JP2010161787A - Variable length coding method and apparatus, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a variable length coding method with which the efficiency of coding can be improved when a coefficient value is coded. <P>SOLUTION: The variable length coding method includes: a coefficient scanning step of changing coefficient values on two-dimensional columns of a block into coefficient values on one-dimensional columns; an RL sequence converting step of converting the coefficient values on one-dimensional column into a combination of run value and level value; a step of determining a total number of non-zero coefficient values; a first coding step of performing variable length coding on the run value; a second coding step of executing variable length coding of the level value until reaching the total number while switching tables; and a third coding step of performing variable length coding on the total number. In the second coding step, if an absolute value of the coefficient value L to be encoded exceeds a threshold of the used table, after switched to a table with a threshold greater than the threshold, a coefficient value L to be encoded next is encoded and switching of the tables is unidirectional. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a variable-length encoding method for encoding coefficient values in each block obtained by frequency-converting image data in a moving image in units of blocks having a predetermined size.

  In the moving image encoding process, the amount of information is generally compressed using redundancy in the spatial direction and temporal direction of a moving image. Here, generally, as a method of using redundancy in the spatial direction, conversion to the frequency domain is used, and as a method of using redundancy in the temporal direction, inter-picture predictive encoding processing is used.

  In order to further increase the encoding efficiency of the conventional MPEG-4 moving image encoding method (see, for example, Non-Patent Document 1), the moving image encoding method currently being standardized has a size of 4 × 4 pixels. After performing frequency conversion in units of blocks, quantization is performed to generate coefficient values. Then, scanning is performed from the DC component toward the high-frequency component, and the number R (Run, hereinafter simply referred to as “R”) of coefficients whose continuous values are “0” and the subsequent coefficient value L (Level, hereinafter simply). A combination of (R, L) is created.

  Then, this (R, L) is converted into a code number using a predetermined code table, and then the code number is converted into a VLC code using one variable length coding (VLC) table. Thus, encoding is performed. In this case, the code table is generally assigned a smaller code number as the occurrence frequency is higher. For example, since a combination with small R and L has a high occurrence frequency, a small code number is assigned. In some VLC tables, a VLC code having a short code length is assigned to a small code number.

ISO / IEC 14496-2: “Information technology-Coding of audio-visual objects-Part2: Visual” 7.4.1, pp.119-120, 1999.12

  However, in the above-described conventional method, when the number of consecutive coefficients R having a coefficient value “0” and the coefficient value L are increased, the code length is increased, leading to a decrease in encoding efficiency. In general, since the coefficient value L of the low frequency component value is a large value, the encoding efficiency when the low frequency component value is encoded is significantly reduced.

  That is, as a result of assigning one unique variable length code to one pair of VLC tables R and L according to the occurrence frequency, a variable length code having a very long code length is always given to a coefficient value L having a large value. Will be converted to. Also, when L is encoded separately from R (one-dimensional encoding of L), when one variable length code table is used, the same problem as when encoding with a pair of R and L occurs. .

  The present invention has been made in view of such problems, and an object of the present invention is to provide a variable-length encoding method and the like that can improve encoding efficiency when a coefficient value L is encoded. .

  In order to achieve the above object, a variable length coding method according to the present invention converts a coefficient value of a two-dimensional sequence of a block obtained by frequency converting image data of a block having a predetermined size into a variable length code. The coding method is to scan the coefficient values of the two-dimensional column of the block in a predetermined order from the low frequency component to the high frequency component, thereby converting the coefficient value of the two-dimensional column of the block to the one-dimensional column. A coefficient scanning step for changing to a coefficient value; a coefficient value of the one-dimensional sequence; a run value indicating the number R of consecutive coefficients having a value of zero; and a level value indicating a non-zero coefficient value L that follows. An RL sequence conversion step for converting into a combination; a step for obtaining a total number of non-zero coefficient values included in the block; a first encoding step for variable-length encoding the run value; and The level value indicating the non-zero coefficient value L is sequentially variable-length encoded from the high frequency component to the low frequency component while switching between a plurality of tables to which different threshold values are assigned, and the encoded non-zero A second encoding step executed until the number of coefficient values reaches the total number of non-zero coefficient values, and a third code for variable-length encoding the total number of non-zero coefficients included in the block In the second encoding step, the absolute value of the non-zero coefficient value L to be encoded is set to the threshold value assigned to the table used for encoding the non-zero coefficient value L. If it exceeds, after switching to a table to which a threshold value greater than the threshold value is assigned, the non-zero coefficient value L to be encoded next to the non-zero coefficient value L to be encoded is encoded, Plural Switching of the table is characterized in that was used prior to the new table is switched table is a one-way switch is not used again.

  In order to achieve the above object, the variable-length encoding method according to the present invention encodes a coefficient value in each block obtained by frequency-converting image data in a moving image in units of blocks having a predetermined size. A long encoding method, wherein a coefficient value scanning step of scanning coefficient values in the block in a predetermined order and a plurality of tables used for encoding are switched, and the coefficient values scanned by the coefficient value scanning step Encoding a variable length code in a predetermined order.

  As a result, a variable-length code having a code length corresponding to the coefficient value can be adapted to each table, so that the coding efficiency can be improved. In other words, when a coefficient value is small in one table, it is encoded to a variable length code having a shorter code length than the other table, and when a coefficient value is large in another table, the code length is shorter than that of the other table. By encoding to a variable length code and switching the table according to the coefficient value, the code length can be drastically shortened.

  Here, the switching direction of each table may be one direction. Accordingly, the table is prevented from being frequently replaced by the coefficient value, and the number of table switching is reduced. Therefore, encoding efficiency can be improved. For example, since the work area of the memory is limited, it is common to place only the next table to be used in the work area. In this case, every time the table is switched, it takes time to read the next table from the ROM and expand it in the work area, so it takes time to encode the next coefficient value. For this reason, the effect of reducing the total number of times until the next coefficient value is encoded can be achieved by limiting the number of table switching operations in one direction.

  The encoding step may be configured to perform encoding by switching the plurality of tables in a block, and the coefficient value is a coefficient value other than one-dimensionalized “0”. It is good also as a structure characterized by these.

  Preferably, the encoding is non-arithmetic encoding. As a result, when a table to be used for encoding is determined, it can be encoded into a variable length code simply by referring to the table.

  In addition, each table has a code length at the minimum value of the coefficient value that increases in the order of the numbers assigned to each table, and the code length at the maximum value of the coefficient value does not increase in the order of the numbers. It is preferable that the change rate of the code length with respect to the numerical value is different. In addition, each table is preferably configured such that the increase rate of the code length with respect to the increase in coefficient value decreases in the order of the numbers assigned to the respective tables. As a result, since a range in which the code length is shortened can be assigned to each table, it is possible to surely improve the coding efficiency.

  Moreover, it is preferable that the said encoding step switches each said table based on the threshold value with respect to the absolute value of the preset coefficient value. As a result, the table switching timing can be easily determined, and the encoding efficiency can be improved.

  The coefficient value scanning step preferably scans the coefficient values in the order from the high frequency component to the low frequency component. As a result, there is a strong tendency that the absolute value of the coefficient gradually increases from around “1”. Therefore, it is easy to determine the table used to encode the first coefficient value in the block, and the configuration of each table can be simplified. It can be determined, and the threshold value can be easily determined.

  Further, when the absolute value of the encoding target coefficient exceeds a threshold value, the encoding step converts the table used for encoding the encoding target coefficient from a table having a larger number than the number assigned to the table. It is preferable to switch and encode the next coefficient value. Thereby, since the code length can be shortened when the next coefficient value is encoded, it is possible to improve the encoding efficiency.

  In addition, the variable length decoding method according to the present invention decodes a variable length code generated by encoding coefficient values in each block obtained by frequency-converting image data in a moving image in units of blocks having a predetermined size. A variable length decoding method, wherein a plurality of tables used for decoding are switched, and a variable length code in each block is decoded into coefficient values in a predetermined order; and the decoding step includes And a coefficient generation step of generating a coefficient value in the block based on the generated coefficient value. As a result, it is possible to correctly decode a highly compressed code.

  Here, the switching direction of each table may be one direction.

  The decoding step may be characterized in that the plurality of tables are switched and encoded in a block.

  The coefficient value may be a coefficient value other than one-dimensionalized “0”.

  The decoding may be non-arithmetic decoding.

  In addition, each table has a code length at the minimum value of the coefficient value that increases in the order of the numbers assigned to each table, and the code length at the maximum value of the coefficient value does not increase in the order of the numbers. The code length change rate may be different from that of the numerical value.

  Each table may be characterized in that the rate of increase of the code length with respect to the increase in coefficient value decreases in the order of the numbers assigned to each table.

  Further, the decoding step may switch each table based on a threshold value with respect to an absolute value of a predetermined coefficient value.

  The coefficient generation step may scan the coefficient values in order from a high frequency component to a low frequency component.

  Further, when the absolute value of the decoded coefficient exceeds a threshold value, the encoding step has a number larger than the number assigned to the table from the table used for decoding the variable length code to be decoded. It may be characterized by switching to a table and decoding the next variable length code into a coefficient value.

  It should be noted that the present invention can be realized not only as such a variable-length encoding method and a variable-length decoding method, but also as a characteristic feature included in such a variable-length encoding method and a variable-length decoding method. It can be realized as a variable-length encoding device using a step as a means, a variable-length decoding device, a variable-length encoding method, a video encoding method using characteristic steps included in a variable-length decoding method, It can also be realized as a 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, the variable length coding method according to the present invention switches between a coefficient value scanning step for scanning coefficient values in a block in a predetermined order, and a plurality of tables used for coding. A coding step of coding the coefficient values scanned by the coefficient value scanning step into a variable length code in a predetermined order, so that a variable length code having a code length corresponding to the coefficient value can be adapted to each table. Thus, encoding efficiency can be improved. In other words, when a coefficient value is small in one table, it is encoded to a variable length code having a shorter code length than the other table, and when a coefficient value is large in another table, the code length is shorter than that of the other table. By encoding to a variable length code and switching the table according to the coefficient value, the code length can be drastically shortened.

  In addition, since the switching direction of each table is set to one direction, it is possible to prevent the table from being frequently replaced by the coefficient value, to reduce the number of table switching, and to improve the encoding efficiency.

  In addition, since the encoding is non-arithmetic encoding, when a table used for encoding is determined, it can be encoded into a variable length code simply by referring to the table.

  In addition, each table has a code length at the minimum value of the coefficient value that increases in the order of the numbers assigned to each table, and the code length at the maximum value of the coefficient value does not increase in the order of the numbers. The code length change rate with respect to the numerical value is configured differently, or each table is configured such that the code length increase rate with respect to the increment of the coefficient value decreases in the order of the numbers assigned to the respective tables. By doing so, it is possible to assign a range in which the code length is shortened for each table, and to realize improvement in coding efficiency.

  In addition, the encoding step can easily determine the timing of table switching by switching each table based on a threshold value with respect to a preset absolute value of a coefficient value, thereby realizing improvement in encoding efficiency. it can.

  In the coefficient value scanning step, the coefficient value is scanned in the order from the high frequency component to the low frequency component, so that the absolute value of the coefficient tends to gradually increase from around “1”. It is possible to easily determine a table used for encoding coefficient values, to easily determine the configuration of each table, and to easily determine a threshold value.

  Further, when the absolute value of the encoding target coefficient exceeds a threshold value, the encoding step converts the table used for encoding the encoding target coefficient from a table having a larger number than the number assigned to the table. By switching and encoding the next coefficient value, it is possible to shorten the code length when encoding the next coefficient value, and to improve the encoding efficiency.

  Also, the variable length decoding method according to the present invention includes a decoding step of switching a plurality of tables used for decoding and decoding variable length codes in each block into coefficient values in a predetermined order, and a decoding step And a coefficient generation step of generating a coefficient value in the block based on the coefficient value generated by the above, so that a code subjected to high compression encoding can be correctly decoded.

