the emphasizing filter.
Further, it is possible to provide an image processing method capable of preventing the occurrence of color fringing.

Further, in the image processing method of the present invention, in the above image processing method, when it is determined that the pixel of interest is a black character edge, an average value of color components other than the black component in the input image data is obtained. And replaces the density value of the pixel of interest with an achromatic color, and for image data representing the black component, obtains the difference between the density value of the pixel of interest and the density value of when the pixel of interest is emphasized, and the difference is calculated. Is distributed to pixels around the pixel of interest.

Therefore, the difference between the density value of the target pixel and the density value when the target pixel is emphasized is distributed to the pixels around the target pixel. As a result, as described above,
In effect, black characters can be emphasized with respect to surrounding pixels.

Further, in the present invention, at this time, the average value of the color components other than the black component in the input image data is obtained and the density value of the pixel of interest is replaced with an achromatic color. The reason why the average value is replaced is to eliminate the density difference between the respective color components such as C, M, and Y, and bring it closer to an achromatic color (black). As a result, black characters can be emphasized while preventing color fringing around the characters.

Therefore, when it is determined that the pixel of interest is a black character edge as a result of the determination, even if the enhancement processing is not performed by the enhancement filter, the effect of the enhancement processing and the factor of image quality deterioration due to the enhancement processing are caused. The effect that both reductions can be achieved is achieved.

Further, in the image processing method of the present invention, in the above image processing method, when it is determined that the enhancement processing by the enhancement filter is not performed as a result of performing the enhancement determination filter processing on the target pixel, the attention processing is performed. It is determined whether or not the pixel is the edge of the extracted character. When it is determined that the pixel is the edge of the extracted character, the density value of the pixel of interest and the density when the emphasis process is performed on the pixel of interest by the emphasis filter. The difference from the value is distributed by subtracting it from the density values of the peripheral pixels of the pixel of interest, while the difference is added to the density values of the peripheral pixels of the pixel of interest when it is determined that it is not the edge of the extracted character. It is a method of allocating by doing so.

Therefore, by making the edge of the extracted character with respect to the peripheral pixels, the effect of the emphasis filter at the edge always works in the decreasing direction with respect to the density of the peripheral pixels, so that the edge density is It is possible to prevent the characters from being crushed due to an excessively higher density than the background density (high density background pixel) and to substantially enhance the blank characters. On the other hand, when the edge is not the outline character, that is, in the case of the normal edge region, the normal edge is applied to the peripheral pixels, and the effect of the enhancement filter at the edge is always applied to the peripheral pixels in the increasing direction. , To prevent white spots and color fringing from occurring due to the edge density becoming lower than the density of surrounding background pixels (background pixels with low density), and to substantially enhance the normal edge. You can

Therefore, by performing the processing suitable for each of the processing method in the normal edge area where white spots and color bleeding may occur and the blank character where the character collapse may occur. Thus, the effect of further reducing the image quality due to the enhancement filter can be minimized.

The image processing apparatus of the present invention, as described above,
As a result of performing the emphasis determination filter processing on the target pixel by the emphasis processing switching determination unit, when it is determined that the emphasis processing by the emphasis filter is not performed, the density value of the target pixel and the emphasis processing by the emphasis filter by the emphasis filter are performed. Spatial filter means is provided for obtaining a difference from the density value in the case of applying the difference and distributing the difference to pixels around the pixel of interest.

Therefore, it is possible to prevent black spots and color bleeding around the black character edge due to the highlighting processing by the highlighting filter, and to prevent small blank characters from being crushed properly, and to appropriately fill black characters, colored characters, blank characters, and blank characters, respectively. This has the effect of being able to appropriately emphasize the edge portion of.

Further, in the image processing apparatus of the present invention, in the above-mentioned image processing apparatus, the character edge determination means is provided with an extraction edge determination section for determining whether or not the pixel of interest is an edge of an extraction character. It is what has been.

Therefore, by dividing the character edge into a normal edge and a blank character, each edge portion can be appropriately emphasized.

The image forming apparatus of the present invention, as described above,
The image processing apparatus described above is provided.

