JPH0334679A - Picture processing device - Google Patents

Abstract

PURPOSE:To eliminate dots in a low-density part, a link of void in a high-density part, and a chain texture in middle and low-density parts by providing an already binarized window in the periphery of a weighted picture element, and changing and correcting the arrangement of a dot pattern in the window based on this dot pattern. CONSTITUTION:A binarized signal correcting circuit 7 provides a 3X3 window for an already binarized signal and identifies whether it is adapted to the pattern. That is, an already generated texture is recognized to check whether it is necessary that the arrangement of the binarized signal is broken to correct the regular texture or not. If it is recognized that the texture to be corrected is generated, a correction pattern is selected and the binarized signal in the window is corrected in accordance with this correction pattern. Thus, dots and a link of void in the error diffusion method and an uncomfortable texture in middle and low-density parts are eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing device that converts multi-value image information into multi-value image information having a smaller amount than the multi-value image information, or into binary image information.

[Prior art] Conventionally, there have been methods that have one or more coloring materials,
Furthermore, there are printers that can print out in binary or 3-4 values by means of modulating the dot diameter.

On the other hand, the number of image information input from an image input device such as an image scanner is generally greater than the number of values of image information that can be output from this printer.

For this reason, conventionally, multivalued image information from this input device must be converted into information with a number of values that can be output by a printer or the like. That is, input multivalued image information from an input device (for example, one color, several pits per pixel, about several tens of bits),
It must be converted into binary or 3-4 value image information that can be output by a printer.

Conventionally, as a general method of converting image information for printing out from such a printer, for example, the means of expressing halftones generally include a dither method, a density pattern method,
A conversion method such as the error diffusion method was adopted.

[Problems to be Solved by the Invention] However, the conventional conversion method has the following problems.

That is, in the dither method, resolution and gradation are contradictory conditions, and if high gradation is to be obtained, a larger dither matrix must be constructed.

Furthermore, in the density pattern method, one pixel of original information is divided into a plurality of pixels, resulting in a decrease in resolution.

Therefore, an error diffusion method has been proposed as a binarization method that satisfies both resolution and gradation to some extent. but,
The disadvantages of this method include the connection of dots and white spots, and the unpleasant chain-like texture that appears regularly, which significantly deteriorates the image quality.

For example, when considering an output device such as an inkjet recording device that holds two types of coloring materials, a dark ink and a light ink, for each color, that is, can output three values per color, differences in ink color, etc. Therefore, the conventional three-level error diffusion method has a problem in that a tone jump occurs in a certain density variation area, resulting in a false contour.

[Means for Solving the Problems] The present invention was developed for the purpose of solving the above-mentioned problems, and includes the following configuration as a means for solving the above-mentioned problems.

That is, in an image processing apparatus that converts multivalued image information into a binary signal using an error diffusion method, a window that has already been converted into a binary value is provided around a pixel of interest, and 1. A change correction means is provided for detecting the dot pattern within the window and changing and correcting the arrangement of the dot pattern within the window based on the detected pattern.

Further, a change correction means provides a window in the vicinity of the pixel of interest of the converted signal, which has already been converted into a small number of pixel signals, detects the dot pattern within the window, and changes and corrects the arrangement of the dot pattern within the window based on the detected dot pattern. Equipped with.

[Operation] In the above configuration, a window is provided in the binary signal after the multilevel image information is binarized by the error diffusion method, and the connection of dots within the window is detected, and the current signal of the pixel of interest is detected. By determining the density level of the image, the texture in the window is disturbed, and by correcting the misalignment of dots caused by the error diffusion method, the connection between dots and white spots in the error diffusion method, which has traditionally been a problem, can be corrected. In addition, unpleasant chain textures in medium and low density areas can be eliminated at low cost and with a simple configuration.

Furthermore, thin lines etc. that are disturbed by the error diffusion method can be faithfully reproduced, and the resolution is significantly improved.

Furthermore, by detecting whether the dot connections are textures or original signals of the original image, good and appropriate binarization can be performed without losing the original image information.

