JPH1173503A - Picture processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To abandon an error extraction of a character for inputted picture data and to more precisely judge and extract a net-point area. SOLUTION: A black peak extraction part 21 extracts a black peak from an L* lightness signal. A white peak extraction part 22 extracts a white peak from the L* lightness signal. In accordance with the above-mentioned black signal, a black peak scoring part 23 compares a picture element under consideration and a black peak pixel existing near it as objects with plural patterns and performs a scoring to the picture element under consideration according to a comparison result. Also, like the black peak scoring part 22, a white peak storing part 24 compares the picture element under consideration and a white peak element existing near it as objects with plural patterns and performs the scoring to the picture element under consideration in accordance with a comparison result. A net-point decision part 25 judges whether or not a pixel unit is a net-point area on the basis of a result (score) from scoring to the pixel unit in the black peak scoring part 23 and the white peak scoring part 24.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for separating a character area such as a character or a line drawing and a picture area such as a photograph or a picture from input image data.

[0002]

2. Description of the Related Art Conventionally, in a copying machine, a facsimile, or the like, when a character area such as a character or a line drawing and a picture area such as a photograph or a pattern are mixed in an input image, when the image is reproduced, It is desirable from the viewpoint of image quality to separate the character area from the picture area, perform processing with emphasis on resolution in the character area, and perform processing with emphasis on gradation in the picture area. Also, when transmitting the image data, it is preferable to separate the character area and the picture area and perform compression processing on each of them by different methods from the viewpoint of image quality and compression ratio.

[0003] In order to adaptively process a process that emphasizes resolution and gradation by using each region of an image, it is necessary to accurately separate a character region and a picture region included in the image (hereinafter, a character region and a pattern region). The separation of a picture area is called image area separation), and various proposals have been made for this image area separation. For example, an image is divided into blocks of a certain size, the maximum density and the minimum density of the pixels included in each block are obtained, and the difference between the maximum density and the minimum density is determined by a predetermined threshold. There is a method in which a block larger than a threshold value is determined as a character area and a block smaller than the threshold value is determined as a picture area. Further, as an example of extracting a halftone dot region, a method of determining an extreme point indicating a peak or a valley of a density change from a density relationship with surrounding pixels and determining whether or not the halftone dot region is based on the number of extreme points in a certain size block. There is.

[0004]

However, in the above-mentioned first prior art, since there is a region having a large density difference also in a picture region including a halftone such as a photograph, an erroneous recognition occurs and the image quality is deteriorated. There was a problem. Further, in the above-described second conventional technique, in a complicated and complicated line drawing, for example, a line drawing or a character created by CAD or the like, since there are many poles in the region, it is erroneously determined to be a halftone dot region. There is a problem.

The present invention has been made in view of the above circumstances, and provides an image processing apparatus capable of eliminating erroneous extraction of characters from input image data and accurately determining and extracting a halftone dot region. It is intended to be.

[0006]

In order to solve the above-mentioned problems, according to the present invention, a character area and a picture area of input image data are separated, and image processing for the image data is performed according to the area type. In the image processing device to be applied, extracting means for extracting a pixel that meets a predetermined condition based on the pixel value information of the image data, and a positional relationship between the pixel extracted by the extracting means and a neighboring pixel of the pixel, Comparing means for comparing with a plurality of predetermined patterns; score calculating means for calculating a score according to the degree of coincidence with the plurality of patterns based on a comparison result compared by the comparing means; Determining means for determining whether or not the target pixel of the image data is a halftone dot region based on the score calculated by the above.

According to the present invention, the extracting means extracts a pixel which satisfies a predetermined condition based on the pixel value information of the image data, and the comparing means compares the extracted pixel with a neighboring pixel of the pixel. Is compared with a plurality of predetermined patterns. Next, the score calculation means calculates a score corresponding to the degree of coincidence with the plurality of patterns based on the comparison result. And the judging means, based on the score,
It is determined whether or not the target pixel of the image data is a halftone area. As a result, the character area and the halftone area can be accurately separated, so that erroneous extraction of characters from the input image data can be eliminated, and only the halftone area can be accurately determined and extracted.

[0008]

Embodiments of the present invention will now be described with reference to the drawings. A. First embodiment A-1. Configuration of First Embodiment FIG. 1 is a block diagram showing a configuration of an image processing device according to a first embodiment of the present invention. In the figure, an image input unit 1
Is an image scanner or the like that reads color image information of an original for each color by a photoelectric conversion element such as a CCD camera and converts the color image information into an electric digital image signal. The input gradation correction unit 2 corrects the gradation of the image signal (for example, RGB, 8 bits for each color) input by the image input unit 1. The color signal conversion unit 3 converts the RGB image signal subjected to the gradation correction into another color signal (for example, L *, a *, b *).
The color signal conversion unit 4 converts the image signals of the color signals L *, a *, and b * into another color signal (for example, YMC) and supplies the color signal to the black plate generation unit 5.

[0009] The black plate generation unit 5 generates a black plate (K) from the YMC image signal, converts the image signal into a YMCK image signal, and supplies it to the spatial filter unit 6. Spatial filter unit 6
Performs a spatial filtering process for each color of the YMCK image signal. Further, the spatial filter unit 6 performs a spatial filtering process by, for example, switching a filtering coefficient or the like in accordance with an area determination result from an image area separating unit 9 described later. The output gradation correction unit 7 performs output gradation correction for each color of the YMCK image signal, which has been subjected to the spatial filtering processing, so as to match the image output gradation characteristic, and supplies the output gradation correction to the image output unit 8. Note that the output tone correction unit 7 also performs output tone correction according to the area determination result from the image area separation unit 9 described later.

The image output unit 8 switches the screen, for example, in accordance with the area determination result from the image area separation unit 9 described later, and outputs the output tone-corrected YMCK image signal as an image on, for example, paper. . Further, the image area separating unit 9
Performs an area determination of a character portion or a picture portion for each pixel according to the L * a * b * image signal output from the color signal conversion section 3. This area determination result is, as described above,
The image data is supplied to the spatial filter unit 6, the output gradation correction unit 7, and the image output unit 8, and is subjected to processing according to the region determination result.

Next, FIG. 2 is a block diagram showing the configuration of the image area separating section 9. As shown in FIG. In the figure, an image signal is input to a character extraction unit 11 and a halftone dot extraction unit 12. Character extraction unit 1
1 extracts a character area such as a character or a line drawing from an input image signal. As a method for extracting the character area, a differential value is obtained and compared with a predetermined threshold value,
There is a method of comparing and extracting the average of the pixel value of interest and its surrounding pixel values. It should be noted that these character extraction methods are not the gist of the present invention, and therefore, detailed description thereof will be omitted. The halftone dot extracting unit 12 determines whether or not the pixel of interest is a halftone area from the input image signal, and outputs the result. The details of the dot extracting unit 12 will be described later.

