JPH11353475A - Image processor and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor and its method which eliminate density irregularity due to the directivity of a process generated at the time of a binarizing process and can obtain an image of high quality enabling top/bottom and right/left symmetrical gradation representation. SOLUTION: Pixels at odd-numbered positions on a raster of a multi-level image are held in a multi-level image memory part 102 and pixels at even- numbered positions are binarized by an image processing part 1 and held in a binary image composition part 104. The odd-numbered pixels are read backward out of the multi-level image memory 102 accessed on an LIFO basis and binarized by the image process part 1, and then put together with the even- numbered pixels by the binary image composition part 104.

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 and method, and more particularly to an image processing apparatus and method for performing a binarization process that reproduces halftone of image data.

[0002]

2. Description of the Related Art In a conventional image processing apparatus which performs print output by handling a multi-valued image, in order to reproduce a halftone in a print image, it was necessary to perform binarization for reproducing a halftone. As the binarization method, for example, a dither method and an error diffusion method are known. The former compares the level of each pixel of the multi-valued image data with matrix data in which a plurality of predetermined thresholds are set, and based on the comparison result,
This is a method of performing a binarization process and outputting binary image data as a result to express a halftone image. In the latter, the result of diffusing the binarization error between the output pixel data and the input pixel data quantized for each pixel of the multi-valued image data into the peripheral pixel data is represented by a binary value based on a set predetermined threshold. This is a method of expressing a halftone image by performing binary processing and outputting binary image data as a result. In particular, in recent years, the error diffusion method has come to be used more often because it is possible to achieve both gradation and resolution as compared with the dither method.

FIG. 8 is a block diagram showing an example of the configuration of a conventional image processing apparatus using this error diffusion method, and its operation will be described below. According to FIG.
It comprises a memory unit 2, an error calculation unit 3, a multi-valued image input unit 4, a density correction unit 5, a binarization unit 6, a binarization error generation unit 7, and a latch unit 8.

The memory unit 2 stores and holds one line or a plurality of lines of error data generated in each pixel, and the error calculation unit 3 calculates an error with respect to the current pixel based on the error data held in the memory unit 2. Calculate the quantity. Then, after adding the error amount calculated by the error calculation unit 3 to the multi-value read data input from the multi-value image input unit 4 in the density correction unit 5, the binarization unit 6 The read data density corrected by the correction unit 5 is
Performs binarization processing and outputs binarized data. The binarization error generating section 7 calculates a binarization error in the pixel based on the read data density-corrected by the density correction section 5 according to the binarization data input from the binarization section 6. In the latch section 8,
The output timing of the binarized error data output from the binarized error generator 7 is adjusted and output to the memory 2 and the error calculator 3.

FIG. 9 shows an example of an error diffusion filter used in error diffusion processing in a conventional image processing apparatus. FIG. 9A shows the distribution ratio of the current pixel indicated by * and its surrounding pixels, and FIG. 9B shows error data of the current pixel density. Here, assuming that the corrected density data with respect to the current pixel density data Gm, n is G'm, n, the error calculation unit 3 and the density correction unit 5 perform a process corresponding to the calculation represented by the following equation (1). Will be

G'm, n = Gm, n + (4 (Em, n-1) +2 (Em-1, n) + (Em-1, n-1) + (Em-1, n + 1) ) / 8 (1) That is, the error with respect to the current pixel density data Gm, n is generated based on the error data of the previous pixel of the previous line generated by the binarization error generation unit 7 and held in the memory unit 2. The amount is calculated by the error calculator 3. Then, the error amount is added to the current pixel density data Gm, n at a predetermined rate by the density correction unit 5 to perform density correction, and the corrected density data G′m, n is Is subjected to the binarization processing in.

Here, if the binarized signal based on the corrected density data G'm, n is Bm, n, the binarization error is Em, n, and the binarization threshold is T, G'm, n ≧ T When Bm, n = 1, Em, n = G'm, n-T When G'm, n <T, Bm, n = 0, Em, n = G'm, n Are diffused sequentially to subsequent pixels as a binarization error.

