KR102453953B1 - Encoding method and apparatus for transformming hierarchical variable block and decoding method and apparatus - Google Patents

Abstract

A transcoding method and apparatus, and a decoding method and apparatus are provided. The transform encoding method may transform the residual block according to predefined or adaptively determined transform type information. And, any one of the transformed residual blocks may be determined as a final block.

Description

Encoding method and apparatus capable of hierarchical variable block transformation, and decoding method and apparatus

The present invention relates to a technique for transcoding by dividing an input image into a plurality of blocks and hierarchically dividing a coding block among the divided blocks.

In general, in the case of performing inter prediction encoding, the transform encoding method applies the same 4×4 or 8×8 transform blocks to all prediction split blocks to perform transform encoding.

For example, the conventional transform encoding method may generate residual blocks by performing motion prediction on 16×16, 16×8, 8×16, 8×8, 4×8, and 4×4 block modes. . In addition, the residual block may be transcoded using a 4×4 or 8×8 transform block. As described above, in the conventional transcoding method, since transcoding is performed equally on all residual blocks regardless of the prediction mode, encoding efficiency may be reduced.

In particular, when using an 8×8 transform block, the conventional transform encoding method has a limitation that transform encoding can be performed only when the prediction modes are 16×16, 16×8, 8×16, or 8×8. .

Accordingly, there is a need for a technology capable of improving encoding efficiency by applying transcoding differently according to characteristics of an image.

The present invention provides a method and apparatus for generating transform type information suitable for a size of a coding block, and a method and apparatus for transform encoding a residual block based on the transform type information, and a decoding method and apparatus.

A coding method according to an embodiment of the present invention includes generating a plurality of prediction blocks by predictively encoding a coding block according to a plurality of prediction mode information, and a plurality of residual blocks using the coding block and the plurality of prediction blocks. generating , transforming each of the residual blocks according to predefined transform type information or adaptively determined transform type information, and determining any one of the transformed residual blocks as a final block. .

In the determining of the final block, any one of the transformed residual blocks having the best encoding performance may be determined as the final block.

The method may further include calculating rate-distortion costs of the transformed residual blocks, respectively. Then, in the determining of the final block, the final block may be determined based on the calculated rate-distortion cost.

In addition, the determining of the final block includes determining the residual block having the minimum calculated rate-distortion cost as the final block, and determining whether to transmit transform type information and prediction mode information applied to the final block may include steps.

In addition, the step of determining whether to transmit transform type information and prediction mode information applied to the final block includes: transform type information and prediction mode information applied to the final block, and transform type information and prediction mode of neighboring blocks adjacent to the coding block Based on the information, it may be determined whether to transmit transform type information and prediction mode information applied to the final block.

In addition, the determining of the final block may generate additional information by combining transform type information and prediction mode information applied to the final block with coded block pattern information (CBP).

In addition, the transforming of each of the residual blocks may include adaptively determining a transform depth using at least one of a size of a unit block, a size of a coding block, and quantization information of a neighboring block, and a size of the coding block. The method may include determining a size of a transform block for each transform depth based on .

In addition, generating a final prediction block by predictive encoding the coding block based on the prediction mode information applied to the final block, generating a final residual block using the final prediction block and the coding block, the final block The method may further include transforming the final residual block based on transform type information applied to , and quantizing and entropy encoding the transformed final residual block.

According to the present invention, encoding efficiency can be improved by transform-coding a residual block based on transform type information suitable for the size of the coding block.

According to the present invention, it is possible to statically transform-code a residual block according to transform type information predefined for a coding block.

According to the present invention, it is possible to adaptively transform and encode a residual block according to transform type information adaptively generated according to the coding block.