The character re-extracting unit 13 extracts a character region again using the output result of the character extracting unit 11 and the output result of the halftone dot extracting unit 12. The character re-extraction unit 13 removes an erroneous determination (erroneous extraction) included in the output result of the character extraction unit 11. That is, when the character extraction accuracy is increased in the character extraction unit 11, an edge region (a region having a large change in pixel value) included in a halftone dot region or the like is erroneously extracted. Therefore, the character re-extraction unit 13 removes the character extraction misjudgment area included in the output result of the character extraction unit 11 by using the halftone dot extraction result of the halftone dot extraction unit 12 and removes the character extraction result without erroneous judgment. obtain. FIG. 3 is a circuit diagram illustrating a configuration example of the character re-extraction unit 13. That is, the logical product (AND) of the NOT (NOT) of the signal of the output result of the dot extraction unit 12 and the signal of the output result of the character extraction unit 11 is calculated, and the result is compared with the output result of the character re-extraction unit 13. I do.

FIG. 4 is a block diagram showing a configuration of the above-mentioned halftone dot extracting section 12. As shown in FIG. Hereinafter, an L * signal (brightness signal) among L * a * b * image signals quantized into 8 bits for each component will be described as an example of an image signal input to the halftone dot extraction unit 12. However, without being limited to this, for example, a G signal of an RGB signal or (α1R +
α2G + α3B) / 3 (where α1, α2, and α3 are multipliers) may be used. It is also possible to input not only one component but also a plurality of components. Note that the L * signal value quantized to 8 bits described in the embodiment of the present invention is 00
h (“0” in hexadecimal, “0” in decimal), FFh (“FF in hexadecimal”, “25 in decimal”
5)) represents black.

In the figure, a black peak extracting unit 21
The black peak is extracted from the brightness signal. The black peak here is conceptually a pixel whose pixel value is larger than the neighboring pixel value. The details of the black peak extracting unit 21 will be described later. Next, the white peak extraction unit 22 calculates L
* Extract the white peak from the brightness signal. The white peak referred to here conceptually refers to a pixel whose pixel value is smaller than the neighboring pixel value, contrary to the above-described black peak.
The details of the white peak extracting unit 22 will be described later.

The black peak scoring unit 23 compares a target pixel and a black peak pixel existing in the vicinity thereof with a plurality of patterns in accordance with the black peak signal.
A point is assigned to the target pixel according to the comparison result. Also,
Similar to the black peak scoring unit 22, the white peak scoring unit 24 compares the target pixel and the white peak pixel present in the vicinity thereof with a plurality of patterns, and scores the target pixel according to the comparison result. I do. Next, the halftone dot determination unit 25 determines whether or not the pixel area is a halftone dot area based on the result (score) scored in pixel units by the black peak scoring unit 23 and the white peak scoring unit 24. judge. The details of the halftone dot determination unit 25 will be described later.

Next, FIG. 5 shows the above-mentioned black peak extraction unit 2.
1 is a block diagram showing a configuration of FIG. In the figure, a neighboring pixel average value calculation unit 31 calculates an average value of neighboring pixels of a target pixel according to an input L * lightness signal. As examples of the neighboring pixels, there are 4 neighborhoods (hatched pixels in FIG. 6A), 8 neighborhoods (hatched pixels in FIG. 6B), and 24 neighborhoods (FIG. 6C ))). In FIGS. 6A to 6C, the pixels painted black are the target pixels.

The adder 32 includes a neighborhood pixel average value calculation unit 31
The average value of the pixels in the vicinity of the pixel of interest calculated by the above is added to a predetermined addition value a, and the addition result is supplied to the comparator 33. It should be noted that the addition of the predetermined addition value a to the average value is performed by simply comparing the average value of the pixel in the vicinity of the pixel of interest with the pixel of interest when determining whether the pixel of interest is the peak pixel. This is to prevent erroneous determination that the target pixel is not the peak because the value is smaller than the value. Therefore, the sum a
May be a negative value. The comparator 33 compares the addition result with the pixel value of interest, and compares the comparison result with the AND circuit 3.
7 to one input terminal. That is, the comparator 33
Outputs an “ON” signal when the target pixel value is larger than the output result of the adder 32, and outputs
If it is smaller than or equal to the output result of No. 2, an "OFF" signal is output.

Further, the maximum value of the neighboring pixel calculating section 34 calculates the maximum value of the neighboring pixels of the target pixel according to the input L * lightness signal. Examples of the neighboring pixels include four neighborhoods and eight neighborhoods as described above. The adder 35 adds the maximum value of the pixel of interest in the vicinity of the pixel of interest calculated by the maximum neighboring pixel calculation unit 34 and a predetermined addition value b, and supplies the addition result to the comparator 36. The reason for adding the predetermined addition value b to the maximum value is the same as the reason for adding the predetermined addition value a to the average value described above. The comparator 36 compares the addition result with the pixel value of interest, and supplies the comparison result to the other input terminal of the AND circuit 37. That is, when the target pixel value is larger than the output result of the adder 35, the comparator 36 sets “O
N "is output, and if the pixel value of interest is smaller than or equal to the output result of the adder 35, an" OFF "signal is output. In the AND circuit 37, the comparator 3
The logical product of the output result of No. 3 and the output result of the comparator 36 is calculated and supplied to the OR circuit 42.

The 0-180 degree black line peak detector 38 detects a black line segment (ridge in the image) extending in the horizontal direction in accordance with the L * lightness signal and supplies it to the OR circuit 42. The 90-270 degree direction black line peak detection unit 39 has a black line segment extending in the vertical direction, and the 45-225 degree direction black line peak detection unit 40 has a black line segment extending in the upper right direction and lower left direction.
35-315 degree direction black line peak detection unit 41, upper left ← →
Detects black line segments extending in the lower right direction. In addition, 0-18
For details of the 0-degree black line peak detector 38, the 90-270 degree black line peak detector 39, the 45-225 degree black line peak detector 40, and the 135-315 degree black line peak detector 41, It will be described later.

The above-mentioned black line peak detectors 38 in the 0-180 degree direction, the black line peak detectors 39 in the 90-270 degree direction,
5-225 degree direction black line peak detectors 40 and 135
When the pixel of interest is a pixel constituting a line segment, the neighborhood pixel average value calculation unit 31 and the neighborhood pixel maximum value calculation unit 34 determine that the pixel of interest is a peak pixel. This is provided to prevent the pixel from being discriminated or from being discriminated as an intermittent peak pixel.

The OR circuit 42 outputs the logical product of the AND circuit 37 and the 0-180 degree direction black line peak detectors 38, 90-
The logical sum of the respective black line peak detection results of the 270-degree black line peak detection unit 39, the 45-225-degree black line peak detection unit 40, and the 135-315-degree black line peak detection unit 41 is calculated. The output of the black peak extraction unit 21 is as follows.

FIG. 7 is a block diagram showing the configuration of the above-described black line peak detector 38 in the 0-180 degree direction.
In the figure, one line memory (line buffer) 51-
1 delays the L * brightness signal by one line period. Also, 1
The line memory (line buffer) 51-2 further delays the L * brightness signal delayed by one line cycle by one line cycle. The latch circuits 52-1 to 52-6 latch the L * brightness signal, the L * brightness signal delayed by one line cycle, and the L * brightness signal delayed by two line cycles, respectively. That is, the one-line memories 51-1 and 51-
2. The latch circuits 52-1 to 52-6 convert the input data into a 3 × 3 pixel data centered on the target pixel Pe as shown in FIG.
The matrix is formed into a pixel group including pixels Pa to Pi constituting a window. Here, the pixel values of the pixels Pa to Pi are Va to Vi, as shown in FIG.