Furthermore, multivalued image information (several bits to lO several bits/
pixel), for example, 3 pixels per pixel using the multilevel error diffusion method.
In the halftone expression means for converting into values and about 4 values, a window is provided for the converted signal, the connection of dots within the window is detected, the generated texture is recognized, and the converted signal within the window is detected. By changing the arrangement of the dots and white spots in the multilevel error diffusion method, which have been a problem in the past, it is possible to eliminate them in a simple manner.

In addition, it is possible to detect and recognize the switching part of a collection of equal-level dots in a window, and similarly switch the arrangement in the window to output 3- or 4-value output, such as a multilevel printer. Problematic false contours can also be reduced by switching the coloring material or the like.

Furthermore, by determining whether the in-window texture is an original signal or one generated by halftone processing, good and optimal halftone processing can be achieved without losing the original image information.

[Embodiment] Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings.

[First Embodiment] FIG. 1 is a diagram showing a block configuration of an embodiment according to the present invention. The image processing device of this embodiment converts a multi-valued image signal into a binary image signal and outputs an image. It is a processing device.

In this image processing apparatus, only the binarization control functional portion of all image processing means is shown, and other necessary components are omitted from the figure. It is also possible to incorporate this image processing device part, for example, into a post-computer section (as hardware or software), an output device such as a printer, or an input device such as an image scanner.

In FIG. 1, l is stored in the ROM4, for example, the second
A central processing unit (CPU) 2 controls the entire apparatus of this embodiment according to the control procedure shown in the figure, and Ilo 2 controls an interface with a connected external device 8 that outputs a multivalued image signal. A multivalued image signal from the external device 8 is input through this l102. 3 is a RAM which is used as a line buffer for errors in the error diffusion method and input multivalued images, and 4 is a ROM in which, in addition to the above-mentioned programs, the diffusion matrix of the error diffusion method is stored.
5 is an error diffusion circuit that performs binarization processing using the error diffusion method according to a predetermined threshold; 6 is a random number generator for generating random numbers and selecting a correction pattern; 7 is a random number generator specific to this embodiment. 6 is a binary signal correction circuit that corrects the binary signal specified by the error diffusion circuit 5 according to the generated random number, and 8 is an external device that outputs a multi-value image signal and the like.

Hereinafter, the image processing of this embodiment having the above configuration will be explained as follows.
This will be explained with reference to the flowchart shown in the figure.

First, in step S1, a multivalued image signal to be binarized is input from the external device 8 and stored in a predetermined line buffer in the RAMB. Then, in the subsequent step S2, the error diffusion circuit 5 adds the errors of adjacent surrounding pixels to the pixel of interest and stores the result in a predetermined line buffer in the RAMB, similarly to the normal error diffusion method.

Then, in step S3, this error and ROMJ
Binarization processing using the error diffusion method is performed using the diffusion matrix of the error diffusion method. The processing of this error diffusion method is well known, and detailed explanation will be omitted.

Next, in step S4, a window is applied to the binarized signal using the error diffusion method. This window processing does not require a pixel of interest and is completely independent from the error diffusion method, so even if the window is scanned immediately after binarization using the error diffusion method In the case where the image data has already been binarized for one page, the image data may be subjected to binarization processing for one page. However, in terms of memory capacity, the former is much more efficient.

In general, in binarization using the error diffusion method, the connections between dots in highlight areas and the connections between white spots in high density areas often occur diagonally; The unpleasant texture is also created by dots lined up diagonally to create a regular pattern.

Therefore, in this embodiment, a window pattern configuration of nine 3×3 matrices as shown in FIG. 3 is adopted. However, this window pattern is not limited to the above example.

FIG. 3(a) is an example in which dots are connected diagonally downward to the right (1: black, 0: white). Figure 3 (b
) is an example of dots connected diagonally downward to the left, the third
FIG. 3(C) shows an example in which white spots are connected in a diagonal direction downward to the right, and FIG. 3(d) shows an example in which white spots are connected in a diagonal direction downward to the left.

Then, in the following step S5, the binary signal correction circuit 7 has already
A 3×3 window is provided for the converted signal, and the third
Identify whether the pattern shown in the diagram is met. In other words, it recognizes the texture that has already occurred, and
Check whether it is necessary to correct the regular texture by disrupting the arrangement of the valued signals. If the window pattern does not match, the 3×3 matrix window is scanned without correction. Therefore, the process immediately advances to step S8.

