JP2008098751A - Arithmetic encoding device and arithmetic decoding device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an arithmetic encoding device capable of putting values of a development probability table back to the state right before arithmetic encoding processing. <P>SOLUTION: The arithmetic encoding device includes: a storage unit 17 which stores a table 0 as one of doubled development probability tables and a table 1 as the other of the doubled development probability tables; an initialization unit 14 which initializes the tables 0 and table 1 in slice units; an arithmetic encoding unit 12 which performs arithmetic encoding processing using the table 0; a table management unit 15 which updates the table 0 based upon the result of the arithmetic encoding processing; a decision unit 13 which decides whether arithmetic encoding processing for one macroblock included in a slice each time the arithmetic encoding processing for the macroblock is completed; and a flag management unit 16 which performs control such that the table 1 have values of the table 0 when it is decided that the arithmetic encoding processing is effective, or the table 0 have values of the table 1 otherwise. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an arithmetic coding apparatus and an arithmetic decoding apparatus. The present invention relates to an arithmetic coding apparatus and an arithmetic decoding apparatus using CABAC defined by H.264 / AVC.

  In the moving picture coding technique, the coding efficiency is improved. As a result, it is becoming possible to realize a smooth-moving videophone or to shoot a high-quality moving image on a mobile phone. In the progress of such coding technology, H.B., which is the international standard for the latest video coding technology. H.264 / AVC defines two systems, CAVLC and CABAC, as syntax element entropy encoding systems. The syntax element is information that is determined to be transmitted according to syntax (data string expression rules), such as DCT coefficients and motion vectors.

  CAVLC is an abbreviation for Context-Adaptive Variable Length Coding (context adaptive variable length coding system). In CAVLC, when a DCT coefficient is encoded, a run and a level, which are continuous zero lengths, are encoded from a direction opposite to the scan direction using a variable length encoding table. CABAC is an abbreviation for Context-based Adaptive Binary Arithmac Coding (context adaptive binary arithmetic coding system). In other words, this is a method of changing the occurrence probability of a coding target that changes with time, and is generally called arithmetic coding (see, for example, Patent Document 1). In CABAC, in addition to normal arithmetic coding, a context is assigned to each code to be compressed, and the occurrence probability is changed for each context.

  FIG. 11 is a block diagram of a conventional binary arithmetic coding apparatus using CABAC. This binary arithmetic coding apparatus includes a binarization unit and a context calculation unit in addition to a normal binary arithmetic coding unit. The binarization unit converts a multilevel signal such as +3 or +4 into a binary signal of 0 and 1. The context calculation unit includes a context variable table. The context variable table is a table in which a context number, a table number indicating the occurrence probability of a context, and a symbol (1 or 0) having a high occurrence probability are associated with each other. The context calculation unit refers to the context variable table to calculate the occurrence probability of the binary signal to be encoded according to the surrounding situation, and updates the occurrence probability table based on the actual encoding result. The binary arithmetic encoding unit encodes a binary signal according to the occurrence probability given from the context calculation unit.

  H. According to H.264 / AVC, an occurrence probability table exists for each context as described above. However, for the sake of simplicity, a plurality of occurrence probability tables are considered as one table. That is, a table in which a plurality of contexts and their occurrence probabilities are associated is referred to as an “occurrence probability table”. This occurrence probability table corresponds to the context variable table described above.

  FIG. 12 is a conceptual diagram of binary arithmetic coding using CABAC. As shown in this figure, in binary arithmetic coding using CABAC, a probability of occurrence of a binary signal is given in advance, and an interval between 0.0 and 1.0 is real according to the input signal. Narrowing down more and more. Here, since the probability of occurrence of “0” in the input signal is 0.8 and the probability of occurrence of “1” is 0.2, when “0” in the input signal is encoded, When “1” is encoded in the lower 0.8 part of the section, the upper 0.2 part of the previous section is updated as a new section.

  Specifically, since “0” is encoded in the first STEP 1, 0.0-0.8 corresponding to the lower 0.8 of 0.0-1.0 of the initial section is updated as a new section. Thereafter, “1” and “0” are sequentially encoded, and the final interval is 0.64 to 0.768. The code word of the arithmetic code is a binary representation of a real value that specifies the last interval. In the case of this example, 0.64 to 0.768 is the final section, and therefore, a real number that falls between them can be 0.75. Since the binary representation of 0.75 is 0.11, “11” below the decimal point is output as a code word except for the first digit that is always 0.

  FIG. 13 is a diagram illustrating a state in which a slice is a basic unit of encoding. For example, the processing time T1 for slice # 3 is the sum of time T2 required for initialization processing of the occurrence probability table and time T3 required for encoding processing for slice # 3. As described above, in binary arithmetic encoding using CABAC, encoding processing is performed for each slice, and the occurrence probability table is updated based on the actual encoding result. That is, the entire occurrence probability table is initialized before encoding for one slice.

  FIG. 14 is a diagram illustrating the generated code amount for each macroblock included in one slice. Here, as shown in FIG. 14A, it is assumed that one macro includes eight macro blocks # 1 to # 8. When arithmetic coding processing is performed on such a slice, as shown in FIG. 14B, the amount of code generated by the arithmetic coding processing is detected for each of the eight macro blocks # 1 to # 8. Is done.

Here, the generated code amount for one macroblock is H.264. The limit value specified in H.264 / AVC must not be exceeded. For example, as shown in FIG. 14B, when the generated code amount for the macro block # 5 exceeds the limit value, the stream data of the macro block # 5 part is replaced with PCM (Pulse Code Modulation) data. Alternatively, it is necessary to perform the arithmetic coding process (re-encoding) on the macroblock # 5 again by adding a restriction to suppress the generated code amount.
JP 2004-135251 A

  However, according to the conventional technique, the value of the occurrence probability table cannot be returned to the state immediately before the arithmetic coding process is performed on the macroblock # 5. That is, since the value of the occurrence probability table is updated based on the actual encoding result, when the stream data of the macro block # 5 portion is replaced with the PCM data, the arithmetic encoding process is performed on the macro block # 5. It is necessary to return the value of the occurrence probability table to the state before execution. However, in the prior art, since the value of the occurrence probability table cannot be returned to such a state, the stream data of the macroblock # 5 portion cannot be replaced with PCM data. Even when re-encoding is performed for macro block # 5, the value in the occurrence probability table cannot be returned to the state immediately before the arithmetic coding process is performed for macro block # 5. Therefore, in this case, it is necessary to initialize the occurrence probability table and then return to the top of the slice and perform the arithmetic coding process again. For example, as shown in FIG. 14B, when the generated code amount for the macroblock # 5 exceeds the limit value, the occurrence probability table is initialized, and the arithmetic code is sequentially started from the macroblock # 1. You must redo the process. Such a problem occurs not only when the generated code amount exceeds the limit value but also when the arithmetic coding process for one macroblock is not effective.

