KR101968268B1 - Apparatus and Method of Context-Adaptive Quantization and Inverse Quantizationfor IPCM block - Google Patents

Abstract

The present invention relates to a method and apparatus for performing context-based adaptive quantization and inverse quantization on a block mode with respect to a block having an IPCM prediction mode in an image encoding / decoding apparatus. The context-based adaptive quantization and dequantization method and apparatus for the IPCM block mode include context information such as block mode and size of surrounding blocks of blocks located around the block to be encoded / decoded and quantization / dequantization parameter values Quantization value / dequantization value of an IPCM block to be encoded / decoded by using the decoded value / dequantization value, and performs quantization and dequantization using the value.

Description

[0001] The present invention relates to a context-based adaptive quantization and inverse quantization method and apparatus for an intra-picture pulse code modulation block,

The present invention relates to an image encoding / decoding apparatus and method, and more particularly, to an image encoding / decoding apparatus and method capable of performing context-based adaptive quantization and inverse quantization on a block mode having an IPCM prediction mode will be.

H.264 / AVC and HEVC include IPCM mode in the intra-picture coding method. The IPCM mode is a method for encoding all pixels in a macroblock or CU (Coding Unit) by PCM. This intra-picture pulse code modulation (IPCM) mode performs PCM on pixels in the block without performing prediction, conversion, quantization, and entropy coding on the input values. For example, H.264 / AVC encodes the pixels of a corresponding macroblock with 384 bytes for 384 pixels in a 4: 2: 0 format image. This IPCM mode solves the problem of generating a large number of bits by limiting the maximum bit amount in the case of having a probability distribution statistic discussed in a specific block in an entropy encoding step of performing coding based on probability distribution statistics.

However, when the image is lossy coded, the surrounding blocks are quantized by the high quantization parameters, and the blocking phenomenon occurs at the block boundary when the IPCM blocks are not quantized. Also, this blocking phenomenon has a problem of lowering the subjective image quality because the filtering is not performed when it is judged as the edge boundary even in the existing deblocking filter. Accordingly, there is a need for a context-adaptive quantization and quantization method and apparatus for quantizing an image by lossy encoding similar to surrounding quantized blocks even in IPCM mode.

The problem to be solved by the present invention is to quantize and dequantize the IPCM mode adaptively to the IPCM mode by using quantization parameters of surrounding quantized blocks or CUs and context information such as the size of blocks or CUs in the IPCM mode And to provide a method and apparatus for carrying out the present invention. There is also provided a method and apparatus for predicting a quantization parameter differential value when performing quantization and inverse quantization on the IPCM mode.

The present invention has been made in view of the above problems, and it is an object of the present invention to provide an apparatus and method for controlling the same.

According to an aspect of the present invention, there is provided an image encoding apparatus comprising: a PPS (Picture Parameter Set) and an IPCM mode quantization unit for transmitting a flag indicating whether or not quantization is applied to a block having an IPCM prediction mode at a slice header level An IPCM block quantization applying flag deriving unit for deriving an IPCM block quantization applying flag to be recorded using a flag to be transmitted, a flag information transmitting unit, and an IPCM block quantization applying flag deriving unit for applying an IPCM block quantization applying flag to an IPCM block A block-based IPCM mode quantization applying judging unit for judging whether or not the IPCM block quantization is applied, a block quantization applying flag judging unit, a context-based determining unit for deriving a quantization parameter adaptively according to a quantization parameter and a block size of a nearby CU or macroblock Adaptive IPCM mode quantization parameter derivation value derivation unit, And an IPCM mode quantization performance unit quantization performance unit for performing quantization for the IPCM block using the derived quantization parameter.

According to an aspect of the present invention, there is provided an apparatus and method for decoding an IPCM block, the apparatus comprising: a decoder for decoding an IPCM block inverse quantization applied to a PIC and a slice header; And an IPCM block dequantization application flag unit for determining whether or not the dequantization of the IPCM block is applied to the IPCM block in units of CU or macroblocks using the decoded derived IPCM mode inverse block quantization applying flag information An adaptive IPCM mode inverse quantization parameter deriving unit for deriving an inverse quantization parameter adaptively according to a quantization parameter and a block size of a neighboring CU or macro blocks, The inverse quantization of the IPCM block is performed using the derived inverse quantization parameter. IPCM mode performs inverse quantization to include parts.

