JP3184197B2 - Image processing device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a system for handling an image such as a document as binary image data. In particular, in a system having a function of inputting or storing an image and displaying the image on a display or outputting from a printer or the like, when displaying an image on an output device having a different resolution from that of the input device, a thinning process may be performed. The present invention relates to an image processing system having a binarization processing method capable of producing a pseudo halftone image with little image quality deterioration.

[Conventional technology]

One of the components of a document is a halftone image such as a photograph or a halftone image often used for printed matter. When these areas are treated as binary image data composed of white and black pixels, a method of binarizing by pseudo halftone processing is widely used.

A pseudo halftone image is an image that expresses shading with the density of black pixels within a certain range. A typical method is an organized dither method (for example, Floyd et al .: Unadaptive).
Algorithm Huo Spersial Grayscale, 1
975, SI Day, International Symposium Digest of Technical Papers, 4.
3, pages 36 to 37, April 1975 (Floyd, R. and Steinbe
rg, L., “An Adaptive Algorithm for SpatialGrey Scal
e "1975 SID International Symposium Digest of Techn
ical Papers, 4.3, pp36-37. Apr, 1975)) and the mean error minimization method (Jabis et al .: Assurance of Technic)
Huo The Display Of Continuas Toon Pictorials On Billy Display,
Computer Graphics and Image Processing, 5, pp. 13-40, 1976 (Jarvis, J.
F., Judice, CN, and Ninke, WH, “A Survey of Techni
ques for the Display of Cintinuous Tone Picture on
Bilevel Displays, "Computer Graphics and Image Pro
cessing, 5, pp13-40, 1976)).

The systematic dither method is a method in which the threshold value for binarizing multi-valued data is periodically changed to make the density of appearing black pixels proportional to the shading value of the image. On the other hand, the average error minimum method is a method of adding the rounding error generated at the time of binarization to the gray value of the surrounding pixels to make the gray value and the black pixel density as a whole as a whole.

As another process, a method called a random dither method such as an average error diffusion method is used.

[Problems to be solved by the invention]

In the systematic dither method, the threshold value periodically changes as described above, so that the output image also has periodicity. for that reason,
When this process is performed on a halftone image having the same periodicity, the image quality in which moire occurs due to the difference between the two periods is significantly deteriorated.

On the other hand, the random dither method such as the average error minimization method has a very strong correlation between neighboring pixels. Therefore, for example, if a thinning process is performed at the time of display, the image quality is significantly deteriorated.

In general, in a system that handles binary images, the input is performed with the highest possible resolution when the image is input. However, CRT
The resolution of an image display device such as a display is lower than the resolution of an image input device such as an image scanner. Therefore, when displaying the entire input image, a thinning process is required to reduce the number of pixels of the output image.

However, when the pseudo halftone image is thinned, the image quality is significantly deteriorated. In particular, the image obtained by the average error minimization method is severely degraded. Therefore, when a pseudo halftone image with a high resolution is displayed, the image quality is displayed lower than when an image having the same resolution as that of the display device is displayed due to deterioration due to thinning. Even existed. No effective means has existed for this problem.

SUMMARY OF THE INVENTION An object of the present invention is to output a multi-valued image that undergoes pseudo halftone processing and output as a binary image, even when the resolution of the input / output device is different, with little deterioration in image quality due to thinning. It is an object of the present invention to realize an image processing apparatus capable of performing such operations.

In addition, when temporarily storing the binarized image, the image processing can select a weighting factor, dither matrix, and the like by assuming in advance the thinning process applied at the time of output by an operator. It is to implement the device.

[Means for solving the problem]

In order to solve the above-mentioned problem, the present invention has means for accumulating a plurality of types of weighting coefficients of the random dither method and selecting an image having little image quality deterioration in accordance with a thinning pattern to be applied. The thinning pattern to be applied is a means for detecting a resolution of an input image and a resolution of a device for outputting an image, a means for calculating a ratio of the resolution of the two places,
The determination can be made by having means for calculating the ratio between the two resolutions.

Also, in the case of an apparatus using the systematic dither method at the time of binarization, similarly, there is provided a means for accumulating a plurality of types of dither matrixes and selecting them according to a thinning pattern.

Further, when one of the input / output devices is not connected, the input / output device has an instruction means for an operator to select a coefficient or dither matrix to be used.

