JP4093200B2 - Data compression method and program, and data restoration method and apparatus - Google Patents

Description

  The present invention relates to a method of compressing data by encoding based on a match with an encoded symbol string, and in particular, it can improve a data compression rate and spend time on data compression and decompression processing. The present invention relates to a data compression method that does not cost.

  In recent years, with the development of information processing technology and the spread of networks such as the Internet, a huge amount of data has been processed and communicated. For example, it is a common practice to cause a printer to execute printing from a host computer via a network, and the processing speed of the printer is improved as needed, but the capacity of print data transmitted from the host computer to the printer is large. In some cases, the communication takes time, and the processing speed of the printer cannot be fully utilized. Therefore, a data compression and decompression technique is required for how data is compressed and transmitted and how it is restored after reception.

  Conventionally, several data compression and decompression techniques have been proposed depending on the purpose and target data. One of them is a method using an LZ77 (Ziv and Lempel (1977)) code. In this method, a symbol string before and after a symbol (data) that is currently encoded is stored in a buffer, and this buffer is referred to as a dictionary, and the longest symbol string that matches the symbol string to be encoded is stored. In the dictionary. Then, the target symbol string is encoded based on the length and position information of the searched symbol string.

  Various variations have been developed for the LZ77 system. For example, a method using an LZSS code described in Non-Patent Document 1 below is frequently used. In this method, when the length of the matching longest symbol string is smaller than a predetermined value, the original symbol is used as it is, that is, the above matching symbol string is not encoded according to the length and position. When the length of the longest symbol string to be performed is larger than a predetermined value, the above-described encoding based on the length and position of the matching symbol string is performed. Then, a flag for distinguishing both is added to the head of the code.

Further, several proposals have been made for such an LZ77 data compression method for the purpose of improving the compression rate (for example, Patent Documents 1 and 2 below).
Japanese Patent Laid-Open No. 5-11973 JP 7-261977 A Tomohiko Uematsu, “Introduction to Document Data Compression Algorithm”, 2nd edition, CQ Publishing, June 1995, p. 145-148

  However, in the above-described LZSS code, the position (hereinafter referred to as the matching position) and the length (hereinafter referred to as the matching length) of the longest matching symbol string are expressed by a fixed-length code (bit string). There was a problem that the data compression rate was not so good.

  Further, Patent Document 1 describes a method for improving the data compression rate by attaching an appearance number to a matching symbol string in the dictionary and outputting the appearance number as the matching position. However, such a method has a problem in that it takes time for the restoration process because it is necessary to obtain the appearance position of the matching symbol string by searching even when the compressed data is decoded.

  Further, in Patent Document 2, a technique for expressing the matching position and the matching length with a code that dynamically changes according to a spray tree and improving the compression ratio is shown. Because of the change, both the compression and decompression processes are complicated, and there is a problem that these processes take time.

  Therefore, an object of the present invention is to provide a data compression method based on the LZ77 code, which can improve the compression rate, and that does not take time for data compression and decompression processing.

  In order to achieve the above object, one aspect of the present invention is to search for a longest match sequence having a maximum length match with a symbol string to be encoded, existing in an encoded symbol string, and to perform the longest match. A data compression method for compressing data by performing encoding based on a match position that is a sequence existing position and a match length that is a length of the longest match sequence, wherein the code generated by the encoding is A first position including information indicating whether the value of the matching position and / or the matching length is the same as the value at the previous encoding, and the matching position not equal to the value at the previous encoding, and / or And a second code representing the value of the coincidence length. Therefore, according to the present invention, since the coincidence position and / or coincidence length information that is the same as the previous encoding value is not added, the data compression rate can be further increased.

  Furthermore, in the above invention, a preferable aspect thereof is characterized in that the second code is a variable-length code, and the first code includes information on the length of the second code. As a result, the data compression rate can be further increased.

  Furthermore, in the above invention, a preferred aspect is that the second code is obtained by removing the highest-order 1 when the value representing the coincidence position and / or the coincidence length is expressed in binary. Features. Thereby, the data compression rate can be further increased.

  In order to achieve the above object, another aspect of the present invention is to search for a longest matching sequence having a maximum length match with a symbol string to be encoded, existing in an encoded symbol string, and to perform the longest match. A data compression method for compressing data by encoding based on a match position that is a sequence existing position and a match length that is a length of the longest match sequence, wherein the length of the longest match sequence is a predetermined length If the value is smaller than the value, the encoding is performed with the first identification information having a predetermined length indicating that the original symbol is smaller than the predetermined value, and the length of the longest matching sequence is the predetermined value. If the value is greater than or equal to the value, information indicating that the value is equal to or greater than the predetermined value, whether the match position and / or the match length value is the same as the previous encoding value, and the previous value Matching position indicating the matching position and / or matching length that is not the same as the encoded value Represents the length of the information and / or the matching length information, the second identification information having the same length as the first identification information, and the matching position and / or the matching length that is not the same as the previous encoding value The encoding is performed using the matching position information and / or the matching length information.

  Furthermore, in the above-mentioned invention, the preferred mode is that the matching position information and / or matching length information is the highest-order 1 when a value representing the matching position and / or the matching length is expressed in binary. It is characterized by being excluded.

  Furthermore, in the above invention, a preferable aspect is that, when the value of the matching position and / or the matching length is the same as the value at the previous encoding, the matching position information included in the second identification information and Alternatively, the length of the matching length information is set to 0.

In the above invention, a preferred aspect is characterized in that the first identification information and / or the second identification information is Huffman-coded.

