KR101519557B1 - Apparatus and method for fast Intra Prediction Algorithm - Google Patents

Abstract

The present invention relates to a method for fast intra prediction coding. This invention divides video blocks for intra prediction coding into the coding unit and the prediction unit. In deciding on an intra prediction coding mode applicable to the prediction unit, one uses coding unit to be split, size information on each coding unit and upper-level coding unit, information on the intra prediction coding mode, In deciding on an intra prediction coding mode applicable to the prediction unit, one can skip additional calculation steps in ways to split the coding units in the initial stage or terminate the split procedure by using information such as the coding unit to be split, the size information on each coding unit and the upper-level coding unit, the information on the intra prediction coding mode, and the rate-distortion cost information calculated based on the sum of absolute transformed difference.

Description

[0001] The present invention relates to a fast intra picture prediction coding method and a fast intra prediction coding method,

The present invention relates to a high-speed intra-picture prediction coding method.

HEVC (High Efficiency Video Coding) is a proposed standard technology for encoding and decoding super high resolution and super high picture with high compression ratio. It is composed of MPEG under ISO / IEC and VCEG under ITU-T It is an international standard technology that has been standardized by JCT-VC. HEVC has improved the coding methods applied in standard technologies such as AVC / H.264 in order to achieve an improved compression rate in image coding, and adopted more advanced coding methods in the intra prediction coding method as standard .

The intra-picture prediction coding technique is a coding process using information in a screen that is known in advance in image compression. In other words, the intra-frame prediction coding technique predicts and generates a reference image block which is similar to the original image block to be currently encoded using the information of the reconstructed image blocks through decoding, The difference value is encoded so that the information amount of the difference value to be encoded can be reduced. HEVC improves the existing AVC / H.264 support for 9 directional modes in intra prediction coding, and supports 35 modes in total.

In general, the rate distortion cost is used to evaluate the performance of the image coding, considering both the bit rate and the distortion. That is, considering the bit rate indicating how much the compression rate is achieved when the encoded bit stream is compared with the original image before encoding and the degree of distortion when the encoded and decoded image is compared with the original image, , And evaluates the performance of image encoding. In the image encoding process, Rate-Distortion Optimization (RDO) is performed to minimize the rate-distortion cost to determine various encoding methods and detailed modes and parameters of each encoding method.

In order to improve the image compression performance, HEVC introduces the concept of a coding unit (Coding Unit) having various sizes from at least 64 × 64 to at least 8 × 8 pixels. The HEVC divides the image to be compressed into units of the most appropriate size according to the characteristics, and determines a coding method to be applied to each coding unit, that is, an inter picture coding or a intra picture coding method. Then, the divided coding unit is divided into prediction units (Prediction Unit), and the detailed operation modes of the coding method are determined for each prediction unit so that optimized coding is performed.

When the intra-picture prediction coding is performed, the HEVC performs rate-distortion optimization to divide the compression target image block into coding units in the most appropriate size, divides each divided coding unit into prediction units, And determines the mode of the optimal intra prediction encoding. However, in order to determine the coding unit division structure and the intra prediction coding mode having the optimal rate-distortion cost as described above, the HEVC divides the compression target image block into a quad tree structure, Distortion cost for each of the intra-picture prediction mode, the intra-picture prediction mode, and the intra-picture prediction mode, compares the rate-distortion cost with each other, and performs an encoding unit and a prediction unit division so as to have a best rate- do. Therefore, the existing HEVC coding unit, prediction unit division structure and intra prediction coding mode determination method have a problem in that they require a lot of time and resources in the coding step due to the complexity of the calculation.

An object of the present invention is to provide a method and an apparatus for dividing an image block to be subjected to intraframe prediction encoding in intra-frame prediction coding of an HEVC into an encoding unit (Coding Unit) and a prediction unit (Prediction Unit) In determining the coding mode, the size information, the intra prediction coding mode information, and the rate-distortion calculated on the basis of SATD (Sum of Absolute Transformed Difference) of each of the coding unit to be divided and the adjacent coding units and upper coding units By omitting the additional calculation process in such a manner that the encoding unit is divided early or the segmentation procedure is terminated early under certain conditions using the cost information, the number of encoding units and prediction unit divisions in all cases and the intra prediction encoding mode The rate-distortion cost is calculated and the rate-distortion optimization is performed, The existing HEVC intra prediction coding method for determining the division structure and the intra prediction coding mode is improved to determine the coding unit and prediction unit division structure so as to have a high rate-distortion performance while effectively reducing the complexity of calculation, Speed intra-picture prediction coding method for determining a coding mode and an apparatus therefor.

According to an aspect of the present invention, there is provided an intra-picture prediction coding method, including: a division target encoding unit belonging to a coding tree unit and having a size smaller than or equal to the size of the encoding tree unit; (Coding Unit); Whether or not to divide the division target coding unit into sub coding units by using the size information of the division target coding unit, the size of the coding units adjacent to the coding tree unit, and the intra prediction coding mode information A first segmentation decision step of early decision; And a coding mode candidate set determining step of determining a predetermined number of intra prediction coding mode candidates for the division target coding unit.

In one embodiment, the determination of the encoding mode candidate set may include generating a reference image block by intra-picture prediction on the intra-picture prediction mode for each intra-picture prediction encoding mode, Distortion cost for the division target encoding unit using the information of the target encoding unit, and performs a predetermined number of intra-picture prediction coding based on the calculated first rate-distortion cost And determines a candidate set of the mode.

In one embodiment, an intra-picture prediction coding method is applied to an intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step, Determining an optimal encoding mode based on the calculated second rate-distortion cost using the information of the target encoding unit, calculating a second rate-distortion cost for the division subject encoding unit, and determining an optimal encoding mode based on the calculated second rate- And further comprising

In one embodiment, the first division determination step may include: dividing the division target encoding unit into the lower encoding units when the division objective encoding unit is determined to divide the division target encoding units; The intra-picture prediction coding is newly set for each of the divided lower coding units in the division target coding unit setting step in the division target coding unit setting step, and when it is decided not to divide the division target coding unit, And proceeds to the encoding mode candidate set determination step.

In one embodiment, the first division determination step determines whether to divide the division object encoding unit into the lower encoding units by the following equation (1).

Equation 1:

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할할지 여부를 결정하는 플래그이고,

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 상측에 바로 인접한 좌단의 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 상측에 바로 인접한 좌단의 부호화 유닛의 화면 내 예측 부호화모드의 값이고,

는 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 화면 내 예측 부호화모드의 값이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is a flag for determining whether to divide the division target encoding unit into the lower encoding units,

Is the length of one side of the division target encoding unit,

Is the length of one side of the leftmost encoding unit immediately adjacent to the upper side of the encoding tree unit,

Is the length of one side of the uppermost encoding unit immediately adjacent to the left side of the encoding tree unit,

Is the value of the intra-picture prediction coding mode of the leftmost coding unit immediately adjacent to the upper side of the coding tree unit,

Is the value of the intra-picture prediction encoding mode of the uppermost encoding unit immediately adjacent to the left of the encoding tree unit,

and

Is a constant of a predetermined size.

In one embodiment, the first rate-distortion cost may be calculated according to Equation 2 and Equation 3 based on Sum of Absolute Transformed Difference (SATD).

Equation 2:

Equation 3:

는 상기 제1 율-왜곡 비용이고,

는 최적화에 사용되는 라그랑지안 상수이고,

는 화면 내 예측 부호화모드 정보를 부호화 하는데 요구되는 비트량이고,

는 상기 제1 율-왜곡 비용을 계산하는 대상 부호화 유닛과 상기 대상 부호화 유닛에 대하여 특정한 화면 내 예측 부호화모드를 적용하여 화면 내 예측을 수행하여 생성된 참조 영상블록 간의 차분값을 하다마드 변환(Hadamard Transform)을 한 것이고,

와

는 상기 하다마드 변환된 결과인 2차원 행열의 인덱스이다.)(here

Is the first rate-distortion cost,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to encode the intra prediction encoding mode information,

Distortion value is calculated by performing a Hadamard transformation on a difference value between a target encoding unit for calculating the first rate-distortion cost and a reference image block generated by performing intra-frame prediction using a specific intra-frame prediction encoding mode for the target encoding unit Transform)

Wow

Is the index of a two-dimensional matrix that is the result of the Hadamard transform.)

