KR20200032681A - Apparatus and Method of Adaptive Quantization Parameter Encoding and Decoder based on Quad Tree Structure - Google Patents

Abstract

Disclosed are an image encoding/decoding method and apparatus. According to the present invention, an image decoding apparatus comprises: an inverse quantization difference value recording block unit determination part that determines a unit of a block in which an inverse quantization difference value is recorded; an inverse quantization parameter prediction block determination part that determines a block to be used for inverse quantization parameter prediction by using position information of a neighboring block placed around the block to decode an inverse quantization parameter assigned to the block; an inverse quantization parameter value deriving part that obtains an inverse quantization parameter value by using an inverse quantization parameter predicted value based on the inverse quantization difference value and the block to be used for the inverse quantization parameter prediction; and an inverse quantization part that performs inverse quantization by using the inverse quantization parameter value.

Description

Apparatus and Method of Adaptive Quantization Parameter Encoding and Decoder based on Quad Tree Structure}

The present invention relates to an image encoding / decoding apparatus and method, and more specifically, for a CU in an LCU, displays a block having a quantization / dequantization difference value based on a quart tree structure, and surrounds a block to be encoded / decoded The present invention relates to an image encoding / decoding method and apparatus for adaptively predicting / decoding quantization / dequantization parameter values using context information of blocks located at.

HEVC encodes / decodes the input image in units of CU (Coding Unit). The CU of the largest size in a frame is called a Large Coding Unit (LCU), and the LCU may be divided into a number of CUs based on quadtree splitting information, and then encoding / decoding may be performed. The quantization parameter value of HEVC is assigned one value in LCU units, and predicts the quantization parameter value of the LCU to be currently encoded in the previously located LCU based on the raster scan order.

H.264 / AVC is encoded / decoded in units of macroblocks, and has a quantization / inverse quantization value in units of macroblocks. The quantization parameter value allocated in units of macroblocks is predicted from the quantization parameter value of the macroblock located to the left in the frame. Coding is performed by writing down a corresponding value in a macroblock to be encoded for a difference value generated after a process of predicting a value of a quantization parameter. The decoder decodes the quantization parameter value by adding the quantization parameter difference value decoded in the entropy decoding step and the quantization parameter value of the macroblock located on the left side.

However, when a relatively large LCU is allocated compared to the size of an input image, bitrate control cannot be effectively performed using quantization parameter values recorded in LCU units. In addition, when a quantization parameter value is assigned in units of CUs, a subjective image quality deterioration problem may occur due to a difference in quantization parameter values from neighboring CUs. Accordingly, quantization parameter values can be assigned to various block sizes from CU to LCU units according to the input image, and it is necessary to optimize the prediction direction of quantization parameters using context information of blocks located around blocks to be encoded. .

The problem to be solved by the present invention is to provide a method and apparatus for encoding / decoding quantization / dequantization parameter values based on various quadtree structures based on partition information of a CU. In addition, the quantization / dequantization parameter encoding / decoding method and apparatus also provide a method and apparatus for predicting a quantization / dequantization parameter value in an effective direction using context information of a neighboring block.

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems that are not mentioned will be clearly understood by those skilled in the art from the following description.

An image encoding apparatus according to an embodiment of the present invention for solving the above problems is a block in which a difference value of a quantization parameter is recorded when an LCU in an image is divided into a plurality of CUs in a quadtree form or is encoded as a single CU A quantization difference value recording unit determining a unit of a block unit determination unit, a quantization unit performing quantization using a quantization value allocated in block units, and context information of neighboring blocks to predict a quantization value used in a block to be encoded A quantization prediction block determination unit to adaptively determine a prediction block, a quantization parameter difference value generation unit to generate a quantization difference value of a block to be encoded using quantization parameters of a prediction block based on context information, and record a quantization difference value Split information for block units and quantization parameters in the block Minutes and a quantization parameter recording unit operable to record the value.

