JP2001169101A - Image discriminating device - Google Patents

Abstract

(57) [Summary] [Problem] To accurately determine whether an input image is a monochrome image or a color image. An image input by an image input unit is separated into a color region pixel, an intermediate region pixel, and a monochrome region pixel by an image region determination unit. The monochrome area pixel 28 is separated into a monochrome area pixel 30 and a base area pixel 31 by a monochrome area determination unit 29. The monochrome area pixel 30 is recognized as a monochrome block by the monochrome block recognition means 32. Pixels at the border of the monochrome block are
Whether or not a color shift has occurred is determined by the attribute determining unit 38, and the determination result is reflected in correction of the count value of the pixel counting unit 33 by the pixel counting correcting unit 39. The image discriminating means 40 derives a discrimination signal indicating whether the image is a monochrome image or a color image based on the counting result of the pixel counting means 33.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image discriminating apparatus for discriminating between a color image and a monochrome image when an image is formed by a color image forming apparatus.

[0002]

2. Description of the Related Art Conventionally, most image data used to form an image by copying or printing is a monochrome image called a monochrome image. However, in recent years, O
With the rapid progress of digitization of the device A and the demand for color image output, color image forming apparatuses such as digital color copying machines and color printers have been widely and widely used.

[0003] In these color image forming apparatuses,
It is also possible in principle to form a black and white image by overlapping a plurality of colors. In black and white images, images that need to have sharp outlines, such as characters, and shades are often expressed with halftone dots,
If a color recording material is to be formed as a black and white image by superimposing a plurality of colors, black characters and thin lines do not become halftone dots due to color misregistration and spectral characteristics of the recording material. Also, considering the process of processing a color image assuming the case of forming a black and white image, in addition to the problem that the quality of the formed image is deteriorated as described above, in spite of being a black and white image, There is also a problem that the processing is time-consuming because the processing is performed as a plurality of components separated by color, and the cost is increased because an image is formed by a plurality of processes.

When forming a color image, an image is formed by converting a plurality of color components, for example, three primary colors of red (R), green (G), and blue (B), cyan (C), magenta (M), and yellow (Y). ) Are processed separately. If the densities of the respective color components are the same, they should be achromatic and become black and white components. However, the above-described problem occurs. For this reason, the input image is a color image containing color information,
It is necessary to determine whether the image is a black-and-white image consisting only of black-and-white information, and appropriately select a process for processing a color image for a color image and a process for processing a black-and-white image for a black-and-white image according to the type of image. is there.

[0005] As a prior art for discriminating such image types, there are, for example, Japanese Patent Application Laid-Open Nos. 3-270380 and 4-282968. JP-A-3-27
Japanese Patent Application Publication No. 0380 discloses a technique of outputting a color image determination signal as a result of comparing the number of times of output of a color determination signal for each pixel with a threshold value. In this prior art, each pixel is determined to be a color pixel or a monochrome pixel, the number of color pixels in the image is counted, and if the count value reaches a certain number or more, the image is displayed. It is determined that the image is a color image, and otherwise, it is determined that the image is a monochrome image.

Japanese Patent Application Laid-Open No. 4-282968 discloses a technique for discriminating a color image based on the number of color lines as a discrimination result for each line forming an image. In this prior art, it is determined whether each pixel is a color pixel or a monochrome pixel, and when the presence of color pixels continuous for a predetermined number or more in the order of input pixels is detected, the number of the continuous color pixels is determined. It is recognized as a color block, and if a predetermined number or more of color blocks exist in one line, the line is counted as a color line.
If a predetermined number of color lines exist in the image, the image is determined to be a color image, and if not, it is determined to be a monochrome image.

However, in these prior arts, no consideration is given to erroneous determination in the determination for each pixel. As an erroneous determination at the time of pixel determination, a case where a pixel is actually determined to be a monochrome pixel despite being a color pixel, and conversely, although a pixel is actually a monochrome pixel,
There is a case where it is determined as a color pixel. Such an erroneous determination cannot be determined only from the influence of color noise mainly occurring in the monochrome edge portion and the density value of one pixel which is problematic in the background portion of the document or in a region called a pastel color. This is due to the way pixels are handled.

[0008] The color noise generated in the monochrome edge portion is a color shift caused by reading an image. When a color shift occurs, there is a possibility that a pixel belonging to the monochrome area is erroneously determined as belonging to the color area. Accordingly, the applicant of the present application has disclosed in
Japanese Patent Publication No. 196282 proposes an image discriminating apparatus capable of solving the above-mentioned problem.

FIG. 16 shows a configuration of an image discriminating apparatus proposed in Japanese Patent Application Laid-Open No. H11-196282. The image discriminating apparatus discriminates an image as follows.

Based on image data read by the image input means 1 and decomposed into a plurality of colors, mutual difference values of the color components constituting the image data are calculated. It is determined whether the image belongs to an area, a black and white area, or an intermediate area that cannot be said to be either a color area or a black and white area.