1 is a flowchart provided to explain a transcoding method according to an embodiment of the present invention.
2 is a flowchart provided to explain a configuration for determining a final block in an encoding apparatus according to an embodiment of the present invention.
3 is a flowchart provided to explain a configuration of encoding a transformed final residual block based on transform type information of the final block in the encoding apparatus according to an embodiment of the present invention.
4 is a diagram illustrating transform type information applicable to a coding block in an encoding apparatus according to an embodiment of the present invention.
5 is a diagram illustrating transform type information when a transform block having a size of a transform block larger than a size of a prediction block is applied in the encoding apparatus according to an embodiment of the present invention.
6 is a block diagram illustrating a detailed configuration of an encoding apparatus according to an embodiment of the present invention.
7 is a flowchart provided to explain a transform decoding method according to an embodiment of the present invention.
8 is a block diagram illustrating a detailed configuration of a decoding apparatus according to an exemplary embodiment of the present invention.
9 is a diagram provided to explain a configuration using pattern information (CBP) of a coding block according to an exemplary embodiment of the present invention.

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited or limited by the examples. Also, like reference numerals in each figure denote like members.

1 is a flowchart provided to explain a transcoding method according to an embodiment of the present invention.

Referring to FIG. 1 , in step 101, the encoding apparatus may divide an input image into a plurality of blocks. For example, the encoding apparatus may divide the input image into a plurality of unit blocks. Here, the unit block is a block having the largest size among the coding blocks, and may mean the highest coding block.

Next, in operation 102 , the encoding apparatus may determine a maximum transform size and a split depth.

For example, the encoding apparatus may determine a maximum transform size and a split depth of a coding block to be applied to a residual block generated by using a unit block and a prediction block of the unit block. In this case, the encoding apparatus may determine the maximum transform size and the division depth based on the size of the unit block. Also, the encoding apparatus may determine the maximum transform size and the split depth by using the quantization information of the neighboring blocks.

Then, in step 103 , the encoding apparatus may divide the unit block into a plurality of encoding blocks. In this case, the encoding apparatus may split the unit block into a plurality of encoding blocks based on the maximum transform size and the split depth.

Next, in operation 104, the encoding apparatus may generate a plurality of prediction blocks for the encoding block. In this case, the encoding apparatus may generate a plurality of prediction blocks by predictively encoding the coding blocks according to the plurality of prediction modes.

As an example, the encoding apparatus may generate the prediction block by predictively encoding the coding block using the inter prediction mode or the intra prediction mode. Here, the inter prediction mode may include prediction modes such as 32×32, 16×16, 16×8, 8×16, 8×8, 8×4, 4×8, and 4×4. In addition, the intra prediction mode may include various directional prediction modes such as a DC prediction mode, a vertical prediction mode, and a horizontal prediction mode.

First, in the case of using the inter prediction mode, the encoding apparatus provides a prediction block when the 16×16 prediction mode is applied, a prediction block when the 16×8 prediction mode is used, ... , a prediction block in the case of using the 4×4 prediction mode may be generated. In addition, when the intra prediction mode is used, the encoding apparatus may generate a prediction block when the DC prediction mode is used, a prediction block when the vertical prediction mode is used, and a prediction block when the horizontal prediction mode is used.

In operation 105, the encoding apparatus may generate a residual block using the prediction block and the encoding block. In this case, the encoding apparatus may generate a residual block through a difference between each prediction block and the encoding block.

For example, when 5 prediction modes are used to generate 5 prediction blocks, the encoding apparatus generates a residual block 1 by subtracting the prediction block 1 from the coding block, and subtracts the prediction block 2 from the coding block to generate a residual block. Create 2, … A residual block 5 may be generated by subtracting the prediction block 5 from the coding block.

Next, in operation 106, the encoding apparatus may transform each of the plurality of residual blocks according to predefined transform type information or adaptively determined transform type information. Here, the transform type information may include a size of a transform block and a transform depth. In this case, the transform type information may be predefined or adaptively determined based on at least one of a size of a unit block, a size of a coding block, and quantization information of a neighboring block. Here, the neighboring block means a block located adjacent to the coding block.

In addition, the transform type information may be predefined or adaptively determined as a single transform type, or may be predefined or adaptively determined as a plurality of transform types. In this case, the transformation type may have a tree structure or a hierarchical structure.