FIG. 1 is a block diagram showing a functional configuration of an encoding apparatus using a variable length encoding method and a moving image encoding method according to Embodiment 1 of the present invention. FIG. 2 is a block diagram showing a detailed functional configuration of the variable length coding unit shown in FIG. FIG. 3 is a schematic diagram for explaining processing executed by the RL sequence generation unit shown in FIG. FIG. 4 is a schematic diagram for explaining the RL sequence generated by the RL sequence generation unit and the rearrangement process executed by the rearrangement unit shown in FIG. FIG. 5 is a diagram showing an example of a code table held by the table storage unit shown in FIG. FIG. 6 is a diagram showing an example of a variable length code table held by the table storage unit shown in FIG. FIG. 7 is a schematic diagram for explaining another example of the RL sequence generated by the RL sequence generation unit and the rearrangement process executed by the rearrangement unit. FIG. 8 is a block diagram showing a functional configuration of a decoding apparatus using the variable length decoding method and the moving picture decoding method according to Embodiment 2 of the present invention. FIG. 9 is a block diagram showing a detailed functional configuration of the variable length decoding unit shown in FIG. FIG. 10 is a schematic diagram for explaining the RL sequence generated by the code conversion unit shown in FIG. 9 and the rearrangement process executed by the rearrangement unit. FIG. 11 is a schematic diagram for explaining processing executed by the coefficient generation unit shown in FIG. FIG. 12 is a schematic diagram for explaining another example of the RL sequence generated by the code conversion unit and the rearrangement process executed by the rearrangement unit. FIG. 13 is a block diagram showing a configuration of an image encoding device according to the third embodiment of the present invention. FIG. 14 is a block diagram showing the internal configuration of the variable length coding unit. FIG. 15 is a schematic diagram schematically showing coefficient blocks output from the quantization unit. FIG. 16 is a schematic diagram schematically showing the RL sequence output from the RL sequence generation unit. FIG. 17 is a transition diagram showing the switching method of the probability table. FIG. 18 is a probability table content display diagram showing the contents of the probability table. FIG. 19 is a block diagram showing a configuration of an image decoding apparatus according to the fourth embodiment of the present invention. FIG. 20 is a block diagram showing the internal configuration of the variable length decoding unit. FIG. 21 is a table showing an example of the binarization table. FIG. 22 is a block diagram showing a functional configuration of an encoding apparatus to which the variable length encoding method and the moving image encoding method according to Embodiment 5 of the present invention are applied. FIG. 23 is a block diagram showing a detailed functional configuration of the variable length coding unit shown in FIG. FIG. 24 is a diagram illustrating an example of the L column and the R column generated by the R column / L column generation unit illustrated in FIG. FIG. 25 is a diagram illustrating a configuration example of each variable length code table held by the storage unit illustrated in FIG. FIG. 26 is a diagram illustrating a configuration example of a threshold value table held by the storage unit illustrated in FIG. FIG. 27 is a flowchart showing variable length code allocation processing executed by the code allocation unit shown in FIG. FIG. 28 is a diagram illustrating a relationship between a variable length code table used for encoding and a threshold value. FIG. 29 is a schematic diagram illustrating a state of the encoding process of the code assigning unit. FIG. 30 is a block diagram showing a functional configuration of a decoding apparatus using the variable length decoding method and the moving picture decoding method according to Embodiment 6 of the present invention. FIG. 31 is a block diagram showing a detailed functional configuration of the variable length decoding unit shown in FIG. FIG. 32 shows a computer system using a flexible disk storing a program for executing the moving picture encoding method of the first, third, and fifth embodiments or the moving picture decoding method of the second, fourth, and sixth embodiments. It is explanatory drawing in the case of implementing by. FIG. 33 is a block diagram showing the overall configuration of a content supply system that implements a content distribution service. FIG. 34 is a diagram showing a mobile phone using the moving picture prediction method, the moving picture encoding apparatus and the picture decoding apparatus according to the present invention. FIG. 35 is a block diagram showing a configuration of a mobile phone according to the present invention. FIG. 36 is a block diagram showing the overall configuration of a digital broadcasting system according to the present invention.

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

(Embodiment 1)
FIG. 1 is a block diagram showing a functional configuration of an encoding apparatus to which a moving image encoding method of the present invention is applied. In the first embodiment, a functional configuration in the case where an input image is subjected to intraframe encoding processing by the moving image encoding method of the present invention is illustrated.

  As shown in the figure, the encoding device 100a includes a block conversion unit 110, a frequency conversion unit 120, a quantization unit 130, and a variable length encoding unit 140. In addition, each part which comprises such an encoding apparatus 100a provides the work area at the time of program execution, ROM which stores CPU and the program and data which are run by CPU beforehand, an input image etc. temporarily. This is realized by a memory or the like for storing.

  The block conversion unit 110 divides the input image into blocks having a size of 4 horizontal pixels and 4 vertical pixels, and outputs the blocks to the frequency conversion unit 120 for each pixel block.

  The frequency conversion unit 120 performs frequency conversion on the input pixel block and converts it to a frequency coefficient. Then, the frequency conversion unit 120 outputs the converted frequency coefficient to the quantization unit 130.

  The quantization unit 130 performs quantization processing on the input frequency coefficient. Here, the quantization process means a process corresponding to a process of dividing a frequency coefficient value by a predetermined quantization value. This quantization value generally differs for each block and for each frequency band. The quantized frequency coefficient is input to the variable length encoding unit 140.

  The variable length coding unit 140 performs variable length coding on the frequency coefficient value of a block of a predetermined size (horizontal 4 pixels, vertical 4 pixels).

  FIG. 2 is a block diagram illustrating a detailed functional configuration of the variable length coding unit 140.

  As shown in FIG. 2, the variable length coding unit 140 includes an RL sequence generation unit 141, a rearrangement unit 142, a code allocation unit 143, and a table storage unit 144.

  The quantized frequency coefficient output from the quantization unit 130 is input to the RL sequence generation unit 141.

  The RL sequence generation unit 141 first converts the quantized frequency coefficient value to one dimension by a predetermined scanning method. The RL sequence generation unit 141 then combines the number R of consecutive coefficient values “0” and the subsequent coefficient values L other than “0” (hereinafter referred to as RL values) with respect to the one-dimensional coefficient values. ) Column (hereinafter referred to as RL column). The example is demonstrated using FIG. 3, FIG.

  FIG. 3A is a diagram illustrating frequency coefficients in one quantized block output from the quantization unit 130. Here, assuming that the coefficient at the upper left indicates a DC component, the frequency component in the horizontal direction becomes higher toward the right, and the frequency component in the vertical direction becomes higher toward the lower. FIG. 3B is a diagram showing a scanning method when the quantized frequency coefficient value is made one-dimensional. The RL sequence generation unit 141 performs one-dimensionalization by scanning from the low frequency region toward the high frequency region.

  Then, the result of generating the RL sequence for the coefficient values made one-dimensional by the RL sequence generating unit 141 is as shown in FIG. In FIG. 4A, EOB (End Of Block) is an identifier indicating that all subsequent coefficient values in the block are “0”. Generally, the higher the frequency component, the easier the coefficient value becomes “0”. Therefore, by scanning from the low frequency region to the high frequency region, the information amount of the RL sequence can be reduced. The generated RL sequence is input to the rearrangement unit 142.

  The rearrangement unit 142 rearranges the input RL sequence in the reverse direction. However, EOB is excluded from the target of rearrangement. The state after the rearrangement is as shown in FIG. The RL sequence rearranged in this way is input to the code assignment unit 143.

  The table storage unit 144 includes a table (a code table, see FIG. 5) in which an RL value is associated with a code number assigned to the RL value, and a plurality of types of tables in which a code number is associated with a variable-length code. (Variable length code table, FIG. 6) and the like are stored in advance.

  The code assigning unit 143 assigns a variable length code to each set of RL sequences using a table held in the table storage unit 144.

  Specifically, the code assignment unit 143 first assigns a code number to the RL value. At this time, conversion into a code number is performed using a predetermined code table (see FIG. 5) held in the table storage unit 144.

  FIG. 5 is a diagram illustrating an example of a code table.

  The code table is generally assigned a smaller code number as the frequency of the RL value is larger. Generally, the smaller the value of both R and L, the higher the frequency, and this code table is configured using this property. When this code table is used, for example, the first RL value (0, −1) is the code number “2”. The second to fifth RL values (1, 1), (0, -2), (0, 3), (0, 4) are code numbers “3”, “8”, “13”, “15”.

  The code assigning unit 143 then converts the code number into a variable length code. In the conversion to the variable length code, conversion to the variable length code is performed using a plurality of variable length code tables (see FIG. 6) held in the table storage unit 144.

  FIG. 6 is a diagram illustrating an example of a variable-length code table.

  In the first embodiment, two types of variable length code tables are held.

  The first variable length code table 1 and the second variable length code table 2 are configured such that the variable length code length increases as the code number increases, while the variable length code table 1 includes the variable length code table 2. Compared to the variable length code table, the variable length code length is shorter when the code number is smaller. The variable length code table 2 has a variable length code length when the code number is larger than the variable length code table 1. Is configured to be shorter. That is, a shorter code is assigned to the variable length code table 1 with a smaller code number, and a shorter code is assigned to the variable length code table 2 with a larger code number.

  For the first RL value, the variable length code table 1 is used. In this case, since the code number of the first RL value is “2”, the variable length code at this time is “011”. Then, conversion to variable length codes is performed in order, and when the absolute value of L exceeds a threshold value, the variable length code table 2 is used after the next RL value. If the threshold value for the absolute value of L is “2”, the threshold value is exceeded at the fourth RL value (0, 3). Therefore, the first to fourth RL values are converted using the variable-length code table 1, and the fifth and subsequent RL values are converted using the variable-length code table 2. Become.

  Here, with the seventh RL value (1, 2), the absolute value of L again becomes less than or equal to the threshold value, but the conversion to the variable length code table VLC1 is not performed, and conversion is performed using the variable length code table 2. . That is, the table switching direction is one direction. Here, one direction means that a table that has been used once is not reused. Accordingly, the table is prevented from being frequently replaced by the coefficient value, and the number of table switching is reduced. When one-dimensionalization is performed from a high frequency component to a low frequency component, the absolute value of L generally tends to increase.

  Therefore, when the absolute value of L once exceeds the threshold value, even if the subsequent absolute value of L again falls below the threshold value, it is often the case that only the coefficient falls below the threshold value. Therefore, even if the absolute value of L again falls below the threshold value, encoding efficiency can be improved by not using the original variable length code table again. For example, since the work area of the memory is limited, it is common to place only the next table to be used in the work area. In this case, every time the table is switched, it takes time to read the next table from the ROM and expand it in the work area, so it takes time to encode the next coefficient value. For this reason, the effect of being able to limit the number of times of table switching and shortening the time until encoding of the next coefficient value as a whole is exhibited by using one direction.

  Further, the RL sequence generation unit 141 scans the coefficient values of the coefficient value sequence in the order from the low frequency component to the high frequency component, and the code allocation unit 143 encodes the variable value code in order from the rear of the coefficient value sequence. The reason is that the absolute value of the coefficient tends to increase gradually from around “1”, so that it is easy to determine the table used to encode the first coefficient value in the block, This is because the configuration can be easily determined and the threshold value can be easily determined.

  As described above, in the variable length coding method according to the first embodiment, the frequency coefficient value in the block is scanned from the low frequency region to the high frequency region to be one-dimensional. Then, with respect to the one-dimensional coefficient values, a series of RL values, which is a combination of the number R of consecutive coefficient values “0” and the subsequent coefficient values L other than “0”, is generated. Then, the RL value is converted into a variable length code in the reverse order of the scanning order. That is, the RL value may be directly converted to a variable length code. When converting to a variable length code, a plurality of variable length code tables are prepared.

  First, conversion is performed using the first variable length code table, and when the absolute value of L exceeds the threshold value, conversion is performed using the second variable length code table for subsequent RL values. Do. In this case, the first variable length code table is configured so that the variable length code length is shorter when the code number is smaller than the second variable length code table. As compared with the first variable length code table, the variable length code length is shortened when the code number is large.

  In general, the lower the frequency component, the larger the absolute value of L. Therefore, when the RL value is converted into a variable length code in the reverse order of scanning from the low frequency region to the high frequency region and generating the RL value, The absolute value increases.

  Therefore, after the absolute value of L exceeds the threshold value, when the code number is large, that is, when the absolute value of L is large, a variable length code table in which the variable length code length is shortened is used, so that the total code amount Can be reduced. That is, by encoding L separately from R and using a plurality of variable length code tables, the total code amount for L can be reduced.

  In the first embodiment, a case where an image is encoded by intra-frame encoding has been described. However, this is a case where inter-frame encoding is performed using motion compensation or the like for moving image input. The same effect can be obtained by the method of this embodiment.

  In the first embodiment, the case where the input image is divided into blocks of 4 horizontal pixels and 4 vertical pixels has been described. However, the size of the blocks may be other sizes.

  In the first embodiment, the scanning method in the block has been described with reference to FIG. 3B. However, this may be another scanning order as long as scanning from a low frequency region to a high frequency region. good.

  In the first embodiment, the example of the code table has been described with reference to FIG. 5, but this may be another code table.

  Further, although the first embodiment has been described with reference to FIG. 6 as an example of the variable length code table, it may be another table.

  In addition, the case where two variable length code tables are used has been described. This is because three or more variable length code tables are used, and a plurality of threshold values are used. You may make it switch.

  In the present embodiment, the variable length code table is switched when the absolute value of L exceeds the threshold value. However, when the code number exceeds the threshold value, the variable length code table is changed. The same effect can be obtained by switching the code table.

  In the present embodiment, the case where EOB is added to the end of the RL sequence has been described. However, this may add the number of RL values to the beginning of the RL sequence. The number of RL values to be encoded and the RL sequence corresponding to FIG. 4 in this case are as shown in FIGS. 7 (a) and 7 (b). In the code table shown in FIG. 5, it is not necessary to assign a code number to the EOB.

(Embodiment 2)
FIG. 8 is a block diagram showing a functional configuration of a decoding apparatus to which the variable length decoding method according to the embodiment of the present invention is applied. Here, it is assumed that a code string generated by the variable length coding method of the present invention described in Embodiment 1 is input.

  As illustrated in FIG. 8, the decoding device 500 a includes a variable length decoding unit 510, an inverse quantization unit 520, an inverse frequency conversion unit 530, and a frame memory 540. In addition, each part which comprises such a decoding apparatus 500a provides a work area at the time of program execution, ROM which stores CPU and the program and data which are beforehand executed by CPU like the encoding apparatus 100a, This is realized by a memory or the like that temporarily stores an input coded sequence or the like.