Here, the pixels determined to be in the black and white area are separated into a base area and a monochrome area by the black and white area determination means (pixel determination means 5) provided in the pixel area determination means 2. The pixel area determination means 2 includes a difference value calculation means 3, a maximum difference value selection means 4, and a pixel determination means 5. The pixel discriminating means 5 classifies the pixel into a color area pixel 6, an intermediate area pixel 7, and a monochrome area pixel 8 based on the discrimination result.

As for the monochrome area pixels 8 which are determined to be monochrome areas, the monochrome area pixels 8 are continuously present, and when the number of continuous pixels is equal to or greater than the threshold value Border, the monochrome block recognition means 9 determines that the area is a monochrome block. It is recognized that there is. Note that the color area pixel 6,
The numbers of the intermediate area pixels 7 and the monochrome area pixels 8 are counted by the color pixel counter 11, the intermediate pixel counter 12, and the monochrome pixel counter 13 of the pixel counting means 10, respectively. The number of pixels in the monochrome block recognized by the monochrome block recognition means 9 is counted by the monochrome block counter 14.

When a monochrome block is detected: 1) If pixels located at both ends of the monochrome block belong to the color area or the intermediate area, the number of pixels is subtracted from the count value of the color pixel counter 11 or the intermediate pixel counter 12. . That is, if a color pixel or an intermediate pixel occurs independently at both ends of the monochrome block, the pixel is regarded as color noise. Therefore, the color pixel counters 11 correspond to the number of pixels in the color area or the intermediate area located at both ends of the monochrome block.
Alternatively, by subtracting from the count value of the intermediate pixel counter 12, color noise can be removed and corrected. 2) When the pixels at both ends of the monochrome block belong to the background area, since the color noise has not occurred, the count values of the color pixel counter 11, the intermediate pixel counter 12, and the monochrome pixel counter 13 are not changed.

In the steps described above, the color noise generated at both ends of the monochrome block can be corrected. This correction is performed by the color noise removing unit 15. The color image discriminating means 16 outputs the discrimination result of the image type, and the final output means 17 performs printing or the like.

[0015]

Problems to be Solved by the Invention Japanese Patent Laid-Open No. 11-1962
In the configuration proposed in Japanese Patent Publication No. 82, when color pixels or intermediate pixels occur at both ends of a monochrome block, even if the pixels are original color pixels,
It is always judged as color noise and removed. In other words, there is a problem that it is not sufficiently determined whether or not the pixels at both ends of the monochrome block are color shift pixels.

It is an object of the present invention to provide an image discriminating apparatus capable of accurately discriminating whether a pixel at a peripheral position of a monochrome block is a color pixel or a monochrome pixel.

[0017]

According to the present invention, there is provided an image input means for decomposing input image information into a plurality of color components for each pixel and deriving the image signal as an image signal; Based on the color component, a pixel region determining unit that determines whether each pixel belongs to a color region, an intermediate region, or a black and white region, and a pixel that is determined to belong to a black and white region by the pixel region determining unit is determined in advance. A black-and-white area determining means for determining whether a pixel belongs to a monochrome area or a background area according to a predetermined criterion, and a pixel determined to belong to a color area / black-and-white area or a monochrome area by a pixel area determining means and a monochrome area determining means. Every time it appears,
The pixel counting means for counting to the color area pixel counter / intermediate area pixel counter or the monochrome area pixel counter, and the monochrome area determination means, when a predetermined number or more of pixels determined to belong to the monochrome area are continuously present, A monochrome block recognizing unit for recognizing a continuous pixel area as a monochrome block, an attribute determining unit for determining an attribute of a pixel located at a boundary of a peripheral edge of the monochrome block recognized by the monochrome block recognizing unit, and an attribute determining unit Means for correcting the count value of the color area pixel counter / intermediate area pixel counter or the monochrome area pixel counter based on the determination result of the means; and image input means based on the count value correction result of the pixel count correction means. Image determination means for determining the type of image input to the An image determination apparatus.

According to the present invention, based on an image signal input by the image input means and decomposed into a plurality of colors, the pixel area determining means determines, for each pixel, a pixel whose pixel belongs to the color area. It is determined whether the pixel belongs to a pixel that belongs to a black-and-white area, or a pixel that belongs to an intermediate area between the color area and the black-and-white area. For pixels determined to belong to the black and white area, the black and white area determination means
It is separated into a base area and a monochrome area. Pixels belonging to each area are counted by a pixel counting means into a color area pixel counter / intermediate area pixel counter or a monochrome area pixel counter. The attribute of the pixel located at the border of the periphery of the pixel recognized as a monochrome block by the monochrome block recognition means is determined by the attribute determination means. According to the determination result of the attribute determination unit, the pixel count correction unit corrects the count value of the color area pixel counter / intermediate area pixel counter or the monochrome area pixel counter. Since the image discriminating unit discriminates the type of the image based on the corrected count value, it is possible to accurately discriminate between a color image and a monochrome image by quantitative discrimination.

