JP2003018014A - Coding method, decoding method, coder, decoder, coding program, decoding program and programs recording medium for them - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a coding method that conducts variable length coding, without losing correlation of components at positions expressed multi-dimensionally, so as to enhance the coding efficiency in the case of applying coding to a value of multi-dimensional matrix. SOLUTION: A valid component discrimination section 103 sequentially detect a valid component of a multi-dimensional matrix and a pre-component coordinate storage section 105 stored multi-dimensional position information. A vector calculation section 106 calculates a difference between the multi-dimensional position information of the valid component, obtained prior to with multi-dimensional position information with a valid component newly, obtained by the valid component discrimination section 103 to obtain a multi-dimensional vector. A last discrimination section 110 discriminates whether or not the valid component obtained by the valid component discrimination section 103 is a final valid component in the multi-dimensional matrix, and provides an output of the result as last information. A vector value last coding section 111 assigns on variable length code to a combination among each component of a multi-dimensional vector obtained by the vector calculation section 106, a valid component obtained by the valid component discrimination section 103 and the last information obtained by the last discrimination section 110.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to variable length coding and variable length decoding in source coding.

[0002]

2. Description of the Related Art Each component of a multidimensional matrix has many same values (A) and different values (B ₁ , B ₂ , ..., B _n ).
If it is composed of and, there is a method of applying run-length coding as a method of coding the components of this matrix. First, the components in the multidimensional matrix are scanned in a predetermined order to form a one-dimensional array. In this one-dimensional array,
Since the same value (A) occupies a large number, the value (A) is often aligned continuously. Therefore, the number of components in which the values (A) are arranged is run-length encoded. Alternatively, the number of arrangement of the value (A) (L _A), before and after the value (A) values (B _{i; i} =
1, ..., if there is a correlation between n) is, (L _A)
There is also a method of performing two-dimensional variable length coding in which one variable length code is assigned to the combination of (B _i ). Here generally run the (L _A), referred to as a level (B _i).

For example, in coding a two-dimensional image, in general,
The coefficient after the discrete cosine transform (DCT) of the pixel is two-dimensionally variable length coded. First, each pixel is divided into blocks each having 8 pixels vertically and horizontally, and two-dimensional DCT is performed for each block. In the two-dimensional DCT, values other than 0 are concentrated in the low frequency part, and conversely, the value 0 is increased in the high frequency part. This D
Consider the CT coefficient as an 8 × 8 two-dimensional matrix.

FIG. 13 is a diagram showing an example of zigzag scanning on 8 × 8 DCT coefficients. Zigzag scanning is performed from the low frequency part to the high frequency part to form a one-dimensional array. Two-dimensional variable-length coding is performed by setting the number of consecutive coefficients having a value of 0 as a run and the value of the subsequent coefficient as a level. A look-up table in which one code is associated with each combination of run and level is created in advance, and two-dimensional variable length coding is performed using this look-up table. For combinations not found in the lookup table, fixed-length coding using escape codes is performed. Alternatively, as a method of performing two-dimensional variable length coding, it is also possible to use arithmetic coding without using a lookup table.

Especially in the image coding system MPEG-1, 2
A two-dimensional variable-length coding lookup table is configured by utilizing the fact that a coefficient of a small value exists around a coefficient of a large value in the dimension DCT. For example, for a run with a large value, a code with a short code length is prepared in combination with a level with a small value.

In addition, the value (B_i) And the value
Length of (A) (L_A) Before two-dimensional variable length coding
And the value (B_i) Is quantized. By quantization
Value (B_i) Is smaller than (BQ_i) Is converted to
Then, the value (BQ_i) And (L _A) Combination 2
Dimension variable length coding. Generally, two-dimensional variable length coding
Then, the value (B_i) Is smaller, the code with shorter code length
, The amount of code can be reduced by quantization.
Become.

Further, when the values (B _i ) and (L _A ) are variable-length coded, it is indicated whether the value (B _i ) is the last component other than the value (A) in the multidimensional matrix. There is a three-dimensional variable length coding method for coding including the last information.

For example, in the image coding system MPEG-4,
When the coefficient other than 0 is encoded among the coefficients subjected to the two-dimensional DCT, the last information is set to 0 if the coefficient is not the last coefficient other than 0, and the last information is set if the coefficient is other than the last 0. Is set to 1, and three-dimensional variable length coding of run, level, and last information is performed. A lookup table in which codes are associated with each run, level, and last is created in advance, and three-dimensional variable length coding is performed with reference to the lookup table.