In order to achieve the above object, another aspect of the present invention is to search for a longest matching sequence having a maximum length match with a symbol string to be encoded, existing in an encoded symbol string, and to perform the longest match. A data compression program for causing a computer to execute a process of encoding and compressing data based on a matching position that is a sequence existing position and a matching length that is the length of the longest matching sequence, the matching position and / or A step of determining whether or not the value of the matching length is the same as the value at the previous encoding, and based on the result of the determination, whether or not the value is the same as the value at the previous encoding If the value of the matching position and / or the matching length is not the same as the value at the previous encoding after the output and the step of outputting the code including the information of the matching position and / or the matching length Output a sign representing the value of It is to perform the up to the computer.

  In order to achieve the above object, still another aspect of the present invention is to search for a longest matching sequence having a maximum length match with a symbol string to be encoded, existing in an encoded symbol string, and A data restoration method for restoring data encoded based on a match position that is a position where a match series exists and a match length that is the length of the longest match series, wherein the data to be restored is the match position And / or the first code including information on whether or not the value of the match length is the same as the value at the previous encoding, and the match position and / or the match not equal to the values at the previous encoding Whether or not the matching position and / or the matching length value is the same as the previous decoding value based on the first code when the second code representing the length value is included. Is determined to be the same as the previous decoding value. The value of the position and / or the matching length is obtained from the value at the previous decoding, and the value of the matching position and / or the matching length determined not to be the same as the value at the previous decoding is calculated from the second code. Obtaining and decoding using the decoded symbol string based on the obtained match position and match length values.

In order to achieve the above object, another aspect of the present invention is to search for a longest matching sequence having a maximum length match with a symbol string to be encoded, existing in an encoded symbol string, and to perform the longest match. A symbol string storage means for storing the restored symbol string in a data restoration device that restores the encoded data based on the matching position that is the position of the series and the matching length that is the length of the longest matching series The previous value storage means for storing the latest value of the match position and the match length value obtained at the time of restoration, and the data to be restored are the match position and / or the match length value. A second code representing the value of the matching position and / or the matching length that is not the same as the previous encoding value Based on the first code. The match position and / or the match length value are determined to be the same as the previous encoding value, and the match is determined to be the same as the previous encoding value. The value of the position and / or the matching length is obtained from the value stored in the previous value storage means, and the value of the matching position and / or the matching length determined not to be the same as the value at the previous encoding, Decoding means that obtains from the second code and performs decoding using the symbol string stored in the symbol string storage means based on the obtained match position and match length value. .

  Further objects and features of the present invention will become apparent from the embodiments of the invention described below.

Embodiments of the present invention will be described below with reference to the drawings. However, such an embodiment does not limit the technical scope of the present invention. In the drawings, the same or similar elements are denoted by the same reference numerals or reference symbols.
FIG. 1 is a configuration diagram according to an embodiment of a data compression apparatus to which the present invention is applied. FIG. 2 is a configuration diagram according to an embodiment of a data restoration apparatus to which the present invention is applied. The data compression apparatus 1 and the data decompression apparatus 2 shown in FIGS. 1 and 2 are apparatuses using the data compression method and the decompression method according to the present invention. The data compression method of the data compression apparatus 1 is based on the concept of the LZ77 code, but the values at the previous processing are stored for the match position and match length values used for encoding, and those values are stored. In the case where the value is the same as the value, the fact that it is included in the code removes the information on the coincidence position and the coincidence length from the code and tries to improve the data compression rate by shortening the code length as much as possible. Further, the data restoration method of the present data restoration device 2 stores the values of the previous processing for the values of the matching position and the matching length in the same manner as the above compression, and adds those previous values added at the time of compression. By using information on whether or not it is the same as the value, information on the matching position and matching length is easily obtained, and the information is decoded into the original symbol string. The process is realized.

  As shown in FIG. 1, the data compression device 1 is composed of an input buffer 11, an encoding unit 12, a dictionary buffer 13, and a previous value buffer 14, and encodes input data to compress the data capacity. is there. Here, input data before encoding is referred to as a symbol (column). In this embodiment, each symbol, for example, “A”, “B” in the input buffer 11 of FIG. "Etc." are assumed to be 8-bit data. Further, the processed code is expressed by a bit string of “0” or “1” in binary.

  The input buffer 11 is a data buffer that sequentially receives and stores symbol strings to be encoded, and sequentially transfers the symbol strings that have been encoded to the dictionary buffer 13. Further, the length of the input buffer 11 (L2 in FIG. 1), that is, the number of symbols that can be stored is assumed to be 257 in the present embodiment. This length L2 means the maximum value of the matching length (for example, 1 in FIG. 1).

  The dictionary buffer 13 is a data buffer that sequentially receives encoded symbol strings from the input buffer 11 and stores them, and is called a so-called dictionary in the LZ77 system. The symbol string stored here is compared with the symbol string to be encoded in the input buffer 11, and the maximum length symbol string that matches the symbol string in the input buffer 11 (hereinafter referred to as the longest matching sequence). Is called). In the dictionary buffer 13, every time encoding in the encoding unit 12 is completed, the encoded symbol string is accepted, and the symbol string at the head (located on the left side in FIG. 1) is unnecessary. Will be exhaled. In addition, the length of the dictionary buffer 13 (L1 in FIG. 1), that is, the number of symbols that can be stored is 16383 (16K) in the present embodiment. This length L1 means the maximum value of the distance (for example, d in FIG. 1) representing the coincidence position. In the example shown in FIG. 1, the longest matching sequence is “ABCD”, the matching length which is the length thereof is l, and the matching position is p.