FIG. 10 shows an example of the detailed configuration of the error calculation section 3. According to the figure, the error calculation unit 3 performs the error calculation based on the above equation (1), so that the binarization error Em, n-1 of the previous pixel input from the latch unit 8 and the binarization error Em, n-1 input from the memory 2 The binarization error Em-1, n-1, Em-1, n, Em-1, n + 1 of the previous line to the multiplier 3a.
Multiply by × 2, × 4, × 1, × 1 with ~ 3d, add the result by adders 3e ~ 3g, and further divide by divider 3h
(× 1/8), and then output the quotient and remainder. The value of this quotient is a correction value for the subsequent density correction. In the example of FIG. 10, since the multiplier and the divisor can all be represented by 2 to the n-th power, the multiplication / division can be easily realized by the bit shift operation.

In the above-mentioned conventional error diffusion method, processing is usually performed in raster order as shown in FIG. In the figure, each horizontal line, for example, lines 10, 11, 12
Represents a pixel data row. In normal raster order,
The lines are processed from top to bottom, ie from line 10. Each line is processed from left to right as indicated by arrows in the figure. Therefore, the line 10 is processed first from left to right, and the next pixel to be processed after processing the rightmost pixel of the line 10 is the leftmost pixel of the line 11, and so on in the same manner in raster order. This order is indicated by the dotted line 13.

As described above, in normal raster order processing, since the accumulation of errors is small at the upper left of the image at the start position of the processing, the density of the upper left is low in the binarized image by error diffusion. An abnormality due to a so-called start delay occurs in which the concentration gradually increases toward the lower right.

Conventionally, a zigzag processing order has been proposed in order to decompose a pattern caused by the processing order. Here, an example of the zigzag processing order is shown in FIG. Again, each horizontal line, such as lines 14, 15, 16 represents a pixel data row, and the transition from one scan line to the next
Shown at 7. The processing in this case is performed in a zigzag manner, such as proceeding from left to right in row 14, then proceeding from right to left in row 15, then proceeding from left to right in row 16, and so on.

[0012]

However, in the above-mentioned conventional example, the binarization processing by the above-described dither method or error diffusion method is sequentially performed on the multi-valued image, for example, from the upper left to the lower right of the image. As a result, density unevenness due to the directionality of processing may occur in an output image.

For example, when the original image is a solid image in which all pixels have the same gradation, the same pattern of the dither matrix appears due to binarization by the dither method. Here, FIG. 6A shows a Bayer pattern as an example of the dither matrix. The pattern shown in FIG. 7A is obtained when the original image is input as 8-bit data.
As shown in FIG. 6B, it is described as a pattern of a threshold value for binarization. In (b) of FIG. 6, for example, when a solid image in which the values of all pixels are “100” is input as an original image and binarized, a binary image as shown in (c) of FIG. 6 is obtained. Is obtained. According to (c) of FIG. 6, since the dither pattern forming the block of the predetermined size is not symmetrical, the uniformity as an image is lost in the binary image obtained as a result of the dither processing, and the specific dither The pattern is noticeable.
In other words, it is difficult to perform a gradation expression symmetrical in the vertical and horizontal directions as a block, and density unevenness due to directionality has appeared in the entire image.

In the above-described binarization by the error diffusion method, by performing the above-described processing in the zigzag processing order, the loss of left-right symmetry in the entire image is eliminated to some extent. However, since the vertical symmetry has not been improved, the problem caused by the start delay has not been basically solved. In addition, the quality of the output image may be deteriorated due to the unnaturalness of the zigzag processing order itself.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and eliminates density unevenness due to the directionality of the processing that occurs during the binarization processing, and provides a vertically and horizontally symmetrical gradation expression. It is an object of the present invention to provide an image processing apparatus and a method capable of obtaining a high quality image that can be obtained.

[0016]

As one means for achieving the above object, the image processing apparatus of the present invention has the following arrangement.