The context-based adaptive quantization and inverse quantization method and apparatus for an IPCM block mode according to an embodiment of the present invention enables effective quantization and inverse quantization for an IPCM mode in a macroblock or quad tree-based block coding and decoding . This context adaptive quantization can improve the subjective image quality by eliminating the blocking phenomenon occurring at the block boundary occurring when the quantization parameter is encoded with a high quantization parameter by matching the distribution of the quantization error with the surrounding quantization block for the IPCM mode, It is possible to improve the compression ratio.

FIG. 1A is a block diagram of a quantization method and apparatus for a macroblock or CU having an IPCM mode in a video encoding apparatus according to a first embodiment of the present invention.
1B is a block diagram of a video decoding apparatus according to a first embodiment of the present invention.
1C illustrates a decoding method and apparatus for a macroblock or CU having an IPCM mode in a video decoding apparatus according to a first embodiment of the present invention.
2A illustrates syntax context construction for quantization and dequantization for an IPCM mode written in a picture parameter set according to a first embodiment of the present invention.
FIG. 2B relates to a context of a suspicion regarding whether to apply quantization and / or quantization for an IPCM mode recorded in a slice header according to the first embodiment of the present invention.
FIG. 3 relates to quantization parameters and a quantization step table for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention.
FIG. 4A relates to quantization performance for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention.
FIG. 4B relates to performing inverse quantization when the quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is zero.
FIG. 5A illustrates quantization performed when a quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is 1. FIG.
FIG. 5B relates to performing inverse quantization when the quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is 1. FIG.
FIG. 6A relates to quantization performed when the quantization step value 310 for a macroblock or CU having an IPCM mode according to the third embodiment of the present invention is 2. FIG.
FIG. 6B relates to performing inverse quantization when the quantization step value 310 for a macroblock or a block for a CU having an IPCM mode according to the third embodiment of the present invention is 2.
FIG. 7A relates to the operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170, 170 according to the first embodiment of the present invention.
7B is an operation example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the first embodiment of the present invention.
8A is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the third embodiment of the present invention.
FIG. 8B illustrates an operation of context-based adaptive IPCM block mode quantization / dequantization parameter value derivation units 120 and 170 according to the third embodiment of the present invention.
FIG. 9A is another example of the operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the fourth embodiment of the present invention.
FIG. 9B is another example of operation of the context-based adaptive IPCM block mode quantization / inverse quantization parameter value derivation unit 120, 170 according to the fourth embodiment of the present invention.
FIG. 10A is another example of the operation of the context-based adaptive IPCM block mode quantization / inverse quantization parameter value derivation unit 120, 170 according to the fifth embodiment of the present invention.
FIG. 10B is another example of operation of the context-based adaptive IPCM block mode quantization / inverse quantization parameter value derivation unit 120, 170 according to the fifth embodiment of the present invention.
FIG. 11A is a flowchart illustrating an operation of the context-based adaptive IPCM block mode quantization / inverse quantization parameter value derivation unit 120, 170 according to the 16th embodiment of the present invention.
FIG. 11B is an example of other operations of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the sixth embodiment of the present invention.
12A is an example of another operation related to the operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the seventh embodiment of the present invention.
12B is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the seventh embodiment of the present invention.
FIG. 13 is another example of operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the eighth embodiment of the present invention.
FIG. 14 shows an example of a prediction operation of a Delta QP in an encoder and a decoder according to the first embodiment of the present invention.

Hereinafter, a context-based adaptive quantization and inverse quantization method and apparatus for a macroblock or CU having an IPCM mode according to an embodiment of the present invention will be described with reference to the accompanying drawings.

FIG. 1A is a block diagram of a quantization method and apparatus for a macroblock or CU having an IPCM mode in a video encoding apparatus according to a first embodiment of the present invention.

Referring to FIG. 1A, a context-based adaptive quantization method and apparatus for a macroblock or CU having an IPCM mode includes an IPCM block mode quantization applying flag information transmitting unit 100, an IPCM block mode quantization applying determining unit 110, A context-based adaptive IPCM mode quantization parameter derivation unit 120, and an IPCM mode quantization performance unit 130 for performing quantization.

The IPCM mode quantization applying flag information transmitting unit 100 transmits a quantization applying flag for a macroblock or CU having an IPCM mode in a picture parameter set and a slice header of an image.

