JPH01216865A - Minification of pixel data - Google Patents

Abstract

PURPOSE:To perform minification processing which does not permit required pixels to be lost, using an extremely simple logic operation by extracting a string of pixels connected obliquely with pixels in a single line before or after (n) pieces of line and including the extracted pixels into the minification results. CONSTITUTION:If n equals 3, pixel data is subdivided into each three blocks in a row direction, and a spacer line L3 is left by thinning out lines L1, L2 of (m)th block. This processing enables the obtaining of a logic sum of all pixel data in lines (n)-1 from (mXn)th line to((m+1)Xn-2)th line as spacer lines of (m)th block. A logic product of the logic sum and the logic negative of pixel data in (mXn-1)th line before a single line of (m)th block is obtained in a row direction. These extracted pixels are independent points which are serially connected obliquely with pixels in a preceding line. In order to leave the extracted independent points P intact in the minified results without spacing, if a logic product of data in((m+1)Xn-1)th line, which is not thinned out and data of the independent point is sought, some minification results are obtained. The above procedures are to be taken in the sequential order of blocks m(1, 2, 3...) to obtain the whole minified image.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for reducing pixel data, and is particularly suitable for use in displaying or printing characters and graphics in a word processor or the like.

[Summary of the invention]

When reducing n lines of dot pixel configuration rows or columns to one line, by extracting pixels that are diagonally continuous to the pixels one line before or after the n lines and including them in the reduction result. , complement the diagonal lines that are interrupted by reduction,
This is a pixel data reduction method that reduces deterioration in pattern quality.

[Conventional technology]

When small character patterns, such as so-called "half-width" characters, are needed in word processors, large character patterns are sometimes used after being reduced. At this time, since there are constraints such as the limit of the CPU's arithmetic processing capacity, methods using only relatively simple processing such as logical operations are often used.

The commonly used methods are “thinning” and
There are two methods using "logical sum".

[Problem to be solved by the invention]

The ``thinning'' method is a method of extracting rows of dot pixels that make up a character pattern and arranging them at equal intervals, and is called this because it involves thinning out unnecessary rows. In this method, if there are independent points or lines in the thinned out and discarded columns, they will appear as omissions in the reduced pattern, degrading the quality of the characters.

The "logical sum" method is a method that takes the logical sum of multiple adjacent columns and arranges them in order to prevent missing points and lines.This method can prevent missing points and lines, but instead This also reduces the quality of the text because the white space between the lines is crushed.

In view of this problem, it is an object of the present invention to perform a reduction process using extremely simple logical operations so that necessary pixels are not missing.

[Means to solve the problem]

Reducing a character or figure with a dot pixel configuration means that one pixel (n=2.3...
・-・-) lines are combined into one line.

Note that one or more lines that are not involved in the reduction process may be interposed between one block consisting of n lines and the next block.

In the reduction method of the present invention, data of one line before or after the above n lines is referred to, and pixels that are diagonally continuous to pixels constituting the line before or after the above n lines are extracted within the n lines. . At least these extracted pixels are used in the reduction result without being removed in the reduction process.

Note that the reduction process may basically be a process of thinning out n-1 lines in one line. If the line to be thinned out has a specific pixel to be saved as described above, it is extracted and added as a pixel to the line that is not to be thinned out.

The line one line before or after the n lines may be a line based on pixel data before being subjected to reduction processing, but if processing is performed sequentially in the reduction direction, the line based on pixel data resulting from reduction may be referred to. You can.

[For production]

Pixels that are diagonally continuous with respect to pixels forming one line before or after one line can be called "independent points," and are highly important when visually recognizing dot-structured characters or figures. If this independent point is lost due to thinning, the character,
Since the diagonal lines that make up the figure are cut off, the quality of the image deteriorates. If independent points are preserved, the diagonal lines will be continuous and quality deterioration will be reduced.

〔Example〕

FIG. 1 shows the principle of the pixel data reduction method of the present invention.

(A) shows a part of the given pixel data; in this case, n=3, and 0 is reduced to 1/3 in the column direction (vertical direction).
Divided into blocks of 3 lines (rows) each in the column direction, for the m-th block, one line ( The reduction process is to summarize the data into rows).

In this example, "combining" means "thinning", and in the example of FIG. 1(A), lines Ll and L of the m-th block
2 is thinned out, leaving line L3.

Figure 1 (B) shows the results of extracting independent points and performing interpolation processing.
Shown below. This process is performed as follows.

First, as shown in FIG. 1(C), all of the n-1 lines (lines L1 and L2 in this example) from the mxn line to the (m+1)Xn-2 line (lines L1 and L2 in this example), which are the lines to be thinned out in the m-th block. Take the logical sum of the pixel data in the column direction (LIUL2)
. Next, logically negate this logical sum and the pixel data of the m X n -1 row, which is one line before the m-th block (1 block)
Performs a logical AND with the column direction. The result is the first (D) pixel data (τOn (LIUL2)), and these extracted pixels are "independent points" that are diagonally continuous with respect to the pixel one line before. This extracted independent point P
In order to leave the data in the reduction result without thinning out, we take the logical OR of the data on the (m+1) ) is obtained.