In the present invention, the image determining means determines the number of pixels determined to belong to a color area / intermediate area or a monochrome area for all pixels of each line of the image input to the image input means. Counting, determining whether the line is a color line including a color area pixel, counting the number of color lines, and determining the type of image based on the counted number of color lines. .

According to the present invention, the determination of the image area is performed for each line forming the image, so that only one buffer is required for the determination processing, and the circuit scale can be reduced. .

In the present invention, the image determining means counts the number of pixels determined to belong to a color area / intermediate area or a monochrome area for all pixels of the image input to the image input means, The type of image is determined based on the counted number of pixels in each area.

According to the present invention, since the determination of the image area is performed for all the pixels of the input image, the number of processing steps can be reduced as compared with the case where the determination is performed for each line.

In the present invention, when the pixel located at the boundary of the peripheral edge of the monochrome block is determined to be a color shift pixel by the attribute determining means, the pixel count correction means belongs to the monochrome area. It is characterized in that the count value is corrected as a pixel.

According to the present invention, if a pixel located at the boundary of the peripheral edge of the monochrome block is determined to be a color shift pixel, it is formed as a pixel belonging to the monochrome area. Without being affected by the influence, the possibility of erroneous determination as a color area pixel is reduced, and the image can be accurately determined.

In the present invention, the attribute discriminating means sets a plurality of color components of the pixel as (XYZ) in density of three primary colors, sets first and second predetermined thresholds as a and b, respectively, and a> b │XY│> a and │YZ│ <b or │YZ│> a and │ZX│ <b or │ZX│> a and │XY The pixel is determined to be a color misregistration pixel on the condition that one of the inequalities of | <b is satisfied.

According to the present invention, a color shift pixel is determined based on the absolute value of the difference between the color components of the pixel. By calculating the absolute value of the difference value by the pixel area determination means, the calculation result can be used for determination of color misregistration. A color shift is a shift in one of a plurality of color components constituting a color pixel. Basically, a position shift of one color component with respect to another color component occurs. is there. Therefore, by using the condition that one of the absolute values of the difference values of the color components is large and the other combination is small, it is possible to reliably determine whether or not the pixel is a color shift pixel.

[0027]

FIG. 1 shows a schematic configuration of an image discriminating apparatus 20 according to an embodiment of the present invention. In the present embodiment, Y, M and C
It is assumed that color image data having the following three color components is read. The image input means 21 is, for example, CC
D (Charge Coupled Device) etc.
This is a color imaging unit using a color filter or the like, and is realized by a color scanner, a color video camera, or the like. The image read by the image input means 21 is
Three color components are given in the scanning direction as image data in pixel units.

The image input means 21 such as a CCD reads a reflected light image as an RGB three primary color analog signal from a document as a source of an image to be read. When the YMC signal is derived as image data from the image input means 21, RG
The B signal is inverted. In the following description, the image data is YM
It is assumed that the signal is a C signal.
Of course, the present invention can be applied to signals as well.

The image data derived from the image input means 21 is subjected to area discrimination by the pixel area discriminating means 22 for each pixel. The pixel area determination unit 22 includes a difference value calculation unit 23, a maximum difference value selection unit 24, and a pixel determination unit 25. The difference value calculation means 23 calculates the mutual difference values YM, MC, and CY between Y, M, and C as the color components constituting the image data for each pixel, and calculates the absolute values. Is calculated. Difference maximum value selection means 24
Are three combinations of the absolute values of the mutual difference values calculated by the difference value calculation means 23 ｛| Y−M |, | M−C |, |
The combination that takes the maximum value of C−Y |
M│, │MC│, │CY│｝ max.

The pixel discriminating means 25 determines predetermined threshold values α, β (α <β, α> 0, β> 0) and ｛| Y
−M│, │MC│, │CY│｝ max, and ｛│YM│, │MC│, │CY│｝ max <
If α, the pixel is determined to be a pixel belonging to the black-and-white area, and || Y−M |, | M−C |, | C−Y |
When ax> β, the pixel is determined to be a pixel belonging to the color area. Also, α ≦ ｛| Y−M |, | MC
When |, | CY |｝ max ≦ β, it is determined that the pixel belongs to an intermediate area that cannot be said to be either a black-and-white area or a color pixel. In this way, the pixel discriminating means 25 discriminates pixels belonging to the color area as color area pixels 26, discriminates pixels belonging to the intermediate area as intermediate area pixels 27, and pixels belonging to the black and white area as black and white area pixels 28. Determine. Here, the aforementioned threshold values α, β
Is, for example, α = 20 and β = 30.

The monochrome area pixel 28 is further separated into a monochrome area pixel 30 and a base area pixel 31 by a monochrome area determination means 29. The black-and-white area determination means 29 is
It has minimum value calculating means for calculating the minimum value of the signal. Minimum value {YMC} min calculated by the minimum value calculation means
Is compared with the background threshold value ω, and an area where {YMC} min <ω is determined to be a background area. The base region is a region in which white pixels or pixels of light chromatic color are formed and are used as a base. The threshold value ω is a value such as ω = 20 or ω = 30.