Of the 3 × 3 pixels, the pixel values Va to Vc are supplied to an adder 53-3, and the pixel values Vd to Vf are supplied to an adder 53-2. Vg to Vi are supplied to the adder 53-1. The adders 53-1 to 53-3 add the three input pixel values. That is, the adder 53-1 is configured as shown in FIG.
The total value of the lowermost pixel values Vg to Vi in the 3 × 3 matrix shown in FIG. The adder 53-
2 calculates the total value of the pixel values Vd to Vf at the middle stage in the 3 × 3 matrix shown in FIG. Then, the adder 53-3 calculates the total value of the uppermost pixel values Va to Vc in the 3 × 3 matrix shown in FIG. 8B.

The subtractor 54-1 subtracts the operation result of the adder 53-1 from the operation result of the adder 53-2, and supplies the subtraction result to the absolute value operation unit 55-1 and the comparator 56-1. I do. That is, the difference between the pixel value sum in the middle stage and the pixel value sum in the lower stage is calculated. The absolute value calculator 55-1 calculates the absolute value of the result of the subtraction, and supplies the result to the comparator 57-1. Further, the subtractor 54-2 subtracts the operation result of the adder 53-3 from the operation result of the adder 53-2, and supplies the subtraction result to the absolute value operation unit 55-2 and the comparator 56-2. . That is, the difference between the total pixel value in the middle stage and the total pixel value in the upper stage is calculated. The absolute value calculator 55-2 calculates the absolute value of the result of the subtraction and supplies the calculated value to the comparator 57-2.

That is, the subtractor 54-1 outputs the result of Expression 1. Further, the absolute value calculator 55-1 is given by the following equation (2).
Outputs the result. The subtractor 54-2 outputs the result of Expression 3. Further, the absolute value calculator 55-2 outputs the result of Expression 4.

[0026]

(Equation 1)

[0027]

(Equation 2)

[0028]

(Equation 3)

[0029]

(Equation 4)

The comparator 56-1 compares the subtraction result of the subtractor 54-1 with a predetermined threshold value TH2.
When the subtraction result of 1 is larger than a predetermined threshold value TH2, an "ON" signal is output, and when the result is smaller or equal, an "OFF" signal is output. Further, the comparator 57
-1 compares the operation result of the absolute value operation unit 55-1 with a predetermined threshold value TH3. If the operation result of the absolute value operation unit 55-1 is smaller than the predetermined threshold value TH3, "O"
An "N" signal is output, and when it is larger or equal, an "OFF" signal is output.

The comparator 56-2 includes a subtractor 54-2.
Is compared with a predetermined threshold value TH2. If the subtraction result of the subtractor 54-2 is larger than the predetermined threshold value TH2, an "ON" signal is output. Output signal. Further, the comparator 57-2 compares the operation result of the absolute value operation unit 55-2 with a predetermined threshold value TH3, and the operation result of the absolute value operation unit 55-2 is higher than the predetermined threshold value TH3. When it is smaller, it outputs an "ON" signal, and when it is larger or equal, it outputs an "OFF" signal.

Next, the AND circuit 59-1 is connected to the comparator 56.
The logical product of the comparison result of −1 and the comparison result of the comparator 56-2 is calculated and supplied to the OR circuit 60. AND circuit 59-
2 calculates the logical product of the comparison result of the comparator 57-1 and the comparison result of the comparator 56-2, and supplies the result to the OR circuit 60.
Further, the AND circuit 59-3 calculates the logical product of the comparison result of the comparator 56-1 and the comparison result of the comparator 57-2,
It is supplied to the OR circuit 60. The OR circuit 60 calculates the ethical sum of the operation results of the respective logical products, and supplies the result to the AND circuit 59-4.

The subtractor 54-3 subtracts the pixel value Ve from the pixel value Vf, and supplies the result of the subtraction (Vf-Ve) to the absolute value calculator 55-3. That is, the difference between the target pixel value Ve and the pixel value Vf in the right lateral direction (0 degree direction) is calculated. The absolute value calculator 55-3 calculates the absolute value of the subtraction result, and supplies the calculation result (| Vf-Ve |) to the maximum value detector 58. Further, the subtractor 54-4 subtracts the pixel value Vd from the pixel value Ve, and the subtraction result (Ve
-Vd) is supplied to the absolute value calculator 55-4. That is, the difference between the target pixel Ve and the pixel value Vd in the left horizontal direction (180-degree direction) is calculated.

The absolute value calculator 55-4 calculates the absolute value of the subtraction result, and supplies the calculation result (| Ve-Vd |) to the maximum value detector 58. The maximum value detector 58 calculates the above calculation result (| Vf-Ve |) and the above calculation result (| Ve-Vd).
) Is detected and supplied to the comparator 57-3. The comparator 57-3 compares the output value of the maximum value detector 58 with a predetermined threshold value TH1.
Is smaller than the threshold value TH1, "O
An "N" signal is output, and when it is larger or equal, an "OFF" signal is output. That is, the target pixel value Ve
If the difference between the pixel value in the horizontal direction (0-180 degrees direction) and the pixel value in the horizontal direction (0-180 degrees direction) is smaller than the threshold value TH1, it can be determined that the pixel of interest may be a part of a line extending in the horizontal direction. .
The AND circuit 59-4 calculates the logical product of the operation result of the above-described OR circuit 60 and the comparison result of the comparator 57-3.
It is output as a calculation result of the black line peak detector 38 in the -180 degree direction. Therefore, as described above, when the pixel of interest is a line segment, the output of the comparator 57-3 is "ON", so that the operation result of the OR circuit 60 is output. Since the output of No. 3 is "OFF",
The operation result of the R circuit 60 is not output.

The processing of the above-described black line peak detecting section 38 in the 0-180 degree direction will be described using equations. AND circuit 59-
4 outputs an “ON” signal when Expression 10 is satisfied,
If not, an "OFF" signal is output.

[0036]

(Equation 5)

[0037]

(Equation 6)

[0038]

(Equation 7)

[0039]

(Equation 8)

[0040]

(Equation 9)

[0041]

(Equation 10)

Next, FIG. 9 shows a 90-270 degree black line peak detector 39, FIG. 10 shows a 45-225 degree black line peak detector 40, and FIG. 11 shows a 135-315 degree black line peak. FIG. 3 is a block diagram illustrating a configuration of a detection unit 41. Note that, in each configuration, the processing content is the same as that of the black line peak detection unit in the 0-180 degree direction, except for the pixels, and a description thereof will be omitted.