For example, when transform type information predefined as a single transform type is used, a size of a transform block corresponding to a 16×16 prediction mode may be predefined as 16×16 and a transform depth of 1 may be predefined. Then, the encoding apparatus may transform the residual block using a 16×16 transform block.

As another example, when transform type information predefined as a plurality of transform types is used, the size of the transform block corresponding to the 16×16 prediction mode is 16×16, 16×8, 8×16, 8×8, and the transform depth is 2 may be predefined. Then, the encoding apparatus may transform the residual block using the 16×16 transform block, transform the residual block using the 16×8 transform block, and transform the residual block using the 8×16 transform block. In addition, when the coding block is 64×64, the size of the transform block may be 64×64, the transform depth is 3, or the size of the transform block is 32×32, the transform depth is 2, and the like may be predefined. And, when the coding block is 32×32, the size of the transform block may be 16×16, the transform depth is 2, or the size of the transform block is 8×8, the transform depth is 1, and the like may be predefined.

As another example, the encoding apparatus may adaptively determine the transform depth based on at least one of a size of a unit block, a size of a coding block, and quantization information of a neighboring block. In addition, the encoding apparatus may determine the size of the transform block for each transform depth based on the size of the coding block. In this case, the encoding apparatus may adaptively determine the transform depth and the size of the transform block based on the pixel value change or motion information in the coding block. For example, the encoding apparatus may determine a smaller block size as the size of the transform block as the change in pixel values in the coding block increases or the movement increases. In addition, the encoding apparatus may determine a larger block size as the size of the transform block as the change in the pixel value in the encoding block or the movement decreases. As such, the encoding apparatus may adaptively determine the size of the transform block according to the characteristics of the image. Then, the residual block may be transformed using the determined transform block.

And, in operation 107, the encoding apparatus may determine any one of the plurality of residual blocks as a final block. In this case, the encoding apparatus may determine one of the plurality of residual blocks as the final block based on the encoding performance of the residual blocks.

For example, the encoding apparatus may determine the final block by using at least one of a rate-distortion cost of the residual block, a difference value of the residual block, and a transform coefficient value of the residual block.

Next, in operation 108 , the encoding apparatus may determine whether to transmit prediction mode information and transform type information applied to the final block.

In this case, the encoding apparatus may determine whether to transmit the transformation type information of the final block to the decoding apparatus based on the transformation type information of the neighboring block and the transformation type information of the final block. Similarly, the encoding apparatus may determine whether to transmit the prediction mode information of the last block to the decoding apparatus based on the prediction mode information of the neighboring block and the prediction mode information of the last block. Here, the final block is a coding block.

As an example, the encoding apparatus may determine whether the encoding block matches with the transformation type information of the neighboring block by comparing the transformation type information. In this case, if they match, the encoding apparatus may not transmit transform type information of the encoding block to the decoding apparatus. Similarly, the encoding apparatus may compare the prediction mode information of the coding block with the prediction mode information of the neighboring blocks to determine whether they match. In this case, if they match, the encoding apparatus may not transmit the prediction mode information of the encoding block to the decoding apparatus. Then, the decoding apparatus may reconstruct the coding block by using the transform type information and the prediction mode information of the neighboring blocks.

In this case, when the transform type information and the prediction mode information of the coding block do not match the transform type information and the prediction mode information of the neighboring blocks, the encoding device may transmit the transform type information and the prediction mode information of the coding block to the decoding device. .

Next, in step 109 , the encoding apparatus may generate and transmit side information.

For example, when transform type information and prediction mode information of a coding block and transform type information and prediction mode information of a neighboring block do not match, the coding apparatus may include coded block pattern information (CBP) of the coding block and the final block. Additional information may be generated by combining transform type information and prediction mode information applied to . Then, the encoding apparatus may transmit the generated additional information to the decoding apparatus.

