JPH03167598A - Method for reducing image - Google Patents

Abstract

PURPOSE:To read a character even when it is reduced by composing plural pieces of thinned data whose thinning position is changed and forming a reduced image. CONSTITUTION:The character is read out from a memory 1 where an image which is displayed is stored based on the instruction of an address generator 2. An address for thinning is generated by the generator 2 and the plural pieces of reduced data whose thinning position is changed is produced from the image. That means, the address is successively generated and the read out odd numbered data is stored in a memory 3. Besides, the even numbered data is stored in a memory 4 and the OR of the data read out every corresponding appearance position is taken by a gate 6 after it is written in the memories 3 and 4. Then, the result is written in a memory 7 according to an address instructed by an address generator 8 and displayed on a display part 9. Thus, the reduced character which can be read is formed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image reduction method for reducing images such as characters. [Prior art] Conventionally, a thinning method was used to create reduced characters. The thinning method means that when a character is composed of m x n pixels vertically and horizontally, for example, the vertical line is thinned out line by line, and the remaining pixel data for each line is transferred to the lth and second lines. , the third line, and so on, by narrowing the line spacing (that is, filling in the thinned out lines) to obtain reduced characters. This allows you to create 1/2 reduced characters. In addition, various reduction methods such as 1/3 reduction can be employed.

[Problem to be solved by the invention] When the above thinning method is adopted, some characters become illegible after reduction. Examples are shown in Figures 2, 3, and 4.
It is shown in the figure. Figure 2 is an example of the Roman letter A, where length x width =
This is an example of a size of 8 x 9. For this character A, two lines,
The right side shows the result when lines 4, 6, and 8 are thinned out, and the remaining lines are filled in by one for each thinned line. The reduced character on the right cannot be read as the letter A. Figure 3 is an example of the Roman letter E. If the same method were used to thin out the letters, the reduced letters on the right would not be legible as the letter E. Figure 4 is an example of the Roman letter B. Similarly, the reduced letters on the right cannot be read as the letter B. An object of the present invention is to provide an image reduction method that makes images legible even when reduced. [Means for Solving the Problems] In the present invention, a plurality of reduced data are created by changing the thinning position from an image, and a reduced image is created by combining the plurality of reduced data. Furthermore, the present invention obtains two reduced data strings by changing the thinning position for each vertical or horizontal pixel string of a character image, and thins out each pixel string for each appearance position with respect to the two reduced data strings. We decided to take the logical sum of the two data and use this logical sum output data as reduced character image data. [Operation] According to the present invention, for images such as characters, even if each of the plurality of thinned-out data does not have characters, they can be combined (including logical sum processing) to complement each other and reduce the size. Images can be formed. [Embodiment] FIG. 5 shows the processing procedure of this embodiment when obtaining a reduced character of the character B. Image I. is the original character B, and its actual size is length x width = 8 x 6.

This character image craft. For, (a) First, the even numbered rows (2nd row, 4th row, 6th row, 8th row)
Thin out. Odd-numbered rows remaining after this thinning (l rows,
3 lines, 5 lines, 7 lines) and create l line as a new line,
Create 2nd, 3rd, and 4th rows.

Let the even-numbered thinned image obtained by condensing these rows be . (mouth) Next, the original image I. odd-numbered rows (1 row,
Lines 3, 5, and 7) are thinned out. After this thinning, even numbered rows (2nd, 4th, 6th, 8th) remain, and by closing these even numbered rows, new rows 2nd, 3rd, 4th, etc.
Create 5 lines. The odd-numbered thinned image obtained by closing this row is I,
Suppose that (c) As is clear from Fig. 5, engineering and I.
are not reproductions of the letter M, B, but completely illegible patterns. Therefore, thinning image craftsman and I! For both, perform a logical OR between pixels at the same appearance position. For example, the presence of pixels is defined as "1" and the absence of pixels is defined as "0". However, if the pixel value of the pixel is displayed in gradation, the pixel value will be according to the gradation level, and will not be a binary value of "1" and "0". The logical sum is not the pixel value of the gradation display, but
Is there a pixel value (“1”) or not (“O”)?
This is data showing. Therefore, if the pixel values are displayed in gradation, it is necessary to perform a process of dividing them into binary values of "1" and "O" based on some standard. This process is not necessary if the display is not a gradation display but a monochrome binary display. Images IU and I with “1” and “O”! If we display , we get Table 1. Table 1 Regarding I1 and engineering in this Table 1, the corresponding appearance position (x
, y). However, X is the horizontal position (indicates one of 1 to 6), and y is the vertical position of 1 m! (1
to 6).

For example, rank! (xt y) = (3.1), place 1! (x+
y) = Do 2) At the position (xt y) (4, 4). If we perform a logical sum on the other values in the same way, we get Table 2. When the output of this logical sum is displayed, the image processing shown in FIG. 5 is obtained. . bawl. Image processing in Figure 5. . As can be seen, the original character B has been almost reproduced as a reduced image.

Table 2 As described above, according to this embodiment, since the logical OR is performed between adjacent rows, there is no removal of pixels between adjacent rows, and there is no difference after reduction. However, the original characters can be read. Note that although the objects of the logical sum are adjacent lines, it is also possible to carry out the logical sum processing of three consecutive lines. However, the reduction will be somewhat rough. It can also be applied to general images instead of characters, and in that case, synthesis using other logic such as logical product is also possible.

There may also be cases where the number of thinnings is 2 or more. In this case, it is preferable that three or more images be combined.

Figure 1 shows an embodiment of the character reduction device of the present invention. Memory 1 is a memory that stores images to be displayed from now on, and characters are read out according to instructions from address generator 2. The address generator 2 generates addresses for thinning. However, the thinning is not performed directly, but addresses are generated continuously, and the read odd-numbered data is stored in the memory 3, and the read even-numbered data is stored in the memory 4.

As a result, memory 3 contains odd-numbered data obtained by thinning out even-numbered data, and memory 4 contains the reverse. Address generation at this time is performed by the address generator 5. After writing to the memories 3 and 4, the logical sum of the data read for each corresponding appearance position is taken by the gate 6, and the result is written to the memory 7 according to the address specified by the address generator 8. The contents of this memory 7 are reduced characters, and the display section (CR
T) Displayed in 9. Note that the memories 1 and 7 may be the same memory. Address generators 2 and 5.8 generate addresses according to instructions from the CPU. This address can be determined by the CPU in various ways depending on the character position and display content. Furthermore, although reduction was performed for each character, if the area in which the character is to be displayed is clear, it is actually possible to reduce the character even if the reduction process is performed on the entire area.

[Effects of the Invention] According to the present invention, even if each of the plurality of thinned-out data cannot form a character, it is possible to form legible reduced characters by combining them to complement each other.

[Brief explanation of the drawing]

FIG. 1 is an embodiment of the present invention, FIGS. 2, 3, and 4 are examples of conventional thinning, and FIG. 5 is a diagram showing reduction processing according to the present invention. 1, 3, 4.7...memory, 2,5.8...address generator.

Claims (1)

[Claims] 1. An image reduction method in which a plurality of reduced data are created by changing thinning positions from an image, and the plurality of reduced data are combined to form a reduced image. 2. For either the vertical or horizontal pixel rows of the character image, change the thinning position for each pixel string to obtain two reduced data strings, and create two data for each appearance position for the two reduced data strings. An image reduction method that takes the logical sum of and uses this logical sum output data as reduced character image data. 3. An image reduction method according to claim 2, in which the number of thinning positions is two or more.