FIG. 12 is a block diagram showing the structure of the white peak extracting section 22. Parts corresponding to those in FIG. 5 are denoted by the same reference numerals, and description thereof is omitted. In the figure,
The comparator 91 compares the addition result with the pixel value of interest,
The comparison result is supplied to one input terminal of the AND circuit 37. That is, the comparator 91 determines that the pixel value of interest is
If the output result of the adder 32 is larger than or equal to the output result of the adder 32, an "ON" signal is outputted.

The minimum neighboring pixel calculating section 92 calculates the minimum value of the neighboring pixels of the target pixel according to the input L * lightness signal. Examples of the neighboring pixels include four neighborhoods and eight neighborhoods as described above. The comparator 93 compares the addition result of the adder 35 with the target pixel value, and supplies the comparison result to the other input terminal of the AND circuit 37. That is, the comparator 93 outputs an “ON” signal when the target pixel value is smaller than the output result of the adder 35, and outputs a “ON” signal when the target pixel value is larger than or equal to the output result of the adder 35. Outputs an "OFF" signal. In the AND circuit 37, the output result of the comparator 91 and the comparator 9
The logical product with the output result of No. 3 is taken and supplied to the OR circuit 42.

The 0-180 degree direction white line peak detecting section 94 detects a white line segment (valley in the image) extending in the horizontal direction in accordance with the L * lightness signal and supplies it to the OR circuit 42. The 90-270 degree direction white line peak detection unit 95 has a white line segment extending in the vertical direction, and the 45-225 degree direction white line peak detection unit 96 has a white line segment extending in the upper right direction and lower left direction.
35-315 degree direction white line peak detection unit 97, upper left ← →
Detects white line segments extending in the lower right direction. In addition, 0-18
Details of the 0-degree white line peak detection unit 94, the 90-270-degree white line peak detection unit 95, the 45-225-degree white line peak detection unit 96, and the 135-315-degree white line peak detection unit 97 will be described later.

The OR circuit 42 outputs the logical product of the AND circuit 37 and the 0-180 degree white line peak detectors 94, 90-
The logical sum of the white line peak detection results of the 270 degree direction white line peak detection unit 95, the 45-225 degree direction white line peak detection unit 96, and the 135-315 degree direction white line peak detection unit 97 is calculated, and the above-described white peak extraction is performed. The output of the unit 22.

FIG. 13 is a block diagram showing the configuration of the white line peak detector 94 in the 0-180 degree direction. In addition,
Portions corresponding to the black line peak detector in the 0-180 degree direction shown in FIG. 7 are denoted by the same reference numerals, and description thereof is omitted. 0-18
The difference from the 0-degree direction white line peak detector is that the subtractor 54-
At 1, the adder 53-2 is calculated based on the addition result of the adder 53-1.
Is subtracted, and the subtractor 54-2 is subtracting the addition result of the adder 53-2 from the addition result of the adder 53-3.

That is, the following equation (11) is obtained from the subtractor 54-1.
Is output, and the result shown in Expression 12 is output from the absolute value calculator 5-1. Further, the result shown in Expression 13 is output from the subtractor 54-2, and the absolute value calculator 55
From -2, the result shown in Expression 14 is output.

[Equation 11]

[0049]

(Equation 12)

[0050]

(Equation 13)

[0051]

[Equation 14] The above-described processing of the white line peak detection unit 94 in the 0-180 degree direction will be described using equations. The AND circuit 59-4 is given by the following equation (2).
When “0” is satisfied, an “ON” signal is output. When “0” is not satisfied, an “OFF” signal is output.

[0052]

(Equation 15)

[0053]

(Equation 16)

[0054]

[Equation 17]

[0055]

(Equation 18)

[0056]

[Equation 19]

[0057]

(Equation 20) Next, FIG. 14 shows a configuration of a 90-270 degree direction white line peak detection unit 95, FIG. 15 shows a configuration of a 45-225 degree direction white line peak detection unit 96, and FIG. FIG. Note that, in each configuration, the processing content is different only for the pixels.
Since it is the same as the 180-degree direction white line peak detection unit 94,
Description is omitted.

Next, FIG. 17 is a block diagram showing the configuration of the black peak scoring section 23. In addition, FIG.
FIG. 18 is a block diagram illustrating a configuration of a white peak scoring unit 24, which is the same as the configuration of the black peak scoring unit 23 illustrated in FIG. Therefore, with regard to FIG. 18, since “black” in the following description may be replaced with “white”, the description is omitted. The black peak scoring unit 23 calculates a score for the target pixel from the positional relationship of the black peak pixels extracted by the black peak extraction unit 21. In the figure,
The matrix generation unit 100 includes, for example, a one-line memory (F
IFO) and a latch circuit. The peak extraction signal supplied from the black peak extraction unit 21 is matrixed into an M × N window size centering on the pixel of interest. In the present embodiment, M = N = 13.

Next, the pattern comparing section 101
It is determined which of a plurality of preset patterns the peak extraction signal matrixed into × 13 matches or is similar to. Specifically, if the pattern matches or is similar to the pattern, the pattern identification signal for identifying the pattern is output. Note that the pattern identification signal is not limited to one, and is not output unless there is a match or similarity with any pattern. If the pattern identification signal matches or is similar with a plurality of patterns, a plurality of the pattern identification signals are output. The pattern comparison unit 1
01 will be described later.

The scoring unit 102 supplies all the scores (points) corresponding to the pattern identification signal to the maximum score detecting unit 103. The details of the scoring unit 102 will also be described later. The maximum score detection unit 103 detects the largest score among all the scores, and supplies the largest score to the selector 105. The maximum score detection unit 1
03 indicates that no score was provided,
Outputs "0".

The continuous peak determining section 104 determines whether or not the target pixel is included and the number of peak pixels is K or more (K> 1) continuous according to the peak extraction signal formed into a matrix of 13 × 13. If they are continuous, an “ON” signal is supplied to the selector 105. If they are not continuous, an “OFF” signal is supplied to the selector 105. The details of the continuous peak determination unit 104 will be described later.

The selector 105 selects the output of the maximum score detection unit 103 when the output signal of the continuous peak determination unit 104 is "OFF", using the output signal of the continuous peak determination unit 104 as a selector signal. When the output signal of the continuous peak determination unit 104 is “ON”, “0” is selected and output as a processing result in the black peak scoring unit 23. That is, when K or more peaks are continuous (when the output signal of the continuous peak determination unit 104 is “ON”), it is determined that the pixel of interest is not likely to be a halftone dot region, and the black peak scoring unit 23 determines Outputs “0”. Conversely, when K or more peaks are not continuous (when the output signal of the continuous peak determination unit 104 is “OFF”), it is determined that the pixel of interest may be a halftone dot area, and black peak scoring is performed. The maximum score detecting unit 10 as an output from the unit 23
An output value of 3 is adopted.

Next, FIG. 19 is a conceptual diagram for explaining the operation of the pattern comparing section 101 described above. FIG.
In (a), each pixel that is formed into a matrix of 13 × 13 by the matrix generation unit 100 is represented by an alphabet and a numeral, and the pixel of interest is set to “A”. In the pattern comparison unit 101, the target pixel “A” is an ON pixel (peak pixel)
, And when all four pixels of the same alphabet from Ci to EEi (i = 1, 2, 3, 4) are "ON", it is determined that the pattern matches the pattern, and an identification signal of the pattern (alphabet) is output. I do.