The IPCM mode quantization application determination unit 110 on a block-by-block basis determines whether to apply quantization for the IPCM mode in units of macroblocks or CUs using the derived quantization flag value to be transmitted. When it is determined by the IPCM mode quantization application determination unit 110 on a block basis that the current macroblock or the IPCM mode of the CU is not to be quantized, the quantization is not performed on the corresponding macroblock or CU.

If it is determined that the IPCM mode quantization application determination unit 110 on a block-by-block basis is to perform quantization on the IPCM mode of the current macroblock or CU, the context-based adaptive IPCM mode quantization parameter derivation unit 120 And derives the quantization parameter for the IPCM mode adaptively using the block size, the position information, and the quantization parameter values of the macroblock or CU around the macroblock or CU.

The IPCM mode quantization performing unit 130 performs quantization using a quantization parameter derived from a macroblock or CU having an IPCM mode.

1B is a block diagram of a video decoding apparatus according to a first embodiment of the present invention.

1B, the image decoding apparatus includes an entropy decoding unit 140, an IPCM mode block decoding unit 141, a reordering unit 142, an inverse quantization unit 143, an inverse transform unit 144, an intra / 145, and a filtering unit 146.

The entropy decoding unit 140 performs entropy decoding on the input bitstream. If the prediction mode of the current macroblock or CU to be decoded is determined to be the IPCM mode through the entropy decoding process, the decoding is performed through the IPCM mode block decoding unit 141. At this time, the reordering unit 142, the inverse quantization unit 143, the inverse transform unit 144, and the intra / inter predictor 145 are not used, and the output value of the IPCM mode block decoding unit 141 is input to the filtering unit 146 ).

If the prediction mode of the current macroblock or CU to be decoded is not the IPCM mode, the differential signal is decoded through the rearrangement unit 142, the inverse quantization unit 1430, and the inverse transform unit 144, Is added to the predictive value obtained through the predictor 145 and the macroblock or CU is first decoded. The decoded macroblock or CU is used as an input to the filtering unit 146. [

1C illustrates a decoding method and apparatus for a macroblock or CU having an IPCM mode in a video decoding apparatus according to a first embodiment of the present invention.

1C, a context-based adaptive inverse quantization method and apparatus for an IPCM block includes an IPCM mode inverse quantization applying flag information decoding unit 150, an IPCM mode inverse quantization applying unit 160 in a block unit, a context- An inverse IPCM mode inverse quantization parameter deriving unit 170, and an IPCM mode inverse quantization performing unit 180 for performing inverse quantization.

The IPCM mode dequantization application flag information decoding unit 150 is located in the entropy decoding unit 140 and includes an IPCM mode dequantization application determination unit 160 for each block and a context based adaptive IPCM mode dequantization parameter derivation unit 160. [ And an IPCM mode inverse quantization performing unit 180 for performing inverse quantization are present in the IPCM mode block decoding unit 141. [

The IPCM mode dequantization application flag information decoder 150 derives and decodes flags regarding whether to apply inverse quantization to a macroblock or a CU having an IPCM mode in a picture parameter set and a slice header of an image.

The IPCM-mode inverse quantization application determination unit 160 on a block-by-block basis determines whether to apply inverse quantization to the IPCM mode in units of macroblocks or CUs using the derived decoded quantization flag values. If the IPCM mode dequantization application determination unit 160 determines that de-quantization is not performed on the IPCM mode of the current macroblock or CU, the de-quantization is not performed on the corresponding macroblock or CU.

The context-based adaptive IPCM mode inverse quantization parameter deriving unit 170 adaptively adjusts the inverse of the IPCM mode using the block size, position information, and inverse quantization parameter values of the macroblock or CU around the macroblock or CU to be decoded. And derives a quantization parameter value.

The IPCM mode inverse quantization performing unit 180 performs inverse quantization on the corresponding macroblock or CU using the inverse quantization parameter values derived for the IPCM block.

FIG. 2A relates to a syntax context configuration regarding whether to apply quantization and inverse quantization for an IPCM mode written in a picture parameter set according to the first embodiment of the present invention.

