JPH10126613A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the image quality by not only smoothing edges of a character line drawing or the like but also enhancing granuality of a gradation image such as a photograph in the case of applying multi-value processing to binary image data. SOLUTION: A pattern detection means 31 that receives binary image data D compares a 2-dimension arrangement pattern for a noted picture element and its surrounding picture elements with various preset patterns for each noted picture element so as to discriminate whether or not they are in matching with each other and provides an output of a pattern code P to a code conversion means 32 when the matching is detected. The code conversion means 32 applies M-value processing (M>2) to the binary image data depending on the pattern code P, decides an area of black picture elements in one picture element and decides phase information within a same picture element of the black picture element and provides an output of the information as multi-value image data D'.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus used in an image forming apparatus such as a digital copying machine, a facsimile apparatus, an image scanner, an image filing apparatus or an image display apparatus. The present invention relates to an image processing apparatus provided with a multi-level converting means for converting binary image data into multi-level image data to improve image quality.

[0002]

2. Description of the Related Art In an image forming apparatus or an image display apparatus as described above, character image data obtained by converting character code data using font data or image image data read by an image scanner or the like is quantized. A binary data is developed in a bit map form (dot matrix form) in a video memory area on a memory (RAM), sequentially read out and sent to an image forming section as video data to form an image on a recording paper, Alternatively, it is sent to a display and displayed on a screen.

In this case, if the image forming object is an analog image, the image can be continuous in any direction. However, a digital bitmap image obtained by binarizing the image and developing it is one dot unit in a direction orthogonal to the dot matrix. Since the direction can be changed only in a step shape, the formed image is distorted. As a result, jaggies occur in which straight lines and smooth curves that are inclined with respect to the direction orthogonal to the dot matrix are formed stepwise, and characters and line drawings (especially contour lines) are formed at or near the original image. It was difficult.

An effective means for solving this problem is to increase the resolution of the bitmap image. For example, doubling the resolution requires four times the memory capacity and four times the data processing capacity. Significant cost increase.
For this reason, various methods have been proposed for reducing image distortion such as jaggies without increasing the memory capacity or increasing the data processing speed.

As one example, Japanese Patent Application Laid-Open No. 5-207282
As disclosed in Japanese Unexamined Patent Application Publication No. H11-157, an edge portion which is a boundary portion between a black dot region and a white dot region of a character line drawing is detected, and the size or phase of a dot (pixel) requiring correction is corrected. Improving the smoothness has been already proposed.

[0006]

However, according to such a conventional correction method, it is possible to smooth the diagonal lines in the character and line drawing of the binary image, but it is not suitable for the binarized pattern or photographic area. Since no processing is performed, no improvement in image quality in the gradation image portion could be expected. SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has as its first object to improve graininess in a gradation image portion of a photograph and the like, and in addition, to smooth edges of a character line drawing and the like. The purpose is also.

[0007]

According to the present invention, there is provided an image processing apparatus having a multi-value processing means for converting binary image data into multi-value image data.
The multi-value processing means converts the binary image data to M-value (M>
2) and phase information output means for outputting phase information within the same pixel (within one dot area). This makes it possible to smooth the edges of the character and line drawing and the like, and to improve the graininess in a gradation image portion such as a photograph.

Further, the multi-value processing means has pattern detection means for detecting, for each target pixel of the binary image data, a pattern of a two-dimensional arrangement of the target pixel and its peripheral pixels, The output means may determine the phase of each pixel obtained by converting the pixel of interest into an M-value in accordance with the detection result of the pattern detection means, and output the phase information. Thereby, the graininess in a gradation image portion such as a photograph can be more appropriately improved.

The multi-value processing means may include means for detecting a line segment shape of the binary image data, and when the means detects a line segment shape inclined with respect to a direction orthogonal to the dot matrix, Means may be provided for shifting the phase of each pixel constituting the line segment in accordance with the inclination angle.
As a result, the processing for smoothing the edges of the character and line drawing is more appropriately performed.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a block diagram showing an example in which the image processing apparatus according to the present invention is applied to a digital copying machine.

In FIG. 2, reference numeral 1 denotes a CCD image sensor (line sensor or two-dimensional sensor) of a scanner section, which reads an image of a document set on a contact glass and performs photoelectric conversion. This CCD image sensor 1
The A / D converter 2 quantizes the image signal converted into an electric signal by the A / D converter 2 and converts the quantized signal into a digital signal. The shading correction unit 3 corrects a variation in light sensitivity of each pixel of the CCD image sensor 1.

