JPH04270561A - Picture processor - Google Patents

Abstract

PURPOSE:To obtain an intermediate tone picture with excellent picture quality by applying proper binarizing processing respectively to each of a binary picture area, an intermediate tone area of non-dot picture and an intermediate tone area of a dot picture. CONSTITUTION:The processing unit consists of a smoothing circuit 1, an MTF correction circuit 2, line buffers 3, 5, a 2-picture element delay circuit 4, a level difference arithmetic circuit 6, selectors 7, 14, 8, 9, a comparator 11, a fixed threshold level binarizing circuit 10, reference value generators 12, 15, 16, a dither processing circuit 13, and an OR circuit 17 or the like, a control means applying intermediate tone binarizing processing to a picture signal outputted from a smoothing processing means for an area discriminated from a dot picture area by a dot discrimination means. Thus, a multi-value picture signal obtained by the dot picture area is subject to smoothing processing and then binarized, then the feature of the dot picture is suppressed and moire stripes appearing in the binarized picture signal obtained after the intermediate tone binarizing processing is suppressed.

Description

Detailed Description of the Invention [Industrial Field of Application] The present invention distinguishes between binary image areas and halftone image areas of a multilevel image signal obtained by reading, and detects the multilevel image area of the area determined to be a binary image area. The present invention relates to an image processing apparatus that binarizes a value image signal using a fixed threshold value and also performs halftone binarization processing on a multivalue image signal of an area determined to be a halftone image area.

[Prior Art] For example, a scanner used as a document reading means of a facsimile machine reads a document image line by line.
The multilevel image signal obtained thereby is binarized using a predetermined threshold,
A binary image signal is formed.

Document images can be broadly divided into so-called binary images, which are black and white images with high contrast, such as character images, and halftone images, such as photographic documents.

When a binary image is binarized, a multilevel image signal is binarized using a fixed threshold, but if the same binarization process is applied to a halftone image, the resulting binary image signal is The image quality when recorded or displayed deteriorates significantly. In addition, below,
A method of performing binarization processing using a fixed threshold value is called fixed binarization processing.

Therefore, the multivalued image signal of the halftone image is subjected to so-called halftone binarization processing in which the multivalued image signal of the halftone image is binarized using a threshold value generated using, for example, a systematic dither matrix. Further, when halftone binarization processing is performed on a multi-level image signal of a binary image, the contrast of the image when the resulting binary image signal is recorded or displayed deteriorates, and the image quality deteriorates.

In this way, when binarizing multivalued image signals obtained from binary images and halftone images, there are methods suitable for each type of binarization processing, and when binary images and halftone images are mixed. When reading a document image, if either one of the binarization processes is performed, the following problems will occur.

That is, when fixed binarization processing is applied to the entire image, when the resulting image signal is recorded or displayed, the image in the halftone image area is significantly degraded.

Furthermore, when halftone binarization processing is applied to the entire area, when the resulting image signal is recorded or displayed, the image in the binary image area is significantly degraded.

Therefore, the binary image area and the halftone image area of the image are discriminated, and the multilevel image signal obtained from the area judged to be the binary image area is subjected to fixed binarization processing, and the area judged to be the halftone image area is Devices that perform halftone binarization processing on the multivalued image signal obtained in the above have been in practical use for some time.

[Problems to be Solved by the Invention] However, such conventional devices have the following disadvantages.

In other words, for halftone images, in addition to true halftone images with extremely high resolution such as hand-drawn pictures and still photographs, gradation is divided into 8 small and small dots (halftone dots) that occupy a certain area. There is a halftone image, like a photograph of a printed matter, that visually expresses shading based on the proportion of the ink adhesion area.

On the other hand, pixels obtained by reading with a scanner have a constant pixel size and have a different pixel structure from a halftone image.

For this purpose, when the multilevel image signal obtained by reading the halftone image with a scanner is subjected to halftone binarization processing, and the resulting image signal is recorded or displayed, the areas where the density of the image changes. This has caused the inconvenience that moiré fringes are formed.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image processing apparatus that can eliminate the disadvantages of conventional apparatuses and obtain halftone images of good quality.