On the other hand, if it is recognized in step S5 that a texture that requires correction has occurred, the process proceeds to step S6, and
The value signal correction circuit 7 generates random numbers using the random number generator 6, selects a correction pattern, and corrects the binary signal within the window according to this correction pattern in step S7.

For example, suppose that the signal within the window matches the pattern shown in FIG. 3, as shown in FIG. 4(a). The data is rearranged using random numbers without changing the uniform density (the number of black rLJ) in the matrix window. Considering the permutations of 3 black data out of 9 data, there are *C5=84 ways, excluding 6 patterns in which 3 black data are arranged in the same column and row ( A pattern is selected from ecs-6)=78 patterns using random numbers as shown in FIG. 4(b).

Then, in step S8, it is determined whether or not all signal window correction processing has been completed, and if it has not been completed, step S1
Return to and continue binarization and correction processing.

FIG. 5 shows an example of processing in which texture is reduced by executing the window correction processing described above.

By sequentially correcting and scanning the 3x3 window matrix according to the control described above, the chain-like texture that occurred in Fig. 5(a) can be removed as shown in Fig. 5(b). This results in a good image broken up into pieces. In this case, the number of 1 (black) is preserved and the density does not change significantly.

In the processing described above, it is also necessary to determine whether the pattern in which dots are connected diagonally is the original signal or a chain of textures resulting from binarization.

In that case, it is best to first determine whether the multi-level data of the current signal to which the error has not been added is at a high concentration level, and then store it together with the binary signal obtained by the error diffusion method. .

In this case, for example, when "0" of the 8-bit multivalued image signal is the white level and r255J is the black level, r200J
It is sufficient to judge whether or not the concentration is high by using the threshold value as "1" if r200J or more, and rOJ if it is not r200J or more. By performing this processing, even if dots are connected diagonally during window processing, the dots will determine whether the current signal is in a high-density area or not, and texture will be generated only if it is not a high-density area. All you have to do is judge that this is the case and apply correction processing.

Also, in the r order of operations shown in the flowchart in Figure 2, the error diffusion method and correction processing are performed serially, but as mentioned above, these two processes are completely independent, so binarization does not finish. However, if parallel processing is possible, it is more efficient to operate in parallel.

[Second Embodiment] FIG. 6 shows a configuration diagram of a window matrix in a second embodiment according to the present invention.

In the figure, (a) is the matrix pattern before correction, (b) ~
(e) shows an example of the pattern after correction.

In the correction in the first embodiment shown in FIG. 4, the corrected patterns were arranged using random numbers.

However, if the random number is too strong, it becomes random noise with no regularity, which is visually undesirable. Therefore, in this embodiment, FIGS. 6(b) to (e)
), which are typical patterns that do not cause dots and crabs to approach each other as much as possible, and these patterns alone cover the correction.

That is, when the pattern in FIG. 6(a) matches the data within the window, correction is performed by sequentially arranging (b), (c), (d), and (e). Even if they are placed in this order, the same pattern will be corrected only once in four times, and it is rare that they will be placed very close together, so new textures may occur. There is little sex. Of course, these four patterns may be extracted using random numbers.

As explained above, according to this embodiment, by sequentially arranging several patterns in this manner, random noise can be suppressed and processing can be performed at high speed.

[Third example] FIG. 7 shows a third embodiment of the present invention.

This third embodiment is a development of the above-described embodiment, and is an invention that attempts to reliably correct the reproduction of thin lines through window processing.

In binarization using the error diffusion method, for example, when rOJ of an 8-bit multilevel image signal is the white level and r255J is the black level, the density level of the thin line is r255J, and the white background is rOJ.
There is no problem if you are at the level. However, except for images such as CG (computer graphics), images input using an image scanner or the like generally have a fine line density lower than r255J. This causes errors and makes it difficult to faithfully reproduce thin lines.

Therefore, in this embodiment, a 4×4 matrix window, for example, is prepared. It is assumed that the already binarized data is applied to a window and that it matches the pattern shown in FIG. 7(a).