FIG. 2 shows an example of the setting state of the pixel area when the image data of each color is represented by 8 bits. α and β
The monochrome region, the intermediate region, and the color region are set based on the threshold value of, and the number of pixels in each region is obtained.

As shown in FIG. 3, the monochrome block 30 in FIG. 1 is recognized by the monochrome block recognizing means 32 as a monochrome block in a portion that occurs continuously. The number of continuous pixels in the monochrome area is x, and a predetermined positive integer Border is a threshold. x and B
The order is compared, and if x ≧ Border, this continuous part is recognized as a monochrome block. Threshold B
The order is, for example, 10.

The pixel counting means 33 shown in FIG. 1 calculates the color area pixel 26, the intermediate area pixel 27 or the monochrome area pixel 30.
Each time any of the above appears, the color pixel counter 3
4. The number of occurrences is counted by the intermediate pixel counter 35 or the monochrome pixel counter 36, respectively. Pixels belonging to a monochrome block recognized by the monochrome block recognition means 32 are counted by the monochrome block counter 37.

The monochrome block recognizing means 32 recognizes an area to which the pixel at the start portion of the monochrome block and the pixel at the end portion of the monochrome block belong. Attribute determination means 3
Reference numeral 8 identifies the attribute of the pixel at the start and end of the monochrome block recognized by the monochrome block recognition means 32 as to whether the pixel is color noise or an original color pixel that is not color noise. The pixel count correction unit 39 corrects the count value of the pixel count unit 33 based on the determination result of the attribute determination unit 38. After correcting the count value by the pixel count correction unit 39, the image determination unit 40 converts the image data read by the image input unit 21 based on the count values of the color pixel counter 34, the intermediate pixel counter 35, and the monochrome pixel counter 36. It is determined whether or not it represents a color image, and a signal representing the determination result is derived.

The pixel area determining means 22, the monochrome block recognizing means 32, the pixel counting means 33, the attribute determining means 38,
The pixel count correction means 39 and the image determination means 40 are realized by a program operation of a computer device. In particular, a digital signal processor can be used as the computer device.

FIG. 4 shows an overall processing procedure of the image discriminating apparatus 20 shown in FIG. Here, for the sake of simplicity, a case will be described where the pixel width of the discrimination target at both ends of the monochrome block is 1, and a total of two pixels are determined at both ends.

After the start of the procedure, in step S1, initial settings are made for the color pixel counter 34, the intermediate pixel counter 35, the monochrome pixel counter 36, the monochrome block counter 37, and the like. In step S2, the current flag referred to during the processing is cleared. In step S3, image data of one pixel is input. In step S4, the area to which the input pixel belongs is determined.
In step S5, an attribute determination process for the pixel is performed.
In step S6, it is determined whether or not the determination for all pixels has been completed. If not, the process returns to step S2. If it is determined in step S6 that the determination of all pixels has been completed, a termination determination process is performed in step S7. In step S8, an image is determined based on the count values of the color pixel counter 34, the intermediate pixel counter 35, the monochrome pixel counter 36, and the monochrome block counter 37, and the overall procedure ends.

As shown in FIG. 5, in the present embodiment, the following two types of flags are used to determine the pixels at both ends of a monochrome block. Front flag: A flag of a pixel at a position facing the current pixel to be determined with a monochrome block interposed therebetween. Current flag: a flag for the current pixel to be determined.

FIG. 6 shows a processing procedure when the initial setting of step S1 in FIG. 4 is performed as a subroutine. In step S11, the count value CO-
Set CNTR to "0". In step S12,
The count value MI-CNTR of the intermediate pixel counter 35 is set to “0”
Set to. In step S13, the count value BL-CNTR of the monochrome pixel counter 36 is set to "0". In step S14, with respect to the current determination target position,
A front color flag FC-FLAG, which is normally "off", is set to "off" as a flag for a color pixel at a position facing the monochrome block.
In step S15, a front intermediate flag FM-FLAG, which is normally "off", is set to "off" as a flag for intermediate pixels at a position facing the current determination target position across the monochrome block. deep. In step S16, a front ground flag FG-F which is normally "off" is set as a background pixel flag at a position opposed to the current determination target position across the monochrome block.
LAG is set to “off”. In step S17, when there is a pixel determined to belong to the monochrome area continuously in the reading order, the count value MO-BLK of the monochrome block counter 37 indicating the number of continuous pixels is set to “0”.
Set to.

FIG. 7 shows a processing procedure when the current flag clear performed in step S2 of FIG. 4 is performed as a subroutine. In step S21, the current color that is used as a flag for a color pixel at the current determination target position and is normally “off” and becomes “on” when it is determined that the current determination target pixel belongs to the color area The flag CC-FLAG is set to “off”. In step S22, a current intermediate flag that is normally “off” as an intermediate pixel flag at the current determination target position and that is “on” when it is determined that the current determination target pixel belongs to the intermediate region. CM-FLAG is set to “off”. In step S23, a flag for a background pixel at the current determination target position, which is normally “off”, and becomes “on” when it is determined that the current determination target pixel belongs to the background region. Flag CG-FL
AG is set to “off”.