  The present invention solves the above-mentioned problem. When the arithmetic coding process for one macroblock is not effective, the value of the occurrence probability table is returned to the state immediately before the arithmetic coding process is performed for the macroblock. It is an object of the present invention to provide an arithmetic coding device capable of performing the above.

  In order to achieve the above object, an arithmetic coding apparatus according to the present invention is an arithmetic coding apparatus that performs an arithmetic coding process using an occurrence probability table, and is one of the duplicated occurrence probability tables. Table storage means for storing a first table that is the second table and a second table that is the other of the duplicated occurrence probability tables, and an initial stage for initializing the first table and the second table in units of slices Included in the slice, an arithmetic encoding unit that performs the arithmetic encoding process using the first table, an update unit that updates the first table based on a result of the arithmetic encoding process, and the slice Determination means for determining whether or not the arithmetic encoding process for the macroblock is valid each time the arithmetic encoding process is completed for one macroblock, and the arithmetic code If it is determined that the encoding process is valid, the value of the second table is controlled to be the value of the first table, and if it is determined that the arithmetic encoding process is not effective, the second table Control means for controlling the value of one table to be the value of the second table. Accordingly, the first table that is one of the duplicated occurrence probability tables is used as the latest table, and the second table that is the other of the duplicated occurrence probability tables is used as a backup table. Therefore, when the arithmetic coding process for one macroblock is not effective, the value of the occurrence probability table can be returned to the state immediately before the arithmetic coding process is performed for the macroblock.

  Here, when the determination unit determines that the arithmetic coding process is not effective, the stream data of the macroblock portion may be replaced with PCM data. As a result, it is possible to replace the stream data of the macro block portion where the arithmetic coding process is not effective with the PCM data.

  Further, the arithmetic coding means may perform the arithmetic coding process for the macroblock again by adding a constraint for suppressing the generated code amount when it is determined that the arithmetic coding process is not effective. As a result, when re-encoding is performed, it is possible to reduce the time for initializing the occurrence probability table and the time for redoing the arithmetic coding process for the macroblock for which the arithmetic coding process has been effective.

  The arithmetic coding apparatus further stores a first flag indicating the first table or the second table in association with a second flag for each entry of the first table and the second table. Flag storing means, the arithmetic encoding means uses a table indicated by the second flag, the updating means updates a table not indicated by the first flag, and the control means includes the first flag When the table not indicated by the flag is updated, the value of the second flag is changed to indicate the table not indicated by the first flag, and when it is determined that the arithmetic encoding process is valid, If the value of the first flag is copied to the second flag and it is determined that the arithmetic coding process is not valid, the value of the second flag is set to the first flag. It may be replicated to. As a result, the table indicated by the first flag is used as a backup table, and the table indicated by the second flag is used as the latest table. Therefore, it is possible to easily switch between the backup table and the latest table simply by changing the flag value.

  The determination unit determines that the arithmetic encoding process is effective when the code amount generated by the arithmetic encoding process does not exceed a predetermined amount, and the code amount generated by the arithmetic encoding process. May exceed the predetermined amount, it may be determined that the arithmetic coding process is not effective. As a result, when the amount of code generated by the arithmetic coding process for one macroblock exceeds a predetermined amount, the value of the occurrence probability table can be returned to the state immediately before the arithmetic coding process is performed for that macroblock. It becomes possible.

  In order to achieve the above object, an arithmetic decoding apparatus according to the present invention is an arithmetic decoding apparatus that performs an arithmetic decoding process using an occurrence probability table, and includes one of the occurrence probability tables that are duplicated. Table storage means for storing a first table that is the second table and a second table that is the other of the duplicated occurrence probability tables, and an initial stage for initializing the first table and the second table in units of slices Included in the slice, an arithmetic decoding unit that performs the arithmetic decoding process using the first table, an updating unit that updates the first table based on a result of the arithmetic decoding process, Determination means for determining whether or not the arithmetic decoding process for the macroblock is valid each time the arithmetic decoding process is completed for the one macroblock; If it is determined that the conversion process is effective, the value of the second table is controlled to be the value of the first table, and if it is determined that the arithmetic decoding process is not effective, the second table Control means for controlling the value of one table to be the value of the second table. Accordingly, the first table that is one of the duplicated occurrence probability tables is used as the latest table, and the second table that is the other of the duplicated occurrence probability tables is used as a backup table. Therefore, when the arithmetic decoding process for one macroblock is not effective, the value of the occurrence probability table can be returned to the state immediately before the arithmetic decoding process is performed for the macroblock.

  Here, when it is determined that the arithmetic decoding process is not valid, the determination unit may replace the stream data of the macroblock portion with PCM data. As a result, it is possible to replace the stream data of the macro block portion where the arithmetic decoding process is not effective with the PCM data.

  In addition, when it is determined that the arithmetic decoding process is not effective, the arithmetic decoding unit may perform the arithmetic decoding process on the macroblock again by adding a constraint that suppresses the generated code amount. As a result, when re-encoding is performed, it is possible to reduce the time for initializing the occurrence probability table and the time for redoing the arithmetic decoding process for the macroblock for which the arithmetic decoding process is effective.

  The arithmetic decoding apparatus further stores a first flag indicating the first table or the second table and a second flag in association with each other for each entry of the first table and the second table. Flag storage means, the arithmetic decoding means uses a table indicated by the second flag, the updating means updates a table not indicated by the first flag, and the control means includes the first flag When the table not indicated by the flag is updated, the value of the second flag is changed to indicate the table not indicated by the first flag, and when it is determined that the arithmetic decoding process is valid, If the value of the first flag is copied to the second flag and it is determined that the arithmetic decoding process is not valid, the value of the second flag is set to the first flag. It may be replicated to. As a result, the table indicated by the first flag is used as a backup table, and the table indicated by the second flag is used as the latest table. Therefore, it is possible to easily switch between the backup table and the latest table simply by changing the flag value.

  The determination means determines that the arithmetic decoding process is valid when the code amount generated by the arithmetic decoding process does not exceed a predetermined amount, and the code amount generated by the arithmetic decoding process. May exceed the predetermined amount, it may be determined that the arithmetic decoding process is not effective. As a result, when the amount of codes generated by the arithmetic decoding process for one macroblock exceeds a predetermined amount, the value of the occurrence probability table can be returned to the state immediately before the arithmetic decoding process is performed for that macroblock. It becomes possible.