As such, the encoding apparatus may reduce the amount of additional information by combining CBP information, transform type information, and prediction mode information. In this case, when a single transform type is used, the encoding apparatus may not transmit additional information indicating the number or type to the decoding apparatus. For example, when the size of the transform block is N×N and the transform depth is 1, since only a single transform type is used, the encoding apparatus decodes additional information including transform type information of the coding block and prediction mode information of the coding block. You don't have to send it to your device.

2 is a flowchart provided to explain a configuration for determining a final block in an encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 2 , in operation 201 , the encoding apparatus may determine encoding performance of each of the residual blocks. In this case, the encoding apparatus may use a rate-distortion cost, a difference value, a transform coefficient value, and the like to determine encoding performance.

In operation 202, the encoding apparatus may determine one of the plurality of transformed residual blocks as a final block based on the determined encoding performance.

For example, when the rate-distortion cost is used, the encoding apparatus may calculate the rate-distortion cost of each of the plurality of transformed residual blocks. Then, the encoding apparatus may determine the residual block having the minimum rate-distortion cost as the final block.

For example, when there are three transformed residual blocks, the encoding apparatus calculates a rate-distortion cost 1 of the transformed residual block 1, calculates a rate-distortion cost 2 of the transformed residual block 2, and the transformed residual block A rate-distortion cost of 3 can be calculated. Next, the encoding apparatus may select the minimum rate-distortion cost from among rate-distortion costs 1 to 3. Then, the encoding apparatus may determine the transformed residual block corresponding to the selected rate-distortion cost as the final block.

As another example, when using a difference value, the encoding apparatus may calculate a difference value of each of the plurality of transformed residual blocks. Then, the encoding apparatus may determine the final block based on the calculated difference value.

As another example, when a transform coefficient value is used, the encoding apparatus may calculate a transform coefficient value of each of the plurality of transformed residual blocks. Then, the encoding apparatus may determine the final block based on the calculated transform coefficient value.

3 is a flowchart provided to explain a configuration of encoding a transformed final residual block based on transform type information of the final block in the encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 3 , first, in step 301 , the encoding apparatus may generate a prediction block based on prediction mode information applied to the final block.

Next, in operation 302, the encoding apparatus may generate a final residual block by using the generated prediction block and the encoding block.

In operation 303, the encoding apparatus may transform the final residual block based on transform type information applied to the final block.

Next, in operation 304, the encoding apparatus may quantize the transformed final residual block and entropy-encode the quantized final residual block. In addition, the encoding apparatus may transmit the bitstream generated through entropy encoding to the decoding apparatus.

In FIG. 3, the final prediction block is generated based on the prediction mode information applied to the final block, the final residual block is transformed based on the transform type information applied to the final block, and the transformed final residual block is quantized and entropy-encoded. has been described, but this corresponds to an embodiment, and the encoding apparatus may omit the configuration of generating the final prediction block and transforming the final residual block, and may directly quantize and entropy encode the final block. In other words, when rate-distortion cost is used to determine encoding performance, in the process of comparing rate-distortion costs of transformed residual blocks, the encoding apparatus determines the residual block having the minimum rate-distortion cost as the final block, , the determined final block may be quantized, and the quantized final block may be entropy-encoded.

4 is a diagram illustrating transform type information applicable to a coding block in an encoding apparatus according to an embodiment of the present invention. In FIG. 4 , the size of the unit block is 8N×8N, and the division depth is 3 .

According to FIG. 4 , when the splitting depth is 3, the unit block 410 is split into 4N×4N coding blocks, the 4N×4N coding block is split into 2N×2N coding blocks, and the 2N×2N coding block is It may be divided into N×N coding blocks.

First, when a maximum transform size of 4N×4N and a transform depth of 3 are predefined for the 4N×4N coding block 430 , a transform type applicable to the 4N×4N coding block 430 may be the same as 431 . In this case, when the maximum transform size of the 4N×4N coding block 430 is predefined as 2N×2N and the transform depth is 2, the transform type applicable to the 4N×4N coding block 430 may be the same as 432 .