According to another aspect of the present invention, there is provided a method of intra-picture prediction coding according to the present invention, the intra-picture prediction coding method comprising: a division target encoding unit that belongs to a coding tree unit, A division target encoding unit setting step of setting a unit (Coding Unit); An encoding mode candidate set determining step of determining a predetermined number of intra prediction encoding mode candidates for the division target encoding unit; A first rate-distortion cost calculated by using the size of the division target encoding unit, information of a reference image block generated by in-frame prediction of the division target encoding unit and information of the division target encoding unit, Distortion cost calculated for the encoding units adjacent to the encoding unit and the first rate-distortion cost calculated for the upper encoding unit before the division-target encoding unit is divided, A division end determination step of early determining whether or not to terminate a procedure of dividing a target coding unit into a sub coding unit (Sub Coding Unit); And an optimal encoding mode determining step of determining an optimal intra prediction encoding mode for the partitioning target encoding unit among the candidate sets determined in the encoding mode candidate set determination step.

In one embodiment, the coding mode candidate set determination step may include a step of generating a reference picture block by performing in-picture prediction on the intra-picture prediction coding mode for the intra-picture prediction coding mode for the partitioning object coding unit, Distortion cost for the division target encoding unit using the information of the encoding unit, and calculates a first rate-distortion cost for the predetermined number of intra-picture prediction encoding modes based on the calculated first rate- Wherein the optimal encoding mode determination step determines an optimal encoding mode by using an intra prediction encoding mode belonging to the candidate set defined in the encoding mode candidate set determination step, Predictive encoding, re-decoding, and using the information of the reconstructed image block and the division target encoding unit Can be characterized by determining an optimum encoding mode based on the distortion cost-than the second division ratio to the target coding unit - calculating a distortion cost, and the calculated second rate.

In one embodiment, after the dividing object encoding unit setting step and before the encoding mode candidate set determining step, the size information of the division object encoding unit, the size of the encoding units adjacent to the encoding tree unit, And a first division determination step of determining whether or not to divide the division target coding unit into the lower coding units by using the mode information.

In one embodiment, the first division determination step may include: dividing the division target encoding unit into the lower encoding units when the division objective encoding unit is determined to divide the division target encoding units; The intra-picture prediction coding is newly set for each of the divided lower coding units in the division target coding unit setting step in the division target coding unit setting step, and when it is decided not to divide the division target coding unit, And proceeds to the encoding mode candidate set determination step.

In one embodiment, the first division determination step determines whether or not to divide the division object encoding unit into the lower encoding units by the following expression (4).

Equation 4:

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할할지 여부를 결정하는 플래그이고,

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 상측에 바로 인접한 좌단의 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 상측에 바로 인접한 좌단의 부호화 유닛의 화면 내 예측 부호화모드의 값이고,

는 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 화면 내 예측 부호화모드의 값이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is a flag for determining whether to divide the division target encoding unit into the lower encoding units,

Is the length of one side of the division target encoding unit,

Is the length of one side of the leftmost encoding unit immediately adjacent to the upper side of the encoding tree unit,

Is the length of one side of the uppermost encoding unit immediately adjacent to the left side of the encoding tree unit,

Is the value of the intra-picture prediction coding mode of the leftmost coding unit immediately adjacent to the upper side of the coding tree unit,

Is the value of the intra-picture prediction encoding mode of the uppermost encoding unit immediately adjacent to the left of the encoding tree unit,

and

Is a constant of a predetermined size.

In one embodiment, the first rate-distortion cost may be calculated according to Equations 5 and 6 based on a Sum of Absolute Transformed Difference (SATD).

Equation 5:

Equation 6:

는 상기 제1 율-왜곡 비용이고,

는 최적화에 사용되는 라그랑지안 상수이고,

는 화면 내 예측 부호화모드 정보를 부호화 하는데 요구되는 비트량이고,

는 상기 제1 율-왜곡 비용을 계산하는 대상 부호화 유닛과 상기 대상 부호화 유닛에 대하여 특정한 화면 내 예측 부호화모드를 적용하여 화면 내 예측을 수행하여 생성된 참조 영상블록 간의 차분값을 하다마드 변환(Hadamard Transform)을 한 것이고,

와

는 상기 하다마드 변환된 결과인 2차원 행열의 인덱스이다.)(here

Is the first rate-distortion cost,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to encode the intra prediction encoding mode information,

Distortion value is calculated by performing a Hadamard transform on a difference value between a target encoding unit for calculating the first rate-distortion cost and a reference image block generated by performing intra-frame prediction using a specific intra-frame prediction encoding mode for the target encoding unit Transform)

Wow

Is the index of a two-dimensional matrix that is the result of the Hadamard transform.)

In one embodiment, the dividing end determination step determines whether to end the procedure for dividing the division target encoding unit into the lower encoding units by the following equation (7).

Equation 7:

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 미리 정해진 크기의 상수이고,

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할하는 절차를 종료할지 여부를 나타내는 플래그이고,

는 상기 분할 대상 부호화 유닛에 대하여 상기 부호화모드 후보 집합 결정 단계에서 정해진 상기 후보 집합에 속한 화면 내 예측 부호화모드를 적용했을 때 가장 작은 값의 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 좌측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 상측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛이 분할되기 바로 이전의 상위 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is the length of one side of the division target encoding unit,

Is a constant of a predetermined size,

Is a flag indicating whether to end the procedure of dividing the division target encoding unit into the lower encoding units,

Is the first rate-distortion cost of the smallest value when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step is applied to the division target coding unit,

Is the first rate-distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode which is optimally determined for the coding unit adjacent to the upper side of the division target coding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode that is optimally determined for the immediately preceding upper-coding unit immediately before the division-target coding unit is divided,

and

Is a constant of a predetermined size.

In one embodiment, the dividing end determination step determines that the division target coding unit is no longer divided into the lower coding unit when it is determined to end the procedure of dividing the division target coding unit, And proceeds to the optimal encoding mode determination step.

In one embodiment, when determining to end the procedure of dividing the partitioning object encoding unit in the partitioning end determination step, the size of the partitioning object encoding unit and the first rate calculated for the partitioning object encoding unit Distortion cost, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate-distortion cost calculated for the upper encoding unit before the division subject encoding unit is divided And a second division determination step of determining whether or not to divide the division target coding unit into the lower coding units by using the second division decision step.

In one embodiment, the second division determination step may determine whether to divide the division object encoding unit by the following equation (8).

Equation 8:

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 미리 정해진 크기의 상수이고,

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할할지 여부를 나타내는 플래그이고,

는 상기 분할 대상 부호화 유닛에 대하여 상기 부호화모드 후보 집합 결정 단계에서 정해진 상기 후보 집합에 속한 화면 내 예측 부호화모드를 적용했을 때 가장 작은 값의 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 좌측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 상측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛이 분할되기 바로 이전의 상위 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is the length of one side of the division target encoding unit,

Is a constant of a predetermined size,

Is a flag indicating whether to divide the division target encoding unit into the lower encoding units,

Is the first rate-distortion cost of the smallest value when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step is applied to the division target coding unit,

Is the first rate-distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode which is optimally determined for the coding unit adjacent to the upper side of the division target coding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode that is optimally determined for the immediately preceding upper-coding unit immediately before the division-target coding unit is divided,

and

Is a constant of a predetermined size.

In one embodiment, the second division determination step may include: dividing the division target encoding unit into the lower encoding units when the division target encoding unit is determined to be divided, and dividing the divided lower encoding units into the lower encoding units The intra-picture prediction coding is newly set for each of the divided lower coding units in the division target coding unit setting step in the division target coding unit setting step, and when it is decided not to divide the division target coding unit, And proceeds to the optimal encoding mode determination step.

In one embodiment, the second rate-distortion cost may be calculated by the following equation (9).

Equation 9:

는 상기 제2 율-왜곡 비용이고,

는 상기 제2 율-왜곡 비용 계산 대상 부호화 유닛과 상기 제2 율-왜곡 비용 계산 대상 부호화 유닛을 화면 내 예측 부호화하고 다시 복호화 하여 복원한 영상블록 간의 차분값에서의 차이의 정도를 나타내는 지표로써, 상기 차분값의 절대값의 합인 SAD(Sum of Absolute Difference) 또는 상기 차분값의 제곱값의 합인 SSD(Sum of Square Difference)를 사용할 수 있고,

는 최적화에 사용되는 라그랑지안 상수이고,

는 화면 내 예측 부호화모드를 적용하여 화면 내 예측 부호화를 하는데 요구되는 비트량이다.)(here

Is the second rate-distortion cost,

Distortion cost calculation target encoding unit and the second rate-distortion cost calculation target encoding unit, and decodes the second rate-distortion cost calculation target encoding unit and the second rate-distortion cost calculation target encoding unit, A Sum of Absolute Difference (SAD), which is the sum of the absolute values of the difference values, or a Sum of Square Difference (SSD), which is the sum of the squares of the difference values,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to perform intra prediction encoding by applying the intra prediction encoding mode.)

