JP2001053968A - Pole pixel detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately detect a pole pixel that corresponds to a halftone dot included in an image by comparing the center pixel existing in a prescribed local area of the image and the center pixel existing in a local area greater than a prescribed local area and around the center pixel of the prescribed local area with the density level of a prescribed peripheral pixel to detect the 1st and 2nd pole pixels and defining the OR of the 1st and 2nd pole pixels as a pole pixel. SOLUTION: A local area is defined as a rectangular mask of a small (3×3) pixel size and also as a rectangular mask of a larger (5×5) pixel size. In regard to these two rectangular masks, the center pixels Lc are defined as the 1st and 2nd pole pixels when the density levels of the pixels Lc are higher than those of the prescribed peripheral pixels L1-L8 by the prescribed threshold. Then the OR of the 1st and 2nd pole pixels is defined as a pole pixel. Thus, the pole pixels can be accurately detected to the halftone dots included in a wide range of the number of lines.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an extreme pixel which is applied to a dot area separating apparatus for separating pixels in an image into a halftone area and other non-halftone areas, such as a digital copying machine and a facsimile. It relates to a detection device.

[0002]

2. Description of the Related Art Continuous tone photographs / pictures represented by halftone dots /
For example, when an image in which line drawings such as characters are mixed is reproduced by a digital copying machine, in order to improve the image quality of the reproduced image, dither processing is performed on the photograph part, moiré removal processing is performed on the halftone dot part, and the line drawing part is processed. Is desirably subjected to a sharpening process.

Further, even in the case of transmitting an image by facsimile or the like, in order to improve the compression ratio of the image data, it is necessary to perform image compression processing suitable for each area of a photograph part, a halftone part, and a line drawing part. desirable.

In order to realize such image processing, it is necessary to accurately separate and extract a photograph portion, a halftone dot portion, and a line drawing portion in an image as pre-processing. Generally, to separate a photographic part from other non-photographic parts (= halftone part + line drawing part) can be relatively easily realized by determining the number of edge pixels in an image. Other non-dots (=
It is difficult to accurately separate (photograph part + line drawing part).

Conventionally, as one method for separating the halftone dot portion from other non-halftone dot portions, for example, Ueno's method (Ueno: "Expression of halftone dot photograph by dot printer", Oki Electric R & D 132 No.53 No.4 pp.71-76). According to this method, an original image is raster-scanned and taken out as digital multi-valued data, and a density difference between light and dark is determined between adjacent pixels on the raster, and the difference is calculated by the following (a) and (b).
When any of the conditions (c) is satisfied, the target pixel is detected as a peak pixel of a peak or a valley of a density change.

(A) When the sign of the difference changes. (B) When the sign of the difference value changes before and after the difference is zero. (C) When the interval from the previous pole reaches a predetermined threshold value. Then, based on the extreme pixel information obtained as described above, a pixel that satisfies any of the following conditions (d) and (e) is detected as a halftone dot area.

(D) When the section length L (i) between the poles satisfies the following equation with respect to predetermined thresholds TH1 and TH2: TH1 <L (i) <TH2 (e) Current section length L (i) ) And the preceding section length L (i-1)
Satisfies the following equation with respect to a predetermined threshold value TH3: | L (i) −L (i−1) | ≦ TH3

[0008]

The above Ueno's method separates the halftone dots in an image on the assumption that the peaks, which are peaks and valleys of the density level, appear regularly in the halftone dots. is there. However, in general, there is a problem that a sufficiently high separation rate cannot be expected because many poles also exist in a photograph part and a line drawing part.

Further, since halftone dots are detected by comparing one-dimensional image information between pixels arranged on a raster scan line, particularly, a halftone dot portion having a low or high area ratio, a document rotating, or the like. Further, in a halftone dot portion where the screen angle is slightly deviated from the horizontal direction, the distance between the extreme points becomes long, and there is a problem that it is difficult to separate from a line drawing portion such as a character.

The present invention has been made under the above circumstances, and an object thereof is to provide an image processing apparatus which has an extreme point corresponding to a halftone dot in an image irrespective of the magnitude of the area ratio of a halftone dot portion and the arrangement state of a document. It is an object of the present invention to provide an extreme pixel detection device capable of accurately detecting pixels.