  The code string is input to the variable length decoding unit 510. The variable length decoding unit 510 decodes the variable length encoded code string. As described above, this code string divides image data into blocks each having a predetermined size, makes the frequency coefficient values of the blocks one-dimensional in a predetermined scanning order, and the continuous coefficient value is 0. It is generated by encoding a sequence of combinations (RL values) of the consecutive number R of coefficients and the subsequent coefficient value L.

  FIG. 9 is a block diagram illustrating a detailed functional configuration of the variable length decoding unit 510.

  As shown in FIG. 9, the variable length decoding unit 510 includes a code conversion unit 511, a table storage unit 512, a rearrangement unit 513, and a coefficient generation unit 514.

  The table storage unit 512 is configured in the same manner as the table storage unit 144, and a plurality of types of tables (variable length code table, FIG. 6) in which code numbers and variable length codes are associated with each other, and the table storage unit 144 A table (code table, see FIG. 5) in which RL values are associated with code numbers assigned to the RL values is stored in advance.

  The code conversion unit 511 first converts the input code string from a variable length code to a code number using a table (a plurality of variable length code tables) held in the table storage unit 512. Conversion to a code number is performed using a plurality of variable length code tables. This variable length code table is held in the table storage unit 512, and is converted into a code number by referring to the table storage unit 512.

  An example of the variable length code table will be described with reference to FIG. Two types of variable length code tables are held. A variable code table 1 is assigned a shorter code with a small code number, and a code with a shorter code is assigned to a variable code table VLC with a larger code number. Here, it is assumed that the code at the head of the input code string is “01100100000100100011100010011”. The variable length code table 1 is used for the first variable length code. Now, referring to the variable length table 1 of FIG. 6, the variable length code “011” matches the input code string, and the code number in this case is “2”.

  Next, the code conversion unit 511 converts the obtained code number into an RL value. At this time, a predetermined code table is used. The code table is held in the table storage unit 512, and is converted into an RL value by referring to the table storage unit 512. An example of the code table is shown in FIG. Since the code number in this case is “2”, the RL value is (0, −1).

  Similarly, when conversion from the variable length code to the code number is performed in order, using the variable length code table 1, the variable length code “00100” becomes the code number “3” and the variable length code “000001” The variable number code “0001110” is converted into the code number “13” into the code number “8”, and each code number is converted into the RL values (1, 1), (0, −2), (0, 3). Converted.

  Here, if the absolute value of L exceeds the threshold value among the obtained RL values, the code conversion unit 511 uses the variable length code table 2 in the subsequent variable length code conversion. If the threshold value for the absolute value of L is “2”, the threshold value is exceeded at the fourth RL value (0, 3). Therefore, the subsequent RL values are converted using the variable length code table 2. Therefore, the next variable length code “0010011” is converted into code number 15 and further converted into an RL value (0, 4).

  Further, even if the absolute value of L among the RL values obtained in the subsequent decoding becomes equal to or less than the threshold value again, switching to the variable length code table 1 is not performed, and the variable length code table 2 is used. Perform conversion. As described above, when RL values for one block are generated (when EOB is detected), they are input to the rearrangement unit 513. Here, it is assumed that the RL sequence in FIG. 10A is generated.

  The rearrangement unit 513 rearranges the input RL sequence in the reverse direction. However, EOB is excluded from the target of rearrangement. The state after the rearrangement is shown in FIG. The RL sequence rearranged in this way is input to the coefficient generation unit 514.

  The coefficient generation unit 514 converts the input RL sequence into coefficient values, and two-dimensionalizes the coefficient blocks by a predetermined scanning method. In the conversion to the coefficient value, the coefficient value “0” is generated by the value indicated by R based on a predetermined scanning order, and then the coefficient value of the value indicated by L is generated. , The RL sequence is converted into coefficient values. Here, when zigzag scanning is performed from the low frequency region to the high frequency region, the RL sequence in FIG. 10B is converted into the coefficient block shown in FIG. The generated coefficient block is input to the inverse quantization unit 520.

  The inverse quantization unit 520 performs inverse quantization processing on the input coefficient block. Here, the inverse quantization process means a process corresponding to a process of adding a predetermined quantized value to a coefficient value. Here, the quantized value generally differs for each block and for each frequency band, and is obtained from a code string or uses a predetermined value. The inversely quantized coefficient block is input to the inverse frequency transform unit 530. The inverse frequency transform unit 530 performs inverse frequency transform on the inversely quantized coefficient block to convert it into a pixel block. The converted pixel block is input to the frame memory 540.

  In the frame memory 540, the decoded pixel blocks are sequentially stored, and when the pixel blocks for one screen are stored, they are output as an output image.

  As described above, in the variable length decoding method of the present invention, the input code string is first decoded using the first variable length table, and the number R of consecutive coefficient values 0, followed by the coefficient values L other than 0, A string of RL values that is a combination of is generated. When the absolute value of L exceeds the threshold value, the second variable length table is used for subsequent decoding of the variable length code. Then, after rearranging the RL values in reverse order, the RL values are converted into coefficient values based on a predetermined scanning order within the block.

  With the above operation, by using the variable length decoding method of the present invention, it is possible to correctly decode a code string encoded using the variable length coding method of the present invention.

  In the second embodiment, the case where the code string generated by the intra-frame coding is decoded has been described. However, this is performed by performing inter-frame coding using motion compensation or the like on the moving image input. The generated code string may be decoded, and the same effect can be obtained by the method of the present embodiment.

  Further, in the second embodiment, the case where the input image is divided and encoded into blocks of 4 horizontal pixels and 4 vertical pixels has been described. However, the size of the blocks may be other sizes. .

  In the second embodiment, the scanning method in the block has been described with reference to FIG. 11. However, this may be another scanning order as long as it is the same as the scanning method used at the time of encoding. .

  In the second embodiment, the example of the code table has been described with reference to FIG. 5, but this may be another code table as long as it is the same as the code table used for encoding.

  In the second embodiment, the variable length code table has been described with reference to FIG. 6 as an example. However, if this is the same as the variable length code table used for encoding, this is another table. Also good. In addition, the case where two variable length code tables are used has been described. This is because three or more variable length code tables are used, and a plurality of threshold values are used. You may make it switch. However, the configuration and threshold value of the variable length table at this time must be the same as those used for encoding.

  In the second embodiment, the case where the variable-length code table is switched when the absolute value of L exceeds the threshold value has been described. This is variable when the code number exceeds the threshold value. The same effect can be obtained by switching the long code table.

  In the second embodiment, a case has been described in which a code string that has been encoded by adding EOB to the end of the RL sequence has been described, but this is the case where the number of RL values is added to the beginning of the RL sequence. The encoded code string may be decoded. The number of RL values obtained by decoding and the RL sequence corresponding to FIG. 10 in this case are as shown in FIGS. 12 (a) and 12 (b). At this time, it is not necessary to assign a code number to the EOB in the code table shown in FIG.

  As described above, in the variable length coding method of the present invention, the frequency coefficient value in the block is scanned from the low frequency region to the high frequency region to be one-dimensional. Then, for a one-dimensional coefficient value, a string of RL values, which is a combination of the number R of consecutive coefficient values 0 and the subsequent coefficient value L other than 0, is generated. Then, the RL value is converted into a variable length code in the reverse order of the scanning order. When converting to a variable length code, a plurality of variable length code tables are prepared.

  First, conversion is performed using the first variable length code table, and when the absolute value or code number of L exceeds the threshold value, the second variable length code table is used for the subsequent RL values. To convert. In this case, the first variable length code table is configured so that the variable length code length is shorter when the code number is smaller than the second variable length code table. As compared with the first variable length code table, the variable length code length is shortened when the code number is large.

  In general, the absolute value and code number of L increase as the low frequency component increases. Therefore, when the RL value is converted into a variable length code in the reverse order from the generation of the RL value by scanning from the low frequency region to the high frequency region. , The absolute value of L increases. Therefore, the total code amount can be reduced by using a variable-length code table in which the variable-length code length is shortened when the code number is large after the absolute value of L exceeds the threshold value.

  In the variable length decoding method of the present invention, the input code string is first decoded using the first variable length table, and the combination of the number R of consecutive coefficient values 0 and the subsequent coefficient values L other than 0 is used. Generate a sequence of certain RL values. When the absolute value or code number of L exceeds the threshold value, the second variable length table is used for subsequent decoding of the variable length code. Then, after rearranging the RL values in reverse order, the RL values are converted into coefficient values based on a predetermined scanning order within the block.

  With the above operation, by using the variable length decoding method of the present invention, it is possible to correctly decode a code string encoded using the variable length coding method of the present invention.

(Embodiment 3)
Hereinafter, an image coding apparatus according to a third embodiment of the present invention will be described with reference to the drawings.

  FIG. 13 is a block diagram showing a configuration of an image encoding device 100b according to the third embodiment of the present invention.

  The image encoding device 100b performs intra-picture encoding processing with improved encoding efficiency on an input image (image data). The block conversion unit 101, the frequency conversion unit 102, and the quantization unit 103 , And a variable length encoding unit 150.

  The block conversion unit 101 divides the input image into pixel blocks each having a size of horizontal 4 × vertical 4 pixels, and outputs the divided pixel block to the frequency conversion unit 102.

  The frequency conversion unit 102 performs frequency conversion on each of the divided pixel blocks to generate a frequency coefficient. Then, the frequency conversion unit 102 outputs the generated frequency coefficient to the quantization unit 103.

  The quantization unit 103 performs a quantization process on the frequency coefficient output from the frequency conversion unit 102. Here, the quantization process means a process corresponding to dividing a frequency coefficient by a predetermined quantization value. In addition, the quantization value generally differs for each pixel block and for each frequency band. Then, the quantization unit 103 outputs the quantized frequency coefficient to the variable length coding unit 150.

  The variable length coding unit 150 performs variable length coding on the frequency coefficient quantized by the quantization unit 103.

  FIG. 14 is a block diagram showing an internal configuration of the variable length coding unit 150.

  As shown in FIG. 14, the variable length encoding unit 150 includes an RL sequence generation unit 201, a rearrangement unit 202, a binarization unit 203, a table storage unit 204, and an arithmetic encoding unit 205.

  The RL sequence generation unit 201 converts the quantized frequency coefficient (hereinafter abbreviated as “coefficient”) output from the quantization unit 103 into a one-dimensional pattern using a predetermined scanning method. The RL sequence generation unit 201 then combines a number R of consecutive coefficient values “0” with a coefficient value L other than “0” (hereinafter “RL”) for the one-dimensionalized coefficients. A column of “value” (hereinafter referred to as “RL column”) is generated. An example will be described with reference to FIGS. 15 and 16.

  FIG. 15A shows a coefficient block composed of a plurality of coefficients output from the quantization unit 103. Here, the upper left frequency coefficient in the coefficient block indicates a DC component, and the frequency component in the horizontal direction increases as it goes to the right, and the frequency component in the vertical direction increases as it goes down.

  FIG. 15B is an explanatory diagram for explaining a scanning method when a plurality of coefficients in the coefficient block are made one-dimensional. The RL sequence generation unit 201 performs one-dimensionalization of the coefficients by scanning the coefficient block from the low frequency region to the high frequency region as indicated by an arrow in FIG.

  FIG. 16A shows an RL sequence output from the RL sequence generation unit 201.

  In FIG. 16A, the first number indicates the number of coefficients. In general, since the coefficient value tends to be “0” in the high frequency region, by scanning from the low frequency region toward the high frequency region, the information amount of the RL sequence (the information amount of the number R of them) is reduced. Can be small. The generated RL sequence is input to the rearrangement unit 202.

  The rearrangement unit 202 rearranges the input RL sequence in the reverse direction. However, the number of coefficients is excluded from the target of rearrangement.

  FIG. 16B shows the RL sequence rearranged by the rearrangement unit 202. By rearranging in this way, the amount of information was reduced as described above, but as a result, the coefficient block was scanned from the high frequency region to the low frequency region to make the coefficients one-dimensional. It will be. Then, the rearranged RL sequence is output to the binarization unit 203.

  The binarization unit 203 binarizes the number of coefficients and each RL value, that is, converts into binarized data including “0” and “1”. Here, the number R and the coefficient value L are binarized separately.

  FIG. 16C shows only the coefficient value L of the RL sequence rearranged by the rearrangement unit 202. For these coefficient values L, their absolute values and positive and negative signs are processed separately. The binarization unit 203 binarizes the absolute values of the number R and the coefficient value L using a predetermined binarization table as shown in FIG. Then, the binarization unit 203 outputs binarized data that has been binarized to the arithmetic coding unit 205.

  The arithmetic coding unit 205 performs binary arithmetic coding on the absolute value of the number R value and coefficient value L expressed as binary data, and also performs coding on the positive / negative sign of the coefficient value L. Do. Here, arithmetic coding of the absolute value of the coefficient value L will be described. The arithmetic coding unit 205 switches between a plurality of probability tables when performing arithmetic coding on the absolute value of the coefficient value L expressed as binarized data. The plurality of probability tables are stored in the table storage unit 204.

  FIG. 17 is a transition diagram showing a probability table switching method.