On the contrary, in the image coding method MPEG-1, it is 2
After encoding the last coefficient of the dimension DCT, a fixed-length code having a unique bit pattern called EOB is output.
This EOB is used for knowing that the coefficient to be decoded has ended when variable-length decoding the two-dimensional DCT coefficient from the bit array on the decoding side. Since this EOB code generally has a long code length, the coding efficiency decreases when the number of coefficients to be coded in the two-dimensional DCT is small.

In the three-dimensional variable-length coding, the output of EOB can be avoided by coding the last information by the variable-length coding. That is, when the number of coefficients to be coded in the two-dimensional DCT is small, it is possible to avoid a decrease in coding efficiency.

There is also a method of using a three-dimensional DCT in three-dimensional image coding. In this case, the three-dimensional DCT coefficient is zigzag scanned to form a one-dimensional array. Then, similarly to the two-dimensional image coding, the two-dimensional variable length coding or the three-dimensional variable length coding is performed on the one-dimensional array.

[0012]

In the multidimensional matrix, when there is a small value component around a large value component, the components to be coded exist. For example, in a two-dimensional DCT in two-dimensional image coding, a coefficient having a small value exists around a coefficient having a large value. If there is such a correlation between the components, the method of once arranging each component of the multidimensional matrix once in a one-dimensional array will impair the correlation between the components.

FIG. 14 is a diagram showing an example of 8 × 8 DCT coefficient positions. For example, a two-dimensional DCT as shown in FIG.
In, consider a case where the values of the coefficients A to G are other than 0 and other components are arranged around the coefficient D. Zigzag scanning makes a one-dimensional array, so coefficient A to coefficient B
Are continuous, and a code with a short code length can be assigned. However, since many coefficients having a value of 0 exist between the coefficient B and the coefficient C, the code length of the assigned variable-length code becomes long or the fixed length using the escape code is used. That is, regarding the coding of adjacent coefficients, a code having a short code length can be assigned when they are adjacent in the scanning direction, but a short code length cannot be assigned when they are not in the scanning direction.

As described above, in the method of converting a multi-dimensional matrix into a one-dimensional array for encoding, the correlation with respect to a direction other than the scan direction used for conversion is impaired. Therefore, the conventional method may cause a decrease in coding efficiency.

[0015]

In order to solve the above problems, the present invention assigns a variable length code word to a combination of a component and a level of a difference vector at a position represented in multidimensional, The most main feature. It is also characterized by quantizing the components of the position difference vector.

Specifically, the first invention is a method for reducing the bit rate in encoding the value of a multidimensional matrix component having a finite number of components, and forming at least one variable length codeword. In, the position difference information between the two components is
It is characterized in that it is used with the value of one of the components.

In the first aspect of the present invention, the position difference between the components is used to encode the position information of the components required as encoded data in the multidimensional matrix. In the case of an N-dimensional matrix, the position of the component P is expressed by the following equation (1).

P (p ₁ , p ₂ , ..., P _n ) (1) The difference in the position of the component Q with respect to the component P is an N-dimensional vector, and this vector V (P, Q) is given by the formula (2 ).

[0019] V (P, Q) = ( q 1 -p 1, q 2 -p 2, ..., q n -p n) (2) Then, in the multi-dimensional matrix, necessary as encoded data One variable length code is assigned to the combination of the component value level and the position difference vector component.

Here, when the codes having the shorter code lengths are assigned in the ascending order of the norm magnitude of the vector V (P, Q),
Since the adjacent components have a small norm, they can be encoded with a short code length. As a result, it is possible to perform efficient coding without losing the correlation with respect to directions other than the scan direction, which is lost in the method of converting into a one-dimensional array.

Further, when the norm of the difference vector is large, there is a correlation that the number of small values increases. Therefore, one variable length code is assigned to the combination of the difference vector component and level. In this way, by performing the variable length coding on the combination of the difference vector component and the level, it is possible to perform the variable length coding more efficiently than by individually coding the difference vector component and the level.

FIG. 1 shows an example (Example 1) of two-dimensional vector coding on 8 × 8 DCT coefficients. For example, two-dimensional DC
In the case of T, if two-dimensional coordinates having the X axis (V _x ) and the Y axis (V _y ) as shown in FIG. 1 are set, the coefficient B to the coefficient C can be expressed as a vector (1,0). By assigning a code having a short code length to a vector having a small norm, the coding efficiency can be improved.