[Means for Solving the Problems] The present invention provides a halftone dot discrimination means for discriminating a halftone image area of a multilevel image signal, a smoothing processing means for smoothing the multilevel image signal, and a halftone dot discrimination means. The area determined to be a halftone image area is provided with control means for performing halftone binarization processing on the image signal output from the smoothing processing means.

[Effect] Therefore, the multivalued image signal obtained in the halftone image area is smoothed and then binarized, so the characteristics of the halftone image are suppressed, and thereby halftone binarization is performed. Moiré fringes appearing in the binarized image signal obtained after processing can be suppressed.

[Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

First, the principle of determining the halftone image area will be explained.

For example, as shown in FIG. 1(a), the density distribution of a halftone image varies greatly on a pixel-by-pixel basis at regular intervals,
When this image is subjected to smoothing processing in which the density level of the pixel of interest is averaged using surrounding pixels, density variations in pixel units are suppressed, as shown in FIG. Note that, as a matter of course, even in this case, the halftones (gradation levels) of the original image are maintained.

On the other hand, when the halftone image is subjected to MTF correction processing to remove the blurred components of the image data and reproduce a binary image, the density level changes in pixel units, as shown in FIG. is emphasized, resulting in a signal that looks like its fluctuation components have been extracted.

Therefore, when a level difference is formed between the signal after the smoothing process and the signal after the MTF correction process, and the level difference is larger than a predetermined value, it is possible to determine that the image area is a halftone image. can.

Here, when considering a 3×5 pixel matrix as shown in FIG. 2(a), an example of a coefficient matrix of a filter for smoothing the pixel of interest H located in the center of this pixel matrix is given below. It is shown in the same figure (b). Further, an example of a coefficient matrix of a filter for performing MTF correction processing on the pixel of interest H is shown in FIG. 2(c).

Further, area identification between a binary image and a halftone image can be performed as follows.

That is, in the image area of the binary image, the black and white contrast of the image is large, and in the image area of the halftone image, the black and white contrast of the image is small.

Therefore, as shown in FIG. 3, two large and small thresholds THb,
THc is set, and the part whose density level is higher than the threshold THb and the part whose density level is lower than the threshold THc are determined to be the image area of the binary image, and the density level is set between the threshold THc and the threshold THb.
Determine the part of the range between.

FIG. 4 shows an image processing device according to an embodiment of the present invention.

In the figure, a multilevel image signal DV output from an image reading device (not shown) is transmitted to a smoothing circuit 1, an MTF correction circuit 2,
It is also added to a line buffer 3 having a FIFO configuration capable of storing one line of multilevel image signal DV of the image reading device.

The output signal of the line buffer 3 is an image signal DVa of the same pixel timing one line before the multilevel image signal DV. It is added to line buffer 5. The output signal of the line buffer 5 is a pixel signal D at the same pixel timing two lines before the multilevel pixel signal DV.
It is added to the image signal smoothing circuit 1 and the MTF correction circuit 2 as Vb.

The smoothing circuit 1 inputs the multilevel image signal DV, the image signal DVa, and the image signal DVb, and executes the smoothing process on the pixel of interest H described above, and the processing result is a smoothed image signal DVc. It is added to the level difference calculation circuit 6 and the input terminal B of the selector 7 as a signal.

The MTF correction circuit 2 inputs the multivalued image signal DV, the image signal DVa, and the image signal DVb, and executes the MTF correction process for the pixel of interest H described above, and the processing result is output as a corrected image signal DVd. , level difference calculation circuit 6,
It is applied to the input terminal A of the comparator 8, the input terminal A of the comparator 9, and the fixed threshold value binarization circuit 10, respectively.

The two-pixel delay circuit 4 forms a pixel delay required for the calculations of the smoothing circuit 1 and the MTF correction circuit 2, and its output signal is applied to the input terminal B of the selector 7.

The level difference calculation circuit 6 calculates the level difference between the smoothed image signal DVc outputted from the smoothing circuit 1 and the corrected image signal DVd outputted from the MTF correction circuit 2, pixel by pixel. The result is held at the input terminal B of the comparator 11 as a level difference signal DF.