  As shown in FIG. 17, the arithmetic encoding unit 205 uses four probability tables and arithmetically encodes the absolute value of the leading coefficient value L using the probability table 1. For the absolute value of the coefficient value L thereafter, the arithmetic encoding unit 205 uses the table number of the probability table used when encoding the absolute value of the immediately previous coefficient value L and the absolute value thereof. Depending on, the probability table to be used is switched. Here, the four probability tables are the probability table 1, the probability table 2, the probability table 3, and the probability table 4, the table number of the probability table 1 is “1”, the table number of the probability table 2 is “2”, and the probability The table number of the table 3 is “3”, and the table number of the probability table 4 is “4”.

  Specifically, when the absolute value of the immediately preceding coefficient value L is encoded using the probability table 1 and the absolute value is “1”, or the absolute value of the immediately preceding coefficient value L uses the probability table 2. If the absolute value is “1”, the probability table 2 is used. The absolute value of the immediately preceding coefficient value L is encoded using the probability table 1 and the absolute value is “2”, or the absolute value of the immediately preceding coefficient value L is encoded using the probability table 2. When the absolute value is “2”, or when the absolute value of the immediately preceding coefficient value L is encoded using the probability table 3 and the absolute value is “2 or less”, the probability table 3 Is used. When the absolute value of the immediately preceding coefficient value L is “3 or more”, or when the absolute value of the immediately preceding coefficient value L is encoded using the probability table 4, the probability table 4 is used.

  As described above, the switching of the probability table is one direction from the probability table with the smaller table number to the probability table with the larger table number, and even if the absolute value of the immediately preceding coefficient value L is equal to or less than the predetermined threshold value, No change in direction is made. This is a point different from the conventional example.

  FIG. 18 is a probability table content display diagram showing the contents of the four probability tables 1 to 4 described above.

  As shown in FIG. 18, each of the probability tables 1 to 4 includes a probability of occurrence of “0” and a probability of occurrence of “1”.

  For example, the probability table 1 includes a probability “0.1” of occurrence of “0” and a probability “0.9” of occurrence of “1”, and the probability table 2 has a probability “0” of occurrence of “0”. 0.2 ”and the probability“ 0.8 ”of occurrence of“ 1 ”.

  That is, if the absolute value of the coefficient value L is “2”, since the binarized value of “2” is “01”, the arithmetic encoding unit 205 uses the probability table 1 to calculate this. When arithmetic coding is performed, the probability “0.1” corresponding to “0” of the “01” and the probability “0.9” corresponding to “1” of the “01” are used, and this “01” is used. "Is arithmetically encoded.

  Here, since the sum of the probability of occurrence of “0” and the probability of occurrence of “1” is 1.0, it is not necessary to hold both probabilities, and even if only one probability is held. good.

  An example of switching the probability table when the absolute value (binarized) of the coefficient value L shown in FIG. 16C is encoded will be described below.

  The arithmetic encoding unit 205 uses the probability table 1 for the absolute value of the first coefficient value L (−2). Here, since the absolute value of the coefficient value L is 2, the arithmetic encoding unit 205 changes the probability table to be used from the probability table 1 to the probability table 3. As a result, the arithmetic encoding unit 205 arithmetically encodes the absolute value of the second coefficient value L (3) using the probability table 3.

  Here, since the absolute value of the coefficient value L is “3”, the arithmetic coding unit 205 causes the probability table 3 to transition from the probability table 3 to the probability table 4. Thus, the arithmetic encoding unit 205 arithmetically encodes the absolute value of the third coefficient value (6) using the probability table 4. Since the probability table used here is transitioned to the probability table 4, the arithmetic encoding unit 205 arithmetically encodes all the absolute values of the subsequent coefficient values L using the probability table 4. For example, the absolute value of the fifth coefficient value L is “2”, but unlike the conventional example, the arithmetic encoding unit 205 performs probability encoding when the absolute value of the sixth and subsequent coefficient values L is arithmetically encoded. The probability table 4 is used without changing the table.

  Further, each probability table is updated as needed depending on whether the input is “0” or “1”, and thus is updated to a probability table adapted to the input.

  As described above, the variable length coding method in the variable length coding unit 150 of the image coding device 100b according to the present invention scans the coefficients in the coefficient block from the low frequency region to the high frequency region and performs one-dimensional scanning. Turn into. Then, an RL value column (RL column) that is a combination of the number R of consecutive coefficient values “0” and the subsequent coefficient value L other than “0” is generated for the one-dimensional coefficient.

  Then, the RL value is converted into a variable length code in the reverse order of the scanning order. When converting to a variable length code, the number R, the absolute value of the coefficient value L, and the sign of the coefficient value L are individually converted. In these conversions, binarization is first performed, and then arithmetic codes are applied. When an arithmetic code is applied to the absolute value of the coefficient value L, a plurality of probability tables are switched. When switching the probability table, the probability table for encoding the next absolute value of the coefficient value L is determined based on the table number of the current probability table and the absolute value of the coefficient value L. The transition of the probability table is only in one direction, and once the absolute value of the coefficient value L exceeds a predetermined value, arithmetic coding is performed with the same probability table thereafter.

  In general, since the absolute value of the coefficient value L is larger in the low frequency region, the absolute value of the coefficient value L often increases in order when scanning is performed in order from the high frequency region to the low frequency region. Therefore, once the absolute value of the coefficient value L exceeds a predetermined value, even if the absolute value of the coefficient value L becomes smaller than the predetermined value after that, only the absolute value of the coefficient value L becomes small. It is very likely that arithmetic coding using the same probability table makes it easier to adapt the update of the probability table to the input, so that each probability table symbol ("0" or binary data) The occurrence probability of “1”) is likely to be biased (that is, the occurrence probability of “0” or “1” is close to 1.0). Arithmetic coding is characterized in that the coding efficiency increases as the probability values in the probability table are biased. Therefore, the coding efficiency can be improved by using the variable length coding method of the present invention.

  The image encoding apparatus according to the present invention has been described above using the present embodiment, but the present invention is not limited to these.

  For example, in the present embodiment, a case where an image is encoded by intra-picture encoding has been described, but this may be a case where inter-picture encoding is performed using motion compensation or the like for moving image input. A similar effect can be obtained.

  In the present embodiment, the case where the input image is divided into pixel blocks of horizontal 4 × vertical 4 pixels has been described, but the size of the pixel blocks may be other sizes.

  In this embodiment, the scanning method in the coefficient block has been described with reference to FIG. 15B. However, this is another scanning order if scanning is performed from the low frequency region to the high frequency region. Also good.

  In the present embodiment, the RL sequence generation unit 201 makes the quantized frequency coefficient one-dimensional by a predetermined scanning method, and the number of continuous coefficient values “0” with respect to the one-dimensional coefficient. Although the case of generating a combination column (RL column) of R and subsequent coefficient values L other than “0” has been described, the number R column and the coefficient value L column may be generated individually. For example, when generating a column of coefficient values L, the rearrangement unit 202 can be omitted if the coefficient values L are generated by scanning from the high frequency region toward the low frequency region and selecting a coefficient having a coefficient value other than 0. .

  Further, in the present embodiment, a case has been described in which four probability tables are used and the probability table changes based on the transition diagram shown in FIG. 17, but the number of probability tables and coefficient values at the time of transition in FIG. The threshold value for the absolute value of L may be another value.

  In the present embodiment, the example of the binarization table has been described with reference to FIG. 21, but this may be another table.

  In the present embodiment, the case where the arithmetic coding unit performs binary arithmetic coding has been described. However, multilevel arithmetic coding may be performed. In this case, the binarization unit 203 can be omitted.

(Embodiment 4)
Hereinafter, an image decoding apparatus according to a fourth embodiment of the present invention will be described with reference to the drawings.

  FIG. 19 is a block diagram showing a configuration of an image decoding device 500b according to the fourth embodiment of the present invention.

  The image decoding apparatus 500b performs an intra-picture decoding process on a code string obtained by intra-picture encoding processing of image data, and includes a variable length decoding unit 601, an inverse quantization unit 602, an inverse frequency A conversion unit 603 and a frame memory 604 are provided. The code string input here is generated by, for example, the variable length coding method of the image coding apparatus 100b according to the third embodiment, and the variable length decoding unit 601 first acquires the code string.

  When the variable length decoding unit 601 acquires the code string, the variable length decoding unit 601 performs variable length decoding on the code string to create a coefficient block including a plurality of coefficients as illustrated in FIG.

  When the inverse quantization unit 602 obtains the coefficient block from the variable length decoding unit 601, the inverse quantization unit 602 performs an inverse quantization process on the coefficient block. Here, the inverse quantization process means that a predetermined quantization value is added to each coefficient of the coefficient block. Here, the quantized value generally differs depending on each coefficient block and frequency band, and is obtained from the code string. Then, the inverse quantization unit 602 outputs the inversely quantized coefficient block to the inverse frequency transform unit 603.

  The inverse frequency transform unit 603 performs inverse frequency transform on the inversely quantized coefficient block to convert the coefficient block into a pixel block. Then, the inverse frequency conversion unit 603 outputs the converted pixel block to the frame memory 604.

  The frame memory 604 sequentially stores the decoded pixel blocks. When pixel blocks for one screen are stored, these pixel blocks are output as an output image.

  Here, the variable length decoding unit 601 described above will be described in detail.

  FIG. 140 is a block diagram showing an internal configuration of the variable length decoding unit 601.

  As shown in FIG. 20, the variable length decoding unit 601 includes an arithmetic decoding unit 701, a multi-value conversion unit 702, a table storage unit 703, a rearrangement unit 704, and a coefficient generation unit 705.

  The table storage unit 703 holds, for example, four probability tables 1 to 4 as shown in FIG.

  When the arithmetic decoding unit 701 acquires the code string, first, the arithmetic decoding unit 701 performs arithmetic decoding on the code string. Here, binary arithmetic decoding for the absolute value (binarized) of the encoded coefficient value L included in the code string will be described.

  When performing arithmetic decoding on the absolute value of the encoded coefficient value L, the arithmetic decoding unit 701 multivalues the absolute value of the immediately preceding coefficient value L that has already been decoded and multi-valued. The probability values 1 to 4 obtained from the conversion unit 702 and stored in the table storage unit 703 according to the absolute value of the coefficient value L are switched as shown in FIG. The binary values of these are binary arithmetic decoded and binarized data corresponding to them are output.

  The multi-value quantization unit 702 multi-values the binarized data output from the arithmetic coding unit 701 by using, for example, a binarization table as shown in FIG. To do. Then, the multilevel unit 702 outputs the absolute value of the coefficient value L to the arithmetic decoding unit 701 and the rearrangement unit 704.

  Detailed operations of the arithmetic decoding unit 701 and the multi-value conversion unit 702 will be described.

  First, the arithmetic decoding unit 701 arithmetically decodes the encoded absolute value of the leading coefficient value L using the probability table 1. Then, the arithmetic decoding unit 701 outputs the binarized data obtained by the arithmetic decoding to the multi-level unit 702. The multi-value conversion unit 702 converts the binarized data into the absolute value of the coefficient value L by using the binarization table, and outputs the absolute value to the arithmetic decoding unit 701 and the rearrangement unit 704. To do.

  Next, the arithmetic decoding unit 701 uses the absolute value of the encoded coefficient value L after that for the binary arithmetic decoding of the absolute value of the immediately previous encoded coefficient value L. The probability table to be used is switched according to the table number of the obtained probability table and the absolute value of the immediately preceding coefficient value L acquired from the multilevel conversion unit 702. As shown in FIG. 17, the absolute value of the immediately preceding coefficient value L is arithmetically decoded using the probability table 1 and the absolute value of the immediately preceding coefficient value L obtained from the multi-value quantization unit 702 is In the case of “1” or the absolute value of the immediately preceding coefficient value L obtained by arithmetic decoding of the absolute value of the immediately preceding encoded coefficient value L using the probability table 2 and obtained from the multi-value quantization unit 702 When “1” is “1”, the probability table 2 is used.

  When the absolute value of the immediately preceding encoded coefficient value L is arithmetically decoded using the probability table 1 and the absolute value of the immediately preceding coefficient value L acquired from the multi-value quantization unit 702 is “2”, or When the absolute value of the immediately preceding encoded coefficient value L is arithmetically decoded using the probability table 2 and the absolute value of the immediately preceding coefficient value L acquired from the multi-value quantization unit 702 is “2”, Alternatively, the absolute value of the immediately preceding encoded coefficient value L is arithmetically decoded using the probability table 3 and the absolute value of the immediately preceding coefficient value L acquired from the multilevel conversion unit 702 is “2 or less”. In this case, the probability table 3 is used.

  When the absolute value of the immediately preceding coefficient value L acquired from the multi-value conversion unit 702 is “3 or more”, or the absolute value of the immediately preceding encoded coefficient value L is arithmetically decoded using the probability table 4. If so, the probability table 4 is used. As described above, the switching of the probability tables 1 to 4 is one direction from the probability table having the smaller table number to the probability table having the larger table number, and the absolute value of the immediately preceding coefficient value L acquired from the multi-value quantization unit 702 is Even when the value is equal to or less than the predetermined threshold value, switching in the reverse direction is not performed. This is a point different from the conventional example.

  An example of switching the probability table in the case where the absolute value of the coefficient value L shown in FIG. 16C is decoded will be described below.