That is, a classifying means for classifying the multi-valued image data into at least two attributes, a binarizing means for performing binarizing processes having different directions on the classified image data of the respective attributes, Synthesizing means for synthesizing the binarized image data of each attribute.

Further, a classifying means for classifying the multi-valued image data into at least two attributes, a determining means for determining a direction of the binarization processing in accordance with the classified attributes, and a determining means in accordance with the determined direction Binarizing means for performing a binarizing process on the multi-valued image data.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment according to the present invention will be described below in detail with reference to the drawings.

<First Embodiment> FIG. 1 is a block diagram showing a main configuration of an image processing apparatus 100 to which an image processing method according to the present embodiment is applied. According to FIG.
Reference numeral 100 denotes a selector 101 that divides multi-valued original image data input in raster units into two images according to odd-numbered pixels and even-numbered pixels, and a multi-valued memory that stores the two divided images. Image memory unit 102, selector 103 for switching between the output of selector 101 and the output of image memory unit 102, selector
An image processing unit 1 for performing binarization by error diffusion processing on the image data selected in 103, a binarizing / combining unit 104 for combining two binary images including odd-numbered and even-numbered pixels,
Consists of

Here, the multi-valued image memory unit 102 is accessed by LIFO format control in which the writing and reading directions are reversed. The binary image synthesizing unit 104 is specifically configured by a binary image memory, and with respect to the image separated into two by the selector 101, the binary image composed of the even-numbered pixels is written to the even-numbered pixels, The binary image composed of the odd-numbered pixels is composed so as to be written in the odd-numbered pixels, thereby synthesizing the two binarized images.

Next, referring to the flowchart of FIG.
The binarization processing in the present embodiment will be described in detail.

First, when the process is started in step S200,
In step S201, the pixel counter X and the raster counter Y are initialized to 0, and in step S202, the selector 103 is switched so that the input from the selector 101 is output. Then, in step S203, reading of the multi-value image data of the raster Y is started. Here, if the pixel counter X is an even number in step S204, that is, if the read image data is an even-numbered pixel, the image processing unit 1 performs binarization by error diffusion processing in step S205, and 2 in the image synthesis unit 104
The data is sequentially written into even-numbered pixel positions from the upper left corner of the value memory. On the other hand, if the pixel counter X is odd in step S204, the image data is stored in the multi-valued image memory unit 102 in step S206.

In step S207, 1 is set to the pixel counter X.
It is determined in step S208 whether or not the pixel counter X is equal to Xend indicating “the number of pixels for one raster−1” of the multi-valued image data. If they are not equal, the processing for one raster has not been completed, and the processing of steps S204 to S207 is repeated. If the pixel counter X is equal to Xend, data processing for one raster has been completed, and 1 is added to the raster counter Y in step S209. In step S210, the raster counter Y is similarly compared with the number of rasters in the sub-scanning direction, and Yend indicating "raster number-1" of the multilevel image data.
If they are not equal, the processing of all images has not been completed yet, so the pixel counter X is cleared to 0 in step S211 and the processing of steps S203 to S209 is repeated. On the other hand, if the raster counter Y is equal to Yend, the entire processing of the multi-valued image data is completed, that is, the even-numbered and odd-numbered pixels on the raster are separated, and the binarization processing is completed for the even-numbered pixels. Will be.

FIG. 3 shows an example of an image separated by the above processing. (A) of FIG. 3 shows multi-valued image data as an original image, and (b) and (c) of FIG. 3 are separated images composed of even-numbered and odd-numbered pixels, respectively. The image composed of the even-numbered pixels shown in (b) of FIG. 3 has already been binarized and is stored in the binary memory of the binary image synthesizing unit 104.
The image composed of the odd-numbered pixels shown in (c) is stored in the multi-valued image memory unit 102.

In the present embodiment, it is necessary to binarize the image composed of the odd-numbered pixels stored in the multi-valued image memory unit 102 and synthesize the image with the even-numbered image. Hereinafter, this processing will be described.