[Action]

In the case of a halftone image, moire does not occur if a random dither method such as an average error minimum method is used as a pseudo halftone processing method.

Image quality degradation during thinning in the random dither method can be suppressed by considering the thinning interval in advance. Specifically, in the average error minimization method, the relative position of a pixel to which an error is to be propagated is switched according to the pixel arrangement after thinning.

Similarly, in the systematic dither method, the degradation of image quality can be suppressed by determining the dither matrix in consideration of the pixel arrangement after thinning.

 The operation principle of the present invention is as follows.

FIG. 2 is a diagram showing an example of an array of pixels to be used when the image data is thinned down to half both vertically and horizontally.

In the figure, white circles, black circles, triangles, and squares indicate pixels, black circles indicate pixels that are images after thinning out, and others indicate pixels that are discarded. In a pseudo halftone image, especially in a random dither image, the image quality is significantly deteriorated due to thinning.
This is because there is a strong correlation between the black pixel and other pixels.

Therefore, by breaking the correlation between the black pixel and other pixels,
One pseudo halftone image can be composed of only black pixels. For example, in the mean error minimization method, an error propagation destination is matched with a thinning interval.

As an example, FIG.
FIG. 3 shows an error weight coefficient matrix in the average error minimum method. FIG. 3A shows an example of a weighting coefficient used in the ordinary mean error minimum method. On the other hand, FIG.
6 is an example of a weight coefficient used when a 1/2 image is subjected to a thinning-out process on a value-converted image. In the figure, * is a pixel to be binarized at that time, and numerical values are coefficients. As shown, 2
The error caused by the binarization does not propagate to adjacent pixels,
Sent to every pixel. As a result, an error generated when binarizing a black dot pixel propagates only to the black pixel and has no effect on other pixels. As a result, a pseudo halftone image with little image quality degradation can be obtained even by thinning. On the other hand, in the case of the systematic dither method, a similar effect can be obtained by setting the dither matrix in units of four pixels, for example, two pixels each in the vertical and horizontal directions.

Generally, when performing thinning processing in an image processing system,
The decimation method has been known in advance. Also, the thinning rate applied at the time of display is limited to a specific rate such as 1/2 in most cases. Therefore, the above-mentioned problem can be solved by storing weighting coefficient matrices corresponding to each thinning method in the binarization processing unit and selecting an appropriate matrix according to the system configuration. .

〔Example〕

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a basic configuration of an embodiment of the present invention. In the figure, reference numeral 10 denotes a document reading device that optically reads a document or the like by a known means and outputs multi-value digital image data, and reference numeral 20 denotes a binarization that binarizes the multi-value digital image data output from the document reading device. A processing unit 25 is a switch for externally instructing the binarization method. This binarization processing unit 20 has a parameter switching unit 27 for switching binarization parameters. As peripheral devices constituting the system, reference numeral 30 denotes a buffer memory for temporarily storing binarized image data, and reference numeral 60 denotes an LB for outputting binary image data.
An output device such as a P (laser beam printer); 70, a display device such as a high-definition CRT (cathode ray tube) for displaying binary image data; 80, a control unit for controlling the entire system;
Is an input unit such as a keyboard for giving instructions to the system.

In order to transmit the setting conditions of these peripheral devices to the binarization processing unit 20, the system includes a signal line 61 for transmitting the resolution of the connected LBP 60, a signal line 71 for transmitting the resolution of the CRT 70, and a signal for transmitting the resolution of the reading unit. The control section 80 has a weight parameter determination section 85 for determining a weight coefficient, and a signal line 21 for transmitting a selection result of a weight parameter to be used to the binarization processing section 20.

