JPH0547155B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image conversion processing device that converts digital image data into digital image data having a desired number of bits.

[Conventional technology]

Conventionally, this type of device includes an image conversion processing device such as a facsimile machine, which converts read digital image data of a predetermined bit into digital image data of a desired number of bits and processes it.

Particularly necessary processing for digital images includes resolution conversion processing to obtain processing of the same size when the recording densities of processed images are different, and scaling processing to obtain processed images of different sizes with the same recording density of processed images. However, it can be said that both processes are based on the same technical idea.

As the resolution conversion process, there is a processing method called a projection method as shown in FIGS. 4 and 5.

FIG. 4 is a schematic diagram illustrating a resolution conversion processing method using a projection method. In this figure, 11 is digital original image data consisting of, for example, 4 bits, and shows a case where it is composed of bit data a ₁ to a ₄ . There is. Reference numeral 12 denotes digitally converted image data, which conceptually shows the case where the digital original image data 11 is converted into 3-bit bit data _b1 to _b3 .

Hereinafter, a case will be described where image conversion is performed using a projection method with a conversion ratio of, for example, 3/4.

For example, when recording an image read with a horizontal resolution of 4 lines/mm and a horizontal resolution of 3 lines/mm, the vertical resolution should be the same and the threshold value should be set to 1/2 (black area ratio 1/2 or more). (black), a conversion table shown in FIG. 5 is obtained. Note that the table value "0" indicates white, and the table value "1" corresponds to black.

FIG. 5 is a schematic diagram illustrating a conversion table based on a projection method, and 21 is a pattern number.

For example, since the digital original image data 11 is composed of 4 bits of bit data _a1 to _a4 , the pattern number 21 is "16".

Therefore, if the bit data a ₁ to _{a 4} of the digital original image data 11 are "1011", the bit data b ₁ to b ₃ of the digital conversion processed image data 12 are converted to "111".

In this way, when the image conversion process using the projection method is executed, the conversion is performed so as to maintain the area ratio of white and black after conversion as much as possible.

[Problem that the invention seeks to solve]

Since the conventional image conversion processing device is configured as described above, as shown in FIG.
If a ₁ to _{a 4} are "1011" or "1101", bit data b ₁ to b ₃ of digital conversion processed image data 12 are converted to "111", and the thin white line cannot be reproduced and becomes black. It is reproduced as an image, and meaningful images such as characters and figures cannot be reproduced. In other words, there was a problem in that the change point information could not be faithfully reproduced.

Note that if the threshold value is set to 1/2 or less, the conversion patterns for pattern number 21 "12" and "14" can reproduce thin white lines such as "101", but the digital original image Bit data a ₁ to _{a 4} of data 11 are "0010", "3" of pattern number 21 of "0100", and bit data b ₁ to b ₃ of digital conversion processed image data 12 of "5" are "111".
, and the opposite problem occurs, where thin black lines are reproduced as white.

This invention was made to solve the above-mentioned problems, and it is possible to convert the digital image data of the original image into adjacent bit data of the immediately preceding digital image data of the original image and adjacent bit data of the digital image data after the immediately preceding conversion process. By converting data into digital image data of a predetermined bit while referring to the data, image conversion processing can be executed while the conversion point information is saved, and digital data that can faithfully reproduce thin white lines and thin black lines can be created. An object of the present invention is to provide an image conversion processing device that can obtain an image.

[Means for solving problems]

The image conversion processing device according to the present invention can process the final bit data of the n-bit digital original image data input immediately before the n-bit digital original image data to be converted, and the last bit data of the n-bit digital original image data input immediately before the n-bit digital original image data to be converted.
m-bit digital processing of the n-bit digital original image data Based on the final bit data of the image data and the n-bit digital original image data to be converted, the n-bit digital original image data is subjected to m-bit digital processing. It is designed to convert into image data.

[Effect]

In the image conversion processing apparatus according to the present invention, the data conversion output means outputs the n-bit digital original image data, the final bit data of the n-bit digital original image data of the immediately preceding block, and the m-bit data of the n-bit digital original image data of the immediately preceding block. Referring to the final bit data of the digitally processed image data of bits, and referring to the number of change points measured by the measuring means and the change point pattern of the digital original image data, the digitally processed image data of each block is sequentially output in units of m bits. do.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings.

