JP3104228B2 - Linear image density conversion device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a binary image linear density conversion device for converting the linear density of a binary image.

[Conventional technology]

In an OA device such as an electronic filing device or a facsimile device, an image is often handled as binary information. When editing such binarized image information, or when performing transmission and reception while converting the density of image information between facsimile machines having different line densities, the original image information may be enlarged or enlarged as necessary. Need to shrink.

Conventionally, conversion methods such as an SPC (Selective Processing Conversion) method, an OR method, a nine-division method, and a high-speed projection method have been used as methods for enlarging or reducing binarized image information. For example, the SPC method is a method in which an original pixel located closest to a conversion pixel is selected, and its density value is used as a conversion pixel value.

FIG. 2 is for explaining the SPC method. Second
In the figure, when P, Q, R, and S are original pixels and D is a converted pixel, the density value of D is the density value of the original pixel located at the closest position. That is, when the conversion pixel D is in the area A1, the density value of the original pixel P becomes the density of the conversion pixel D. Similarly, when in the area A2, the density value of Q becomes the density value of the conversion pixel D, and when it is in the area A3, the density value of S becomes the density value of the conversion pixel D.

[Problems to be solved by the invention]

In the conventional methods for enlarging or reducing the binarized image information including other methods as described above, any of the conventional methods uses a logical operation or the like of the density of one pixel closest to the conversion pixel or the four nearest original pixels. The pixel density is determined. For this reason, when the original image is a halftone image, moire occurs when interference occurs between the period of the halftone dot and the period of the calculation for converting the density.

Further, since there is no function of correcting the density difference between the original image and the converted image, there is a disadvantage that the density of the converted image with respect to the original image changes.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a binary image linear density conversion device capable of reducing occurrence of moire and obtaining a converted image having a small density change with respect to an original image.

[Means for solving the problem]

According to the first aspect of the present invention, there is provided (a) an original image extracting means for extracting an original image block of an area corresponding to a desired conversion magnification from original image information composed of binarized image information; (C) correction means for adding a predetermined density correction value to the density sum value obtained by the sum calculation means and outputting the corrected density value as a corrected density value; (D) conversion pixel density determination means for binarizing the correction density value output by the correction means and outputting a conversion pixel density for the original image block;
(E) a density error calculating means for calculating a density error between a value obtained by weighting the converted pixel density according to the conversion magnification and a corrected density value output by the correcting means; and (f) a density error calculating means. The linear density converter for a binary image is provided with feedback means for feeding back the obtained density error to the correction means as the density correction value.

That is, the first aspect of the present invention is to determine the density of a converted image by feeding back a density error between an original image and a converted image weighted so as to be equal to the original image as a density correction value.

EXAMPLES Hereinafter, the present invention will be described in detail with reference to examples.

FIG. 1 schematically shows the configuration of a linear image linear density conversion apparatus according to an embodiment of the present invention.

The binary image linear density converter includes an original image cutout unit 11. The original image extracting section 11 extracts an original image block 13 composed of a set of pixels corresponding to a conversion magnification specified by a magnification specifying section (not shown) from an original image memory 12 in which binarized image information is stored, and calculates a sum. Supply to the unit 14. Conversion magnification is vertical p
When the magnification is doubled and the width is set to q times, the original image cutout unit 11 cuts out the pixels of the original image composed of 1 / p in height and 1 / q in width as the original image block 13.

The sum total calculation unit 14 calculates the density sum value 16 of the original image block 13 by setting the density to “1” when the pixel of the original image is a black pixel indicating black and setting the density to “0” when the pixel is a white pixel. Correction unit 17
Is configured to correct the total density value 16 by adding a density correction value 19 input from the error buffer 18 and to supply the corrected density value 21 to the converted pixel density determination unit 22.