In one embodiment, after determining the optimal coding mode, a third division determination step of determining whether to end the division procedure of the division target coding unit and whether to divide the division target coding unit into the lower coding unit is performed And further comprising

In one embodiment, if the division procedure of the division target coding unit belonging to the coding tree unit is completed after the third division determination step, the division structure of the divided coding tree unit and the prediction unit (Prediction Unit) And a prediction unit determining step of determining a final intra prediction encoding mode to be applied to each of the prediction units.

In one embodiment, when the depth value of the quadtree formed by dividing the coding tree unit into the division subject coding units is equal to a predetermined maximum depth, the third division determination step determines that the division target coding unit And if the depth value is smaller than the maximum depth, dividing the division target encoding unit into the lower encoding units, and dividing the divided lower encoding units into the division target encoding unit setting step And the intra-picture prediction coding is newly performed for each of the divided sub-coding units.

In one embodiment, in the case where it is determined that the encoding tree unit is to end the division procedure in all of the encoding units generated by dividing the encoding tree unit in the third division decision step, Determining a division structure of the coding tree unit by using the second rate-distortion cost for the units, determining the divided coding unit as the prediction unit, and determining a final Picture predictive encoding mode is determined.

According to another aspect of the present invention, there is provided an intra-picture prediction encoding apparatus including a Coding Tree Unit and dividing the Coding Tree Unit into a Coding Unit, An intra picture prediction unit for determining an intra picture prediction coding mode of the coding unit and generating a reference picture block according to the determined intra prediction coding mode; And a difference block generator for generating a difference block between the reference image block generated by the intra-picture prediction unit and the divided encoding unit, wherein the intra-picture prediction unit belongs to the encoding tree unit, Sets a coding object unit (Coding Unit) having a size equal to or smaller than that of the coding tree unit; Whether or not to divide the division target coding unit into sub coding units by using the size information of the division target coding unit, the size of the coding units adjacent to the coding tree unit, and the intra prediction coding mode information The first decision is made; Calculates a first Rate-Distortion Cost for the division target encoding unit by using the information of the reference image block generated by performing in-frame prediction on the division target encoding unit and the information of the division target encoding unit Determining a predetermined number of intra prediction encoding mode candidate sets based on the calculated first rate-distortion cost; An intra-picture prediction coding mode belonging to the determined intra-picture prediction coding mode is applied to the intra-picture prediction coding mode of the partitioning object coding unit, the intra-picture prediction coding is performed again using the information of the reconstructed image block and the partitioning object coding unit, Calculating a second rate-distortion cost for the division target encoding unit, determining an optimal intra prediction encoding mode on the basis of the calculated rate-distortion cost; Decides to divide the division target coding unit into the lower coding units if the depth value of the quadtree formed by dividing the coding tree unit into the division target coding units is smaller than a predetermined maximum depth, And decide to end the procedure of dividing the division target encoding unit into the lower encoding units if the maximum depth is the same.

In one embodiment, the intra-picture prediction unit predicts the size of the partitioning object encoding unit and the size of the partitioning object encoding unit before determining the optimal intra-picture prediction encoding mode after determining the prediction set of the predictive encoding mode, Distortion cost calculated for the coding units adjacent to the division target coding unit and the first rate-distortion cost calculated for the coding units adjacent to the division target coding unit and the first rate-distortion cost calculated for the upper coding unit before the division target coding unit is divided Determining whether to end the procedure of dividing the segmentation subject encoding unit into the lower segmentation unit using the first rate-distortion cost; The first rate-distortion cost calculated for the division target encoding unit, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate- Further comprising a step of determining as a second order whether or not to divide the division subject coding unit into the lower coding units by using the first rate-distortion cost calculated for the upper coding unit before the coding unit is divided .

In one embodiment, the intra-picture prediction unit may divide the division target coding unit into the lower coding units if it is determined to divide the division target coding unit into the lower coding units, Units are set as the division target encoding units, the operation of the intra-picture prediction unit is progressed with respect to each of the divided sub-encoding units, and if the encoding tree unit is divided into all of the generated encoding units Determines the division structure of the coding tree unit using the second rate-distortion cost for the divisionally generated coding units, and when the divided coding unit is determined as the prediction unit And determines a finalization to be applied to each of the prediction units Plane predictive encoding mode is determined.

In one embodiment, the intra-picture prediction unit may be configured such that the first rate-distortion cost is based on a Sum of Absolute Transformed Difference (SATD) using a Hadamard Transform, And the second rate-distortion cost is calculated based on a Sum of Absolute Difference (SAD) or a Sum of Square Difference (SSD), and using the reconstructed image block.

According to the present invention, in the intra-picture prediction coding method of the HEVC, the rate-distortion cost is calculated for all the number of coding units, the prediction unit division and the intra-picture prediction coding mode, and the rate- And the existing HEVC intra prediction coding method for determining the prediction unit division structure and the intra prediction coding mode are determined to determine the coding unit and prediction unit division structure so as to have a high rate-distortion performance while effectively reducing the complexity of calculation By determining the intra-picture prediction coding mode, intra-picture prediction coding can be performed at high speed.

1 is a flowchart illustrating an operation of an intra prediction encoding method according to an embodiment of the present invention.
2 is a flowchart illustrating an operation of a first partition determination step of an intra prediction encoding method according to an embodiment of the present invention.
FIG. 3 is a flowchart illustrating an operation of the intra-picture prediction coding method according to an exemplary embodiment of the present invention.
FIG. 4 is a flowchart illustrating an operation of a second division determination step of the intra prediction encoding method according to an embodiment of the present invention.
5 is a block diagram illustrating an operation of an image encoding apparatus including an intra prediction encoding apparatus according to an embodiment of the present invention.
6 is a reference diagram for explaining a process of dividing an encoding tree unit according to an intra prediction encoding method according to an embodiment of the present invention.
FIG. 7 is a reference diagram illustrating an encoded tree unit according to an intra-frame prediction encoding method according to an embodiment of the present invention. Referring to FIG.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are used to designate the same or similar components throughout the drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. In addition, the preferred embodiments of the present invention will be described below, but it is needless to say that the technical idea of the present invention is not limited thereto and can be variously modified by those skilled in the art.

The intra-picture prediction coding method and apparatus according to an embodiment of the present invention will be described in detail below. In particular, the intra-picture prediction coding method and apparatus according to an embodiment of the present invention will be described in detail below. And can be applied to a standard technology HEVC (High Efficiency Video Coding).

1 is a flowchart illustrating an operation of an intra prediction encoding method according to an embodiment of the present invention.

The intra-picture prediction coding method according to an exemplary embodiment of the present invention includes a setting of a division target encoding unit S110, a first division decision step S120, an encoding mode candidate set determination step S130, a division end decision step S140, A second segmentation decision step S150, an optimal encoding mode decision step S160, a third segmentation decision step S170 and a prediction unit decision step S180.

The intra-picture prediction coding method according to still another embodiment of the present invention may further include a first division decision step (S120), a division end determination step (S140), and a second division decision step The second division decision step S150, the optimal encoding mode decision step S160, the third division decision step S170 and the prediction unit decision step S180 may be selectively included or omitted. For example, another embodiment of the present invention is a method of encoding a moving picture, comprising the steps of setting a segmentation object encoding unit (S110), a first segmentation decision step (S120), an encoding mode candidate set determination step (S130), an optimal encoding mode determination step The encoding mode candidate set determination step S130, and the division end determination step S170. The decoding unit may be configured to include a decoding target encoding unit setting step S110, an encoding mode candidate set determining step S130, S140, an optimal encoding mode determination step S160, a third segmentation decision step S170, and a prediction unit determination step S180, or may be configured to include a segmentation target encoding unit setting step S110, A candidate division determining step S130, a division end determination step S140, a second division decision step S150, an optimal encoding mode decision step S160, a third segmentation decision step S170, a prediction unit decision step S180, As shown in FIG.

Hereinafter, each step of the intra-picture prediction coding method according to the best mode will be described in more detail.