FIG. 8 shows a processing procedure when the pixel determination in step S4 in FIG. 4 is performed as a subroutine. In step S31, it is determined whether or not the maximum value of the absolute value of the mutual difference between the color components is smaller than a first threshold value α. When it is determined that the maximum value of the absolute values of the mutual difference values is smaller than α, it is determined in step S32 whether or not the minimum value of YMC is smaller than a threshold value ω for determining the background area.
If it is determined that it is smaller than the threshold value ω, step S3
In step 3, the current background flag CG-FLAG is set to "ON".

If it is determined in step S32 that the minimum value of YMC is not smaller than the threshold value ω, the process proceeds to step S32.
In step S34, the count value BL-CNTR of the monochrome pixel counter is increased by one, and in step S35, the count value MO-BLK of the monochrome block counter 37 is increased by one.

When it is determined in step S31 that the maximum value among the absolute values of the mutual difference values of YMC is not smaller than the threshold value α, it is determined in step S36 whether the maximum value of the absolute value is larger than β. Judge. When the maximum value of the absolute value is larger than β, the count value CO-CNTR of the color pixel counter 34 is increased by 1 in step S37. In step S38, the current color flag CC-
Set FLAG to “ON”. When it is determined in step S36 that the maximum value of the absolute value is not larger than the threshold value β, the count value MI-CNTR of the intermediate pixel counter 35 is increased by 1 in step S39. In step S40, the current intermediate flag CM-FLAG is set to "ON". When the procedure of step S33, step S35, step S38 or step S40 ends, the subroutine ends and the program returns to the original program.

In the pixel discrimination shown in FIG. 8, as the input image data, a combination in which the absolute value of the mutual difference between Y, M, and C is maximized is ｛| Y−M |, | M−C |, │CY
The value of |｝ max is compared with the two pixel discrimination thresholds α and β to determine whether the pixel belongs to a color area, an intermediate area, or a monochrome area. Pixels belonging to the black-and-white region are compared with a threshold value ω for background determination to determine whether the pixel belongs to a background region that can be regarded as white or a monochrome region that can be regarded as black. It is determined whether there is. If it is determined that the pixel belongs to the color area or the intermediate area, the corresponding color pixel counter 34 or the intermediate pixel counter 35 is incremented by one, and the current color flag CC-FLAG or the current intermediate flag CM is increased.
-Turn FLAG "ON" respectively. If it is determined that the pixel belongs to the monochrome area, the monochrome pixel counter 36 is increased by one, and the monochrome block counter 37 is also increased by one. If the pixel belongs to the background area, the current background flag CG-FLAG is turned on.

FIG. 9 shows a procedure for processing the attribute discrimination process performed in step S5 of FIG. 4 as a subroutine.
In step S51, the current color flag CC-FL
It is determined whether any of the AG, the current intermediate flag CM-FLAG, and the current base flag CG-FLAG is “ON”. When it is determined that either of them is "ON", in step S52, it is determined whether or not the count value MO-BLK of the monochrome block counter 37 is equal to or greater than a preset monochrome block threshold value BORDER. When the condition is satisfied, a flag check is performed in step S53. After the flag check in step S53 or when the condition is not satisfied in step S52,
In step S54, the current color flag CC-FLAG
Is set in the front color flag FC-FLAG. In step S55, the current background flag CG-
The state of FLAG is set to the front base flag FG-FLAG. In step S57, the count value MO-BLK of the monochrome block counter 37 is set to "0". After the end of step S57, or when the condition is not satisfied in step S51, the attribute determination processing ends.

In the attribute determination process shown in FIG. 9, the current color flag CC-FLAG and the current intermediate flag CM-FLAG
When either the FLAG or the current background flag CG-FLAG is “ON”, the count value MO-BLK of the monochrome block counter 37 is compared with the monochrome block threshold value BORDER, and the count value MO− of the monochrome block counter is determined. BLK is monochrome block threshold BO
If it is equal to or greater than RDER, it is determined that color noise may have occurred at both ends of the monochrome block, and attribute determination of the pixels at both ends is performed. The count value MO-BLK of the monochrome block counter is the monochrome block threshold value BOR.
If the value is below DER, no attribute determination is performed. This is because a continuous monochrome area having no more than a certain size may be an isolated point or simply noise, so that the attribute discrimination processing is applied only to a monochrome block having a certain size or more. is there.

FIG. 10 shows a processing procedure when the flag check in step S53 of FIG. 9 is performed as a subroutine. In step S61, the front color flag FC
Determine whether FLAG is "on"; If it is "ON", a color misregistration determination process is performed in step S62. In step S61, the front color flag FC-
If it is determined that the FLAG is not “ON”, it is determined in a step S63 whether or not the front intermediate flag FM-FLAG is “ON”. If it is "ON", a color misregistration determination process is performed in step S62. If it is determined in step S63 that the front intermediate flag FLAG is not “ON”, the process proceeds to step S64. After the end of the color misregistration determination processing in step S62, the process proceeds to step S64.