The second invention is a method for reducing the bit rate in encoding the value of a multidimensional matrix component having a finite number of components, and in the formation of at least one variable-length codeword, there is a difference between two components. It is characterized in that the difference information of the position of is used together with the value of one of the components and the last information indicating whether it is the last component to be encoded.

According to the second invention, the difference vector component, the value of the component to be encoded, and the last component indicating whether or not the component to be encoded in the multidimensional matrix is the last component. One variable length code can be assigned to a combination of information. As a result, it is possible to avoid a decrease in coding efficiency that occurs when the number of components to be coded in the multidimensional matrix is small.

In a third aspect of the present invention, a variable-length codeword created by an encoding method for reducing the first or second bit rate is used in encoding the value of a multidimensional matrix component having a finite number of components. It is characterized in that it is recorded in a lookup table in advance and variable length coding is performed using the lookup table.

According to the third invention, the first or second
By previously creating the variable-length code word in the invention described above and recording it in the lookup table, it becomes possible to perform variable-length encoding while referring to the lookup table in the encoding of the multidimensional matrix.

A fourth invention is characterized in that, in the above-mentioned coding method, the difference information of the position between two components to be variable-length coded is obtained in a coordinate system different from that of the multidimensional matrix.
A coordinate system different from this multidimensional matrix can be changed for each position of the origin component for which the difference is calculated.

According to the fourth aspect of the invention, in encoding a multidimensional matrix, the difference vector can be obtained by converting it into a coordinate system different from that of the multidimensional matrix. Figure 2 is 8x8D
An example (Example 2) of two-dimensional vector coding on CT coefficients is shown. For example, in the two-dimensional DCT, the coordinate system as shown in FIG.
A coordinate system having an X axis (V _x ) and a Y axis (V _y ) as shown in FIG. 2 can be used. Thereby, the difference vector of the coefficient A to the coefficient B can be expressed as (1, 0), and a code having a shorter code length can be assigned.

Further, when the difference vector is converted into a coordinate system different from the multidimensional matrix in the fourth aspect of the invention, the coordinate system can be changed for each position of the origin of the difference vector. For example, in the two-dimensional DCT, it is possible to use the coordinate system shown in FIG. 2 for the coefficient A and the coordinate system shown in FIG. 1 for the coefficient B.

A fifth aspect of the present invention is characterized in that, in the above encoding method, the variable length encoding is performed after the difference information of the position between the two components to be variable length encoded is quantized.

According to the fifth aspect of the invention, in encoding a multidimensional matrix, variable-length encoding can be performed after quantizing the values of the components of the difference vector. This makes it possible to further improve the coding efficiency when performing variable-length coding in which a code having a shorter code length is assigned as the norm of the difference vector is smaller.

A sixth aspect of the present invention is to decode at least each component of a multidimensional vector and an effective component from the variable length code created by the above encoding method in decoding the value of a multidimensional matrix component having a finite number of components. It is characterized in that the value of the multi-dimensional matrix component is decoded by variable-length decoding the values including and, and calculating the position information in the multi-dimensional matrix using the components of the multi-dimensional vector. Here, the variable-length code word created by the above-described encoding method can be recorded in a look-up table in advance, and variable-length decoding can be performed using the look-up table.

According to the sixth aspect of the invention, in decoding a multidimensional matrix, the bit array output by the encoding method according to the first to fifth aspects of the invention can be decoded using a look-up table or the like. it can.

Further, it is possible to configure a coding apparatus that implements the above first to fifth coding methods and a decoding apparatus that implements the above sixth decoding method. In the encoding device, it is determined whether the value of each component of the multidimensional matrix is a value that satisfies a predetermined criterion, and if it satisfies, an effective component determination unit that determines as an effective component is provided. It is possible to judge the effective component from the above and encode only the effective component. The following are examples of criteria for determining whether an active ingredient is included. (A) The case where the value is the reference value T or more is valid. When T is 1, it can be regarded as an effective component when T is other than 0. (B) Valid if the value is even.

The above encoding method and decoding method can be realized not only by circuitized hardware, but also by a computer and a software program installed and executed in the computer. Can be stored in an appropriate recording medium such as a computer-readable portable medium memory, a semiconductor memory, or a hard disk.

[0036]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. As an example, a configuration and an operation example of an image coding apparatus and an image decoding apparatus that perform two-dimensional DCT on a two-dimensional vertical and horizontal 8-pixel image and perform variable length coding of DCT coefficients will be shown. In this embodiment, the value of the DCT coefficient and the position difference vector are quantized and then variable length coded. In addition, the variable length coding uses the coefficient value, the difference vector component, and the last information.