The segmentation target encoding unit setting step S110 is a step of receiving a Coding Tree Unit (CTU) image block to be subjected to intraframe prediction coding and dividing the inputted coded tree unit into quad tree structures , A segmentation target encoding unit (Coding Unit, CU) to be segmented is set. Herein, the coding tree unit may be referred to as another largest coding unit (Largest Coding Unit, LCU), and may have a maximum size of 64x64. The encoding unit (CU) is a unit which is a unit of image encoding. The encoding tree unit is divided, the size is determined, the size is at least 8x8, and a method of image encoding is determined for each encoding unit.

Here, the division target encoding unit setting step (S110) first sets the encoding tree unit as the division target encoding unit in order to repeatedly divide the encoding tree unit into a quad tree form. In the first division decision step (S120), the second division decision step (S150) or the third division decision step (S170) described below, the division object coding unit is divided into a sub coding unit If it is determined that the sub-encoding unit has been divided, the sub-encoding unit may be input to the sub-encoding unit. As the divided lower coding unit is again set as the division target coding unit, the coding tree unit can be repeatedly divided and divided into coding units of a quad tree structure having a certain depth.

The first division decision step (S120) uses the size information of the division object encoding unit, the size of the encoding units adjacent to the encoding tree unit, and the intra-picture prediction mode information to set the division object encoding unit as a lower encoding unit Decide early on whether to split. Here, the intra prediction coding mode means a mode value for determining the direction and the mode of intra prediction in intra prediction coding.

Here, the first division decision step (S120) decides to divide the division target encoding unit early without actually calculating the rate-distortion cost for the division target encoding unit, thereby reducing the calculation complexity of intra picture coding. The operation of the first segmentation decision step S120 will be described in more detail with reference to FIG.

Here, the first division decision step (S120) may be optionally included or omitted, and if it is omitted, the encoding mode candidate set decision step (S130) may be performed next.

The encoding mode candidate set determination step (S130) generates a reference image block by intra-picture prediction for each intra-picture prediction encoding mode with respect to the division target encoding unit, Distortion cost for the division target encoding unit and calculates a predetermined number of intra prediction encoding mode candidate sets on the basis of the calculated first rate- .

Here, it is preferable that the encoding mode candidate set determination step (S130) determines a certain number of intra prediction encoding modes as the elements of the candidate set of the encoding mode in order of the first rate-distortion cost.

It is preferable that the first rate-distortion cost is calculated according to Equations (1) and (2) based on a Sum of Absolute Transformed Difference (SATD). The rate-distortion cost originally used for rate-distortion optimization is calculated by using difference values between the reconstructed image block and the original image block before encoding by encoding and re-decoding the original image block, but the first rate- Calculates the rate-distortion cost using the in-picture predicted reference image block rather than the reconstructed image block as below, thereby reducing the calculation complexity in the rate-distortion optimization.

는 상기 제1 율-왜곡 비용이고,

는 최적화에 사용되는 라그랑지안 상수이고,

는 화면 내 예측 부호화모드 정보를 부호화 하는데 요구되는 비트량이고,

는 상기 제1 율-왜곡 비용을 계산하는 대상 부호화 유닛과 상기 대상 부호화 유닛에 대하여 특정한 화면 내 예측 부호화모드를 적용하여 화면 내 예측을 수행하여 생성된 참조 영상블록 간의 차분값을 하다마드 변환(Hadamard Transform)을 한 것이고,

와

는 상기 하다마드 변환된 결과인 2차원 행열의 인덱스이다.)(here

Is the first rate-distortion cost,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to encode the intra prediction encoding mode information,

Distortion value is calculated by performing a Hadamard transformation on a difference value between a target encoding unit for calculating the first rate-distortion cost and a reference image block generated by performing intra-frame prediction using a specific intra-frame prediction encoding mode for the target encoding unit Transform)

Wow

Is the index of a two-dimensional matrix that is the result of the Hadamard transform.)

The dividing end determination step (S140) may include determining a size of the division target encoding unit, a first rate that is calculated using the information of the reference image block generated by in-frame prediction of the division target encoding unit and the information of the division target encoding unit Distortion cost, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate-distortion cost calculated for the upper encoding unit before the division subject encoding unit is divided, It is determined early whether or not to terminate the procedure for dividing the division object encoding unit into sub coding units by using the distortion cost.

Here, the division end determination step (S140) reduces the calculation complexity in intra picture coding by omitting the procedure of further dividing the division target encoding unit. The operation of the division end determination step S140 will be described in more detail with reference to FIG.

Here, if the second division decision step (S150) is omitted, the second division decision step (S150) can be performed. The encoding mode determination step S160 may be performed.

The second division decision step (S150) includes a step of determining the size of the division target encoding unit, the first rate-distortion cost computed for the division subject encoding unit, and the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit Distortion cost and the first rate-distortion cost calculated for an upper-level encoding unit before the division-target encoding unit is divided, whether to divide the division-target encoding unit into the lower encoding units Decide early.

Here, the second division decision step (S150) decides to divide the division target encoding unit early in the division objective encoding unit without actually calculating the second rate-distortion cost to be described below, Thereby reducing computational complexity. The operation of the second division decision step S150 will be described in more detail with reference to FIG.

Here, the second segmentation decision step S150 may be selectively included or omitted, and if it is determined that the segmentation end determination step S140 determines not to end the segmentation target encoding unit, If it is omitted, then the optimal encoding mode determination step (S160) may be performed.

The optimal encoding mode determination step S160 is a step of performing intra prediction encoding using the intra prediction encoding mode belonging to the candidate set defined in the encoding mode candidate set determination step S130 and re- Distortion cost for the division target encoding unit for each encoding mode using information of the block and the division target encoding unit, and calculates a second rate-distortion cost for the division target encoding unit based on the calculated second rate- And determines an optimal intra-picture prediction encoding mode for the unit.

Here, it is preferable that the optimal encoding mode determination step (S160) determines the intra prediction encoding mode having the smallest value of the second rate-distortion cost among the candidate sets as the optimal intra prediction encoding mode.

Here, the second rate-distortion cost is preferably calculated by the following equation (3).

는 상기 제2 율-왜곡 비용이고,

는 상기 제2 율-왜곡 비용 계산 대상 부호화 유닛과 상기 제2 율-왜곡 비용 계산 대상 부호화 유닛을 화면 내 예측 부호화하고 다시 복호화 하여 복원한 영상블록 간의 차분값에서의 차이의 정도를 나타내는 지표로써, 상기 차분값의 절대값의 합인 SAD(Sum of Absolute Difference) 또는 상기 차분값의 제곱값의 합인 SSD(Sum of Square Difference)를 사용할 수 있고,

는 최적화에 사용되는 라그랑지안 상수이고,

는 화면 내 예측 부호화모드를 적용하여 화면 내 예측 부호화를 하는데 요구되는 비트량이다.)(here

Is the second rate-distortion cost,

Distortion cost calculation target encoding unit and the second rate-distortion cost calculation target encoding unit, and decodes the second rate-distortion cost calculation target encoding unit and the second rate-distortion cost calculation target encoding unit, A Sum of Absolute Difference (SAD), which is the sum of the absolute values of the difference values, or a Sum of Square Difference (SSD), which is the sum of the squares of the difference values,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to perform intra prediction encoding by applying the intra prediction encoding mode.)

The third division determination step (S170) determines whether to end the division procedure of the division target encoding unit and whether to divide the division target encoding unit into the lower encoding units.

Here, the third segmentation decision step (S170) is a step of determining whether or not the value of the quadtree formed by dividing the encoding tree unit up to the division target encoding unit is equal to a predetermined maximum depth or ending the division in the division end determination step (S140) , It is determined to end the division procedure of the division target encoding unit.

For example, when the size of the coding tree unit is 64x64 and the depth is 0, the lower coding unit generated by dividing the coding tree unit by one time has a size of 32x32, a depth of 1, and a predetermined maximum depth of 2 , The partitioning procedure for the partitioning object coding unit is ended when the size of the partitioning object coding unit is 16x16 and the depth becomes 2.

The maximum depth may be a maximum depth of a quadtree formed when the encoding tree unit is divided by a size smaller than a minimum size of a predetermined encoding unit. This is because, in the prediction unit determination step (S180) to be described below, the prediction unit can be set to a size smaller than the minimum size of the encoding unit. Therefore, the segmentation target encoding unit can be divided into a smaller size than the minimum size of the predetermined encoding unit and can be determined as a prediction unit.

For example, when the size of the encoding tree unit is 64x64 and the minimum size of the predetermined encoding unit is 8x8, the prediction unit can be set to 4x4 size. In this case, the depth of the encoding tree unit of 64x64 size is 0 Since the depth value is incremented by 1 each time it is divided, the depth of the 4x4 block is 4, so that the maximum depth is 4.