[0011]

According to the first aspect of the present invention, the extreme pixel is determined by comparing the density level of a central pixel in a predetermined local area of an image converted into digital multi-valued data with a predetermined surrounding pixel. A first extreme pixel detecting means for detecting, and a second extreme pixel detecting the extreme pixel by comparing a density level of a central pixel in a local area wider than the predetermined local area with a predetermined peripheral pixel with the central pixel as a center. A detection unit; and a logical sum unit that outputs a logical sum of a detection result of the first pole pixel detection unit and a detection result of the second pole pixel detection unit as a pole pixel.

In the invention according to claim 2, the predetermined local area of the first extreme pixel detection means is an area composed of 3 × 3 pixels centered on the center pixel, and
The wide local area of the extreme pixel detection means is an area composed of 5 × 5 pixels.

According to the third aspect of the present invention, the wide local area of the second extreme pixel detecting means is masked every other row and column adjacent to the center pixel as the first extreme pixel. The first extreme pixel detection means and the second extreme pixel detection means detect the extreme pixels by comparing the density levels of predetermined surrounding pixels at the same position with the same area as the local area of the detection means.

According to the fourth aspect of the present invention, the extreme point detection condition is set such that when the density level of the central pixel in the local area is maximum or minimum compared to the density level of a predetermined surrounding pixel, the central pixel is determined as the extreme pixel. Detected as

According to the fifth aspect of the present invention, the extreme point detection condition is set when the density level of the center pixel in the local area is larger or smaller than a predetermined threshold value compared to the density level of predetermined surrounding pixels. The center pixel is detected as a pole pixel.

According to the sixth aspect of the present invention, the extreme point detection condition is such that the density level of the central pixel in the local area is maximum or minimum as compared with the density level of a predetermined surrounding pixel, and the density level of the central pixel is When the absolute value of the difference between the center pixel and the average value of the density levels of the other two pixels located at the point target position with the center pixel as the center is larger than a predetermined threshold value (TH2), the center pixel is set to the extreme point. Detected as pixels.

In the present invention, pixels arranged on a straight line passing through the center pixel and extending in the up, down, left, and right directions are used as the predetermined surrounding pixels to be compared with the center pixel. In the invention according to claim 8, the threshold value is changed according to the density level of the center pixel or the density level of predetermined surrounding pixels.

According to the ninth aspect of the present invention, the threshold value in the case of detecting the peak pixel of the mountain where the density level of the center pixel is higher than the density level of the predetermined surrounding pixels is large when the density level of the center pixel is high. And the threshold value in the case of detecting a pole pixel in the valley where the density level of the center pixel is lower than the density level of the predetermined surrounding pixels is changed so as to become smaller when the density level of the center pixel becomes smaller. .

According to the tenth aspect of the present invention, an average value of the density levels of predetermined surrounding pixels is obtained, and the threshold value is changed according to the average value.

[0020]

FIG. 1 is a block diagram of a halftone dot area separating apparatus to which a pole detecting apparatus according to the present invention is applied. In the figure,
1 is an input image signal unit, 2 is an extreme pixel detection unit according to the present invention, 3 is a halftone dot candidate region detection unit, and 4 is a halftone dot region detection unit. Hereinafter, a specific configuration example of each of the circuits 1 to 4 will be described separately for each circuit. Note that the original image to be processed in the following example is an image in which a photograph portion, a halftone dot portion, and a line drawing portion are mixed. The halftone portion included in the image is, for example, about 65 to 200 lines, and the characters constituting the line drawing portion are of the seventh class or higher.

[I] Input Image Signal Unit 1 The input image signal unit 1 is a circuit that reads an original image and converts it into digital multi-value data. As the input image signal unit 1, for example, a scanner is used. In the input image signal section 1, the original image is converted into digital multi-valued data of, for example, about 400 dpi and about 64 gradations, and data for predetermined scan lines is stored and stored in a built-in memory.

If the reproduced image is monochrome, the processing may be executed using a luminance signal. If the reproduced image is color, the RGB signals after color separation (or YM after color correction)
By using the C signal), the area separation processing of the present invention may be performed in parallel for each of the RGB colors, and it may be finally determined whether or not the area is a halftone dot area by majority logic of the separation result for each color.

