JP3336046B2 - Image area separation device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image area separating apparatus for separating a character area with high precision.

[0002]

2. Description of the Related Art Conventionally, an image is divided into image areas and adaptively divided into two.
In the case of performing the binarization processing, generally, an “image area separation method for a mixed image of characters / patterns (dots, photographs)” (Transactions of the Institute of Electronics, Information and Communication Engineers, Vol.
47, January 1992), the text area and the picture area are separated, for example, MTF correction is applied to the text area, then binarized with a fixed threshold, and the dither processing is applied to the picture area. Alternatively, binarization processing by the error diffusion method is performed to achieve both character resolution and picture gradation.

[0003]

However, the above-described image area separation method has a drawback that a grid-like pattern that should be originally determined as a character area is erroneously determined as a picture area. In other words, as shown in FIG. 19, the "+" part and the center part of the "□" of the lattice pattern are set as the halftone dot peak pixels (the density level of the center pixel is higher or lower than that of the surrounding pixels). It is erroneously detected, and the entire grid pattern is erroneously determined to be a halftone dot region having a high peak pixel density, and a process for a picture is performed.

The same processing as described above is performed on the image shown in FIG. 20 where the lines are dense (depending on the SN of the scanner). That is, in a part where the lines are dense, one line cannot be read at a certain uniform density level.
In some cases, the ridge portion of the line shown in FIG. 20 is erroneously detected as a peak pixel. by this,
The entire area of such a dense pattern of lines is erroneously determined to be a halftone dot area having a high peak pixel density, and a process for a picture is performed.

Further, the present applicant has previously proposed an image area separating apparatus which combines edge detection and white background detection (Japanese Patent Application No. 3-35752). In this image area separation device,
If the target pixel is an edge region and a white background exists in the vicinity, the target pixel is determined to be a character region.

This device is provided with a white background area detection circuit, and a white background detection pattern shown in FIG. 21C is prepared in the circuit. Then, depending on the size of the white pattern to be prepared (the larger the size of this pattern), the above-mentioned “lattice pattern” (FIG. 21)
(A)) and “dense lines” (FIG. 21 (b))
Since it is determined that there is no white background, there is a problem that the image is erroneously processed as a pattern.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an image area separating apparatus in which the accuracy of separation is improved by determining a dense portion of a grid pattern and lines as a character area.

[0008]

In order to achieve the above object, according to the first aspect of the present invention, a target pixel of an image has an edge.
First detecting means for detecting whether the area is an area or not;
Second detection means for detecting whether or not the eye pixel is a halftone dot area
And detecting whether the target pixel is a bar-shaped region.
Third detecting means and the first detecting means detect an edge area.
And the second and third detecting means respectively detect a halftone area.
When the bar-shaped area is not detected, the target pixel is set to the character area.
And that the third detection means has detected the rod-shaped region.
Irrespective of the detection results of the first and second detection means.
Determining means for determining the target pixel as a character area;
It is characterized by.

According to the second aspect of the present invention, a pixel of interest of an image
Detecting means for detecting whether or not is an edge area
And a second step for detecting whether or not the pixel of interest is a white background area.
4 means for detecting whether the pixel of interest is a bar-shaped area
Detecting means for detecting the first and fourth detection means,
The step detects an edge area and a white background area, respectively, and
When the detecting means does not detect the rod-shaped area, the target pixel
Is determined to be a character region, and the third detection means
When detecting, the detection results of the first and fourth detecting means
Determining means for determining the pixel of interest as a character area regardless of
It is characterized by having.

[0010]

The edge region detecting section detects continuous portions of white pixels and continuous portions of black pixels, and outputs a portion where both are present as an edge region. The halftone dot region detecting portion detects a peak pixel. Then, an area having a peak pixel density equal to or higher than a predetermined density is output as a halftone area, and the rod-shaped area detection unit detects a rod-shaped pattern by pattern matching. When the bar-shaped pattern is detected by the bar-shaped region detection unit,
The target pixel is determined to be a character area regardless of the detection results of the edge area detection unit and the halftone dot area detection unit. As described above, when information with a high probability of being a character area is obtained,
Since the information is used with priority and other information is not used, the character area can be separated with high precision.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram of a first embodiment of the present invention. The image area separating apparatus according to the present invention mainly includes four processing blocks. That is, the edge area detection unit 1, the halftone dot area detection unit 2, the bar-shaped area detection unit 3,
And it is composed of a comprehensive judgment section 4.