In addition, when the maximum transform size of the 2N×2N coding block 440 is predefined as 2N×2N and the transform depth is 2, the transform type applicable to the 2N×2N coding block 440 may be the same as 441 .

In addition, when the maximum transform size of the N×N coding block 450 is predefined as N×N and the transform depth is 1, the transform type applicable to the N×N coding block 450 may be equal to 451 . In other words, the size of the transform block applicable to the N×N coding block 450 may be four.

5 is a diagram illustrating transform type information when a transform block having a size of a transform block larger than a size of a prediction block is applied in the encoding apparatus according to an embodiment of the present invention.

According to FIG. 5 , a 4N×4N coding block 510 may be divided into three 2N×2N blocks and four N×N blocks. In this case, when a maximum transform size of 4N×4N and a transform depth of 3 are predefined for the 4N×4N coding block 510 , a transform type applicable to the 4N×4N coding block 510 may be the same as 520 .

6 is a block diagram illustrating a detailed configuration of an encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 6 , the encoding apparatus 600 includes a prediction block generator 610 , a residual block generator 620 , a transform unit 630 , a final block determiner 640 , a quantizer 650 , and entropy encoding. A portion 660 may be included.

First, the prediction block generator 610 may generate a plurality of prediction blocks by predictively encoding a coding block according to a plurality of prediction mode information.

Here, the prediction mode information may include an inter prediction mode and an intra prediction mode. And, the inter prediction mode may include prediction modes such as 16×16, 16×8, 8×16, 8×8, 8×4, 4×8, and 4×4. And, the intra prediction mode may include a DC prediction mode, a vertical prediction mode, a horizontal prediction mode, and the like.

The residual block generator 620 may generate a plurality of residual blocks by using the coding block and the plurality of prediction blocks. For example, the residual block generator 620 may generate a plurality of residual blocks by subtracting each of the coding block and the plurality of prediction blocks.

The transform unit 630 may transform each of the residual blocks according to predefined transform type information or adaptively determined transform type information. In this case, the transform unit 630 may transform each of the residual blocks according to predefined or determined transform type information using at least one of the size of the unit block, the size of the coding block, and quantization information of the neighboring block.

Here, the transform type information may include a size of a transform block and a transform depth. In addition, the transform type information may be predefined or determined as a single transform type or a plurality of transform types. In this case, when a plurality of transform types are used, the transform type information may have a tree structure or a hierarchical structure.

For example, the transform unit 630 may determine the transform depth by using at least one of a size of a unit block, a size of a coding block, and quantization information of a neighboring block. In addition, the transform unit 630 may determine the size of the transform block for each transform depth based on the size of the coding block. Next, the transform unit 630 may transform the coding block using the determined size of the transform block.

The final block determiner 640 may determine encoding performance of each of the transformed residual blocks. For example, the final block determiner 640 may determine the encoding performance using a rate-distortion cost, a difference value, a transform coefficient value, and the like. In addition, the final block determiner 640 may determine the residual block having the best encoding performance as the final block.

As an example, when the rate-distortion cost is used, the final block determiner 640 may calculate the rate-distortion cost of each of the transformed residual blocks. In this case, the final block determiner 640 may determine that the lower the rate-distortion cost, the better the encoding performance. Then, the final block determiner 640 may determine the residual block having the minimum rate-distortion cost as the final block.

Then, when the final block is determined, the prediction block generator 610 may generate the final prediction block by predictively encoding the coding block according to the prediction mode information applied to the final block. Then, the residual block generator 620 may generate a final residual block by subtracting the generated prediction block and the coding block. Then, the transform unit 630 may transform the final residual block according to transform type information applied to the final block.

Next, the quantization unit 650 may quantize the transformed final residual block. In addition, the entropy encoder 650 may entropy-encode the quantized final residual block to output a bitstream.

Meanwhile, the final block determiner 640 may determine whether to transmit prediction mode information of the coding block and transform type information of the coding block to the decoding apparatus.