An image decoding apparatus according to an embodiment of the present invention for solving the above-described problem is an inverse quantization parameter difference value block division flag derivation unit, a decoded inverse quantization, which decodes information on a block having an inverse quantization parameter difference value in an LCU unit. Inverse quantization difference value recording block unit determining unit determining inverse quantization difference value in the LCU using the parameter difference value block division flag, inverse quantization parameter difference value Inverse quantization parameter difference value according to the block division flag Inverse quantization parameter difference value derivation unit to decode, Inverse quantization parameter value prediction block determination unit to determine a block used for prediction using context information of a neighboring block to decode an inverse quantization parameter value of a block to be decoded, an inverse quantization process To decode the inverse quantization parameter value used in Inverse quantization parameter value derivation unit, using the decoded inverse quantization parameter value and comprising an inverse quantizer that performs inverse quantization.

An adaptive quantization / dequantization parameter encoding and decoding method and apparatus based on a quadtree structure according to an embodiment of the present invention enables to allocate quantization parameter difference values at various levels when a block is divided into a quadtree structure . The allocation of the difference values of the quantization parameters of various levels enables a more precise bit amount adjustment than the allocation of a single quantization parameter value in LCU units. In addition, when predicting / decoding a quantization / dequantization parameter value in a block unit, quantization with a neighboring block by adaptively determining a prediction direction using context information of a neighboring block as well as a zigzag scan method based on a quadtree structure The problem of subjective deterioration that can occur due to a large difference in value can also be solved.

1A is a method and apparatus for adaptively assigning a quantization parameter value to a block having a quadtree structure and encoding it in a video encoding apparatus according to a first embodiment of the present invention.
1B illustrates a method and apparatus for adaptive inverse quantization parameter decoding based on a quadtree structure in an image decoding apparatus according to a first embodiment of the present invention.
2 is a configuration of an image decoding apparatus according to a first embodiment of the present invention.
3 is a context for controlling a quantization difference value based on a quadtree structure recorded in a sequence parameter set according to a first embodiment of the present invention.
4 is for variables set as initial values in slice data according to the first embodiment of the present invention.
5 is related to the context of a quantization / dequantization parameter difference value recorded in CU units and a condition in which the difference value exists according to the first embodiment of the present invention.
6 is related to the operation of the quantization difference value recording block unit determination unit and the inverse quantization difference value recording block unit determination unit according to an embodiment of the present invention.
7 relates to a quantization parameter value prediction block determination unit and an inverse quantization parameter value prediction block determination unit according to the first embodiment of the present invention.
8 relates to a quantization parameter value prediction block determination unit and an inverse quantization parameter value prediction block determination unit according to a second embodiment of the present invention.
9 relates to a quantization parameter value prediction block determination unit and an inverse quantization parameter value prediction block determination unit according to a third embodiment of the present invention.
10 relates to a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a fourth embodiment of the present invention.
11 relates to a quantization parameter value prediction block determination unit and an inverse quantization parameter value prediction block determination unit according to a fifth embodiment of the present invention.

Hereinafter, an adaptive quantization / dequantization parameter encoding and decoding apparatus based on a quadtree structure according to an embodiment of the present invention will be described with reference to the accompanying drawings.

1A is a method and apparatus for adaptively assigning a quantization parameter value to a block having a quadtree structure and encoding it in a video encoding apparatus according to a first embodiment of the present invention.

Referring to FIG. 1A, a method and apparatus for encoding an adaptive quantization parameter based on a quadtree structure include a quantization difference value recording block unit determination unit 100, a quantization unit 101, a quantization parameter value prediction block determination unit 102, and a quantization parameter And a difference value generator 103 and a quantization parameter recorder 104.

The quantization difference value recording block unit determination unit 100 determines a block unit to record a quantization difference value for each CU unit or a plurality of CU bundles based on CU partition information of a corresponding LCU in an LCU unit in an image. do. The information of the block that records the quantization difference value may be composed of a quadtree structure.