Returning to FIG. 2, in step S212, the pixel counter X
Then, the raster counter Y is initialized to 0, and the selector 103 is switched in step S213 so that the input from the multi-value image memory unit 102 is output. Then, in step S214, reading of the multi-value image data of the raster Y from the multi-value image memory unit 102 is started. Here, reading of image data from the multi-valued image memory unit 102 is performed in the LIFO direction in the direction opposite to the writing direction.
Since the reading is performed in the format, the image is read from the lower right to the upper left. The image data read from the multi-valued image memory unit 102 is subjected to binarization by an error diffusion process in the image processing unit 1 in step S215,
The data is sequentially written to the odd-numbered position from the lower right corner of the binary memory in the value image synthesizing unit 104. In other words, the binary image
On the binary memory, the data of the odd-numbered pixels that have been binarized are written in the opposite direction to the even-numbered pixels that have been binarized in step S205 described above so as to fill in the spaces.

Then, at step S216, 1 is set to the pixel counter X.
In step S217, it is determined whether or not the pixel counter X is equal to Xend indicating “the number of pixels for one raster−1” of the multilevel image data. If they are not equal, the processing for one raster has not been completed, and the processing of steps S215 to S216 is repeated. If the pixel counter X is equal to Xend, data processing for one raster has been completed, and 1 is added to the raster counter Y in step S218. Here, in step S219, the raster counter Y and Yend indicating "number of rasters-1" of the multi-valued image data are similarly compared for the number of rasters in the sub-scanning direction.
If not equal, the processing of all images has not been completed yet, so the pixel counter X is cleared to 0 in step S220, and the processing of steps S214 to S218 is repeated. On the other hand, if the raster counter Y is equal to Yend, it means that all the processes of the multi-valued image data have been completed in step S221, that is, all the binarization processes have been completed.

FIG. 4 shows a state of the binarization processing in the present embodiment.
Shown in FIG. 4A shows the order in which the image composed of the even-numbered pixels shown in FIG. 3B is binarized, and FIG.
(C) shows the order in which the image composed of the odd-numbered pixels is binarized. The dotted arrows 18 and 19 in FIG. 4 indicate the direction of the raster order to be processed. In the present embodiment, two images sequentially binarized from these two directions are finally combined by the binary image combining unit 104 as one image.

In the present embodiment, an example has been described in which a multi-level image memory 102 and a binary image memory in the binary image synthesizing unit 104 are provided to create a vertically symmetric halftone image. However, in the present embodiment, a halftone image that is further symmetrical in the left and right directions can be easily created by adding the following modifications. That is, the multi-valued image memory 102 is provided as a multi-valued raster memory, and the binary image synthesizing unit 104 is provided as a binary valued raster memory. Process in the same direction as the form (left to right), and for the image composed of odd-numbered pixels shown in FIG.
The binarization may be performed in the direction from the right side to the left side, and then the images may be combined.

As described above, according to the present embodiment, each of the divided multi-valued images is subjected to binarization with different directionality and synthesized, so that the binary values appearing in the generated halftone image are obtained. Density unevenness due to the directionality of the image processing can be eliminated, and a halftone image symmetrical in the vertical and horizontal directions can be obtained.

In the present embodiment, an example has been described in which a multi-valued image to be subjected to halftone processing is divided into two, binarized, and then combined, but is divided into three or more images. May be binarized with different directions. That is, it suffices if unevenness due to the binarization directionality can be removed.

<Second Embodiment> Hereinafter, a second embodiment according to the present invention will be described.

FIG. 5 is a block diagram showing a main configuration of an image processing apparatus 300 to which the image processing method according to the second embodiment is applied. In the figure, reference numeral 301 denotes a tone number conversion unit,
The multi-valued image data input in raster units is converted into the number of tones to be actually reproduced. For example, if the dither matrix to be used is 4 × 4 as shown in FIG. 6A, the multivalued image data is quantized to values of 0 to 16.