At this time, as in the embodiment described above, a fixed threshold value (for example, r2
When the signal for determining whether or not it is a high concentration level binarized with 00J) was as shown in Figure 7(b),
It turns out that the deviation of the dots in a) is not due to the deviation from the current signal, but is due to the addition of the error of binarization by the error diffusion method.

Therefore, in this example, based on both data (a) and (b),
The correction is performed as shown in (C), and control is performed to compensate for the drawbacks of the error diffusion method.

By performing the above-described correction processing, it is possible to sufficiently improve the resolution even with the error diffusion method.

As explained above, according to the above embodiments, in the error diffusion method, problems such as dots in low-density areas, connections between white spots in high-density areas, and medium and low-density areas, which have traditionally been very problematic, can be solved. The unpleasant chain-like texture can be eliminated at low cost and with a simple configuration.

Further, if the correction processing is performed in parallel with the error diffusion method, there is no extra time loss and correction can be performed at high speed.

Furthermore, thin lines etc. that are disturbed by the error diffusion method can also be faithfully reproduced, and the resolution is significantly improved.

Furthermore, by detecting whether the dot connections are due to texture or the current signal of the original image, good and appropriate binarization can be performed without losing the original image information.

[Fourth Embodiment] In the first to third embodiments described above, processing for converting multivalued image information into binary image information has been described, but the present invention is not limited to the above examples. Needless to say, the present invention can be applied to processing for converting multivalued image information into multivalued image information having a value equal to or less than the input value.

Hereinafter, an example will be described in which the present invention is used to convert an input image signal into image information having a multi-value number equal to or less than the input multi-value number.

Each of the following embodiments can also be realized, for example, by the image processing apparatus having the hardware configuration not shown in FIG. 1 described above.

The following description will be made based on a wiring diagram etc. that schematically show image processing using the above-described block configuration.

FIG. 8 shows a block diagram of main parts of the image processing apparatus of this embodiment.

In the image processing apparatus of this embodiment, only the halftone expression control function part of all the image processing means is shown, and other components are omitted from the figure. This part can be incorporated, for example, into a host computer unit (as hardware or software), an output device such as a printer, or an input device such as an image scanner.

In this embodiment, an example in which an input multilevel signal is ternarized will be described as halftone expression.

Figure 8 shows an example of performing array conversion within a window in synchronization with the conventional multilevel error diffusion method, and please refer to the flowchart in Figure 9 which shows the specific control procedure of Figure 8. The configuration and control of this embodiment will be explained below.

First, as shown in step Sll in FIG. 9, a multivalued image signal (several bits to lO several bits) is input. This input signal signal □ is sent to the input correction means 101. For this reason, the input correction means 101 performs input correction in step S12 by adding the errors caused by surrounding neighboring pixels to the pixel of interest in a process similar to the normal error diffusion method, and performs a correction signal processing. is output to the multivalue conversion means 103 and the error distribution calculation means 105.

Multi-value conversion means 103 receives this correction signal Ixy.
is the input signal I' as in step S13.

is compared with the multi-value threshold 102 (T, , ), and converts the input multi-value signal into a 3-value signal (or 4-value signal, etc.),
P) is output as Ml and is output to the threshold value determination means 109 and the error distribution calculation means 105.

This is true even when the connected output device is a printer that can output up to 3 or 4 values for each color using, for example, dark and light ink, or dot modulation, instead of 2 values for each color. , correspondingly, it is configured to be able to express halftones.

For example, when performing three-value output, the multi-value conversion means 103 compares the input signal with a threshold value and converts it into three types of signals (P, , ) of (1, station, 0), and the actual signal is ( 2, l, 0
), which expresses intermediate tones.

At the same time, the difference calculation means 104 outputs this three-value signal (or four-value signal, etc.) P8. The difference between the input signal I'xy and the error distribution calculation means 10 is calculated as an error signal E
Output to 5. The error distribution calculating means 105 distributes the weighting of the error signal Elly using the diffusion matrix shown at reference numeral 106 in FIG. sent to.

Since the processing of the multilevel error diffusion method described above is well known, detailed explanation will be omitted.