If the ipcm_quantization_flag value (200) is 1 in the picture parameter set, quantization and inverse quantization can be selectively applied to the macroblock or CU having the IPCM mode on a slice basis. When the ipcm_quantization_flag value 200 is 0 in the picture parameter set, quantization and inverse quantization are not applied to the macroblock or CU having all the IPCM modes in the picture.

If the ipcm_qstep_table_flag value 201 is 1 in the picture parameter set, the mapping table of the quantization parameter value and the quantization step value for IPCM mode quantization can be transmitted. The transmitted table is used in quantization and inverse quantization in the encoder and decoder. When the corresponding value is 0, the encoder and the decoder use a predetermined mapping table.

If the ipcm_qstep_table_flag value 201 is 1 in the picture parameter set, the quantization step value mapped according to the QP value is encoded and decoded through the ipcm_qauntization_step value 202.

FIG. 2B is a syntax context of quantization and inverse quantization applied to the IPCM mode recorded in the slice header according to the first embodiment of the present invention.

When the ipcm_quantization_flag value 200 is 1 in the picture parameter set, the disable_ipcm_quantization_flag value 210 is additionally coded and decoded in units of slices. When the disable_ipcm_quantization_flag value 210 is 1, the encoder performs quantization and inverse quantization according to the present invention in the encoder and the decoder in all the IPCM modes occurring in the slice. If the corresponding value is 0, the encoder does not perform quantization and inverse quantization according to the present invention for all IPCM modes occurring in the slice. In the decoder, if the disable_ipcm_quantization_flag value 210 is 1 in units of slices, inverse quantization is performed on all IPCM blocks in the slice. When the corresponding value is 0 in the decoder, the decoder does not perform inverse quantization on all IPCM blocks generated in the slice.

FIG. 3 relates to quantization parameters and a quantization step table for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention.

3, in accordance with a QP value 300 adaptively determined through a CU size and quantization parameters of a macroblock having an IPCM mode or a neighboring macroblock to be quantized for a CU or a block or a coding unit of the CU, Lt; RTI ID = 0.0 > 310 < / RTI > The quantization parameter and the quantization step table can be transmitted to the decoder according to the value of ipcm_qstep_table_flag in FIG.

FIG. 4A relates to quantization performance for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention.

The quantization process for the macroblock or CU having the IPCM mode is performed according to the quantization step value 310 for the macroblock or CU having the IPCM mode. In FIG. 4A, the quantization input values 400 of the IPCM mode are output to the quantization output values 410 of the IPCM mode after the quantization process. If the quantization step value 310 for the macroblock or the CU is 0, the input pixel value 420 in the IPCM mode is outputted as it is to the same value 430 without quantization, and it operates in the same manner as the existing IPCM encoding .

FIG. 4B relates to performing inverse quantization when the quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is zero.

In FIG. 4B, the inverse quantization input values 450 of the IPCM mode are output to the inverse quantization output values 460 of the IPCM mode after the inverse quantization process. When the quantization step value 310 for the macroblock or CU having the IPCM mode is 0, the inverse quantization input value 470 is output as the inverse quantization output value 480 having the same brightness value.

FIG. 5A illustrates quantization performed when a quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is 1. FIG.

In FIG. 5A, the quantization input values 500 of the IPCM mode block are quantized and output to the quantization output values 510 of the IPCM mode block. If the quantization step value 310 for the IPCM mode block is 1, then the two brightness levels of the input are quantized to one brightness level. Two of the quantization input values in the IPCM mode block 520 are quantized into one output value 530. [

FIG. 5B illustrates inverse quantization performed when the quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is 1. FIG.

In FIG. 5B, the inverse quantization input values 550 of the IPCM mode block are output to the inverse quantization output values 560 of the IPCM mode block after undergoing inverse quantization. Quantization step value 310 for the IPCM mode block is 1, it is inversely quantized to a value 580 twice of the inverse quantization input value 570. [

FIG. 6A illustrates quantization performed when a quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is 2. FIG.

In FIG. 6A, the quantization input values 600 of the IPCM mode block are quantized and output to the quantization output values 610 of the IPCM mode block. If the quantization step value 310 for the IPCM mode block is 2, then the three brightness levels of the input are quantized to one brightness level. Three of the brightness values 620 of the IPCM mode block quantization input values are quantized into one output value 630.

6B illustrates inverse quantization performed when the quantization step value 310 for a macroblock or CU having an IPCM mode according to the first embodiment of the present invention is 2. FIG.