In step S64, it is determined whether the current color flag CC-FLAG is "ON".
If it is "ON", a color misregistration determination process is performed in step S65. Note that the current background flag CG-FLA
G and front ground flag FG-FLAG are "ON"
Even in the case of, color noise does not exist in the background portion of the image for convenience, and is ignored. In step S64,
When it is determined that the current color flag CC-FLAG is not “on”, in a step S66, it is determined whether or not the current intermediate flag CM-FLAG is “on”. If it is "ON", a color misregistration determination process is performed in step S65. After the end of the color misregistration determination processing in step S64, or when it is determined in step S66 that the current intermediate flag CM-FLAG is "OFF", or when the color misregistration determination processing in step S65 is completed, the entire procedure is also performed. finish.

The color misregistration determination is performed to determine whether a color pixel or an intermediate pixel to be determined is color noise. In this determination, it is determined according to the following first to third equations whether or not the target pixel has a color shift. │YM│> a and │MC│ <b ... (1) │MC│> a and │CY│ <b ... (2) │CY│> a and │YM │ <b… (3)

However, in the first to third equations, a> b
> 0, which is a predetermined threshold based on various images. The first to third expressions are based on the concept described below. That is, when a monochrome pixel is formed by additive color mixture, the color noise due to color misregistration is such that the component of one of the YMC colors is displaced from the other color components and protrudes to the edge of the monochrome block. This is due to the occurrence of color pixels. Therefore, the color noise pixel has a feature that the value of the protruding color is larger than that of the other colors, and the values of the other color components are substantially equal. Note that, for a pixel generated at the end of the monochrome block, if the image is shifted so as to protrude to the downstream side of the YMC, only the value of the protruding color becomes large color noise. When one color of YMC is shifted to the upstream side, only the shifted color becomes small color noise. That is, in the first to third expressions, color noise is detected when two of the three colors of YMC have substantially the same value and the other color has a value far apart. . Therefore, when any of the first to third expressions is satisfied, the target pixel is determined to be color noise due to color shift,
It is counted as a monochrome area.

FIG. 11 shows the internal structure of the attribute judging means 38 of FIG. 1 for judging color misregistration. Color noise detection circuits 41, 42, and 43 are provided in the attribute determination means 38. The color noise detection circuits 41, 42, and 43 supply | YM |, | MC |, | C-
Of the three signal outputs Y |, two combinations are respectively input. Each of the color noise detection circuits 41, 42,
43, two comparators 41a, 41b;
2b; 43a and 43b are provided;
Are compared with threshold values a and b, respectively. Two comparators 41a, 41b; 42a, 42b; 43a, 43
The output of b is an AND circuit 41c, 4 having two inputs.
2c and 43c. Three AND circuits 41c,
The outputs from 42c and 43c are input to a three-input OR circuit 44, from which a color misregistration determination signal is derived. That is, in the color noise detection circuit 41, the comparator 41a outputs an “ON” signal when | Y−M |> a, and the comparator 41b outputs “ON” when | MC− <b. Output a signal. The AND circuit 41c includes a comparator 41
a and the comparator 41b output an "ON" signal when both output "ON" signals, that is, when | YM |> a and | YM | <b. The color misregistration determination signals output from the OR circuit 44 are output from the color noise detection circuits 41 to 41.
43 is output as a valid signal “1” when at least one of the signals 43 is “ON”, and the color noise detection circuit 4
When all of the outputs 1 to 43 are “off”, they are output as the invalid signal “0”.

FIG. 12 shows a processing procedure when the color misregistration determination processing performed in step S62 or S65 in FIG. 10 is performed as a subroutine. In step S71 of the color misregistration determination processing, it is determined whether or not the color misregistration determination signal from the attribute determination unit 38 in FIG. 11 is valid. If the color misregistration determination signal is valid, even if the target pixel whose attribute has been determined is color noise, it is actually regarded as a monochrome pixel. Therefore, in step S72, the count value BL-CNTR of the monochrome pixel counter 36 is determined. Is increased by one. In step S73, the front color flag FC-
It is determined whether FLAG or the current color flag CC-FLAG is “ON”. Either is "on"
If so, in step S74, the count value CO-CNTR of the color pixel counter 34 is decreased by one. Step S
If it is determined in 73 that none of the flags is "ON", the count value MI-CNTR of the intermediate pixel counter 35 is decreased by 1 in step S75. Step S
When it is determined at 71 that the color misregistration determination signal is not valid,
Alternatively, when the procedure of step S74 or step S75 ends, the process returns to the original routine.

FIG. 13 shows a processing procedure in the case where the termination determination processing in step S7 in FIG. 4 is performed as a subroutine.
The end determination process is a process of performing attribute determination on a pixel which still remains as a monochrome block. In step S81, the count value M of the monochrome block counter 37 is set.
It is determined whether or not O-BLK is equal to or greater than the monochrome block threshold BORDER. When the condition is satisfied, a flag check is performed in step S82 according to a procedure as shown in FIG. When step S82 ends,
Alternatively, when the condition is not satisfied in step S81, the termination determination processing ends.