Here, the operation of the above-described pattern comparing section 101 will be described with reference to FIG. First, in step S101, whether the target pixel “A” is “ON”
Check if it is "OFF". The target pixel “A” is “O”
If "N", the comparison process with the pattern is performed in step S102 and thereafter, and if "OFF", the pattern comparison process for the pixel of interest is terminated.

When the target pixel “A” is “ON”,
First, in step S102, pixels (represented by squares) that are patterns to be determined are initialized to “C”. Next, in step S103, □ i (i = 1, 2, 3, 4) are all “O
N ”is determined. If all are "ON", a pattern match identification signal corresponding to □ is output in step S104.

Next, in step S105, it is determined whether or not there is another pattern to be determined. If there is another pattern to be determined, the target is changed to that pattern in step S106, and the process proceeds to step S103. Return. Hereinafter, step S103 and subsequent steps are executed, all patterns are compared, and
Each time a match is made, a pattern match identification signal is output. If there is no other pattern to be determined, the pattern comparison process for the pixel of interest ends.

For example, in FIG. 19B, the target pixel "A" is "ON", and all four pixels K1 to K4 are "O".
N ”, a pattern match identification signal indicating that the pattern matches the pattern K is output. In FIG. 19C, the target pixel “A” is “ON”, all the four pixels K1 to K4 that are relatively close to the target pixel are “ON”, and DD1 is relatively far from the target pixel. Since four pixels .about.DD4 are "ON", a pattern matching identification signal indicating that the pattern K matches the pattern K is output.

A second processing example of the pattern comparing section 101 will be described with reference to FIG. First, in step S201, whether the target pixel “A” is “ON” or “O”
FF ”is determined. When the target pixel “A” is “ON”, the comparison process with the pattern is performed in step S202 and thereafter, and when the target pixel “A” is “OFF”, the pattern comparison process for the target pixel is ended.

If the target pixel "A" is "ON", in step S202, a pixel (indicated by a square) which is a pattern to be determined is initialized to "C". Next, step S
In 203, among □ i (i = 1, 2, 3, 4), “O
Check the number of "N" pixels. If all four pixels are “ON”, a pattern matching identification signal corresponding to □ is output in step S204. If the number of “ON” pixels is three, in step S205, A pattern similarity identification signal is output.

Next, in step S206, it is determined whether or not there is another pattern to be determined. If there is another pattern to be determined, the determination target is changed to that pattern in step S207, and the process proceeds to step S203. Return. Hereinafter, step S203 and subsequent steps are executed, and all patterns are compared.
A pattern matching identification signal is output each time they match, and a pattern similar identification signal is output each time they match. If there is no other pattern to be determined, the pattern comparison process for the pixel of interest ends.

For example, in FIG. 19D, the target pixel "A" is "ON", and among K1 to K4, K1, K3,
Since the three pixels K4 are "ON", a pattern similarity identification signal indicating that the pattern K is similar to the pattern K is output.

Further, a third processing example of the pattern comparing section 101 will be described with reference to FIG. First, in step S301, whether the target pixel “A” is “ON”
It is determined whether it is "OFF". Then, when the target pixel “A” is “ON”, a comparison process with the pattern is performed in step S302 and thereafter, and when the target pixel “A” is “OFF”,
The pattern comparison processing for the pixel of interest ends.

If the target pixel "A" is "ON", in step S302, a pixel (represented by a square) which is a pattern to be determined is initialized to "C". Next, step S
In 303, among □ i (i = 1, 2, 3, 4), “O
Check the number of "N" pixels. If all four are "ON", a pattern matching identification signal corresponding to □ is output in step S305. If the number of “ON” pixels is 1 to 3, in step S304,
It is determined whether at least one “ON” pixel exists in the vicinity of all the “i” pixels that are “OFF”, and if so, in step S306, a pattern similarity identification signal corresponding to Output. This neighborhood is, for example, 4
A neighborhood or eight neighborhoods can be used.

Next, in step S307, it is determined whether or not there is another pattern to be determined. If there is another pattern to be determined, the pattern to be determined is changed to that pattern in step S308, and the process proceeds to step S303. Return. Hereinafter, step S303 and subsequent steps are executed, and all the patterns are compared.
A pattern matching identification signal is output each time they match, and a pattern similar identification signal is output each time they match. If there is no other pattern to be determined, the pattern comparison process for the pixel of interest ends.

For example, in FIG. 19 (e), the target pixel “A” is “ON”, and in DD1 to DD4,
Since DD1 is “ON”, Z2 near DD2 at four, Z3 near DD3 at four, and PP4 near DD4 at four are “ON”,
A pattern similarity identification signal indicating that the pattern is similar to the pattern DD is output.

Although the above-described three examples have been given as examples of processing of the pattern comparing section 101, they may be combined. That is, when the determination target patterns are C to G, the processing described in the second example may be performed, and when the determination target patterns are H to EE, the processing described in the third example may be performed. .

Next, the scoring unit 102 will be described. The scoring unit 102 is, for example, a RAM
Data (score) based on the pattern matching identification signal or the pattern similarity identification signal supplied from the pattern comparing unit 101.
And outputs the data. Here, FIG.
FIG. 3 is a conceptual diagram illustrating an example of score data stored in a scoring unit. For example, the pattern comparing unit 10
When the pattern matching identification signal of “K” from 1 is supplied, “13” is output, and when the pattern similar identification signal of “DD” is supplied, “21” is output.

The score data stored in the scoring unit 102 may be an arbitrary value, but is desirably a value proportional to or approximately proportional to the square of the distance from the pixel of interest. In the example of the score data shown in FIG. 23, the square of the Euclidean distance between the target pixel and the pattern is used for the pattern matching identification signal, and
The value obtained by rounding up half of the square of the Euclidean distance between the target pixel and the pattern is used as score data. Further, regarding the scoring unit 102, in the above-described example, the RAM
Or ROM, the distance is calculated directly from the input pattern match identification signal or pattern similar identification signal using a multiplier or an adder, and score data is output. Is also good.

FIG. 24 is a circuit diagram showing a configuration of the above-mentioned continuous peak judging section 104. In the continuous peak determination unit 104, the 1 supplied from the matrix generation unit 100
K or more peak pixels including the pixel of interest are used (K> 1) by using the peak extraction signal matrixed into 3 × 13.
It is determined whether or not they are continuous, and if they are continuous, an “ON” signal is supplied to the selector 105 if they are not continuous. In FIG. 24, K
= 4 (4 or more consecutive peak pixels) are shown as an example. The alphabet in the figure is
This is a symbol indicating the position of a peripheral pixel when “A” shown in FIG.