Subsequently, in step S14, a window shown at 107 is applied to the signal ternarized by the error diffusion method in step S13. This window processing does not particularly require a pixel of interest and is completely independent from the multilevel error diffusion method, so window scanning can be performed immediately after ternarization using the multilevel error diffusion method as in this embodiment. I can do it. Furthermore, if the memory capacity is large, the window may be scanned for one page of image data that has already been ternarized. However, in terms of processing speed and memory capacity efficiency, the former is more effective.

In this embodiment, this window is composed of 16 4×4 matrices as shown in FIG. 8 107, for example.

In general, ternarization using the error diffusion method has the same problems as binarization, although there are differences in degree.

Therefore, in this embodiment, in order to solve this problem, the first
Consider a window pattern as shown in figure O. 10th
Figures (a-1) and (a-2) show examples in which dots are connected diagonally downward to the right.

("l"...black, rOJ...white, "%"...gray) In Figure 10 (b-1) and (b-2), dots are connected diagonally downward to the left. An example is shown in Figure 10 (c-
1), (c-2) are examples of white spots (including gray) connected diagonally downward to the right, and Figure 10 (d-1), (
d-2) shows an example in which white spots (including gray) are connected diagonally downward to the left.

In this way, in this embodiment, the signal that has already been ternarized is
x4 windows are provided, and in step S15 LUT (
A look-up table) is used to identify whether the pattern shown in FIG. 10 is met. That is, it is determined whether or not to correct the regular texture by recognizing the texture that has already been generated and destroying the three-value previous arrangement. If the window pattern does not match, the process proceeds to step S18 without making any correction, and the 4×4 matrix window is scanned.

On the other hand, if it is recognized that the window pattern is suitable and a texture that requires correction has occurred, the process proceeds to steps S16 and S17, and a correction process for the multi-level signal within the window is performed.

First, in step S16, random numbers are generated by the random number generator 109 and a correction pattern is selected. And step S17
Correct the 300 million issues in the window. For example, the 11th
Assume that the signal within the window as shown in Figure (a) matches the pattern shown in Figure 10. This data is rearranged within the window using random numbers without changing the equidiscretion density (number of 1 (black), number of % (gray)) within the window of this matrix.

For example, in the example of the window shown in Figure 10, considering a combination of 4 1 (black) data from 16 data, l@C
4 = 1820 patterns, excluding 8 patterns in which 4 1 (black) data are arranged in the same column and row (18C
4-s) = 1s A pattern is selected from 12 random numbers as shown in Fig. 4Cb). As a result, the 8th
The window in FIG. 107 is
8, the array of each element is converted, resulting in the window shown in FIG. 8 110.

Then, in step S18, it is determined whether the entire signal window correction process has been completed by a small amount, and if it has not been completed, the process returns to Sll.

The operating procedure shown in the flowchart in Figure 9 serially performs the error diffusion method and the correction process, but as mentioned above, these two processes are completely independent (unless the ternarization process is completed). However, if parallel processing is possible, it is more efficient to operate in parallel.

[Fifth Example] In the correction according to the fourth example shown in FIG. 11, the corrected patterns were arranged using random numbers.

However, if the random number is too strong, it will generate random noise with no regularity, which is visually undesirable.

In order to improve this point, it is desirable to list some representative patterns in which the bottom of the drive is as close as possible and to control the correction using only these patterns.

A fifth embodiment of the present invention that controls in this manner will be described below.

FIG. 12 is a configuration diagram of a window matrix in a fifth embodiment according to the present invention.

In the figure, (a) is the matrix pattern before correction, (b) ~
(e) shows an example of the pattern after correction.

In this embodiment, several representative patterns are used in which the dots do not come close to each other as much as possible, such as the examples shown in FIGS. 12(b) to 12(e), and these patterns alone cover the correction.

That is, when the pattern in FIG. 12(a) matches the data within the window, the patterns are not arranged in one type of pattern, but in sequence as shown in FIG. 12(b).

Correction is performed by sequentially arranging them as shown in (c), (d), and (e).

In this way, even if they are arranged in order, the same pattern will be corrected only once in four times in this example.
Since they are rarely placed in close proximity, there is little possibility that new textures will be generated.