  Next, the encoding unit 12 is a part that performs general control related to the input, encoding, and output of the symbol string in the data compression apparatus 1, mainly referring to the dictionary buffer 13, A process of encoding the symbol string in the input buffer 11 is executed. The encoding process performed by the encoding unit 12 is characterized, and the specific contents thereof will be described later. The encoding unit 12 may be configured by a program indicating a processing procedure, a CPU that executes processing according to the program, or may be configured by a hardware circuit.

  The previous value buffer 14 is a part that holds the value of the matching position and the matching length at the time of the previous encoding process using the matching position and the matching length in the encoding unit 12. Each time the encoding process using the matching length is performed, the value held is updated. One of the features of the data compression apparatus 1 is that the value held in the previous value buffer 14 is used for the encoding process in the encoding unit 12.

  As shown in FIG. 2, the data restoration device 2 has the same configuration as that of the data compression device 1, and includes an input buffer 21, a decoding unit 22 (decoding unit), a dictionary buffer 23 (symbol string storage unit), and a previous value. A buffer 24 (previous value storage means) is provided. The data restoration device 2 is a device for restoring the data (code) compressed by the data compression device 1 to the original symbol string data. The input buffer 21 is a data buffer that sequentially receives and stores codes to be processed.

  The dictionary buffer 23 is a data buffer that sequentially receives and stores decoded symbol strings from the decoding unit 22. The symbol string stored here is referred to during the decoding process by the decoding unit 22 and is used for the decoding process. In the dictionary buffer 23, every time the processing in the decoding unit 22 is completed, the decoded symbol string is accepted, and the corresponding symbol string at the head (located on the right side in FIG. 2) is spouted as an unnecessary symbol. Further, the length (size) of the dictionary buffer 23 is the same as the length L1 of the dictionary buffer 13 of the data compression apparatus 1.

  Further, the state in the dictionary buffer 23 is the same as the state of the dictionary buffer 13 when the code to be processed in the decoding unit 22 at that time is encoded in the data compression apparatus 1. In the example shown in FIG. 1, when the symbol string “ABCD” is encoded based on the matching position p (or d) and the matching length l, when the code is decoded by the data restoration device 2. The state in the dictionary buffer 23 is as shown in FIG. That is, the symbol strings (expressed in the reverse direction in FIG. 2) are stored in the order of “ABCD” from the position p at a distance d from the left end of the dictionary buffer 23. Therefore, the decoding unit 22 can decode “ABCD” based on the matching position p (or d) acquired from the code and the matching length l.

  The decoding unit 22 is a part that performs general control related to code input, decoding, and symbol output in the data restoration device 2, and mainly refers to the code in the input buffer 21 while referring to the dictionary buffer 23. The process which decodes is performed. Specific contents of the decoding process performed by the decoding unit 22 will be described later. The decoding unit 22 may be configured by a program showing a processing procedure, a CPU that executes processing according to the program, or may be configured by a hardware circuit.

  The previous value buffer 24 is a part that holds the value of the match position and the match length acquired during the previous process in the decoding unit 22 of the code encoded using the match position and the match length. Yes, each time such processing is performed, the value held is updated. One of the characteristics of the data restoration device 2 is that the value held in the previous value buffer 24 is used for the decoding process in the decoding unit 22.

  FIG. 3 is a flowchart illustrating the contents of the process performed by the encoding unit 12 of the data compression apparatus 1. Hereinafter, based on FIG. 3, the specific content of the compression process performed by this data compression apparatus 1 is demonstrated. First, the encoding unit 12 initializes the coincidence position and coincidence length values (the values of d and l in the example of FIG. 1) held in the previous value buffer 14 (step S1). Specifically, both values may be 0, or may be values that frequently appear during encoding. Next, the symbol strings to be processed are sequentially read into the input buffer 11 (step S2). Thereafter, the encoding unit 12 searches for a match with the symbol string stored in the dictionary buffer 13 with respect to the symbol string from the head position (leftmost position in FIG. 1) stored in the input buffer 11 (step S3). ). That is, the longest matching sequence described above is searched.

  Then, it is checked whether or not the length of the searched longest matching sequence (matching length) is greater than or equal to a predetermined value (step S4). For example, the predetermined number is “3”. As a result, if the match length is 2 or less (No in step S4), the encoding unit 12 does not represent the symbol at the head position of the input buffer 11 with the match position of the longest match sequence and the match length. In addition, the symbol is encoded by a method of expressing the symbol as it is in the binary system. And the code | symbol is output (step S5). Specifically, in the present embodiment, since the symbol is 8-bit data, the 8-bit data as the original symbol is represented by 9 bits, which is used as the code of the symbol.

  FIG. 4 is a diagram for explaining codes and the like generated by the data compression apparatus 1. FIG. 4A schematically shows a code generated by the data compression apparatus 1, and the diagram shown in the upper part (denoted [mismatch]) is the code output in step S5. It is. In other words, the code in the case where there is no match in the dictionary buffer 13 for a predetermined number of symbol strings from the head position of the input buffer 11 is shown. As described above, since this code originally represents 8 bits of data in 9 bits, the most significant bit is always “0”. As will be described later, when the matching length is greater than or equal to the predetermined number in step S4, the most significant bit is always “1”, so the most significant bit is “mismatched” for the predetermined number of symbol strings. ". The subsequent 8-bit data is the original symbol itself.

  Therefore, the code is identified by the most significant bit as being “mismatched”, that is, indicating the symbol itself in the subsequent bit string. In other words, an expression that can take a value of 0 to 511 is used to express a value of 0 to 255, thereby identifying the original symbol as it is. The entire code in the case of “mismatch” including the “mismatch” and the original symbol is referred to as identification information (first identification information) in the case of “mismatch”.

