JPS595902B2 - Error checking method when reproducing character data in phototypesetting machine - Google Patents

Description

Detailed Description of the Invention The present invention provides a phototypesetting machine that reproduces and prints a character pattern based on character data for one character read from a memory. It relates to an error checking method applied to determine whether

The applicant of this application is Japanese Patent Application No. 49-143161 (Japanese Unexamined Patent Publication No. 5
1-69329), Japanese Patent Application No. 50-14854 (Japanese Unexamined Patent Publication No. 51-90529), Japanese Patent Application No. 50-47585 (
In Japanese Patent Application Laid-Open No. 123635 (1973), characters are separated into a large number of pixels, several of these pixels are grouped together to form segments, and the combination of "white" and "black" of each pixel in each segment is In addition, when segments of the same pattern are repeated many times, or in order to synthesize a predetermined segment pattern, segment patterns are We have proposed a system in which a binary code is determined when adding a segment pattern, or when reversing the "white" or "black" of a segment pattern, and this binary code is stored in memory. In this way, when the number of pixels is very large, such as characters used for general printing, the effect of repeating the segment pattern is large, and the character data can be stored more easily than when allocating 1 bit to 1 pixel. You can expect a significant reduction in memory capacity.
However, on the other hand, when reproducing the binary code into the original segment pattern, errors may occur in the binary code itself during the reproduction process, the reproduced segment pattern may not be reproduced as the original correct pattern, or there may be omissions, etc. If this occurs, the reproduced character pattern will be significantly damaged because repetition, addition, etc. are performed as described above. Among such erroneous reproductions, errors in the binary code itself that occur during the reproduction process can be checked using conventional methods such as parity checking, but errors in the reproduced segment pattern can be checked. , cannot be checked using this method. In view of the above points, this invention has been developed to
This was done to make it possible to check for errors in both the decimal code itself and the reproduced segment pattern itself, and to provide a method that does not increase storage capacity due to this checking. be.

In the error check in this invention, for each segment row in the original image, pixels at a certain position in each segment are
By writing the segment position where the segment changes from "white" to "black" or from "black" to "white" in advance in memory, and determining whether it matches the result of actual playback. It will be done. Next, an embodiment of the present invention will be described with reference to the drawings. The square section shown in Figure 1 indicates the area where an arbitrary character ("tiger" in this example) is displayed, and the inside of the ring of this character is actually black;
The outside shall be white. This section is divided into a large number of pixels aligned in the X and Y directions, and a certain number of pixels consecutive in the Y direction constitutes one segment. 1st
The solid arrows shown in the figure correspond to each segment, and the first segment 1-1 to the m-th segment 1-m connected in the X direction constitute n segment rows. Information indicating "white" or "black" of each pixel is binary encoded for each segment to which each pixel belongs as a pattern signal indicating the type of pattern of "white" or "black" for this segment, The pattern signals of each segment are sequentially arranged in series from the first segment 1-1 of the first segment column to the m-th segment 1-m of the n-th segment column and stored in the memory. FIG. 2 shows the structure of a frame corresponding to one character.

In this frame, one unit is a sequence of n segments made up of a series of segments, and a check signal is assigned to each unit. Now, if we look at the pixels at a certain position from the top in order in the X direction for each segment that makes up each segment column, we will see the points where they change from "white" to "black" or from "black" to "white". be. For example, if the 6th segment column is this pixel column, 4 of A, b, c, and d
This change occurs in two positions. In this invention, in the sixth segment string, the a-th, b-th, c-th, and d-th segments are defined as check segments, and their coordinates are stored in the memory as a check signal before the pattern signal. let FIG. 3 shows a mechanism for reproducing a character pattern based on this character signal.

The memory 1 stores a predetermined number of character data having the structure shown in FIG. 2, and the character data stored at an address designated by a print command is read out. This character data is temporarily stored in the memory 2 for one character, and the check signal is sent to the check signal register 3.
Furthermore, the pattern signals are respectively supplied to the reproduction circuit 4. The reproduction circuit 4 reproduces a black and white pattern for one segment based on the pattern signal sent from the memory 2 for one character, and this reproduction pattern signal is sequentially stored in the reproduction pattern memory 5. Furthermore, each time the processing of each segment is completed, the reproduction circuit 4 generates one pulse, and this pulse is counted by the segment counter 6. That is, the count number of the segment counter 6 represents the number of the reproduced segment (ie, the coordinates on the display screen). When the reproduction of one character is completed, the reproduction pattern memory 5 is transferred to the pattern generating means, for example, CR.
Supply a pattern signal to T to display a character pattern,
This character pattern is printed on the photosensitive material through a lens system. This operation is repeated for each character, and phototypesetting is performed here. On the other hand, the check signal register 3 stores a check signal applied to one segment string each time the signal of one segment string is read out from the memory 2 for one character.