FIG. 3 is a diagram showing a configuration example of the image coding apparatus according to the present embodiment. The image encoding device 1 has a D
A value quantizer 101 for quantizing the values of CT coefficients, and a DCT
A multi-dimensional matrix reading unit 102 that reads the coefficient values in the order shown in FIG. 13, and an effective component determination unit 1 that determines whether the read value is 0 and that if it is other than 0, considers it as an effective component.
03, a value storage unit 104 for storing the values of effective components,
The previous component coordinate storage unit 105 that stores the two-dimensional position information of the effective component, the two-dimensional position information stored in the previous component coordinate storage unit 105, and the two-dimensional position of the effective component obtained by the effective component determination unit 103. A vector calculation unit 106 that calculates the difference between information and obtains a two-dimensional vector, a vector component quantization unit 107 that quantizes the components of the two-dimensional vector, and a two-dimensional unit that dequantizes the components of the quantized two-dimensional vector. Vector component dequantization unit 1 that calculates position information, obtains position information of the previous component based on the calculated position information, and stores the position information in the front component coordinate storage unit 105.
09, a vector storage unit 108 for storing the vector components obtained by the vector calculation unit 106, and an effective component determination unit 1
If the effective component is not obtained up to the end of the DCT coefficient in 03, 1 is output as the last information, and if it is obtained, 0 is output as the last information, and the vector component quantization unit 107 A vector value that assigns one variable length code to the combination of each component of the obtained two-dimensional vector, the value of the effective component obtained by the effective component determination unit 103, and the last information obtained by the last determination unit 110. A last encoding unit 111 and a vector value last table unit 112 that accumulates a look-up table to which one variable-length code is assigned for the values of the multidimensional vector and the effective component and the last information are provided.

FIG. 4 is a diagram showing an example of two-dimensional vector coordinates. The vector coordinate system used in the vector calculation unit 106 is changed from the three types shown in FIG. 4 according to the position information stored in the previous component coordinate storage unit 105. For the matrix element (x, y) (x, y = 0, ..., 7), x
Is even or 0 and y is 0, the coordinate system of FIG. 4A is applied. Otherwise, when x + y is an even number, FIG.
If the coordinate system of (b) is applied and x + y is an odd number, then FIG.
The coordinate system of (c) shall be applied.

In the vector value last table section 112,
It is assumed that the lookup table shown in FIG. 5 has been accumulated. The lookup table has a configuration in which the code word is referred to by a combination of the vector V _x component, the vector V _y component, the quantized DCT coefficient value, and the last information.

The value quantizer 101 quantizes the values of all the coefficients according to the equation (3). I indicates an input value and O indicates an output value.

O = I / 2 (3) The vector component quantization unit 107 uses the position difference vector (I ₁ , I ₂ ) as the input vector and the output vector (O
₁ , O ₂ ), and the components of the two-dimensional vector are quantized according to equation (4). However,
It is assumed that x and y respectively indicate the positions stored in the previous component coordinate storage unit 105. It is assumed that (0, 0) is previously stored as an initial value in the previous component coordinate storage unit 105.

O _i = (I _i +1) / ((x + y + 7) / 7) (4) The vector component dequantization unit 109 is assumed to dequantize according to equation (5).

O _i = I _i * ((x + y + 7) / 7) −1 (5) FIG. 6 is a diagram showing an example of the input DCT coefficient in the present embodiment. Based on the above assumptions, we already have two-dimensional D
When the two-dimensional DCT coefficient shown in FIG. 6 which has been subjected to CT is obtained, the image encoding device 1 operates as follows. First, the value quantization unit 101 quantizes the value of the DCT coefficient according to the equation (3). FIG. 7 is a diagram showing an example of quantized DCT coefficients. When the DCT coefficient of FIG. 6 is quantized, it becomes as shown in FIG. Multidimensional matrix reading unit 102
Reads out the quantized values in the order shown in FIG.

A procedure for coding the coefficients in the order of reading by the multidimensional matrix reading unit 102 will be described. The value of the matrix element position (x, y) will be referred to as A (x, y). The effective component determination unit 103 determines that the effective component because A (1,0) is other than 0. A (1,0) is stored in the value storage unit 104
Accumulate in.

The vector calculation unit 106 obtains the difference between the position (0,0) stored in the previous component coordinate storage unit 105 and the coordinate position (1,0) of the current component. At this time, FIG. 4A is used as the coordinate system of the vector. The difference vector can be calculated as (1,0).