Of course, these four patterns may be extracted using random numbers.

As explained above, by sequentially arranging several patterns in this way, random noise can be suppressed and processing can be performed at high speed.

[Sixth Embodiment] FIG. 13 is a block diagram of main parts of a sixth embodiment according to the present invention. Components similar to those of the fifth embodiment shown in FIG. 8 are given the same numbers and detailed explanations will be given. Omitted.

The sixth embodiment shown in FIG. 13 differs from the fifth embodiment used in FIG. 8 in that even when dots are connected diagonally within the window, the original pixel signal is originally image information of diagonal lines. A process is added to determine whether it is a texture in which dots are connected diagonally by chance due to ternarization using the error diffusion method.

That is, the multi-value conversion means Ill is added, and the multi-value conversion means 111 first compares the original human power signal Zoo (1,,) to which the error has not been added with the fixed threshold value group (T'). do. This fixed threshold value group is, for example, input signal I. When is an 8-bit multi-value signal, settings are made so that high-density areas and low-density areas can be clearly divided among the signals from 00H (white level) to FF and (black level).

For example, two threshold values are prepared: T'm,=dcH and T', ,=lE.

Then, in the case of dc engineering≦IX,≦FF, P”,,
=1 (10 as a signal).

On the other hand, if 1E□≦E□dcH, p'
,,=% (01 as a signal).

Furthermore, if O≦Ixy<IEs, P”, , =O (00 as a signal).

Then, the signal P'o, which is the output multi-value signal from the multi-value conversion circuit 111, is further dot-shifted and converted into P by an adder. In addition to the above signal, it is stored in memory as 4-bit/pixel information based on two types of ternary signals, and window processing is performed.

That is, in this embodiment, two types, the p', signal and the P,y signal, are subjected to window processing. For this reason,
Even if the P and y signals form a texture in which dots are connected in a diagonal direction, the P″8° signal will change the connection of dots in the diagonal direction from that of the original signal.
Alternatively, it is possible to determine whether the texture appears due to ternarization.

In this embodiment, if it is determined that it is due to the original signal,
Naturally, the array inside the window is not converted.

Since the other processes are the same as those in the fifth embodiment, explanations of the other processes will be omitted.

By adding the processing of this embodiment described above, the required memory capacity increases slightly, but it prevents the straight line of the original signal from being erroneously disturbed.

[Seventh Embodiment] FIG. 14 shows an example of a configuration diagram of a window to be corrected according to a seventh embodiment of the present invention.

In the seventh embodiment, not only the texture within the window is recognized, but also the pseudo contour within the window is recognized as a pattern, and the arrangement is converted.

For example, in window processing, (a) to
Consider a case where a pattern like (C) occurs. In such a case, when a multilevel printer is connected and the output is output to the printer, false contours often occur as tone jumps when changing coloring materials, etc. even when the colors are the same.

Therefore, in this example, the patterns shown in FIGS. 14(a) to (C) (dots of equal level form a cluster,
Similar to the above-mentioned embodiment, the L
Save it in UT (actually, it is shown in Figures 14 (a) to (c)).
Although there are many other types of patterns stored in the window, in order to simplify the explanation, the following explanation will be limited to the illustrated example).

Here, if it matches the pattern within the window, the array is converted using random numbers and a predetermined correction pattern, as in the fourth embodiment in Fig. 8 and the fifth embodiment in Fig. 12. , breaking the occurrence of false contours (tone jumps) in the window.

In this embodiment, as in the embodiment shown in FIG. 12, a simple ternary signal using a fixed threshold value is also stored, and window processing is performed to determine whether or not the original signal is an edge portion. This is to prevent the straight line of the original signal from being disturbed.

As explained above, according to the above-described embodiment, the connection between dots in low-density areas and white spots in high-density areas, which has traditionally been a serious problem in the multilevel error diffusion method, can be solved.
It is possible to eliminate unpleasant chain-like textures in medium and low density areas at low cost and with a simple configuration.

Further, if the correction processing is performed in parallel with the error diffusion method, there is no extra time loss and correction can be performed at high speed.