  Returning to FIG. 3, when the matching length is equal to or larger than the predetermined number in step S4 (Yes in step S4), the encoding unit 12 sets the matching position and the matching length of the searched longest matching sequence. The number of bits (bit string length) necessary to represent is obtained (step S6). In the present embodiment, the matching position takes a value of 1 to 16383 based on the size of the dictionary buffer 13 described above, so the number of bits necessary to represent the matching position is a value of 1 to 14. Further, since the matching length takes a value of 3 to 257 from the size of the input buffer 11 and the condition in step S4, the number of bits necessary to represent the matching length is a value of 1 to 8. .

  Next, the encoding unit 12 checks whether or not the value of the matching position of the longest matching sequence searched this time is the same as the value of the matching position held in the previous value buffer 14 (step S7). . If they are the same (Yes in step S7), the number of bits necessary to represent the obtained matching position is set to 0 (step S8). On the other hand, if they are not the same, that is, if they are different (No in step S7), the number of bits necessary to represent the obtained matching position is not changed.

  Next, the encoding unit 12 checks whether or not the match length value of the longest match sequence searched this time is the same as the match length value held in the previous value buffer 14. (Step S9). If they are the same (Yes in step S9), the number of bits necessary to represent the obtained matching length is set to 0 (step S10). On the other hand, if they are not the same, that is, if they are different (No in step S9), the number of bits necessary to represent the obtained matching length is not changed.

  When the number of bits of the matching position and the matching length is determined in this way, the encoding unit 12 determines the number of bits of the matching position and the matching length and the dictionary buffer 13 for a predetermined number of symbol strings from the head position of the input buffer 11. A code indicating that there is a “match” is generated and output (step S11). Such a code is referred to herein as identification information (first code, second identification information) in the case of “match”. Specifically, as an example, a value such as the following formula (1) is generated as identification information.

Identification information (in the case of “match”) = 256 + BL × (maximum BP + 1) + BP (1)
Where BL is the number of bits required to represent the matching length
BP: Number of bits required to represent the matching position
Maximum BP: Maximum number of bits necessary to represent the matching position In the present embodiment, as described above, the maximum BP has a value of 14, and BP and BL have the above values. Thus, the identification information (in the case of “match”) has a value of 256 to 390. Therefore, the identification information (in the case of “match”) is also expressed by 9 bits as a binary code. The four codes (denoted [match]) shown in the lower part of FIG. 4A indicate codes that are generated and output when there is a “match” of the predetermined number of symbol strings. The 9 bits on the left side correspond to the identification information.

  As described above, since the identification information takes a value of 256 or more, it is possible to identify that there is a “match”, and to match the position and match by the value obtained by subtracting 256 from the value of the identification information. The number of bits required to represent the length is shown. In other words, the most significant bit of the code represented by 9 bits always has a value of “1”, which indicates that there is a “match”, and the lower 8 bits indicate the match position and match length. The number of bits necessary to represent Further, when the number of bits necessary to represent the coincidence position and the coincidence length is 0, each of those values is the same as the value held in the previous value buffer 14, that is, the previous time, This indicates that the match position and match length values are the same when a predetermined number or more of the symbols match.

  As described above, in the code according to the present embodiment, the identification information is expressed by 9 bits both in the case of “match” and “mismatch”, depending on whether the value is 256 or more. In other words, it is possible to identify whether or not encoding is performed based on the matching position and the matching length by the most significant bit. In addition, after the identification, the lower 8 bits can be used to know the original symbol itself in the case of “mismatch”, and in the case of “match”, the matching position and the matching length that follow the identification information can be obtained. It can be known that the length of the code to be represented, or the value of the matching position and the matching length are the same as the previous value.

  When the output of the identification information is finished, the encoding unit 12 checks whether or not the number of bits necessary to represent the coincidence position is 0 (step S12), and if not (0 in step S12). ), The matching position information (second code) indicating the matching position of the searched longest matching sequence is output as a code (step S13). Specifically, a lower-order bit string excluding the most significant “1” bit when the value representing the matching position is expressed in binary notation is output as matching position information. This is self-evident that the most significant bit is “1”, and is for the purpose of reducing the amount of data as much as possible. For example, when the value of the coincidence position is 9, although “1001” is obtained in the binary system, “001” is output as the coincidence position information.

Further, if the matching position and the matching length are expressed in comparison with the case of the LZSS code expressing the fixed-length bit string, the matching position is expressed by the maximum length of 14 bits, and the MSB (Most Significant) of the bit string is expressed. Bits) obtained by removing the “0” bit continuously present from the “Bit” side and the next “1” bit are used as matching position information.
On the other hand, when the number of bits necessary to represent the coincidence position is 0 (Yes in step S12), the coincidence position information is not output.

  Next, the encoding unit 12 checks whether or not the number of bits necessary to represent the matching length is 0 (step S14). If the number is not 0 (No in step S14), the search is performed. The matching length information (second code) indicating the matching length of the longest matching sequence is output as a code (step S15). Similarly to the match position information, the match length information is output as a match length information by subordinate bit strings excluding the most significant “1” when the value representing the match length is expressed in binary. . On the other hand, if the number of bits necessary to represent the matching length is 0 (Yes in step S14), the matching length information is not output.

  In the “match” code illustrated in the lower part of FIG. 4A, “match position information” and “match length information” shown on the right side correspond to each information output in steps S13 and S15. Among them, the code indicated by (difference, difference) indicates the code output when the value of the matching position and the value of the matching length are not the same as the value of the previous value buffer 14. . Hereinafter, similarly, the signs indicated by (difference, same), (same, different), and (same, same) are respectively the same as the value of the previous value buffer 14 only in the value of the matching length. In this case, the code output when only the value of the matching position is the same as the value of the previous value buffer 14 and when the value of the matching position and the value of the matching length are the same as the value of the previous value buffer 14 Is shown. In this way, the same information as the previous value is not output, and further data compression is achieved.

