JPS63290070A - Picture processor - Google Patents

Abstract

PURPOSE:To obtain a picture which is faithful to an original picture by using specifically distributed dither matrixes in which respective elements are sequentially and equally ditributed in plural divided areas, converting digital picture signals into dithering picture signals, and executing a multi-levelling processing on them so as to turn them into multi-levelling pictures. CONSTITUTION:Picture information are converted into the digital picture signals, and the converted digital picture signals are converted into the dithering picture signals by using the distributed dither matrix Dn in which respective elements are sequentially and equally distributed in plural divided areas and which is shown by an expression I. The converted dithering picture signals are multi- levelling processed, and the multi-levelling signal are obtained. The size of the matrix in the distributed dither matrix Dn is set to 2nX2n. The size of the matrix in Tn is set to nX2n and respective elements take the values in the range of 1-nX2n, whereby all the values of respective elements differ. An arrangement is shown in an optional expression II. The size of a matrix Un is nX2n and all the values of respective elemented are '1', whereby it is shown by an expression III. Thus, the faithful characteristic of reproducing gradation can be obtained.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to an image processing device, and particularly to an image processing device that has a simple [11-way configuration and has tone reproduction characteristics faithful to an input image]. The present invention relates to an image processing device α that can output an image.

(Prior Art) Conventionally, in image processing apparatuses, images having gradation, such as photographic images, have been represented using pseudo gradation expression through dither processing. At this time, systematic decoding is particularly used because of the simplicity of the circuit configuration in the dither processing section.

By the way, there are many requests for image processing of a diIf-converted image in units of di(a) matrices.For example, when using a dither separately excited FjI signal as a prediction function for an original image signal, or when calculating an average density. .

When performing such processing, it is necessary to perform the density prediction process of the original image from the deep image using the number of line buffers proportional to the size of the diaphragm matrix used in the deep image processing.

(Problems to be Solved by the Invention) Conventionally, dither processing was performed using a large dither matrix in order to produce an output image with tone characteristics more faithful to the input image. If you increase the size of the enhancer matrix with emphasis on
There is a problem in that it is difficult to downsize and lower the cost of the entire A model.

The present invention has been made in view of the above, and an object of the present invention is to provide an image processing device capable of outputting an image with highly faithful gradation reproduction characteristics using a simple circuit configuration. It is in.

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above problems, the image processing device of the present invention includes the following:
An input processing section converts image information into a digital image signal, and the digital image signal converted by this human processing section is digitally processed using a distributed dither matrix in which each element is sequentially and evenly distributed in a plurality of divided areas. a dither processing unit that converts the dither image signal into a dither image signal; and a multi-value processing unit that performs multi-value processing on the di-1 image signal converted by the dither processing unit to convert the multi-value image signal into a multi-value image signal. Write the main point as 1.

(Function) The image processing apparatus I (・) of the present invention converts digital image 9 (i) into digital image @
This digitized image number is converted into a signal and subjected to multi-value processing.

(Example) Hereinafter, the present invention will be explained in detail using the drawings.

FIG. 1 is a block diagram showing a circuit configuration of an image processing apparatus according to an embodiment of the present invention. The υ image processing device shown in the figure inputs an image of a document (not shown) into one image by an input processing unit 21.
The digital image signal 11 is sampled at a line density of 6 trees/IRIIl and converted into an 8-bit digital image signal 11, and this digital image @false code 11 is supplied to a dither processor 1jll group 22. The dip processing unit 22 uses a distributed dither matrix in which each element is sequentially and evenly distributed in a plurality of divided areas, as will be described later, to generate the digital image 8! The signal 11 is dithered to form a 1-pit dithered image signal 12. .

This 1-bit dithered image signal 12 is lit! i
The image is stored in the image storage unit 24 to be saved as fi information, and is also supplied to the image processing unit 23 where it is subjected to image processing such as image editing such as enlargement/reduction and rotation.