  Note that the present invention can be realized not only as such an arithmetic coding device or arithmetic decoding device, but also as an arithmetic step including steps characteristic of the arithmetic coding device or arithmetic decoding device. It can also be realized as an encoding method or an arithmetic decoding method, or as a program for causing a computer to execute these steps. Needless to say, such a program can be distributed via a recording medium such as a CD-ROM or a transmission medium such as the Internet.

  As is clear from the above description, according to the arithmetic coding apparatus of the present invention, the first table that is one of the occurrence probability tables that are duplicated is the latest table and the occurrence that is duplicated. Since the second table, which is the other of the probability tables, is used as a backup table, if the arithmetic coding process for one macroblock is not valid, the immediately preceding arithmetic coding process for that macroblock is performed. It becomes possible to return the value of the occurrence probability table to the state. In addition, if a configuration in which control is performed using two flags is employed, it is possible to easily switch between the backup table and the latest table simply by changing the value of the flag.

  As a result, it is possible to replace the stream data of the macro block portion where the arithmetic coding process is not effective with the PCM data. Further, when re-encoding is performed, it is possible to reduce the time for initializing the occurrence probability table and the time for performing the arithmetic encoding process again for a macroblock for which the arithmetic encoding process is effective. When the processing time is thus reduced, the power consumption is also reduced. Therefore, it is effective to apply the present invention to an apparatus that needs to suppress the power consumption.

  Similarly, according to the arithmetic decoding apparatus of the present invention, the first table that is one of the duplicated occurrence probability tables is the latest table and the other one of the duplicated occurrence probability tables. Since the second table is used as a backup table, if the arithmetic decoding process for one macroblock is not valid, the occurrence probability table of the occurrence table immediately before the arithmetic decoding process for the macroblock is set. The value can be returned. In addition, if a configuration in which control is performed using two flags is employed, it is possible to easily switch between the backup table and the latest table simply by changing the value of the flag.

  As a result, it is possible to replace the stream data of the macro block portion where the arithmetic decoding process is not effective with the PCM data. Also, when re-encoding is performed, it is possible to reduce the time for initializing the occurrence probability table and the time for performing the arithmetic decoding process again for the macroblocks for which the arithmetic decoding process is effective. When the processing time is thus reduced, the power consumption is also reduced. Therefore, it is effective to apply the present invention to an apparatus that needs to suppress the power consumption.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(Embodiment 1)
FIG. 1 is a diagram for explaining the outline of the present invention. As already described, according to the prior art, when the generated code amount for the macroblock # 5 exceeds the limit value, the value of the occurrence probability table is in a state immediately before the arithmetic coding process is performed for the macroblock # 5. Therefore, the stream data of the macro block # 5 cannot be replaced with PCM data. Further, when re-encoding is performed for the macroblock # 5, as shown in FIG. 14B, after the occurrence probability table is initialized, the arithmetic coding process must be performed again in order from the macroblock # 1. . On the other hand, in the present invention, as shown in FIG. 1B, even when the generated code amount for the macroblock # 5 exceeds the limit value, the arithmetic coding process is performed for the macroblock # 5. The value in the occurrence probability table can be returned to the previous state.

  FIG. 2 is a configuration diagram of the arithmetic coding apparatus 10 according to Embodiment 1 of the present invention. The arithmetic coding device 10 is a device that performs arithmetic coding processing using an occurrence probability table, and includes an analysis unit 11, an arithmetic coding unit 12, a determination unit 13, an initialization unit 14, and a table. A management unit 15, a flag management unit 16, and a storage unit 17 are provided. Here, illustration of the binarization unit is omitted. The binarization unit may be in the preceding stage of the analysis unit 11, and may be in the arithmetic encoding device 10 or outside the arithmetic encoding device 10.

  The analysis unit 11 analyzes the input signal from the binarization unit. When the head of the slice appears in the input signal, the fact is output to the initialization unit 14, and when the end of the macroblock appears in the input signal, the fact is outputted to the flag management unit 16. The input signal from the binarization unit is input to the arithmetic encoding unit 12 via the analysis unit 11 (or directly without going through the analysis unit 11).

  The storage unit 17 is an example of a table storage unit and a flag storage unit according to the present invention, and is, for example, a RAM (Random Access Memory). The storage unit 17 stores a first flag and a second flag in association with each entry of the table 0 and the table 1. The table 0 is an example of the first table or the second table according to the present invention, and is one of the duplicate occurrence probability tables. Table 1 is an example of the first table or the second table according to the present invention, and is the other of the duplicate occurrence probability tables. The first flag is a flag indicating the table 0 or the table 1. If the value of the first flag is 0, the first flag indicates table 0, and if the value of the first flag is 1, the first flag indicates table 1 Become. Similarly, the second flag is a flag indicating the table 0 or the table 1. When the value of the second flag is 0, the second flag indicates table 0, and when the value of the second flag is 1, the second flag indicates table 1 Become.

  The initialization unit 14 is an example of an initialization unit according to the present invention, and initializes table 0 and table 1 in units of slices. The function of the initialization unit 14 is the same as the conventional one except that the duplicated table 0 and table 1 are initialized. When the initialization unit 14 initializes the table 0 and the table 1, the table management unit 15 is used.

  The arithmetic encoding unit 12 is an example of arithmetic encoding means according to the present invention, and performs arithmetic encoding processing using a table indicated by the second flag. For example, when the value of the second flag is 0, the table 0 is used, and when the value of the second flag is 1, the table 1 is used. The function of the arithmetic encoding unit 12 is the same as the conventional one except that the arithmetic encoding process is performed using the table 0 or the table 1. Note that when the arithmetic coding unit 12 uses the table 0 or the table 1, the table management unit 15 is used.

  The table management unit 15 is an example of an updating unit according to the present invention, and updates a table not indicated by the first flag, in other words, a table opposite to the table indicated by the first flag based on the result of the arithmetic encoding process. To do. For example, when the value of the first flag is 0, the table 1 is updated, and when the value of the first flag is 1, the table 0 is updated. The function of the table management unit 15 is the same as the conventional one except that the table 0 or the table 1 is updated.

  The determination unit 13 is an example of a determination unit according to the present invention, and whether or not the arithmetic encoding process for the macroblock is valid each time the arithmetic encoding process is completed for one macroblock included in the slice. Determine. For example, if the code amount generated by the arithmetic encoding process does not exceed a predetermined amount, it is determined that the arithmetic encoding process for the macroblock is valid, and the code amount generated by the arithmetic encoding process is a predetermined amount. Is exceeded, it is determined that the arithmetic coding process for the macroblock is not effective. The predetermined amount here is H.264. It means a limit value defined in H.264 / AVC. The determination unit 13 outputs a determination result signal indicating whether or not the generated code amount exceeds the limit value to the flag management unit 16. In addition, when a discard instruction signal described later is received from the flag management unit 16, the result of the arithmetic coding process for the macroblock is discarded.