Now, a system with a CRT with a resolution of 200 dpi
Consider a case where a dot picture is input at a resolution of 00dpi. First, as an example showing the features of the present invention, a case where a halftone picture document is input on the entire surface will be described. In this case, the binarization processing unit 20 binarizes the multi-value data input from the document reading unit 10 by the average error minimum method. The binarization method can be selected by the mode switch 25 or the like, which will be described later. The weighting coefficient used in the binarization processing unit 20 is selected from a plurality of types set in advance in the parameter switching unit 27 according to the thinning rate applied at the time of display. The thinning rate is determined by the ratio between the resolution of the image reading unit 10 and the resolution of the high-definition CRT 70. The resolution of the image reading unit 10 is transferred to the parameter setting unit 85 in the control unit 80 via the signal line 11 and the resolution of the high definition CRT 70 is transferred via the signal line 71. The parameter setting unit 85 compares the two, determines a thinning rate, and transfers a parameter code corresponding to the thinning rate to the parameter switching unit 27 in the binarization processing unit 20 via the signal line 21. Instead of transferring the parameter code, a method of storing the weighting factor itself in the parameter determination unit 85 and transferring it to the binarization processing unit 20 according to the resolution at the time of input / output can also be executed. In this case, the parameter switching unit 27 has a function of rewriting the contents. In addition, instead of a signal from the connected input / output device, the thinning rate can be set by an instruction from the input unit 90, a switch connected to the control unit 80, or the like.

The binarized image data is stored in the buffer memory 30 and displayed on the high-definition CRT 70. The binary image data displayed on the high-definition CRT 70 is an image obtained by thinning out the data in the buffer memory by half both vertically and horizontally. Here, the display on the high-definition CRT is described as an example, but the same applies to the case where the printer 60 and an external output device are connected through a communication terminal though not shown in the figure.

As a result, it is possible to display a thinned image with little image quality deterioration on a 200 dpi CT while storing 400 dpi binary image data in the buffer memory 30. Therefore, for example, when the resolution of the CRT and the LBP connected to the system are different, for example, when the CRT is 200 dpi and the LBP is 400 dpi, 400 dpi
By inputting an image at a resolution of i and creating a binary image of 400 dpi by this method, both output devices can output a suitable image.

On the other hand, a large amount of binary image data stored in the buffer memory 30 can be stored as an image database using, for example, an optical disk. FIG. 4 shows a configuration example in that case. In the figure, the code decoding unit 45 uses a known means.
A part for encoding and decoding binary image data, a data storage part
Reference numeral 40 denotes a large-capacity data storage device such as an optical disk. The binary data is transferred bidirectionally between the buffer memory and the data storage unit. When displaying an already stored image, the data in the data storage unit 40 is called,
After decryption, it is expanded and displayed in the buffer memory.

Next, the internal configuration of the binarization processing unit 20 for implementing the above processing will be described. FIG. 5 shows an example of the configuration of the binarization processing section 20.

This example shows a configuration example in the case where pseudo halftone processing is performed by a minimum average error method as a binarization method. Other pseudo halftone methods and a method of selecting one of a plurality of types of binarization methods can also be realized, which will be described later.

The multi-valued image data Y ₁ input from the document reading unit 10 is input to an adder 210 via a signal line 201. The adder 210 adds the input image data Y ₁ (x, y) and the error data E (x, y) input from the signal line 289 to obtain multi-value data F (x, y).
y) is output. The multi-valued data F (x, y) is compared with a threshold value T by a comparator 220, and the binary image data Y ′ ₁ (x, y) is determined based on the result. Here, the threshold T is provided by the register 235.

In the case of the minimum average error method, it is necessary to calculate error data E (x, y) for binarization processing. The error data E (x, y) is obtained by adding a predetermined weighting coefficient δ (to the error ε generated when the multivalued data F near the binarized (x, y) is binarized. m, n). By selecting the weight coefficient δ (m, n) in accordance with the system setting conditions, it is possible to prevent image quality deterioration due to thinning. Next, the operation of each unit in the process of calculating the error data E (x, y) will be described. When the weighting factor is (a) in FIG. 3, E (x, y) is obtained by the following equation.
In the figure, * is a pixel to be binarized at that time, and corresponds to the pixel at the coordinates (x, y) described here.

E (x, y) = 1/8 [ε (x−1, y−1) + ε (x + 1, y−1) + ε (x−2, y) + ε (x, y−2) + 2εε (x , y−1) + ε (x−1, y)}] In the case of FIG. 3 (b), it is obtained by the following equation.