In FIG. 6B, the inverse quantization input values 650 of the IPCM mode block are output to the inverse quantization output values 660 of the IPCM mode block after undergoing inverse quantization. If the quantization step value 310 for the IPCM mode block is 2, the inverse quantization input value 670 is multiplied by 3 and then minus 1 to dequantize the addition value 680.

FIG. 7A relates to the operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170, 170 according to the first embodiment of the present invention.

The context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120, 170 and 170 encode the current CU 720 in the IPCM mode based on the quad tree, and the CUs 710, 711 and 712 and the quantization / inverse quantization parameter values of the CUs 700, 701 and 702 located at the upper boundary and the block size information of the CU are used to adaptively determine the QP value 300, And determines a quantization / dequantization step 310 for the mode block. The quantization value / dequantization value of the current CU 720 in the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170, 170 is the value of the quantization value / dequantization value of the current CU 72002 (Tc) QP value and the QP value of the largest CU (712; Lc) in the left CUs. In the low complexity coding and decoding mode, it may be determined as an average value of the quantization / inverse quantization parameter values of the leftmost CU (700, Ta) in the upper CUs and the uppermost CU (710, La) in the left CUs in the current CU 720 have.

7B is another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the first embodiment of the present invention.

The context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120 and 170 may be configured to encode the current CU 750 in the IPCM mode on a quadtree basis, in the LCU (Largest Coding Unit) The QP value 300 is adaptively determined using the quantization / dequantization parameter values of the CUs 730, 731, and 732 located at the upper boundary and the CUs 740, 741, and 742 located at the upper boundary and the CU block size information And uses this value to determine the quantization / dequantization step 310 for the IPCM mode block. The context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 determines whether the quantization / dequantization value of the current CU 750 is greater than the QP value of the largest CU 742 (Tc) in the upper CUs And the average value of the QP values of the largest CU (730, La) in the left CUs. In the low complexity coding and decoding mode, the leftmost CU 740 (Ta) in the upper CUs and the quantization / dequantization parameter values of the uppermost CU 730 (La) in the left CUs are determined to be the current values of the current CU 750 .

8A is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

When context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120 and 170 use the context information of CUs located on the left and above, the largest CUs are selected and the quantization / dequantization parameter And determines the quantization / dequantization value of the current CU as an average value of the values. In this case, if there are several CUs in one direction, the leftmost CU (800; Ta) is selected in the upper boundary, and the CU (820; La) And determines the quantization / dequantization parameter value of the current CU 840 as an average value of the quantization / dequantization parameter values of the two CUs.

8B is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

The context-based adaptive IPCM block mode quantization / dequantization parameter value derivation units 120 and 170 select the largest CUs when using the context information of CUs located on the left and upper sides in the LCU, / The dequantization value of the current CU 890 is determined as an average value of the dequantization values. In this case, if there are several largest CUs, select the leftmost CU (850; Ta) in the upper boundary, select the uppermost CU (870; La) The quantization / dequantization parameter value of the current CU 890 is determined as an average value of the quantization / dequantization parameter values of the CU.

9A is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

When the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120 and 170 refer to the context information of neighboring CUs, there is no CUs on the left side in the left border of the image. In this case, the quantization / dequantization value of the CU having the largest block size among the CUs 900, 910, and 920 (Ta, Tb, Tc) located at the upper side is used as the quantization / dequantization parameter value of the current CU 930.

9B is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the first embodiment of the present invention.

The context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120 and 170 may determine that the CUs 940, 950 and 960 The quantization value / inverse quantization value of the CU 960 (Tc) having the largest block size among the quantized values (Ta, Tb, Tc) is used as the quantization / inverse quantization parameter value of the current CU 970.

10A is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

If context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120 and 170 can use only context information of CUs located at the upper side, CUs 1000, The quantization value / inverse quantization value of the leftmost CU (1000; Ta) from the upper side is used as the quantization / inverse quantization parameter value of the current CU 1020 when the number of the input values is Ta and Tb.

FIG. 10B is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the first embodiment of the present invention.

