JPS5816819B2 - Data compression method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a data compression method for compressing and transmitting data in order to improve the transmission and reception efficiency in the data transmission and reception method.

Transmission efficiency can be improved by compressing data, and various data compression methods have already been proposed for this purpose, one example of which will be described.

As shown in FIG. 1, the transmitting device 11 refers to the transmission conversion table 12 to determine whether or not the input data is compressible data. For example, if one character is 8 bits, the data can be compressed. Characters (hereinafter referred to as compressed characters) are compressed into 4-bit codes (hereinafter referred to as compressed codes) corresponding to the characters, and characters other than compressed characters are compressed into 8-bit codes specified in the transmission conversion table 12. Convert and send.

However, if an odd number of compressed characters are consecutive, the last character is converted into an 8-bit code defined in the transmission conversion table 12 and transmitted.

The receiving device 13 reproduces the received data by referring to the receiving conversion table 14. It checks the received data 8 bits at a time, and if both the upper 4 bits and the lower 4 bits are compressed codes, the It is determined that 8 bits are two compressed characters, and the upper 4 bits and lower 4 bits are converted into corresponding compressed characters.

Also, if both or one of the upper 4 bits and lower 4 bits is not a compressed code, it is determined that those 8 bits are not the result of compressing two compressed characters, and the 8 bits are inversely converted using the reception conversion table 14. and restore.

Figure 2 shows an example of a conventional conversion table, in which the compressed characters are from column 0 to column 6 of row F, A, D, E.
, G, I, L, N, 0. R, S, T, U, and the corresponding compressed code is a 4-bit hexadecimal number P O"~1
1 C11.

The codes of characters other than compressed characters are represented by a total of 8 bits, with the hexadecimal number displayed on the left side of the conversion table being the upper 4 bits, and the hexadecimal number displayed on the upper side being the lower 4 bits.

For example, when character B in row D and column 8 is converted using this conversion table, it becomes an 8-bit "D8 pattern."

FIG. 3 is an explanatory diagram when data is compressed using the conversion table shown in FIG.
Assuming ULEJ, R and E are compressed characters, so they are converted to their corresponding 4-bit compressed codes "9" and °"3", and Q, /, and M are not compressed characters, so their two corresponding 8-bit code CD expressed in hexadecimal
""EB", "'DE" is converted, and 0.D, U, and L are compressed characters, so the corresponding 4-bit compressed code 1
Convert to 18I+, “12TI, IIcIZI “6”.

The next E is also a compressed character, but as mentioned above, it is the last character in an odd number of consecutive compressed characters, so it is converted to an 8-bit F3'', and / is not a compressed character, so it is an 8-bit “
Convert to 'EB'.

That is, as shown by the arrows, each character is converted and the more characters are compressed and encoded, the shorter the transmitted data becomes, and the transmission efficiency improves.

The receiving device checks the transferred data 8 bits at a time, determines whether both the upper 4 bits and lower 4 bits are compressed characters, and performs inverse conversion. First 8-bit part 9 “II
Both 3I+ are compressed codes, so 9" is set to H, and 3
'' are respectively inversely converted to E and restored.

The next 8-bit "CD" is converted back to the character Q corresponding to the 0 row of the conversion table in FIG. 2, since D is not a compressed code.

Similar processing is performed for the following, and the characters are restored by inverse conversion.

However, in such conventional methods, the number of compressed characters is limited, and therefore sufficient compression efficiency cannot be obtained.

The present invention improves the above-mentioned drawbacks, and aims to increase the number of compressed characters by setting a plurality of groups of characters that tend to appear consecutively, and also identifies switching of character groups to be compressed. The purpose of this is to improve compression efficiency.

Examples will be described in detail below.

FIG. 4 is a block diagram of a transmitting device according to an embodiment of the present invention, in which 411 is a transmitting device, 412 is a control circuit, 413, 4
16, 422, 426, 428 are signal lines, 414 is a buffer, 415 is an address register, 417 is a conversion circuit,
418,421°425.427 are buffer registers,
419 is a display register that displays whether or not the current character is a compressed character; 420 is a display register that displays whether the current character is a compressed character;
A compressed character group display register that displays the compressed character group to which the compressed character belongs, 429 a register that displays whether or not the previous character is an odd-numbered compressed character, 423 a line buffer, and 424 a register that displays the compressed character group immediately before the compressed character. This is a compressed character group display register that displays the compressed character group to which it belongs.