  The flag management unit 16 is an example of a control unit according to the present invention, and when a table not indicated by the first flag is updated, the value of the second flag is changed to indicate a table not indicated by the first flag. . Further, when it is determined that the arithmetic encoding process is effective, such as when a determination result signal indicating that the generated code amount does not exceed the limit value is received from the determination unit 13, the value of the first flag is set to the second value. Duplicate the flag. On the other hand, when it is determined that the arithmetic encoding process is not effective, such as when a determination result signal indicating that the generated code amount exceeds the limit value is received from the determination unit 13, the value of the second flag is set to the first flag. Duplicate to. When the value of the first flag is copied to the second flag in this way, a signal (discard instruction signal) to discard the result of the arithmetic coding process for the macroblock is output to the determination unit 13.

  FIG. 3 is a diagram showing a context arrangement in the occurrence probability table. The occurrence probability table is a table in which contexts and their occurrence probabilities are associated with each other. In the present invention, the occurrence probability table is duplicated, one is called table 0 and the other is called table 1. The characters shown in the figure indicate syntax elements, and one record in the figure corresponds to one bit. “Mb_skip_run”, “mb_skip_flag”, and the like are 1-bit syntax elements and are therefore described in one record, and “mb_type” is a 4-bit syntax element and are described in 4 records. Since the context (entry) is information corresponding to each code constituting the syntax element, one record in the figure corresponds to one context. Note that the arrangement of contexts shown in FIG. 3 is merely an example, and is not actually such an arrangement.

  FIG. 4 is a flowchart showing a procedure of processing performed by the arithmetic coding apparatus 10. The operation for one slice, which is the basic unit of arithmetic coding, will be described below.

  First, when the head of a slice appears in the input signal, the initialization unit 14 initializes table 0 and table 1 (step S1). Thereafter, the processes in steps S2 to S8 are repeated for the number of macroblocks in the slice.

  That is, when it becomes necessary to refer to (use) the occurrence probability table, the arithmetic encoding unit 12 refers to the table indicated by the second flag (step S2). Then, an arithmetic coding process is performed using the referenced occurrence probability value (step S3), and a table not indicated by the first flag is updated based on the result of the arithmetic coding process (step S4). As described above, the table reference (step S2) and the table update (step S4) are performed via the table management unit 15. Thereafter, the flag management unit 16 changes the value of the second flag so as to indicate a table not indicated by the first flag (step S5).

  When the processes in steps S2 to S5 are repeated for all the context numbers, the determination unit 13 determines whether or not the generated code amount for the macroblock exceeds the limit value (step S6). If it is determined that the generated code amount does not exceed the limit value, the flag management unit 16 copies the value of the second flag to the first flag (step S7). On the other hand, when it is determined that the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag (step S8).

When the above processing is completed for all slices, the entire processing is completed.
When it is determined that the generated code amount exceeds the limit value (step S6), the determination unit 13 replaces the stream data of the macroblock portion with PCM data. As a result, it is possible to replace the stream data of the macro block portion where the generated code amount exceeds the limit value with the PCM data. When it is determined that the generated code amount exceeds the limit value (step S6), the arithmetic encoding unit 12 adds a constraint to suppress the generated code amount, and the arithmetic encoding unit 12 re-executes the macroblock. Encoding may be performed. In this case, it is possible to reduce the time for initializing the occurrence probability table and the time for redoing the arithmetic coding process for the macroblock whose generated code amount did not exceed the limit value.

  Here, the table indicated by the first flag can be considered as an occurrence probability table immediately before the arithmetic coding process is performed on the macroblock being subjected to the arithmetic coding process, that is, a backup table. The table indicated by the second flag can be considered as an occurrence probability table updated during the arithmetic encoding process, that is, the latest table.

  Therefore, when referring to the table during the arithmetic encoding process, the latest table, that is, the table indicated by the second flag is referred to (step S2). When updating the table during the arithmetic coding process, a table that is not a backup table, that is, a table not indicated by the first flag is updated (step S4).

  When a table that is not a backup table is updated in this way, the value of the second flag is changed to indicate a table that is not a backup table (step S5). This is because if the value of the second flag is changed to indicate a table that is not a backup table, the table that is not a backup table can be considered as the latest table thereafter.

  If the generated code amount does not exceed the limit value, the value of the second flag is copied to the first flag (step S7). This is because if the value of the second flag is copied to the first flag, the same effect as that obtained when the latest table value is copied to the backup table can be obtained. On the other hand, if the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag (step S8). This is because if the value of the first flag is duplicated to the second flag, the same effect as when the value of the backup table is duplicated to the latest table can be obtained.

  FIG. 5 is a diagram for explaining the table reference and the table update in detail. Hereinafter, the flag value before the table update, the table reference update mode, and the flag value after the table update will be described in detail. The combination of the value of the first flag and the value of the second flag before the table update is expressed as (0, 0) (0, 1) (1, 0) There are four types of (1, 1). Therefore, here, the case where the combination of the value of the first flag and the value of the second flag before the table update is (0, 0) is case 1, the case of (0, 1) is case 2, and (1, 0). Case 3 is referred to as case 3, and case (1,1) is referred to as case 4.

  In case 1, the values of the first flag and the second flag before the table update are (0, 0). The value of the first flag being 0 means that the backup table is table 0. In addition, the value of the second flag being 0 means that the latest table is table 0. Therefore, when referring to the table during the arithmetic encoding process, the latest table, that is, the table 0 is referred to. When updating the table during the arithmetic encoding process, the table that is not a backup table, that is, the table 1 is updated. When the table 1 is updated in this way, the value of the second flag is changed so as to indicate the table 1. As a result, the value of the first flag and the value of the second flag after the table update are (0, 1).

  In Case 2, the values of the first flag and the second flag before the table update are (0, 1). The value of the first flag being 0 means that the backup table is table 0. Further, the value of the second flag being 1 means that the latest table is table 1. Therefore, when referring to the table during the arithmetic encoding process, the latest table, that is, the table 1 is referred to. When updating the table during the arithmetic encoding process, the table that is not a backup table, that is, the table 1 is updated. When the table 1 is updated in this way, the value of the second flag is changed so as to indicate the table 1. As a result, the value of the first flag and the value of the second flag after the table update are (0, 1). In Case 2, since the value of the second flag is originally 1, the value of the second flag is not substantially changed.