In FIG. 1, a line memory 1 temporarily stores input digital original image data. 2 is a memory that stores the digital original image data stored in the line memory 1 in the order of storage in units of 1 block (n
For example, 4-bit digital original image data _a1 to _a4 is output to the arithmetic circuit 4, which serves both as a measuring means and a data conversion output means of the present invention. Reference numeral 3 denotes a flip-flop (holding means) which sequentially stores the final pixel (final bit data) of the digital original image data for the immediately preceding block, and outputs the stored final bit data _a0 of the previous block to the arithmetic circuit 4. The arithmetic circuit 4 stores m bits (3 bits in this embodiment) of digitally processed image data (bit data) b ₁ to b 3 corresponding to the immediately preceding book stored in the flip-flop (holding means) 6, and the last bit of the immediately processed image data b1 to _b3 . Data b ₀ is input, and these 6 bits (n bits + 2 bits)
m-bit digitally processed image data b ₁ to b ₃ corresponding to this block are not shown.
The third memory stored in ROM (read-only memory)
Reading from the conversion table shown in the figure, the digitally processed image data of each block is sequentially outputted to the memory 5 in units of m bits while referring to the measured number of change points and the change point pattern of the n-bit digital original image data. A line memory 7 sequentially stores the digitally processed image data in units of m bits stored in the memory 5.

Note that the capacity of the ROM provided in the arithmetic circuit 4 is as follows, assuming that the conversion pattern shown in FIG. 3 is stored.
From the above number of bits, it becomes 2n ⁺² ×m bits.

FIG. 2 is a schematic diagram illustrating the structure of input/output digital data of the arithmetic circuit 4 shown in FIG. 1.

FIG. 3 is a schematic diagram illustrating a conversion table secured in the arithmetic circuit 4 shown in FIG. 1.

In this figure, _a0 indicates the last bit data of the previous block, and _b0 indicates the last bit data of the immediately preceding process.

Note that if the values of the last bit data a ₀ of the previous block and the last bit data b ₀ of the immediately preceding process are the same, the number of change points in the original image may not be retained (pattern number 21 is "6", "11"). do.

That is, for patterns in which there are two or more consecutive pixels of the same value in digital original image data (bit data) a ₁ to a ₄ , bit data b ₁ to _{b 3}
If the number of pixels is reduced by one pixel and allocated to one location somewhere in the area, the number of change points can be maintained.

Therefore, (a ₀ , b ₀ ) = (0, 0) and (a ₁ , a ₂ ,
a ₃ , a ₄ )=(1, 0, 1, 0), bit data b ₁ to b ₃ are set so that the number of change points is maximized.
is set to (1, 0, 1).

Also, (a ₀ , b ₀ )=(1, 1) and (a ₁ , a ₂ ,
a ₃ , a ₄ ) = (0, 1, 0, 1), the bit data b ₁ to b ₃ are changed to (0,
1, 0).

If the values of the previous block final bit data a ₀ and the immediately preceding process final bit data b ₀ are different, the number of change points in the original image may not be retained (pattern indicated by * in the figure).

In this case as well, the bit data b ₁ to b ₃ are set so that the number of change points is maximized. This results in
Previous block last bit data a ₀ bit data b ₁
can be reflected in

Next, the operation will be explained.

Note that for convenience of explanation, the case where n is "4" and m is "3" will be explained below as an example.

The digital original image data (bit data) _a1 to _a4 read out in blocks (n bits) from the line memory 1 are sequentially read out to the memory 2.
It is output to the arithmetic circuit 4 at the subsequent stage.

For example, if the digital original image data 11 with the pattern number 21 of "3" is input, the pattern number 21 is "0010".
The bit data a ₁ to a ₄ are input to the arithmetic circuit 4. At this time, the arithmetic circuit 4 outputs the last bit data a ₀ of the previous block output from the flip-flop 3.
Then, the most suitable m-bit digitally converted image data 12 is determined by referring to the last processed final bit data _b0 outputted from the flip-flop 6 and by referring to the number of change points of the bit data _a1 to _a4 .
(Bit data b ₁ to b ₃ ) are read from the ROM having the conversion table shown in FIG. 3 and output to the memory 5.

