JP3093235B2 - Image processing method and apparatus - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for optimally processing and reproducing an image in which photographs, halftone dots, and line drawings are mixed for each area.

[Conventional technology]

In general, an image called a document image contains a mixture of continuous tone photographs, halftone dots of black and white dots, and line drawings such as characters and lines. When such an image is reproduced by a digital copier or facsimile, etc., to improve the image quality of the reproduced image, the line drawing must be reproduced clearly so that the lines are not broken. For halftone dots to be reproduced with emphasis on tone (gradation), it is desirable to remove moiré and reproduce with emphasis on gradation.

As one method for achieving the above object, for example, there is a method of performing the following image processing. That is, the line drawing area is subjected to the binarization processing after the edge enhancement.

A vortex-type dither process, which is one of the systematic dither methods, is applied to the photograph region.

The halftone dot region is subjected to the same spiral dithering process as after the smoothing process.

[Problems to be solved by the invention]

By the way, in order to realize the above method, it is necessary to separate all pixels of the input image into a photograph area, a halftone dot area, and a line drawing area. However, regarding the separation of characters, most of the related art targets characters on a white background with no background, and it is difficult to accurately separate characters existing on halftone dots. Therefore, there is a problem in that the above-described adaptive processing is not performed on characters existing on the halftone dots, and the reproduced image becomes unclear.

On the other hand, as one method of performing image processing without separating an input image into regions as described above, an error between image information of a target pixel and image information of a target pixel after dither processing is distributed to peripheral pixels. A so-called average error minimization method for performing dither processing according to the above-mentioned method, or a method using an error diffusion method similar thereto. This method is characterized in that a reproduced image at a certain level can be obtained without separating an input image into regions. Characters on halftone dots are easier to see than the above-described region separation method. Characters that exist on the ground,
Sharpness (resolution) especially for small characters on a white background
Was insufficient.

Further, in the case of the above-described conventional region separation method, although a halftone dot is subjected to a spiral dithering process after a smoothing process, a halftone image having a low number of lines (for example, 65 lines or 85 lines) is used.
In the case of (line), since the space between the halftone dots is wide, a fixed smoothing filter of about 3 × 3 pixels or about 5 × 5 pixels which is conventionally used is not sufficiently smoothed, and moire may not be completely removed. There was also a problem that would occur.

The present invention has been made on the basis of the above circumstances, and its purpose is to make clear the line drawing such as characters and lines irrespective of the white background or on halftone dots, and to emphasize the gradation property for photographs, and It is an object of the present invention to provide an image processing apparatus which can remove moiré and emphasize gradation, and can reproduce an image adaptively. It is still another object of the present invention to provide an image processing apparatus capable of adaptively changing processing between halftone dots having a high screen ruling and halftone dots having a low screen ruling to remove moire regardless of the number of halftone screen lines.

[Means for solving the problem]

A device for reproducing an image in which a photograph, a halftone dot, and a line image are mixed; a means for converting an input image into a digital multi-valued image signal; a dither processing unit for performing systematic dither processing on the converted digital multi-valued image signal; A sharpening processing unit for performing a sharpening process on the converted digital multi-valued image signal, and each pixel in the input image based on pixel information in a predetermined local area of the converted digital multi-valued image signal. An area separating unit for separating into a halftone area or a line drawing area; and determining the number of lines of the halftone area in the input image based on pixel information in a predetermined local area of the converted digital multivalued image signal. A halftone frequency determining unit, the dither processing unit when the area is separated as a photographic area by the area separation unit, the sharpening processing unit when the area is separated as a line drawing area, and a halftone area. Ah When the halftone frequency determining unit determines that the number of lines is high, the dither processing unit is determined.When the halftone frequency is determined to be a halftone area, the halftone frequency determining unit determines that the number of low lines is low. And a selection output unit for selecting and outputting the sharpening processing unit in the case where the processing has been performed.

[Action]

Each pixel of the input image is separated into one of a photograph, a halftone dot, and a line drawing area in the region separating unit, and it is determined whether the halftone dot is a low screen ruling halftone dot or a high screen ruling halftone dot. Dither processing is performed in the case of a region or a high screen ruling, and sharpening processing is selected and output in the case of a line drawing region or a low screen ruling, respectively. As a result, it is possible to realize optimal image reproduction according to the image area while simplifying the circuit configuration.