Hereinafter, each component will be described in detail. <Edge Area Detecting Unit> This detecting unit is described in, for example, the above-mentioned paper (p4
2 (see FIG. 10). FIG. 2 is a block diagram of the edge area detection unit. In the ternarization circuit 11, the input image data is subjected to edge enhancement, and then ternarized using two types of fixed thresholds. Black pixel continuity detector 12
And the white pixel continuity detecting unit 13 detects an area in which black and white pixels after ternarization are continuous by pattern matching.

FIG. 3 shows a pattern for detecting continuity of black pixels, and FIG. 4 shows a pattern for detecting continuity of white pixels. A continuous portion of white pixels and a continuous portion of black pixels matching these patterns are detected, and the AND circuit 14 outputs a portion where both pixels exist as an edge region.

<Dot Area Detecting Unit> This detecting unit is also described in, for example,
41 (see FIG. 6). FIG. 5 is a block diagram of the halftone dot area detection unit. Peak pixel detection unit 21
Then, a peak pixel as shown in FIG. 6 is detected. Here, the peak pixel means that the density level L of the center pixel is higher or lower than that of all surrounding pixels, and
The density levels a and b of the diagonal pixels located diagonally across L and the center pixel are | 2 × Lab−> TH for all four pairs.
When (fixed threshold), the central pixel is set as the peak pixel.

The dot area detection circuit 22 calculates 4 × 4 pixels as 1
If there are two or more blocks including a peak pixel in four blocks, the block of interest is determined as a halftone dot candidate area, and the halftone area correction circuit 23 determines the nine blocks around the block of interest. If four or more blocks are halftone dot candidates (that is, a halftone dot region having a certain size exists), the target block is set to a halftone dot region; otherwise, the target block is set to a non-halftone dot region. The correction process is performed.

As described above, a "lattice pattern" or a "dense line" often matches such a peak pixel.

<Rod Region Detection Unit> Before describing the details of the rod region detection unit, experimental facts will be described first. When a document image is read by a 400 dpi, 6-bit scanner (for example, a monochrome luminance signal, a color green signal), FIG.
A bar-shaped pattern (1) as shown in (b), (c), and (d)
(A bar-shaped pattern with a pixel width) rarely appears in a picture (photograph, halftone dot).

Furthermore, if the pattern is limited to vertical and horizontal patterns, there is almost no matching in the picture. What is particularly problematic in the bar pattern detection in FIG. 7 is a halftone original. However, since a halftone dot generally has a specific screen angle, the problem of frequently matching a bar pattern is greatly reduced if the halftone dot is limited to the vertical and horizontal directions.

For the same reason, if a "+" pattern as shown in FIG. 8 is prepared, only the grid pattern of the line drawing can be detected, and it is not erroneously detected in the picture. In the example of FIG. 7, the pattern is a bar pattern having a length of 5 pixels. By making this a longer bar pattern such as 7 pixels or 9 pixels, the pattern is matched with such a pattern in the picture. The number of locations is reduced (however, the cost increases by preparing a long bar pattern such as 7 pixels or 9 pixels).

The vertical and horizontal bar patterns shown in FIG.
A photograph or a halftone dot document that matches the "+" character pattern is, for example, a "sign written on a white signboard" in a picture.
Is a vertical and horizontal bar on a white background, and is therefore a place where character processing is originally desired.

From the above background, in the present invention, as shown in FIGS. 7 and 8, the region belonging to the bar pattern in the vertical and horizontal directions (ie, the target pixel and its neighboring pixels in FIGS. 7 and 8) Is based on the concept of determining a character area regardless of other information, for example, a detection result of a halftone dot area.

FIG. 9 is a block diagram of the bar-shaped area detecting section. The binarization circuit 31 binarizes the signal from the scanner with a fixed threshold. Note that MTF is used as preprocessing.
It is also effective to apply a correction filter (for example, the filter shown in FIG. 11 of the above-mentioned paper). Further, detecting a black pixel and a white pixel by performing ternarization using two different thresholds can also be expected to have an effect of reducing erroneous determination in a picture (especially a photograph).

In the pattern matching circuit 32, FIG.
A bar-shaped pixel is detected using the “|” pattern in (a), the “−” pattern in (b), the “+” pattern in FIG. 8, and the patterns shown in FIGS. 10 (a) to (d). In addition, these patterns are patterns for detecting fine lines on a white background, but by reversing the black and white of the prepared pattern,
White characters in solid black can also be detected. When the pattern matches, the target pixel is output as a bar-shaped pixel.

In the expansion circuit 33, the pixel area subjected to the pattern matching as described above is divided into 5 × pixels as shown in FIG.
Inflate with a 5 mask. That is, 1 in a 5 × 5 mask
If there is even a bar-shaped pixel, the target pixel is set to a bar-shaped pixel.

The reason why such expansion processing is performed is that the pixel position to be matched by the pattern matching circuit 32 is the part of the back (ridge) of the character, and the part that is actually determined to be a character area is the character edge. The expansion process as described above is required.

<Comprehensive Judgment Unit> In the overall judgment unit 4, the following three detection results are obtained for each pixel: edge (1) / non-edge (0) in the edge area detection unit; Non-dot (0) The attribute (character area / picture area) of the pixel of interest is determined based on the rod shape (1) / non-rod shape (0) in the rod-shaped region detection unit.

FIG. 12 shows a determination table. The feature of this table is that if the detection result of the bar-shaped area detection unit is a bar (output signal is 1), the pixel of interest is determined to be a character area regardless of other detection results.

As described above, according to the present invention, when determining the character area / picture area, the information with high accuracy from the rod-shaped area detecting section is used with priority, so that the character area is separated with high accuracy. be able to.

FIG. 13 is a block diagram showing another embodiment of the present invention. The image area separating apparatus of this embodiment is obtained by replacing the halftone dot area detecting section of the above-described embodiment with a white area area detecting section 5, and the edge area detecting section and the bar-shaped area detecting section are the same as those described above. Therefore, the white background area detection unit 5 and the comprehensive judgment unit 4 will be described.

As the detection unit, for example, the detection unit previously proposed by the present applicant (Japanese Patent Application No. 3-35752) is used. FIG. 14 shows the configuration of the first embodiment of the white background area detection unit 5. The white background area is an area where a cluster of white pixels of a predetermined size (1 × 5 or 5 × 1) exists near the target pixel.

After the input image signal is sharpened by the MTF correction circuit 51, the image signal is binarized to white / non-white by a predetermined threshold value by a binarization circuit 52, and is input to a white pixel block pattern matching circuit 53. . FIG. 15 shows a white pixel block pattern in which 1 × 5 or 5 × 1 are all white pixels.
The white pixel block pattern matching circuit 53 detects a block of white pixels near the pixel of interest by pattern matching.
When matching with the pattern shown in FIG. 15, the target pixel is set as the active pixel. The expansion circuit 54 counts the number of active pixels in, for example, a 5 × 5 block centered on the target pixel from the output of the white pixel block pattern matching circuit 53, and if there is at least one active pixel in the block, The entire block is determined and output as a white background area.

FIG. 16 shows a second embodiment of the white area detection section. In this embodiment, a correction circuit is provided. Other configurations are the same as those of the first embodiment. Input image signal is M
After the sharpening process by the TF correction unit, the image signal is binarized by a binarization circuit into white / non-white using a predetermined threshold value, and is input to a white pixel block pattern matching circuit. In the white pixel block pattern matching circuit, a block of white pixels near the target pixel is detected by pattern matching, and when the white pixel block pattern matches the pattern shown in FIG. 15, the target pixel is set as a temporary active pixel.

After detecting a cluster of white pixels, the correction circuit
As shown in FIG. 17, when there are clusters of white pixels A and B at a distance L pixels left and right from the target pixel in the main scanning direction, the target pixel is output as a true active pixel. I do. The expansion circuit counts the number of active pixels in, for example, a 5 × 5 block centered on the target pixel, and if there is at least one active pixel in the block, determines and outputs the entire block as a white background area.

As described above, the definition of the white background (cluster of white pixels) (in the above example, the explanation is made with 5 × 1 or 1 × 5, but in practice, unless a larger mask is prepared) Depending on the situation, the “grating pattern” or “dense line” may be matched.

<Comprehensive Judgment Unit> In the overall judgment unit 4, the following three detection results are obtained for each pixel: Edge (1) / non-edge (0) in the edge area detection unit White (1) / non-white background in the white area detection unit (0) The attribute (character area / picture area) of the pixel of interest is determined based on the rod shape (1) / non-rod shape (0) in the rod-shaped region detection section.

FIG. 18 shows a determination table. The feature of this table is that, as in the previous table, if the detection result of the bar-shaped area detection unit is a bar (output signal is 1), the pixel of interest is determined to be a character area regardless of other detection results. Is a point.

[0037]

As described above, according to the present invention , when the information that the character area is obtained is obtained with certainty, the information is given priority and other information is not used. Good separation can be performed. In addition, since they are separated by local processing, they can be realized by a small-scale circuit configuration.

According to the present invention , by performing pattern matching using a pattern such as a bar-shaped pattern or a "+" character pattern, information that a character area is definitely obtained, the information is given priority, and other information is used. As a result, character area separation with high precision is possible.

According to the present invention , when the information of the character area is surely obtained, the information of the halftone dot separation is not used. , Dense lines) can be covered, and accurate separation can be achieved.

According to the present invention , when the information of the character area is surely obtained, the information of the white background separation is not used. , Dense lines) can be covered, and accurate separation can be achieved.

[Brief description of the drawings]

FIG. 1 is a block diagram of a first embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration of an edge area detection unit.

FIG. 3 shows a pattern for detecting continuity of black pixels.

FIG. 4 shows a pattern for detecting continuity of white pixels.

FIG. 5 is a block diagram of a halftone dot area detection unit.

FIG. 6 is a diagram illustrating a peak pixel.

FIGS. 7 (a), (b), (c) and (d) are examples of a bar-shaped pattern.

FIG. 8 is an example of a “+” character pattern.

FIG. 9 is a block diagram of a bar-shaped area detection unit.

FIGS. 10 (a), (b), (c) and (d) are other examples of a bar-shaped pattern.

FIG. 11 is a 5 × 5 dilation mask.

FIG. 12 is a comprehensive judgment table according to the first embodiment.

FIG. 13 is a block diagram of another embodiment of the present invention.

FIG. 14 shows a configuration of a first embodiment of a white background area detection unit.

FIG. 15 shows a white pixel block pattern in which all 1 × 5 or 5 × 1 are white pixels.

FIG. 16 shows a second embodiment of the white background area detection unit.

FIG. 17 is a diagram illustrating detection of a cluster of white pixels at a distance left and right from a target pixel by L pixels.

FIG. 18 is a comprehensive determination table according to the second embodiment.

FIG. 19 is an example of a grid-like pattern that is erroneously determined to be a dot area.

FIG. 20 is an example of a dense pattern of lines erroneously determined as a halftone dot area.

21A is an example of a lattice pattern that is erroneously determined to have no white background, FIG. 21B is an example of a dense pattern of lines that are erroneously determined to have no white background, and FIG. This is a white background pattern.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Edge area | region detection part 2 Halftone dot area | region detection part 3 Rod-shaped area | region detection part 4 Comprehensive judgment part 5 White background area detection part

──────────────────────────────────────────────────続 き Continued on front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N 1/40

Claims (2)

(57) [Claims] 2. A method according to claim 1, wherein said pixel of interest is an edge area.
First detecting means for detecting whether the pixel of interest is a halftone dot
Second detection means for detecting whether or not the area is
Third detection means for detecting whether or not a pixel is a bar-shaped area
The first detecting means detects an edge area, and
Second and third detecting means detect the dot area and the bar area, respectively.
If not, the target pixel is determined to be a character area,
When the third detecting means detects the rod-shaped area, the first,
The target pixel is sentenced regardless of the detection result of the second detection means.
Determining means for determining a character area
Image area separation device. 2. Whether or not a pixel of interest in an image is an edge area
First detecting means for detecting whether the target pixel is a white background area.
A fourth detecting means for detecting whether or not the area is
Third detection means for detecting whether or not a pixel is a bar-shaped area
And the first and fourth detecting means are respectively an edge area and
The white area is detected, and the third detecting means detects the rod-shaped area.
If not, the target pixel is determined to be a character area,
When the third detecting means detects the rod-shaped area, the first,
Regardless of the detection result of the fourth detection means, the pixel of interest is
Determining means for determining a character area
Image area separation device.