Here, the third segmentation decision step (S170) may be such that when the depth value of the encoding tree unit is smaller than the maximum depth and it is decided not to end the segmentation in the segment end determination step (S140), the segmentation procedure of the segmentation objective encoding unit I decide to proceed further. At this time, the third division decision step (S170) divides the division target encoding unit into the lower encoding units and sets the divided lower encoding units as the division target encoding units in the division target encoding unit setting step (S110) , So that the intra-picture prediction coding can be progressed for each of the divided lower coding units.

Here, in the third segmentation decision step S170, if the segmentation end determination step S140 is omitted, the determination as to whether or not the segmentation end decision step S140 is decided to end the segmentation is omitted.

Here, the third segmentation decision step S170 may be optionally included or omitted as necessary.

In the prediction unit determination step (S180), when it is determined to end the division procedure in the third division decision step (S170) for all of the encoding units generated by dividing the encoding tree unit, the division structure of the encoding tree unit and the prediction And determines a unit (Prediction Unit). Here, the prediction unit is a unit in which the intra-picture prediction coding mode is determined, and one prediction unit performs intra-picture prediction coding in one intra-picture prediction coding mode.

Herein, the prediction unit determination step (S180) may further include the step of determining, using the second rate-distortion cost calculated corresponding to the optimal intra-picture prediction coding mode determined in the optimal coding mode determination step (S160) The structure in which the encoding tree unit is divided into the encoding unit and the prediction unit is determined. That is, the sum of each of the second rate-distortion costs of the four lower-layer encoding units, and the second rate-distortion cost of the upper-level encoding unit, from the encoding unit divided into the smallest size to the highest- And the second rate-distortion cost in the coding tree unit is minimized by selecting the one having the smaller rate-distortion cost as compared with the first rate-distortion cost.

Herein, the prediction unit determination step S180 determines each of the encoding units for which the division structure is determined as a prediction unit as described above, and determines an optimal encoding mode determination step (S160) for the encoding unit corresponding to each of the prediction units, Picture prediction mode to be applied to the prediction unit is determined to be the final intra-picture prediction coding mode to be applied to the prediction unit. At this time, the size of the prediction unit may be one step smaller than the coding unit of the predetermined minimum size. Therefore, an encoding unit having a predetermined minimum size can be further divided into one step and determined as a prediction unit.

Here, the prediction unit determination step S180 may be optionally included or omitted.

FIG. 2 is a flowchart illustrating an operation of a first subdivision determination step (S120) of an intra prediction encoding method according to an exemplary embodiment of the present invention.

이 미리 정해진 크기의 상수

보다 크거나 같은 경우는 단계 S122를 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할하지 않기로 결정한다. 여기서 상기

는 32로 하는 것이 바람직하다.In step S121, the length of one side of the division target encoding unit

Constants of this predetermined size

The process proceeds to step S122. Otherwise, it is determined not to divide the division target encoding unit. Here,

Is preferably set to 32.

가 상기 분할 대상 부호화 유닛의 한 변의 길이

의 반보다 작고 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 한 변의 길이

가 상기

의 반보다 작은 경우는 단계 S123을 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할하지 않기로 결정한다.In step S122, the length of one side of the leftmost encoding unit immediately adjacent to the upper side of the encoding tree unit to which the division target encoding unit belongs

Is the length of one side of the division target encoding unit

The length of one side of the upper encoding unit immediately adjacent to the left side of the encoding tree unit

Gt;

, The process proceeds to step S123. Otherwise, it is determined not to divide the division target encoding unit.

과 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 화면 내 예측 부호화모드의 값

의 차이값의 절대값이 미리 정해진 크기의 상수

보다 큰 경우는 단계 S124를 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할하지 않기로 결정한다. 여기서 상기

는 5로 하는 것이 바람직하다.In step S123, the value of the intra-picture prediction coding mode of the leftmost coding unit immediately adjacent to the upper side of the coding tree unit

And the value of the intra prediction encoding mode of the upper encoding unit immediately adjacent to the left side of the encoding tree unit

The absolute value of the difference value is a constant of a predetermined size

The process proceeds to step S124. Otherwise, it is determined not to divide the division target encoding unit. Here,

Is preferably 5.

In step S124, the division target encoding unit is divided into four sub-encoding units having the same size, and the divided sub-encoding units are input to the division target encoding unit setting step (S110) So that the intra-picture prediction coding is progressed for each of the divided lower coding units.

If it is determined that the division target encoding unit is not to be divided, the first division decision step (S120) is terminated and the process proceeds to the next step.

Here, the first division decision step (S120) may determine whether to divide the division target encoding unit by the following expression (4). The operation from the step S121 to the step S124 may be expressed by the following equation (4).

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할할지 여부를 결정하는 플래그이고,

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 상측에 바로 인접한 좌단의 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 한 변의 길이이고,

는 상기 부호화 트리 유닛의 상측에 바로 인접한 좌단의 부호화 유닛의 화면 내 예측 부호화모드의 값이고,

는 상기 부호화 트리 유닛의 좌측에 바로 인접한 상단의 부호화 유닛의 화면 내 예측 부호화모드의 값이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is a flag for determining whether to divide the division target encoding unit into the lower encoding units,

Is the length of one side of the division target encoding unit,

Is the length of one side of the leftmost encoding unit immediately adjacent to the upper side of the encoding tree unit,

Is the length of one side of the uppermost encoding unit immediately adjacent to the left side of the encoding tree unit,

Is the value of the intra-picture prediction coding mode of the leftmost coding unit immediately adjacent to the upper side of the coding tree unit,

Is the value of the intra-picture prediction encoding mode of the uppermost encoding unit immediately adjacent to the left of the encoding tree unit,

and

Is a constant of a predetermined size.

는 32로 하고, 상기

는 5로 하는 것이 바람직하다.Here,

Is set to 32,

Is preferably 5.

3 is a flowchart illustrating an operation of determining an end of division (S140) in the intraframe prediction method according to an exemplary embodiment of the present invention.

이 미리 정해진 크기의 상수

보다 크거나 같은 경우 단계 S142를 진행하고, 그 이외의 경우는 단계 S143을 진행한다. 여기서 상기

은 32로 하는 것이 바람직하다.In step S141, the length of one side of the division target encoding unit

Constants of this predetermined size

The process proceeds to step S142, and otherwise, the process proceeds to step S143. Here,

Is preferably set to 32.

을, 상기 분할 대상 부호화 유닛의 좌측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용인

와, 상기 분할 대상 부호화 유닛의 상측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용인

중 더 큰 값으로 나누고, 그 나눈 값이 미리 정해진 크기의 상수

보다 작은 경우는 단계 S144를 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할종료하지 않기로 결정한다.In step S142, when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step (S130) is applied to the division target coding unit, the first rate-distortion cost

Distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit

Distortion cost for the intra-picture prediction coding mode optimally determined for the coding unit adjacent to the upper side of the division target coding unit

, And the divided value is divided by a constant of a predetermined size

The process proceeds to step S144. Otherwise, it is determined that the division target encoding unit is not to be divided.

을, 상기 분할 대상 부호화 유닛이 분할되기 바로 이전의 상위 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용

으로 나눈 값이 미리 정해진 크기의 상수

보다 작은 경우는 단계 S144를 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할종료하지 않기로 결정한다. In step S143, when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step (S130) is applied to the division target coding unit, the first rate-distortion cost

Distortion cost for the intra-picture prediction coding mode optimally determined for the immediately preceding upper coding unit before the division target coding unit is divided,

Is a constant of a predetermined size

The process proceeds to step S144. Otherwise, it is determined that the division target encoding unit is not to be divided.

In step S144, the procedure for dividing the division target encoding unit is ended and it is determined that the division is not further divided into the lower encoding unit.

If it is determined that the segmentation target encoding unit is not to be segmented, the segmentation end determination step S140 is terminated and the second segmentation decision step S150 is performed. If the second segmentation decision step S150 is omitted It is determined that the division target coding unit is no longer divided into the lower coding unit if it is determined to end the division, and the optimal coding mode determination step (S160) is performed do.

Here, the division end determination step S140 may determine whether to end the procedure of dividing the division target encoding unit into the lower encoding units by the following equation (5). The operation from the step S141 to the step S144 may be expressed by the following equation (5).