For example, the final block determiner 640 may compare whether transform type information and prediction mode information of a coding block match with transform type information and prediction mode information of a neighboring block. And, if they match, the final block determiner 640 may not transmit transform type information and prediction mode information of the coding block to the decoding apparatus. And, if they do not match, the final block determiner 640 may transmit transform type information and prediction mode information of the coding block to the decoding apparatus. In this case, in order to reduce the bit amount of additional information transmitted to the decoding apparatus, the encoding apparatus may generate additional information by combining transform type information and prediction mode information of the encoding block, and pattern information (CBP) of the encoding block. . Then, the encoding apparatus may transmit the generated additional information to the decoding apparatus.

As such, the transform type is not fixed according to the size of the prediction block, but can be flexibly adapted to the image transformation by hierarchically variable block transformation of the residual block. For example, referring to FIG. 10 , a variable block transformation 1020 is possible such that a 4×4, 8×8 transform type is mixed and applied or an 8×16 transform type is applied within a 16×16 coding block. do. In other words, in the 16×16 coding block, since only two types of transform types are fixed, 4×4 and 8×8, it is possible to perform variable block transformation in various sizes according to the characteristics of the image without performing the division transformation 1010 .

7 is a flowchart provided to explain a transform decoding method according to an embodiment of the present invention.

Referring to FIG. 7 , first, in step 701 , the decoding apparatus may receive a bitstream from the encoding apparatus.

Next, in operation 702, the decoding apparatus may obtain transform type information and prediction mode information of the coding block. Here, the transform type information of the coding block may be transform type information applied to the final block, and the prediction mode information of the coding block may be prediction mode information applied to the final block. Transformation depth may be included.

For example, the decoding apparatus may receive additional information from the encoding apparatus. Here, the additional information may be a combination of transform type information of the coding block, prediction mode information of the coding block, and pattern information of the coding block. Then, the decoding apparatus may obtain transform type information of the coding block and prediction mode information of the coding block from the additional information.

As another example, when transform type information and prediction mode information of a coding block match with transform type information and prediction mode information of a neighboring block, and thus transform type information and prediction mode information of a coding block are not received from the encoding device, the decoding apparatus may obtain transform type information and prediction mode information of a neighboring block.

Next, in operation 703, the decoding apparatus may reconstruct the residual block using the obtained transform type information and prediction mode information. In this case, the decoding apparatus may reconstruct the residual block by using the transform type information and the prediction mode information of the final block. Also, the decoding apparatus may reconstruct the residual block by using the transform type information and the prediction mode information of the neighboring blocks.

For example, the decoding apparatus may inverse quantize the entropy-decoded block. Then, the decoding apparatus may inverse transform the inverse quantized block according to the obtained transform type information to reconstruct the residual block.

And, in operation 704, the decoding apparatus may generate a prediction block of the entropy-decoded block by using the obtained prediction mode information.

Next, in operation 705, the decoding apparatus may reconstruct the coding block by adding the prediction block of the decoding block and the residual block generated through inverse transform.

8 is a block diagram illustrating a detailed configuration of a decoding apparatus according to an exemplary embodiment of the present invention.

Referring to FIG. 8 , the decoding apparatus may include an entropy decoding unit 810 , an information obtaining unit 820 , an inverse quantization and inverse transform unit 830 , a prediction block generating unit 840 , and a restoration unit 850 . can

The entropy decoding unit 810 may entropy-decode the bitstream received from the encoding apparatus.

Then, the information obtaining unit 820 may obtain transform type information of the final block and prediction mode information of the final block from the entropy-decoded block. Here, the transform type information may include a size of a transform block and a transform depth.

The inverse quantization and inverse transform unit 830 may inverse quantize the entropy-decoded block. The inverse quantization and inverse transform unit 830 may inversely transform the inverse quantized block according to transform type information of the final block to reconstruct the residual block.

And, the prediction block generator 840 may generate the prediction block of the entropy-decoded block based on the prediction mode information of the final block.