〔Example〕

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows a first embodiment of the present invention. For the sake of simplicity, a case where the reproduced image is a monochrome binary image will be described as an example. If the reproduced image is color, for each color signal YMC after color separation and color correction of the original image,
Image processing described below may be performed in parallel, and finally, the respective results may be combined.

In FIG. 1, the input image signal is a digital multi-valued image signal obtained by reading the original image using a means for converting the original image into a digital multi-valued image signal, for example, a scanner of about 400 dpi and 64 gradations.

The sharpening processing circuit 1 is a circuit that emphasizes an edge of an input digital multilevel image signal. Binarization processing circuit 2
Is a circuit for binarizing an image signal edge-emphasized by the sharpening processing circuit 1 with a predetermined threshold value. The sharpening processing circuit 1 and the binarization processing circuit 2 constitute a line drawing area processing unit. As the sharpening processing circuit 1, for example, an MTF correction filter as shown in the following equation (1) is used.

The average error minimum processing circuit 3 is a circuit that performs dither processing on the input digital multilevel image signal by the average error minimum method. The average error minimum processing circuit 3 constitutes a halftone dot processing unit. The average error minimization method determines a threshold value for binarization for each pixel position so that the density error between the original image and the binarized dot image is minimized as an average, and digitally multiplies according to the threshold value. This is a method of binarizing the value image signal.

That is, as shown in FIG.
_{When x, y} and the output image signal are g _{x, y} , the threshold value T _{x, y} for binarization is determined by the following equation (2).

Here, α _{k, L} is a weighting coefficient, and e _{k, L} is a density error between the input image signal and the output image signal. The weight coefficient α _{k, L} is
For example, it is given as a matrix as shown in the following equation (3) _, and a pixel closer to the currently focused pixel fx _{, y} is given a larger weight.

The dither processing circuit 4 is a circuit that performs systematic dither processing such as a spiral dither suitable for a photograph. The dither processing circuit 4 constitutes a photograph area processing unit.

The area separating circuit 5 separates each pixel in the input image into one of a photograph area, a halftone dot area, and a line drawing area based on pixel information in a predetermined local area of the digital multilevel image signal, as described later. Circuit.

The image signal selection circuit 6 is a circuit that selects and outputs any one of the output signals of the binarization processing circuit 2, the average error minimum processing circuit 3, or the dither processing circuit 4 according to the separation result of the area separation circuit 5.

The operation of the first embodiment will be described. A digital multi-valued image signal is input to a segmentation circuit 5, a sharpness processing circuit 1, an average error minimum processing circuit 3, and a dither processing circuit 4, respectively.

The digital multi-valued image signal input to the sharpening processing circuit 1 is edge-enhanced by the sharpening processing circuit 1 and then binarized by the binarization processing circuit 2 to obtain a line image portion such as a character or a line in the input image. Is output as a sharpened image signal.

Further, the digital multi-valued image signal input to the average error minimum processing circuit 3 is binarized by the above-described average error minimum method, and the moire pattern is applied to halftone dots in the input image and characters (including lines) on the halftone dots. It is removed and output as an image signal that has been subjected to gradation processing.

Further, the digital multi-valued image signal input to the dither processing circuit 4 is binarized by a spiral type dither suitable for a photograph, and is output as an image signal obtained by pseudo-grading a photograph portion in the input image.

On the other hand, the area separation circuit 5 separates each pixel of the input image into a line drawing area, a halftone dot area, or a photograph as described later, and outputs a result of separation of each pixel to the image signal selection circuit 6.

The image signal selection circuit 6 outputs the output signal of the binarization processing circuit 2 in the case of the line drawing area and the average error minimum in the case of the halftone dot area based on the area separation result sent from the area separation circuit 5. The output signal of the processing circuit 3 is selected, and in the case of a photographic area, the output signal of the dither processing circuit 4 is selected.
Output.

Therefore, the image signal selection circuit 6 outputs an image signal that has been subjected to optimal processing according to each area of a photograph, a halftone dot, and a line drawing in the input image. When an image is reproduced using the image signal output from the image signal selection circuit 6, not only characters on a white background but also characters existing on a halftone dot are sharpened. Emphasis on sex,
With respect to the halftone dots, an image in which moiré is removed and tone is emphasized is reproduced.