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 미리 정해진 크기의 상수이고,

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할하는 절차를 종료할지 여부를 나타내는 플래그이고,

는 상기 분할 대상 부호화 유닛에 대하여 부호화모드 후보 집합 결정 단계(S130)에서 정해진 상기 후보 집합에 속한 화면 내 예측 부호화모드를 적용했을 때 가장 작은 값의 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 좌측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 상측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛이 분할되기 바로 이전의 상위 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is the length of one side of the division target encoding unit,

Is a constant of a predetermined size,

Is a flag indicating whether to end the procedure of dividing the division target encoding unit into the lower encoding units,

Is the first rate-distortion cost of the smallest value when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step (S130) is applied to the partitioning object coding unit,

Is the first rate-distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode which is optimally determined for the coding unit adjacent to the upper side of the division target coding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode that is optimally determined for the immediately preceding upper-coding unit immediately before the division-target coding unit is divided,

and

Is a constant of a predetermined size.

은 32로 하는 것이 바람직하다.Here,

Is preferably set to 32.

FIG. 4 is a flowchart illustrating an operation of a second segmentation determination step (S150) of the intra-picture prediction coding method according to an embodiment of the present invention.

이 미리 정해진 크기의 상수

보다 크거나 같은 경우 단계 S142를 진행하고, 그 이외의 경우는 단계 S143을 진행한다. 여기서 상기

은 32로 하는 것이 바람직하다.In step S151, the length of one side of the division target encoding unit

Constants of this predetermined size

The process proceeds to step S142, and otherwise, the process proceeds to step S143. Here,

Is preferably set to 32.

을, 상기 분할 대상 부호화 유닛의 좌측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용인

와, 상기 분할 대상 부호화 유닛의 상측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용인

중 더 큰 값으로 나누고, 그 나눈 값이 미리 정해진 크기의 상수

보다 큰 경우는 단계 S154를 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할하지 않기로 결정한다.In step S152, when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step (S130) is applied to the division target coding unit, the first rate-distortion cost

Distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit

Distortion cost for the intra-picture prediction coding mode optimally determined for the coding unit adjacent to the upper side of the division target coding unit

, And the divided value is divided by a constant of a predetermined size

The process proceeds to step S154. Otherwise, it is determined not to divide the division target encoding unit.

을, 상기 분할 대상 부호화 유닛이 분할되기 바로 이전의 상위 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용

으로 나눈 값이 미리 정해진 크기의 상수

보다 큰 경우는 단계 S154를 진행하고, 그 이외의 경우는 상기 분할 대상 부호화 유닛을 분할하지 않기로 결정한다. In step S153, when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step (S130) is applied to the division target coding unit, the first rate-distortion cost

Distortion cost for the intra-picture prediction coding mode optimally determined for the immediately preceding upper coding unit before the division target coding unit is divided,

Is a constant of a predetermined size

The process proceeds to step S154. Otherwise, it is determined not to divide the division target encoding unit.

In step S154, the division target encoding unit is divided into four sub-encoding units having the same size, and the divided sub-encoding units are input to the division target encoding unit setting step (S110) So that the intra-picture prediction coding is progressed for each of the divided lower coding units.

If it is determined that the division target encoding unit is not to be divided, the second division decision step (S150) is terminated and the process proceeds to the optimal encoding mode decision step (S160), which is the next step.

Here, the second division decision step (S150) may determine whether the division target encoding unit is divided by the following equation (6). The operation from the step S151 to the step S154 may be expressed as the following equation (6).

은 상기 분할 대상 부호화 유닛의 한 변의 길이이고,

는 미리 정해진 크기의 상수이고,

은 상기 분할 대상 부호화 유닛을 상기 하부 부호화 유닛으로 분할할지 여부를 나타내는 플래그이고,

는 상기 분할 대상 부호화 유닛에 대하여 부호화모드 후보 집합 결정 단계(S130)에서 정해진 상기 후보 집합에 속한 화면 내 예측 부호화모드를 적용했을 때 가장 작은 값의 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 좌측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛의 상측에 인접한 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

는 상기 분할 대상 부호화 유닛이 분할되기 바로 이전의 상위 부호화 유닛에 대하여 최적으로 결정된 화면 내 예측 부호화모드에 대한 상기 제1 율-왜곡 비용이고,

과

는 미리 정해진 크기의 상수이다.)(here

Is the length of one side of the division target encoding unit,

Is a constant of a predetermined size,

Is a flag indicating whether to divide the division target encoding unit into the lower encoding units,

Is the first rate-distortion cost of the smallest value when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step (S130) is applied to the partitioning object coding unit,

Is the first rate-distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode which is optimally determined for the coding unit adjacent to the upper side of the division target coding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode that is optimally determined for the immediately preceding upper-coding unit immediately before the division-target coding unit is divided,

and

Is a constant of a predetermined size.

는 32로 하는 것이 바람직하다.Here,

Is preferably set to 32.

5 is a block diagram illustrating an operation of an image encoding apparatus including an intra prediction encoding apparatus according to another embodiment of the present invention.

The intra-picture prediction coding apparatus may include an intra-picture prediction unit 100 and a difference block generation unit 200.

The intra picture prediction unit 100 receives a coding tree unit, divides the coding tree unit into a coding unit, determines an intra prediction coding mode of the divided coding unit, And generates a reference image block according to the encoding mode.

The difference block generating unit 200 generates the difference image between the reference image block generated by the intra prediction unit 100 and the divided encoding unit.

Here, the intra prediction unit 100 may perform intra prediction coding according to the intra prediction coding method according to an embodiment of the present invention described above with reference to FIG. 1 to FIG. Hereinafter, the same elements as those in the above description will be omitted.

The intra prediction unit 100 sets a coding object unit (Coding Unit) belonging to the coding tree unit and having a size equal to or smaller than the coding tree unit;

Whether or not to divide the division target coding unit into sub coding units by using the size information of the division target coding unit, the size of the coding units adjacent to the coding tree unit, and the intra prediction coding mode information The first decision is made;

In the intra-picture prediction encoding mode for each of the intra-picture prediction encoding modes, generates the reference picture block by performing intra-picture prediction on the intra-picture prediction encoding mode for the partitioning object encoding unit, Calculating a first rate-distortion cost for the division target encoding unit and determining a predetermined number of intra prediction encoding mode candidate sets based on the calculated first rate-distortion cost;

The first rate-distortion cost calculated for the division target encoding unit, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate- Determining whether to end the procedure of dividing the division target encoding unit into the lower encoding units by using the first rate-distortion cost calculated for the upper encoding unit before the encoding unit is divided;

The first rate-distortion cost calculated for the division target encoding unit, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate- Determining as a second order whether or not to divide the division subject coding unit into the lower coding units by using the first rate-distortion cost calculated for the upper coding unit before the coding unit is divided;

An intra-picture prediction coding mode belonging to the determined intra-picture prediction coding mode is applied to the intra-picture prediction coding mode of the partitioning object coding unit, the intra-picture prediction coding is performed again using the information of the reconstructed image block and the partitioning object coding unit, Calculating a second rate-distortion cost for the division target encoding unit, determining an optimal intra prediction encoding mode on the basis of the calculated rate-distortion cost;

If the depth value of the quadtree in which the coding tree unit is divided is smaller than a predetermined maximum depth, decides to divide the division target coding unit into the lower coding units, and if the depth value is equal to the maximum depth, It may decide to end the procedure of dividing the target encoding unit into the lower encoding units.

If the intra-picture prediction unit 100 determines to divide the division target coding unit into the lower coding units, the intra prediction unit 100 divides the division target coding unit into the lower coding units and divides the divided lower coding units And the intra-picture prediction unit 100 is set to the divided object coding unit so that the operation of the intra-picture prediction unit 100 is progressed for each of the divided lower coding units,

If it is determined that the division procedure is to be ended for all of the division target encoding units generated by dividing the encoding tree unit, the encoding rate of the encoding tree unit, using the second rate-distortion cost of the division target encoding units, Determines a structure to be divided into the coding units, sets a prediction unit (Prediction Unit) according to each coding unit for which the division is determined, and determines the final intra-picture prediction coding mode to be applied to each of the prediction units .

Here, the intra prediction unit 100 may calculate the first rate-distortion cost based on a Sum of Absolute Transformed Difference (SATD) using a Hadamard Transform, and calculate the first rate-distortion cost using the generated reference image block And the second rate-distortion cost is calculated based on a Sum of Absolute Difference (SAD) or a Sum of Square Difference (SD), and can be calculated using the reconstructed image block.

