Video decoding method and video encoding method
Video decoding method and video encoding method
 CN 107,580,230 B
 Filed: 01/21/2013
 Issued: 08/20/2021
 Est. Priority Date: 01/20/2012
 Status: Active Grant
First Claim
Patent Images
1. A video decoding method, comprising:
 determining a prediction scheme of a quantization matrix used in the inverse quantization; and
decoding a quantization matrix used in inverse quantization according to a prediction scheme of the quantization matrix,wherein the prediction scheme of the quantization matrix is any one of a prediction scheme between coefficients in the quantization matrix and a quantization matrix copy scheme,wherein when the flag information is 1, the prediction scheme of the quantization matrix is determined as a prediction scheme between coefficients in the quantization matrix, and when the flag information is 0, the prediction scheme of the quantization matrix is determined as a quantization matrix copy scheme,wherein when the prediction scheme of the quantization matrix is a quantization matrix copy scheme, the quantization matrix is determined based on the identification information,wherein when the identification information is 0, the quantization matrix is determined as a default quantization matrix, and when the identification information is not 0, the identification information is used to determine a reference quantization matrix,wherein when the identification information is not 0, coefficient values of the quantization matrix are determined to be the same as coefficient values of a reference quantization matrix derived by using the identification information,wherein the quantization matrix is specified according to a prediction mode, a color component type and a size of a block using the quantization matrix,wherein the identification information indicates a difference between matrixID indicating a prediction mode used in a quantization matrix to be decoded and a color component type of the quantization matrix and RefMatxID indicating a prediction mode used in the reference quantization matrix and a color component type of the reference quantization matrix,wherein the quantization matrix (ScalingList [ sizeID ] [ matrixID ] [ i ]) to be decoded is determined from the reference quantization matrix (ScalingList [ sizeID ] [ RefMatrixId ] [ i ]) by;
ScalingList[sizeID][matrixID][i]＝
ScalingList[sizeID][RefMatrixId][i]。
Chinese PRB Reexamination
Abstract
The present invention relates to a video decoding method and a video encoding method. The video decoding method includes: determining a prediction scheme of a quantization matrix used in the inverse quantization; and decoding the quantization matrix used in the inverse quantization according to a prediction scheme of the quantization matrix, wherein the prediction scheme of the quantization matrix is any one of a prediction scheme between coefficients in the quantization matrix and a quantization matrix copy scheme.
10 Claims

1. A video decoding method, comprising:

determining a prediction scheme of a quantization matrix used in the inverse quantization; and decoding a quantization matrix used in inverse quantization according to a prediction scheme of the quantization matrix, wherein the prediction scheme of the quantization matrix is any one of a prediction scheme between coefficients in the quantization matrix and a quantization matrix copy scheme, wherein when the flag information is 1, the prediction scheme of the quantization matrix is determined as a prediction scheme between coefficients in the quantization matrix, and when the flag information is 0, the prediction scheme of the quantization matrix is determined as a quantization matrix copy scheme, wherein when the prediction scheme of the quantization matrix is a quantization matrix copy scheme, the quantization matrix is determined based on the identification information, wherein when the identification information is 0, the quantization matrix is determined as a default quantization matrix, and when the identification information is not 0, the identification information is used to determine a reference quantization matrix, wherein when the identification information is not 0, coefficient values of the quantization matrix are determined to be the same as coefficient values of a reference quantization matrix derived by using the identification information, wherein the quantization matrix is specified according to a prediction mode, a color component type and a size of a block using the quantization matrix, wherein the identification information indicates a difference between matrixID indicating a prediction mode used in a quantization matrix to be decoded and a color component type of the quantization matrix and RefMatxID indicating a prediction mode used in the reference quantization matrix and a color component type of the reference quantization matrix, wherein the quantization matrix (ScalingList [ sizeID ] [ matrixID ] [ i ]) to be decoded is determined from the reference quantization matrix (ScalingList [ sizeID ] [ RefMatrixId ] [ i ]) by; ScalingList[sizeID][matrixID][i]＝
ScalingList[sizeID][RefMatrixId][i]。


2. the video decoding method of claim 1, wherein the identification information is used for a first function of deriving a reference quantization matrix and a second function of indicating that the quantization matrix is determined as a default quantization matrix.

3. The video decoding method of claim 1, wherein the default quantization matrix is defined in a decoder, and the identification information is scaling _ list _ pred _ matrix _ id _ delta.

4. The video decoding method according to claim 1, wherein in the decoding step of the quantization matrix,
when the prediction scheme of the quantization matrix is a prediction scheme between coefficients in a quantization matrix, a difference value between the coefficients in the quantization matrix is acquired and the quantization matrix is decoded based on the difference value between the coefficients in the quantization matrix, and the difference between the coefficients in the quantization matrix is the difference between the decoded target coefficient value and the previously decoded coefficient value in the quantization matrix.

5. The video decoding method according to claim 4, wherein in the decoding step of the quantization matrix, when the prediction scheme of the quantization matrix is a prediction scheme between coefficients in the quantization matrix, the DC matrix coefficient value is acquired for a quantization matrix having a predetermined size of 16x16 or 32x32, and an initial value for the quantization matrix coefficient is set to a constant value.

6. The video decoding method of claim 5, wherein the DC matrix coefficient value is scaling _ list _ DC _ coef _ minus8, and the constant value for the initial value of the quantization matrix coefficient is one of 8 or 16.

7. The video decoding method of claim 1, wherein the identification information indicates whether to use the reference quantization matrix or the default quantization matrix as the quantization matrix.

8. A video encoding method, comprising:

determining a quantization matrix to be used in quantization and performing quantization; determining a prediction scheme of a quantization matrix; and encoding information on the quantization matrix according to a prediction scheme of the quantization matrix, wherein the prediction scheme of the quantization matrix is any one of a prediction scheme between coefficients in the quantization matrix and a quantization matrix copy scheme, wherein the flag information is set to 1 if the prediction scheme of the quantization matrix is determined as the prediction scheme between coefficients in the quantization matrix, the flag information is set to 0 if the prediction scheme of the quantization matrix is determined as the quantization matrix copy scheme, wherein when the prediction scheme of the quantization matrix is a quantization matrix copy scheme, the quantization matrix is determined based on the identification information, wherein the identification information is set to 0 if the quantization matrix is determined as a default quantization matrix, is not set to 0 if the identification information is used to determine a reference quantization matrix, wherein when the identification information is not 0, coefficient values of the quantization matrix are determined to be the same as coefficient values of a reference quantization matrix derived by using the identification information, wherein the quantization matrix is specified according to a prediction mode, a color component type and a size of a block using the quantization matrix, wherein the identification information indicates a difference between matrixID indicating a prediction mode used in a quantization matrix to be encoded and a color component type of the quantization matrix and RefMatxID indicating a prediction mode used in the reference quantization matrix and a color component type of the reference quantization matrix, wherein the quantization matrix (ScalingList [ sizeID ] [ matrixID ] [ i ]) to be encoded is determined from the reference quantization matrix (ScalingList [ sizeID ] [ RefMatrixId ] [ i ]) by; ScalingList[sizeID][matrixID][i]＝
ScalingList[sizeID][RefMatrixId][i]。


9. the video coding method of claim 8, wherein the identification information is used for a first function of deriving a reference quantization matrix and a second function of indicating that the quantization matrix is determined as a default quantization matrix.

10. The video coding method of claim 8, wherein the identification information indicates whether to use the reference quantization matrix or the default quantization matrix as the quantization matrix.
Specification(s)