JPH10215382A - Color image processing unit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To discriminate colored characters in an original with high accuracy. SOLUTION: A colored character detection means 1 detects colored characters from a color image signal in the case of preliminary scanning. A frequency detection means 2 calculates a frequency of colored characters in all images by dividing an area of a domain detected as the colored characters with a total area of image data. A parameter selection means 3 selects a parameter corresponding to the frequency. A chromatic achromatic discrimination means 4 discriminates a chromatic achromatic domain of the image data obtained in the case of the main scanning by using the parameter selected in the case of the preliminary scanning.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color image processing apparatus for automatically determining whether a character in a document is chromatic or achromatic, and performing processing according to the result of the determination.

[0002]

2. Description of the Related Art In an image processing apparatus such as a digital color copier or a digital copier, when reproducing an image in which characters and patterns are mixed, a high gradation is applied to the pattern in order to obtain a high-quality reproduced image. It is desirable to perform an appropriate process, and to perform a process that emphasizes the resolution on characters. In order to realize such processing, it is necessary to separate a character area and a picture area in an image with high accuracy. That is,
For example, a color image processing apparatus is provided which includes means for detecting the hue of a document and means for determining whether the document is a character area or a halftone area, and reproduces a single black color when the document is achromatic and a character area. (Refer to Japanese Patent Application Laid-Open No. Sho 64-72662), however, high-quality reproduction is possible by reproducing black characters in a single black color.

However, it is not easy to detect an achromatic region due to a problem in the performance of the scanner, and many color characters are detected as black characters. On the other hand, if it is attempted to determine a color character with high accuracy, the determination accuracy of a black character is reduced, so that the black character is colored and the image quality is degraded. Since black characters are frequently used in commonly used documents, it is necessary to determine black characters with high accuracy.
In the chromatic / achromatic judgment criteria, the judgment of black characters is given priority, and the color characters are judged next.

By the way, in a map or the like, there are many cases where the color itself has a meaning rather than the fact that black characters are always a single black color. For example, purple or brown characters and line drawings are used. However, it is difficult to discriminate between black and black characters and line drawings. In particular, the topographic map issued by the Geospatial Information Authority of Japan has contour lines drawn with brown lines, and copying the contour lines causes a problem that most of the contour lines are output in black.

In order to solve such a problem,
For example, as described in JP-A-2-249365 (color image processing apparatus), a means for determining an achromatic color area of an input image and a means for correcting an achromatic color area signal (that is, a means for correcting an achromatic color area signal). When detected, the chromatic color area is not determined to be an achromatic color area).
It is necessary to improve the chromatic / achromatic color determination accuracy.

[0006]

However, in the above-mentioned prior art, it is necessary to take into account the processing accuracy of the scanner, the position correction of the line sensor, the difference in the MTF characteristics of each of the RGB color components, and the like, and the cost increases. Inevitable.

Further, there is a color image processing apparatus in which a color reading signal is classified into an achromatic color and a chromatic color, and an area in which the achromatic color and the chromatic color are classified is changed according to a mode set from the outside. (Japanese Patent Laid-Open No. 2-2
No. 80573). In this apparatus, when a photograph mode or a character mode is set, an area determined as an intermediate color is assigned to a chromatic area in the photograph mode, and to an achromatic area in the character mode. However, in this device, in the character mode, colored characters are easily output as black characters, and as a result, the purple characters and brown contour lines described above are reproduced more and more in a single black color. Still not resolved.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a color image processing apparatus for accurately determining a color character in a document.

[0009]

In order to achieve the above object, according to the present invention, there is provided means for reading an original as a digital image signal, and determining whether the image signal is chromatic or achromatic. A color image processing apparatus having at least a determination unit, a unit configured to detect a color character satisfying a predetermined condition from the image signal, a unit configured to detect an appearance frequency of the color character, Means for controlling the criterion of the determination means.