The input image signal unit 1 not only converts the original image into digital multi-valued data, but also converts the converted digital multi-valued data as preprocessing for the next extreme pixel detection in the extreme pixel detection unit 2. Smoothing may be performed with a predetermined weight count. By this smoothing processing, (a) removal of digital noise in the image, and (b) in the next detection processing of the extreme pixel in the extreme pixel detection unit 2, a mask having a small pixel size (for example, 3 ×
Large halftone dots (for example, 65 pixels)
It is possible to detect the extreme pixel of (line class).

FIG. 2 shows an example of the smoothing process. FIG. 2 (a)
Indicates the density level state of the halftone dot before smoothing. In this figure, the pixels having the maximum density level constituting the poles extend over three pixels. For example, when the pole detection is performed using a small mask having a size of 3 × 3 pixels, the density difference between the pixels in this mask is determined. Since it does not occur, it is impossible to detect an extreme pixel. However, when the smoothing process is performed, as shown in FIG. 2B, the density level of the central pixel becomes the maximum, and the density levels of the pixels before and after the central pixel decrease. For this reason, even with a small mask having a size of 3 × 3 pixels, a density difference occurs between the pixels, and the center pixel can be detected as a pole pixel.

Further, using the original digital multivalued data that has not been smoothed and the digital multivalued data that has been smoothed, pole detection is performed based on each data, and the logical sum (OR) of the detection results is obtained. As a result, it is possible to separate halftone dots having a wider number of lines with a mask of the same size. That is, an extreme pixel of a small halftone dot (for example, 100 lines or more) is detected from the original digital multilevel data, and a large halftone dot (for example, 65 to 100 lines) is obtained from the smoothed digital multilevel data. An extreme pixel can be detected.

[II] Regarding the extreme pixel detecting unit 2 The extreme pixel detecting unit 2 converts the density information in a predetermined local area of the input image, which is converted from the input image signal unit 1 into digital multivalued data and sent. The circuit detects the peak pixels of peaks and valleys, which are the points of change in the density in the local region, based on the detected values. That is, the density level of the central pixel in the predetermined local area and the density level of the predetermined surrounding pixel are compared, and the extreme pixel is determined based on the predetermined extreme detection condition.

For the sake of simplicity, a case where a peak pixel of a mountain giving the maximum level position of density is detected will be described as an example. In order to detect the peak pixel of the valley that gives the minimum level position of the density, it has the opposite property to the peak pixel of the peak. Therefore, the words in <> in the detection conditions described below may be adopted.

The condition that the pixel of interest is the peak pixel of the mountain is that the density level of the pixel of interest is higher than the density level of the surrounding pixels. Therefore, the following can first be adopted as the extreme point detection conditions.

When the density level of the central pixel is maximum <minimum> compared to the density level of the peripheral pixels in the predetermined local area, the central pixel is determined to be a pole pixel. 3A to 3D show specific examples of the extreme point detection conditions. FIG. 3A shows a 3 × 3 pixel size rectangular mask, FIG. 3B shows a cross-shaped mask, and FIG.
FIG. 3C shows a cross-shaped mask, and FIG. 3D shows a rectangular mask having a size of 5 × 5 pixels. In each mask, the hatched pixel is the central pixel. In these masks, when the density level of the central pixel is greater than any of the other pixels in the mask,
What is necessary is just to detect a center pixel as a pole pixel.

Further, the following can be adopted as other extreme point detection conditions. When the density level of the center pixel is larger than the threshold value TH1 (smaller) than the density level of the surrounding pixels in the predetermined local area, the center pixel is set as the peak pixel of the mountain.

FIG. 4A shows a specific example of the extreme point detection condition.
It is shown in (B). FIG. 4A employs a rectangular mask having a size of 3 × 3 pixels as a local area, and FIG.
A rectangular mask having a pixel size is adopted. In each of these masks, when the central pixel Lc satisfies the following conditions with respect to the predetermined pixels L1 to L8 around it, the central pixel Lc may be detected as a pole pixel.