FIG. 3 shows a specific configuration example of the area separating circuit 5 in FIG. In the figure, reference numeral 501 denotes a photographic area separation circuit that separates each pixel in an input image into a photographic area and a non-photographic area, and 502 denotes a halftone area separation that separates each pixel in the input image into a halftone area and a non-halftone area Circuit.

In order to separate each pixel in the input image into three regions of a photograph, a halftone dot, and a line drawing, each pixel in the input image is first separated into a photographic region and a halftone dot region by the two circuits 501 and 502. Then, an AND logic of the image area determined to be a non-dot area by the photo area separation circuit 501 and an image area determined to be a non-dot area by the halftone area separation circuit 502 are taken. The image region determined as the non-dot region may be set as the line drawing region.

As the photographic area separation circuit 501 in FIG. 3, for example, a circuit disclosed in Japanese Patent Application Laid-Open No. 61-225974 can be used. Further, as the dot region separation circuit 502, for example, a circuit as shown in FIG. 4 can be employed.

Hereinafter, the dot area separating circuit 502 in FIG. 4 will be described.

The pole detection unit 503 is a circuit that detects a peak or valley pixel (hereinafter, a “peak pixel”) that forms a halftone dot. For example, in the case of detecting a peak peak pixel, for example, a mask having a size of 3 × 3 pixels as shown in FIG. 5 is used, and a mask between the center pixel Lc and the surrounding pixels L1 to L8 in the mask is as follows. (4) When the expressions (5) are simultaneously satisfied, the center pixel Lc is detected as the peak pixel. In order to detect a valley peak pixel, the direction of the inequality sign in equation (4) may be reversed.

The halftone dot candidate area detecting unit 504 detects a pixel to be a halftone dot candidate area based on the peak pixels of the peaks and valleys detected by the extreme point detecting unit 503. That is, for example, as shown in FIG. 6 (a), when processing is performed in units of blocks, the number of peak pixels of peaks and valleys in the block of interest is counted, and the total is larger than a predetermined threshold value. When it is larger, all pixels (hatched pixels) in the target block are detected as halftone dot candidate areas. In addition, as shown in FIG. 6 (b), when processing is performed in pixel units, when the total number of peak pixels of peaks and valleys in the target block is larger than a predetermined threshold value, Pixels of interest (hatched pixels)
Is detected as a dot candidate area.

The halftone dot area determination unit 505 finally determines whether or not each pixel is a halftone area based on the detection result of the halftone dot candidate area detection unit 504. That is, as shown in FIG. 7 (a), when processing is performed in block units, the number of halftone dot candidate blocks in a target block and a plurality of predetermined surrounding blocks is counted, and the number is determined in advance. When the number is larger than the predetermined number, all pixels (hatched pixels) in the target block are determined to be a halftone dot area.

On the other hand, as shown in FIG. 7 (b), when processing is performed on a pixel basis, the number of halftone dot candidate pixels in surrounding pixels within a predetermined range including the target pixel is counted, and this number is determined in advance. If the number is larger than the predetermined number, the pixel of interest (hatched pixel) is determined to be a halftone dot area.

By executing each of the above-described processes for all the pixels of the input image, all the pixels can be separated into a halftone dot region and a non-halftone dot region. Therefore, if the area separating circuit shown in FIG. 3 is configured by using the halftone area separating circuit shown in FIG. 4, each pixel in the input image can be separated into three areas of a photograph, a halftone dot, and a line drawing. it can.

FIG. 8 shows a second embodiment of the present invention. In the second embodiment, a halftone dot and a photograph are combined into one to form a picture area, and an input image is divided into two parts, a picture area and a line drawing area, and processed. The circuit configuration is further simplified.

In FIG. 8, the average error minimum processing circuit 3 used in FIG. 1 is used as a picture area processing unit for processing a picture area consisting of halftone dots and photographs. As described above, when the average error minimization method or the error diffusion method is used, a reproduced image is reproduced at a certain level for photographs, halftone dots, and characters on halftone dots, excluding characters (including lines) on a white background. Obtainable. Therefore, halftone dots and photos in the input image
The same processing effect as that of the first embodiment can be obtained even if the picture areas are collectively processed by the average error minimum processing circuit 3.