FIG. 5 is a block diagram of a receiving device according to an embodiment of the present invention, in which 451 is a receiving device, 452 is a control circuit, 453, 4
59 is a signal line, 454 is a line buffer, 455 is a buffer register, 456 and 458 are address registers, 45
7 is a compressed character group display register, 460 is a conversion circuit, 46
1 is a buffer.

FIG. 6 shows an example of the conversion table memory provided in the conversion circuits 417 and 460, and the compressed characters are English capital letters 5.
Compressed character group 11 consisting of A, D, E, G, I, L, N, O, R, S, T, U numbers 0-9. Compressed character group consisting of *, ■, English lowercase letter '? ) Ha Hd t et
It is divided into three groups: a compressed character group ■ consisting of g + ij19m90trts and ttu.

Since the compressed character group is divided into three groups consisting of English capital letters, numbers, and English hand letters, the transmitter 411
It is necessary to determine which compressed character group is currently performing data compression, and the receiving device 451 also needs to determine which compressed character group the transferred compressed code belongs to.

Therefore, the first character belonging to different compressed character groups is transmitted without being compressed into a code, thereby making it possible to identify the compressed character groups.

An overview of the operations on the sending and receiving sides is shown below.

On the sending side, characters other than compressed characters are transferred in 8 bits,
For compressed strings with multiple consecutive compressed characters, (1
) If the compressed character string and the compressed character immediately before the compressed character string belong to the same compressed character group, and the number of characters in the compressed character string is an even number, convert the compressed characters belonging to the compressed character string into 4 bits. If the compressed character string is compressed and transferred, and the number of compressed characters belonging to the compressed string is an odd number, only the last character of the compressed string is transferred in 8 bits, and the other compressed characters are compressed to 4 bits. Forward.

(2) If the compressed character string and the compressed character immediately before the compressed character string belong to different compressed character groups, the first character of the compressed character string is transferred in 8 bits, and the remaining characters are an even number. If the number of remaining compressed characters is an odd number, all remaining compressed characters are compressed to 4 bits and transferred, and if the number of remaining characters is an odd number, the last character is
Other compressed characters are compressed to 4 bits and transferred.

On the other hand, on the receiving side, the upper 4 bits and lower 4 bits of each 8-bit character are checked to determine whether or not they are both compressed codes.

(a) If both are compressed codes, each is converted into an 8-bit compressed character.

(b) If either one is not a compressed code, it is assumed to be an 8-bit character and the following determination is made.

(bl) Characters that cannot be compressed, or (b2) Characters that can be compressed, but
Is it the last odd-numbered character of the compressed string (b3)?
At the beginning of a compressed character string, it is determined whether the character belongs to a compressed character group different from the compressed character group to which the immediately preceding compressed character belongs.

Next, the operations of the transmitting device and the receiving device will be explained in detail with reference to FIGS. 4 and 5.

The transfer data set in the buffer 414 is transferred from the control circuit 412 via the signal line 413 as a transfer signal to the buffer 4.
14, transfer data of 8 bits, that is, one sentence, is set from the buffer 414 to the address register 415.

When the conversion circuit 417 receives a conversion instruction signal from the control circuit 412 via the signal line 416, the conversion circuit 417 converts the address register 415
Depending on the configuration of the conversion table memory, the conversion code and the code of the transfer data set in the address register 415 may be the same, and the data set in the address register 415 may be the same as the transfer data code. It may happen.

), sets a conversion code in the buffer register 418, and sets information on whether the transferred character is a compressed character in the buffer register 419, and if it is a compressed character, which compressed character group the compressed character belongs to. information is set in the compressed character group display register 420, and a 4-bit compressed code corresponding to the transferred character is set in the buffer register 42.
Set to 1.

Further, the compressed character group display register 424 stores compressed character group information indicating which compressed character group the transmitting device 411 is currently performing data compression on.

Next, the control circuit 412 controls the contents of the display register 419, which displays whether or not the current transferred character is a compressed character, and the register 429, which displays whether or not the previous transferred character is an odd-numbered compressed character. If the transferred character is an odd-numbered compressed character, °1' is set, and if it is not an odd-numbered compressed character, "0" is set.) The contents are read and their contents are determined.