The reference value generator 12 uses the method described above to generate a smoothed image signal DVc and a corrected image signal DVd for determining a halftone image.
The threshold signal TTa is applied to the input terminal A of the comparator 11.

Comparator 11 receives threshold signal TT applied to input terminal A.
a, the level difference signal DF applied to the input terminal B
When the value becomes large, the output signal S1 is changed to a logic H level, and the output signal S1 is applied to the input terminal S of the selector 7.

When the signal S1 applied to the input terminal S is at the logic L level, the selector 7 selects the 2-pixel delayed multi-level image signal DVa applied to the input terminal A, and selects the multilevel image signal DVa applied to the input terminal S. When the signal S1 is at the logic H level, the smoothed image signal DVc applied to the input terminal B is selected, and the selected signal is applied to the halftone image signal DVt.
The signal is output to the dither processing circuit 13 as a signal.

The dither processing circuit 13 performs halftone binarization processing on the input halftone image signal DVt by applying, for example, a threshold of a systematic dither matrix of size 8×8, and the processing result is a halftone image signal DVt. It is applied to the input end B of the selector 14 as a halftone area binary signal HT obtained by binarizing the image signal of the area.

The reference value generator 15 and the reference value generator 16 generate a threshold value TH for discriminating between the binary image area and the halftone image area described above.
It outputs threshold signals TTb, TTc corresponding to the signals TTb and THc, and the threshold signals TTb and TTc are applied to the input end B of the comparator 8 and the input end B of the comparator 9, respectively.

The comparator 8 receives the corrected image signal DV applied to the input terminal A.
When d is larger than the threshold signal TTb applied to the input terminal B, the output signal S2 is changed to a logic H level, and the output signal S2 is sent to the OR circuit 17.
is applied to one input end of the .

The comparator 9 receives the corrected image signal DV applied to the input terminal A.
When d is smaller than the threshold signal TTc applied to the input terminal B, the output signal S3 is changed to the logic H level, and the output signal S3 is sent to the OR circuit 17.
is applied to the other input end of the .

Therefore, the output signal S3 of the OR circuit 17 changes to the logic H level in the binary image area, and the signal S3 is applied to the input terminal S of the selector 14.

The fixed threshold binarization circuit 10 binarizes the corrected image signal DVd using a threshold value in the binary image area, and the processing result is obtained by binarizing the image in the binary image area. It is applied to the input terminal A of the selector 14 as a binary area binary image signal BW.

When the signal S3 applied to the input terminal S is at the logic L level, the selector 14 selects the halftone binary image signal HT applied to the input terminal B, and when the signal S3 is at the logic H level. When , the binary area binary image signal BW applied to the input terminal A is selected, and the selected signals are outputted to the next stage device as the binary image signal BV.

With the above configuration, the multilevel image signal DV read by the image reading device is sequentially applied to the smoothing circuit 1 and the MTF correction circuit 2, and is sequentially accumulated in the line buffer 3 and the line buffer 5.

When the multivalued image signal DV is accumulated up to the line buffer 5, the smoothing circuit 1 and the MTF correction circuit 2 execute the above-mentioned smoothing process and MTF correction process, and the smoothed image signal DVc of the pixel of interest H, respectively. and corrected image signal DVd
is output. At this time, the two-pixel delay circuit 4 outputs the multilevel image signal DVa of the pixel of interest H.

Here, when the pixel of interest H belongs to the binary image area,
Either one of the output signals S2 and S3 of the comparator 8 or the comparator 9 rises to the logic H level, thereby causing the signal S
3 rises to the logic H level, the selector 7 outputs the binary area binary image signal BW applied to the input terminal A as the binary image signal BV.

As a result, the binary area binary image signal BW obtained by binarizing the corrected image signal DVd after the MTF correction process of the multilevel image signal DV of the binary image area is converted into a binary image signal BW using a predetermined threshold value. It is output as an image signal BV.

On the other hand, when the pixel H of interest belongs to a halftone image area and the halftone image area is a halftone image area, the level difference signal DF output from the level difference calculation circuit 6 is the threshold value signal TTa.
, the output signal S1 of the comparator 11 rises to the logic H level, and the selector 7 selects the smoothed image signal DVc applied to the input terminal A.
It is output to the dither processing circuit 13 as a halftone image signal DVt.