Lc-L1> TH1 & Lc-L2> TH1 & Lc-L3> TH1 & Lc-L4> TH1 & Lc-L5> TH1 & Lc-L6> TH1 & Lc-L7> TH1 & Lc-L8> TH1

Further, the following can be adopted as other extreme point detection conditions. In the predetermined local area, the density level of the central pixel is the maximum <minimum> compared to the density level of the predetermined surrounding pixels, and the density level of the central pixel and the central pixel are located at the position of the point object. When the absolute value of the difference from the average value of the density levels of the other two pixels is larger than a predetermined threshold value TH2, the center pixel is set as the extreme pixel.

FIG. 5 shows a specific example of the extreme point detection condition. FIG. 5 employs a rectangular mask having a size of 3 × 3 pixels as the local area.
When c satisfies the following conditions with respect to the predetermined pixels L1 to L8 around it, the center pixel Lc may be detected as the extreme pixel.

Lc> L1 & Lc> L2 & Lc> L3 & Lc> L4 & Lc> L5 & Lc> L6 & Lc> L7 & Lc> L8 and | 2 × Lc−L1 + L8 |> TH2 & | 2 × Lc− L2 + L7 |> TH2 & | 2 * Lc-L3 + L6 |> TH2 & | 2 * Lc-L4 + L5 |> TH2

Further, the following can be adopted as other extreme point detection conditions. In a small area that is smaller than the predetermined local area and that includes the central pixel of the local area, the density level of the central pixel is smaller than or equal to the density level of the predetermined surrounding pixels, and the density level of the central pixel. Is smaller than the density level of a predetermined pixel in a local area excluding the small area by a predetermined threshold value TH3 or more, the central pixel is detected as a peak pixel of the mountain.

FIG. 6 shows a specific example of the extreme point detection condition. In the 5 × 5 pixel size mask of FIG. 6, when the center pixel Lc satisfies the following conditions with respect to the surrounding pixels L1 to L16, the center pixel Lc may be detected as a pole pixel.

Lc ≧ L1 & Lc ≧ L2 & Lc ≧ L3 & Lc ≧ L4 & Lc ≧ L5 & Lc ≧ L6 & Lc ≧ L7 & Lc ≧ L8 and Lc−L9> TH3 & Lc−L10> TH3 & Lc−L11 > TH3 & Lc-L12> TH3 & Lc-L13> TH3 & Lc-L14> TH3 & Lc-L15> TH3 & Lc-L16> TH3

By changing the values of the threshold values TH1, TH2, and TH3 under the above-described extreme point detection conditions in an adaptive manner according to the density level of the central pixel or the surrounding pixels, it is possible to change the value of the character or the like. It is possible to prevent the extreme pixel of the line drawing portion from being erroneously detected as the extreme pixel of the halftone portion, and it is possible to more reliably detect only the extreme pixel of the halftone portion. Next, a method for changing the threshold value adaptively will be described.

When two threshold values are prepared and the way of the peak of the peak and the valley, for example, the density gradient near the peak of the valley and the valley is different due to the characteristics of the scanner constituting the input image signal unit 1, etc. Use different thresholds for peaks and valleys.

In general, when a halftone dot portion is digitized, near the extreme point of the peak of the halftone dot portion, the larger the density level of the extreme point becomes <smaller>, the greater the difference in density level between the neighboring pixels becomes. There is. Therefore, using this property, the threshold may be changed according to the height of the mountain. For example, taking the above-mentioned extreme point detection condition as an example, two values A and B are set as the threshold value TH1.
(A> B) is prepared, and if Lc ≧ α, then TH1 = A, if Lc <α, then TH1 = B (α: switching judgment value), such as according to the density level of the center pixel Lc. The threshold value TH1 may be changed in two stages. Note that, instead of Lc in the above-mentioned determination formula, the average density level of the pixels around the central pixel, for example, the peripheral pixels L1 to L8 in the mask of FIG.
The threshold value TH1 can be switched using the average density level of.

Further, any two or more of the above detection conditions to are prepared.
After detecting the extreme pixel according to each detection condition, the extreme pixel may be detected by finally calculating the logical sum (OR) of the detection results. For example,
Taking FIG. 4 as an example, an extreme pixel of a small halftone dot is detected using a mask having a small pixel size of 3 × 3 in FIG.
An extreme pixel of a large halftone dot may be detected with a mask having a large pixel size of 5 × 5 in FIG. 4B, and a logical sum (OR) of these detection results may be obtained. As a result, it is possible to accurately detect an extreme pixel with respect to a halftone dot having a wide number of lines.