In the case of the example of FIG. 8, the area separation circuit 5 'needs to separate all pixels of the input image into two parts, a line drawing area and a picture area. To realize this, a circuit as shown in FIG. Adopt it. The area separating circuit of FIG. 9 has the same circuit configuration as the area separating circuit 5 of FIG.
The OR logic of the photo area determination output 501 and the halftone area determination output of the halftone area separation circuit 502 is obtained, and the logical result (= photo area + halftone area) is determined as a picture area. .

The image signal selection circuit 6 outputs the output signal of the binarization processing circuit 2 when the separation result is a line drawing area, and minimizes the average error when the determination result is a picture area according to the separation result of the area separation circuit 5. An output signal of the processing circuit 3 is selected and output. Therefore, it is possible to obtain an optimum reproduced image corresponding to the two areas of the line drawing and the picture.

In the embodiments of FIGS. 1 and 8 described above, the halftone error minimum method is used as the halftone dot area processing unit and the picture area processing unit, but an error diffusion method may be used instead. Things.

FIG. 10 shows a third embodiment of the present invention. In the third embodiment, a line drawing such as a character and a halftone dot having a low screen ruling (for example, 85 lines or less) are combined into a line drawing / low screen ruling halftone area. By combining points (for example, 100 lines or more) into a single photo / high screen frequency halftone dot area, and processing the input image separately into these two areas, moiré can be generated regardless of the number of halftone dots. It is one that can be removed.

In FIG. 10, a region separation circuit 5 is a circuit for separating all pixels in an input image into three regions of a line drawing, a photograph, and a halftone dot. As the area separating circuit 5, the area separating circuit 5 used in FIG. 1 can be used.

The line drawing / low ruling halftone dot processing unit is a circuit for performing image processing suitable for a line drawing region such as a character and a halftone dot region having a low ruling, and includes a sharpening processing unit 1 and a binarization processing unit 2. It is composed of As the sharpening processing section 1 and the binarization processing section 2, the sharpening processing circuit 1 and the binarization processing circuit 2 used in FIG. 1 can be used.

The photograph / high ruling halftone dot area processing unit is a circuit for performing image processing suitable for a photographic area and a high ruling halftone dot area,
It comprises a smoothing processing circuit 7 and a systematic dither processing circuit 8 such as a spiral dither. As the smoothing processing circuit 7, for example, when halftone dots are separated into 100 lines or more and 85 lines or less and processed, a smoothing filter such as the following equation (6) is used.

The smoothing filter in the smoothing processing circuit 7 may be properly used depending on the number of halftone dots to be separated. For example, when halftone dots are separated at 65 lines, the following ( What is necessary is just to use the smoothing filter like Formula 7). Alternatively, a plurality of smoothing filters such as the equations (6) and (7) may be prepared, and may be adaptively selected according to processing.

As will be described later, the halftone frequency determining circuit 9 is a circuit that determines, for a halftone portion included in the input image, whether the number of lines is 100 lines or more or 85 lines or less, for example.

The comprehensive determination unit 10 is a circuit that determines a processing signal to be selected based on the determination result of the halftone frequency determining unit 9 and the area separation result of the area separation unit 5 in accordance with the following determination conditions. That is, when a line drawing area signal is output from the area separation circuit 5, a halftone area signal is output from the area separation circuit 5, and the halftone frequency determination circuit 9 determines that the halftone area has a low number of halftone dots (85 lines). If it is determined that the following conditions are satisfied, the image signal selection circuit 6 is instructed to select the processing signal of the line drawing / low screen number halftone dot area processing unit. When the halftone area signal is output from the area separation circuit 5 and the halftone frequency determining circuit 9 determines that the halftone area is a halftone (100 or more) halftone screen, , And instructs the image signal selection circuit 6 to select a processing signal of the photograph / high screen ruling halftone dot region processing section.

The operation of the third embodiment will be described. The digital multi-valued image signal is divided into an area separating circuit 5, a halftone frequency determining circuit 9,
Each of the sharpening processing circuit 1 and the smoothing processing circuit 7 is input.