The vector component quantizer 107 uses the equation (4)
Quantize using to get (1,0). This is stored in the vector storage unit 108. Vector component inverse quantizer 10
9 is inversely quantized using equation (5) to obtain (1,0).
Since the vector coordinate system is as shown in FIG. 4C, the vector component is coordinate-converted into a matrix component to obtain (1,0), and the previous component coordinate storage unit 105 stores (0,0) + (1 , 0)
= (1,0) is accumulated.

The last determination unit 110 is the effective component determination unit 1
Since a valid value was obtained in 03, 0 is output as the last information. The vector value last encoding unit 111 does not operate when only the last information is obtained.

Then, the effective component determination section 103
Since (0,1) is other than 0, it is determined to be an effective component.
A (0, 1) is stored in the value storage unit 104. The vector calculation unit 106 obtains the difference between the positions (1,0) and (0,1) stored in the previous component coordinate storage unit 105. At this time, FIG. 4C is used as the coordinate system of the vector. The difference vector can be calculated as (1,0). The vector component quantizer 107 quantizes using the equation (4) to obtain (1,0). This is stored in the vector storage unit 108. The vector component dequantization unit 109 obtains (1,0) by dequantization using equation (5). Then, the vector component is coordinate-converted into a matrix component to obtain (−1,1), and (1,0) + (− 1,1) = (0,1) is stored in the previous component coordinate storage unit 105. To do.

The last determination unit 110 is the effective component determination unit 1
Since a valid value was obtained in 03, 0 is output as the last information. The vector value last encoding unit 111 stores the value A (0, 1) stored in the value storage unit 108, the (1, 0) stored in the vector storage unit 108, and the last information 0 as the vector value last. Variable length coding is performed using the lookup table recorded in the table unit 112.

By repeating such processing, the DCT coefficients are variable length coded in the order shown in FIG. FIG. 8 shows an example of the order of DCT coefficients to be encoded. After encoding the last coefficient A (3,5), the effective component determination unit 103
Cannot obtain a valid component in the DCT coefficient. In this case, the last determination unit 110 outputs 1 as the last information. The vector value last encoding unit 111 converts the value A (2,6) stored in the value storage unit 104, (1,0) stored in the vector storage unit 108, and 1 of the last information into a vector. Variable length coding is performed using the lookup table recorded in the value last table unit 112.

A list of the value A (x, y), the vector component and the last information obtained by repeating the above processing is shown in FIG. FIG. 9 shows an example of a combination encoded by the vector value last encoding unit 111. FIG. 10 shows an example of codewords output by the vector value last encoding unit 111. For each combination of FIG. 9, the lookup table (FIG. 7) of the vector value last table unit 112 is referred to. Thus, the code word is determined, and a bit string of variable length coding as shown in FIG. 10 is output.

FIG. 11 is a diagram showing a configuration example of the image decoding apparatus according to the present embodiment. The image decoding device 2 includes a vector value last table unit 201 and a vector value last decoding unit 202 that detects one variable length code and decodes each component of the multidimensional vector, the value of the effective component, and the last information.
And a vector component dequantization unit 2 that dequantizes the components of the multidimensional vector obtained by the vector value last decoding unit 202.
03, a pre-component coordinate accumulating unit 204 for accumulating position information of components of the multi-dimensional matrix, a multi-dimensional vector component obtained by the vector component dequantizing unit 203, and a pre-component coordinate accumulating unit 2
A position information calculation unit 205 that calculates the position information in the multidimensional matrix from the position information accumulated in 04, and a value dequantization unit 206 that dequantizes the value of the effective component obtained by the vector value last decoding unit 202. And a value substituting unit 207 for substituting the value of the effective component obtained by the value dequantizing unit 206 into the DCT coefficient of the position obtained by the position information calculating unit 205, and the last information obtained by the vector value last decoding unit 202. Is set to 1, the non-effective component value setting unit 208 that sets the value of the coefficient for which a value has not been obtained to 0.

The previous component coordinate storage unit 204 stores the position (0,
0) is stored in advance. The position information calculation unit 205 converts the coordinate system of the vector obtained by the vector component dequantization unit 203 according to the position information stored in the previous component coordinate storage unit 204 to calculate the position information. The vector coordinate system is to be changed from the three types shown in FIG. Matrix element (x, y) (x, y = 0, ...
, 7), if x is an even number or 0 and y is 0, the coordinate system of FIG. 4 (a) is applied. Otherwise, if x + y is an even number, the coordinate system of FIG. 4 (b) is applied. The system is applied, and when x + y is an odd number, the coordinate system of FIG. 4C is applied.