Image processing unit 23 F-image processed deep image signal 1
3 is supplied to the density prediction processing section 25 that constitutes the multi-value processing section. This vitality prediction process III! The unit 25 predicts the image density from the input dithered image signal 13,
A multi-value output image signal 14 is formed. This multi-value output image signal 14 is supplied to an output processing section 26, such as a 17J1! type thermal transfer printer having a resolution of 16 lines/mm, and this The output processing unit 26 reproduces an image that is faithful to the original image and outputs it.
'In other words, in the image processing apparatus of this embodiment, the 8-bit digital image formed by the input processing section 21 ÷ 3
11 is dithered in the dip processing unit 22 using a distributed dilintrix and converted into a 1-bit dithered image signal 12, thereby making various image processing faster and easier. be. 1 That is, this digital image signal 12 is stored in an image storage section 24 consisting of, for example, a 20 MB magnetic disk device, and then subjected to various image processing, for example, image data composition, under the control of the image processing section 2J3. , image processing such as affine transformation, and image processing such as creation/editing of text images according to the requests of users. The dithered image signals 1; 3 are subjected to the above image processing and then sent to the density prediction processor 1!1j (2
5.

Here, the distributed dither matrix used in the dither processing section 22 will be explained.

FIG. 3(a) shows a normal lumped dither matrix, and FIG. 3(b) shows a distributed 1% lIT dither matrix used in the present invention. Both dither matrices and b are shown as dither matrices of size 4×4. As can be seen from the comparison of both gazer matrices, Figure 4 (
In accordance with a), in the centralized diary matrix, C is
Each element consisting of 1 to 16 shown in a 4×4 matrix is sequentially numbered from 1 to 2.3. ......, each successive element is arranged next to each other, that is, attached to each other, and spreads out from approximately the center of the matrix in a spiral pattern, whereas in Fig. 4(b) In the distributed dither matrix, each element consisting of 1 to 16 is arranged evenly distributed one after another without being adjacent to each other.

More specifically, as shown in FIG. 4(b), in the distributed dither matrix, a 4×4 matrix is divided into two, an upper half divided area A and a lower half divided area [3].
Each element is sequentially and evenly distributed to both divided areas A and B starting from 1. That is, when element 1 is distributed to divided area △, the next element 2 is distributed to divided area B, and the next element 3 is distributed to rear A, and in the same way, the next element 3 is distributed to divided area A, B. C is distributed alternately and sequentially.

As a result of each element 1 to 16 being uniformly distributed sequentially to the divided areas, density prediction can be performed with little density error and without moiré even if density prediction is performed in units of divided areas. ing. That is, when performing density prediction from the dithered image signal 12 that has been subjected to deep processing in the dip processing section 22, it has conventionally been possible to predict the density in an area of the same size as the dither matrix used in the dither processing section 22. Although it was necessary, by using a distributed dither matrix, it is possible to predict a dense sentence in the personality area of the divided area used in this distributed dither matrix.

Specifically, for d2, the 41st ffl(a), (
b) As you can see, if only element 1.2 is a black dot with a diagonal line as density 2, and the density is predicted based on the size of the original 4x4 dili matrix, the result is shown in Figure 4. For the case of the lumped dip matrix in (a), the fourth
Naturally, the density is 2 in the case of the distributed dither matrix shown in Figure (b). However, each diagonal matrix is now divided into two, the upper divided area and the lower 2r divided area B, and 11 density calculations are performed for each divided area. That is, as shown in FIG. 5, the density calculation is performed in the reference pixel range shown surrounded by a broken line with respect to the J marker pixel indicated by "△".

The results of this vague calculation are shown in Figures 4(a) and 4(b), respectively, and the density calculation results in Figure 4<a) for the lumped dither matrix in Figure 3(a) are as follows: Moiré occurs in the sub-scanning direction at a period of width 172 of the D+1 matrix, such that the upper half area S, 11 area A has a density of 4, and the lower half area 81 area B has a density of O. On the other hand, in the concentration calculation result of FIG. 4(b) for the distributed dither matrix of FIG. 3(b), the density is 2 for both the upper half divided area and the lower half divided area B, and the density It can be seen that a uniform image can be obtained.