  As described above, when the output of the match length information is finished, the encoding unit 12 searches the match position and the match length value held in the previous value buffer 14 respectively for the longest match searched in the current process. The sequence is updated to match position and match length values (step S16).

  As described above, when the code generation and output process in the case of “mismatch” or “match” is completed, the symbol strings stored in the input buffer 11 and the dictionary buffer 13 are slid (step S17). ). Specifically, the symbol (sequence) generated by the code generated by the above processing and output is moved from the head portion of the input buffer 11 to the tail portion of the dictionary buffer 13. Then, a new symbol (sequence) for the movement is input to the input buffer 11, and a symbol (sequence) for the movement is discharged from the dictionary buffer 13. In the case of “mismatch”, only one symbol is encoded, so there is only one moving symbol. In the case of “match”, three or more symbols are encoded. The moving symbol is 3 or more.

  Thereafter, it is determined whether or not the compression process to be performed this time has been completed (step S18). If the compression process has not been completed (No in step S18), the processes from step S3 described above are repeated. When it is determined that the compression process has been completed (Yes in step S18), the series of data compression processes is terminated.

  FIG. 5 is a flowchart illustrating the contents of the process performed by the decryption unit 22 of the data restoration device 2. Hereinafter, based on FIG. 5, the specific content of the restoration process performed by the data restoration apparatus 2 will be described. First, the decoding unit 22 initializes the value of the matching position and the matching length held in the previous value buffer 24 (step S21). Specifically, the value is the same as the value at the time of initialization of the previous value buffer 14 when the data to be decoded is encoded. Thereafter, the codes to be processed are sequentially read into the input buffer 21 (step S22). Next, the decoding unit 22 reads the first 9 bits of the code, that is, identification information from the input buffer 21 and interprets the information (step S23).

  Then, it is determined whether the current processing target code is the case of “mismatch” or “match” at the time of encoding (step S24). As described above, this determination can be made based on the most significant bit of the identification information. If the most significant bit is “0”, it is determined that the case is “mismatch”, and the most significant bit is “ If it is “1”, it is determined that the case is “match”. In other words, if the value of the read identification information is 0 to 255, it is determined that the value represents the original symbol itself, and if the value of the identification information is 256 or more, this identification information follows. It is determined that the number of bits of information on the matching position and the matching length can be known.

  If it is determined that the result is “mismatch” (No in step S24), since the read identification information represents the original symbol itself, the original symbol is restored and output. (Step S25). More specifically, 9-bit identification information is output as 8 bits.

  On the other hand, if it is determined that it is the case of “match” (Yes in step S24), it is necessary to represent the number of bits and the match length necessary to represent the match position from the read identification information. The number of bits is obtained (step S26). Specifically, since the identification information is generated by the above-described equation (1), 256 is subtracted from the value of the read identification information, and the subsequent value is divided by a value (14 + 1) larger by 1 than the maximum BP. The remainder at that time is the number of bits necessary to represent the coincidence position, and the quotient at that time is the number of bits necessary to represent the coincidence length.

  Next, the decoding unit 22 checks whether or not the number of bits necessary to represent the obtained coincidence position is 0 (step S27). If it is not 0 (No in step S27), the reading is performed. Following the identification information, the number of codes (bits) obtained by subtracting 1 from the number of bits necessary to represent the obtained matching position is read from the input buffer 21 (step S28). Thereafter, the decoding unit 22 obtains the value of the matching position from the read matching position information (step S29). Specifically, a value obtained by adding a bit of “1” to the most significant bit string of the read matching position information is set as the matching position value. In other words, the value of the matching position is obtained by adding the value of the matching position information to the power of 2 (the number of bits necessary to represent the matching position minus 1).

  On the other hand, if the number of bits necessary to represent the obtained matching position is 0 (Yes in step S27), the value of the matching position is set as the value of the matching position held in the previous value buffer 24 (step S30). ).

  Subsequently, the decoding unit 22 checks whether or not the number of bits necessary to represent the obtained matching length is 0 (step S31). If not, it is not 0 (No in step S31). Following the identification information, the number of codes (bits) obtained by subtracting 1 from the number of bits necessary to represent the obtained matching length is read from the input buffer 21 (step S32). Thereafter, the decoding unit 22 obtains a value of the matching length from the read matching length information (step S33). Specifically, a value obtained by adding 2 to the value obtained by adding the bit “1” to the most significant bit string of the read match length information is set as the match length value. In other words, a value obtained by adding the value of the match length information and 2 to the power of 2 (the number of bits necessary to represent the match length minus 1) is set as the match length value. Here, 2 is added, as described above, the value of the matching length takes a value of 3 to 257, and 2 is subtracted from the value of the matching length in order to express it with 8 bits at the time of encoding. This is because the values (1 to 255) are encoded.

  On the other hand, if the number of bits necessary to represent the obtained match length is 0 (Yes in step S31), the match length value is set as the match length value held in the previous value buffer 24. (Step S34).

  Next, the decoding unit 22 restores the original symbol string from the value of the obtained matching position and matching length, and outputs the restored symbol string (step S35). Specifically, a symbol string corresponding to the coincidence length of the obtained value is read from the coincidence position of the obtained value of the symbol string stored in the dictionary buffer 23, and the read symbol string is used as the original symbol string. Output.