The digital multi-valued image signal input to the sharpening processing circuit 1 is edge-enhanced by the sharpening processing circuit 1 and then binarized by the binarization processing circuit 2 to obtain a line drawing area and a low line count of 85 lines or less. It is output as an image signal that has been subjected to image processing suitable for the halftone dot area.

The digital multi-valued image signal input to the smoothing processing circuit 7 is smoothed by the smoothing filter as shown in the above equation (6) in the smoothing processing circuit 7 and then the vortex type dither or the like in the dither processing circuit 8. And then output as an image signal that has been subjected to image processing suitable for a photographic area and a halftone dot area having a high screen ruling of 100 lines or more.

On the other hand, the area separating circuit 5 separates each pixel in the input image into one of a line drawing area, a halftone dot area, and a photograph area, and sends a result of the separation to the overall determination unit 10. Further, the halftone frequency determining circuit 9 determines whether the number of halftone dots in the input image is 100 lines or more or 85 lines or less, as described later, and determines the overall result. Send to Part 10.

The overall judgment unit 10 judges the image area in accordance with the judgment conditions 1 to 4 and sends the judgment signal to the image signal selection circuit 6. The image signal selection circuit 6 selects an image signal output from the binarization processing circuit 2 in the case of a line drawing area and a halftone dot area having a low number of lines of 85 lines or less based on the determination signal of the comprehensive determination unit 10. In the case of a photograph area and a halftone dot area having a high screen ruling of 100 lines or more, the image signal output from the dither processing circuit 8 is selected and output. Therefore, in the case of the third embodiment, the halftone dot image having a low line frequency, which may cause moire due to the fixed smoothing filter having a small size, is processed by the line drawing / low frequency line halftone dot region processing unit. , The moire does not occur in a halftone dot portion having a low line frequency as in the related art.

A specific configuration example of the halftone frequency determining circuit 9 in FIG.
Figure 11 shows.

The halftone frequency determining circuit 9 in FIG. 11 is in principle the same circuit as the halftone region separation circuit 502 in FIG. 4, and the pole detection unit 901 is, for example, a 3 × 3 pixel size in FIG. When the following equations (8) and (9) are simultaneously satisfied using the above mask, the central pixel is detected as a peak pixel. In order to detect a valley peak pixel, the direction of the inequality sign in equation (8) may be reversed.

The pole pixel density calculation unit 902, based on the peak pixel detection results of the peaks and valleys in the pole detection unit 901, a predetermined small block, for example, the number of peak peak pixels P ₁ in a small block of 16 × 16 pixel size. Count the number P ₂ of valley peak pixels,
The number of peak pixels and the number of peak pixels are determined.

The halftone frequency determining unit 903 compares the larger one of the _two count values P ₁ and P ₂ obtained by the extreme pixel density calculating unit 902 with a predetermined threshold TH3 prepared in advance, and calculates P <TH3. , It is determined that the small block (or the representative pixel of the small block) is a halftone dot having a low screen ruling (in the above example, a halftone dot having 85 lines or less). When P ≧ TH3, the small block is determined. (Or the representative pixel of the small block) is determined to be a halftone dot having a high number of lines (100 or more halftone dots in the above example).

Note that the size of the small block for calculating the number of peak pixels in the extreme pixel density calculating unit 902 and the halftone frequency determining unit 90
By changing the value of the threshold value TH3 for the determination in step 3, halftone dots can be separated and determined at an arbitrary number of lines.

The halftone frequency determining circuit illustrated in FIG. 11 is for determining the halftone frequency for each small block composed of a predetermined local area. It is also possible to adopt a configuration in which the halftone frequency is determined globally from a set of small blocks. For example, the 16 ×
It is sufficient to collect 16 × 16 blocks of the determination result of the small block having the size of 16 pixels, take a majority decision on the entire 16 × 16 blocks, and determine the number of halftone dots.

〔The invention's effect〕

As is clear from the above description, when reproducing an image in which a photograph, halftone dots and line drawings are mixed, an input image is divided into a line drawing region and a halftone dot region with a low frequency, and a photographic region and a halftone dot with a high frequency. Since the image is divided into regions and image processing suitable for each region is performed, moiré can be generated even for halftone dot images with a low number of lines, which may cause moiré due to a fixed and small smoothing filter. Can be prevented, and higher-quality image reproduction can be realized.