[0010] The invention according to claim 2 is characterized in that the color character detecting means includes means for detecting a character and means for detecting a color area under a predetermined condition.

According to a third aspect of the present invention, the frequency detecting means detects a frequency based on a total area of the document area and a total area of the area detected as the color character.

According to a fourth aspect of the present invention, the frequency detecting means detects a frequency based on a total area of the area detected as the character and a total area of the area detected as the color character. I have.

In the invention according to claim 5, the frequency detecting means performs a first calculation based on a total area of the document area and a total area of a region detected as a character, and executes the first calculation based on the total area of the region detected as a character. The second operation is performed based on the total area of the area and the total area of the area detected as the color character, and the frequency is detected based on the first and second operation results.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be specifically described below with reference to the drawings. <Embodiment 1> FIG. 1 shows the structure of Embodiment 1 of the present invention.
In the figure, 1 is a color character detecting means, 2 is a frequency detecting means for detecting the frequency of a color character, 3 is a parameter selecting means for selecting a parameter according to the detected frequency, and 4 is a color using the selected parameter. This is a chromatic / achromatic determining means for determining a chromatic / achromatic region of image data.

In this embodiment, first, a prescan is performed.
Scan the original using an image input device such as a scanner,
Obtain an image signal. The color character detection means 1 performs color character detection using this image signal.

FIGS. 5A and 5B show the structure of the color character detecting means. In the configuration of (a), after the specific color detection unit 41 detects a specific color, the character detection unit 42 performs character detection only on an area having that color, and in the configuration of (b), the character detection After that, only characters of a specific color are detected for only the character area.

FIG. 6 shows a specific color area detected by the specific color detecting means 41. That is, for the RGB image signal, the minimum value x = min (RGB) of RGB is in the range of [a, b] (a and b are predetermined values), and the linear expression R
The maximum value max (RGB) and the minimum value min (RG
When the difference d of B) satisfies d <m1x + n1, d> m2x + n2, it is detected as a specific color (m1, m2, n)
1, n2 is a predetermined constant).

As the character detecting means, for example, Ouchi, Imao, Yamada, IEICE Transactions vol. J7
5-D-2 No. 1 “Image area separation method for mixed images of characters / patterns (dots, photos)” described in 1992-01
Is used. That is, the character area has many high-level and low-level pixels (hereinafter, referred to as black pixels and white pixels),
At the edge portion, the black pixel and the own pixel are continuous. A character edge is detected based on the continuity of each of the black and white pixels. FIG. 8 shows the configuration of the edge area determination means.

In the ternarization processing section, the input image signal is ternarized using two fixed thresholds (TH1, TH2) (white pixel <TH1, TH1 ≦ intermediate density pixel <TH2, TH2).
≤ black pixels). Here, TH1 and TH2 are thresholds used for the ternarization processing, for example, when the input data is 0
When represented by 256 gradations (0 is white) from 255 to 255, values of TH1 = 20 and TH2 = 80 are taken.

The continuity detecting section detects a portion where the ternary black and white pixels continue by pattern matching. FIG. 9 shows an example of a pattern to be used. Pixels matching these patterns are detected as black continuous pixels and self-continuous pixels.

The neighborhood detecting section determines whether the black continuous pixel and the white continuous pixel are in the vicinity. When one or more black continuous pixels and one or more white continuous pixels are present in a block having a size of 5 × 5 pixels, the block is determined to be an edge region, otherwise, the block is determined to be a non-edge region. “1” is output from a pixel determined to be an edge area.

Returning to FIG. 1, the frequency detecting means 2 calculates the frequency of the color characters in the entire image by dividing the area of the area detected as the color characters by the total area of the image data.
The parameter selection means 3 selects a parameter to be output to the chromatic / achromatic determination means 4 according to the value of the frequency.

Next, using the image data obtained at the time of the main scan, the chromatic / achromatic determining means 4 determines a chromatic achromatic region of the image. The parameters used at this time are the parameters selected at the time of the pre-scan.