Then, the frequency degradation of the input image signal is corrected by the filter processing means 4, and
Pixels in the main scanning direction are enlarged / reduced in accordance with the designated magnification. Further, the gamma correction unit 6 converts the gamma characteristic of the input image signal into the output gamma characteristic, and the gradation processing unit 7 performs optimal gradation processing in the image mode according to the image output device (image forming unit) 15. And outputs binary image data A or multi-valued image data B.

The binary image data A is data-compressed by the compression / decompression means 8, and the compressed
It can be stored and stored in the external storage means 9 such as a floppy disk device. When outputting the stored compressed image data, the compressed image data is read from the external storage means 9 and decompressed by the compression / decompression means 8 to restore the original binary image data. In this way, a large amount of image data can be stored without increasing the storage capacity of the external storage unit 9 too much.

The selector 11 determines which of the binary image data A restored by the compression / expansion means 8 and the binary image data E input from the external interface (I / F) 10 is to be transmitted to the subsequent stage. , Gradation processing means 7
And the multi-valued image data F from the external interface 10 are transmitted to the subsequent stage by a selection signal from the system controller 16, respectively.

The binary image data D selected and output by the selector 11 is multi-valued by a multi-value processing means 13 and used as one input of a selector 14.
The multivalued image data G selected and output by the selector 14 becomes the other input of the selector 14. Therefore, the selector 14 selects one of the multivalued image data according to the selection signal from the system controller 16 and sends it to an image output unit (image forming unit), where an image is formed on a sheet. Here, the multi-value processing means 13 is a part according to the present invention.

FIG. 3 is a block diagram showing an example of the gradation processing means 7 in FIG. This gradation processing means 7
Is a 1-bit output data A obtained by binarizing the input multi-valued original image data (8 bits) by the error calculation / binarization means 21 by the error diffusion method, and performs no processing on the input original image data. And outputs 8-bit output data B without performing.

The error generated by the binarization of each pixel by the error calculation / binarization means 21 is stored in a FIFO memory 22. After the coefficients of the error matrix 23 are multiplied by the error summing means 24, the error sum is calculated. Output data is calculated by the arithmetic / binarization means 21. The error that occurs here is
At the next timing, it is used for the product-sum operation with the coefficient of the error matrix 23 and stored in the FIFO memory 22.

FIG. 1 is a block diagram showing a first example of the configuration of the multi-value processing means 13 according to the present invention in FIG. The multi-value processing means 13 includes a pattern detection means 31
And code conversion means 32. Selector 11 of FIG.
The binary image data D output from the above is input, and the pattern detection unit 31 determines the pattern of the two-dimensional arrangement of the target pixel and its surrounding pixels for each target pixel of the binary image data,
A comparison is made with various types of patterns set in advance to determine whether or not they match, and when a match is detected, the pattern code P is output to the code conversion means 32.

Therefore, the pattern detecting means 31 includes a memory capable of storing the input binary image data two-dimensionally,
A memory for storing various patterns to be compared with the memory, a CPU for comparing a bit pattern in a predetermined range around the pixel of interest of the input binary image data with the previously stored various patterns to determine whether or not they match each other. Having.

The code conversion means 32 converts the binary image data into an M-value (M> 2) according to the pattern code P.
Then, the area of the black pixel in one pixel is determined, and the phase information in the same pixel (in the area of one dot) of the black pixel portion is also determined. Then, such information is output as multi-valued image data D '. That is, the code conversion unit 32 serves both as a unit that converts the binary image data into an M-value (M> 2) and a phase information output unit that outputs the phase information in the same pixel.

The code conversion means 32 may look up a conversion table (look-up table) created and stored in advance by using a pattern code P, or may convert the conversion table (look-up table) according to a pattern detected for each pixel of interest. Alternatively, the pixel area and the phase information may be calculated and obtained.

FIG. 4 shows an example of a pattern detected by the pattern detecting means 31 (left side) and a corresponding multi-valued image data pattern obtained by converting the binary image data of each pattern by the code converting means 32 (right side). Is shown. (a) to (e) show an image of a two-dimensional area of 5 × 3 dots, and (f) shows an image of a two-dimensional area of 5 × 5 dots, where the center dot is the target pixel.

In this example, as shown on the left side of FIGS. 4A to 4C, an isolated black pixel of one dot or two dots is regarded as a gradation image (halftone image) of a photograph or the like. In the case of an existing pattern, as shown on the right side, the area of a black pixel is divided into two pixels each having a half of the area of one pixel, and the phase within one pixel of each black pixel portion is changed. , The center, the left half, and the right half.