The quantization unit 101 quantizes an input block using a value of a quantization parameter assigned to the block.

The quantization parameter value prediction block determiner 102 uses the context information of the CU located around the CU to predict the quantization value to effectively encode the quantization value assigned to each CU or an arbitrary CU. Decide. As the context information, a block size, a block position, and a block prediction mode may be used.

The quantization parameter difference value generation unit 103 generates a quantization parameter difference value by subtracting the quantization value of the current block from the quantization parameter value of the quantization parameter prediction block determined by the quantization parameter value prediction block determination unit 102.

The quantization parameter recorder 104 is used to indicate the segmentation information of a block including a sequence parameter set, flag information about application / non-application in slice units, and a quantization difference value in an LCU for adaptive quantization parameter encoding based on a quadtree structure. It serves to entropy code the flag information and the quantization parameter difference value.

1B illustrates a method and apparatus for adaptive inverse quantization parameter decoding based on a quadtree structure in an image decoding apparatus according to a first embodiment of the present invention.

Referring to FIG. 1B, a method and apparatus for decoding an adaptive inverse quantization parameter based on a quadtree structure include an inverse quantization parameter difference value block division flag derivation unit 120, an inverse quantization difference value recording block unit determination unit 121, and an inverse quantization parameter It includes a difference value derivation unit 122, an inverse quantization parameter value prediction block determination unit 123, an inverse quantization parameter value derivation unit 124, and an inverse quantization unit 125.

The dequantization parameter difference value block division flag derivation unit 120 decodes a block division flag for a block having an inverse quantization parameter difference value in LCU units from a sequence parameter set and slice data.

The inverse quantization difference value recording block unit determination unit 121 determines blocks in which the inverse quantization parameter difference value is recorded using the decoded inverse quantization parameter difference value block splitting flag and the CU splitting flag. The information of the block that records the quantization difference value may be composed of a quadtree structure.

The dequantization parameter difference value deriving unit 122 derives an inverse quantization parameter difference value for each block determined by the inverse quantization difference value recording block unit determination unit 121.

The inverse quantization parameter value prediction block determination unit 123 determines a block to be adaptively referenced using context information of neighboring blocks when performing inverse quantization. As the context information, a block size, a block position, and a block prediction mode may be used.

An inverse quantization parameter value derivation unit 124 is an inverse quantization parameter value of the prediction block calculated by the inverse quantization parameter value prediction block determination unit 123 and an inverse quantization calculated by the inverse quantization parameter difference value derivation unit 122 By adding a parameter difference value, an inverse quantization parameter value to be used in the inverse quantum unit 125 is derived.

The inverse quantization unit 125 performs inverse quantization on the input block using parameters calculated by the inverse quantization parameter value deriving unit 124.

2 is a configuration of an image decoding apparatus according to a first embodiment of the present invention.

Referring to FIG. 2, the image decoding apparatus includes an entropy decoding unit 200, a quadtree-based inverse quantization parameter derivation unit 210, a reordering unit 220, an inverse quantization unit 230, and an inverse discrete cosine transform encoding unit 240. , An intra / inter prediction unit 250 and a filtering unit 260.

The entropy decoding unit 200 inverse quantization parameter difference value block division flag derivation unit 120 for deriving a block division flag used for adaptive inverse quantization based on a quadtree structure, quantization parameter from the derived block division flag An inverse quantization difference value recording block unit determination unit 121 for determining a recorded block, and an inverse quantization parameter difference value derivation unit 122 for decoding an inverse quantization parameter difference value recorded in the corresponding block.

The quadtree-based inverse quantization parameter derivation unit 210 inverse quantization parameter value for determining a prediction block referred to when decoding the inverse quantization parameter The prediction block determination unit 123 and the inverse quantization parameter of the prediction block and the derived inverse quantization parameter It includes an inverse quantization parameter value derivation unit 124 that calculates an inverse quantization parameter by adding a difference value, and the corresponding block serves to restore an inverse quantization parameter for a quadtree structured block.