Reference numeral 302 denotes a dither pattern ROM storing a dither pattern, which stores a Bayer dither pattern shown in FIG. 6A and a pattern shown in FIG. 7 obtained by rotating the Bayer dither pattern by 180 degrees. . In the dither pattern ROM 302, the lower three bits of the output of the pixel counter 303 indicating the pixel position on the raster and the lower two bits of the output of the raster counter 303 indicating the raster position are connected as an address signal. The pattern shown in FIG. 6A and the pattern shown in FIG. 7 are switched and output according to the least significant bit of the output. That is, 1
The pattern is switched for each odd-numbered and even-numbered pixel on the raster. In addition, the vertical pattern is switched by the lower two bits of the output of the pixel counter 303, and the horizontal pattern is switched by the lower two bits of the raster counter 304.
Output from

From the dither pattern ROM 302, 5-bit dither pattern data is read and input to the comparison binarization unit 305. In the comparison binarization unit 305, the gradation number conversion unit 30
The image data quantized by 1 is compared with the dither pattern. If the image data is larger than the dither pattern data, 1 is output as a binarized output, and if it is smaller, 0 is output. As a result, the binarization process in the second embodiment is performed.

As described above, according to the second embodiment, binarization is performed by a dither pattern having a 180-degree pattern different between even-numbered pixels and odd-numbered pixels on a raster. Halftone images can be created.

In the second embodiment, the dither pattern data output from the dither pattern ROM 302 has the same number of bits as the input multi-valued image data, so that the gradation number conversion unit 301 is omitted. be able to.

Similarly, 5 bits of the output of the gradation number conversion unit 301 are used as an input to the address signal of the dither pattern ROM 302, and 3 bits from the pixel counter 303 are stored in the dither pattern ROM 302.
Together with the lower 2 bits of the bit and raster counter 304
Data may be created as a table that outputs 1-bit data, and 1-bit binary data may be output. By doing so, the comparison binarization unit 305 can be omitted.

<Other Embodiments> Even if the present invention is applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), an apparatus composed of one device (for example, For example, the present invention may be applied to a copying machine, a facsimile machine, and the like.

Another object of the present invention is to supply a storage medium storing a program code of software for realizing the functions of the above-described embodiments to a system or an apparatus, and to provide a computer (or CPU) of the system or the apparatus.
And MPU) read and execute the program code stored in the storage medium.

In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention.

As a storage medium for supplying the program code, for example, a floppy disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD
-R, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also an OS (Operating System) running on the computer based on the instruction of the program code. ) May perform some or all of the actual processing, and the processing may realize the functions of the above-described embodiments.

Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, based on the instructions of the program code, It goes without saying that the CPU provided in the function expansion board or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

[0046]

As described above, according to the present invention, 2
It is possible to eliminate high-density unevenness due to the directionality of the processing that occurs at the time of the value conversion processing, and to obtain a high-quality image capable of expressing symmetrically up, down, left, and right gradations.

[0047]

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating a binarization process according to the embodiment.

FIG. 3 is a diagram for explaining an image divided in the present embodiment.

FIG. 4 is a diagram for describing a processing direction for a divided image in the present embodiment.

FIG. 5 is a block diagram illustrating a configuration of an image processing device according to a second embodiment of the present invention.

FIG. 6 is a diagram illustrating an example of a Bayer dither matrix pattern according to the second embodiment.

FIG. 7 is a diagram illustrating a pattern obtained by rotating a Bayer dither matrix pattern in the second embodiment by 180 degrees.

FIG. 8 is a block diagram illustrating a configuration of a conventional image processing apparatus.

FIG. 9 is a diagram illustrating an example of an error diffusion filter used in a conventional error diffusion process.

FIG. 10 is a block diagram illustrating a configuration of an error calculation unit in a conventional image processing apparatus.

FIG. 11 is a diagram showing a normal raster processing order in a conventional error diffusion method.

FIG. 12 is a diagram showing a zigzag raster processing order in a conventional error diffusion method.