  Thereafter, the decoding unit 22 updates the match position and match length values held in the previous value buffer 24 to the match position and match length values obtained in the current process, respectively (step S36).

  As described above, when the decoding process in the case of “mismatch” or “match” is completed, the codes and symbol strings stored in the input buffer 21 and the dictionary buffer 23 are slid (step S37). Specifically, the code restored by the above processing is deleted from the input buffer 21, and a new code corresponding to the code is input to the input buffer 21. Further, the symbol string restored by the above processing is added to the tail portion of the dictionary buffer 23, and the corresponding symbol string at the head portion is discharged from the dictionary buffer 23 as an unnecessary symbol.

  Thereafter, it is determined whether or not the restoration process to be performed this time has been completed (step S38). If the restoration process has not been completed (No in step S38), the processes from step S23 described above are repeated. If it is determined that the restoration process has been completed (Yes in step S38), the series of data restoration processes is terminated.

  FIG. 6 is a diagram showing a specific example of processing performed by the data compression apparatus 1 and the data restoration apparatus 2. The example shown in the figure is a case where there is a “match” of the predetermined number of symbols or more at the time of data compression, the match position at that time is 300, and the match length is 21. Furthermore, this example is a case where the value of the matching length is the same as the value of the matching length at the previous processing. FIG. 6A is a diagram schematically showing the matching position (300) and the matching length (21) with the maximum number of bits. FIG. 6B is a binary representation of the matching position (300) with the maximum number of bits.

  When the encoding by the encoding unit 12 described above is performed on this state, a code as shown in FIG. 6C is generated and output. The left part of the code shown in the figure is the above-described identification information (in the case of “match”), and the calculation shown in the figure is performed according to the equation (1). That is, since the value of the match length is the same as the value held in the previous value buffer 14, the number of bits necessary to represent the match length is 0, and the value is multiplied by 15 (maximum BP + 1) The number of bits 9 necessary for binary representation of 300, which is the value of the matching position, and 256 are added to the value to obtain identification information of 265.

  Also, the 8 bits in the right part of the code shown in FIG. 6C is the matching position information, and the most significant “1” in the bit string expressing 300 in binary is removed. Accordingly, 8 bits obtained by removing all “0” s on the MSB side and the most significant “1” from the 14-bit representation shown in FIG. 6B are output as codes. Further, as described above, since the value of the matching length is the same as the value held in the previous value buffer 14, the matching length information is not output.

  When such a code is received by the data restoration device 2 and decoded by the decoding unit 22, the identification information is interpreted, and the number of bits necessary to represent the matching position and the matching length is 9 and 0, respectively. It is judged that there is. Then, according to the processing contents described above, as shown in FIG. 6D, first, the (9-1) bit after the identification information is read, and the bit “1” is added to the most significant bit. Is done. Thereby, it is determined that the value of the coincidence position is 300. In addition, since the number of bits necessary to represent the matching length is 0, the value in the previous value buffer 24 is determined as the matching length value. In this case, since the previous value buffer 24 holds 21 as the value of the matching length, the value of the matching length is also acquired correctly. Thus, the decoding unit 22 extracts and outputs a predetermined symbol string from the dictionary buffer 23, and ends the decoding process for the code.

  As described above, in the data compression method and the decompression method according to the present embodiment, when there is no longest matching sequence of a predetermined length or more in the dictionary (buffer 13), a constant including the fact and the original symbol. The long identification information is output as a code. On the other hand, if there is a longest matching sequence of a predetermined length or longer in the dictionary, this is indicated, and the number of bits necessary to represent the matching position and the matching length are indicated. First, identification information having the same length as that of the identification information including the number of bits required is output, and thereafter, matching position information and matching length information are output as necessary. If the number of bits necessary to represent the matching position and the matching length is 0, it means that those values are the same as the previous value. In such a case, the matching is the same. Information about position and matching length is not output. The output match position information and match length information are expressed as a bit string excluding the most significant bit when the values indicating the match position and the match length are displayed in binary.

  Therefore, when encoding is performed based on the matching position and the matching length, it is not necessary to always represent the information representing them with a bit string of the maximum fixed length (14 and 8 in the present embodiment), and the above identification Even if information is added, the length of the code as a whole can be shortened on average, and the data compression rate can be improved as compared with the conventional case. Further, in the compression method according to the present embodiment, when the matching position and the matching length values are the same as those in the previous processing, the information indicating them is not output, and information indicating that they are the same. Since the code does not become longer due to the addition, this can further improve the compression rate.

  FIG. 4B schematically shows a code example when the LZSS code is used when the maximum value of the matching position and the matching length is the same as in the present embodiment. If there is no longest matching sequence of a predetermined length or longer in the upper dictionary (“mismatch”), “0” indicating that fact and 8 bits of the original symbol are output. If the longest matching sequence having a predetermined length or longer exists in the lower dictionary (“match”), “1” indicating that, a fixed-length match position information (14 bits), and match length information (8 bits) is output. Compared to the case of the present embodiment shown in FIG. 4A, the data length is the same in the case of “mismatch”, but in the case of “match”, the present embodiment example In the case of, the evaluation that the data length becomes shorter on average is obtained.

  Also, when restoring the code, first, the identification information of the same length is interpreted, it is determined whether the encoding based on the matching position and the matching length has been made, and the coding has been made In addition, the matching position information and the matching length information can be easily read from the identification information. In addition, it is possible to know whether or not the value of the matching position and the matching length is the same as the value at the previous processing by the identification information, and since the value at the previous processing is held, it is the same as the previous processing. Even in this case, it is possible to easily obtain the value of the matching position and the matching length. Therefore, the process of reading the code of the coincidence position or the coincidence length from the input buffer is not required during the decoding process, and the process does not take time.