[Brief description of the drawings]

FIG. 1 is a block diagram of a first embodiment of the present invention, FIG. 2 is an explanatory diagram of dither processing by the average error minimizing method, and FIG. 3 shows a specific example of a region separating circuit for the first embodiment. FIG. 4 is a block diagram showing a specific example of a halftone dot separation circuit, FIG. 5 is an explanatory diagram of peak pixel detection, FIG. 6 is an explanatory diagram of halftone dot candidate region detection, and FIG. FIG. 8 is a block diagram showing a second embodiment of the present invention, and FIG. 9 is a block diagram showing a specific example of a region separating circuit for the second embodiment. FIG. 10 is a block diagram of a third embodiment of the present invention, and FIG. 11 is a block diagram showing a specific example of a halftone dot number determination circuit for the second embodiment. DESCRIPTION OF SYMBOLS 1 ... Sharpening processing circuit 2 ... Binarization processing circuit 3 ... Average error minimum processing circuit 4 ... Dither processing circuit 5, 5 '... Region separation circuit 6 ... Image signal selection circuit 7 ... Smoothing processing circuit 8 Dither processing circuit 9 Halftone frequency determination circuit 10 General determination unit

Continuation of the front page (56) References JP-A-63-250274 (JP, A) JP-A-63-263974 (JP, A) JP-A-62-165477 (JP, A) JP-A-61-133470 (JP) JP-A-1-133470 (JP, A) JP-A 64-73972 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 1/40-1/409

Claims (5)

(57) [Claims] 1. An apparatus for reproducing an image in which a photograph, a halftone dot, and a line image are mixed, means for converting an input image into a digital multi-valued image signal, and performing systematic dither processing on the converted digital multi-valued image signal. A dither processing unit, a sharpening processing unit for performing a sharpening process on the converted digital multi-valued image signal, and a sharpening processing unit for converting each of the input digital images based on pixel information in a predetermined local area of the digital multi-valued image signal. Pixels in the photo area,
An area separating unit for separating into a halftone area or a line drawing area; and determining the number of lines of the halftone area in the input image based on pixel information in a predetermined local area of the converted digital multivalued image signal. A halftone frequency determining unit, the dither processing unit when the area is separated as a photographic area by the area separation unit, the sharpening processing unit when the area is separated as a line drawing area, and a halftone area. When the halftone frequency determining unit determines that the number of lines is high in the case where the halftone frequency is separated, the dither processing unit, when the halftone frequency determining unit is separated as a halftone area, the low frequency. And a selection output unit that selects and outputs the sharpening processing unit when the determination is made. 2. An image processing apparatus according to claim 1, wherein said area separating section determines first and second determination means for determining whether the input image is a photographic area or a non-photographic area according to a local feature. A second determining means for determining whether the input image is a halftone area or a non-halftone area in accordance with a local feature; and the second determining means which determines that the input image is a non-photographic area by the first determining means. An image processing apparatus comprising: a third determination unit that determines a region determined as a non-dot region by the unit to be a line drawing region. 3. The image processing apparatus according to claim 1, wherein the halftone frequency determining unit performs the determination based on a density of pixels corresponding to halftone dots in a predetermined area. 4. The image processing apparatus according to claim 1, wherein the halftone frequency determining unit determines a pixel corresponding to a peak of a halftone dot and a valley peak based on pixel information of a predetermined local area of the digital multilevel image signal. , The number of pixels corresponding to a peak peak and the number of pixels corresponding to a valley peak in a predetermined small area are counted, and the line frequency of a halftone dot is determined based on the larger one of the two count values. An image processing apparatus characterized in that: 5. An input image is converted into a digital multi-valued image signal, and each pixel in the input image is converted into a photographic region based on pixel information in a predetermined local region of the converted digital multi-valued image signal.
The number of halftone dots in the input image is determined based on pixel information in a predetermined local region of the converted digital multi-valued image signal, which is separated into either a halftone dot region or a line drawing region. If it is determined that the image is separated, the systematic dither processing is performed.If the image is separated as the line drawing area, the sharpening processing is performed. Is an image processing method characterized by performing an organized dithering process and, when it is determined that the number of lines is low, performing a sharpening process and outputting the result.