[Description of Signs] 1 Image processing unit 100 Image processing device 101 Selector 102 Multi-valued image memory unit 103 Selector 104 Binary image synthesizing unit

Claims (18)

[Claims] A classification unit configured to classify multi-valued image data into at least two attributes; a binarization unit configured to perform binarization processing with different directions on the classified image data of each attribute; A synthesizing unit for synthesizing the binarized image data of each attribute. 2. The image processing apparatus according to claim 1, wherein the binarizing unit performs a first binarizing process on the image data classified by the classifying unit into a first attribute, and performs image processing on the image data classified by the second attribute. 2. The image processing apparatus according to claim 1, wherein a second binarization process is performed on the second binarization process in a processing order different from that of the first binarization process. 3. The image processing apparatus according to claim 1, wherein said binarizing means performs binarization by dither processing. 4. The binarizing means according to claim 2, wherein said binarizing means performs binarization by error diffusion processing.
3. The image processing apparatus according to claim 1, wherein the image processing apparatus performs the binarization. 5. The method according to claim 2, wherein the classifying unit classifies the multi-valued image data into the first attribute and the second attribute according to a pixel position on a raster of the multi-valued image data.
The image processing device according to any one of the above. 6. The classification means classifies the multi-valued image data into the first attribute and the second attribute according to whether the pixel position on the raster is odd or even. 6. The image processing device according to claim 5, wherein: 7. The image processing apparatus according to claim 1, further comprising: holding means for holding the image data classified by the classifying means into a second attribute, wherein the binarizing means stores the image data stored in the holding means in a second attribute. 7. The image processing apparatus according to claim 2, wherein a binarization process of 2 is performed. 8. The image processing apparatus according to claim 7, wherein the holding unit is accessed in a LIFO format. 9. A classifying means for classifying multi-valued image data into at least two attributes; a determining means for determining a direction of a binarization process according to the classified attributes; To the multi-valued image data
An image processing apparatus comprising: a binarizing unit that performs a binarizing process. 10. The binarization means according to claim 2, wherein said binarization means performs binarization by dither processing.
10. The image processing device according to claim 9, wherein value conversion is performed. 11. The image processing apparatus according to claim 10, wherein said determination means determines a dither matrix to be used according to the attribute classified by said classification means. 12. The image processing apparatus according to claim 11, wherein said determining means selects one of a first dither matrix and a second dither matrix obtained by rotating said first dither matrix. apparatus. 13. The image processing apparatus according to claim 9, wherein the classification unit classifies each pixel of the multi-valued image data into an attribute according to a position of the pixel. 14. The image processing apparatus according to claim 13, wherein the classification unit classifies the multi-valued image data into odd-numbered pixels and even-numbered pixels on a raster. 15. A classification step of classifying multi-valued image data into at least two attributes, and a first binarization step of performing a first binarization process on the image data classified into a first attribute And a second 2 in a processing order different from the first binarization processing for image data classified into attributes other than the first attribute.
An image processing method comprising: a second binarization step of performing a binarization process; and a synthesis step of synthesizing the image data binarized by the first and second binarization steps. . 16. A classifying step of classifying multi-valued image data into at least two attributes; a determining step of determining a direction of a binarization process according to the classified attributes; To the multi-valued image data
An image processing method, comprising: a binarization step of performing a binarization process. 17. A recording medium on which a program code for image processing is recorded, wherein a code of a classification step for classifying multi-valued image data into at least two attributes, and a code for image data classified into a first attribute are provided. A code of a first binarization step of performing a first binarization process, and a processing order different from the first binarization process for image data classified into attributes other than the first attribute. By the second 2
A code of a second binarization step of performing a binarization process, and a code of a synthesis step of synthesizing the image data binarized by the first and second binarization steps. Recording medium. 18. A recording medium storing a program code for image processing, comprising: a code for a classification step of classifying multi-valued image data into at least two attributes; and a binarization process in accordance with the classified attributes. And a code for a determining step of determining the direction of the multi-valued image data according to the determined direction.
A code for a binarization step of performing a binarization process.