  In Case 3, the value of the first flag and the value of the second flag before the table update are (1, 0). That the value of the first flag is 1 means that the backup table is table 1. In addition, the value of the second flag being 0 means that the latest table is table 0. Therefore, when referring to the table during the arithmetic encoding process, the latest table, that is, the table 0 is referred to. When updating the table during the arithmetic coding process, a table that is not a backup table, that is, table 0 is updated. When the table 0 is updated in this way, the value of the second flag is changed so as to indicate the table 0. As a result, the value of the first flag and the value of the second flag after the table update are (1, 0). In Case 3, since the value of the second flag is originally 0, the value of the second flag is not substantially changed.

  In Case 4, the values of the first flag and the second flag before the table update are (1, 1). That the value of the first flag is 1 means that the backup table is table 1. Further, the value of the second flag being 1 means that the latest table is table 1. Therefore, when referring to the table during the arithmetic encoding process, the latest table, that is, the table 1 is referred to. When updating the table during the arithmetic coding process, a table that is not a backup table, that is, table 0 is updated. When the table 0 is updated in this way, the value of the second flag is changed so as to indicate the table 0. As a result, the value of the first flag and the value of the second flag after the table update are (1, 0).

  FIG. 6 is a diagram for explaining in detail the process of copying the flag value. As described above, when it is determined that the generated code amount does not exceed the limit value, the value of the second flag is copied to the first flag, and it is determined that the generated code amount exceeds the limit value The value of the first flag is copied to the second flag. In FIG. 6A, it is assumed that the value of the second flag is duplicated in the first flag, and the values of the first flag and the second flag before and after the duplication are shown. In FIG. 6B, it is assumed that the value of the first flag is duplicated to the second flag, and the values of the first flag and the second flag before and after the duplication are shown. There are four combinations (0, 0) (0, 1) (1, 0) (1, 1) as combinations of the value of the first flag and the value of the second flag before duplication. The case where the combination is (0, 0) is referred to as case 1, the case where (0, 1) is referred to as case 2, the case where (1, 0) is referred to as case 3, and the case where (1, 1) is referred to as case 4.

First, the case where the value of the second flag is copied to the first flag will be described with reference to FIG.
In case 1, the value of the first flag and the value of the second flag before duplication are (0, 0). The value of the first flag being 0 means that the backup table is table 0. In addition, the value of the second flag being 0 means that the latest table is table 0. When it is determined that the generated code amount does not exceed the limit value in this state, the value of the second flag is copied to the first flag. As a result, the value of the first flag and the value of the second flag after copying are (0, 0). In Case 1, since the value of the first flag is originally 0, the value of the first flag is not substantially changed. Thereafter, the next macroblock is processed with the table 0 as the backup table and the table 0 as the latest table.

  In Case 2, the value of the first flag and the value of the second flag before duplication are (0, 1). The value of the first flag being 0 means that the backup table is table 0. Further, the value of the second flag being 1 means that the latest table is table 1. When it is determined that the generated code amount does not exceed the limit value in this state, the value of the second flag is copied to the first flag. As a result, the value of the first flag and the value of the second flag after duplication are (1, 1). Thereafter, the next macroblock is processed with the table 1 as a new backup table and the table 1 as the latest table.

  In Case 3, the value of the first flag and the value of the second flag before duplication are (1, 0). That the value of the first flag is 1 means that the backup table is table 1. In addition, the value of the second flag being 0 means that the latest table is table 0. When it is determined that the generated code amount does not exceed the limit value in this state, the value of the second flag is copied to the first flag. As a result, the value of the first flag and the value of the second flag after copying are (0, 0). Thereafter, the next macroblock is processed with the table 0 as a new backup table and the table 0 as the latest table.

  In Case 4, the value of the first flag and the value of the second flag before duplication are (1, 1). That the value of the first flag is 1 means that the backup table is table 1. Further, the value of the second flag being 1 means that the latest table is table 1. When it is determined that the generated code amount does not exceed the limit value in this state, the value of the second flag is copied to the first flag. As a result, the value of the first flag and the value of the second flag after duplication are (1, 1). In Case 4, since the value of the first flag is originally 1, the value of the first flag is not substantially changed. Thereafter, the next macroblock is processed with the table 1 as the backup table and the table 1 as the latest table.

Next, a case where the value of the first flag is copied to the second flag will be described with reference to FIG.
In case 1, the value of the first flag and the value of the second flag before duplication are (0, 0). The value of the first flag being 0 means that the backup table is table 0. In addition, the value of the second flag being 0 means that the latest table is table 0. In this state, when it is determined that the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag. As a result, the value of the first flag and the value of the second flag after copying are (0, 0). In Case 1, since the value of the second flag is originally 0, the value of the second flag is not substantially changed. Thereafter, the next macroblock is processed with the table 0 as the backup table and the table 0 as the latest table.

  In Case 2, the value of the first flag and the value of the second flag before duplication are (0, 1). The value of the first flag being 0 means that the backup table is table 0. Further, the value of the second flag being 1 means that the latest table is table 1. In this state, when it is determined that the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag. As a result, the value of the first flag and the value of the second flag after copying are (0, 0). Subsequently, the next macroblock is processed with the table 0 as the backup table and the new table 0 as the latest table.

  In Case 3, the value of the first flag and the value of the second flag before duplication are (1, 0). That the value of the first flag is 1 means that the backup table is table 1. In addition, the value of the second flag being 0 means that the latest table is table 0. In this state, when it is determined that the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag. As a result, the value of the first flag and the value of the second flag after duplication are (1, 1). Subsequently, the next macroblock is processed with the table 1 as the backup table and the new table 1 as the latest table.

  In Case 4, the value of the first flag and the value of the second flag before duplication are (1, 1). That the value of the first flag is 1 means that the backup table is table 1. Further, the value of the second flag being 1 means that the latest table is table 1. In this state, when it is determined that the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag. As a result, the value of the first flag and the value of the second flag after duplication are (1, 1). In this case 4, since the value of the second flag is originally 1, the value of the second flag is not substantially changed. Thereafter, the next macroblock is processed with the table 1 as the backup table and the table 1 as the latest table.

  FIG. 7 is a diagram showing state transitions of flags and tables. Here, the state transition when it is determined that the generated code amount does not exceed the limit value will be described.

  FIG. 7A shows a state immediately after the table 0 and the table 1 are initialized. In the state immediately after the table 0 and the table 1 are initialized, the value of the first flag and the value of the second flag are (0, 0). This case corresponds to case 1 in FIG. That is, the value of the first flag being 0 means that the backup table is table 0. In addition, the value of the second flag being 0 means that the latest table is table 0. Although not shown here, the same occurrence probability value (initial value) is set to the same entry in Table 0 and Table 1.