This is because, with a concentrated dither matrix, accurate density information can only be obtained if the reference pixel range is the same size as the dither matrix, whereas with a distributed dither matrix, the dither matrix is divided in the sub-scanning direction. ]/2, and there is always a black point in each straw division area, so even if the reference pixel range is selected to be 1/2 the width in the sub-scanning direction of the dither matrix, there will be a degree error from the original image. is the minimum. 1, that is, when using a distributed dither matrix, the error is O when the density jα level is an even number, and the error is ±1/1 for an odd number and 1.
There is a possibility that it will be 6. In contrast, the centralized dither “7
The concentration error of 1-lix varies depending on the concentration level,
In this example, concentration? (The difference is that IA is 5/16.

In addition, in a distributed dither matrix, the larger the size of the dither matrix, the smaller the error becomes (for example, ±1/n2 in an nxn dither matrix), but in a lumped dither matrix, the error becomes smaller. It doesn't change.

Here, if we write the distributed dither matrix [)n-, which can be applied in the present invention, and the size of the matrix is 2nx21), it becomes as follows. Here, Tn has a matrix size of 0×2
0, each element takes a value in the range of 1 to nx2n, the value of each element is different by 1, and the arrangement is arbitrary.

Further, LJn is a matrix of l1fi with a matrix size of nx2n and each element having a value of 1.

The dither processing section 22 converts the digital image signal 11 into a dithered image signal 12 using the above-described distributed dither matrix, and supplies this dithered image signal 12 to the image processing section 23 and the image storage section 24. After image processing, density prediction processing 8tA is performed as a deep image signal 13.
Input to S25.

As shown in detail in FIG. 2, the density prediction processing section 25 converts the dithered image signal 13 into a line buffer for two lines. This memory 251 is stored in the memory 251 consisting of
In order for the dithered image signal stored in the dithered image signal to predict the density of the original input image, in order to reproduce the original input image by performing multi-value processing on the dithered image signal,
First, the dithered image signal from the memory 251 is supplied to the sample pixel selection section 252, and sample pixels are sequentially selected from the dithered image signal. Next, a reference pixel is determined in the reference pixel selection section 253 based on the selected sample pixel. Then, a density calculation section 254 determines the density of the sample pixel from the density of the reference pixel selected by the reference pixel selection section 253. By repeating the above processing, density prediction is performed for all pixels.

More specifically, the sample pixel selection unit 252 sequentially determines the sample pixel (Of(i, j)) every time the address of the memory 251 where the dithered image is temporarily stored is added. Next, the reference pixel selection board 253'c determines a reference pixel range centered on the determined sample pixel (Of(i,j)). The reference pixel range obtained in this way is shown by the broken line in FIG. Real flI! In example i, the reference pixel range is a 2×4 pixel area.

Furthermore, in the density calculation unit 254, each of the determined reference pixels)
The density of the sample pixel (Dq(
i, j)). 'gA degree (Dq
(i, j)) can be expressed by a linear equation.

Dq (i, j) = (ΣDo (i, j)/M)
X^2...<4) Here, M represents the number of picture primes (16 in this example) in the diddy matrix, and 8 represents the result of the reference pixel.

When determining this density, by accumulating image data in a two-line line buffer memory, it becomes possible to extract an average of 11 degrees from the 16 pieces of data.

Therefore, conventional dither processing using a concentrated dither matrix requires a 4-line line buffer memory, but the average density of a single dither matrix can be determined using 2 line buffer memory.

The multivalued output image signal 14 determined as above is
It is supplied from the temperature extraction section 254 of the density prediction processing section 25 to the back-shaped thermal transfer printer constituting the output processing section 26, and an image output with faithful gradation reproduction F1 to the input image is obtained. .