In the figure, an AND circuit 111-1 calculates a logical product of inputted signals of four pixels H 3, D 3, B 3, and A, and supplies the result to an OR circuit 115. Similarly, the AND circuit 111-2 calculates the logical product of the signals of D3, B3, A, and B1,
The AND circuit 111-3 is a logical product of the signals of B3, A, B1, and D1, and the AND circuit 111-4 is a logical product of A, B1, D1, and H1.
And outputs the result to the OR circuit 115.
The OR circuit 115 calculates the logical sum of the input four signals and supplies the result to the OR circuit 119. That is, in the AND circuits 111-1 to 111-4, while shifting one pixel at a time, it is determined whether four consecutive pixels in the horizontal direction including the target pixel A are "ON", in other words, the target pixel A
, It is determined whether or not the “ON” signal is continuous for four pixels, and the OR result of each determination result is calculated by the OR circuit 115. Therefore, the output of the OR circuit 115 becomes "O
N ”,“ O ”in the horizontal direction including the pixel of interest A
"N" signal indicates that four pixels are continuous, and "OFF"
In the case of, it indicates that they are not continuous.

Similarly, AND circuits 112-1 to 112-
4 and the OR circuit 116 determine whether or not the “ON” signal is continuous for four pixels in the vertical direction including the pixel of interest A. The AND circuits 113-1 to 113-4 and O
The R circuit 117 determines whether or not the “ON” signal is continuous for four pixels in the lower left to right upper direction including the target pixel A. Further, the AND circuits 114-1 to 114-4 and the OR circuit 118 determine whether or not the "ON" signal is continuous for four pixels in the upper left direction to the lower right direction including the target pixel A.

The OR circuit 119 comprises the OR circuits 115 to 115
The logical sum of the output of 118 is calculated. That is, when the “ON” signal including the target pixel A is continuous for four pixels in any of the vertical and horizontal directions, the OR circuit 119 outputs “O”.
N "signal, and when four pixels are not consecutive,
An "OFF" signal is output.

In the present embodiment, the horizontal / vertical / lower left ← → upper right / upper left ← →
Although the four-pixel continuous determination in the four lower right directions is performed, a four-pixel continuous peak determination of, for example, “A”, “B1”, “E1”, and “I1” may be performed.

Next, FIG. 25 shows the above-described halftone dot determination section 25.
FIG. 3 is a block diagram showing the configuration of FIG. In the figure, the maximum score determination unit 121 determines the larger one of the scoring result for the black peak supplied from the black peak scoring unit 23 and the scoring result for the white peak supplied from the white peak scoring unit 24. The judgment is made and the larger score signal is supplied to the matrix generator 122. The matrix generation unit 122 includes, for example, a one-line memory (FI
FO) and a latch circuit, and converts the score signal into a matrix of K × L window size centering on the pixel of interest. In the present embodiment, the following description is made on the assumption that K = L = 25.

The score adding section 123 converts the score signals (6
(25 pixels), and the result of the addition is supplied to the score comparison unit 124. The score comparing section 124 compares the result of the addition with a predetermined threshold value TH25, and outputs the result of the comparison (halftone dot determination result). That is, when the output value from the score adding unit 123 is larger than the predetermined threshold value TH25, the target pixel is determined to be a halftone dot region, and an “ON” signal is output from the score comparing unit 124, If it is smaller than or equal to the threshold value TH25, it is determined that the pixel of interest is not a halftone dot area, and the “OFF” signal is output from the score comparison unit 124.

Next, FIG. 26 shows the above-described halftone dot determination section 25.
FIG. 13 is a block diagram showing another example of the configuration. Parts corresponding to those in FIG. 25 are denoted by the same reference numerals, and description thereof is omitted.
In the figure, an adder 125 adds the scoring result for the black peak supplied from the black peak scoring unit 23 and the scoring result for the white peak supplied from the white peak scoring unit 24, that is, the same result. The scores for the target pixel are added together, and the addition result is supplied to the matrix generation unit 122. Subsequent processing is the same as the above-described configuration, and thus will be briefly described.
The output of the adder 125 is output by the matrix generation unit 122 to 25
The score signal that is matrixed into a window size of × 25 and is matrixed is supplied to the score adding unit 123. In the score adding unit 123, 25 × 25 = 625
All the score signals for the pixels are added, and the score comparison unit 124 compares the addition result with a predetermined threshold value TH26, and outputs a comparison result (halftone dot determination result).

FIG. 27 is a block diagram showing the halftone judgment unit 2 described above.
It is a block diagram which shows the other example of a structure of FIG. In the figure,
The matrix generation unit 122-1 converts the scoring result for the black peak supplied from the black peak scoring unit 23 into a matrix having a window size of 25 × 25. The score adding unit 123-1 calculates the above 25 × 25 = 62.
All the score signals for the five pixels are added, and the result of the addition is supplied to the score comparison unit 124-1. Score comparison unit 1
24-1 is the score added value and a predetermined threshold value TH2
7 is compared with the threshold value TH.
If it is larger than 27, it outputs an "ON" signal, and if it is smaller than or equal to a predetermined threshold value TH27, it outputs an "OFF" signal.

On the other hand, the matrix generation section 122-2 converts the scoring result for the white peak supplied from the white peak scoring section 24 into a matrix of a window size of 25 × 25. The score addition unit 123-2 adds all the score signals for the above 25 × 25 = 625 pixels, and supplies the addition result to the score comparison unit 124-2. The score comparing unit 124-2 compares the above-mentioned score addition value with a predetermined threshold value TH28, and when the score addition value is larger than the predetermined threshold value TH28, “ON”
It outputs a signal, and outputs an "OFF" signal when it is smaller than or equal to a predetermined threshold value TH28.

The OR circuit 126 is connected to the score comparing section 12
The logical sum of the comparison result of 4-1 and the comparison result of the score comparison unit 124-2 is calculated, and the calculation result is output as the halftone dot determination result. As described above, the black and white peaks are extracted from the input image signal, the scores are obtained from the black and white peak signals, and the obtained two score values are used. It becomes possible.

A-2. Operation of First Embodiment Next, the operation of the above-described image area separation unit 9 will be described.
First, the character extracting unit 11 obtains a differential value by a known method, for example, a method of extracting the differential value by comparing it with a predetermined threshold value,
A character area such as a character or a line drawing is extracted from the input image signal by a method of comparing and extracting the average of the pixel value of interest and its neighboring pixel values.

Next, the dot extracting unit 12 determines whether or not the pixel of interest is a dot area from the input image signal.
Output the result. Specifically, the black peak extracting unit 21 in the halftone dot extracting unit 12 extracts a black peak from the L * lightness signal. The black peak is supplied to the black peak scoring unit 23. The black peak scoring unit 23 first determines to which of a plurality of preset patterns the peak extraction signal matrixed into 13 × 13 around the pixel of interest matches or is similar. The score corresponding to the degree of coincidence and the degree of similarity is scored, and it is determined whether or not K or more black peaks are continuous. When K or more black peaks are continuous, it is determined that the pixel of interest is unlikely to be a halftone dot area, and “0” is determined.
On the other hand, if K or more black peaks are not continuous, it is determined that the pixel of interest may be a halftone dot area,
The maximum value among the scores according to the degree of coincidence and the degree of similarity is supplied to the dot determining unit 25.