FIG. 6 is a reference view for explaining a process of dividing the encoding tree unit according to the intra prediction encoding method and the intra prediction encoding method according to an embodiment of the present invention. FIG. A coding method and a reference diagram for explaining the coding tree unit divided according to the apparatus.

6, the coding tree unit having a size of 32x32 is first set as the division target coding unit and divided into four sub-coding units, and the divided sub-coding units are again set as the divisional target coding units 4 The lower coding unit can be divided into the lower coding unit. As described above, the size of a side is reduced by half and the depth in a quad tree is increased by one every time the coding tree unit is divided. The intra-picture prediction coding method according to an embodiment of the present invention, and a case where the coding tree unit is finally divided as shown in FIG. 6 according to the apparatus, has a final coding unit and a prediction unit division structure as shown in FIG.

It is to be understood that the present invention is not limited to these embodiments, and all elements constituting the embodiment of the present invention described above are described as being combined or operated in one operation. That is, within the scope of the present invention, all of the components may be selectively coupled to one or more of them.

In addition, although all of the components may be implemented as one independent hardware, some or all of the components may be selectively combined to perform a part or all of the functions in one or a plurality of hardware. As shown in FIG. In addition, such a computer program may be stored in a computer readable medium such as a USB memory, a CD disk, a flash memory, etc., and read and executed by a computer to implement an embodiment of the present invention. As the recording medium of the computer program, a magnetic recording medium, an optical recording medium, a carrier wave medium, and the like can be included.

Furthermore, all terms including technical or scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined in the Detailed Description. Commonly used terms, such as predefined terms, should be interpreted to be consistent with the contextual meanings of the related art, and are not to be construed as ideal or overly formal, unless expressly defined to the contrary.

It will be apparent to those skilled in the art that various modifications, substitutions and substitutions are possible, without departing from the scope and spirit of the invention as disclosed in the accompanying claims. will be. Therefore, the embodiments disclosed in the present invention and the accompanying drawings are intended to illustrate and not to limit the technical spirit of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments and the accompanying drawings . The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

S110: CISPR setting unit setting step
S120: First partition determination step
S130: encoding mode candidate set determination step
S140: Split end determination step
S150: second split determination step
S160: Optimum encoding mode determination step
S170: Third split determination step
S180: prediction unit determination step

Claims (26)

An intra-picture prediction coding method for image coding,
A division target encoding unit setting step of setting a division target encoding unit (Coding Unit) belonging to a Coding Tree Unit and having a size equal to or smaller than that of the Coding Tree Unit;
Whether or not to divide the division target coding unit into sub coding units by using the size information of the division target coding unit, the size of the coding units adjacent to the coding tree unit, and the intra prediction coding mode information A first segmentation decision step of early decision; And
And a coding mode candidate set determining step of determining a candidate set of a predetermined number of intra prediction coding modes for the division target coding unit. The method of claim 1, wherein the determining of the encoding mode candidate set comprises:
A prediction unit for generating a reference image block by performing in-picture prediction on the intra-picture prediction coding mode for the intra-picture prediction coding mode for the partitioning object coding unit and using the information of the generated reference picture block and the partitioning object coding unit, Distortion cost, and determines a predetermined number of intra-prediction coding mode candidates based on the calculated first rate-distortion cost. The method of claim 1, Way. The method according to claim 1,
An intra prediction coding unit that applies the intra prediction coding mode belonging to the candidate set determined in the coding mode candidate set determination step to perform intra prediction coding on the division target coding unit, Distortion cost of the division target encoding unit and determining an optimal encoding mode based on the calculated second rate-distortion cost, characterized by further comprising: To-be-predicted encoding method. 2. The method according to claim 1,
And when the division target encoding unit is determined to be divided, dividing the division target encoding unit into the lower encoding units and dividing the divided lower encoding units into the division object encoding unit in the division object encoding unit setting step So that the intra-picture prediction coding is progressed for each of the divided lower coding units,
And when it is determined not to divide the division target encoding unit, proceeding to the encoding mode candidate set determination step. The method according to claim 1,
Wherein the first division determination step determines whether or not to divide the division target coding unit into the lower coding units by the following Equation (1).
Equation 1:

(here

Is a flag for determining whether to divide the division target encoding unit into the lower encoding units,

Is the length of one side of the division target encoding unit,

Is the length of one side of the leftmost encoding unit immediately adjacent to the upper side of the encoding tree unit,

Is the length of one side of the uppermost encoding unit immediately adjacent to the left side of the encoding tree unit,

Is the value of the intra-picture prediction coding mode of the leftmost coding unit immediately adjacent to the upper side of the coding tree unit,

Is the value of the intra-picture prediction encoding mode of the uppermost encoding unit immediately adjacent to the left of the encoding tree unit,

and

Is a constant of a predetermined size. 3. The intra prediction encoding method of claim 2, wherein the first rate-distortion cost is calculated according to Equation 2 and Equation 3 based on Sum of Absolute Transformed Difference (SATD).
Equation 2:

Equation 3:

(here

Is the first rate-distortion cost,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to encode the intra prediction encoding mode information,

Distortion value is calculated by performing a Hadamard transformation on a difference value between a target encoding unit for calculating the first rate-distortion cost and a reference image block generated by performing intra-frame prediction using a specific intra-frame prediction encoding mode for the target encoding unit Transform)

Wow

Is the index of a two-dimensional matrix that is the result of the Hadamard transform.) An intra-picture prediction coding method for image coding,
A division target encoding unit setting step of setting a division target encoding unit (Coding Unit) belonging to a Coding Tree Unit and having a size equal to or smaller than that of the Coding Tree Unit;
An encoding mode candidate set determining step of determining a predetermined number of intra prediction encoding mode candidates for the division target encoding unit;
A first rate-distortion cost calculated by using the size of the division target encoding unit, information of a reference image block generated by in-frame prediction of the division target encoding unit and information of the division target encoding unit, Distortion cost calculated for the encoding units adjacent to the encoding unit and the first rate-distortion cost calculated for the upper encoding unit before the division-target encoding unit is divided, A division end determination step of early determining whether or not to terminate a procedure of dividing a target coding unit into a sub coding unit (Sub Coding Unit); And
And determining an optimal intra picture prediction coding mode for the partitioning object coding unit among the candidate sets determined in the coding mode candidate set determination step. 8. The method of claim 7, wherein the determining of the encoding mode candidate set comprises:
In the intra-picture prediction mode for each intra-picture prediction coding mode, generates a reference picture block by performing in-picture prediction on the intra-picture prediction coding mode for the partitioning object coding unit, and generates a reference picture block using the generated reference picture block and information on the partitioning object coding unit Distortion cost, and determines a predetermined number of intra-picture prediction coding mode candidates based on the calculated first rate-distortion cost,
Wherein the optimal encoding mode determination step comprises:
An intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step is applied to perform intra-prediction coding on the division target coding unit, and the information on the reconstructed image block and the division target coding unit is decoded Distortion cost for the division target encoding unit, and determines an optimal encoding mode based on the calculated rate-distortion cost calculated by the second rate-distortion cost calculation unit. 8. The method of claim 7,
Using the size information of the partitioning object coding unit, the size of the coding units adjacent to the coding tree unit and the intra prediction coding mode information, after the partitioning object coding unit setting step and before the coding mode candidate set determining step, Further comprising a first division determination step of determining early whether or not to divide the division target coding unit into the lower coding units. 10. The method according to claim 9,
And when the division target encoding unit is determined to be divided, dividing the division target encoding unit into the lower encoding units and dividing the divided lower encoding units into the division object encoding unit in the division object encoding unit setting step So that the intra-picture prediction coding is progressed for each of the divided lower coding units,
And when it is determined not to divide the division target encoding unit, proceeding to the encoding mode candidate set determination step. 10. The method of claim 9,
Wherein the first division determination step determines whether or not to divide the division target coding unit into the lower coding units by the following equation (4).
Equation 4:

(here

Is a flag for determining whether to divide the division target encoding unit into the lower encoding units,

Is the length of one side of the division target encoding unit,

Is the length of one side of the leftmost encoding unit immediately adjacent to the upper side of the encoding tree unit,

Is the length of one side of the uppermost encoding unit immediately adjacent to the left side of the encoding tree unit,

Is the value of the intra-picture prediction coding mode of the leftmost coding unit immediately adjacent to the upper side of the coding tree unit,

Is the value of the intra-picture prediction encoding mode of the uppermost encoding unit immediately adjacent to the left of the encoding tree unit,

and

Is a constant of a predetermined size. 8. The intra prediction encoding method of claim 7, wherein the first rate-distortion cost is calculated according to Equation (5) and Equation (6) based on Sum of Absolute Transformed Difference (SATD).
Equation 5:

Equation 6:

(here

Is the first rate-distortion cost,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to encode the intra prediction encoding mode information,

Distortion value is calculated by performing a Hadamard transformation on a difference value between a target encoding unit for calculating the first rate-distortion cost and a reference image block generated by performing intra-frame prediction using a specific intra-frame prediction encoding mode for the target encoding unit Transform)

Wow

Is the index of a two-dimensional matrix that is the result of the Hadamard transform.) 8. The method according to claim 7,
And determines whether to end the procedure of dividing the division target encoding unit into the lower encoding units by the following equation (7).
Equation 7:

(here

Is the length of one side of the division target encoding unit,

Is a constant of a predetermined size,

Is a flag indicating whether to end the procedure of dividing the division target encoding unit into the lower encoding units,

Is the first rate-distortion cost of the smallest value when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step is applied to the division target coding unit,

Is the first rate-distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode which is optimally determined for the coding unit adjacent to the upper side of the division target coding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode that is optimally determined for the immediately preceding upper-coding unit immediately before the division-target coding unit is divided,

and

Is a constant of a predetermined size. 8. The method according to claim 7,
When it is determined to end the procedure of dividing the division target encoding unit, it is determined that the division target encoding unit is no longer divided into the lower encoding unit, and the optimal encoding mode determination step is performed To-be-predicted encoding method. 8. The method of claim 7,
Wherein, when it is decided not to end the procedure of dividing the division target encoding unit in the division end determination step, the size of the division target encoding unit, the first rate-distortion cost calculated for the division target encoding unit, Using the first rate-distortion cost calculated for the encoding units adjacent to the target encoding unit and the first rate-distortion cost calculated for the upper encoding unit before the segmentation subject encoding unit is divided, Further comprising a second division determination step of determining early whether or not to divide the target coding unit into the lower coding units. 16. The method of claim 15,
Wherein whether to divide the division target encoding unit is determined by the following Expression (8).
Equation 8:

(here

Is the length of one side of the division target encoding unit,

Is a constant of a predetermined size,

Is a flag indicating whether to divide the division target encoding unit into the lower encoding units,

Is the first rate-distortion cost of the smallest value when the intra-picture prediction coding mode belonging to the candidate set defined in the coding mode candidate set determination step is applied to the division target coding unit,

Is the first rate-distortion cost for the intra prediction encoding mode best determined for the encoding unit adjacent to the left of the division target encoding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode which is optimally determined for the coding unit adjacent to the upper side of the division target coding unit,

Is the first rate-distortion cost for the intra-picture prediction coding mode that is optimally determined for the immediately preceding upper-coding unit immediately before the division-target coding unit is divided,

and

Is a constant of a predetermined size. 16. The method of claim 15,
And when the division target encoding unit is determined to be divided, dividing the division target encoding unit into the lower encoding units and dividing the divided lower encoding units into the division object encoding unit in the division object encoding unit setting step So that the intra-picture prediction coding is progressed for each of the divided lower coding units,
And when it is determined not to divide the division target coding unit, proceeding to the optimal coding mode determination step. The intra-picture prediction coding method according to any one of claims 3 and 7 to 17, wherein the second rate-distortion cost is calculated by the following equation (9).
Equation 9:

(here

Is the second rate-distortion cost,

Distortion cost calculation target encoding unit and the second rate-distortion cost calculation target encoding unit, and decodes the second rate-distortion cost calculation target encoding unit and the second rate-distortion cost calculation target encoding unit, A Sum of Absolute Difference (SAD), which is the sum of the absolute values of the difference values, or a Sum of Square Difference (SSD), which is the sum of the squares of the difference values,

Is the Lagrangian constant used for optimization,

Is the amount of bits required to perform intra prediction encoding by applying the intra prediction encoding mode.) 18. A method according to any one of claims 3 or 7 to 17,
And a third division decision step of deciding whether to end the division procedure of the division target encoding unit and whether to divide the division object coding unit into the lower coding unit after the optimal coding mode decision step. To-be-predicted encoding method. 20. The method of claim 19,
Determining a division structure and a prediction unit (Prediction Unit) of the divided coding tree unit when the division procedure of the division target coding unit belonging to the coding tree unit is completed after the third division determination step, Further comprising a prediction unit determination step of determining a final intra prediction coding mode to be applied to the prediction units. 20. The method of claim 19,
When the depth value of the quadtree formed by dividing the encoding tree unit into the division target encoding units becomes equal to a predetermined maximum depth, it is determined to end the division procedure of the division target encoding unit,
And when the depth value is smaller than the maximum depth, dividing the division target encoding unit into the lower encoding units and setting the divided lower encoding units as the division target encoding units in the division target encoding unit setting step And the intra-picture prediction coding is newly performed for each of the divided lower coding units. 21. The method of claim 20,
Wherein if the coding tree unit is determined to end the division procedure in all of the generated coding units in the third division decision step, the second rate-distortion cost for the divisionally generated coding units is used Wherein the dividing structure of the coding tree unit is determined to determine the divided coding unit as the prediction unit and a final intra prediction coding mode to be applied to each of the prediction units is determined. / RTI > In the intra prediction encoding apparatus,
A coding unit for receiving a coding tree unit and dividing the coding unit into a coding unit, determining an intra-picture coding mode of the divided coding unit, and outputting a reference picture block according to the determined intra- An intra picture prediction unit for generating a prediction picture; And
And a differential block generator for generating the reference image block generated by the intra prediction unit and the differential block of the divided coding unit,
The intra-
Setting a coding object unit (Coding Unit) belonging to the coding tree unit and having a size equal to or smaller than that of the coding tree unit;
Whether or not to divide the division target coding unit into sub coding units by using the size information of the division target coding unit, the size of the coding units adjacent to the coding tree unit, and the intra prediction coding mode information The first decision is made;
Calculates a first Rate-Distortion Cost for the division target encoding unit by using the information of the reference image block generated by performing in-frame prediction on the division target encoding unit and the information of the division target encoding unit Determining a predetermined number of intra prediction encoding mode candidate sets based on the calculated first rate-distortion cost;
An intra-picture prediction coding mode belonging to the determined intra-picture prediction coding mode is applied to the intra-picture prediction coding mode of the partitioning object coding unit, the intra-picture prediction coding is performed again using the information of the reconstructed image block and the partitioning object coding unit, Calculating a second rate-distortion cost for the division target encoding unit, determining an optimal intra prediction encoding mode on the basis of the calculated rate-distortion cost;
Decides to divide the division target coding unit into the lower coding units if the depth value of the quadtree formed by dividing the coding tree unit into the division target coding units is smaller than a predetermined maximum depth, And determines to end the procedure of dividing the division target encoding unit into the lower encoding unit if the maximum depth is the same. 24. The apparatus of claim 23, wherein the intra-picture prediction unit predicts, after determining a candidate set of the predictive encoding mode and before determining the optimal intra-
The first rate-distortion cost calculated for the division target encoding unit, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate- Determining whether to end the procedure of dividing the division target encoding unit into the lower encoding units by using the first rate-distortion cost calculated for the upper encoding unit before the encoding unit is divided;
The first rate-distortion cost calculated for the division target encoding unit, the first rate-distortion cost calculated for the encoding units adjacent to the division subject encoding unit, and the first rate- Further comprising determining as a second order whether to divide the division subject coding unit into the lower coding units by using the first rate-distortion cost calculated for the upper coding unit before the coding unit is divided And outputs the intra-picture predictive coding. 24. The apparatus of claim 23, wherein the intra-
If it is determined to divide the division target encoding unit into the lower encoding units, the division target encoding unit is divided into the lower encoding units, and the divided lower encoding units are respectively set as the division target encoding units The intra-picture prediction unit performs an operation of the intra-picture prediction unit with respect to the divided lower-
If it is decided to end the division procedure for all of the encoding units generated by dividing the encoding tree unit, using the second rate-distortion cost for the divisionally generated encoding units, And determines a final intra-picture prediction encoding mode to be applied to each of the prediction units. The intra-picture prediction encoding apparatus according to claim 1, The apparatus of any one of claims 23 and 24, wherein the first rate-distortion cost is based on a Sum of Absolute Transformed Difference (SATD) using a Hadamard Transform The second rate-distortion cost is calculated using the generated reference image block, and the second rate-distortion cost is calculated based on a Sum of Absolute Difference (SAD) or a Sum of Square Difference (SSD) To-predictive encoding unit.

Images