  Note that although the multiplier in the equation (1), that is, the multiplier by which the number of bits of the matching length information is multiplied, is 15, the multiplier may be 16. As a result, the multiplication at the time of encoding and the division at the time of decoding need only be a bit shift, and the compression and decompression processes can be further accelerated. Also in this case, it is not necessary to increase the number of bits of identification information, and the amount of data does not increase.

  Further, the fixed-length identification information in this embodiment may be Huffman encoded. Thereby, the data compression rate can be further increased.

  Next, another embodiment according to the present invention will be described. The configurations of the data compression apparatus and the data restoration apparatus according to this embodiment are the same as those in the above-described embodiment shown in FIGS. In this embodiment, at the time of compression, the match position information and the match length information are each expressed by a fixed-length code (bit string), but the values of the match position and the match length are the same as the values at the previous processing. Is encoded by a method in which a code (first code) indicating whether or not the same is added to the head and information on the same match position and match length is not output. In other words, the code includes a code indicating whether the value of the match position and the match length is the same as the value at the time of the previous processing, and the required fixed-length match position information and match length information (second code). Consists of.

  FIG. 7 is a diagram exemplifying codes after compression processing output according to the present embodiment. (A), (b), (c), and (d) in the figure, respectively, when the value of the matching position and the value of the matching length are not the same as the previous value, only the value of the matching length is the previous value. Indicates the code that is output when the value is the same as the value, only when the value at the match position is the same as the previous value, and when the value at the match position and the value of the match length are the same as the previous value. ing. In each code, the first two bits are codes indicating whether or not the values of the matching position and the matching length are the same as the values at the time of the previous processing, and are the four types (a) to (d). The case is distinguished.

  When decoding data encoded in this way, the data restoration apparatus according to the present embodiment uses the first two bits of the code to match the match position and match length values to the values at the time of the previous process. For values that are the same, obtain the value using the previous value that was held, and for values that are not the same, obtain the value from the fixed-length code that follows the 2 bits. To do.

  Thus, in this embodiment, when the value of the matching position and the matching length is the same as the previous value, such information is not output, and a code added to indicate that fact. Is not so long, it is possible to increase the data compression rate in the data compression method using the matching position and matching length of the longest matching sequence. Moreover, the decompression | restoration of the code | cord | chord compressed by this method is also easy as above-mentioned, and processing does not take time.

  Next, application examples of the data compression method and the decompression method according to the above-described two exemplary embodiments will be described. FIG. 8 is a schematic configuration diagram of a printing system using the data compression method and the decompression method. As shown in FIG. 8, the printing system includes a host computer 3 and a printer 4. A print request is issued from the application 31 of the host computer 3, and the printer driver 32 that receives the print request generates print data and transmits it to the printer 4. In the printer 4, the controller 41 receives the print data, performs predetermined processing, and sends the data to the engine 42. The engine 42 executes printing on a print medium based on the data.

  The printing system is a so-called host-based system and performs up to halftone processing (screen processing) on the host computer 3 side. Accordingly, as shown in FIG. 8, the printer driver 32 includes an image generation unit 33 that generates image data, a color conversion unit 34 that performs color conversion processing, a halftone processing unit 35 that performs halftone processing, and a halftone. A compression unit 36 is provided for compressing the processed print data. Note that these portions can be configured by a driver program showing the contents of each process and a CPU (control device) that executes the process according to the driver program.

  On the other hand, the print data compressed by the host computer 3 is transmitted to the controller 41 of the printer 4 and is provided with a decompression unit 43 for restoring the print data.

  The compression unit 36 and the decompression unit 43 of the printing system configured as described above use the data compression device 1 and the data restoration device 2 according to one of the above-described two exemplary embodiments, respectively. Accordingly, the compression unit 36 encodes the print data representing the binarized dot image by the method described above. The encoded print data is transmitted to the printer 4 and decoded by the method described in the decompression unit 43.

  In the present printing system, the data compression method and the restoration method according to the above-described embodiment are used in this way, so that the print data compression rate can be increased and the transmission time from the host computer 3 to the printer 4 can be shortened. it can. Accordingly, it is possible to follow the data transmission in the processing by the host computer 3 and the printer 4 which are increased in speed, and the throughput as the printing system can be improved. In particular, since the print data after halftone processing tends to appear repeatedly in the same way as in this application example, it can be said that the present data compression method is suitable for improving the compression rate. For the data after halftone processing, when this compression method is used so that the smaller matching value is selected when the longest matching sequence is the same, the matching position information is an average of 7-8. It has been evaluated that the match length information can be expressed by an average of 2 to 3 bits in bits, and even if the match position and the match length value are not the same as the previous value, it can be said that the average of the entire code is about 19 bits. . Therefore, the data can be shortened as compared with the 23-bit code by the LZSS system (see FIG. 4).

  Further, the control device (CPU) of the printer 4 has a desire to avoid the use of a high-speed control device as compared with the host computer 3 constituted by a personal computer or the like. If the restoration method is used, as described above, the restoration process performed by the printer 4 is facilitated and meets the demand. In addition, in the case of print data after halftone processing, it is suitable for Huffman coding of the identification information, which can further improve the compression rate.

  The protection scope of the present invention is not limited to the above-described embodiment, but covers the invention described in the claims and equivalents thereof.