In the white peak extracting section 22 in the halftone dot extracting section 12, a white peak is extracted from the L * lightness signal. The white peak is supplied to the white peak scoring unit 24. The white peak scoring unit 24 compares the target pixel and the white peak pixel existing in the vicinity thereof with a plurality of patterns in the same manner as the above-described black peak scoring unit 23, and responds to the comparison result. In addition, scoring is performed on the target pixel, and it is determined whether or not K or more white peaks continue.
When K or more white peaks are continuous, it is determined that the pixel of interest is unlikely to be a halftone dot area, and “0” is output.
The pixel of interest is determined to be a halftone dot area, and the maximum value among the scores according to the coincidence and similarity is supplied to the halftone dot determination unit 25.

The halftone determination section 25 determines which of the maximum value for the black peak and the maximum value for the white peak is larger, and then forms a matrix into a window size of 25 × 25. The sum of the scores is calculated, and if the score is larger than a predetermined threshold value, the pixel of interest is determined to be a halftone dot region,
If they are equal, it is determined that the target pixel is not a halftone dot area.

The character re-extraction unit 13 removes the output result of the character extraction unit 11 and the halftone dot extraction unit 12 in order to remove erroneous judgments (erroneous extraction) included in the output result of the character extraction unit 1.
The character area is extracted again by using the output result of.
That is, when the character extraction accuracy is increased in the character extraction unit 11, an edge region (a region having a large change in pixel value) included in a halftone dot region or the like is erroneously extracted. Therefore, the character re-extraction unit 13 removes the character extraction misjudgment area included in the output result of the character extraction unit 11 by using the halftone extraction result of the halftone extraction unit 12, so that character extraction without erroneous judgment is performed. The result is obtained.

B. Second Embodiment Next, a second embodiment of the present invention will be described. Book second
The embodiment relates to the black peak scoring unit 23 illustrated in FIG. 4 described in the first embodiment.
Here, FIG. 28 is a block diagram illustrating another configuration example of the black peak scoring unit 23. Note that parts corresponding to those in the first embodiment are given the same reference numerals and description thereof is omitted.

In the figure, the matrix generation unit 100
The peak extraction signal supplied from the black peak extraction unit 21 is formed into a matrix with a window size of, for example, 13 × 13 centered on the pixel of interest. Pattern comparison unit 101
A value of -1 compares whether the peak extraction signal matrixed into 13 × 13 matches or is similar to a plurality of predetermined patterns. If the pattern matches or is similar to the pattern, a pattern identification signal for identifying the pattern is output. Note that the pattern identification signal is not limited to one, and is not output unless it matches or is similar to any pattern, and if it matches or is similar to a plurality of patterns, the corresponding number (plural) is output. You.

The scoring section 102-1 supplies all the scores corresponding to all of the pattern identification signals to the maximum score detecting section 103-1. Maximum score detection unit 103-
1 detects the largest score from all scores and supplies it to the selector 105-1. Note that the maximum score detection unit 103-1 outputs “0” when nothing is input.

On the other hand, the continuous peak judging section 104-1 includes the pixel of interest in accordance with the peak extraction signal formed into a matrix of 13 × 13 by the above-described matrix generating section 100,
It is determined whether or not K or more (K> 1) consecutive peak pixels are present. If the peak pixels are consecutive, an “ON” signal is output. If not, an “OFF” signal is output to the selector 105-1.
To supply.

The selector 105-1 uses the output signal of the continuous peak determination section 104-1 as a selector signal, and when the output signal of the continuous peak determination section 104-1 is "OFF", the maximum score detection section 103-1 The output is selected. Conversely, when the output signal of the continuous peak determination unit 104-1 is “ON”, “0” is selected and supplied to the selector 134.

The pixel inverting section 130 inverts all “ON” and “OFF” of the 13 × 13 matrix of the peak extracted signals in the matrix generating section 100, and outputs the continuous peak determining section 104-2. And pattern comparison unit 10
1-2.

The pattern comparing section 101-2, the scoring section 102-2, the maximum score detecting section 103-2, the continuous peak judging section 104-2, and the selector 105-2 are respectively composed of a pattern comparing section 101-1 and a scoring section. 102-
1, maximum score detection unit 103-1, continuous peak determination unit 1
04-1 and the selector 105-1, and the description is omitted.

Also, the matrix generation unit 100 calculates 13 × 1
The peak extraction signal matrixed into 3 is also supplied to the pixel count section 131. However, the peak extraction signal supplied to the pixel number counting unit 131 does not need to be all 13 × 13 pixels, and some pixels at the center (for example, a 9 × 9 matrix centered on the pixel of interest is A peak extraction signal for the constituent pixel group) may be used. Hereinafter, the peak extraction signal supplied to the pixel number counting unit 131 is
Assume that the signals are pixels of a 9 × 9 matrix.
The pixel count unit 131 calculates the number of pixels that are “ON” in the peak extraction signal composed of 9 × 9 pixels, and supplies the result to the comparator 132.

The comparator 132 includes a pixel number counting unit 131
Number of “ON” pixels supplied from the memory and a predetermined threshold value TH2
9 and supplies the result of the comparison to the selector 134. Note that this predetermined threshold value is determined by the pixel number counting unit 1.
It is desirable that the number of pixels supplied to the pixel 31 is about half. That is, in the present embodiment, the pixel number counting unit 13
Since the number of pixels supplied to 1 is 9 × 9 = 81, the above-mentioned predetermined threshold value is preferably set to about “40”.

The selector 134 selects and outputs either the output result of the selector 105-1 or the output result of the selector 105-2 according to the signal supplied from the comparator 132. That is, when the number of peak pixels around the target pixel is small (when the output of the pixel number counting unit 131 is smaller than the predetermined threshold TH29), the score is calculated using the peak extraction signal from the black peak extraction unit 21. On the contrary, if the number of peak pixels around the target pixel is large (the output of the pixel number
If it is greater than or equal to 9), a score is calculated using a signal obtained by inverting ON and OFF of the peak extraction signal from the black peak extraction unit 21.

Although the description is omitted, the white peak scoring unit 24 also performs the above-described black peak scoring unit 23.
Similarly to the configuration described above, the configuration described above can be adopted.

As described above, the black peak and the white peak are extracted from the input image signal, the score is obtained from the black and white peak signals and the inverted signals thereof, and the score is obtained from the ratio of the number of peak pixels. By switching the score and determining the halftone dot, the scale of the process is larger than in the first embodiment, but more accurate halftone dot determination is possible.

[0107]

As described above, according to the present invention, the extraction means extracts a pixel which satisfies a predetermined condition based on the pixel value information of the image data.
The positional relationship between the extracted pixel and a neighboring pixel of the pixel is
After comparing with a plurality of predetermined patterns, the score calculation unit calculates a score corresponding to the degree of coincidence with the plurality of patterns based on the comparison result, and then determines the image based on the score by the determination unit. Since it is determined whether the target pixel of the data is a halftone area or not, the character area and the halftone area can be accurately separated, so that erroneous extraction of characters from input image data is eliminated and only the halftone area is used. Can be determined and extracted with high accuracy.

[Brief description of the drawings]

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

FIG. 2 is a block diagram illustrating a configuration of an image area separation unit 9.