3 is a context for controlling a quantization difference value based on a quadtree structure recorded in a sequence parameter set according to a first embodiment of the present invention.

When the value of cu_qp_delta_enabled_flag (300) is 1 in the sequence parameter set, it means that quantization / dequantization parameter difference values can be controlled for various quadtree blocks from the smallest CU to the largest CU in all slices in the sequence. do.

4A is for variables set as initial values in slice data according to the first embodiment of the present invention.

When a slice is divided into a quadtree structure and encoded / decoded, it is first divided into LCU units, which are quadtrees of the largest size, and encoding / decoding is performed in the order of sequential scanning methods for the LCUs. In the case of encoding / decoding each LCU, the corresponding LCU may be divided into a plurality of CU regions based on a quadtree structure, and such a splitting process may be performed until the size of the minimum CU.

In FIG. 4A, isCuQpDeltaCoded 400 is a variable for controlling a quantization / dequantization parameter difference value that can be recorded in each CU when an arbitrary CU is split into N CUs again. This value is always initialized to 0 before encoding / decoding for each LCU in the slice.

In FIG. 4A, the coding_tree 401 function is a function that performs encoding / decoding for one LCU in a slice. The meaning of the fourth argument value of this function is a flag indicating whether a quantization / dequantization parameter difference value exists in the CU, and at least one quantization / dequantization parameter difference value is recorded in the LCU unit, so each LCU in a slice is recorded. This value is always called as 1 before encoding / decoding.

4B is a context for inverse quantization parameter difference value block division recorded in a coding tree block according to the first embodiment of the present invention.

The coding tree block represents a context for a CU, and a CU of 2N × 2N size can be encoded / decoded after being divided into four CUs of N × N size according to split_coding_unit_flag 420. Alternatively, it may be no longer split into a CU with a smaller size, and may be encoded / decoded into a CU having a current size of 2N × 2N.

The current CU receives as input cu_qp_delta_exist_flag 421, which is a flag for whether a quantization / dequantization parameter difference value exists in a higher level CU. In addition, when the current CU of 2N × 2N size is divided into a CU of N × N size according to the value of split_coding_unit_flag (420), split_qp_delta_flag (422) is additionally encoded / decoded. This additional segmentation information is encoded / decoded only when the cu_qp_delta_enable_flag (300) value recorded in the sequence parameter set and the cu_qp_delta_exist_flag (421) input from the upper CU are 1. The split_qp_delta_flag (422) value is encoded / decoded only when the current CU of 2N × 2N size is divided into N × N, and the corresponding value is input as the value of cu_qp_delta_exist_flag (421) when encoding / decoding the lower CUs of N × N size. .

If the split_qp_detla_flag (422) value is 0, the current CU of 2N × 2N size is divided into CUs of N × N size, but the size of the block that records the quantization / dequantization parameter difference value is 2N × 2N to N × N size It means that it is no longer divided into blocks. When the split_qp_detla_flag (422) value is 0, the IsCuQpDeltaCoded (400) value is additionally initialized to 0, so that the quantization / dequantization parameter difference value is recorded only in the first N × N CU when divided into N × N CUs. Plays a role.

5 is related to the context of a quantization / dequantization parameter difference value recorded in CU units and a condition in which the difference value exists according to the first embodiment of the present invention.

If the CU is not in the skip mode, quantization / dequantization parameter difference values may be recorded in the corresponding CU. When the value of cu_qp_delta_exist_flag (500; 421) is 1, it means that the difference value of the quantization / dequantization parameter exists in the current CU. In this case, the value of cu_qp_delta (501) according to the value of cu_qp_delta_enabled_flag (300) recorded in the sequence parameter set It can be recorded in CU units. For example, when the value of cu_qp_delta_exist_flag (500; 421) is 1, when the value of cu_qp_delta_enabed_flag (300) is 0, cu_qp_delta 501 is not recorded.