E (x, y) = 1/8 [ε (x−2, y−2) + ε (x + 2, y−2) + ε (x−4, y) + ε (x, y−4) +2 ｛ε (x , y−2) + ε (x−2, y)}] On the other hand, the error ε of each pixel is either the difference between the multi-value data F and 0 or the difference between the maximum value Fmax that F can take. . This value is obtained as follows. At some point,
The comparator 220 outputs the multivalued data F (x ₂ −1, y ₂ ) of the coordinates (x ₂ −1, y ₂ )
_{When 2} ) is binarized, F (x ₂ −1, y ₂ ) is input to a comparator 220, a differentiator 225 and a selector 227. Differentiator
225 is the maximum possible value of the converted image data F (x, y)
The difference between Fmax and F (x ₂ -1, y ₂ ) is output and sent to selector 227. For example, when S ₂ (x, y) takes a value from 0 to 255, Fmax = 255, and the differentiator 225 outputs 255−S ₂ (x ₂ −1, y ₂ ). Also, here, if F (x ₂ -1, y ₂ )> Fmax, the differentiator 225 outputs 0. Selector 227
Is F (x ₂ −1, y ₂ ) or Fmax−F according to the following conditions:
(X ₂ -1, y ₂ ) is output as an error ε (x ₁ -1, y ₁ ).

ε (x, y) = F (x, y): Q (x, y) = 0 Fmax−F (x, y): Q (x, y) = 1 Output ε (x ₂ −1, y ₂ ) is sent to the line buffer 250 or the like. The process of obtaining E (x, y) from the error ε is as follows.
This is executed by the latch and line buffer and the product-sum operation circuit 280. Therefore, a part of the parameter switching unit 27 in FIG. 1 also exists in the product-sum operation circuit 270. 243,261,262 indicated as [Latch × 2] in the figure
Is obtained by connecting two latches in series.

Next, the operation of each part when obtaining Q (x, y) will be described. When the binary data Q (x−1, y) is determined by the binarization processing unit, the error ε (x,
y), ε (x−1, y), ε (x−2, y), ε (x−4, y),
ε (x, y−1), ε (x−1, y−1), ε (x + 1, y−
1), ε (x, y-2), ε (x−2, y−2), ε (x + 2,
y-2) and ε (x, y-4) are input. Product-sum operation circuit 28
0 is the sum of 11 types of input multi-value data and 8 * E
(X, y) is calculated and sent to the shift register 287.

Switching of the parameters is realized by changing the weight coefficient δ used in the product-sum operation circuit 280 according to the thinning rate. Accordingly, the parameter switching unit 27 shown in FIG. 1 exists in the product-sum operation circuit 280. In addition.
The changing means will be described later.

The shift register 287 shifts the input multi-value data to obtain E (x, y). E (x, y) obtained
Is input to the adder 210 via the signal line 289.

When x = 1 or y = 1, a part of ε required to obtain E (x, y) may not exist. In this case, for example, E (x, y) is determined using a value recorded in the line memories 250 to 270 in advance.

FIG. 6 shows an example of the product-sum calculator 280 having a function of changing the weight coefficient according to an external instruction. In the figure, shift registers A310 and A320 receive an input error ε (x−1,
y) and ε (x, y-1) are always doubled. The shift registers B315 and 330 output the errors ε (x−2, y) and ε (x, y) according to the output of the parameter determination unit 85 sent through the signal line 21.
-2) 1 or 2 times. Meanwhile, adder
Eight gates 311 to 341 placed before 350 control whether or not to pass data according to an instruction from the signal line 21. When data is not passed, 0 is input to the adder 350 from the gate. By using the weighting factor of FIG. 3 (a) which does not consider the thinning process at the time of binarization,
The operation of each unit will be described by taking binarization as an example. "0" is output from the signal line 21 and the shift registers B315 and B330
Outputs the input value without shifting it. On the other hand, among the eight gates, four of 311,321,322,323 are input error data ε (x−2, y), 2 × ε (x, y−1), ε (x−1, y−).
1), ε (x−1, y + 1), and 312, 331, 332, 341 output a value “0”. The shift registers B315 and 330 output ε (x−2, y) and 2 × ε (x, y−1). Here, the adder 350 receives the above six types of error data, performs addition, and outputs error data 8 × E (x, y).

On the other hand, when the binarization process using the weighting coefficients shown in FIG. 3B is performed, "1" is output from the signal line 21, and accordingly, the shift register B and each gate operate. Thus, the adder 350 has six types of errors 2 × ε
(X−2, y), ε (x−4, y), 2 × ε (x, y−2), ε
(X−2, y−2), ε (x + 2, y−2), ε (x, y−4)
Is entered. That is, the shift register and the gate to which the signal line 21 is connected correspond to the parameter switching unit 27 in FIG.