Based on the respective determination results, one of the processes (1) to (4) shown below is performed, and then the above-described process is performed again.

(1) If the current transfer character is a compressed character and the previous transfer character is an odd-numbered compressed character (register 429
1'), the contents of the compressed character group display register 424 and the contents of the compressed character group display register 420 are compared, and if they match, the compressed code set in the buffer register 42L427 is transferred to the line buffer 423. Thereafter, 0'' is set in the register 429 which indicates whether or not the previous transferred character is an odd-numbered compressed character.

In addition, if the comparison does not match, the buffer register 418,
After transferring the 8-bit code set in 425 to the line buffer 423, the register 429 is set to "0".

(2) If the current transferred character is a compressed character and the previous transferred character is not an odd-numbered compressed character (register 429 = 'O
), the contents of the compressed character group display registers 420 and 424 are compared, and if the comparison is negative, the buffer register 4
The 4-bit compressed code set in 21 is transferred to the buffer register 42γ, and the 8-bit code set in the buffer register 418 is transferred to the buffer register 42γ.
At the same time, the register 429 is set to 1.

If the comparison does not match, the 8-bit code set in the buffer register 418 is transferred to the line buffer 423, and the contents of the compressed character group display register 424 are rewritten to the contents of the compressed character group display register 420.

In this case, the contents of register 429 are held at "0".

(3) If the current transferred character is not a compressed character and the previous transferred character is an odd-numbered compressed character (register 429
="1'), the contents of buffer registers 418 and 425 are transferred to line buffer 423, and then register 4 is transferred to line buffer 423.
29 is set to "O'".

(4) If the current transfer data is not a compressed character and the previous transfer data is not an odd-numbered compressed character (register 4
29='0'), the contents of the buffer register 418 are transferred to the line buffer 423.

In this case, the contents of register 429 are held as 0.

As mentioned above, uncompressed characters are sent as 8-bit conversion codes by the conversion circuit 417, and compressed characters are sent as 8-bit conversion codes for the first character belonging to a compressed character group, and for subsequent characters belonging to the same compressed character group. is sent as a compressed code.

For example, when using the conversion table memory shown in FIG.
When sending transfer data rDATA*53adJ, '
F2', "1", "13 +.

"F1', "EB", "5", "3', 'Di'"
D2".

In the receiving device, the received data is transferred to the line buffer 454.
The control circuit 452 sends a transfer signal to the line buffer 454 via the signal line 453, and the line buffer 4
The received data set in 54 is set in the buffer register 455 8 bits at a time.

In the control circuit 452, the upper 4 bits and lower 4 bits of the received data set in the buffer register 455 are set from 0 to C.
It is checked in the same way as in the conventional method whether or not the code is a compressed code.

If both of these determination results are compressed codes, the control circuit 452
sends a control signal to the buffer register 455, sets the upper 4 bits of the received data set in the buffer register 455 to the address register 456, and also sets the current data compression value set in the compressed character group display register 457 in the method of the present invention. The type of the target compressed character group is set in the address register 458, and a conversion instruction signal is sent to the conversion circuit 460 via the signal line 459.

Upon receiving the conversion instruction signal, the conversion circuit 460 converts the 4-bit compressed code set in the address register 456 into compressed characters according to the information indicating the type of compressed character group set in the address register 458, and sets the compressed characters in the buffer 461. .

Next, the lower 4 bits of the received data set in the buffer register 455 are set in the address register 456,
From the control circuit 452 to the conversion circuit 46 via the signal line 459
Sends a conversion instruction signal to 0.

Upon receiving the conversion instruction signal, the conversion circuit 460 converts the compressed code set in the address register 456 into compressed characters according to the contents set in the address register 458, and sets the compressed characters in the buffer 461.

The received data set in the buffer register 455 is checked, and if both or one of the upper 4 bits and lower 4 bits is not a compressed code, the control circuit 452 checks the upper 4 bits to determine whether or not this received data is a compressed character. It is determined whether the lower 4 bits of E and F are 0 to C. If it is a compressed character, the received data set in the buffer register 455 and the compressed character group set in the compressed character group display register 457 are used. Compare with the type of .