In the vector value last table section 201,
It is assumed that the lookup table shown in FIG. 5 has been accumulated. The value dequantization unit 206 quantizes the values of all the coefficients according to equation (6). I indicates the input value,
O indicates the output value.

O = I * 2 (6) The vector component inverse quantization unit 203 is assumed to perform inverse quantization according to equation (5).

Under such a premise, first, the vector value last decoding unit 202 detects a variable length code word from the bit string in the order of FIG. 10, and the vector value last table unit 201 stores the variable length code word for each variable length code word. The recorded lookup table is referenced to decode the vector component, valid value, and last information.

For the first variable length code word, the vector (1, 0), the value 5 and the last information 0 are obtained. The vector component inverse quantization unit 203 inversely quantizes the vector according to the equation (5) to obtain (1,0). Since (0,0) is stored in the previous component coordinate storage unit 204, the position information calculation unit 205 uses the coordinate system of FIG.
Is converted to (1,0) and added to (0,0) stored in the previous component coordinate storage unit 204. by this,
The position information is obtained as (0,0) + (1,0) = (1,0). This position information is stored in the previous component coordinate storage unit 204. Next, the value dequantization unit 206 dequantizes the value 5 according to the equation (6) to obtain the value 10. The value substitution unit 207 is a DCT
The value 10 is substituted for the coefficient position (1,0).

Subsequently, the vector value last decoding unit 202
Obtains vector (1,0), value 5 and last information 0 for the next variable length codeword. Vector component inverse quantizer 2
03 dequantizes the vector according to equation (5) (1,
0) is obtained. Since the position information calculation unit 205 stores (1, 0) in the previous component coordinate storage unit 204,
Using the coordinate system of (c), the vector (1,0) is converted into (-1,
It is converted to 1) and added to (1, 0) stored in the previous component coordinate storage unit 204. This gives (1,0) +
The position information is obtained as (−1,1) = (0,1). This position information is stored in the previous component coordinate storage unit 204. Next, the value dequantization unit 206 dequantizes the value 5 according to the equation (6) to obtain the value 10. The value substituting unit 207 substitutes the value 10 for the DCT coefficient position (0, 1).

By repeating such processing, the bit string is variable length decoded to the end. When the vector value last decoding unit 202 decodes the last variable length codeword, 1 is obtained as the last information. The ineffective component value setting unit 208 substitutes the value 0 into the coefficient for which a value has not been obtained.

FIG. 12 shows a DC obtained by the image decoding device 2.
It is an example of a T coefficient. As shown in FIG. 12, in this embodiment, since the vector between effective components is quantized, the DCT
Although there is a coefficient that the phase of is shifted, it can be set so that there is a possibility that it will be shifted only at a high frequency as in equations (4) and (5). That is, since the high-frequency distortion is generally less visually deteriorated, the image quality deterioration due to the phase shift can be reduced.

Further, in this embodiment, the lookup table is recorded in advance and the variable length coding and decoding are performed by referring to the lookup table. It is also possible to perform variable length coding and decoding using arithmetic coding without using this table. In this case, the image encoding device 1
Instead of the vector value last encoding unit 111 and the vector value last table unit 112, a vector value last arithmetic encoding unit (not shown) is provided. Further, the image decoding device 2 is provided with a vector value last arithmetic decoding unit (not shown) instead of the vector value last decoding unit 202 and the vector value last table unit 201.

Further, in the present embodiment, a method of quantizing the value of the effective component but not quantizing is also possible. In this case, the image coding apparatus 1 does not include the value quantization unit 101. The image decoding device 2 does not include the value dequantization unit 206.

Further, although the position information between effective components is quantized in the present embodiment, a method without quantization is also possible.
In this case, the image coding apparatus 1 does not include the vector component quantization unit 107 and the vector component dequantization unit 109.
Further, the image decoding device 2 does not include the vector component inverse quantization unit 203.

Further, in this embodiment, the variable length coding is performed by including the last information, but a method not including the last information is also possible. In this case, the last determination unit 1 is added to the image encoding device 1.
10, the value storage unit 104, and the vector storage unit 108 are not provided. The image decoding device 2 does not include the ineffective component value setting unit 208.