FIG. 3 is a circuit diagram showing a configuration of a character re-extraction unit 13;

FIG. 4 is a block diagram illustrating a configuration of a halftone dot extracting unit 12.

FIG. 5 is a block diagram illustrating a configuration of a black peak extraction unit 21.

FIG. 6 is a conceptual diagram for explaining the relationship between a target pixel and its neighboring pixels.

FIG. 7 is a block diagram showing a configuration of a 0-180 degree direction black line peak detection unit 38;

FIG. 8 is a conceptual diagram illustrating pixels of a 3 × 3 matrix centered on a target pixel and pixel values thereof.

9 is a block diagram illustrating a configuration of a 90-270 degree direction black line peak detection unit 39. FIG.

FIG. 10: 45-225 degree direction black line peak detector 40
FIG. 3 is a block diagram showing the configuration of FIG.

FIG. 11: 135-315 degree direction black line peak detector 4
1 is a block diagram showing a configuration of FIG.

FIG. 12 is a block diagram illustrating a configuration of a white peak extracting unit 22.

FIG. 13 is a block diagram illustrating a configuration of a white line peak detection unit 94 in the 0-180 degree direction.

14 shows a 90-270 degree direction white line peak detector 95. FIG.
FIG. 3 is a block diagram showing the configuration of FIG.

FIG. 15: 45-225 degree direction white line peak detector 96
FIG. 3 is a block diagram showing the configuration of FIG.

FIG. 16: White line peak detector 9 in the direction of 135 to 315 degrees
7 is a block diagram showing a configuration of FIG.

FIG. 17 is a block diagram illustrating a configuration example of a black peak scoring unit 23.

FIG. 18 is a block diagram illustrating a configuration example of a white peak scoring unit 24.

FIG. 19 is a conceptual diagram showing an example of a 13 × 13 matrix of pixels centered on a pixel of interest and a peak extraction result in the matrix generation unit 100.

FIG. 20 is a flowchart illustrating a first processing example of the pattern comparison unit 101;

FIG. 21 is a flowchart illustrating a second processing example of the pattern comparison unit 101.

FIG. 22 is a flowchart for explaining a third processing example of the pattern comparing unit 101;

FIG. 23 is a conceptual diagram showing an example of data (score) stored in the scoring unit 102.

FIG. 24 is a circuit diagram showing a configuration of a continuous peak determination unit 104.

FIG. 25 is a block diagram illustrating a first configuration example of a halftone dot determining unit 25.

FIG. 26 is a block diagram illustrating a second configuration example of the halftone dot determination unit 25.

FIG. 27 is a block diagram illustrating a third configuration example of the halftone dot determination unit 25.

FIG. 28 is a block diagram illustrating a configuration of a black peak scoring unit 23 according to the second embodiment of the present invention.

[Description of Signs] 21 Black peak extracting unit (extracting means, black peak extracting means) 22 White peak extracting unit (extracting means, white peak extracting means) 23 Black peak scoring unit (comparing means, score calculating means) 24 white peak Scoring section (comparing means, score calculating means) 25 halftone determining section (determining means) 100 matrix generating section 101 pattern comparing section (comparing means) 102 scoring section (point calculating means) 103 maximum score detecting section (point calculating means) 104 continuous peak determination unit 105 selector 101-1 pattern comparison unit (first comparison unit) 102-1 scoring unit (first score calculation unit) 103-1 maximum score detection unit (first score calculation unit) 101 -2 Pattern comparing section (second comparing means) 102-2 Scoring section (second score calculating means) 103-2 Maximum score A Detection unit (second score calculation means)

Claims (11)

[Claims] An image processing apparatus that separates a character area and a picture area of input image data and performs image processing on the image data according to an area type, wherein a predetermined area is determined based on pixel value information of the image data. Extracting means for extracting a pixel that satisfies the condition; comparing means for comparing the positional relationship between the pixel extracted by the extracting means and a pixel adjacent to the pixel with a plurality of predetermined patterns; Based on the comparison result,
A score calculation unit that calculates a score according to the degree of coincidence with the plurality of patterns, based on the score calculated by the score calculation unit,
An image processing apparatus comprising: a determination unit configured to determine whether a target pixel of the image data is a halftone area. 2. The extracting means compares the pixel value of interest with an average value of neighboring pixel values of the pixel of interest,
2. The image processing apparatus according to claim 1, wherein pixels that meet the predetermined condition are extracted. 3. The method according to claim 1, wherein the extracting unit extracts a pixel that satisfies the predetermined condition by comparing a pixel value of interest with a maximum value or a minimum value of neighboring pixel values of the pixel of interest. The image processing device according to claim 1. 4. The black peak extracting means for extracting a pixel meeting the predetermined condition by comparing a pixel value of interest with an average value or a maximum value of neighboring pixel values of the pixel of interest. And a white peak extracting means for extracting a pixel that meets the predetermined condition by comparing the target pixel value with an average value or a minimum value of neighboring pixel values of the target pixel. The image processing device according to claim 1. 5. The extracting unit extracts a pixel constituting a line segment in an image as a pixel that meets the predetermined condition by comparing a pixel value of interest with a pixel value near the pixel of interest. The image processing apparatus according to claim 1, wherein: 6. The image processing apparatus according to claim 1, wherein the comparing unit compares a positional relationship of a target pixel and a total of five pixels including at least four pixels near the target pixel with the plurality of patterns. 7. The image processing apparatus according to claim 1, wherein the comparing unit compares a positional relationship of a pixel of interest and at least three pixels in the vicinity of the pixel of interest with a total of four pixels with the plurality of patterns. 8. The image processing apparatus according to claim 1, wherein the comparing unit compares the positional relationship between the pixel of interest and at least four neighboring pixels near the pixel of interest with the plurality of patterns. . 9. The score calculation means, when calculating a score corresponding to the degree of coincidence with the plurality of patterns based on a comparison result compared by the comparison means, compares the score with the comparison means. 2. The image processing apparatus according to claim 1, wherein the value is proportional to the square of the Euclidean distance between the target pixel and a pixel near the target pixel. 10. The score calculation unit according to claim 1, wherein the score calculation unit changes the score when the pixel extracted by the extraction unit includes a target pixel and continues for a predetermined number or more. Image processing device. 11. The comparing means comprises: first comparing means for comparing a positional relationship between a pixel extracted by the extracting means and a pixel adjacent to the pixel with a plurality of predetermined patterns; Inverting the pixel extracted by the above, and a second comparison means for comparing the positional relationship between the inverted pixel and a neighboring pixel of the pixel with a plurality of predetermined patterns, the score calculation means, First score calculation means for calculating a score corresponding to the degree of coincidence with the plurality of patterns based on a comparison result compared by the first comparison means; and a comparison result compared by the first comparison means. And a second score calculation means for calculating a score according to the degree of coincidence with the plurality of patterns, based on the number of pixels, the determination means according to the number of pixels extracted by the extraction means, It is determined whether or not the target pixel of the image data is a halftone dot area based on either the score calculated by the first score calculation means or the score calculated by the second score calculation means. The image processing apparatus according to claim 1, wherein: