JPH02248160A - Dot area separation system - Google Patents

Abstract

PURPOSE:To avoid mis-separation near a dot area ratio 50% with a certain line number whose pole detection has been difficult and to separate a dot area and a line picture efficiently by detecting a part of a specific area rate of a dot. CONSTITUTION:An input picture signal section 1 converts an input picture signal into a digital multi-gradation input signal and sets the signal by M1 lines. A dot detection section 2 detects the pole of a picture element and sets the signal by N1 line to the detection section 4 to detect a dot candidate block. Moreover, the signal section 1 sets the input picture signal by M2 line and an area rate 50% detection section 3 detects the picture element whose area rate is 50% and sets the signal by N2 line to a 1st dot area detection section 4. The detected dot block is ORed by a 2nd dot area detection section 5 to obtain s dot candidate block number P in plural blocks Bo-Bn each consisting of N1XN1 picture elements and whether or not the number P is a prescribed number Po or more is discriminated. When the number is Po or over, a signal taking the center picture element of the noted block Bo as a character is outputted. The processing above is repeated and all picture elements of the input picture are separated into the dot part and the line picture part.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention improves the quality of the reproduced image when reproducing an image in which a halftone image such as a photograph or picture expressed by halftone dots and a line drawing such as text is mixed. The present invention relates to a halftone dot area separation method for separating a halftone dot area and a line drawing area and performing different image processing for each in order to improve image quality.

[Conventional technology]

In order to improve the quality of reproduced images in copying machines and facsimile machines, halftone areas such as halftone photographs and line drawing areas such as characters are separated, and the halftone areas are subjected to moiré removal processing and the line drawing areas are processed. It is desirable to apply a sharpening process to the image. Furthermore, even when transmitting an image containing a mixture of halftone drawings and line drawings, encoding processing is performed after processing according to the characteristics of each image area in order to improve the data compression rate. This is desirable.

One method for separating halftone areas from an image containing a mixture of halftone drawings and line drawings is the ``Reproduction of halftone photographs using a dot printer (Oki Electric R&D No. 1)'' proposed by Ueno.
32 Vol. 1.53 No. 4). This method focuses on the fact that the image signal level of a halftone photograph changes periodically, and uses a simple method to microscopically detect the halftone dot area and halftone period from the interval between the poles of the periodically occurring image signal. However, in the halftone dot area, smoothing is performed by averaging the distance between the poles to suppress the occurrence of moiré.

That is, as shown in Fig. 7, the processing flow is as follows: First, the image data of the original image is extracted as a digital multi-tone signal by rask scanning, and the difference in brightness and darkness between adjacent pixels on the rask is calculated. A difference signal is generated, and from this difference signal, a pixel that does not satisfy any of the following conditions (1) to (2) is detected as an extreme point indicating a peak or valley of density change, as shown in FIG.

(2) The pixel where the sign of the difference changes before and after is set as the extreme point (Fig. 8(a)).

(2) The pixel where the sign of the difference value changes before and after the difference is zero is set as the extreme point (FIG. 8(b)).

(2) A pixel whose distance from the previous pole reaches a predetermined threshold is set as a pole.

The processing interval is defined as the interval between consecutive poles from the pole obtained in this way, and the following conditions
When (2) is at the same time full length, it is determined to be a halftone dot area, and this area is used as the area to be smoothed.

■ The section length L(+) between the poles is a predetermined threshold THI,
, TH2 when the following formula is satisfied.

Tl+ 1 < L (i) < Tl+ 2 ■ The difference between the current section length L (i) and the previous section length L (i-1) is the predetermined threshold value TI (for 3, the following formula % formula % Based on this determination result, for the sections to be smoothed, the average value of that section is used as the image signal of the section, and for the sections that are not to be smoothed, no averaging is performed.The processed image obtained in this way is
Value.

According to this method, in halftone photographs, extreme points occur at a period corresponding to the halftone dot period, so by averaging over this interval, halftone dots can be removed and moiré can be suppressed. Furthermore, since the edge portion is determined to be a non-halftone area, the edge portion is not smoothed, and blurring of the edge portion can be suppressed and moire can be suppressed.

[Problem to be solved by the invention]

The above-mentioned conventional method separates halftone dots from an image on the premise that peaks and valleys of density levels regularly appear in the halftone dot area. Since there are many extreme points in a continuous tone photographic area, the above-mentioned conventional method has the disadvantage that a high separation rate cannot be expected.

In addition, since the conventional method performs detection by one-dimensional comparison of pixels before and after arranged on the rask scan line,
In halftone areas with a low halftone ratio, high halftone ratios, or halftone areas where the screen angle deviates from the horizontal direction due to rotation of the document, the length of the interval between poles becomes longer. There is also the problem that it becomes difficult to separate the text from the character part.

Therefore, in order to ensure the separation between the halftone dot area and the line drawing area, the applicant previously proposed an algorithm that detects the peaks and valleys of each halftone dot using a local two-dimensional pattern (particularly Application No. 63-268312: Title of invention: ``lr halftone area separation method''). As shown in FIG. 9, this method is based on the density level of a center pixel Lc and two other pixels La and Lb located at symmetrical positions with respect to the center pixel Lc within a 3×3 pixel mask, for example. When the absolute value of the difference between the density level and the average value of When the following is satisfied, the center pixels I and c are detected as the extreme points.

Although this method can improve the separation rate between halftone dot areas and line drawing areas, the following disadvantages still remain. In other words, the number of scanners used in this method was 400.
dots/1nch, the number of separable halftone dots is 1
00 to 175 lines, however, it may become impossible to detect the pole in a portion where the area of a certain number of lines is around 50%. The reason for this is that, as shown in Figure 10, there are two
This is because the density levels are approximately equal in units of ×2 pixels, and the diagonal directions are connected with approximately equal density.

The present invention has a dot area ratio of 5 for a certain number of lines where it becomes difficult to detect the poles.
The purpose is to eliminate erroneous separation near 0%.

[Means to solve the problem]

The present invention binarizes a human image signal using a fixed threshold, counts the number of pixels in the local region S1 that has a density ff equal to or higher than the fixed threshold, and if the counted value is within a predetermined range, the The center pixel of area S1 is detected as a halftone dot pixel constituting a part of a halftone dot, and the detected halftone dot pixels are counted in local area 52 (31<S2) as a unit, and based on the counted value, the above-mentioned local The entire area S2 or the center pixel of the local area S2 is detected as a halftone area.

[For production]

The number of pixels having a density level equal to or higher than a fixed threshold is counted using the local two-dimensional area S1, and if the counted value is within a predetermined threshold range, the central pixel is expressed around a predetermined area ratio. This detected pixel is further counted within the local two-dimensional area S2 (Sl<S2), and depending on the counted value, the detected pixel is detected as a pixel in the entire area S2 or the area S
By determining the center pixel of No. 2 as a halftone dot area or a line drawing area, it is possible to eliminate erroneous separation near a predetermined area ratio of a certain number of lines, which makes it difficult to detect a pole.

〔Example〕

FIG. 1 is a block diagram showing an embodiment of a halftone dot area separation device constructed by applying the method of the present invention.

In FIG. 1, an input image signal section 1 generates a digital multi-gradation input image signal consisting of a luminance signal corresponding to the density level by mask-scanning an original image in which a halftone image such as a halftone photograph and a line image such as characters are mixed. This circuit stores at least the scan lines necessary for subsequent separation processing in a line memory or the like. If the original image is monochrome, a luminance signal is input, and if it is color, each color signal after color separation is input.

The pole detection unit 2 sequentially applies a mask consisting of predetermined MlXM and pixels to each pixel of the input image signal sent from the input image signal unit 1, and detects when the center pixel of the mask detects the peak or valley of the density change. It is detected from the density relationship with surrounding pixels whether or not it is the extreme point shown.

The area ratio 50% portion detection unit 3 detects pixels expressing the area ratio of halftone dots around 50% from the input image signal. The number of pixels having a density value is counted, and if the counted value is within a predetermined range, the central pixel is detected as a pixel expressing an area ratio of around 50%.

The first halftone area detection unit 4 detects the peak pixels indicating the peaks and the peak pixels indicating the valleys detected by the pole detection unit 2 in units of blocks each having a size of Nl XN pixels (Ml <Nl). The block is counted, and the larger count value is set as the count value of this block. When the calculated count value satisfies a predetermined condition, this block is set as a halftone dot candidate block. Furthermore, the pixels expressing the vicinity of 50% area ratio of halftone dots detected by the 50% area ratio detection unit 3 are expressed as N2XN, pixels (M 2 <
A block having a size of N 2 ) is counted as a unit, and when the counted value satisfies a predetermined condition, this block is determined as a halftone dot candidate block. Note that in this embodiment, for convenience of explanation, NI=N2.

When moving the process pixel by pixel, the second halftone area detection unit 5 selects the center pixel of the block of interest as a halftone point when the number of halftone candidate blocks among the block of interest and surrounding blocks is greater than or equal to a predetermined number. If processing is performed block by block, pixels in the block of interest are determined to be halftone dots when the number of halftone candidate blocks among the block of interest and surrounding blocks is equal to or greater than a predetermined number.

The output image signal unit 6 is a circuit that outputs the output of each pixel as a signal corresponding to a halftone dot or a signal corresponding to a line drawing based on the halftone dot area detection results detected by the halftone dot area detection units 4 and 5.

Next, the operation of this embodiment will be explained with reference to the flowchart shown in FIG.

The input image signal unit 1 masks-scans the original image and converts it into a digital multi-gradation input image signal consisting of a luminance signal, and sets this image signal for M lines in order to perform the pole pixel detection process (step SL). ).

The pole detection unit 2 applies a 3×
Masks consisting of three pixels (M+=3) are applied sequentially, and the center pixel m of each mask is applied. It is detected whether or not is the extreme point of concentration change (step S2).

The conditions for detecting this pole are: ■ Center pixel m. The density level of other pixels m1 to m
maximum or minimum compared to the concentration level of s.

■ Center pixel m. and the density level of this central pixel m. The other two pixels m a at symmetrical positions with
When the concentration level of m b (m+, mn, mz. m7, ...) has a relationship of 2mo - ma - mb ≧thl.

This is when both conditions are satisfied. .

Next, the poles detected in this way are set in the first halftone dot area detection section 4 for N1 lines (step S3), N,
Each block consisting of N pixels is counted as a unit, and the poles indicating a mountain and the poles indicating a valley are counted separately, and the larger count value is determined as the count value C1 of this block (step S4), and C3
If the condition ≧th5 is satisfied, this block is set as a halftone dot candidate block (step S5).

In addition, the input image signal section 1 sets the input image signal for M2 lines in order to perform the detection process of the 50% area ratio section (
Step S6).

The area ratio 50% detection unit 3 counts the number of pixels having a density level equal to or higher than a threshold th3 within a mask of M2×M2 pixels near the pixel of interest, and when the counted value C2 is within a predetermined threshold th4 to th5. For example, if the count value C2 is 12 or 13 in the 5×5 pixel (M2=5) mask shown in FIG. S7).

Next, the pixels detected in step S7 are set in the first halftone area detection unit 4 for N2 lines (step S8),
A block consisting of N2×N2 pixels is counted as a unit to obtain a count value C3 (step S9). Count value C3 is C3
If the condition ≧th5 is satisfied, this block is designated as a halftone dot candidate block (step 510).

Next, the halftone dot candidate blocks detected in step S5 and step SIO are ORed by the second halftone dot area detection unit 5 (step 511), and as shown in FIG.
1 block N+ xN, multiple blocks B each consisting of pixels
. - The number P of halftone candidate blocks in B8 is determined (step 512), and the number P of blocks is a predetermined value P. It is determined whether or not this is the case (step 513). The number of blocks P is a predetermined value P.

If the above is the case, the block of interest Bo is output from the output image signal unit 6.
Step 514) outputs a signal indicating that the center pixel of the block is a halftone dot part, and if the value is PG or less, the target block B. Outputs a signal indicating that the center pixel of is the character part (step 5
15).