According to the above method, the average error minimizing method can be executed while arbitrarily changing the weighting coefficient. In this example,
The weight coefficient was changed by the gate and shift register,
Alternatively, a ROM or a RAM (Random Access Memory) can be used. In particular, when a RAM is used, the parameter determination unit can output the weight coefficient itself and load it through the signal line 21 in addition to outputting the parameter code.

On the other hand, in the case of the pseudo halftone processing by the systematic dither method, the same effect can be obtained by changing the threshold value for binarization according to the thinning rate. In this case, an example of the system configuration is shown in FIG. The configuration of the system is almost the same as FIG. Here, the switch 25 instructs the binarization processing unit 20 to perform the binarization mode based on the systematic dither method. The matrix determination unit 86 outputs a code for selecting a dither matrix to be calculated and applied from the resolution of the input / output device in the same manner as the parameter determination unit in FIG. On the other hand, the matrix selection section 26 in the binarization processing section 20 has a plurality of dither matrices, and selects one of them according to the code input from the matrix determination section 86 to execute binarization.

FIG. 8 shows an example of a binarization processing section according to this method. In the figure, a ROM 230 is a threshold ROM storing a plurality of types of dither matrixes. The multivalued image data sent from the document reading unit 10 is input to the comparator 220 via the signal line 201. The comparator 220 compares the multi-valued image data with a threshold value input from the threshold value ROM, and outputs the result of the comparison as binary data. The threshold ROM outputs a threshold signal using the lower information of the image address input from the control unit 80 through the signal lines 204 and 205 as an address. Here, by using the matrix code input through the signal line 21 as an address, the dither matrix to be used can be switched by the output of the matrix determination unit 86.

FIG. 9A shows an example of the threshold value used in the present invention.
This figure shows Baye, widely known as dither matrix.
The dither matrix called r (Bayer) type is 8
× 8 pixels as a unit. The range of the input data is 65 levels from 0 to 64. FIGS. 9B to 9D show output images obtained by binarizing the images having the luminances of 16, 32, and 48 using the threshold value of FIG. The threshold is set for the purpose of evenly dispersing the density. When the image binarized by the matrix is thinned out to 1/2, the distribution of black pixels is as shown in FIG. As a result, the original shading is not expressed at all.

Therefore, when it is assumed that the thinning process is performed at the time of display, the threshold value is determined so that the image thinned by half becomes a reduced image of the original image. FIG. 11 (a) shows an example of the threshold value. FIG. 11 (b) shows an output image obtained by binarizing an image having a luminance of 16, 32, or 48 using the threshold value.
12 (b) to 12 (d) show images in which each figure is thinned down to half length and width. The white / black ratio of the image after thinning is
Since it is the same as before thinning, unnatural image quality does not occur.

Next, the parameter determination unit 85 in FIG. 1 will be described. The parameter determination unit has a function of inputting the resolution of an input / output device connected to the system and outputting a code specifying a weighting coefficient according to the combination. Therefore,
The matrix determination unit 86 in FIG. 7 can be realized with the same configuration.

Although there are many means for realizing this function, a method using a ROM will be described as an example.

FIG. 13 shows the contents of the ROM used as the parameter determination unit 85. In this case, the number of input / output devices to be connected is four, such as a document reading device, a high-definition CRT, and a printer, and the resolutions are two types, 200 dpi and 400 dpi. F (SCN), F (C
RT), F (PRT), and F (TRM) indicate codes representing the resolution of a scanner, a CRT, a printer, and an external output device connected through a communication line, respectively.
In this case, it is set to “1”. On the other hand, F (out) indicates the value of the output code.

Now, if the resolution of the high-definition CRT 70 is 200 dpi, the flag signal F (CRT) indicating the resolution becomes “0” indicating 200 dpi. On the other hand, the resolution of the input device is 400 dpi, and the flag signal F (SCN) indicating the resolution is “1”. In this case, the parameter determination unit 85 outputs a code “1” indicating the thinning process according to FIG. Similarly, by setting codes corresponding to various combinations of resolutions, it is possible to support various combinations of devices.

Further, the parameters can be set not only as described above but also manually by a switch connected to the input unit 90 or the control unit 80, for example.