[III] Regarding Halftone Candidate Area Detection Unit 3 The halftone dot candidate area detection unit 3 determines a predetermined pixel in the small area based on the existence state of the peak pixels of the peaks and valleys in the predetermined small area. This is a circuit for determining whether the area is a halftone dot candidate area or a non-halftone dot candidate area. The process of detecting a halftone dot candidate region can be performed in block units or pixel units.

First, the following conditions can be adopted as the halftone dot candidate area detecting condition in the halftone dot candidate area detecting section 3. The peak pixels of the peaks and valleys detected by the pole pixel detection unit 2 are counted in a small area S of a predetermined size, and the total value of the peak pixels of the peaks and valleys is defined as the number P of the pole pixels of the small area.
When the number of pole pixels P is equal to or greater than a predetermined threshold,
-All pixels in the small area S are set as halftone dot candidate areas (in the case of processing in block units) or-The central pixel in the small area S is set as halftone dot candidate areas (in the case of processing in pixel units) .

FIG. 7A shows a specific example of this detection condition.
It is shown in (B). FIG. 7A shows an example in which processing is performed in block units, and FIG. 7B shows an example in which processing is performed in pixel units. When the above conditions are satisfied, the hatched pixels are determined as the halftone dot candidate areas.

Further, in order to minimize the influence of the area ratio of the halftone dots, the number of pole pixels P
Alternatively, the following extreme pixel number P obtained by or can be adopted.

The peak pixels at the peaks and the valleys are counted separately, and the larger value of the two count values is set as the number P of the pole pixels of the small area S. The peak pixels of the peaks and the valleys are counted separately, and the sum of the two count values is defined as the number P of the pole pixels of the small area S.

Further, in order to prevent erroneous detection of an extreme pixel in a photograph portion or a line drawing portion as a halftone dot candidate area, the property that the extreme portion pixels in the halftone portion exist uniformly and in large numbers is used. The obtained pole pixel number P can also be adopted.

The small area S is divided into a plurality of areas Si, and the number of extreme pixels is counted for each area Si. When the number of areas Si in which the number of extreme pixels is equal to or less than a predetermined value is equal to or more than a predetermined value, The number of extreme pixels in the small area S is set to zero, and when smaller than a predetermined value, the sum of the number of extreme pixels in each area Si is defined as the number P of extreme pixels in the small area S.

FIG. 8 shows a specific example of this detection condition. Figure
8 indicates that the small area S is divided into four areas S₁, S_Two, S_Three, S_FourTo
Divide and, for each region, the number of extreme pixels P₁, P_Two,
P_Three, P _FourAnd if any one of the counts is zero,
If the number of extreme pixels in the small area S is zero,
The number of extreme pixels in the area P₁, P_Two, P_Three, P_FourSubtotal of
It is assumed that the number S of pole pixels in the area S is P.

Further, the following conditions can be adopted as the conditions for detecting the pole candidate region. The small area S is divided into a plurality of areas Si, and the number of extreme pixels is counted for each area Si. When the number of areas Si in which the number of extreme pixels is equal to or less than a predetermined value is equal to or less than a predetermined value, Detect all the pixels in the area S as the halftone dot candidate area (in the case of processing in block units) or • Detect the central pixel in the small area S as the halftone dot candidate area (in the case of processing in pixel units).

A specific example of the detection condition will be described with reference to FIG. 8. If the small area B is divided into four areas S ₁ , S ₂ , S ₃ , and S ₄
Is determined for each of the divided areas. When the extreme pixels exist in all the areas S ₁ , S ₂ , S ₃ , and S ₄ , all the pixels in the small area S ( The halftone dot candidate area may be a pixel (in the case of processing in block units) or a center pixel (in the case of processing in pixel units).

[IV] Regarding the Halftone Region Detecting Unit 4 The halftone region detecting unit 4 is a circuit that finally detects a halftone region by using the detection result of the halftone candidate region detecting unit 3. The process of detecting a halftone dot region can be performed in block units or pixel units. That is,

In the case of processing in block units: The distribution state of the halftone dot candidate areas in the small area of interest and the surrounding small areas is determined, and it is determined whether the small area of interest is a halftone area. In the case of processing in pixel units, the distribution state of the halftone dot candidate region in the target pixel and its surrounding pixels is determined, and it is determined whether or not the target pixel is a halftone region.

When the processing is performed in units of blocks, the following dot area detection conditions can be specifically adopted. When the number of small areas determined to be a halftone dot candidate area is equal to or more than a predetermined number among the small area of interest and its surrounding small areas, all pixels in the area are regarded as halftone dot areas.

When processing is performed on a block basis, processing is performed by representing each small area with the center pixel of each small area, and it is determined whether or not the small area of interest is a halftone dot area according to the result. It may be determined. This makes it possible to simplify and speed up the processing.

FIG. 9 shows a specific example of this detection condition. FIG. 9 shows that, in the small area of interest and the small areas before and after it,
Pixels marked with “x” are selected as representative pixels of each area, and halftone dot areas are detected by determining whether or not each of these “x” pixels is a halftone dot candidate area. Then, for example, when all of the three X-marked pixels are halftone dot candidate areas, all hatched pixels in the small area of interest are determined as halftone dot areas.

Further, when processing is performed on a pixel-by-pixel basis,
Specifically, the following or the following dot area detection conditions can be adopted. When there are a predetermined number or more of pixels in the halftone dot candidate area among the pixel of interest and surrounding pixels, the pixel of interest is defined as a halftone dot area.

As the surrounding pixels, a pixel at an arbitrary position such as a pixel following the pixel of interest or a pixel at a predetermined distance from the pixel of interest can be adopted. FIGS. 10 and 11 show specific examples of the detection conditions.
Shown in FIG. 10 adopts pixels at discrete positions indicated by crosses, and when there are a predetermined number or more of pixels in the halftone region among these cross-point pixels, the hatched pixel of interest is determined as the halftone region. Is what you do. FIG. 11 employs pixels at consecutive positions indicated by crosses, and when there are a predetermined number or more of pixels in the halftone dot candidate region among the consecutive crosses, the hatched target pixel is replaced with a halftone dot region. Is determined.

Each of the examples shown in FIGS. 9 to 11 shows a case where the surrounding small area or surrounding pixels are accessed in the main scanning direction (horizontal direction). The access method may be determined according to the amount of the line buffer and the separation accuracy.

The specific examples of the circuits 1 to 4 in FIG. 1 have been described above. FIG. 13 is a flowchart of the processing operation of the halftone dot region separating apparatus of FIG. The outline of the operation of the apparatus in FIG. 1 will be described below with reference to this flowchart. In addition,
The processing conditions of the extreme pixel detection unit 2, the halftone dot candidate region detection unit 3, and the halftone dot region detection unit 4 in this flowchart are as follows (a), (b), and (c).

(A) The detection of an extreme pixel in the extreme pixel detection unit 2 is performed in an M × M block (for example, FIG.
(A) 3 × 3 pixel size mask). (B) The detection of a halftone dot candidate area in the halftone dot candidate area detection unit 3 is performed in block units using N × N blocks (N> M) (for example, the 4 × 4 pixel size in FIG. 7A). Small area S). (C) The detection of the halftone dot area by the halftone dot area detection unit 3 is performed by arranging K N × N blocks in the main scanning direction as shown in FIG.

When the processing is started, the input image signal unit 1 raster-scans the original image, converts the original image into digital multi-valued data, and stores it in an internal buffer. The extreme pixel detection unit 2 reads digital multivalued data for N lines required for processing from the input image signal unit 1 (step 3).
1) A mask of M × M pixel size, for example, a mask of 3 × 3 pixel size in FIG. 3A is sequentially applied to each pixel, and for example, the density level of the center pixel in the mask is compared with the density level of surrounding pixels. The maximum pixel or the minimum pixel is sequentially detected as the extreme pixel (step 3).
2).

Next, in step 33, the halftone dot candidate area detecting section 3 calculates the N-th pole pixels of the N lines detected as described above for every N × N blocks, for example, as shown in FIG.
The pole pixel number P is counted for each small area S (hereinafter, referred to as “block S”) of 4 × 4 pixel size in FIG. Then, in step 34, for example, a block S in which the obtained pole pixel number P is equal to or larger than a predetermined threshold value is detected as a halftone dot candidate area.

Next, the halftone dot area detecting section 4 executes step 3
In step 5, as shown in FIG. 13, K blocks S are arranged in the main scanning direction, and the number B of halftone dot candidate area blocks in the K blocks is obtained. Then, step S3
In 6, the block number B is equal to or a predetermined number B _TH or more, is greater than the predetermined number B _TH determines the target block and halftone dot portion (step 37), also given It is smaller than the number B _TH is the block of interest is determined as Hiami point unit (step 38).

By repeating the above processing (steps 39 and 40), all the pixels in the input image are separated into halftone portions and non-halftone portions. The halftone dot region separating apparatus of the present invention described above separates all the pixels in an image into a halftone dot portion and a non-halftone dot portion (= photo portion + line drawing portion). The halftone dot separating apparatus of the present invention and a known photographic area separating apparatus (for example, see Japanese Patent Application Laid-Open No.
By combining this with (3), a so-called three-region separating device for separating an image into three regions of a photograph portion, a halftone dot portion, and a line drawing portion can be easily configured. That is, to configure a three-region separating device, a halftone dot region separating device of the present invention and a known photographic region separating device are used, and (a) a halftone region separating device → a halftone portion (b) a photographic region separating device → Photograph part (c) Using the results of (a) and (b), a configuration may be adopted in which a non-halftone dot area and a non-photograph part is determined to be a line drawing part. The halftone dot separation process (a) and the photographic region separation process (b) can be performed in either parallel or serial processing.

FIG. 14 shows an example of a copying machine constructed using the three-region separating apparatus. In the figure, 11 is a three-region separating device, 12 is an MTF correction circuit for a line drawing region, 13 is a binarization processing circuit for a line drawing region, 14 is a smoothing processing circuit for a dot region, and 15 is a dot processing region. A vortex dither processing circuit emphasizing gradation, 16 is a vortex dither processing circuit emphasizing gradation for a photographic region, and 17 is a three-region separating device 1.
1 is an image signal selection circuit for selecting one of the processing signals of the processing circuits 13, 15, 16 according to the judgment output of 1.

The input image signal is supplied to the three-region separating device 11, M
The signals are input to the TF correction circuit 12, the smoothing processing circuit 14, and the dither processing circuit 16 in parallel. The three-region separating device 11 separates all pixels of the input image into halftone dots / photographs / line drawings, and sends the determination output to the image signal selection circuit 17.

The image signal selection circuit 17 selects the output image of the dither processing circuit 16 when the determination output of the three-region separating device 11 is a photographic area signal, and selects the output image of the dither processing circuit 16 when the determination output is a halftone area signal. An output image is selected, and in the case of a line drawing area signal, an output image of the binarization processing circuit 13 is selected and output. As a result, in the photographic part of the input image, the image subjected to the pseudo halftone processing by the vortex type dither processing circuit 16 is removed by the smoothing processing circuit 14 and the vortex type dither processing circuit 15 in the halftone dot part. In the line image area, an image whose edge is enhanced by the MTF correction circuit 12 and the binarization processing circuit 13 is output. Therefore, if an image is reproduced using the image signal output from the image signal selection circuit 17, a high-quality reproduced image obtained by performing image processing by the best method for each of the halftone, photographic and line drawing areas in the image Is obtained.

[0070]

According to the present invention, the extreme pixel is detected by comparing the density level of the central pixel with that of a predetermined peripheral pixel for each of a predetermined local area centered on the central pixel and a predetermined local area wider than the local area. Since the logical sum of the result is output as the extreme pixel, the extreme pixel can be accurately detected irrespective of the magnitude of the area ratio of the halftone dot portion or the arrangement state of the document.

[Brief description of the drawings]

FIG. 1 is a block diagram of a halftone dot area separating apparatus to which a halftone dot detecting apparatus according to the present invention is applied.

FIG. 2 is a diagram illustrating a state of halftone dots before and after smoothing.

FIG. 3 is a diagram illustrating an example of a local region and an extreme pixel.