Such halftone dot separation processing is repeatedly executed for all pixels of the input image, and all pixels of the input image are separated into halftone dot portions and line drawing portions.

In the above embodiment, the original image is composed of a halftone part such as a halftone photograph and a line drawing part such as characters, but in reality, the original image includes a continuous tone photograph. There may be cases where continuous tone parts such as images and pictures are mixed. In such a case, the method of the present invention may be applied after first differentiating the input image signal and removing the continuous tone portion from its edge density (see, for example, Japanese Patent Laid-Open No. 115975/1982). By doing so, it becomes possible to separate the halftone dot area, two character areas, and the continuous tone area.

In addition, in the above-mentioned embodiment, the polar point detection method and the halftone area ratio of 50% are used.
Although the latter method is used in combination with the partial detection method, it is also possible to separate limited halftone dot areas by using only the latter method.

For example, within a 5×5 pixel mask, the above-mentioned count value C
If 2 is set to 5, an area ratio of 20% will be detected.

FIG. 6 shows an example of a copying machine configured using the above-mentioned dot area separation device, and in the figure, the dot area separation device 10 corresponds to the above-mentioned dot area separation device (FIG. 1). . This device further includes a sharpening processing circuit 11 for character area processing, a Bayer-type dither processing circuit 12 that emphasizes resolution in order to sharpen characters, and a smoothing processing circuit 1 for halftone area processing.
3. A spiral dither processing circuit 14 that emphasizes gradation to produce halftones in halftone photographs, etc., whichever dither processing circuit 12 or 14 receives the halftone determination signal from the halftone area separation device 10. The image signal selection circuit 15 selects one of the output signals.

The input image signal is judged by the halftone dot area separation device 10 for each pixel as to whether it belongs to a halftone dot area or not, and the result of the judgment is sent to the image signal selection circuit 15. The image signal selection circuit 15 selects and outputs the image signal on the dither processing circuit 12 side when the judgment signal is not a halftone dot area signal, and outputs the image signal on the dither processing circuit 12 side when the judgment signal is a halftone dot area signal. Select and output an example image signal.

As a result, in the line drawing area, image signals such as characters that have been sharply processed by the dither processing circuit 12 are selected and output, and in the halftone area, image signals such as halftone photographs that have been subjected to pseudo halftone processing in the dither processing circuit 14 are selected. Output. Therefore, if the image signal outputted from the image signal selection circuit 15 is subjected to necessary processing and then copied and reproduced, line drawings such as characters will be made clearer, and halftone portions such as halftone photographs will have a natural appearance. A high-quality binary image subjected to pseudo-halftone processing is obtained.

〔Effect of the invention〕

According to the present invention, by detecting a part of a certain area ratio of halftone dots, it is possible to eliminate the erroneous separation around 50% of the halftone area ratio of a certain number of lines, which was difficult to detect the extreme point, and to It becomes possible to efficiently separate the line drawing area.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing one embodiment of the present invention, and FIG.
Figure 3 is a diagram showing M, XM, and pixel masks, Figure 4 is a diagram showing M2
A diagram showing a ×M2 pixel mask, FIG. 5 is an example of a matrix consisting of multiple blocks, FIG. 6 is a configuration diagram of a copying machine to which the present invention is applied, FIG. 7 is an explanatory diagram of a conventional method, and FIG. 8 is a conventional method. Figure 9 is a diagram illustrating an example of polar point detection using the conventional method. Figure 10 is a diagram showing an MXM pixel mask using the conventional method.
The figure shows an example in which the conventional method cannot detect a pole. Patent applicant Riko Co., Ltd. MIXM 1 pixel mask Figure 3 k)

Claims (1)

[Claims] Binarize the input image signal with a fixed threshold, count the number of pixels in the local area S1 that have a density value equal to or higher than the fixed threshold,
If the count value is within a predetermined range, the center pixel of the local area S1 is detected as a halftone pixel forming a part of a halftone dot, and the detected halftone pixel is further added to the local area S2 (
A halftone dot region separation method, characterized in that the whole local region S2 or the center pixel of the local region S2 is detected as a halftone dot region based on the counted value (S1<S2).