By the means described above, pseudo halftone processing can be realized using a weight coefficient or a threshold value according to the difference in resolution. Next, a case where a document in which various types of eye media are mixed will be described as an applied embodiment.

FIG. 14 is an example of a configuration of a system for a document in which various areas such as characters and photographs are mixed. In the figure, reference numeral 15 denotes an area determination unit for identifying each area of an image from a document image by an existing method. The areas to be identified are a halftone area in which the pseudo halftone processing is to be executed at the time of binarization, and a line graphic area to be simply binarized. Also, by identifying a halftone area into a halftone image area such as a photograph and a halftone area,
Further effective processing can be realized. In the figure, reference numeral 22 denotes a binarization processing unit (1) for performing binarization processing by the average error minimization method, reference numeral 23 denotes a binarization processing unit (2) for applying an organized dither method, and reference numeral 24 denotes a simple target for a line figure. A binarization processing unit (3) for executing the binarization processing is provided with a selector 29 for selecting one of the three types of binarization processing from the result. The selector 29 is the mode switch 25
It operates according to an instruction from the device and an output of the area determination unit 15.

FIG. 15 shows an example of the binarization processing section 20 using the output of the area identification section. In this example, a binary image with a fixed threshold value, a systematic dither image, and a binary image with a minimum average error method are output from signal lines 291, 292, and 293, respectively.
Reference numeral 295 selects one of the three types of input binary image data and outputs it from a signal line 299. Selection is made by the area judgment unit 1
This is performed by the area code input from the signal line 5 through the signal line 19.

In this case, the mode indicating the input of the mixed document is instructed by the switch 25, and the selector 295 operates according to the area determination result. The relationship between the determination result and the output image is as follows. The line graphic area is a binary image with a fixed threshold,
The halftone image selects the output of the systematic dither method or the minimum error method by setting the switch 21 for the halftone image, and selects the output of the minimum error method. The method of the binarization processing by each method is the same as in FIG. 5 described above.

〔The invention's effect〕

According to the present invention, in the image processing system, even when the resolution of the output device is lower than the resolution of the input unit,
Even if the pseudo halftone image is output at high speed by the thinning process, high image quality can be obtained. In particular, by the thinning-out processing, a random dither method such as the average error minimizing method, which has been significantly deteriorated, can be output at high speed. For this reason, in the image processing system, the pseudo halftone processing by the random dither method can be applied, so that a pseudo halftone image free from moire with respect to the halftone dot image can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an example of a basic configuration of an image processing system using the present invention, FIG. 2 is a diagram illustrating a thinning process of image data, and FIG. 3 is a feature of the present invention.
FIG. 4 shows an example of a weighting factor for realizing the average error minimization method with no correlation with adjacent pixels. FIG. 4 shows an example of a mode in which the document image processing system proposed in the present invention operates at the time of image input. The diagram shown in FIG. 5, FIG.
FIG. 6 is a diagram showing a configuration of a binarization processing unit, FIG. 6 is a diagram showing operation details of a product-sum operation unit, FIG. 7 is a block diagram showing an example of a configuration when an organized dither method is used, and FIG. FIG. 9A shows an example of a dither matrix and a binarization processing result using the dither matrix. FIG. 9A shows a dither matrix in which image quality hardly deteriorates even in 1/2 thinning processing. To (d) show an example of a result of the binarization processing using the same, FIG. 10 shows a distribution of black pixels when the matrix shown in FIG. 9 is used, and FIG. 11 shows the present invention. A dither matrix that does not easily cause image quality degradation even during 1/2 thinning processing used in
FIG. 12 is a diagram showing an example of a binarization processing result using the same, FIG. 12 is a diagram showing an example of a result obtained by thinning out an image binarized by a dither matrix used in the present invention by half, and FIG. , A diagram for explaining the operation of the parameter determination unit, FIG. 14,
FIG. 11 is a block diagram showing an example of a configuration in the case of using area identification,
FIG. 15 is a diagram illustrating a binarization processing unit when an area identification signal is used.