Further, in the present embodiment, the position information between effective components is made into a two-dimensional vector and its coordinate system is converted, but it is also possible to use a method in which it is not converted.

As in this embodiment, in the image coding apparatus 1 and the image decoding apparatus 2 according to the present invention, the smaller the norm of the vector between the effective components, the shorter the variable length code can be assigned. Become.

The present invention can be applied not only to the image coding but also to the coding of the value of the multidimensional matrix component having a finite number of components, and the bit rate can be similarly reduced. .

[0068]

According to the invention described above, a variable length codeword is assigned to a combination of a component and a level of a difference vector at a position represented in multiple dimensions. Therefore, variable length coding can be performed without impairing the correlation of each component of the position represented in multidimensional. In particular, the smaller the difference between the positions of the two components in the multidimensional matrix, the shorter the variable length code can be assigned, and the bit rate can be reduced.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of two-dimensional vector coding on 8 × 8 DCT coefficients.

FIG. 2 is a diagram showing another example of two-dimensional vector coding on 8 × 8 DCT coefficients.

FIG. 3 is a diagram showing a configuration example of an image coding apparatus according to the present embodiment.

FIG. 4 is a diagram showing an example of two-dimensional vector coordinates.

FIG. 5 is a diagram showing an example of a lookup table accumulated in a vector value last table unit.

FIG. 6 is a diagram showing an example of input DCT coefficients in the present embodiment.

FIG. 7 is a diagram showing an example of quantized DCT coefficients.

FIG. 8 is a diagram showing an example of the order of DCT coefficients to be encoded.

[Fig. 9] Fig. 9 is a diagram illustrating an example of a combination encoded by a vector value last encoding unit.

[Fig. 10] Fig. 10 is a diagram illustrating an example of a codeword output by a vector value last encoding unit.

FIG. 11 is a diagram showing a configuration example of an image decoding apparatus according to the present embodiment.

FIG. 12 is a diagram illustrating an example of DCT coefficients obtained by the image decoding device.

FIG. 13 is a diagram showing an example of zigzag scanning on 8 × 8 DCT coefficients.

FIG. 14 is a diagram showing an example of 8 × 8 DCT coefficient positions.

[Explanation of symbols]

1 Image coding device 101 value quantizer 102 Multi-dimensional matrix reading unit 103 Active ingredient determination unit 104 value storage 105 previous component coordinate storage 106 Vector calculator 107 vector component quantizer 108 Vector storage 109 vector component inverse quantizer 110 Last judgment section 111 Vector value last encoding unit 112 Vector value last table part 2 Image decoding device 201 Vector value last table part 202 Vector value last decoding unit 203 vector component inverse quantizer 204 Pre-component coordinate storage 205 Location information calculator 206-value inverse quantizer 207 Value substitution unit 208 Non-effective component value setting section

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Naoki Kobayashi             2-3-1, Otemachi, Chiyoda-ku, Tokyo             Inside Telegraph and Telephone Corporation F-term (reference) 5C059 MA23 MC11 ME01 SS20 UA02                       UA05 UA39                 5J064 AA02 BA09 BA16 BB01 BC14                       BD03 BD04

Claims (18)