Then, the reconstruction unit 850 may reconstruct the coding block by adding the generated prediction block and the reconstructed residual block.

Meanwhile, the information obtaining unit 820 may adaptively determine transform type information based on at least one of the size of the unit block, the size of the coding block, and quantization information of the neighboring block. In this case, the transform type information may be predefined based on at least one of a size of a unit block, a size of a coding block, and quantization information of a neighboring block. Then, the information obtaining unit 820 may obtain transform type information of the last block and prediction mode information of the last block based on predefined or adaptively determined transform type information.

In this case, when the encoding apparatus transmits additional information generated by combining transform type information and prediction mode information of the final block and pattern information (CBP) of the encoding block, the information obtaining unit 820 determines the information of the encoding block from the additional information. Transform type information and prediction mode information of the final block may be obtained by separating the pattern information (CBP). For example, the information obtaining unit 820 may obtain transform type information and prediction mode information of the final block by extracting pattern information of the coding block as shown in Table 1 below.

for(j=0; j<4; j++)
{
if(cbp&(1<<j))
{
T8_flag[j] = T8_flag;
}else{
T8_flag[j] = 0;
}
}

In Table 1, j is the number of the transform block (0: upper left block of FIG. 9, 3: lower right block of FIG. 9), T8_flag is transform type information included in side information transmitted from the encoding apparatus, and T8_flag[] is encoding Transform type information corresponding to each 8×8 block using the block pattern information (CBP).

According to Table 1, the transmitted 2-bit additional information may be restored to a total of 4 bits for each transform block using pattern information of the coding block.

As described above, although the present invention has been described with reference to limited embodiments and drawings, the present invention is not limited to the above-described embodiments, and those skilled in the art to which the present invention pertains can make various modifications and variations from these descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the following claims as well as the claims and equivalents.

610: prediction block generator
620: difference block generating unit
630: conversion unit
640: final block decision unit
650: quantization unit
660: entropy encoding unit

Claims (3)

a prediction block generator generating a prediction block by performing prediction on the coding block;
a residual block generator for generating a residual block for the coding block based on the coding block and the prediction block; and
an encoder for encoding the residual block;
The encoding unit,
When the generation of the prediction block and the transform on the residual block are not omitted, the transform is performed based on transform type information for performing the transform and the maximum transform size, and quantization is performed on the residual block, Entropy encoding is performed on the residual block,
The transform type information includes a size of a transform block,
The maximum transform size is a maximum size allowed for the transform block and is determined to be equal to a maximum size allowed for the coding block or smaller than a maximum size allowed for the coding block.
a decoding unit that obtains a coded residual block with respect to the coded block from the bitstream and decodes the coded residual block;
a prediction block generator configured to generate a prediction block by performing prediction on the coding block; and
and a reconstruction unit for reconstructing the coding block based on the decoded residual block and the generated prediction block;
When entropy decoding is performed on the residual block, inverse quantization is performed on the residual block, and generation of the prediction block and inverse transform on the residual block are not omitted, based on the transform type information and the maximum transform size performing the inverse transformation,
The transform type information includes a size of a transform block,
The maximum transform size is a maximum size allowed for the transform block and is determined to be equal to a maximum size allowed for the coding block or smaller than a maximum size allowed for the coding block.
A computer-readable recording medium for storing a bitstream generated by a video encoding apparatus, the video encoding apparatus comprising:
a prediction block generator generating a prediction block by performing prediction on the coding block;
a residual block generator for generating a residual block for the coding block based on the coding block and the prediction block; and
an encoder for encoding the residual block;
The encoding unit,
When the generation of the prediction block and the transform on the residual block are not omitted, the transform is performed based on transform type information for performing the transform and the maximum transform size, and quantization is performed on the residual block, Entropy encoding is performed on the residual block,
The transform type information includes a size of a transform block,
The maximum transform size is a maximum size allowed for the transform block, and is determined to be the same as the maximum size allowed for the coding block or smaller than the maximum size allowed for the coding block.

Images