As a result, the smoothed image signal DVc is subjected to halftone binarization processing, and the resulting halftone binarized image signal HT is applied to the selector 14.

On the other hand, in this case, since the output signals S2 and S3 of the comparators 8 and 9 are both at the logic L level, the signal S3 becomes the logic L level. The halftone binary image signal HT added to B is output as a binary image signal BV.

In this way, in the halftone image area, the halftone binarized image signal HT obtained by performing the halftone binarization process on the smoothed image signal DVc obtained by the smoothing process by the smoothing circuit 1 is generated. ,
It is output as a binary image signal BV.

Further, when the pixel H of interest belongs to a halftone image area and the halftone image area is a non-halftone image area, the level difference signal DF output from the level difference calculation circuit 6 is the threshold value signal TT.
a or less, the output signal S1 of the comparator 11 becomes the logic L level, and the selector 7 selects the multilevel image signal DVa applied to the input terminal B, and sends it to the dither processing circuit 13 as the halftone image signal DVt. Output.

As a result, the multilevel image signal DVa is subjected to halftone binary processing, and the resulting halftone binary image signal HT is applied to the selector 14.

In this case, in the same manner as described above, the signal S3 is at the logic L level and the selector 14 selects the input terminal B, so the halftone binary image signal HT is output as the binary image signal BV. be done.

In this way, in the non-halftone image area, the halftone binary image signal HT obtained by performing halftone binary processing on the multilevel image signal DVa obtained by the image reading device is converted into a binary image. It is output as a signal BV.

As described above, in this embodiment, appropriate binarization processing is performed for each of the binary image area, the halftone area of a non-halftone image, and the halftone area of a halftone image. The image quality of the image obtained when the digitized image signal BV is recorded or displayed can be maintained at high quality.

Note that in the above-described embodiment, the smoothing process and the MTF correction are realized by filter calculation using a 3×5 coefficient matrix, but the content of the process is not limited to this.

Further, as a means for performing halftone binarization processing, it is also possible to use something other than the dither processing circuit described above.

[Effects of the Invention] As explained above, according to the present invention, a binary image area,
Appropriate binarization processing is performed for the halftone area of the non-halftone image and the halftone area of the halftone image, so the image quality of the image obtained when recording or displaying the binary image signal is improved. This has the effect of being able to maintain high quality.

[Brief explanation of the drawing]

FIG. 1(a) is a graph diagram illustrating the state of density change of the halftone dot image, FIG. 1(b) is a graph diagram illustrating the state of density change after smoothing processing of the image of FIG. FIG. 2(c) is a graph illustrating the state of density change after MTF correction processing of the image in FIG. 2(a), and FIG. Figure (b) is a schematic diagram showing an example of a coefficient matrix for smoothing processing, Figure (c) is a schematic diagram showing an example of a coefficient matrix for MTF correction processing, and Figure 3 is a schematic diagram showing an example of a coefficient matrix for MTF correction processing. A graph diagram for explaining the discrimination method, and FIG. 4 is a block diagram showing an image processing apparatus according to an embodiment of the present invention. DESCRIPTION OF SYMBOLS 1... Smoothing circuit, 2... MTF correction circuit, 3, 5... Line buffer, 4... 2 pixel delay circuit, 6... Level difference calculation circuit, 7, 14... Selector, 8, 9, 11... Comparator, 10...
Fixed threshold value binarization circuit, 12, 15, 16... reference value generator, 13... dither processing circuit, 17... OR circuit. Agent Patent Attorney Makoto Monda

Claims (1)

[Claims] The binary image area and the halftone image area of the multivalued image signal obtained by reading are determined, and the multivalued image signal of the area determined to be the binary image area is binarized using a fixed threshold. In an image processing device that performs halftone binarization processing on a multivalued image signal of a determined area, a halftone discrimination means that determines a halftone image area of the multivalued image signal, and a smoothing unit that performs smoothing processing on the multivalued image signal. An image characterized by comprising a processing means and a control means for performing halftone binarization processing on the image signal outputted from the smoothing processing means for the area determined to be a halftone image area by the halftone dot discrimination means. Processing equipment.