It is a block diagram concerning the example of an embodiment of a data compression device to which the present invention is applied. It is a block diagram concerning the example of an embodiment of a data restoration device to which the present invention is applied. It is the flowchart which illustrated the content of the process which the encoding part 12 performs. It is a figure for demonstrating the code | symbol etc. which are produced | generated by this data compression apparatus. It is the flowchart which illustrated the content of the process which the decoding part 22 performs. It is the figure which showed the specific example of the process which the data compression apparatus 1 and the data decompression | restoration apparatus 2 perform. It is the figure which illustrated the code | symbol after the compression process output by this Example. It is a schematic block diagram of the printing system using the said data compression method and the decompression | restoration method.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Data compression apparatus, 2 Data decompression | restoration apparatus, 3 Host computer, 4 Printer, 11 Input buffer, 12 Encoding part, 13 Dictionary buffer, 14 Previous value buffer, 21 Input buffer, 22 Decoding part (decoding means), 23 Dictionary buffer (Symbol string storage means), 24 previous value buffer (previous value storage means), 31 application, 32 printer driver, 33 image generation unit, 34 color conversion unit, 35 halftone processing unit, 36 compression unit, 41 controller, 42 engine 43 Defroster

Claims (10)

The longest matching sequence that matches the maximum length with the encoding target symbol sequence that exists in the encoded symbol sequence is searched, and the matching position that is the location of the longest matching sequence and the length of the longest matching sequence are searched. A data compression method for encoding data based on a certain matching length and compressing data,
The code generated by the encoding is
A first code including information on whether the value of the matching position and / or the matching length is the same as the previous encoding value;
A data compression method comprising: the second code representing the coincidence position and / or the coincidence length value which is not identical to the previous encoding value. In claim 1,
The second code is a variable-length code;
The data compression method, wherein the first code includes information on the length of the second code. In claim 2,
The data compression method according to claim 2, wherein the second code is obtained by removing the highest-order 1 when the value representing the matching position and / or the matching length is expressed in binary. The longest matching sequence that matches the maximum length with the encoding target symbol sequence that exists in the encoded symbol sequence is searched, and the matching position that is the location of the longest matching sequence and the length of the longest matching sequence are searched. A data compression method for encoding data based on a certain matching length and compressing data,
When the length of the longest matching sequence is smaller than a predetermined value, the encoding is performed with the first identification information having a predetermined length, representing the fact that it is smaller than the predetermined value and the original symbol,
When the length of the longest matching sequence is equal to or greater than the predetermined value,
It is the same as the value at the time of the previous encoding, and the information whether the value of the matching position and / or the length of the matching is the same as the value at the previous encoding, and the value at the previous encoding. Non-matching position information representing the matching position and / or matching length and / or length of matching length information, second identification information having the same length as the first identification information,
The data compression method, wherein the encoding is performed by using the matching position information and / or matching length information indicating the matching position and / or the matching length which is not the same as the previous encoding value. In claim 4,
Data compression characterized in that the coincidence position information and / or coincidence length information is obtained by removing the most significant 1 when the value representing the coincidence position and / or the coincidence length is expressed in a binary system. Method. In either claim 4 or 5,
When the value of the matching position and / or the matching length is the same as the previous encoding value, the length of the matching position information and / or the matching length information included in the second identification information is set to 0. A data compression method characterized by: In any one of Claims 4 thru | or 6,
The data compression method, wherein the first identification information and / or the second identification information is Huffman encoded. Search for the longest matching sequence that has the longest match with the symbol sequence to be encoded and exists in the encoded symbol sequence. A data compression program for causing a computer to execute a process of encoding and compressing data based on a certain matching length,
Determining whether the value of the matching position and / or the matching length is the same as the previous encoding value;
Based on the result of the determination, outputting a code including information on whether or not the value is the same as the previous encoding value;
After the output, if the value of the match position and / or the match length is not the same as the previous encoding value, outputting a code representing the match position and / or the match length value; A data compression program that is executed by the computer. The longest matching sequence that matches the maximum length of the encoding target symbol string that exists in the encoded symbol string is searched, and the matching position that is the position of the longest matching sequence and the length of the longest matching sequence are searched. A data restoration method for restoring encoded data based on a certain matching length,
The restoration target data includes a first code including information on whether the value of the matching position and / or the matching length is the same as the previous encoding value, and the previous encoding value. And a second code representing the value of the matching position and / or the matching length that are not identical to
Based on the first code, determine whether the value of the matching position and / or the matching length is the same as the previous decoding value,
The match position and / or match length value determined to be the same as the previous decoding value is obtained from the previous decoding value,
A match position and / or match length value determined not to be the same as the previous decoding value is obtained from the second code,
A data restoration method, wherein decoding is performed using a decoded symbol string based on the acquired match position and match length values. The longest matching sequence that matches the maximum length of the encoding target symbol string that exists in the encoded symbol string is searched, and the matching position that is the position of the longest matching sequence and the length of the longest matching sequence are searched. A data restoration device for restoring data encoded based on a certain matching length,
Symbol string storage means for storing the restored symbol string;
Previous value storage means for storing the latest value of the matching position and the matching length obtained at the time of the restoration;
The restoration target data includes a first code including information on whether the value of the matching position and / or the matching length is the same as the previous encoding value, and the previous encoding value. And a second code representing the value of the matching position and / or the matching length that are not identical to
Based on the first code, it is determined whether the value of the matching position and / or the matching length is the same as the value at the previous encoding, and the same as the value at the previous encoding. A match position and / or match length value determined to be present is acquired from a value stored in the previous value storage means, and a match position and / or match value determined not to be the same as the previous encoding value A decoding unit that obtains a length value from the second code, and performs decoding using the symbol string stored in the symbol string storage unit based on the acquired matching position and matching length value A data restoration device comprising:

Images