FIG. 14 shows the concept of termination processing in the processing procedure shown in FIG. After the monochrome area in which some pixels continue, if the processing of the image data ends as the pixel belonging to the monochrome area as the pixel at the end as well, the pixels still remaining in the continuous monochrome area In other words, it is not determined whether or not to perform attribute determination. The determination for this part is performed as the termination determination shown in FIG. This process is performed by counting the count value MO- of the monochrome block counter 37 at the end point.
BLK is compared with a monochrome block threshold value BORDER to determine whether or not to perform attribute determination. If so, attribute determination processing is performed based on the state of each front flag. In this case, since it is the end, the state of each current flag can be ignored.

The image discriminating means 40 shown in FIG. 1 discriminates an image based on the count values of the color pixel counter 34, the intermediate pixel counter 35, and the monochrome pixel counter 36 which are finally obtained as a result of the above-described discrimination.

FIG. 15 shows the internal structure of the image discriminating means 40. The image determination means 40 includes a determination unit 45, a color line counter 46, and a comparator 47. The image discrimination by the image discriminating means 40 is performed based on the count value from each counter input to the judgment unit 45. The determination process is performed for each line, and the determination unit 45 receives a count value for each line.

In the judgment section 45, the color pixel counter 34
Is compared with a predetermined threshold value. If the count value is equal to or greater than the threshold value, the line is determined to be a color line, and the color line counter 46 is incremented by one. When the count value of the color pixel counter 34 is smaller than the threshold value, the count value of the monochrome pixel counter 36 is compared with a predetermined threshold value. Is determined as a monochrome line. When neither condition is satisfied, the above-described determination result is corrected in consideration of the count value of the intermediate pixel counter 35. For example, an intermediate pixel counter 35 such as 1/2 or 1/3 which is weighted with the count value of the color pixel counter 34
Is compared with a predetermined threshold value, and if the total value exceeds the threshold value, the line is determined as a color line, and the color line counter is incremented by one.
If the total value does not exceed the threshold, the line is determined to be a monochrome line.

After all lines have been determined for the image input to the determination unit 45, the color line counter 4
6 count value is compared in a comparator 47 with a predetermined threshold value TH _CL, and outputs a color image or a monochrome image or a discrimination signal. In the case where the determination is not made for each line as described above and the determination is made based on the total number of pixels of the input image, the image determination means 40 is composed of only the determination unit 45 and the result of the determination unit 45 May be output based on the determination.

In this embodiment described above, three components of Y, M, and C are used as input color components. However, actually, not only these components but also three components of R, G, and B are used. The same is true for In the present embodiment, the target pixel widths for attribute determination at both ends of the monochrome block are each set to 1
Although pixels are used and pixels at both ends are used, two pixels at both ends are not actually specified. In an image in which color noise is predicted to occur over a wide range, the target width for attribute determination is 2 pixels, and four pixels are used at both ends. It is also possible to use According to the image discrimination as in the present embodiment, it is possible to discriminate color noise occurring near the monochrome edge, which is a problem at the time of performing the image discrimination, so that more accurate image discrimination can be performed. At the same time, instead of using a buffer for storing image data when removing color noise, the determination can be made by increasing or decreasing the pixel counter, so that the processing circuit can be made compact. In addition, since it does not require time to access many image data many times, the processing can be speeded up.

[0061]

As described above, according to the present invention, it is quantitatively determined whether a pixel located at the boundary of the peripheral portion of a monochrome block belongs to a color area, an intermediate area, or a color shift pixel. Therefore, a color image and a monochrome image can be accurately determined.

According to the present invention, since the image is determined for each line, the capacity of a buffer and the like required for the determination processing can be reduced, and the circuit scale can be reduced.

Further, according to the present invention, the process of determining the image area is performed for all the pixels of the read image, so that the number of processing steps can be reduced as compared with the case of performing the determination for each line. Can be.

A color pixel is usually formed by superimposing a plurality of color components RGB or CMYK. Depending on the input accuracy of the image input means and mechanical fluctuations,
A color shift may occur in any of the color components constituting the color pixel, and the color pixel may be identified as a pixel in the color area even though the pixel is originally in the monochrome area. When a color area pixel or an intermediate area pixel and a monochrome block in which a predetermined number or more of monochrome area pixels are adjacent to each other, the pixel at the boundary of the peripheral edge of the monochrome block is likely to be a color area pixel, and Pixel may be erroneously determined as a color area pixel due to color misregistration. According to the present invention, the attribute of such a pixel is determined, and the pixel determined as a color misregistration pixel is counted as a pixel in a monochrome area, so that an image can be accurately determined. it can.

Further, according to the present invention, the absolute value of the difference value of the color component of the pixel is one set larger than the first threshold value a and the absolute value of the other difference value is smaller than the second threshold value b. If the condition is satisfied, the pixel is determined to be a color shift pixel. In this case, there is a high possibility that this pixel is a pixel in which only one color component is displaced, thereby making it possible to reliably determine whether or not this pixel is a color-displaced pixel.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of an image discriminating apparatus 20 according to an embodiment of the present invention.