FIG. 7 shows the configuration of the chromatic / achromatic determining means.
As shown in FIG. 7, a circuit 51 for calculating the absolute value of the color difference,
In steps 52 and 53, the absolute values of the color differences of R, G, and B are obtained, and the output max (| R−G |, | GB−,
| BR |) is compared with a predetermined threshold value th by a comparator 55. If the predetermined threshold value is not exceeded, the color is determined to be achromatic, and if the threshold value is exceeded, the color is determined to be chromatic. When the frequency of the color characters is high, the parameter selection unit 3 selects th1 as the threshold value, and when the frequency of the color characters is low, the parameter selection unit 3 selects t1 as the threshold value.
h2 (> th1) is selected. This makes it possible to accurately determine a brown character or the like that has been erroneously determined to be a black character as a chromatic color.

The output of the chromatic / achromatic determining means is given to, for example, a UCR circuit constituting a color image processing apparatus. That is, when it is determined to be chromatic, Y, M, C, K
Is output, and when it is determined to be achromatic,
The outputs of Y, M, and C are set to 0, and only K is output (monochrome).

The above-mentioned chromatic / achromatic determination means is to make a determination on a pixel-by-pixel basis. The determined pixels are divided into blocks (for example, 8 × 8 pixels) composed of a plurality of pixels, and the chromatic / achromatic The color may be determined.

<Embodiment 2> FIG. 2 shows the configuration of Embodiment 2 of the present invention. The color character detecting means of the second embodiment includes the specific color detecting means 21, the character detecting means 22, and the AND gate 23 described with reference to FIG. First, prescanning is performed, and an original is read using an image input device such as a scanner to obtain an image signal. The specific color detecting means 21 performs specific color detection using the image signal, and performs character detection using the character detecting means 22. The AND gate 23 outputs a color character when a specific color and a character are detected.

The frequency detecting means 24 calculates the frequency of the color characters in the entire image by dividing the area of the area detected as the color characters by the total area of the image data. The parameter selecting unit 25 selects a parameter to be output to the chromatic / achromatic determining unit 26 according to the value of the frequency.

Next, a character is detected by the character detecting means 22 using the image data obtained at the time of the main scan, and a chromatic / achromatic determination is performed by the chromatic / achromatic determining means 26 using the parameters preselected in the prescan. Perform color determination. That is, when the character detecting means 22 detects a character and the chromatic / achromatic determining means 26 determines a chromatic color, the AND gate 27
Outputs a chromatic color character, and when the character detecting means 22 detects a character and the chromatic / achromatic determining means 26 determines an achromatic color, the AND gate 27 outputs an achromatic color character.
In this embodiment, the common character detection means 22 can be used at the time of pre-scan and at the time of main scan, so that the configuration is simplified and there is an advantage in cost.

<Embodiment 3> FIG. 3 shows the structure of Embodiment 3 of the present invention. The difference from the second embodiment is that the character detection unit 22
Is directly input to the frequency detecting means 24. The specific color detecting means 21 and the character detecting means 22 function in the same manner as in the second embodiment. Then, at the time of the pre-scan, the frequency detecting means 24 determines the area of the area detected as the color character output from the AND gate 23 by the character detecting means 22.
By dividing by the total area of the region determined to be a character, the frequency of the color character in the character region of the image is calculated. The parameter selecting unit 25 selects a parameter to be output to the chromatic / achromatic determining unit 26 according to the value of the frequency.

Next, a character is detected by the character detecting means 22 using the image data obtained at the time of the main scan.
The chromatic / achromatic determination is performed by the chromatic / achromatic determination unit 26 using the parameters selected in advance in the prescan. As in the second embodiment, when the character detecting unit 22 detects a character and the chromatic / achromatic determining unit 26 determines a chromatic color,
The D gate 27 outputs a chromatic color character. When the character detecting means 22 detects a character and the chromatic / achromatic determining means 26 determines an achromatic color, the AND gate 27 outputs an achromatic color character. I do.