  The arithmetic decoding unit 701 arithmetically decodes the absolute value of the first encoded coefficient value L (−2) into binarized data “01” using the probability table 1. Here, since the arithmetic decoding unit 701 obtains “2” multi-valued from the binarized data “01” from the multi-valued unit 702, the probability table to be used is determined from the probability table 1. Transition to the probability table 3. As a result, the arithmetic decoding unit 701 arithmetically decodes the absolute value of the second encoded coefficient value L (3) into binary data “001” using the probability table 3.

  Here, the arithmetic decoding unit 701 acquires “3” multi-valued from the binarized data “001” from the multi-value conversion unit 702, so that the probability table to be used is obtained from the probability table 3. Transition to the probability table 4. As a result, the arithmetic decoding unit 701 arithmetically decodes the absolute value of the third encoded coefficient value L (6) into binary data “000001” using the probability table 4. Since the probability table used here is transitioned to the probability table 4, the arithmetic decoding unit 701 performs arithmetic decoding of all the absolute values of the encoded coefficient values L using the probability table 4. For example, the absolute value of the fifth encoded coefficient value L is decoded and multi-valued, and the result is “2”. Unlike the conventional example, the arithmetic decoding unit 701 performs the sixth and subsequent operations. When arithmetically decoding the absolute value of the encoded coefficient value L, the probability table 4 is used without transitioning the probability table.

  When the absolute value and the number R of the coefficient values L for one coefficient block and the sign of the coefficient value L are generated by the operation as described above, these are input to the rearrangement unit 704 as an RL sequence.

  The rearrangement unit 704 rearranges the input RL sequence in the reverse direction. However, the number of coefficients is excluded from the target of rearrangement. The rearranged state is as shown in FIG. Then, rearrangement section 704 outputs the RL sequence thus rearranged to coefficient generation section 705.

  The coefficient generation unit 705 converts the input RL sequence into a coefficient block. At this time, the coefficient generation unit 705 generates coefficients having the value “0” by the number indicated by the number R based on a predetermined scanning order, and then the coefficient having the value indicated by the coefficient value L. By repeating the generation of the RL sequence to the coefficient block. Here, the coefficient generation unit 705 performs zigzag scanning from the low frequency region to the high frequency region as illustrated in FIG. 15B, and the RL sequence illustrated in FIG. The coefficient block shown in a) is converted. Then, the coefficient generation unit 705 outputs the coefficient block generated in this way to the inverse quantization unit 602.

  As described above, the variable length decoding method in the variable length decoding unit 601 of the image decoding apparatus 500b according to the present invention is performed when arithmetic decoding of the absolute value of the coefficient value L in the input code string is performed. Switch between multiple probability tables. When switching the probability table, the probability table for decoding the absolute value of the next coefficient value L is determined based on the table number of the current probability table and the absolute value of the coefficient value L obtained by decoding. At this time, the transition of the probability table is only in one direction, and once the absolute value of the coefficient value L obtained by decoding exceeds a predetermined value, arithmetic decoding is performed with the same probability table thereafter.

  Thus, by using the variable length decoding method of the present invention, it is possible to correctly decode the code string encoded using the variable length coding method of the present invention.

  The image decoding apparatus according to the present invention has been described above using the present embodiment, but the present invention is not limited to these.

  For example, in the present embodiment, a case has been described in which a code string generated by intra-picture encoding is decoded. This is generated by inter-picture encoding using motion compensation or the like for moving image input. It is also possible to decode the processed code string, and the same effect can be obtained.

  In this embodiment, a description has been given of a code string in which image data is encoded by being divided into pixel blocks of horizontal 4 × vertical 4 pixels, but the size of the pixel block is other size. Also good.

  Further, in the present embodiment, a case has been described in which four probability tables are used and the probability table changes based on the transition diagram shown in FIG. 17, but the number of probability tables and coefficient values at the time of transition in FIG. The threshold value for the absolute value of L may be another value.

  In this embodiment, the scanning method in the coefficient block has been described with reference to FIG. 15B. However, if this is the same as the scanning method used at the time of encoding, another scanning order is used. May be.

  In this embodiment, the binarization table has been described with reference to FIG. 21 as an example. However, if this is the same as the binarization table used for encoding, other tables may be used. good.

  In the present embodiment, the case where the arithmetic decoding unit 701 performs binary arithmetic decoding has been described. However, multi-level arithmetic decoding may be performed. In this case, the multi-value conversion unit 702 can be omitted.

  Next, still another embodiment according to the present invention will be described with reference to the drawings.

(Embodiment 5)
FIG. 22 is a block diagram showing a functional configuration of an encoding apparatus to which the variable length encoding method and the moving image encoding method using the variable length encoding method of the present invention are applied. In the fifth embodiment, as in the first and third encoding apparatuses 100a and 100b, the functional configuration when an input image is subjected to intraframe encoding processing by the moving image encoding method of the present invention is also used. It is shown in the figure. In addition, each unit configuring such an encoding device 100c includes a CPU, a ROM that stores programs and data executed by the CPU in advance, a work area when executing the program, and an input image or the like temporarily. This is realized by a memory or the like for storing.

  As shown in FIG. 22, the coding apparatus 100c according to the fifth embodiment includes a block transform unit 110, a frequency transform unit 120, a quantization unit 130, and a variable length coding unit 160.

  Here, the encoding apparatus 100a according to the first embodiment is configured to encode a pair of R and L using a plurality of variable length code tables (VLC tables), and the code according to the third embodiment. In the encoding apparatus 100b, L is configured to be arithmetically encoded using a plurality of probability tables separately from R. However, in the encoding apparatus 100c according to the fifth embodiment, L is separately from R. In addition, it is different from the encoding devices 100a and 100b in that the encoding is performed using a plurality of variable length code tables. Therefore, the encoding device 100c is configured using a variable length encoding unit 160 instead of the variable length encoding units 140 and 150 of the encoding devices 100a and 100b. Since other configurations are the same as those of the encoding devices 100a and 100b, the description thereof will be omitted, and the variable length encoding unit 160 will be described in detail.

  The variable length coding unit 160 individually generates an L column and an R column based on the frequency coefficient quantized by the quantization unit 130, and uses the one-dimensional VLC switching method to obtain the absolute value | L | A code string is generated.

  FIG. 23 is a block diagram illustrating a detailed functional configuration of the variable length coding unit 160.

  As shown in FIG. 23, the variable length coding unit 160 includes an R column / L column generation unit 161, a code allocation unit 163, and a table storage unit 164.

  The R column / L column generation unit 161 zigzags scans the quantized frequency coefficients (hereinafter also simply referred to as “coefficients”) from the lower frequency to the higher frequency to generate the L column and the R column separately. To do.

  Specifically, when the coefficient of the block illustrated in FIG. 3A is input, the R-row / L-row generation unit 161 performs zigzag scanning as illustrated in FIG. Then, as shown in FIG. 24A, the R column / L column generator 161 first determines the number m of Ls whose coefficient values are other than “0” and the absolute value of the coefficient | The sequence of L | and the code sequence of the coefficient are obtained. This is because R depends on L, whereas L does not depend on R and can be acquired independently. Next, the R column / L column generation unit 161 generates an R column (R column) as illustrated in FIG.

  The table storage unit 164 includes a plurality of (for example, eight) variable-length code tables 1641a to 1641g for variable-length encoding the absolute value | L | of each coefficient in the L column, and the absolute value | L | And a threshold value table 1642 for adaptively switching the variable length code tables 1641a to 1641g in accordance with the absolute value | L | of the coefficient.

  FIG. 25 is a diagram illustrating a configuration example of each of the variable length code tables 1641a to 1641g. Note that each of the variable length code tables 1641a to 1641g is actually configured by associating the absolute value | L | of the coefficient with its binarized code, but in FIG. Yes.

  Here, a smaller code number is assigned as the appearance frequency of the coefficient absolute value | L | is higher, but in general, the appearance frequency is higher as the absolute value | L | of the coefficient is smaller. This is because the maximum value of the absolute value | L | of the coefficient is dispersed depending on the image and the screen, and the frequency of appearance of the same value is low, whereas the minimum value of the absolute value | L | This is because values, that is, high-frequency components, tend to concentrate on about “1” or “2”, and the appearance frequency of the same value increases. On the other hand, if the absolute value | L | of the coefficient is associated with the binarized code with only one variable length code table, the larger the absolute value | L | of the coefficient, the longer the code length. For this reason, variable length code tables 1641a to 1641g to be applied in accordance with the absolute value | L | of the coefficient are prepared in advance so that the code length does not become too long even when the absolute value | L | .

  In each of the variable length code tables 1641a to 1641g, the code length at the minimum value of the coefficient value becomes longer in the order of the number k assigned to each table, and the code length at the maximum value of the coefficient value becomes shorter in the order of the number k. Thus, the code length change rate with respect to the coefficient value is made different.

  More specifically, the variable length code table 1641a has the shortest code length when the absolute value | L | of the coefficient is small and the longest code length when the absolute value | L | A long table. That is, the variable length code table 1641a is the table having the largest code length change rate with respect to the coefficient absolute value | L | among the variable length code tables 1641a to 1641g, and the coefficient absolute value | L | Is suitable for application to a small case (for example, “1” to “3”).

  The variable length code table 1641g is a table having the longest code length when the absolute value | L | of the coefficient is small and the shortest code length when the absolute value | L | of the coefficient is large. . That is, the variable length code table 1641g is a table having the smallest code length change rate with respect to the coefficient absolute value | L | among the variable length code tables 1641a to 1641g, and the coefficient absolute value | L | This is suitable for application when the value is large (for example, “193”).

  In the variable length code tables 1641b to 1641f, the code length gradually increases when the coefficient absolute value | L | is small, and when the coefficient absolute value | L | is large, the code length is 1641b to 1641f. It is a table that gradually becomes shorter. That is, the variable length code tables 1641b to 1641f are tables in which the change rate of the code length with respect to the absolute value | L | of the coefficient gradually decreases in the order of 1641b to 1641f. Used when the value | L | is, for example, “4” to “6”, but applied when the absolute value | L | of the coefficient is, for example, “7” to “12” in the variable-length code table 1641c. It is suitable for each.

  As a result, a variable-length code having a code length corresponding to the coefficient value can be adapted to each table, so that the coding efficiency can be improved. In other words, when a coefficient value is small in one table, it is encoded to a variable length code having a shorter code length than the other table, and when a coefficient value is large in another table, the code length is shorter than that of the other table. By encoding to a variable length code and switching the table according to the coefficient value, the code length can be drastically shortened. Furthermore, since a range in which the code length is shortened can be assigned to each table, the encoding efficiency can be improved. Since the encoding is non-arithmetic encoding, that is, the VLC method, complicated processing such as arithmetic encoding is not required, and once a table used for encoding is determined, it is easy to simply refer to that table for variable length codes. Can be encoded.

  FIG. 26 is a diagram illustrating a configuration example of the threshold value table 1642.

  The threshold value table 1642 is preset according to the characteristics of the variable length code tables 1641a to 1641g, and holds a plurality of threshold values used for switching the variable length code tables 1641a to 1641f. For example, the threshold value for switching the variable length code tables 1641a and 1641b is “4”, the threshold value for switching (transition) of the variable length code tables 1641b and 1641c is “7”,..., The variable length code table 1641f , 1641g is set to a threshold value “193”. This makes it possible to easily determine the table switching timing and to switch to the optimum table corresponding to the absolute value | L | of the coefficients.

  The code allocation unit 163 uses the variable length code tables 1641a to 1641g and the threshold value table 1642 held by the table storage unit 164 to calculate the absolute value | L | of the coefficient output from the R column / L column generation unit 161. Variable length coding is performed separately from the R sequence, and a binary code is assigned. That is, the code allocation unit 163 performs one-dimensional encoding on the absolute value | L | of the coefficient.

  Next, an encoding operation in the encoding device 100c will be described. Since the operations of the block transform unit 110 to the quantization unit 130 are the same as the operations in the above-described encoding devices 100a and 100b, the description of the operation is omitted, and the variable-length encoding unit 160 performs variable-length encoding. The operation will be described in detail.

  The frequency coefficient quantized by the quantization unit 130 is input to the R-sequence / L-sequence generation unit 161 of the variable length encoding unit 160.

  First, as in the case of FIG. 3B, the R / L column generation unit 161 first zigzags scans the quantized frequency coefficient values in the block from the DC component region to the high frequency component region. Thus, the frequency coefficient value is made one-dimensional. Next, the R column / L column generation unit 161 includes a column of coefficient values L other than “0” (hereinafter also referred to as “L column”) and a column of the number R of consecutive coefficient values “0” (hereinafter, “L”). Are also created individually. An example of the generated L and R columns is shown in FIG. The L column is divided into the number m of coefficients, the absolute value | L | of the coefficients, the sign of the coefficient, and the like. For the positive / negative sign of the coefficient, for example, positive is associated with “0” and negative is associated with “1”.

  Here, since the coefficient value generally tends to be “0” as the frequency component is high, the coefficient value of the L column approaches “1” by scanning from the low frequency region toward the high frequency region.

  The code allocation unit 163 encodes each L value of the L sequence generated by the R sequence / L sequence generation unit 161 in the reverse order to the zigzag scan, that is, in order from the highest frequency. That is, the code assigning unit 163 sequentially obtains a Huffman code (variable length code) corresponding to the absolute value | L | of the coefficient using the variable length code tables 1641a to 1641g in order from the rear of the L column.