  FIG. 7B shows a state immediately after table 0 and table 1 are initialized and table reference and table update are performed. In a state immediately after the table 0 and the table 1 are initialized, the backup table is the table 0, and the latest table is the table 0. Therefore, as described in case 1 of FIG. 5, when referring to the table during the arithmetic encoding process, the latest table, that is, table 0 is referred to. When updating the table during the arithmetic encoding process, the table that is not a backup table, that is, the table 1 is updated. When the table 1 is updated in this way, the value of the second flag is changed so as to indicate the table 1. As a result, the value of the first flag and the value of the second flag after the table update are (0, 1).

  Here, the state in which the top entry is updated as a result of referring to the top entry is shown, but the entry to be referred to and the entry to be updated are not necessarily the same. For example, as a result of referring to the top entry, the top entry and the second entry from the top may be updated. Such table reference and table update are repeated by the number of all contexts. Here, for the sake of simplicity, it is assumed that the table reference and table update are performed once.

  FIG. 7C shows a state immediately after it is determined that the generated code amount does not exceed the limit value after table reference and table update are performed. In a state immediately after the table reference and the table update are performed, the backup table is the table 0, and the latest table is the table 1. This case corresponds to Case 2 in FIG. That is, when it is determined that the generated code amount does not exceed the limit value, the value of the second flag is copied to the first flag. As a result, the value of the first flag and the value of the second flag after duplication are (1, 1). Thereafter, the next macroblock is processed with the table 1 as a new backup table and the table 1 as the latest table.

  FIG. 7D shows a state immediately after the table reference and the table update are performed after the value of the second flag is copied to the first flag. In a state immediately after the value of the second flag is copied to the first flag, the backup table is table 1, and the latest table is table 1. Therefore, as described in case 4 of FIG. 5, when the table is referred to during the arithmetic encoding process, the latest table, that is, the table 1 is referred to. When updating the table during the arithmetic coding process, a table that is not a backup table, that is, table 0 is updated. When the table 0 is updated in this way, the value of the second flag is changed so as to indicate the table 0. As a result, the value of the first flag and the value of the second flag after the table update are (1, 0).

  FIG. 8 is a diagram showing state transitions of flags and tables. Here, the state transition when it is determined that the generated code amount exceeds the limit value will be described.

  The contents of FIG. 8A are the same as those of FIG. That is, it shows a state immediately after the table 0 and the table 1 are initialized. The value of the first flag being 0 means that the backup table is table 0. In addition, the value of the second flag being 0 means that the latest table is table 0.

  The content of FIG. 8 (B) is the same as FIG. 7 (B). That is, when referring to the table during the arithmetic encoding process, the latest table, that is, the table 0 is referred to. When updating the table during the arithmetic encoding process, the table that is not a backup table, that is, the table 1 is updated. When the table 1 is updated in this way, the value of the second flag is changed so as to indicate the table 1. As a result, the value of the first flag and the value of the second flag after the table update are (0, 1).

  FIG. 8C shows a state immediately after it is determined that the generated code amount exceeds the limit value after table reference and table update are performed. In a state immediately after the table reference and the table update are performed, the backup table is the table 0, and the latest table is the table 1. This case corresponds to Case 2 in FIG. That is, when it is determined that the generated code amount exceeds the limit value, the value of the first flag is copied to the second flag. As a result, the value of the first flag and the value of the second flag after copying are (0, 0). Subsequently, the next macroblock is processed with the table 0 as the backup table and the new table 0 as the latest table.

  FIG. 8D shows a state immediately after the table reference and the table update are performed after the value of the first flag is copied to the second flag. In a state immediately after the value of the first flag is copied to the second flag, the backup table is table 0, and the latest table is table 0. Therefore, as described in case 1 of FIG. 5, when referring to the table during the arithmetic encoding process, the latest table, that is, table 0 is referred to. When updating the table during the arithmetic encoding process, the table that is not a backup table, that is, the table 1 is updated. When the table 1 is updated in this way, the value of the second flag is changed so as to indicate the table 1. As a result, the value of the first flag and the value of the second flag after the table update are (0, 1).

  As described above, according to the arithmetic coding apparatus according to the present invention, the table indicated by the first flag is used as a backup table, and the table indicated by the second flag is used as the latest table. Therefore, if the amount of code generated by the arithmetic coding process for one macroblock exceeds the limit value, the value of the occurrence probability table may be returned to the state immediately before the arithmetic coding process is performed for that macroblock. It becomes possible. In addition, since the control is performed using two flags, it is possible to easily switch between the backup table and the latest table simply by changing the value of the flag.

  As a result, it is possible to replace the stream data of the macro block portion in which the generated code amount exceeds the limit value with the PCM data. Also, when re-encoding is performed, it is possible to reduce the time for initializing the occurrence probability table and the time for performing the arithmetic coding process again for a macroblock whose generated code amount does not exceed the limit value. When the processing time is thus reduced, the power consumption is also reduced. Therefore, it is effective to apply the present invention to an apparatus that needs to suppress the power consumption.

  Here, the control is performed using the two flags of the first flag and the second flag, but the present invention is not limited to this. That is, if a configuration in which the table 0 is fixedly used as the latest table and the table 1 is fixedly used as a backup table, the same effect can be obtained without using a flag. However, when such a configuration is adopted, instead of the process of mutually copying the flag values between the first flag and the second flag, the table values are mutually exchanged between the table 0 and the table 1. It is necessary to copy the file. That is, when the generated code amount exceeds the limit value, it is necessary to copy the value of the table 1 as the backup table to the table 0 as the latest table. On the other hand, when the generated code amount does not exceed the limit value, it is necessary to copy the value of the table 0 as the latest table to the table 1 as the backup table. Needless to say, it is more preferable in terms of processing speed to duplicate the flag value than to duplicate the table value.

  Further, here, it has been described that the effect of the present invention can be obtained when the generated code amount exceeds the limit value, but the present invention is not limited to this. That is, if the arithmetic coding process for one macroblock is not effective, the same effect as described above can be obtained.

  In FIG. 4, step S5 for changing the value of the second flag is repeated for all the context numbers, but the present invention is not limited to this. That is, as shown in FIG. 9, the timing at which the value of the second flag is changed is only immediately after the determination process is performed and immediately after the table update is performed after the determination process is performed. The reason why the value of the second flag changes immediately after the determination process is performed is due to duplication in step S8 and is not related to step S5. Therefore, step S5 for changing the value of the second flag may be executed only immediately after the table update is performed after the determination process is performed.

(Embodiment 2)
In the first embodiment, the case where the present invention is applied to an arithmetic coding apparatus that encodes data using CABAC has been described. In the second embodiment, data encoded using CABAC is decoded. A case will be described in which the present invention is applied to an arithmetic decoding device to be realized.