The pixel area is, for example, 0 pixels: 001 /
Two pixels: 01 One pixel: 11 is ternary, and the phase information in one pixel is added as ternary information, for example, right half: 01, center: 11, left half: 10 I do. As a result, the converted multi-valued image data becomes 4-bit data such as "0110" or "0101". However, it is also possible to convert the data into 8-bit data or 16-bit data by finer multi-value conversion. (D) of FIG.
In the case of the pattern shown on the left side of (f), since it is regarded as a character line drawing, the pattern of the image is not changed, but one black pixel is converted into 4 bits from "1" to "1111". I do.

Here, the phase (drawing position within a pixel) in black pixel drawing will be further described. FIG. 4A shows a halftone highlight portion in which only one black pixel is present in a 5 × 3 pixel window. As shown on the right side, half the density (pixel area) is shown. Are drawn in two pixels. Here, the phase of the left pixel is the left phase, and the right pixel is the right phase, so that the distance between the two pixels is maximized. The multi-valued image data of the divided two pixels is, for example, “0110” “0101”.

Next, as shown in FIG. 4 (b), when only two black pixels are separated by one pixel within a 5 × 3 pixel window, as shown on the right side thereof, The image is divided into four pixels of a dot having a density (pixel area). Here, as for the phase, the rightmost pixel is drawn as the right phase, two pixels between them are drawn as the center phase, and the leftmost pixel is drawn as the left phase. By doing so, the distance between the black pixels is equalized.

As shown in FIG. 4C, when two black pixels are arranged side by side in a 5 × 3 pixel window, a half density (pixel area) ) Are drawn with 2 pixels of dots and 3 pixels of dots of 2/2 density between them. Here, as for the phase, the rightmost pixel is drawn as the right phase, one pixel in between is drawn as the center phase, and the leftmost pixel is drawn as the left phase. By doing so, the distance between the black pixels is equalized. As described above, according to this embodiment, it is possible to improve the graininess in a gradation image portion such as a photograph.

Next, a second example of the configuration of the multi-value generating means according to the present invention will be described with reference to FIG. FIG. 5 is a block diagram showing a second configuration example of the multi-value processing means 13 in FIG. The multi-level converting means 13 includes a line shape detecting means 33 and a second code converting means in addition to a pattern detecting means 31 'and a first code converting means 32' similar to the multi-level converting means shown in FIG. It comprises means 34, third code conversion means 35, and selector 36.

The multi-value conversion means 13 is provided by the selector 1 shown in FIG.
Different multivalue processing is performed depending on whether the image of the binary image data D input from 1 is a halftone such as a photograph, a diagonal line such as a character line drawing or the like.

The pattern detecting means 31 ', like the pattern detecting means 31 shown in FIG.
For each pixel of interest, the pattern of the two-dimensional arrangement of the pixel of interest and its surrounding pixels is compared with various preset patterns (in this case, a pattern of a halftone image) to determine whether or not they match, Is detected, the pattern code P
a is output to the first code conversion means 32 '. further,
When a match with any of the patterns is detected, the halftone detection signal S1 output to the selector 36 is turned “ON”. When a match with any of the patterns cannot be detected, the halftone detection signal S1 is "OF".
F ".

The first code conversion means 32 'converts the binary image data into M-valued data according to the pattern code Pa output from the pattern detection means 31', similarly to the code conversion means 32 of FIG. M> 2) to determine the area of the black pixel in one pixel, and also determine the phase information of the black pixel portion in the same pixel, and use that information as multi-valued image data D1.
And outputs it to the selector 36.

The line segment shape detecting means 33 converts the input binary image data into two in the same manner as the pattern detecting means 31 and 31 '.
A memory that stores two-dimensionally, a memory that previously stores various line-segment image patterns,
It is composed of a CPU or the like for determining mismatch.

Then, for each pixel of interest of the input binary image data D, various patterns in which the pattern of the two-dimensional arrangement of the pixel of interest and its surrounding pixels are set in advance (in this case, an oblique line such as a character line drawing). The pattern code Pb including the information of the inclination angle is output to the second code conversion means 34 when the match is detected. Further, when a match with any of the patterns is detected, that is, when a diagonal line is detected, the diagonal line detection signal S2 output to the selector 36 is turned “ON”.

When a match with any of the patterns cannot be detected, that is, when a diagonal line cannot be detected, the diagonal line detection signal S2 is "OFF". Here, the “oblique line” refers to a line segment shape that is inclined with respect to the direction orthogonal to the dot matrix as shown in FIG. In the case of a curved line, a small portion is regarded as an oblique line, and is included in the oblique line here.

When the pattern code Pb including the information of the inclination angle of the oblique line is input from the line segment shape detecting means 33, the second code conversion means 34 changes the phase of each pixel constituting the line segment to the inclination. The pixel area and the phase information are determined so as to be shifted according to the angle, and the determined information is output as multi-valued image data D2 and input to the selector 36.