  The reason why the order of the zigzag scanning is reversed is that the convergence is converged to the vicinity of the coefficient value “1” other than “0” in the high frequency region, and the first table used for encoding can be easily determined. This is because it is easy to create ~ 1641g and to easily determine the threshold value.

  The code assigning unit 163 assigns a variable length code to each L in the L sequence and each R in the R sequence using various tables held in the table storage unit 164. The code assigning unit 163 assigns a variable length code for the number m of coefficients. Here, a process for assigning a variable length code to the absolute value | L | of the coefficient will be described.

  FIG. 27 is a flowchart illustrating variable length code allocation processing executed by the code allocation unit 163.

The code assigning unit 163 sets the number m of coefficient values output from the R column / L column generating unit 161 when starting to encode the intra-block coefficient value (absolute value | L |) of the block (S101). Next, the code assignment unit 163 sets “0” to the table number k as an initial value of the variable-length code table to be referred to (S102). Next, the code assignment unit 163 refers to the threshold value table 1642 and sets threshold value = 4 (S103).

  When the setting of the number m of coefficients, the reference destination of the variable length code table (in this case, the variable length code table 1641a), and the threshold value is completed, the code assignment unit 163 outputs the R sequence / L sequence generation unit 161. The absolute value | L | of the coefficient is read from the back (S104), and the absolute value | L | of the coefficient read using the variable length code table of the set number is encoded into a variable length code (S105). When the encoding is completed, the code assigning unit 163 stores the binary code obtained by the encoding in a buffer (not shown) (for example, a FIFO buffer) (S106), and sets the number m of coefficients to “1”. It is decremented (S107), and it is determined whether or not the decremented number m is “0”, that is, whether or not all the coefficients included in the L column have been encoded (S108).

  If the number m of coefficients is not “0” (No in S108), it is determined whether or not the absolute value | L | of the immediately preceding coefficient exceeds a threshold value (S109). If not exceeded (No in S109), the code allocation unit 163 reads the absolute value | L | of the next coefficient from the back (S104), and executes steps S105 to S108 and the like. That is, the absolute value | L | of the next coefficient is encoded using the same variable length code table as before.

  When the absolute value | L | of the immediately preceding coefficient exceeds the threshold value (No in S109), the code assignment unit 163 increments the table number k by “1” (S110). As a result, when encoding the absolute value | L | of the next coefficient, the variable length code table suitable for encoding the absolute value | L | (For example, if the previous variable-length code table is 1641a with k = 0, 1641b with k = 1) is referred to.

  When the table number k is incremented, the code assignment unit 163 refers to the threshold value table 1642 and updates it to the next threshold value (for example, “7” if the previous threshold value is “4”). (S111). As a result, only when the absolute value | L | of the coefficient exceeds a new threshold value, the rate of change of the code length is smaller, and the following is suitable for encoding the absolute value | L | Transition to the variable length code table is possible.

  More specifically, the absolute value | L | of the immediately preceding coefficient exceeds the threshold value “4” between the variable length code table 1641a with the table number “0” and the variable length code table 1641b with the table number “1”. In this case, as shown in FIG. 28, the reference destination is switched from the variable-length code table 1641a to the variable-length code table 1641b as a reference destination when the absolute value | L | of the next coefficient is encoded. “7” is set.

  Even when the absolute value | L | of the immediately preceding coefficient exceeds the thresholds “7” to “193” between the variable length code tables 1641b to 1641g of the table numbers “1” to “6”, the threshold “ As in the case of “4”, as a reference destination in encoding the absolute value | L | of the next coefficient, the variable length code table 1641b of the table number “1” is changed to the variable length code table 1641c of the table number 2. ..., sequentially switched to the variable length code table 1641g. This is shown in FIG.

  Here, the table switching direction is one direction and does not return. Accordingly, the table is prevented from being frequently replaced by the coefficient value, and the number of table switching is reduced. Therefore, encoding efficiency can be improved. For example, since the work area of the memory is limited, it is common to place only the next table to be used in the work area. In this case, every time the table is switched, it takes time to read the next table from the ROM and expand it in the work area, so it takes time to encode the next coefficient value. For this reason, the effect of reducing the total number of times until the next coefficient value is encoded can be achieved by limiting the number of table switching operations in one direction.

  When the increment of the table number and the updating of the threshold value are completed, the code assignment unit 163 reads the absolute value | L | of the next coefficient from the back (S104), and executes steps S105 to S108, etc. . That is, encoding is performed using a variable-length code table suitable for a coefficient whose absolute value | L | is larger than before.

  Such processing (S104 to S111) is repeatedly executed until the number m of coefficients becomes “0”. When the number m of coefficients becomes “0”, the absolute value | L | of the coefficient in the block is encoded. finish.

  Specifically, the columns of the absolute value | L | of the coefficients in the block are “1”, “1”, “2”, “3”, “4”, “12”, “2”, “3” from the back. ”,“ 31 ”,“ 22 ”,“ 5 ”,“ 9 ”,“ 38 ”, the code assigning unit 163 first uses the variable length code table 1641a as shown in FIG. ”,“ 1 ”,“ 2 ”,“ 3 ”,“ 4 ”,“ 12 ”in this order, binary codes“ 1 ”,“ 1 ”,“ 010 ”,“ 011 ”,“ 00100 ”,“ 0001100 ” Are encoded respectively. Since the threshold value “4” was exceeded when coding the absolute value of the coefficient | L | = “12”, the code allocating unit 163 sets the table used for encoding to the variable length of the table number k = 1. Switch to the code table 1641b.

  Next, the code assigning unit 163 uses the switched variable length code table 1641b to convert the absolute value | L | = “2”, “3”, “31” of the next coefficient into the binary code “11”, It encodes to “0100” and “0000100000”, respectively. Since the threshold value “7” was exceeded when coding the absolute value of the coefficient | L | = “31”, the code allocation unit 163 sets the table used for encoding to a variable length of the table number k = 2. Switch to the code table 1641c.

  Further, the code assigning unit 163 encodes the absolute value | L | = “22” of the next coefficient into the binarized code “0011001” using the switched variable length code table 1641c. Since the threshold value “13” has been exceeded when coding the absolute value of the coefficient | L | = “22”, the code allocation unit 163 sets the table used for encoding to the variable length of the table number k = 3. Switch to the code table 1641d.

  Next, the code allocating unit 163 uses the variable length code table 1641d to switch the absolute value | L | = “5”, “9”, “38” of the next coefficient in this order using the variable length code table 1641d. The binary codes “1100”, “010000”, and “00101101” are respectively encoded.

  As a result, the binarization code “1101001100100000110011010000000001000000000011001110001000000101101” is stored in the buffer.

  The buffer also stores the number m of encoded L-sequence coefficients, the sign of the coefficient, and the R-value binary code of the R-sequence, and the encoded L-sequence stored in the buffer. The number of coefficients m, the binarization code of the absolute value | L |, the sign of the coefficient, and the R value binarization code of the R column are a recording medium such as a CD or a transmission medium such as the Internet or satellite broadcasting. To the decoding device.

  Here, the absolute value of the L-column coefficients | L | “1”, “1”, “2”, “3”, “4”, “12”, “2”, “3”, “31”, Assuming that “22”, “5”, “9”, and “38” are encoded using only the variable length code table 1641a, the binarized codes are “1”, “1”, “010”, “ “011”, “00100”, “0001100”, “010”, “011”, “0000011111”, “0000110110”, “00101”, “001001”, “0000010110”, which is a 64-bit code length.

  On the other hand, in the encoding method of the fifth embodiment, the maximum value of the absolute value | L | of the coefficient in the block is relatively low and the absolute value | L | of the coefficient does not gradually increase. However, the code length is 61 bits, and the encoding efficiency can be increased. This is because when the absolute value of the coefficient | L | is, for example, “22” or “38”, the variable-length code table 1641a requires 9 bits “000010110” and 11 bits “00000100110”. In this method, 7 bits of “0011001” and 8 bits of “00101101” contribute greatly. Therefore, when the maximum value of the absolute value | L | of a coefficient in a normal block is relatively high and the absolute value | L | of the coefficient gradually increases, the coding efficiency can be dramatically increased. it can.

  In the fifth embodiment, when the absolute value | L | of the immediately preceding coefficient exceeds the threshold value (Yes in S109), the table number k is incremented by 1 (S110), and the variable length of the next number is changed. The encoding is performed using the code table (see FIG. 28), but the variable length code table adapted to the absolute value | L | of the coefficient by the absolute value | L | of the coefficient immediately before exceeding the threshold value You may make it jump and switch. That is, for example, when the absolute value | L | of the coefficient immediately before encoding with reference to the table of the number k = 1 is “20”, the absolute value | L | of the next coefficient is larger than “20”. Since the possibility is high, the absolute value | L | of the next coefficient may be encoded by referring to a table of k = 3. In this case, the threshold value (for example, 25) that matches the variable length code table may be set.

  Moreover, although the case where eight variable-length code tables were used was demonstrated, this used two to seven, or eight or more variable-length code tables, and exceeded each threshold value using a plurality of threshold values. The variable length code table may be switched each time.

  In the fifth embodiment, the coefficient value is encoded by dividing it into an absolute value and a positive / negative sign, and each variable length code table for the absolute value of the coefficient value is configured with no positive / negative sign (absolute value). A numerical value may be encoded with a positive / negative sign, and in that case, a binary code may be configured with a positive / negative sign. In this case, for example, one bit of the plus / minus code may be added to the LSB bit of the variable length code.

  In the fifth embodiment, the case where an image is encoded by intra-frame encoding has been described. However, this is also the case where inter-frame encoding is performed using motion compensation or the like for moving image input. The same effect can be obtained by the method of this embodiment.

  In the fifth embodiment, the case where the input image is divided into blocks each having 4 horizontal pixels and 4 vertical pixels has been described. However, the size of the blocks may be other sizes.

  In the fifth embodiment, the scanning method in the block has been described with reference to FIG. 3B. However, this may be another scanning order as long as scanning from a low frequency region to a high frequency region. good.

  Further, although the fifth embodiment has been described with reference to FIG. 25 as an example of the variable length code table, it may be another table.

  In the fifth embodiment, the case where the number of L values is added to the beginning of the L column has been described. However, EOB may be added to the end of the L column.

(Embodiment 6)
FIG. 30 is a block diagram showing a functional configuration of a decoding apparatus to which the variable length decoding method and the video decoding method using the same according to the embodiment of the present invention are applied. Here, a description will be given assuming that a code string generated by the variable length coding method of the present invention described in Embodiment 5 is input.

  As illustrated in FIG. 30, the decoding device 500c includes a variable length decoding unit 560, an inverse quantization unit 520, an inverse frequency conversion unit 530, and a frame memory 540. In addition, each part which comprises such a decoding apparatus 500c provides a work area at the time of program execution, ROM which stores CPU and the program and data which are beforehand executed by CPU similarly to the encoding apparatus 100c, This is realized by a memory or the like that temporarily stores an input coded sequence or the like. Further, since the inverse quantization unit 520, the inverse frequency conversion unit 530, and the frame memory 540 are the same as those of the decoding devices 500a and 500b, the description thereof will be omitted, and the configuration of the variable length decoding unit 560 will be described. Is described in detail.

  The variable length decoding unit 560 includes a code conversion unit 561, a table storage unit 562, and a coefficient generation unit 564.

  The table storage unit 562 stores a plurality of types (eight) of variable length code (decoding) tables 5621a to 5621g in which variable length codes and absolute values | L | of coefficients are associated with each other, a threshold value table 5622, and the like. Hold. The variable length code tables 5621a to 5621g are configured the same as the variable length code tables 1641a to 1641g shown in FIG. 25, and the threshold value table 5622 is configured the same as the threshold value table 1642 shown in FIG. Is done.

  The code conversion unit 561 uses the tables (variable length code tables 5621a to 5621g, threshold value table 5622, and the like) held in the table storage unit 562 to convert the variable length code to L The number m of coefficients in the column, the absolute value of the coefficient | L |, and the R value in the R column are converted. Conversion to the absolute value | L | of the coefficient is performed using the variable length code tables 5621a to 5621g.

  The coefficient generation unit 564 converts the coefficient values into coefficient values based on the input L and R columns, and two-dimensionalizes them using a predetermined scanning method. In the conversion to the coefficient value, the coefficient value “0” is generated by the value indicated by R based on a predetermined scanning order, and then the coefficient value of the value indicated by L is generated. Are repeated to convert the L and R columns into coefficient values. Here, if the zigzag is scanned from the low frequency region toward the high frequency region, the coefficient block shown in FIG. 11 is converted. The generated coefficient block is input to the inverse quantization unit 520.

  Next, the decoding operation in each unit of the variable length decoding unit 560 will be described.

  Here, the codes of the binarized code input code string input to the code conversion unit 561 are “1”, “1”, “010”, “011”, “00100”, “0001100”, in order from the top. The description will be made assuming that “11”, “0100”, “0000100000”, “0011001”, “1100”, “010000”, and “00101101”.