Therefore, it is possible to prevent white spots and color bleeding around the black character edge caused by the highlighting process by the highlighting filter, and to prevent the small blank characters from being crushed.
By dividing the case into white characters, the respective edge portions can be appropriately emphasized, so that it is possible to provide an image forming apparatus capable of outputting a high-quality image.

[Brief description of drawings]

FIG. 1 is a diagram showing an embodiment of an image processing method of an image processing apparatus provided in an image forming apparatus of the present invention, and an image is obtained by referring to a black character / color character edge determination signal and an emphasis switching determination signal. It is a flow chart which shows processing at the time of performing processing.

FIG. 2 is a block diagram showing an overall configuration diagram of the image processing apparatus.

FIG. 3 is a block diagram showing a schematic configuration of a character edge determination unit of the image processing apparatus.

FIG. 4 is a diagram showing a 5 × 5 mask centered on a pixel of interest, which is used in the processing of the character edge determination unit.

FIG. 5 is a block diagram showing a specific configuration example of a character edge detection / color determination unit that constitutes the character edge determination unit.

FIG. 6A is a diagram showing a 3 × 3 pixel mask centered on a pixel of interest, which is used in the processing of the character edge detection / color determination unit of the character edge determination unit, and FIG. It is a figure which shows an example of the main scanning direction Sobel filter used for the process of a character edge detection / color determination part, (c) is a subscanning direction used for the process of the said character edge detection / color determination part. It is a figure which shows an example of a Sobel filter.

FIG. 7: Feature amount calculation that constitutes the character edge determination unit
It is a block diagram showing a concrete example of composition of a judgment part.

FIG. 8 is used for the processing of the feature amount calculation / determination unit 2
It is a figure which shows the example of a value mask.

FIG. 9 is a block diagram showing a specific configuration example of a threshold value selection unit that constitutes the feature amount calculation / determination unit.

FIG. 10 is a block diagram showing a configuration of a cut edge determination unit in the character edge determination unit.

FIG. 11 is a block diagram showing a configuration of an emphasis processing switching determination unit in the character edge determination unit.

FIG. 12A is a schematic diagram of an edge on a low-density background, and FIG. 12B illustrates a contour generated around a character when an emphasis determination filter is applied to the entire edge of FIG. FIG. 4C is a schematic diagram for doing so, FIG. 6C is a schematic diagram of a cut edge on a high-density background, and FIG. It is a schematic diagram for demonstrating how it is crushed.

FIG. 13 is a diagram illustrating an example of an emphasis determination filter applied to an emphasis process switching determination unit in the character edge determination unit.

FIG. 14A is a diagram showing a density distribution of a character area, and FIG. 14B is a diagram showing a result of performing the emphasis determination filter processing on FIG.

15A and 15B are diagrams showing the principle of black generation and are schematic diagrams of density changes before and after black generation / undercolor removal processing.

16 (a) to (e) are diagrams showing a set of densities for explaining a specific example of the flowchart shown in FIG.

17 (a) and 17 (b) are diagrams showing a density change before and after processing at a black character edge for which emphasis processing is determined to be possible.

18A to 18D are diagrams showing density changes before and after processing at a black character edge that is determined to be incapable of emphasis processing.

FIG. 19 is a diagram showing a result of an emphasis filter process at a color character edge determined to be capable of an emphasis process.

20A to 20C are diagrams showing a density change before and after processing on a color character edge determined to be incapable of emphasis processing.

21 (a) and 21 (b) are diagrams illustrating a specific example in which a determination signal is considered for allocated pixels.

22A to 22E are diagrams showing density changes before and after processing at a black character edge when a determination signal is taken into consideration in assigned pixels.