As another example, when a 2N × 2N CU is divided into 4 N × N CUs in a quadtree structure, only one value of the quantization / dequantization parameter difference value may be recorded. In this case, the quantization / dequantization parameter difference value is recorded in the first CU among the 4 CUs, and the quantization / dequantization parameter difference value is not recorded in the remaining 3 CUs. At this time, since the split_qp_delta_flag (422) value that is encoded / decoded in the 2N × 2N CU is 0, the value of cu_qp_delta_exist_flag (500) input to the N × N CU is 0. Therefore, there is basically no cu_qp_delta (501) in the divided 4 N × N CUs, but by using the IsCuQpDeltaCoded variable, the cu_qp_delta (501) value can be recorded in the first N × N CU. For the other three CUs, the cu_qp_delta (501) value is not recorded because the first CU decodes cu_qp_delta (501) and then changes the IsCuQpDeltaCoded value to 1.

Even in this case, the condition of the value of cu_qp_delta_enabled_flag (300) recorded in the sequence parameter set is checked at the same time, and only when the corresponding value is 1, the value of cu_qp_delta (501) can be recorded.

6A relates to the operation of the quantization difference value recording block unit determination unit 100 and the inverse quantization difference value recording block unit determination unit 121 according to the first embodiment of the present invention.

A 2N × 2N LCU to be encoded / decoded can be divided into 4 N × N CUs, and each CU is further divided and processed. Even if the 2N × 2N sized LCU is divided into multiple CUs and encoded / decoded, when the split_qp_delta_flag (422) value is 0 as shown in FIG. 6A, the difference value of the quantization / dequantization parameter is recorded in the first CU of the LCU. For the remaining CUs of the LCU, the quantization / dequantization parameter values restored from the first CU are used as they are.

6B relates to the operation of the quantization difference value recording block unit determination unit 100 and the inverse quantization difference value recording block unit determination unit 121 according to the second embodiment of the present invention.

This is a case where a 2N × 2N LCU to be encoded / decoded is divided into four CUs in the first step, and the second CU of the CUs is further divided. In this case, when the flag for splitting is encoded / decoded in the first step, split_qp_delta_flag (422; 630), which is a split flag for quantization / dequantization parameter values, is additionally encoded / decoded. If the corresponding value is 1, it means that all four partitioned CUs have a quantization / dequantization parameter difference value, so the split flag split_qp_delta_flag (422; 631) for the quantization / dequantization parameter value is additionally encoded in the corresponding step. / Decrypted.

Even in this case, the second N × N CU is split until the third step, but since the split_qp_delta_flag (422; 631) block split flag for quantization / inverse quantization is 0, only one quantization / dequantization parameter value for multiple CUs Assigned.

6C relates to the operation of the quantization difference value recording block unit determination unit 100 and the inverse quantization difference value recording block unit determination unit 121 according to the third embodiment of the present invention.

6C shows a case where a 2N × 2N LCU to be encoded / decoded is divided into four CUs in the first step, and the second CU of the CUs is further divided. Even when the second CU is divided into 4 CUs and the divided CU is divided into 4 CUs again, the size of a block in which quantization / dequantization parameter difference values are recorded can be determined through the split_qp_delta_flag (422; 661) value. have. 6C shows that the CU has been divided up to three depth information, but the relative quantization / dequantization parameter difference value has up to two depth information.

7A relates to a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to the first embodiment of the present invention.

The encoder allocates quantization parameter values to CU blocks determined through the quantization difference value recording block unit determination unit 100, and uses the previously used quantization parameter values as it is for the remaining CU blocks. At this time, the quantization parameter value is predicted from the quantization parameter values of the previous CU blocks, and then only the quantization parameter difference value is actually encoded.