  FIG. 10 is a configuration diagram of the arithmetic decoding device 20 according to the embodiment of the present invention. The arithmetic decoding device 20 is a device that performs arithmetic decoding processing using an occurrence probability table, and includes an analysis unit 21, an arithmetic decoding unit 22, a determination unit 23, an initialization unit 24, and a table. A management unit 25, a flag management unit 26, and a storage unit 27 are provided. As will be described below, the configuration of the present arithmetic decoding device 20 is basically the same as that of the arithmetic coding device 10 in the first embodiment except that arithmetic coding or arithmetic decoding is performed. It is.

  The analysis unit 21 analyzes the input signal that has been arithmetically encoded. When the head of the slice appears in the input signal, the fact is output to the initialization unit 24, and when the end of the macroblock appears in the input signal, the fact is outputted to the flag management unit 26. The arithmetically encoded input signal is input to the arithmetic decoding unit 22 via the analysis unit 21 (or directly without going through the analysis unit 21).

  The storage unit 27 is an example of a table storage unit and a flag storage unit according to the present invention, and is a RAM, for example. The storage unit 27 stores a first flag and a second flag in association with each entry of the table 0 and the table 1.

  The initialization unit 24 is an example of an initialization unit according to the present invention, and initializes table 0 and table 1 in units of slices. When the initialization unit 24 initializes the table 0 and the table 1, the table management unit 25 is used.

  The arithmetic decoding unit 22 is an example of the arithmetic decoding unit according to the present invention, and performs arithmetic decoding using the table indicated by the second flag. In addition, when the arithmetic decoding unit 22 uses the table 0 or the table 1, it goes through the table management unit 25.

  The table management unit 25 is an example of an updating unit according to the present invention, and updates a table not indicated by the first flag, in other words, a table opposite to the table indicated by the first flag based on the result of the arithmetic decoding process. To do.

  The determination unit 23 is an example of a determination unit according to the present invention, and whether or not the arithmetic decoding process for the macroblock is valid each time the arithmetic decoding process is completed for one macroblock included in the slice. Determine. For example, if the code amount generated by the arithmetic decoding process does not exceed a predetermined amount, it is determined that the arithmetic decoding process for the macroblock is valid, and the code amount generated by the arithmetic decoding process is a predetermined amount. Is exceeded, it is determined that the arithmetic decoding process for the macroblock is not valid. The determination unit 23 outputs a determination result signal indicating whether or not the generated code amount exceeds the limit value to the flag management unit 26. When the discard instruction signal is received from the flag management unit 26, the result of the arithmetic decoding process for the macroblock is discarded.

  The flag management unit 26 is an example of a control unit according to the present invention, and when a table not indicated by the first flag is updated, the value of the second flag is changed to indicate a table not indicated by the first flag. . Also, when it is determined that the arithmetic decoding process is valid, such as when a determination result signal indicating that the generated code amount does not exceed the limit value is received from the determination unit 23, the value of the first flag is set to the second value. Duplicate the flag. On the other hand, when it is determined that the arithmetic decoding process is not effective, such as when a determination result signal indicating that the generated code amount exceeds the limit value is received from the determination unit 23, the value of the second flag is set to the first flag. Duplicate to. When the value of the first flag is copied to the second flag in this way, a discard instruction signal for the macroblock is output to the determination unit 23.

  As described above, according to the arithmetic decoding apparatus according to the present invention, the first table that is one of the duplicated occurrence probability tables is the latest table, and the other one of the duplicated occurrence probability tables. Since the second table is used as a backup table, if the arithmetic decoding process for one macroblock is not effective, the occurrence probability immediately before the arithmetic decoding process is performed for that macroblock. It is possible to return the value of the table. In addition, since the control is performed using two flags, it is possible to easily switch between the backup table and the latest table simply by changing the value of the flag.

  As a result, it is possible to replace the stream data of the macro block portion where the arithmetic decoding process is not effective with the PCM data. Also, when re-encoding is performed, it is possible to reduce the time for initializing the occurrence probability table and the time for performing the arithmetic decoding process again for the macroblocks for which the arithmetic decoding process is effective. When the processing time is thus reduced, the power consumption is also reduced. Therefore, it is effective to apply the present invention to an apparatus that needs to suppress the power consumption.

  The present invention relates to an arithmetic coding apparatus or the like that needs to return the value of the occurrence probability table to a state immediately before performing the arithmetic coding process for the macroblock when the arithmetic coding process for one macroblock is not effective. It can be applied for use.

The figure for demonstrating the outline | summary of this invention Configuration diagram of arithmetic coding apparatus according to Embodiment 1 of the present invention The figure which shows the context arrangement in the occurrence probability table The flowchart which shows the procedure of the process which an arithmetic coding apparatus performs Diagram for explaining table reference and table update in detail Diagram for explaining in detail the process of copying the flag value Diagram showing state transition of flags and table Diagram showing state transition of flags and table The figure which shows the transition of the combination of the value of a 1st flag, and the value of a 2nd flag Configuration diagram of arithmetic decoding apparatus according to Embodiment 2 of the present invention Configuration diagram of a conventional binary arithmetic coding apparatus using CABAC Schematic diagram of binary arithmetic coding using CABAC Diagram showing how a slice is the basic unit of encoding The figure which shows the generated code amount for every macroblock contained in 1 slice

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Arithmetic coding apparatus 11 Analyzing part 12 of arithmetic coding apparatus Arithmetic coding part 13 of arithmetic coding apparatus 13 Determination part 14 of arithmetic coding apparatus Initialization part 15 of arithmetic coding apparatus Table management part 16 of arithmetic coding apparatus Flag management unit 17 of arithmetic coding device Storage unit 20 of arithmetic coding device 20 Arithmetic decoding device 21 Analyzing unit 22 of arithmetic decoding device Arithmetic coding unit 23 of arithmetic decoding device Determination unit 24 of arithmetic decoding device Arithmetic decoding Initialization unit 25 of the decoding device Table management unit 26 of the arithmetic decoding device Flag management unit 27 of the arithmetic decoding device Storage unit of the arithmetic decoding device

Claims (14)