The conversion pixel density determination unit 22 includes a comparator 23 and a threshold value calculation unit 24. The comparator 23 binarizes the corrected density value 21 with the threshold value 26 input from the threshold value calculation unit 24, and outputs it as a converted pixel density value 27. The converted pixel density value 27 is stored in the converted original image block 13 in the converted pixel memory 28.
Is stored in a predetermined area corresponding to. Threshold calculator 24
Is configured to calculate a threshold value according to a magnification specified by a magnification specification unit (not shown) or to output a threshold value set by an operator.

The conversion density error calculation section 30 includes a weight multiplication section 31.
The weight multiplication unit 31 adds (1 / p) · (1 / q) to the converted pixel density value 27.
And the density value of one pixel after conversion is corrected to the same weight as the density value of the original block 13 composed of (1 / p) · (1 / q) pixels. The differentiator 33 obtains a difference between the corrected density value 21 before conversion and the corrected converted pixel density 32, and outputs a density error 34.

The density error 34 is supplied to a density error distribution unit 36 and distributed to the peripheral area of the converted original image block. Density error 34
Is obtained from the original image blocks 13 to be distributed by the distribution peripheral block determination unit 37 and the distribution ratio. The allocation ratio multiplying unit 38 multiplies the original block obtained by the allocation peripheral block determining unit 37 by a weight corresponding to the allocation ratio, and outputs an allocation density 39. In this embodiment, the conversion error density of the original image block of l rows and m columns
34 is added to the original block of (l + 1) rows and m columns by 50%, (l
The original picture blocks of (+1) rows and (m + 1) columns are distributed at a rate of 50%.

Each distribution density 39 output from the density error distribution unit 36 is stored in each area of the error buffer 18 corresponding to the distributed original image block. The total value of the error densities 39 for each original block is supplied to the correction unit 17 as a density correction value 19 for converting the original block.

Next, the operation of the linear image conversion apparatus for binary images configured as described above will be described. Although the linear density conversion rate of the binary image linear density conversion apparatus can take an arbitrary value, a case where the conversion magnification is 1/2 will be described for simplification of the description.

In order to increase the linear density by a factor of two, it is sufficient to convert four pixels of the original image into one pixel of the converted image. Therefore, the operator operates a magnification designating section (not shown) to set the magnification to half vertically and half horizontally and set the conversion magnification.

When the conversion magnification ２ is set by a magnification designation unit (not shown), the original image extracting unit 11 extracts an original image block 13 composed of four pixels from the original image memory 12. The original image block 13 cut out by the original image cutout unit 11 has l rows and m columns. The original image block 13 is supplied to the total sum calculation unit 14, and the total density value 16 is calculated. In the case of the original picture memory 12 schematically shown in FIG. 1, the original picture block 13 of l rows and m columns has one black pixel and three white pixels, so that the density sum value 16 is "1". is there. The total density value 16 is supplied to the correction unit 17.

The correction unit 17 adds the density correction value 19 “0.4” supplied from the error buffer 18 to the total density value 16 “1”, obtains a corrected density error 21 “1.4”, and outputs this. The density correction value 19 "0.4" for the original block 13 of l rows and m columns is (l-1)
This is the total value of the distribution densities 39 obtained by converting the original image blocks of the row, the m-th column, the l-th row, and the (m-1) -th column.

Corrected density value 21 of original image block 13 output by correction unit 17
Is input to the comparator 23 of the conversion pixel density determination unit 22. The comparator 23 binarizes the corrected density value 21 with a threshold value 26 input from the threshold value calculation unit 24. comparator
The original block 13 obtained by the threshold calculator 24 is
Of the number of pixels “4”, that is, “2” is input as the threshold 26. Since the corrected density value 21 is “1.4”, the converted pixel density value 27 for the original image block 13 is obtained as “0”. That is, l rows, m output from the conversion pixel density determination unit 22
The converted pixel after the original image block 13 in the column has been converted by a factor of two is determined as "white". The converted pixel density value 27 “0” is stored in an area of l rows and m columns, which is an area of the converted pixel memory 28 corresponding to the converted original image block 13.