  When starting decoding of the variable length code, the code conversion unit 561 decodes the number m of coefficients output from the encoding device 100c, and sets the number m of decoded coefficients. Next, the code conversion unit 561 sets “0” to the table number k as an initial value of the variable-length code table to be referred to. Next, the code assignment unit 163 refers to the threshold value table 5622 and sets threshold value = 4 (S103). When the setting of the number m of coefficients, the reference destination of the variable-length code table (in this case, the variable-length code table 5621a) and the threshold value is completed, the code conversion unit 561 sets the coefficients in the order output from the encoding device 100c. The absolute value | L | is read from the front (that is, the high frequency side), and the variable length code read using the variable length code table of the set number is subjected to variable length decoding to the absolute value | L | of the coefficient. When decoding is completed, the code conversion unit 561 stores the absolute value | L | of the coefficient obtained by decoding in a buffer (not shown) (for example, a FILO buffer), and sets the number m of coefficients to “1”. It is decremented and it is determined whether or not the number m after decrement is “0”, that is, whether or not all the coefficients included in the L column have been decoded.

  If the number m of coefficients is not “0”, it is determined whether or not the absolute value | L | of the coefficient immediately before variable length decoding exceeds a threshold value. If not exceeded, the code conversion unit 561 reads the next variable length code from the front and decodes it to the absolute value | L | of the coefficient using the same variable length code table as before.

  When the absolute value | L | of the coefficient immediately before the variable length decoding exceeds the threshold value, the code conversion unit 561 increments the table number k by “1”. As a result, when encoding the absolute value | L | of the next coefficient, the variable length code table suitable for encoding the absolute value | L | (For example, if the previous variable-length code table is 5621a, 5621b) will be referred to. When the table number k has been incremented, the code conversion unit 561 refers to the threshold value table 5622 and updates it to the next threshold value (for example, “7” if the previous threshold value is “4”). ). Thus, only when the absolute value | L | of the coefficient exceeds a new threshold value, the next variable length suitable for decoding the variable length code of the coefficient having a long code length with a smaller code length change rate. Transition to the code table is possible.

  More specifically, the variable length code table 5621a of k = 0 is referred to for the first variable length code. Now, referring to the variable length table 5621a, the variable code “1” matches the input code string, and the absolute value | L | of the coefficient in this case is “1”. Similarly, when conversion from the variable length code to the absolute value | L | of the coefficient is sequentially performed using the variable length code table 5621a, the variable length code “1” becomes the absolute value of the coefficient | L | = “ 1 ”, the variable length code“ 010 ”is the coefficient absolute value | L | =“ 3 ”, the variable length code“ 00100 ”is the coefficient absolute value | L | =“ 4 ”, and the variable length code“ 0001100 ”. Are converted into the absolute value of the coefficient | L | = “12”, respectively.

  If the threshold value for the absolute value | L | of the coefficient is “4”, the absolute value | L | = “12” of the sixth coefficient is exceeded. Accordingly, the code conversion unit 561 uses the next variable length code table 5621b in which k = 1 in the subsequent conversion to the absolute value | L | of the coefficient of the variable length code, sets 7 as the threshold, The absolute value | L | of the coefficient is converted. Therefore, the next seventh variable length code “11” is converted to the absolute value of the coefficient | L | = “2”.

  The next eighth variable length code “0100” is converted to the coefficient absolute value | L | = “3”, and the ninth variable length code “0000010000” is converted to the coefficient absolute value | L | = “31”. Is done. Here, if the threshold value for the absolute value | L | of the coefficient is “7”, the absolute value | L | = “31” of the ninth coefficient is exceeded. Therefore, the code conversion unit 561 uses the next variable length code table 5621b in which k = 2 in the subsequent conversion to the absolute value | L | of the coefficient of the variable length code, sets 13 as the threshold, The absolute value | L | of the coefficient is converted. Even if the absolute value | L | of the coefficient obtained in the decoding to the absolute value | L | of the seventh coefficient is equal to or less than the threshold value 7 again, switching to the variable-length code table 5621a is not performed. Conversion is performed using the variable length code table 5621b.

  When the above processing is repeated and absolute values | L | of coefficients for one block (m) are generated, they are rearranged in reverse order by first-in last-out by the FILO buffer. Also, the positive and negative signs of the coefficients are rearranged in the reverse order by the first-in last-out by the FILO buffer. However, the number is excluded from the target of sorting. Here, it is assumed that a column having the same order as the L column shown in FIG. 24A (that is, the sequence from low frequency to high frequency) is generated. The absolute value | L | of each coefficient of the L columns rearranged in this way is input to the coefficient generation unit 564. Note that the code conversion unit 561 decodes each R in the R sequence by the same processing as the coefficient absolute value | L |, and outputs the R sequence shown in FIG. 24A to the coefficient generation unit 564.

  The coefficient generation unit 564 converts the coefficient values into coefficient values based on the input L and R columns. At this time, based on a predetermined scanning order, the coefficient value “0” is generated by the value indicated by R, and then the coefficient value of the value indicated by L is generated by adding a positive / negative sign. Are repeated to convert the L and R columns into coefficient values. Here, when the zigzag scanning is performed from the low frequency region to the high frequency region as shown in FIG. 11, the L column shown in FIG. 24A and the R column shown in FIG. 24 are converted into coefficient blocks. Will be. The generated coefficient block is input to the inverse quantization unit 520.

  As described above, in the variable length decoding method according to Embodiment 6 of the present invention, first, in the decoding step, a plurality of variable length code (decoding) tables used for decoding are switched while switching in one direction. Using the variable length code table, the variable length codes of the code string are decoded into coefficient values other than “0” in the frequency domain in a predetermined order. Next, in the coefficient value conversion step, the coefficient values in the block are converted based on the generated coefficient values. Here, each table has a coefficient value such that the code length at the minimum coefficient value is increased in the order of the numbers assigned to each table, and the code length at the maximum coefficient value is decreased in the order of the numbers. The rate of change in code length with respect to is different.

  The threshold value is set based on the adaptive characteristics of each table in which the code length corresponding to the coefficient value is shorter than that of the other tables. The variable length codes of the code string are arranged in the order from the high frequency component to the low frequency component, and the decoding step decodes the coefficient value in the order of the code string arrangement, and the decoded coefficient value of the code string A coefficient value sequence of coefficient values is generated by sequentially outputting from the rear of the arrangement, and the coefficient generation step scans the coefficient values in the order of arrangement of the coefficient value sequences.

  When the absolute value of the decoded coefficient exceeds a threshold value, the encoding step converts the table used for decoding the variable length code to be decoded into a table having a number larger than the number assigned to the table. Switch to decode the next variable length code into coefficient values.

  With the above operation, by using the variable length decoding method of the present invention, it is possible to correctly decode a code string encoded using the variable length coding method of the present invention.

  In the sixth embodiment, when the absolute value | L | of the immediately preceding coefficient decoded in the same manner as in the fifth embodiment exceeds the threshold value, one table number k is set. Increment and decode using the variable number code table of the next number, but if the table switching method used at the time of encoding is the same, the absolute value of the coefficient immediately before the threshold is exceeded Depending on the value | L |, the variable length code table adapted to the absolute value | L | In this case, a threshold value that matches the variable length code table may be set.

  In the sixth embodiment, the variable length code table has been described with reference to FIG. 25. However, this may be another table as long as it is the same as the table used for encoding. Moreover, although the case where eight variable-length code tables were used was demonstrated, this used two to six or eight or more variable-length code tables, and exceeded each threshold value using a plurality of threshold values. The variable length code table may be switched each time. However, the configuration and threshold value of the variable length table at this time must be the same as those used for encoding.

  In the sixth embodiment, the coefficient value is encoded by dividing it into an absolute value and a positive / negative code, and each variable length code table for the absolute value of the coefficient value is configured with no positive / negative code (absolute value). The numerical value may be encoded with a positive / negative sign, and in that case, the binarized code may be configured with a positive / negative sign. In this case, for example, one bit of the plus / minus code may be added to the LSB bit of the variable length code.

  In the sixth embodiment, the variable length code table is switched when the value of L exceeds the threshold value. However, the order of the absolute value of the coefficient | L | In order), when the absolute value | L | of the decoded coefficient is less than the threshold value, the variable-length code with the smaller number is decoded. You may make it switch to a table.

  Further, in the sixth embodiment, the case where the code string generated by the intra-frame coding is decoded has been described, but this is performed by performing inter-frame coding using motion compensation or the like on the moving image input. The generated code string may be decoded, and the same effect can be obtained by the method of the present embodiment.

  In the sixth embodiment, the case where the input image is divided and encoded into blocks of 4 horizontal pixels and 4 vertical pixels has been described. However, the size of the blocks may be other sizes. .

  In the sixth embodiment, the scanning method in the block has been described with reference to FIG. 11. However, this may be another scanning order as long as it is the same as the scanning method used at the time of encoding. .

(Embodiment 7)
Next, the variable length coding method, variable length decoding method, variable length coding device, variable length decoding device, moving image coding method, moving image decoding method, moving image coding device, and moving image according to the present invention An example in which the image decoding apparatus is realized in another form will be described.

  By recording a program for realizing the configuration of the encoding device or the decoding device shown in each of the above embodiments on a storage medium such as a flexible disk, the processing shown in each of the above embodiments is performed. It can be easily implemented in an independent computer system.

  FIG. 32 shows a flexible disk storing a program for performing the processing of the moving picture encoding method of the first, third, and fifth embodiments or the moving picture decoding method of the second, fourth, and sixth embodiments. It is explanatory drawing in the case of implementing by a computer system.

  FIG. 32B shows the appearance, cross-sectional structure, and flexible disk as seen from the front of the flexible disk, and FIG. 32A shows an example of the physical format of the flexible disk that is the recording medium body. The flexible disk FD is built in the case F, and a plurality of tracks Tr are formed concentrically on the surface of the disk from the outer periphery toward the inner periphery, and each track is divided into 16 sectors Se in the angular direction. ing. Therefore, in the flexible disk storing the program, a moving image encoding device as the program is recorded in an area allocated on the flexible disk FD.

  FIG. 32C shows a configuration for recording and reproducing the program on the flexible disk FD. When the program is recorded on the flexible disk FD, the moving picture encoding apparatus or moving picture decoding apparatus as the program is written from the computer system Cs via the flexible disk drive. Further, when the moving picture coding apparatus is constructed in a computer system by a program in a flexible disk, the program is read from the flexible disk by a flexible disk drive and transferred to the computer system.

  In the above description, a flexible disk is used as the recording medium, but the same can be done using an optical disk. Further, the recording medium is not limited to this, and any recording medium such as an IC card or a ROM cassette capable of recording a program can be similarly implemented.

  Furthermore, application examples of the moving picture prediction method, the moving picture encoding apparatus, and the moving picture decoding apparatus described in the above embodiment and a system using the same will be described.

  FIG. 33 is a block diagram showing an overall configuration of a content supply system ex100 that implements a content distribution service. The communication service providing area is divided into desired sizes, and base stations ex107 to ex110, which are fixed radio stations, are installed in each cell.

  The content supply system ex100 includes, for example, a computer ex111, a PDA (personal digital assistant) ex112, a camera ex113, a mobile phone ex114, a camera via the Internet ex101, the Internet service provider ex102, the telephone network ex104, and the base stations ex107 to ex110. Each device such as the attached mobile phone ex115 is connected.

  However, the content supply system ex100 is not limited to the combination as shown in FIG. 33, and any of the combinations may be connected. Further, each device may be directly connected to the telephone network ex104 without going through the base stations ex107 to ex110 which are fixed wireless stations.

  The camera ex113 is a device capable of shooting a moving image such as a digital video camera. The mobile phone is a PDC (Personal Digital Communications) system, a CDMA (Code Division Multiple Access) system, a W-CDMA (Wideband-Code Division Multiple Access) system, or a GSM (Global System for Mobile Communications) system mobile phone, Alternatively, PHS (Personal Handyphone System) or the like may be used.

  In addition, the streaming server ex103 is connected from the camera ex113 through the base station ex109 and the telephone network ex104, and live distribution or the like based on the encoded data transmitted by the user using the camera ex113 becomes possible. The encoded processing of the captured data may be performed by the camera ex113 or may be performed by a server or the like that performs data transmission processing.

  Further, the moving image data shot by the camera 116 may be transmitted to the streaming server ex103 via the computer ex111. The camera ex116 is a device that can shoot still images and moving images, such as a digital camera. In this case, the encoding of the moving image data may be performed by the camera ex116 or the computer ex111. The encoding process is performed in the LSI ex117 included in the computer ex111 and the camera ex116. Note that image encoding / decoding software may be incorporated into any storage medium (CD-ROM, flexible disk, hard disk, etc.) that is a recording medium readable by the computer ex111 or the like. Furthermore, you may transmit moving image data with the mobile telephone ex115 with a camera. The moving image data at this time is data encoded by the LSI included in the mobile phone ex115.

  In this content supply system ex100, the content (for example, video shot of music live) captured by the user with the camera ex113, camera ex116, etc. is encoded and transmitted to the streaming server ex103 as in the above embodiment. On the other hand, the streaming server ex103 distributes the content data to the requested client. Examples of the client include a computer ex111, a PDA ex112, a camera ex113, a mobile phone ex114, and the like that can decode the encoded data. In this way, the content supply system ex100 can receive and reproduce the encoded data at the client, and also realize personal broadcasting by receiving, decoding, and reproducing in real time at the client. It is a system that becomes possible.

  The image encoding device or the image decoding device described in the above embodiments may be used for encoding and decoding of each device constituting this system.

  A mobile phone will be described as an example.