In addition, by detecting whether the dot connections are a texture or the original signal of the original image, it is possible to perform good and appropriate multilevel conversion (ternarization, quaternization, etc.) without losing the original image information. Can be processed.

[Effects of the Invention] As explained above, according to the present invention, connections between dots in low-density areas and white spots in high-density areas in the error diffusion method, and unpleasant chain-like textures in medium and low-density areas can be eliminated. cheaply,
It can be eliminated with a simple configuration.

Furthermore, by detecting whether the dot connections are textures or original signals of the original image, good and appropriate binarization or higher multivalue conversion can be performed without losing the original image information.

Furthermore, false contours that tend to occur can be similarly reduced by the multilevel error diffusion method.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment according to the present invention. FIG. 2 is a flowchart showing image processing control in this embodiment. FIG. 3 is an example of a pattern that requires window correction in this embodiment. Figure 4 is a diagram showing an example of correction using random numbers in this embodiment, Figure 5 is a diagram showing an example of execution in which the texture of a binary signal in this embodiment is removed by correction, and Figure 6 is a diagram showing an example of correction using random numbers in this embodiment. FIG. 7 is a diagram showing an example of the window correction pattern of the second embodiment of the invention, FIG. 7 is a diagram showing the correction process of the third embodiment of the invention, and FIG. 8 is a block diagram of the fourth embodiment of the invention. 9 is a flow chart showing the image processing procedure of the fourth embodiment according to the present invention. FIG. 10 is a diagram showing an example of a pattern requiring window correction in the fourth embodiment. FIG. 12 is a diagram showing an example of the window correction pattern of the fifth embodiment according to the present invention. FIG. 13 is the block configuration of the sixth embodiment according to the present invention. FIG. 14 is a diagram showing an example of a pattern requiring window correction according to a seventh embodiment of the present invention. In the figure, 1... central processing unit (CPU), 2...
・Ilo, 3...RAM, 4...ROM, 5...
Error diffusion circuit, 6... Random number generator, 7... Binary signal correction circuit, 8... External equipment, lOl... Input correction means, 103° 111... Multi-value conversion (ternary conversion) ) means, 10
4... Difference calculation means, 105... Error allocation calculation means, 106... Diffusion matrix to peripheral adjacent pixels, 10
7.110...Window, 108...Window array conversion means, 109...Random number generator. Figure (a) (b) 3 Figure (c) (b) (d) 0001000 1000100 00001 0001 0001 0 10000100010001 00101 000100010 00010001000100 000 10000 10001 0 00000101 000001 10001 000 100001 00100100100100 000100010000 10 Figure (C) ( a-1) (b-1) ψ-2) (b-2) (c-1) (cl-1) Figure 10 Q::i I Figure 1 (b) (C) CG) Figure 12 Figure 14 (C) Procedural amendment (spontaneous) 17th January 1999, Commissioner of the Japan Patent Office, 11 indicated patent applications No. 2, Title of the invention Image processing device Canon Co., Ltd. Specification of patent claims Scope column and Detailed description of the invention column ρ゛) 6. Contents of amendment (1) “Several tens of bits” in the second line of page 5 of the specification was changed to rlO
"Several bits". (2) Replace “employed” on page 5, line 9 of the specification with r.
"It was adopted," he corrected. (2) "Several bits" on page 9, line 3 of the specification is 64. Correct by a few bits j. ′-゛“1 pixel” in the 4th line of page 9 of the specification is 111 colors 1
and correct it. “Gain error” on page 13, line 2 of the specification! Correct it as j. "For white" on page 14, line 14 of the specification is corrected to "r white spot". (7) "Permutation" on page 16, line 4 of the specification is corrected to "combination 1." (8) Specification, page 17, line 10, page 21, lines 14 to 15, The "current signal" on the 5th line of page 23 is corrected to "r original signal". (9) Correct "error distribution calculation means 105" on page 25, line 15 of the specification to be j difference calculation means 104j. (10) "Output threshold value determination means 109 and error distribution calculation means 105" on page 26, line 4 to line 5 of the specification
At the same time, P xy is corrected by the difference calculation means 104J. (11) ``Input signal and threshold value'' on page 26, line 12 of the specification is corrected to ``r input signal divided into levels and threshold value for each level''. (12) Correct "3-value previous issue" on page 29, line 11 of the specification to r3-value signal 1. (13) "Figure 4" on page 31, line 2 of the specification
1 Correct as shown in Figure 1. (14) "This result" on page 31, line 3 of the specification is corrected to r, that is, j. (15) Change “small amount” in line 8 of page 31 of the specification to “end 1”
and correct it. (16) Correct the "dot" in the first line of page 36 of the specification to r bit 1. (17) The scope of claims is as attached. Correction of the scope of claims of Japanese Patent Application No. 1-167081 (1) In an image processing device that converts multilevel image information into a binary signal using an error diffusion method, a window that has already been binarized around a pixel of interest is detecting the dot pattern in the window and changing the arrangement of the dot pattern in the window based on the detected pattern;
An image processing apparatus characterized by having a change correction means for correcting. (2) The image processing apparatus according to claim 1, wherein the change correction means adds, in addition to the in-window detected dot pattern, a signal that has been binarized using a predetermined fixed threshold. (3) The image processing apparatus according to claim 1, wherein the change correction means changes the arrangement of the dot patterns within the window using random numbers. (4) The image processing apparatus according to claim 1, wherein the change/correction means changes the arrangement of the dot patterns within the window by preparing several types of correction patterns and sequentially using the correction patterns. (5) In an image processing device using an error diffusion method that converts multi-value image information into multi-value image information that is smaller than the multi-value image information, a window is provided near the pixel of interest of the converted signal that has already been converted into a smaller number of pixel signals. An image processing modification device characterized by comprising a change correction means for detecting a dot pattern within the window and changing and correcting the arrangement of the dot pattern within the window based on the detected dot pattern. (6) In addition to detecting a dot pattern obtained by converting the array conversion condition of the dot pattern within the window using a multilevel error diffusion method, the change correction means includes a window in the pixel signal that is simply multileveled using a predetermined fixed threshold value in the array conversion condition. The image processing device according to claim 5, characterized in that: (7) The image processing apparatus according to claim 5, wherein the change correction means converts the arrangement of the dot patterns within the window using generated random numbers. (8) The image processing apparatus according to claim 5, wherein the change/correction means sequentially selects a plurality of types of correction patterns provided in advance with arrangement conversion of dot patterns in the window, and performs the processing according to the selected correction pattern.