<Embodiment 4> FIG. 4 shows the configuration of Embodiment 4 of the present invention. The frequency detecting means of the fourth embodiment includes a first calculating means 29, a second calculating means 28, and two comparators 30 and 31. The other configuration is the same as that of the third embodiment, and the description is omitted.

At the time of pre-scan, the first calculating means 29 calculates the area of the character area detected by the character detecting means 22 as
An operation of dividing by the total area of the image data is performed. Second calculation means 2
In step S8, an operation of dividing the area of the area detected as a color character by the area of the character area detected by the character detection unit 22 is performed. The comparators 31 and 30 decode the first calculation result and the second calculation result to a frequency of 2 bits by comparing them with thresholds th3 and th4, respectively. The parameter is selected by the parameter selecting means 25 according to the value of the frequency.
For example, if both the first operation result and the second operation result are th
If the third calculation result is larger than th4, a threshold for determining a chromatic color (FIG. 7) is selected as a parameter. If the first calculation result is larger than th3 and the second calculation result is smaller than th4, an achromatic color is used as a parameter. Is selected.

Characters are detected by the character detecting means using the image data obtained at the time of the main scan, and the chromatic / achromatic determination is performed by the chromatic / achromatic determining means using the parameters selected in advance in the prescan. That is, chromatic characters and achromatic characters are detected based on the character detection result and the chromatic / achromatic determination result.

[0035]

As described above, according to the first and second aspects of the present invention, the chromatic and achromatic judgment criteria are controlled in accordance with the frequency of color characters and black characters contained in the original. Therefore, the chromatic / achromatic determination process that matches the characteristics of the document can be accurately and automatically performed.

According to the third, fourth and fifth aspects of the present invention, the appearance frequency of the color characters is normalized by the size of the document, the ratio of the characters, and the like, so that the color characters can be detected with high accuracy.

[Brief description of the drawings]

FIG. 1 shows a configuration of a first exemplary embodiment of the present invention.

FIG. 2 shows a configuration of a second exemplary embodiment of the present invention.

FIG. 3 shows a configuration of a third embodiment of the present invention.

FIG. 4 shows a configuration of a fourth embodiment of the present invention.

FIGS. 5A and 5B show a configuration of a color character detection unit.

FIG. 6 shows a specific color area detected by a specific color detecting means.

FIG. 7 shows a configuration of a chromatic / achromatic determining unit.

FIG. 8 shows a configuration of an edge area determination unit.

FIG. 9 shows an example of a pattern for detecting continuity of black pixels and white pixels.

[Explanation of symbols]

 1 color character detecting means 2 frequency detecting means 3 parameter selecting means 4 chromatic / achromatic determining means

Claims (5)

[Claims] 1. A color image processing apparatus comprising: a unit for reading a document as a digital image signal; and a unit for determining whether the image signal is a chromatic color or an achromatic color. Means for detecting a color character satisfying a predetermined condition; means for detecting the frequency of appearance of the color character; and means for controlling a criterion of the determination means according to the frequency of appearance. Color image processing device. 2. A color image processing apparatus according to claim 1, wherein said color character detecting means comprises means for detecting a character and means for detecting a color area under a predetermined condition. 3. The color image processing apparatus according to claim 1, wherein the frequency detection unit detects a frequency based on a total area of the document area and a total area of the area detected as the color character. . 4. The frequency detection unit according to claim 1, wherein the frequency detection unit detects a frequency based on a total area of the area detected as the character and a total area of the area detected as the color character. The color image processing apparatus as described in the above. 5. The frequency detecting means performs a first operation based on a total area of the document area and a total area of a region detected as a character, and calculates a total area and a color character of the region detected as a character. 3. The color image processing apparatus according to claim 1, wherein a second operation is performed based on a total area of the detected region, and a frequency is detected based on the first and second operation results. .