The converted pixel density value 27 is input to the weight multiplication unit 31 of the conversion density error calculation unit 30. Since the weight of the density of one converted pixel is equivalent to four pixels of the original image, the weight multiplying unit 31 multiplies the converted pixel density value 27 by the weight “4” to obtain the corrected converted pixel density 32
Is input to the differentiator 33. A differentiator 33 calculates a difference between the corrected density value 21 before conversion and the corrected converted pixel density 32, and calculates a density error 34.
Is output. Correction conversion pixel density 32 is “0”, correction density value 21
Is "1.4", the density error 34 of the original block 13 of l rows and m columns is "1.4".

The density error 34 obtained by the conversion density error calculation unit 30 is supplied to a density error distribution unit 36, and the converted original image block of l rows and m columns is provided.
The peripheral blocks are allocated to the 13 peripheral areas by the peripheral block determining unit 37. In the present embodiment, the value “0.7” obtained by multiplying the density error 34 “1.4” by a weight of で in the distribution ratio multiplying unit 38 is located in the (l + 1) th row, the mth and lth rows, and the (m + 1) th row. It is added to the error buffer 18 as the distribution density 39 of the original image block. Error buffer
The density correction value 19 in l rows of 18 and (m + 1) columns is (l-1)
Distribution density by conversion of original picture block of row, (m + 1) column
39 This is a value “0.9” obtained by adding “0.2” to the distribution density “0.7” of l rows and m columns. Also, (l + 1) rows of the error buffer 18,
The density correction value 19 in the m-th column is obtained by adding the distribution density 39 in the (l + 1) -th row and the (m-1) -th row to the distribution density 39 "0.7" in the l-th row and the m-th column.

The density correction value 19 obtained in this way is
To the corresponding original picture blocks
Correct the total density value 16 of 13.

In the embodiment described above, the conversion error density of the original block of l rows and m columns is changed to the original block of (l + 1) rows and m columns.
50%, 50 (1 + 1) rows and (m + 1) columns
However, in the present invention, the peripheral blocks to which the conversion error density is distributed and the distribution ratio are not particularly limited. For example, the conversion error density of l rows and m columns is set to (l + 1) rows, the original image block of n columns is 40%, (l +
1) 40% in the original block of the row and (m + 1) column, (l +
1) The original picture blocks in the rows and (m + 1) columns may be distributed at a rate of 30%, and may be distributed to the periphery of these blocks at a predetermined rate.

〔The invention's effect〕

As described above, according to the first aspect of the invention, since the density error is obtained by weighting the converted image so as to be equal to the original image, a converted image closer to the density of the original image can be obtained. That is, the density preservation of the converted image with respect to the original image is further enhanced.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a linear image linear density conversion apparatus according to an embodiment of the present invention, and FIG. 2 is an explanatory diagram for explaining magnification conversion by the SPC method. 11: Original image extraction unit, 14: Total calculation unit, 17: Correction unit, 19: Error buffer, 22: Conversion pixel density determination unit, 30: Conversion density error calculation unit, 36: Density error Calculation unit.

Claims (1)

(57) [Claims] An original image extracting means for extracting an original image block of an area corresponding to a desired conversion magnification from original image information composed of binarized image information, and a density of the original image block extracted by the original image extracting means. Total sum calculating means for obtaining a total sum value; correcting means for adding a predetermined density correction value to the total density value obtained by the sum calculating means to output a corrected density value; and correcting the corrected density value output by the correcting means. Conversion pixel density determining means for outputting a converted pixel density for the original block, and a density error between a value obtained by weighting the converted pixel density according to the conversion magnification and a corrected density value output by the correction means And a feedback unit that feeds back the density error obtained by the density error calculation unit to the correction unit as the density correction value. Line density conversion apparatus of the binary image, characterized in that.