In the decoder, the inverse quantization parameter difference value deriving unit 122 decodes the inverse quantization parameter difference value, and then adds the difference value to the inverse quantization parameter value of the block used for prediction to calculate the inverse quantization parameter value.

The quantization parameter value prediction block determination unit 102 and the inverse quantization parameter value prediction block determination unit 123 each determine a neighboring block to be referred when the quantization parameter value of the current block is predicted by the encoder and the decoder.

When the quantization / dequantization parameter value is assigned to the CU / 720 to be encoded / decoded in FIG. 7A, CUs 710,711,712 and La, Lb positioned adjacent to the left boundary of the current CU 720 to predict the corresponding value , Lc) among CUs 700,701,702; Ta, Tb, Tc located adjacent to the upper boundary, the average value and the minimum value of quantization parameter values of CUs 712,702; Lc, Tc having the largest block size at each boundary. , Maximum value, etc.

7B is another example of a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to the first embodiment of the present invention.

Even if it is split and encoded / decoded into multiple CUs within the LCU, after selecting the CU with the largest block size among the CUs located adjacent to the current CU 750 to be encoded / decoded from the left border and the upper border, The average, minimum, and maximum values of the quantization parameter values used by the two CUs are used to predict the quantization parameter values of the current CU.

8A is related to a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a second embodiment of the present invention.

When predicting the quantization parameter value of the current CU 840 to be encoded / decoded, the CU of the largest block among CUs located adjacent to the CU is referenced. In this case, when there are a large number of CUs of the maximum size, the CU (820; La) located closest to the top in the left boundary is selected, and the CU (800; Ta) located closest to the left in the upper boundary. First, an average value, a maximum value, and a minimum value of quantization parameter values of the CU are used with reference.

8B is another example of a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to the second embodiment of the present invention.

Even when the LCU is divided into a plurality of CUs and is encoded / decoded, the CUs having the largest block size among CUs located adjacent to the current CU 890 to be encoded / decoded are referred to. At this time, if there are multiple CUs having the largest block size on the left border, the CU (870; La) positioned adjacent to the top is used as a reference block. When multiple CUs having the largest block size exist in the upper boundary, the CU (850; Ta) positioned adjacent to the leftmost side is used as a reference block. After determining the blocks to be referenced in the left and up directions, the maximum, minimum, and average values of the quantization values of the two blocks are obtained, and then the quantization parameter values of the CU 890 to be encoded / decoded are predicted using the values.

9A is related to a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a third embodiment of the present invention.

When predicting the quantization parameter value of the current CU 920 to be encoded / decoded, all referenceable CUs (900,901,902,910,911,912; Ta, Tb, Tc, La, Lb, Lc) located adjacent to the CU are selected as reference blocks. . The encoder predicts the quantization parameter value of the CU 920 to be encoded by using the maximum, minimum, and average values of the quantization parameter values of all CUs that can be referenced, and then codes the quantization parameter difference value.

The decoder decodes the inverse quantization parameter value of the CU 920 to be decoded by adding the decoded quantization parameter difference value to the maximum, minimum, and average values of the inverse quantization parameter values of all CUs that can be referenced.

9B is another example of a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a third embodiment of the present invention.

Quantization parameter values using all referenceable CUs (930,931,932,940,941,942; La, Lb, Lc, Ta, Tb, Tc) located adjacent to the periphery of the CU even when the CU 950 for encoding / decoding is located in the LCU Predict.

The encoder predicts the quantization parameter value of the CU 950 to be encoded from the average value, maximum value, and minimum value of quantization parameter values of all neighboring referenced CUs, and encodes only the difference value.

When the inverse quantization parameter value of the CU 950 to be decoded is restored by the decoder, the inverse quantization parameter value is decoded by adding the average value, the maximum value, and the minimum value of the dequantization parameter values of all surrounding referenced CUs.

10A is related to a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a fourth embodiment of the present invention.