[Claims] 1. A method for reducing a bit rate in encoding a value of a multi-dimensional matrix component having a finite number of components, which comprises:
An encoding method characterized in that the difference information of the positions between the components is used together with the value of one of the components. 2. A method for reducing a bit rate in encoding a value of a multi-dimensional matrix component having a finite number of components, which comprises:
A coding method characterized in that the difference information of the positions between the components is used together with the value of one of the components and the information indicating whether it is the last component to be encoded. 3. The encoding method according to claim 1 or 2, wherein a variable length codeword created using at least position difference information between the two components is recorded in advance in a lookup table. Every other, variable-length coding is performed using the lookup table. 4. The encoding method according to any one of claims 1 to 3, wherein the difference information of the positions between the two components to be variable length encoded is in a coordinate system different from that of the multidimensional matrix. An encoding method characterized by obtaining. 5. The encoding method according to any one of claims 1 to 4, wherein variable length encoding is performed after quantizing difference information of positions between two components to be variable length encoded. An encoding method characterized by the above. 6. Decoding a value of a multidimensional matrix component having a finite number of components, at least multidimensional from a variable length code created by the encoding method according to any one of claims 1 to 5. Decodes the value of the multidimensional matrix component by performing variable length decoding on the value including each component of the vector and the value of the effective component and calculating the position information in the multidimensional matrix using the components of the multidimensional vector A decoding method characterized by the above. 7. A multidimensional matrix reading unit for reading the values of the respective components of a multidimensional matrix in a coding device for the values of multidimensional matrix components having a finite number of components,
It is determined whether the read value is a value that satisfies a predetermined criterion, and if it is, the effective component determination unit that determines the effective component, and the previous component coordinate storage unit that stores the multidimensional position information of the effective component. And a vector calculation unit for calculating a difference between the multidimensional position information accumulated in the preceding component coordinate storage unit and the multidimensional position information of the effective component obtained by the effective component determination unit, and obtaining a multidimensional vector, A vector value encoding unit that assigns one variable length code to a combination of each component of the multidimensional vector obtained by the vector calculation unit and the value of the effective component obtained by the effective component determination unit; An encoding device, comprising: a vector value table unit for accumulating a look-up table in which one variable-length code is assigned to the values of multi-dimensional vector components and effective components. 8. A multidimensional matrix reading unit for reading the values of the respective components of a multidimensional matrix in a predetermined order in a coding device for the values of multidimensional matrix components having a finite number of components,
It is determined whether the read value is a value that satisfies a predetermined criterion, and if it is, the effective component determination unit that determines the effective component, and the previous component coordinate storage unit that stores the multidimensional position information of the effective component. And a vector calculation unit for calculating a difference between the multidimensional position information accumulated in the preceding component coordinate storage unit and the multidimensional position information of the effective component obtained by the effective component determination unit, and obtaining a multidimensional vector, The effective component obtained by the effective component determination unit determines whether or not the last effective component in the multidimensional matrix and outputs this as the last information, and the multidimensional calculated by the vector calculation unit. A vector value last encoding unit that assigns one variable length code to each combination of each component of the vector, the value of the effective component obtained by the effective component determination unit, and the last information obtained by the last determination unit. When The relative value and the last information of the component and the active ingredient of the multidimensional vector, the encoding apparatus characterized in that it comprises a vector value last table unit for storing a look-up table assigned one variable length code. 9. The encoding device according to claim 7, wherein the vector calculation unit obtains a multidimensional vector in a coordinate system different from that of the multidimensional matrix. Device. 10. The encoding device according to claim 7, further comprising a vector component quantizer that quantizes the component value of the multidimensional vector obtained by the vector calculator. An encoding device characterized by. 11. A decoding device for decoding a value of a multidimensional matrix component having a finite number of components, and stores a look-up table in which one variable length code is assigned to a value of a multidimensional vector component and a valid component value. Vector value table section
Using the lookup table, a vector-value decoding unit that decodes each component of a multidimensional vector and the value of an effective component from one variable-length code, and a pre-component that stores position information of the components of a multidimensional matrix A coordinate storage unit, a position information calculation unit that calculates position information in a multidimensional matrix from the components of the multidimensional vector obtained by the vector value decoding unit and the position information stored in the previous component coordinate storage unit. A decoding device characterized by the above. 12. A decoding device for a value of a multidimensional matrix component having a finite number of components, a last of which indicates a multidimensional vector component, a valid component value, and the last valid component in the multidimensional matrix. A vector-valued last table unit for accumulating a look-up table to which one variable-length code is allocated for information, and the look-up table,
A vector value last decoding unit that decodes each component of a multidimensional vector, a value of a valid component, and last information from one variable-length code, and pre-component coordinate storage that stores position information of components of a multidimensional matrix. Unit, a position information calculation unit that calculates position information in a multidimensional matrix from the components of the multidimensional vector obtained by the vector value last decoding unit, and the position information accumulated in the previous component coordinate accumulation unit, and the vector When the last information obtained by the value last decoding section indicates that the obtained effective component value is the last in the multidimensional matrix, the value of the component whose value is not obtained in the matrix is predetermined. And a non-effective component value setting unit for setting a value. 13. The decoding device according to claim 11 or 12, wherein the position information calculation unit calculates position information using a coordinate system different from a multidimensional matrix. apparatus. 14. The decoding device according to claim 11, wherein after decoding each component of the multidimensional vector, each component of the obtained vector is inversely quantized. Decoding device. 15. An encoding program for causing a computer to execute the encoding method according to any one of claims 1 to 5. 16. A decryption program for causing a computer to execute the decryption method according to claim 6. 17. A recording medium for an encoding program, which stores a program for causing a computer to execute the encoding method according to any one of claims 1 to 5. 18. A recording medium for a decoding program, which stores a program for causing a computer to execute the decoding method according to claim 6.