FIG. 4 is a diagram illustrating an example of a local region and an extreme pixel.

FIG. 5 is a diagram illustrating an example of a local region and an extreme pixel.

FIG. 6 is a diagram illustrating an example of a local region and an extreme pixel.

FIG. 7 is a diagram illustrating an example of detection of a halftone dot candidate area.

FIG. 8 is a diagram illustrating an example of division of a small area.

FIG. 9 is a diagram illustrating an example of detection of a halftone dot region.

FIG. 10 is a diagram illustrating an example of detection of a halftone dot region.

FIG. 11 is a diagram illustrating an example of halftone dot region detection.

FIG. 12 is a flowchart of a processing operation.

FIG. 13 is a diagram illustrating an example of a method for counting halftone dot candidate areas.

FIG. 14 is a diagram illustrating an example of a copying machine configured using the apparatus of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Input image signal part 2 Pole pixel detection part 3 Halftone dot candidate area detection part 4 Halftone dot area detection part

Claims (10)

[Claims] A first extreme pixel detecting means for comparing a density level of a central pixel in a predetermined local region of an image converted into digital multi-valued data with a density level of a predetermined surrounding pixel to detect an extreme pixel; A second extreme pixel detecting means for comparing a density level of a central pixel in a local area wider than the predetermined local area with a predetermined peripheral pixel with respect to a pixel to detect an extreme pixel, and a first extreme pixel detecting means; An extreme pixel detection apparatus, comprising: a logical sum unit that outputs a logical sum of a detection result and a detection result of the second extreme pixel detection unit as an extreme pixel. 2. The method according to claim 1, wherein the predetermined local area of the first extreme pixel detecting means is an area composed of 3 × 3 pixels centered on the central pixel, and the wide local area of the second extreme pixel detecting means is selected. The extreme pixel detection apparatus according to claim 1, wherein is an area composed of 5 x 5 pixels. 3. The local area of the first extreme pixel detection means by masking the wide local area of the second extreme pixel detection means every other row and column pixels adjacent to the center pixel. The first extreme pixel detection means and the second extreme pixel detection means compare the density levels of predetermined surrounding pixels at the same position to detect an extreme pixel. 3. The extreme pixel detection device according to 1 or 2. 4. The extreme point detection condition is such that the central pixel is detected as an extreme pixel when the density level of the central pixel in the local area is maximum or minimum compared to the density level of a predetermined surrounding pixel. The extreme pixel detection apparatus according to claim 1, 2 or 3, wherein: 5. The extreme point detection condition is set such that the central pixel is defined as an extreme pixel when the density level of the central pixel in the local area is larger or smaller than a predetermined threshold value compared to the density level of a predetermined surrounding pixel. 4. The extreme pixel detection apparatus according to claim 1, wherein the detection is performed. 6. The extreme point detection condition is such that the density level of the central pixel in the local area is maximum or minimum compared to the density level of a predetermined peripheral pixel, and the density level of the central pixel and the central pixel are When the absolute value of the difference from the average value of the density levels of the other two pixels at the point target position is larger than a predetermined threshold value (TH2), the center pixel is detected as a pole pixel. The extreme pixel detection apparatus according to claim 1, 2 or 3, wherein: 7. A method according to claim 1, wherein the predetermined peripheral pixels to be compared with the central pixel are pixels arranged on a straight line passing through the central pixel and extending in the up, down, left and right directions.
6. The extreme pixel detection device according to any one of the above. 8. The extreme pixel detection apparatus according to claim 1, wherein the threshold value is changed according to the density level of the central pixel or the density level of predetermined surrounding pixels. . 9. The threshold value in the case of detecting a peak pixel of a mountain where the density level of a central pixel is higher than the density level of a predetermined surrounding pixel is changed so as to increase when the density level of the central pixel increases. The threshold value in the case of detecting the extreme pixel of the valley where the density level of the center pixel is lower than the density level of the predetermined surrounding pixels is changed so as to become smaller when the density level of the center pixel becomes smaller. 9. The extreme pixel detection apparatus according to claim 8, wherein: 10. The extreme pixel detecting apparatus according to claim 8, wherein an average value of density levels of predetermined surrounding pixels is obtained, and a threshold value is changed according to the average value.