(In the conventional method, the upper 4 bits are F and the lower 4 bits are 0.
~C) If the comparison matches, this is the last character of an odd number of consecutive compressed characters, so the upper 4 bits of the received data set in the buffer register 455 are set in the address register 458. Then, the lower 4 bits are set in the address register 456, and a conversion instruction signal is sent from the control circuit 452 to the conversion circuit 460 via the signal line 459.

Upon receiving the conversion instruction signal, the conversion circuit 460 reads the upper 4 bits and lower 4 bits of the received data set in the address registers 456 and 458, and stores the data in the buffer register 4.
The character is converted into a character corresponding to the 8-bit code set to 55 and set in the buffer 461.

If the comparison does not match, the type of the compressed character group set in the compressed character group display register 451 is changed to the type of the compressed character group to which the compressed character set in the buffer register 455 belongs, and the type of the compressed character group set in the buffer register 455 is changed. The upper 4 bits and lower 4 bits of the set received data are set in address registers 458 and 456.

The control circuit 452 sends a conversion instruction signal to the conversion circuit 460 via a signal line 459. Upon receiving this conversion instruction signal, the conversion circuit 460 reads the code set in the address registers 456 and 458, and sets the code in the buffer register 455. Convert the 8-bit code into the corresponding character,
Set in buffer 461.

The control circuit 452 determines whether the received data set in the buffer register 455 is a compressed character, and if the determination result is not a compressed character, the buffer register 45
The upper 4 bits of the received data set to 5 are set in the address register 458, the lower 4 bits are set in the address register 456, and the conversion circuit 460 converts them into characters corresponding to the 8-bit code set in the buffer register 455. , and set it in the buffer 461.

As explained above, the present invention provides a plurality of compressed character groups that are the targets of data compression, and when transferring data, determines whether the transferred character is a compressed character, which compressed character group it belongs to, and Compare the compressed character group targeted for data compression and the compressed character group to which the transfer character belongs,
The first character of consecutive transfer characters belonging to the same compressed character group is sent without being converted into a compressed code, and the next character is sent as a compressed code based on the conversion table, and on the receiving side, the received code is a compressed code. , and which compressed character group the inversely converted compressed character belongs to.Based on the result of this determination, it is determined which compressed character group is the compressed character group on the sending side that is currently the target of data compression. judge,
Perform data restoration.

Therefore, since the number of compressed characters can be increased, compression efficiency can be improved, and it is very effective when applied to batch processing or the like in which a large number of characters with a high probability of appearing consecutively are output.

[Brief explanation of the drawing]

Figure 1 is a block diagram of a conventional data compression method, Figure 2 is an explanatory diagram of an example of a conventional conversion table, and Figure 3 is an explanatory diagram of an example of data compression using the conversion table of Figure 2. , FIG. 4 is a block diagram of a transmitting device according to an embodiment of the present invention, and FIG. 5 is a block diagram of a receiving device according to an embodiment of the present invention.
FIG. 6 is an explanatory diagram of an example of a conversion table memory used in the method of the present invention. 11.411 is a transmitting device, 12 is a transmission conversion table, 1
3,451 is a receiving device, 14 is a receiving conversion table, 41
2,452 is a control circuit, 413°416.422,42
6,428,453,459 are signal lines, 414,461
is a buffer, 415°456.458 is an address register, 417°460 is a conversion circuit, 418, 421, 42
5°427.455 is a buffer register, 423°45
4 is the line buffer, 419 is the display register, 420.4
24, 457 is a compressed character group display register, and 429 is a register.

Claims (1)

[Claims] 1. In a data compression method that compresses and decompresses data using a conversion table, a plurality of compressed character groups are provided in the conversion table, in which characters with a high probability of appearing consecutively are grouped into arbitrary numbers, and the sending side the conversion table, means for determining whether each character of the transfer data belongs to which group of the compressed character groups, and storage means for storing the compressed character group to which the characters of the previous transfer data belong; a comparison means for comparing the contents of the storage means and the determination result of the determination means, the first character of consecutive transfer data belonging to the same compressed character group is transmitted without being compressed; transmitting a compressed code according to the conversion table from the next character of A data compression method characterized in that data is restored using the conversion table based on the result.