FIG. 2 is a graph showing an example in which a read image is divided into a color area, an intermediate area, and a monochrome area and the number of pixels is counted.

FIG. 3 is a diagram illustrating a concept of a monochrome block according to the present invention.

FIG. 4 is a flowchart illustrating an overall processing procedure of image discrimination in the embodiment of FIG. 1;

FIG. 5 is a diagram showing flags at both ends of a monochrome block.

FIG. 6 is a flowchart showing a processing procedure of an initial setting shown in FIG. 4;

FIG. 7 is a flowchart showing a processing procedure of a current flag clear shown in FIG. 4;

FIG. 8 is a flowchart illustrating a procedure of a pixel determination process illustrated in FIG. 4;

FIG. 9 is a flowchart illustrating a procedure of an attribute determination process in FIG. 4;

FIG. 10 is a flowchart showing a procedure of a flag check shown in FIG. 9;

11 is a block diagram showing a configuration of an attribute discriminating means 38 of FIG.

FIG. 12 is a flowchart illustrating a procedure of a color misregistration determination process illustrated in FIG. 10;

FIG. 13 is a flowchart illustrating a procedure of a termination determination process illustrated in FIG. 4;

FIG. 14 is a diagram illustrating the concept of the termination determination processing in FIG.

FIG. 15 is a block diagram showing a configuration of an image determining unit 40 shown in FIG.

FIG. 16 is a block diagram showing a configuration of a conventional image discriminating apparatus.

[Explanation of symbols]

 Reference Signs List 20 image discriminating device 21 image input means 22 pixel area discriminating means 23 difference value calculating means 24 maximum difference value selecting means 25 pixel discriminating means 26 color area pixel 27 intermediate area pixel 30 monochrome area pixel 31 background area pixel 32 monochrome block recognizing means 33 Pixel counting means 34 Color pixel counter 35 Intermediate pixel counter 36 Monochrome pixel counter 37 Monochrome block counter 38 Attribute discrimination means 39 Pixel count correction means 40 Image discrimination means 41, 42, 43 Color noise detection circuit 44 OR circuit 45 Judgment unit 46 Color line Counter 47 Comparator

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) LA10 LA31 MA01 NA29 PA01 5L096 AA02 AA03 AA06 BA07 EA35 GA19 GA28 GA36

Claims (5)

[Claims] An image input means for decomposing input image information into a plurality of color components for each pixel and deriving an image signal, and for each pixel based on a color component of an image signal from the image input means. A pixel area determining means for determining whether the pixel area belongs to a color area / intermediate area or a black-and-white area; and a pixel determined to belong to the black-and-white area by the pixel area determining means. A black-and-white area discriminating means for discriminating which pixel belongs to a color area, a pixel area discriminating means and a black-and-white area discriminating means. The pixel counting means for counting to the counter / intermediate area pixel counter or the monochrome area pixel counter and the monochrome area determining means (B) when a predetermined number or more of pixels determined to belong to the area exist continuously, a monochrome block recognizing means for recognizing the area of the continuous pixels as a monochrome block; and a monochrome block recognized by the monochrome block recognizing means. Attribute determining means for determining the attribute of the pixel located at the boundary of the periphery of the pixel, and a pixel for correcting the count value of the color area pixel counter / intermediate area pixel counter or the monochrome area pixel counter based on the determination result of the attribute determining means An image discriminating apparatus comprising: a count correcting unit; and an image discriminating unit that discriminates a type of an image input to the image input unit based on a count value correction result of the pixel count correcting unit. 2. The image determining means counts the number of pixels determined to belong to a color area / intermediate area or a monochrome area for all pixels of each line of an image input to the image input means. Determining whether or not the line is a color line including color area pixels, counting the number of color lines, and determining the type of image based on the counted number of color lines. 2. The image discriminating apparatus according to 1. 3. The image determination means according to claim 1, wherein all pixels of the image input to the image input means are provided with a color area and a color area.
The image discriminating apparatus according to claim 1, wherein the number of pixels determined to belong to the intermediate area or the monochrome area is counted, and the type of the image is determined based on the counted number of pixels of each area. 4. The pixel count correction unit, when the attribute determination unit determines that a pixel located at a boundary of a peripheral edge of a monochrome block is a color shift pixel, sets the pixel as a pixel belonging to a monochrome region. The image discriminating apparatus according to any one of claims 1 to 3, wherein the count value is corrected when there is an error. 5. The attribute discriminating means sets a plurality of color components of the pixel to three primary color densities (XYZ), and sets first and second predetermined thresholds as a and b, respectively, where a> b.
When b> 0, │XY│> a and │YZ│ <b or │YZ│> a and │ZX│ <b or │ZX│> a and │X- The image discriminating apparatus according to any one of claims 1 to 4, wherein the pixel is judged to be a color misregistration pixel on the condition that any inequality Y | <b is satisfied.