In addition, in the above-mentioned actual example, the case where the modified Dili matrix shown in FIG. If it is within the range, other thresholds (direct head and C
It's good.

Furthermore, in the embodiment described above, the image j of image @accumulation 81S 24
゛-Although the example of image editing was shown in the evening, the image processing unit 23 may edit the input image of the number of shots, or the image processing head can be used as a copy output of the input image without image editing. May be used. Also, it can be applied to facsimile etc.
By performing density prediction processing on the received two-candy image data, it is also possible to reproduce an image that is faithful to the original image.

[Effects of the Invention] As explained above, according to the present invention,? Using a distributed dither matrix in which each element is sequentially and evenly distributed in 82 divided areas, a digital image signal is converted into a dithered image signal, and multi-value processing is performed on this dithered image signal. As a result, the number of temporary storage elements used in multilevel processing can be reduced to the number corresponding to the divided areas of the distributed dither matrix, making the device smaller and cheaper. Can be done.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an image processing apparatus according to an embodiment of the present invention, FIG. 2 is a block diagram showing the detailed configuration of a density prediction processing section used in the image processing apparatus of FIG. figure(
Figures 4(a) and 4(b) are diagrams showing the concentrated dither matrix and the distributed dither matrix used in this example, respectively, and Figures 4(a) and 4(b) are respectively the same as Figure 3(a).
), (El>), Figure 5 is a diagram showing an example of the sample pixel and reference pixel range 1111 in density prediction processing, Figure 6 is a diagram showing an example of the sample pixel and reference pixel range 1111 in the density prediction process.
The figure is a diagram illustrating a method for determining reference pixels in density prediction processing. 11... Digital image signal 22... Dither processed product 23... Image processing section 24.
...Both image storage section 25...Density prediction processing section 26...
・Output processing unit 252...sample pixel selection unit 253・
...Reference picture progressive board part 254... Vitality performance part p::,', Ha411: Miyoshi 11- Nono No. 3
Figure (a) Distributed 2-party call 7-trix (4X4) 3rd copy (b
) Centralized key number 7 tricks λ (4X4) Figure 4 (a) Distributed key number matrix (4 squares 4) Figure 4 (b) ○ 00000

Claims (4)

[Claims] (1) An input processing unit that converts image information into a digital image signal, and a distributed dither matrix in which the digital image signals converted by this input processing unit are sequentially and evenly distributed into multiple divided areas. It has a dither processing section that converts into a dithered image signal, and a multi-value processing section that performs multi-value processing on the dithered image signal converted by the dither processing section to obtain a multi-value image signal. An image processing device characterized by: (2) The distributed dither matrix has a size of 2n
×2n is a matrix that has ▲mathematical formulas, chemical formulas, tables, etc.▼, where Tn is a matrix whose size is n×2n and each element is 1−n×2n
The matrix shown below has a range of values, the values of each element are all different, and the arrangement is arbitrary. ▲ Contains mathematical formulas, chemical formulas, tables, etc. ▼ Un is a matrix whose size is n x 2n, and the size of each element is 2. The color image processing device according to claim 1, wherein all values are 1, and the color image processing device is a matrix of values such that ▲There are mathematical formulas, chemical formulas, tables, etc.▼. (3) The multivalue processing unit includes a sample pixel selection unit that selects a sample pixel from the dithered image signal, and an original pixel group of a predetermined size that includes the sample pixel selected by the sample pixel selection unit. The image processing according to claim 1, further comprising a reference pixel selection unit that selects a reference pixel as a reference pixel, and a density calculation unit that calculates an average density of the reference pixels selected by the reference pixel selection unit. Device. (4) The reference pixel selection unit has means for selecting the divided areas in the distributed dither matrix as a reference range and the original pixel group of the reference range as the reference pixels. The image processing device described.