Claims (8)

[Claims] (1) In an image processing device that converts multivalued image information into a binary signal using the error diffusion method, a window that has already been binarized is provided around the pixel of interest, and the dot pattern within the window is detected and converted into a detection pattern. Change the arrangement of dot patterns in the window based on,
An image processing apparatus characterized by having a change correction means for correcting. 2. The image processing apparatus according to claim 1, wherein the change correction means adds, in addition to the in-window detected dot pattern, a signal binarized using a predetermined fixed threshold. (3) The image processing apparatus according to claim 1, wherein the change correction means changes the arrangement of the dot patterns within the window using random numbers. (4) The image processing apparatus according to claim 1, wherein the change/correction means prepares several types of correction patterns and sequentially uses the correction patterns to change the arrangement of the dot patterns within the window. (5) In an image processing conversion device using an error diffusion method that converts multi-value image information into multi-value image information that is less than the multi-value image information, attention is paid to the converted signal that has already been converted into fewer pixel signals. 1. A pixel processing device comprising: a window provided near a pixel; and change/correction means for detecting a dot pattern within the window and changing and correcting the arrangement of the dot pattern within the window based on the detected dot pattern. (6) In addition to detecting a dot pattern obtained by converting the arrangement conversion condition of a dot pattern within a window using a multilevel error diffusion method, the change correction means includes a window in a pixel signal that is simply multileveled using a predetermined fixed threshold value in the arrangement conversion condition. The image processing device according to claim 5, characterized in that: (7) The image processing apparatus according to claim 5, wherein the change correction means performs arrangement conversion of the dot pattern within the window using generated random numbers. (8) The image processing apparatus according to claim 5, wherein the change/correction means sequentially selects a plurality of types of correction patterns provided in advance with arrangement conversion of the dot patterns within the window, and performs the processing according to the selected correction pattern.