When encoding / decoding a CU 1020 located at an LCU boundary, a CU 1012 having the largest block size among CUs 1010, 1011, 1012, and La, Lb, Lc adjacent to the left boundary of the corresponding CU ; Tc). At this time, if there is one or more CUs having the largest block size among the CUs that can be referenced in the left border, the CU positioned adjacent to the top of the blocks is selected as a reference block.

The encoder predicts the quantization parameter value of the CU 1020 to be encoded as the quantization parameter value of the CU 1012 (Tc) selected as a reference block at the left boundary, and encodes the difference value.

The decoder decodes the inverse quantization parameter value by adding the inverse quantization parameter value of the CU 1012 (Tc) selected as a reference block at the left boundary after decoding the inverse quantization parameter difference value of the CU 1020 to be decoded.

10B is another example of a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to the fourth embodiment of the present invention.

When the LCU is divided into a plurality of CUs and is encoded / decoded, the largest block size among CUs 1030, 1031, 1032; La, Lb, and Lc located adjacent to the left boundary of the CU 1050 to be encoded / decoded See CU 1030; La. At this time, if there is one or more CUs having the largest block size on the left border, the CU 1030, La located adjacent to the top of the blocks is selected as a reference block.

The encoder predicts the quantization parameter value of the CU 1050 to be encoded as the quantization parameter value of the CU 1030 (La) selected as a reference block at the left boundary and encodes the difference value.

The decoder decodes the inverse quantization parameter value by decoding the inverse quantization parameter difference value of the CU 1050 to be decoded and adding the inverse quantization parameter value of the CU 1030 (La) selected as a reference block at the left boundary.

11A is related to a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a fifth embodiment of the present invention.

When encoding / decoding the CU 1120 located at the LCU boundary, all of the referenced CUs 1110, 1111, 1112; La, Lb, and Lc adjacent to the left boundary of the CU are used as reference blocks.

The encoder predicts the quantization parameter value of the CU 1120 to be encoded using the average, maximum, and minimum values of the quantization parameter values of the referenced CUs 1110, 1111, 1112; La, Lb, and Lc adjacent to the left boundary. After that, only the difference value is encoded.

The decoder decodes the dequantization parameter difference value of the CU 1120 to be decoded, and then averages and maximizes the inverse quantization parameter values of referenced CUs 1110, 1111, 1112 and La, Lb, Lc adjacent to the left boundary. , Decode the inverse quantization parameter value of the CU 1120 to be decoded by adding the minimum value and the like.

11B is another example of a quantization parameter value prediction block determination unit 102 and an inverse quantization parameter value prediction block determination unit 123 according to a fifth embodiment of the present invention.

When the LCU is divided into multiple CUs and is encoded / decoded, all referenceable CUs located adjacent to the left boundary of the CU 1150 to be encoded / decoded are used as reference blocks.

The encoder predicts a quantization parameter value of the CU 1150 to be encoded as an average value, a minimum value, and a maximum value of quantization parameter values of all referenceable CUs located adjacent to a left boundary, and then encodes the difference value.

The decoder decodes the inverse quantization parameter difference value of the CU 1150 to be decoded, and then adds an inverse quantization parameter value by adding an average value, a minimum value, and a maximum value of quantization parameter values of all referenceable CUs located adjacent to the left boundary. Decrypt.

Claims (3)