[Explanation of symbols]

6 Black Generation / Under Color Removal Section 7 Spatial Filter Section (Spatial Filter Means) 10 Character Edge Judgment Section (Character Edge Judgment Section) 20 Character Edge Detection / Color Judgment Section (Character Edge Judgment Section, Character Edge Judgment Section) 21 Edge Detection Circuit 22 Color determination circuit 30 Feature amount calculation / determination unit (character edge determination means, character edge determination unit) 31 First feature amount calculation circuit 32 Second feature amount calculation circuit 33 Judgment circuit 33 34 AND circuit 40 Switching emphasis process Judgment unit 41 Emphasis judgment filter unit 42 Comparator 43 Adder (ADD) 44 Comparator 45 Difference device (SUB) 46 Adder (ADD) 47 Comparator 50 Extracted edge judgment unit 51 Chromatic color / achromatic judgment unit 52 Density judgment Unit 53 AND circuit 54 Adder circuit (ADD)

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) H04N 1/46 H04N 1/40 F 1/60 B41J 3/00 AF term (reference) 2C262 AB11 AB13 BA05 BA07 BA12 BA13 BB03 BB08 BB15 BC11 DA02 DA03 DA09 DA13 EA07 2H027 DB01 EA18 EB02 EC11 EC20 5B057 AA11 BA30 CA01 CA08 CA12 CA16 CB01 CB08 CB12 CB16 CE03 CE06 CE11 CE12 CH04 CH04 CH12 DA12 DA17 DB02 DB06 DB02 DB06 DB09 DB04 DB07 DB02 DB06 DB0 DB02 DB06 DB09 DB04 PP03 PP25 PP27 PP28 PP32 PP33 PP39 PP51 PP68 PQ08 PQ12 PQ24 TT06 5C079 HB01 HB03 HB12 LA02 LA03 LA06 LA15 LA21 LA24 LA31 LB12 MA03 MA11 NA02 NA07 PA02

Claims (6)

[Claims] 1. A block composed of a plurality of pixels including a target pixel is extracted from input image data composed of a plurality of color components,
Based on the pixel information in the block, it is determined whether the pixel of interest is a character edge. If it is determined that the pixel of interest is a character edge, the pixel of interest is subjected to emphasis determination filter processing, and the pixel of interest is selected based on the result. In the image processing method for determining whether or not to perform the enhancement filter processing, when it is determined that the enhancement processing by the enhancement filter is not performed as a result of performing the enhancement determination filter processing on the target pixel, the density value of the target pixel and the An image processing method, characterized in that a difference from a density value when an enhancement process is performed on the pixel of interest is calculated, and the difference is distributed to pixels around the pixel of interest. 2. When the pixel of interest is determined to be a black character edge, the average value of color components other than the black component in the input image data is calculated to replace the density value of the pixel of interest with an achromatic color, and the black component The difference between the density value of the pixel of interest and the density value when the pixel of interest is emphasized is calculated for the image data representing, and the difference is distributed to the pixels around the pixel of interest. The image processing method according to claim 1. 3. As a result of performing the emphasis determination filter processing on the target pixel, when it is determined that the emphasis processing by the emphasis filter is not performed, it is determined whether the target pixel is the edge of the extracted character. , When it is determined to be the edge of the omitted character, the difference between the density value of the target pixel and the density value when the target pixel is subjected to the enhancement processing by the enhancement filter is calculated from the density values of the peripheral pixels in the target pixel. 3. The distribution is made by subtraction, while the distribution is made by adding the difference to the density values of the peripheral pixels in the target pixel when it is determined that the edge is not the outline character. Image processing method. 4. A block composed of a plurality of pixels including a target pixel is extracted from input image data composed of a plurality of color components,
A character edge determination unit that determines whether or not the pixel of interest is a character edge, and whether or not to perform emphasis determination filter processing on the pixel of interest and whether or not to perform emphasis processing by the emphasis filter on the pixel of interest based on the result. In an image processing apparatus including a character edge determination unit that includes an enhancement processing switching determination unit that performs determination, as a result of performing the enhancement determination filtering process on the pixel of interest by the enhancement processing switching determination unit, the enhancement processing by the enhancement filter is performed. When it is determined that there is no such difference, the difference between the density value of the target pixel and the density value when the target pixel is subjected to the enhancement processing by the enhancement filter is calculated, and the difference is distributed to the pixels around the target pixel. An image processing apparatus comprising means. 5. The image processing according to claim 4, wherein the character edge determination means includes a cut edge determination unit that determines whether or not the pixel of interest is an edge of a cut character. apparatus. 6. An image forming apparatus comprising the image processing apparatus according to claim 4.