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor, Tsuguo Takahashi 2880 Kozu, Kozuhara, Odawara-shi, Kanagawa Prefecture Inside the Odawara Plant, Hitachi, Ltd. 72) Inventor Hidefumi Masuzaki 2880 Kozu, Odawara City, Kanagawa Prefecture, Japan Odawara Plant, Hitachi, Ltd. (72) Inventor Yasuo Kurosu 292, Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture, Japan Microelectronics Equipment Development Laboratory, Hitachi, Ltd. 56) References JP-A-60-223377 (JP, A) JP-A-1-1611971 (JP, A) JP-A-1-3055663 (JP, A) JP-A-1-258555 (JP, A) 63-155950 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 1/40-1/409 H04N 1/46 H 04N 1/60

Claims (9)

(57) [Claims] 1. A means for binarizing data read as a multi-value digital image, means for reading a multi-value digital image, first resolution detecting means for detecting the resolution of the read image, Means for binarizing the read multi-valued image by an average error minimization method, means for accumulating a plurality of types of coefficients used in the average error minimization method, means for outputting an image after binarization, and image output A second resolution detecting means for detecting the resolution of the input / output image, a means for calculating the ratio of the resolution of the input and output images obtained by the first and second resolution detecting means, and the plurality of types stored based on the ratio of the resolutions. Means for selecting one of the coefficients of the average error minimization method. 2. An apparatus for binarizing data read as a multi-valued digital image, means for binarizing the multi-valued image by an average error minimization method, and storing a plurality of types of coefficients used in the average error minimization method. Means for specifying one of the plurality of coefficients of the average error minimization method stored from outside, and means for selecting one of the coefficients stored by specification from the outside. Image processing apparatus. 3. An apparatus for binarizing data read as a multivalued digital image, means for reading the multivalued digital image, first resolution detecting means for detecting the resolution of the read image, Means for binarizing the read multi-valued image by an average error minimization method, means for accumulating a plurality of types of coefficients used in the average error minimization method, means for outputting an image after binarization, and image output A second resolution detecting means for detecting the resolution of the input / output image, a means for calculating the ratio of the resolutions of the input and output images obtained by the first and second resolution detecting means, and an average error minimizing method using the ratio of the resolutions. An image processing apparatus comprising means for selecting a coefficient of (i). 4. An apparatus for binarizing data read as a multi-value digital image, comprising: an apparatus for reading a multi-value digital image; and means for binarizing the read multi-value image by an average error minimizing method. Means for accumulating a plurality of types of coefficients used in the average error minimization method, means for outputting a binarized image, means for externally inputting information, and coefficients used in the average error minimization method based on the information. Selecting means for selecting an image. 5. The image processing apparatus according to claim 1, wherein said image processing apparatus has a coefficient for setting an error propagation coefficient to a pixel adjacent to at least one of the coefficients used in the average error minimization method to be zero. Characteristic image processing device. 6. The image processing apparatus according to claim 5, wherein
An image processing apparatus comprising: means for transmitting an error to a neighboring image when performing a binarization process by pseudo halftone processing; means for transmitting an error every other pixel in the vertical and horizontal directions. 7. A means for reading data read as a multi-valued digital image, comprising: a device for reading a multi-valued digital image; a means for detecting the resolution of the read-out image; Means for binarizing the image data by the average error minimization method, means for accumulating a plurality of types of coefficients used in the average error minimization method, means for setting the resolution of the device that outputs the image, Means for calculating the ratio of the resolution of the input image, means for selecting one of the plurality of coefficients of the average error minimization method stored from the resolution ratio, and means for storing the binarization processing result. An image processing apparatus comprising: 8. A means for binarizing data read as a multi-valued digital image, means for externally setting the resolution of the multi-valued image, and means for converting the read multi-valued image to a binary value by an average error minimizing method. Means for accumulating a plurality of types of coefficients used in the average error minimization method, a device for detecting the resolution of a device for outputting the image, and a device for detecting the detected resolution and the resolution of the set input image. A means for calculating a ratio, a means for selecting one of the plurality of stored coefficients from the resolution ratio, and a means for displaying or storing the binarization processing result. Image processing device. And means for inputting the image as a multi-valued image when the image is read and stored as digital data, means for binarizing the multi-valued image by pseudo halftone processing based on a minimum mean error method, Means for storing a binarized image, means for setting the resolution of the device for outputting the stored image, and means for calculating the ratio of the set resolution to the resolution of the input image, Means for switching the coefficient of the average error minimization method when binarizing the multi-valued image by the pseudo halftone processing based on the calculation result, and executing the binarization processing using a coefficient suitable for a preset output device An image processing apparatus for storing the binary image.