As already mentioned, the check signal is, for example, an 8-bit binary code indicating the coordinates of the a-th, b-th, c-th, and d-th segments for the sixth segment string. As a result, signals indicating the coordinates of four segments in the sixth segment string are stored in the check signal register. These four coordinates are then read out one by one by the matching register 7 and stored here. The inversion detector 9 extracts only the pixels at a predetermined position of the segment reproduced by the reproducing circuit 4, and changes the predetermined pixel of the previously reproduced segment and the predetermined pixel of the currently reproduced segment from "white" to "white". This is detected when the color changes to ``black'' or from ``black'' to ``white.''

Now, when the inversion detector 9 detects the inversion of black and white and sends a detection signal to the coincidence detection circuit 8, if the contents of the segment counter 6 and the collation register 7 match, this means that normal reproduction is being performed. This shows that

Further, if they do not match, it indicates that the reproduced segment pattern is different from the original one. In other words, the coincidence detection circuit 8 determines whether the binary code read from the memory 1, passed through the memory 2 for one character, and sent to the reproduction circuit 4 has become a code different from the original code during this process. , or when the original segment pattern is not reproduced when reproduced by the reproduction circuit 4, this is detected and a check signal is generated to inform the outside that an error has occurred. Further, if no check signal is generated, this means that no error has occurred, and the subsequent printing operation will be performed. The number of bits required for error checking using the method of this invention is determined by the number of check points to be set.If the black and white inversion point is selected as the check point,
It depends on the number of black and white inversion points.

If the segment rows are taken in the horizontal direction of the characters, Japanese characters have about 2.5 black and white inversions in one row on average. For each segment column of multiple columns (for example, 32 columns),
Assuming that the coordinates of each segment are expressed in one byte, the capacity of the check signal for one character is only 80 bytes.
Here, 8 bits are used to represent the coordinates of the segments in each segment column, but if the number of pixels for one character is 512 x 512, then the number of segments in each segment column is also 512. , 8 bits is insufficient,
9 bits are required.

However, even if 2 bits are omitted and the coordinates are represented by 8 bits, it is sufficient that black and white inversion occurs between the two segments, and this does not cause much trouble. As you proceed with this kind of thinking, the accuracy will gradually become lower,
It can be said that even if the coordinates are expressed in 4 bits, there will not be much trouble. In this case, it is sufficient that black and white inversion occurs somewhere within the 32 segments. As described above, according to the present invention, by adding a check signal with a small number of bits, it is possible to reliably detect errors associated with pattern signal processing.

Therefore, various inconveniences caused by printing with markedly inferior character quality can be prevented. Note that if an error is detected, the segment string is read from the beginning, and if an error is still detected, control is performed to read the data again from the original memory. If there is still an error, it means that there is an error in the original data, and in this case, processing such as not printing this character is performed.

[Brief explanation of the drawing]

Figure 1 is an explanatory diagram showing the position of check points on a character pattern used for error detection in this invention, Figure 2 is an explanatory diagram showing the arrangement of character signals for one character, and Figure 3 is an explanatory diagram showing the character signal processing. FIG. 1...Memory, 2...1 character memory, 3...Check signal register, 4...
...Reproduction circuit, 5...Reproduction pattern memory,
6... Segment counter, 7... Verification register, Total... Match detection circuit, 9...
・Reversal detector.

Claims (1)

[Claims] 1. Divide the area where a character is formed into a large number of pixels, define a predetermined number of pixels as one segment, and use digital codes that correspond to the "white" and "black" patterns of the pixels in each segment as character data. In a phototypesetting machine that stores predetermined character data in a memory, reads predetermined character data from the memory in response to a printing command, converts this character data into a character pattern signal, and then prints the data, the segments are classified into a plurality of segment rows. , for each segment column
In the process of reproducing the character data as a pattern signal for each segment, the reversal point of white and black pixels is set as a checkpoint, and this reproduction signal is compared with the pixel located on the coordinates of the checkpoint. An error checking method during reproduction of character data in a phototypesetting machine, characterized by detecting errors in character data.