Determining whether an encoding block has an inverse quantization difference value, and when it is determined that the encoding block has an inverse quantization difference value, a parameter related to whether the inverse quantization difference value is encoded is set to a first value;
Determining whether the coding block has an inverse quantization difference value, and when it is determined that the coding block has an inverse quantization difference value, a parameter related to whether the inverse quantization difference value is coded is set to a first value ;
Decoding flag information indicating whether the coding block is divided into a plurality of sub-blocks, the plurality of sub-blocks being four sub-blocks obtained by dividing the coding block using a quadtree structure;
Obtaining an inverse quantization difference value for the coding block, when the parameter has the first value and the coding block is divided into a plurality of sub-blocks, a first sub-block among the plurality of sub-blocks For, the inverse quantization difference value is decoded, and when the inverse quantization difference value is decoded for the first sub-block, the parameter is changed from the first value to the second value, and the parameter is changed to the second value. Thereafter, the inverse quantization difference value is not decoded for the remaining sub-blocks except the first sub-block;
Obtaining an inverse quantization parameter prediction value using a neighboring block of the coding block, the neighboring block is determined using position information of a neighboring block located around the coding block, and the neighboring block is a left side of the coding block A left peripheral block located at and an upper peripheral block located above the coded block; And
And obtaining an inverse quantization parameter using the inverse quantization difference value and the inverse quantization parameter prediction value,
The inverse quantization parameter prediction value,
An image decoding method characterized in that it is derived using an inverse quantization parameter of the left peripheral block and an inverse quantization parameter of the upper peripheral block. Determining whether to encode an inverse quantization difference value for an encoding target block, and when it is determined to encode an inverse quantization difference value for the encoding target block, a parameter related to whether an inverse quantization difference value is encoded is a first value. Set to;
Determining whether the encoding target block is divided into a plurality of sub-blocks, the plurality of sub-blocks are four sub-blocks obtained by dividing the encoding target block using a quadtree structure;
Obtaining an inverse quantization parameter prediction value using a neighboring block of the encoding target block, the neighboring block is determined using location information of a neighboring block located around the target block, and the neighboring block is the target of encoding A left peripheral block located on the left side of the block and an upper peripheral block located above the block to be coded;
Obtaining an inverse quantization difference value for the encoding target block using an inverse quantization parameter and the inverse quantization parameter prediction value; And
Encoding an inverse quantization difference value for the encoding target block, when the parameter has the first value, and the encoding target block is divided into a plurality of sub-blocks, a first sub of the plurality of sub-blocks When the inverse quantization difference value is coded for a block, and when the inverse quantization difference value is coded for the first sub-block, the parameter is changed from the first value to a second value, and the parameter is the second value. After changing to, the inverse quantization difference value is not encoded for the remaining sub-blocks except the first sub-block,
Including,
The inverse quantization parameter prediction value,
An image encoding method characterized by being derived by using an inverse quantization parameter of the left peripheral block and an inverse quantization parameter of the upper peripheral block. A computer readable recording medium that stores a bitstream received by a video decoding apparatus and decoded to be used to reconstruct an image, wherein the bitstream includes flag information about splitting of an encoding block and information about an inverse quantization difference value. and,
The flag information indicates whether the coding block is divided into a plurality of sub-blocks, and the plurality of sub-blocks are four sub-blocks obtained by dividing the coding block using a quadtree structure.
When it is determined that the encoding block has an inverse quantization difference value, a parameter related to whether or not the inverse quantization difference value is encoded is set to a first value, the parameter has the first value, and the encoding block has a plurality of values. When divided into sub-blocks, information about the inverse quantization difference value is used to decode the inverse quantization difference value for the first sub-block among the plurality of sub-blocks, and the information for the first sub-block is When an inverse quantization difference value is decoded, the parameter is changed from the first value to the second value, and after the parameter is changed to the second value, the inverse quantization difference is applied to the remaining subblocks except the first subblock. The value is not decrypted,
The inverse quantization difference value and the inverse quantization parameter prediction value are used to obtain an inverse quantization parameter, the inverse quantization parameter prediction value is obtained using a neighboring block of the coding block, and the neighboring block is located around the coding block. It is determined using location information of a neighboring block, and the neighboring block includes a left neighboring block located on the left side of the coding block and an upper neighboring block located above the coding block,
The inverse quantization parameter prediction value,
A computer-readable recording medium storing a bitstream, characterized in that it is derived using an inverse quantization parameter of the left peripheral block and an inverse quantization parameter of the upper peripheral block.

Images