Therefore, from the relationship between the previous block final bit data a ₀ and the immediately preceding processing final bit data b ₀ output from the flip-flop 6, the memory 5 stores the following data.
Either "010", "010", "101", or "001" is output.

Therefore, the bit data a ₁ to _{a 4} of the digital original image data 11 are "1011" and "1101", that is, "12" of the pattern number 21 shown in FIG.
In this example, the conversion pattern for "14" is such that bit data b ₁ to _{b 3} are "101", "010", "101",
It is converted to "010" and can be converted to digital image data that can reproduce thin white lines.

In addition, in the conversion table shown in FIG. 3, (a ₀ , a ₁ , a ₂ , a ₃ , a ₄ , b ₀ ) is (1, 0, 0,
1, 1, 1), the bit data b ₁ to b ₃ are set to (011) based on the criterion that the changing point position is selected to be the closest position to the original image, but they may also be set to (001).

Similarly, (a ₀ , a ₁ , a ₂ , a ₃ , a ₄ , b ₀ ) is (0,
1, 1, 0, 0, 0), the bit data b ₁ to b ₃ are set to (110) based on the standard of selecting the change point position to the position closest to the original image, but (100)
You can also use it as

Furthermore, in this embodiment, emphasis is placed on black pixels, and the third
The conversion table shown in the figure is stored in the ROM of the arithmetic circuit 4, but if there are two closest positions, a bit data pattern may be used that takes the position that changes quickly.

In addition, two-dimensional processing, for example from n×n to m×
When conversion to m is required, line memory 7
It is sufficient to prepare the output for (n+1) lines, read out (n+1) pixels in the vertical direction, and execute the above processing. Note that (n+1) includes a line memory for inputting the last pixel of the immediately preceding block. It goes without saying that a line memory is required to store the final converted pixels of the vertical conversion process.

〔Effect of the invention〕

As described above, according to the present invention, the final bit data of the n-bit digital original image data input immediately before the n-bit digital original image data to be converted, and the n-bit digital original image data input immediately before Based on the final bit data of the m-bit digitally processed image data for the digital original image data and the n-bit digital original image data to be converted, the n-bit digital original image data is converted into m-bits digitally processed image data. Since it is configured to convert into , fine white lines and fine black lines can be faithfully reproduced, and high-quality resolution conversion processing and reduction conversion processing can be performed.

[Brief explanation of the drawing]

FIG. 1 is a block diagram explaining the configuration of an image conversion processing device showing an embodiment of the present invention, FIG. 2 is a schematic diagram explaining the configuration of input/output digital data of the arithmetic circuit shown in FIG. Figure 3 is a schematic diagram explaining the conversion table secured in the arithmetic circuit shown in Figure 1, Figure 4 is a schematic diagram explaining the resolution conversion processing method using the projection method, and Figure 5 is a schematic diagram explaining the conversion table using the projection method. It is a schematic diagram for explaining. In the figure, 1 and 7 are line memories, 2 and 5 are memories, 3 and 6 are flip-flops (holding means),
4 is an arithmetic circuit (data conversion output means). In addition, in the figures, the same reference numerals indicate the same or equivalent parts.

Claims (1)

[Claims] 1 Input n-bit digital original image data, and convert the n-bit digital original image data into
In an image conversion processing device that outputs digitally processed image data that has been converted into (m<n) bits and subjected to resolution conversion processing or reduction conversion processing, the n-bit digital original image data input immediately before the n-bit digital original image data to be converted is used. holding means for holding final bit data of n-bit digital original image data and holding final bit data of m-bit digitally processed image data for the n-bit digital original image data input immediately before;
Data for converting the n-bit digital original image data into m-bit digitally processed image data and outputting the data based on each final bit data held by the holding means and the n-bit digital original image data to be converted. An image conversion processing device comprising: conversion output means.