When the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120 and 170 use only the context information of CUs located at the upper positions in the LCU, the CUs 1050 The quantization value / dequantization value of the CU 1050 (Ta) located at the leftmost position from the upper side is used as the quantization / dequantization parameter value of the current CU 1070. [

11A is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

When only the context information of the CUs 1100, 1110, 1120 (La, Lb, Lc) located on the left side can be used in the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170, The quantization value / dequantization value of the CU 1120, Lc having the largest block size among the CUs is used as the quantization / dequantization parameter value of the current CU 1130.

11B is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

The context-based adaptive IPCM block mode quantization / dequantization parameter value deriving units 120 and 170 can use only the context information of the CUs 1150, 1160, 1170 (La, Lb, Lc) located on the left side in the LCU The quantization / dequantization value of the CU 1170 (Lc) having the largest block size among the CUs is used as the quantization / dequantization parameter value of the current CU 1180.

12A is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 according to the first embodiment of the present invention.

When only the context information of the CUs 1200, 1210 (La, Lb) located at the left in the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 can be used, When there are a plurality of CUs having the largest block size, the quantization / dequantization values of the uppermost CU 1200 (La) among the CUs on the left side are used as the quantization / dequantization parameter values of the current CU 1220 .

12B is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

When the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation units 120 and 170 can use only the context information of CUs 1250 and 1260 (La and Lb) located on the left side in the LCU, When there are a plurality of CUs having the largest block size among the CUs, the quantization / dequantization values of the uppermost CU 1250 (La) among the CUs on the left side are converted into the quantization / dequantization values of the current CU 1270 Value.

FIG. 13 is an example of another operation of the context-based adaptive IPCM block mode quantization / dequantization parameter value deriving unit 120, 170 according to the first embodiment of the present invention.

The context-based adaptive IPCM block mode quantization / dequantization parameter value derivation unit 120, 170 derives the quantization / dequantization parameter values of the current CU at the boundary of the LCU or in the context of the surrounding CUs within the LCU , The quantization value / dequantization value of the largest CU on the left and top is used. In this case, when the reference CUs 1330 and 1350 (Tc and Lc) are coded / decoded in the IPCM mode and the quantization / dequantization in the IPCM block mode is applied, the quantization / Inverse quantization parameter value 300 calculated in the operation of the adaptive IPCM block mode quantization / dequantization parameter value derivation unit 110 and 170 of the corresponding CUs as the quantization parameter values of the corresponding CUs.

FIG. 14 relates to a prediction operation of a Delta QP in an encoder and a decoder according to the first embodiment of the present invention.

The QP value is assigned in units of CU, and the difference value of the QP value allocated in each CU unit is written as actual bits. In the process of predicting the QP value, when the current CU 1460 predicts the quantization value / dequantization value from the CU 1450 in which the IPCM quantization / dequantization has been performed, the CU 1460 performing the IPCM quantization / 1450 are quantized / dequantized parameter values 300 that are contextually determined in the context-based adaptive IPCM block mode quantization / dequantization parameter value derivation units 1210, 170.

Claims (1)

Obtaining inverse quantization parameters from a plurality of neighboring coding units adjacent to a current coding unit;
Deriving an inverse quantization parameter of the current coding unit using the inverse quantization parameters of the neighboring coding units;
Determining whether the inverse quantization is applied to the current coding unit based on information on whether or not the inverse quantization is applied, which is derived from the input bitstream; And
And decoding the current coding unit by selectively performing inverse quantization based on the inverse quantization parameter of the current coding unit according to the determination as to whether the inverse quantization is applied,
Wherein the plurality of neighboring coding units include an upper neighboring coding unit and a left neighboring coding unit adjacent to the current coding unit,
Wherein the upper neighboring coding unit is a coding unit located at the leftmost of a plurality of neighboring coding units existing at an upper end of the current coding unit when a plurality of neighboring coding units exist at an upper end of the current coding unit,
The left neighboring coding unit is a coding unit located at the top of a plurality of neighboring coding units existing on the left side of the current coding unit when a plurality of neighboring coding units exist on the left side of the current coding unit,
Wherein deriving the dequantization parameter of the current coding unit comprises:
Calculating an inverse quantization parameter predictive value of the current coding unit using an inverse quantization parameter of the upper neighboring coding unit and an inverse quantization parameter of the left neighboring coding unit; And
And deriving an inverse quantization parameter of the current coding unit using the inverse quantization parameter prediction value of the current coding unit and the inverse quantization parameter difference value of the current coding unit.

Images