Note that if a simple thinning process is performed without interpolating independent points, the independent points P in FIG. 1(B) will be missing, and the continuity of dots in the diagonal direction will be lost.

The above processing is performed in block m (1, 2, 3-...
-) to obtain the entire reduced image. Note that printing by the serial printer is performed in the increasing direction of m.

Here, if m=O, that is, for one line of the first block, there is no line called the 11th line, so
- Assume a blank line as the first line.

In this case, the result is the same as calculating the logical OR of all up to the first n rows.

Note that almost the same result can be obtained by referring to the reduced line in FIG. 1(B) and extracting the independent points of the next thinned block m+l.

This is more convenient if the image data to be reduced is contained in a sequential file.

In addition, if L2 and L3 are lines to be thinned out in FIG. 1(A), an independent point is extracted with reference to line L4 one line after the m-th block. In this case, a memory is required to store each line of the m-th block and one line thereafter.

Figure 2 is a character pattern diagram of the kanji ``geki'' that has been reduced in various ways to half-width (1/2 ratio); (A)
is the original pattern, (B) is a pattern after thinning processing, (C) is a pattern obtained by interpolation processing of independent points to which the method of the present invention is applied, and (D) is a pattern obtained by logical sum processing. As is clear from this figure, in (B) the dots are significantly missing, and in (D) the dots are crushed and difficult to see. In the process of (C), the defects of (B) and (D) are improved and the quality is improved.

FIG. 3 shows a concrete example of the reduction circuit. This circuit is for one column (
1 bit), and are actually provided in parallel as many as the number of columns. Input pixel data is sequentially taken in to three stages of series latch circuits 1a to 1c at the timing of the pixel clock CK. Continuous pixel data in the column direction enters the latch circuits 1a and 111C in the order of lines LO1L1 and L2, and the pixel data of the current line L3 appears on the input line 2. The latch circuit tb is made independent by ANDing the output of the OR circuit 3, which uses the output of the IC, and the output of the NOT circuit 4, which inverts the output (one line before) of the latch circuit 1a, using the AND circuit 5. Pixel data of a point is extracted.

Interpolation processing is performed by ORing the output of the AND circuit 5 and the current input in the OR circuit 6. By inputting the output of the OR circuit 6 into the latch circuit 7 using a clock frequency divided by the 1/3 frequency divider 8, pixel data Q as a result of reduction is obtained. Note that in order to extract independent points using the previous reduction result, the output of the latch circuit 7 may be supplied to the NOT circuit 4 as shown by the dotted line in FIG.

FIG. 4 is a block diagram of a Japanese word processor to which the present invention can be applied. As is well known, word input from the keyboard 10 is processed by the CPU 1, and is processed by the RAM 12.
is memorized. Further, the document formed in the RAMI2 is stored on the disk 13.

Data in the RAM 12 is converted into dot pixel data by a character generator 14 having a ROM configuration, and displayed on the display 15. The dot pixel data is also sent to the line printer 16 to print the document. The dot pattern obtained from the character generator 14 is reduced and displayed, for example, as half-width characters on the display 15 or printer 16.
It can be displayed (printed) with . The reduction process described above is c
Software processing using pUll and RAM 12 is also possible. As described above, the processing process is executed by extremely simple logical operations, so even if the processing capacity of the CPU 11 is low, it does not place much burden on the CPU 11.

〔Effect of the invention〕

As described above, the present invention extracts diagonally continuous dot pixels that are lost due to reduction and retains them in the reduction result, so that it is possible to interpolate the diagonal lines that make up characters and figures that are interrupted due to reduction. Therefore, deterioration in the quality of the reduced pattern can be reduced, and easy-to-read reduced display or printing can be obtained. Moreover, since the interpolation process can be performed by simple logical operations, the circuit does not become large-scale and the burden of processing steps on the processor does not increase.

[Brief explanation of the drawing]

FIG. 1 is a dot pattern diagram of pixel data showing the reduction process of the present invention, FIG. 2 is a dot pattern diagram when applied to half-width reduction of Chinese characters, and FIG. 3 is a circuit diagram showing an embodiment of the reduction circuit. FIG. 4 is a block diagram of a word processor to which the present invention is applied. In addition, in the symbols used in the drawings, 10.-----1.--Keyboard 1
1・−・−・−・−・・−・CPU 12・−・・−・・−・・−RAM 13・−・−・−・・−・・Disk 14・・・−
・−−−−１−−−−−・−・Character generator 15・−・−・・・・−・・−Display 1fy−１−−・・−・−・・・・−・−7 ” IJ 7 ta. @Figure 3 Figure 4

Claims (1)

[Claims] In a pixel data reduction method for reducing pixel data arranged in row and column directions in the row or column direction, ，
3. When reducing the line (…………) to one line,
Referring to pixel data one line before or after the n lines, the method includes a process of extracting pixels in the n lines that are continuous in a diagonal direction with respect to the pixels one line before or after the n lines; A pixel data reduction method characterized by including pixel data in a reduction result.