  FIG. 34 is a diagram illustrating the mobile phone ex115 using the video prediction method, the video encoding device, and the image decoding device described in the above embodiment. The cellular phone ex115 includes an antenna ex201 for transmitting and receiving radio waves to and from the base station ex110, a camera such as a CCD camera, a camera unit ex203 capable of taking a still image, a video shot by the camera unit ex203, and an antenna ex201. A display unit ex202 such as a liquid crystal display for displaying data obtained by decoding received video and the like, a main body unit composed of a group of operation keys ex204, an audio output unit ex208 such as a speaker for audio output, and audio input To store encoded data or decoded data such as a voice input unit ex205 such as a microphone, captured video or still image data, received mail data, video data or still image data, etc. Recording medium ex207, and slot portion ex206 for enabling the recording medium ex207 to be attached to the mobile phone ex115. Have. The recording medium ex207 stores a flash memory element which is a kind of EEPROM (Electrically Erasable and Programmable Read Only Memory) which is a nonvolatile memory that can be electrically rewritten and erased in a plastic case such as an SD card.

  Further, the cellular phone ex115 will be described with reference to FIG. The cellular phone ex115 controls the power supply circuit unit ex310, the operation input control unit ex304, and the image encoding with respect to the main control unit ex311 configured to comprehensively control each unit of the main body unit including the display unit ex202 and the operation key ex204. Unit ex312, camera interface unit ex303, LCD (Liquid Crystal Display) control unit ex302, image decoding unit ex309, demultiplexing unit ex308, recording / reproducing unit ex307, modulation / demodulation circuit unit ex306, and audio processing unit ex305 via a synchronization bus ex313 Are connected to each other.

  When the end of call and the power key are turned on by a user operation, the power supply circuit ex310 starts up the camera-equipped digital mobile phone ex115 in an operable state by supplying power from the battery pack to each unit. .

  The cellular phone ex115 converts the audio signal collected by the audio input unit ex205 in the voice call mode into digital audio data by the audio processing unit ex305 based on the control of the main control unit ex311 including a CPU, a ROM, a RAM, and the like. The modulation / demodulation circuit unit ex306 performs spread spectrum processing, and the transmission / reception circuit unit ex301 performs digital analog conversion processing and frequency conversion processing, and then transmits the result via the antenna ex201. In addition, the cellular phone ex115 amplifies the received signal received by the antenna ex201 in the voice call mode, performs frequency conversion processing and analog-digital conversion processing, performs spectrum despreading processing by the modulation / demodulation circuit unit ex306, and analog audio by the voice processing unit ex305 After being converted to a signal, this is output via the audio output unit ex208.

  Further, when an e-mail is transmitted in the data communication mode, text data of the e-mail input by operating the operation key ex204 on the main body is sent to the main control unit ex311 via the operation input control unit ex304. The main control unit ex311 performs spread spectrum processing on the text data in the modulation / demodulation circuit unit ex306, performs digital analog conversion processing and frequency conversion processing in the transmission / reception circuit unit ex301, and then transmits the text data to the base station ex110 via the antenna ex201.

  When transmitting image data in the data communication mode, the image data captured by the camera unit ex203 is supplied to the image encoding unit ex312 via the camera interface unit ex303. When image data is not transmitted, the image data captured by the camera unit ex203 can be directly displayed on the display unit ex202 via the camera interface unit ex303 and the LCD control unit ex302.

  The image encoding unit ex312 has a configuration including the image encoding device described in the present invention, and an encoding method using the image data supplied from the camera unit ex203 in the image encoding device described in the above embodiment. The encoded image data is converted into encoded image data by compression encoding, and sent to the demultiplexing unit ex308. At the same time, the cellular phone ex115 sends the sound collected by the audio input unit ex205 during imaging by the camera unit ex203 to the demultiplexing unit ex308 as digital audio data via the audio processing unit ex305.

  The demultiplexing unit ex308 multiplexes the encoded image data supplied from the image encoding unit ex312 and the audio data supplied from the audio processing unit ex305 by a predetermined method, and the multiplexed data obtained as a result is a modulation / demodulation circuit unit Spread spectrum processing is performed in ex306, digital analog conversion processing and frequency conversion processing are performed in the transmission / reception circuit unit ex301, and then transmitted via the antenna ex201.

  When receiving data of a moving image file linked to a home page or the like in the data communication mode, the received signal received from the base station ex110 via the antenna ex201 is subjected to spectrum despreading processing by the modulation / demodulation circuit unit ex306, and the resulting multiplexing is obtained. Is sent to the demultiplexing unit ex308.

  In addition, in order to decode the multiplexed data received via the antenna ex201, the demultiplexing unit ex308 separates the multiplexed data to generate an encoded bitstream of image data and an encoded bitstream of audio data. The encoded image data is supplied to the image decoding unit ex309 via the synchronization bus ex313, and the audio data is supplied to the audio processing unit ex305.

  Next, the image decoding unit ex309 is configured to include the image decoding device described in the present invention, and a decoding method corresponding to the encoding method described in the above embodiment for an encoded bit stream of image data. To generate playback moving image data, which is supplied to the display unit ex202 via the LCD control unit ex302, thereby displaying, for example, moving image data included in the moving image file linked to the homepage . At the same time, the audio processing unit ex305 converts the audio data into an analog audio signal, and then supplies the analog audio signal to the audio output unit ex208. Thus, for example, the audio data included in the moving image file linked to the home page is reproduced. The

  Note that the present invention is not limited to the above-described system, and recently, digital broadcasting using satellites and terrestrial waves has become a hot topic. As shown in FIG. 36, at least the image encoding device or image of the above-described embodiment is included in a digital broadcasting system. Any of the decoding devices can be incorporated. Specifically, in the broadcasting station ex409, the encoded bit stream of the video information is transmitted to the communication or broadcasting satellite ex410 via radio waves. Receiving this, the broadcasting satellite ex410 transmits a radio wave for broadcasting, and receives the radio wave with a home antenna ex406 having a satellite broadcasting receiving facility, such as a television (receiver) ex401 or a set top box (STB) ex407. The device decodes the encoded bit stream and reproduces it.

  In addition, the image decoding apparatus described in the above embodiment can also be implemented in a playback apparatus ex403 that reads and decodes an encoded bitstream recorded on a storage medium ex402 such as a CD or DVD that is a recording medium. is there. In this case, the reproduced video signal is displayed on the monitor ex404. Further, a configuration in which an image decoding device is mounted in a set-top box ex407 connected to a cable ex405 for cable television or an antenna ex406 for satellite / terrestrial broadcasting, and this is reproduced on the monitor ex408 of the television is also conceivable. At this time, the image decoding apparatus may be incorporated in the television instead of the set top box. It is also possible to receive a signal from the satellite ex410 or the base station ex107 by the car ex412 having the antenna ex411 and reproduce a moving image on a display device such as the car navigation ex413 that the car ex412 has.

  Furthermore, the image signal can be encoded by the image encoding device shown in the above embodiment and recorded on a recording medium. As a specific example, there is a recorder ex420 such as a DVD recorder that records an image signal on a DVD disk ex421 or a disk recorder that records on a hard disk. Further, it can be recorded on the SD card ex422. If the recorder ex420 includes the image decoding device described in the above embodiment, the image signal recorded on the DVD disc ex421 or the SD card ex422 can be reproduced and displayed on the monitor ex408.

  The configuration of the car navigation ex 413 is, for example, the configuration shown in FIG. 36 excluding the camera unit ex203, the camera interface unit ex303, and the image encoding unit ex312. The same applies to the computer ex111 and the television (receiver). ) Ex401 can also be considered.

  In addition to the transmission / reception type terminal having both the encoder and the decoder, the terminal such as the mobile phone ex114 has three mounting formats: a transmitting terminal having only an encoder and a receiving terminal having only a decoder. Can be considered.

  As described above, the variable length coding method, the variable length decoding method, the variable length coding apparatus, the variable length decoding apparatus, the moving picture coding method, and the moving picture decoding using these methods described in the above embodiments. It is possible to use the encoding method, the moving image encoding device, and the image decoding device in any of the devices / systems described above, and by doing so, the effects described in the above embodiments can be obtained.

  In addition, about all embodiment of this invention, this invention is not limited to this said embodiment, A various deformation | transformation or correction is possible without deviating from the scope of the present invention.

  In the variable length coding apparatus and the variable length decoding apparatus of the first to sixth embodiments, the coefficient values are scanned in the order from the low frequency component to the high frequency component. However, as a modification of the present invention, The coefficient values may be scanned in the order from the high frequency component to the low frequency component. In this case, processing such as rearrangement of coefficient values can be omitted.

  The variable length coding method, variable length coding apparatus, and the like of the present invention can be applied to a computer system that compresses and encodes moving images, a mobile phone, and the like.

100c encoder 110 block converter 120 frequency converter 130 quantizer 160 variable length encoder 161 R sequence / L sequence generator 163 code allocation unit 164 table storage unit 500c decoder 560 variable length decoder 561 code Conversion unit 562 Table storage unit 564 Coefficient generation unit 520 Inverse quantization unit 530 Inverse frequency conversion unit 540 Frame memory 1641a to 1641g Variable length code table 1642 Threshold table

Claims (3)

A coding method for variable-length coding coefficient values of a two-dimensional sequence of a block obtained by frequency-converting image data of a block having a predetermined size,
A coefficient scanning step of changing the coefficient value of the two-dimensional column of the block to the coefficient value of the one-dimensional column by scanning the coefficient value of the two-dimensional column of the block in a predetermined order from the low frequency component to the high frequency component; ,
An RL sequence conversion step for converting the coefficient values of the one-dimensional sequence into a combination of a run value indicating the number R of consecutive coefficients having a value of zero and a level value indicating a non-zero coefficient value L that follows.
Determining the total number of non-zero coefficient values contained in the block;
A first encoding step of variable length encoding the run value;
A level value indicating the non-zero coefficient value L is sequentially variable-length encoded and encoded from a high frequency component to a low frequency component while switching between a plurality of tables to which different threshold values are assigned. A second encoding step performed until the number of non-zero coefficient values reaches the total number of non-zero coefficient values;
A third encoding step of variable length encoding the total number of non-zero coefficients included in the block;
In the second encoding step,
When the absolute value of the non-zero coefficient value L to be encoded exceeds the threshold value assigned to the table used for encoding the non-zero coefficient value L, the table to which a threshold value larger than the threshold value is assigned After being switched, the non-zero coefficient value L to be encoded is followed by a non-zero coefficient value L to be encoded next,
The encoding method according to claim 1, wherein the switching of the plurality of tables is a one-way switching in which a table used before the new switched table is not used again. An encoding device that performs variable length encoding on a coefficient value of a two-dimensional sequence of a block obtained by frequency-converting image data of a block having a predetermined size,
Coefficient scanning means for changing the coefficient value of the two-dimensional column of the block to the coefficient value of the one-dimensional column by scanning the coefficient value of the two-dimensional column of the block in a predetermined order from the low frequency component to the high frequency component; ,
RL sequence conversion means for converting the coefficient value of the one-dimensional sequence into a combination of a run value indicating the number of consecutive R coefficients having a value of zero and a level value indicating a subsequent non-zero coefficient value L;
Means for determining the total number of non-zero coefficient values included in the block;
First encoding means for variable-length encoding the run value;
The level value indicating the non-zero coefficient value L is sequentially variable-length encoded and encoded from the high frequency component toward the low frequency component while switching between a plurality of tables to which different threshold values are assigned. Second encoding means that executes until the number of non-zero coefficient values reaches the total number of non-zero coefficient values;
Third encoding means for variable length encoding the total number of non-zero coefficients included in the block;
The second encoding means includes
When the absolute value of the non-zero coefficient value L to be encoded exceeds the threshold value assigned to the table used for encoding the non-zero coefficient value L, the table to which a threshold value greater than the threshold value is assigned After being switched, the non-zero coefficient value L to be encoded is followed by a non-zero coefficient value L to be encoded next,
Switching between the plurality of tables is a one-way switching in which a table used before the new switched table is not used again. A program for variable-length encoding a coefficient value of a two-dimensional sequence of the block obtained by frequency-converting image data of a block having a predetermined size,
A coefficient scanning step of changing the coefficient value of the two-dimensional column of the block to the coefficient value of the one-dimensional column by scanning the coefficient value of the two-dimensional column of the block in a predetermined order from the low frequency component to the high frequency component; ,
An RL sequence conversion step for converting the coefficient values of the one-dimensional sequence into a combination of a run value indicating the number R of consecutive coefficients having a value of zero and a level value indicating a non-zero coefficient value L that follows.
Determining the total number of non-zero coefficient values contained in the block;
A first encoding step of variable length encoding the run value;
A level value indicating the non-zero coefficient value L is sequentially variable-length encoded and encoded from a high frequency component to a low frequency component while switching between a plurality of tables to which different threshold values are assigned. A second encoding step performed until the number of non-zero coefficient values reaches the total number of non-zero coefficient values;
Causing the computer to execute a third encoding step of variable-length encoding the total number of non-zero coefficients included in the block;
In the second encoding step,
When the absolute value of the non-zero coefficient value L to be encoded exceeds the threshold value assigned to the table used for encoding the non-zero coefficient value L, the table to which a threshold value larger than the threshold value is assigned After being switched, the non-zero coefficient value L to be encoded is followed by a non-zero coefficient value L to be encoded next,
The switching of the plurality of tables is a one-way switching in which a table used before a new table that has been switched is not used again.

Images