For example, the input binary image data is
In the case of the oblique line pattern shown in FIG. 9A, the line segment shape detecting means 33 detects the pattern and sends a pattern code Pb including the inclination information (2/1) to the second code converting means. The second code conversion means 34 converts the pixel area within one dot (shown by a small square in the figure) of each pixel so as to convert the image pattern as shown in FIG. The phase information indicating the position is represented by multi-valued image data D
Output as 2. This makes it possible to smooth edges of character and line drawings.

The third code conversion means 35 receives the input 2
Without changing the pattern of the value image data, the data is converted into data having the same number of bits as the multivalued image data D1 and D2, and output. For example, the binary image data to be input is shown in FIG.
In the case of the patterns shown in (d) to (e), the black pixel “1” is “1111” and the white pixel “0” is “000”.
The data is converted as "0" and output as multi-valued image data D3.

The selector 36 is provided with multivalued image data D1, D2 and D3 from each code conversion means, a halftone detection signal S1 from the pattern detection means 31 'and an oblique line detection signal S2 from the line shape detection means 33. And selects one of the multi-valued image data D1, D2 and D3 according to the combination of “ON” and “OFF” of the halftone detection signal S1 and the oblique line detection signal S2 as shown in Table 1. Output as multi-valued image data D '.

[0039]

[Table 1]

That is, the multi-valued image data D1 from the first code conversion means is used when the halftone is detected and no oblique line is detected, and the second code is used when the halftone where the oblique line is detected is not detected. When both the halftone and the oblique line are detected or both are not detected, the multivalued image data D3 from the third code conversion means is selected and output.

FIG. 7 shows a case where the input binary image data includes a gradation image (halftone) portion such as a photograph, an oblique line such as a character line drawing, and other (in this example, a vertical line) images. An example of multi-value conversion is shown. (A) is a binary image before multi-value conversion,
(B) is a multi-value image after multi-value conversion. As described above, according to this embodiment, it is possible to improve the graininess in a gradation image portion such as a photograph and to smooth the edge of a character / line image or the like.

[0042]

As described above, according to the present invention, when converting binary image data into multi-valued image data, it is possible to improve the graininess in a gradation image portion such as a photograph. . At the same time, it is possible to smooth the edges of the character and line drawing and the like, so that the image quality can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a first configuration example of a multi-value processing unit 13 in FIG. 2;

FIG. 2 is a block diagram showing an example in which the image processing apparatus according to the present invention is applied to a digital copying machine.

FIG. 3 is a block diagram showing a configuration example of a gradation processing means 7 in FIG. 2;

FIG. 4 is a diagram showing an example of an image pattern before and after multi-value conversion by the gradation processing means 7 shown in FIG. 3;

FIG. 5 is a block diagram showing a second configuration example of the multi-value processing means 13 in FIG. 2;

6 is a diagram showing an example of an image pattern before and after multi-value conversion of an oblique line by the second code conversion means 34 in FIG. 5;

7 is a diagram illustrating an example of an image pattern before and after multi-value conversion of image data in which various types of images are mixed by the multi-level conversion processing unit 13 illustrated in FIG. 5;

[Explanation of symbols]

 1: CCD 2: A / D converter 3: Shading correction means 4: Filter processing means 5: Zooming means 6: Gamma correction means 7: Gradation processing means 8: Compression / expansion means 9: External storage means 10: External I / F 11, 12, 14, 36: Selector 13: Multi-value processing means 15: Image output device 31, 31 ': Pattern detection means 32: Code conversion means 32': First code conversion means 33: Line segment Shape detection means 34: Second code conversion means 35: Third code detection means

Claims (3)

[Claims] 1. An image processing apparatus comprising a multi-value processing means for converting binary image data into multi-value image data, wherein the multi-value processing means converts the binary image data into M-valued data (M>
2) An image processing apparatus comprising: means for performing phase information; and phase information output means for outputting phase information in the same pixel. 2. The image processing apparatus according to claim 1, wherein
The multi-value processing means includes, for each pixel of interest of the binary image data, pattern detection means for detecting a pattern of a two-dimensional arrangement of the pixel of interest and its surrounding pixels, and the phase information output means comprises: An image processing apparatus comprising: means for determining a phase of each pixel obtained by converting the pixel of interest into an M-value in accordance with a detection result of a pattern detecting means and outputting the phase information. 3. The image processing apparatus according to claim 1, wherein said multi-value processing means includes means for detecting a line segment shape of binary image data, and said means for detecting a line segment shape of said binary image data in a direction orthogonal to a dot matrix. An image processing apparatus comprising: means for shifting the phase of each pixel constituting the line segment in accordance with the angle of inclination when the shape of the line segment that is inclined is detected.