An arithmetic encoding device that performs arithmetic encoding processing using an occurrence probability table,
Table storage means for storing a first table that is one of the duplicated occurrence probability tables and a second table that is the other of the duplicated occurrence probability tables;
Initialization means for initializing the first table and the second table in units of slices;
Arithmetic coding means for performing the arithmetic coding processing using the first table;
Updating means for updating the first table based on a result of the arithmetic encoding process;
A determination unit that determines whether or not the arithmetic coding process for the macroblock is valid each time the arithmetic coding process is completed for one macroblock included in the slice;
When it is determined that the arithmetic encoding process is valid, the value of the second table is controlled to be the value of the first table, and when it is determined that the arithmetic encoding process is not effective. And an arithmetic coding apparatus comprising: control means for controlling the value of the first table to be the value of the second table. 2. The arithmetic coding apparatus according to claim 1, wherein the determination unit replaces the stream data of the macroblock portion with PCM (Pulse Code Modulation) data when determining that the arithmetic encoding processing is not effective. . 2. The arithmetic coding unit according to claim 1, wherein, when it is determined that the arithmetic coding process is not effective, the arithmetic coding process is performed again on the macroblock with a constraint for reducing a generated code amount. 3. The described arithmetic coding device. The arithmetic encoding device further includes:
For each entry of the first table and the second table, a flag storage means for storing a first flag indicating the first table or the second table and a second flag in association with each other is provided.
The arithmetic encoding means uses a table indicated by the second flag,
The updating means updates a table not indicated by the first flag;
When the table not indicated by the first flag is updated, the control means changes the value of the second flag so as to indicate the table not indicated by the first flag, and the arithmetic coding process is effective. If it is determined that there is, the value of the first flag is copied to the second flag, and if it is determined that the arithmetic coding process is not valid, the value of the second flag is set to the first flag. The arithmetic coding apparatus according to claim 1, wherein the arithmetic coding apparatus is duplicated. The determination unit determines that the arithmetic encoding process is effective when the code amount generated by the arithmetic encoding process does not exceed a predetermined amount, and the code amount generated by the arithmetic encoding process is determined. The arithmetic coding apparatus according to claim 1, wherein the arithmetic coding process is determined not to be effective when the amount exceeds a fixed amount. An arithmetic decoding device that performs an arithmetic decoding process using an occurrence probability table,
Table storage means for storing a first table that is one of the duplicated occurrence probability tables and a second table that is the other of the duplicated occurrence probability tables;
Initialization means for initializing the first table and the second table in units of slices;
Arithmetic decoding means for performing the arithmetic decoding process using the first table;
Updating means for updating the first table based on a result of the arithmetic decoding process;
Determination means for determining whether or not the arithmetic decoding process for the macroblock is valid each time the arithmetic decoding process is completed for one macroblock included in the slice;
When it is determined that the arithmetic decoding process is valid, the value of the second table is controlled to be the value of the first table, and when it is determined that the arithmetic decoding process is not effective. And an arithmetic decoding device comprising: control means for controlling the value of the first table to be the value of the second table. 7. The arithmetic decoding device according to claim 6, wherein, when it is determined that the arithmetic decoding process is not effective, the determining unit replaces the stream data of the macroblock portion with PCM (Pulse Code Modulation) data. . 7. The arithmetic decoding unit according to claim 6, wherein, when it is determined that the arithmetic decoding process is not effective, the arithmetic decoding process is performed again for the macroblock with a restriction for suppressing a generated code amount added. 7. The arithmetic decoding device described. The arithmetic decoding device further includes:
For each entry of the first table and the second table, a flag storage means for storing a first flag indicating the first table or the second table and a second flag in association with each other is provided.
The arithmetic decoding means uses a table indicated by the second flag;
The updating means updates a table not indicated by the first flag;
When the table not indicated by the first flag is updated, the control means changes the value of the second flag to indicate the table not indicated by the first flag, and the arithmetic decoding process is effective. If it is determined that there is, the value of the first flag is copied to the second flag, and if it is determined that the arithmetic decoding process is not valid, the value of the second flag is set to the first flag. The arithmetic decoding device according to claim 6, wherein the arithmetic decoding device is duplicated. The determination means determines that the arithmetic decoding process is valid when the code amount generated by the arithmetic decoding process does not exceed a predetermined amount, and the code amount generated by the arithmetic decoding process is determined. The arithmetic decoding device according to claim 6, wherein if the amount exceeds a fixed amount, it is determined that the arithmetic decoding process is not effective. An arithmetic encoding method for performing an arithmetic encoding process using an occurrence probability table,
An initialization step of initializing, in slice units, a first table that is one of the duplicated occurrence probability tables and a second table that is the other of the duplicated occurrence probability tables;
An arithmetic encoding step of performing the arithmetic encoding process using the first table;
An update step of updating the first table based on a result of the arithmetic encoding process;
A determination step of determining whether or not the arithmetic encoding process for the macroblock is valid each time the arithmetic encoding process is completed for one macroblock included in the slice;
When it is determined that the arithmetic encoding process is valid, the value of the second table is controlled to be the value of the first table, and when it is determined that the arithmetic encoding process is not effective. And a control step of controlling the value of the first table to be the value of the second table. An arithmetic decoding method for performing an arithmetic decoding process using an occurrence probability table,
An initialization step of initializing, in slice units, a first table that is one of the duplicated occurrence probability tables and a second table that is the other of the duplicated occurrence probability tables;
An arithmetic decoding step for performing the arithmetic decoding process using the first table;
An update step of updating the first table based on a result of the arithmetic decoding process;
A determination step of determining whether or not the arithmetic decoding process for the macroblock is valid each time the arithmetic decoding process is completed for one macroblock included in the slice;
When it is determined that the arithmetic decoding process is valid, the value of the second table is controlled to be the value of the first table, and when it is determined that the arithmetic decoding process is not effective. And a control step for controlling the value of the first table to be the value of the second table. A program for performing arithmetic coding using an occurrence probability table,
An initialization step of initializing, in slice units, a first table that is one of the duplicated occurrence probability tables and a second table that is the other of the duplicated occurrence probability tables;
An arithmetic encoding step of performing the arithmetic encoding process using the first table;
An update step of updating the first table based on a result of the arithmetic encoding process;
A determination step of determining whether or not the arithmetic encoding process for the macroblock is valid each time the arithmetic encoding process is completed for one macroblock included in the slice;
When it is determined that the arithmetic encoding process is valid, the value of the second table is controlled to be the value of the first table, and when it is determined that the arithmetic encoding process is not effective. A program for causing a computer to execute a control step of controlling the value of the first table to be the value of the second table. A program for performing an arithmetic decoding process using an occurrence probability table,
An initialization step of initializing, in slice units, a first table that is one of the duplicated occurrence probability tables and a second table that is the other of the duplicated occurrence probability tables;
An arithmetic decoding step for performing the arithmetic decoding process using the first table;
An update step of updating the first table based on a result of the arithmetic decoding process;
A determination step of determining whether or not the arithmetic decoding process for the macroblock is valid each time the arithmetic decoding process is completed for one macroblock included in the slice;
When it is determined that the arithmetic decoding process is valid, the value of the second table is controlled to be the value of the first table, and when it is determined that the arithmetic decoding process is not effective. A program for causing a computer to execute a control step of controlling the value of the first table to be the value of the second table.

Images