WO2023123358A1

WO2023123358A1 - Encoding and decoding methods, code stream, encoder, decoder, and storage medium

Info

Publication number: WO2023123358A1
Application number: PCT/CN2021/143688
Authority: WO
Inventors: 徐陆航
Original assignee: Oppo广东移动通信有限公司
Priority date: 2021-12-31
Filing date: 2021-12-31
Publication date: 2023-07-06

Abstract

Disclosed in embodiments of the present application are encoding and decoding methods, a code stream, an encoder, a decoder, and a storage medium. The method comprises: analyzing a code stream, and determining absolute values of reconstruction coefficients of the current block; scanning, according to a preset scanning sequence, the absolute values of the reconstruction coefficients of the current block, and determining, on the basis of the preset scanning sequence, the last coefficient group of the current block; and performing sign prediction on the last coefficient group to determine a predicted sign value for the last coefficient group. Therefore, the invention not only expands the scope of application of the sign prediction technology, but also improves the accuracy of prediction of positive and negative signs.

Description

Codec method, code stream, encoder, decoder and storage medium

technical field

The embodiments of the present application relate to the technical field of video encoding and decoding, and in particular, relate to an encoding and decoding method, a code stream, an encoder, a decoder, and a storage medium.

Background technique

With the improvement of people's requirements for video display quality, computer vision-related fields have received more and more attention. In recent years, image processing technology has been successfully applied in various industries. For the encoding and decoding process of video images, at the encoding end, the image data to be encoded will be compressed and encoded through the entropy encoding unit after transformation and quantization, and the code stream generated after the entropy encoding process will be transmitted to Decoder; then analyze the code stream, and after inverse quantization and inverse transformation processing, the original input image data can be restored.

However, for a non-zero quantized coefficient obtained through transformation and quantization, these non-zero quantized coefficients can be positive or negative, and sign prediction technology can be used to predict the sign of these coefficients. However, for the sign prediction of non-zero quantized coefficients at the image edge and block edge, due to the imperfection of related technologies, there may be defects such as limited transform block size and unreasonable cost calculation during sign prediction, resulting in inaccurate sign prediction results. low degree.

Contents of the invention

Embodiments of the present application provide a codec method, a code stream, an encoder, a decoder, and a storage medium, which can not only expand the scope of application of sign prediction technology, but also improve the accuracy of sign prediction for positive and negative signs.

The technical solutions of the embodiments of the present application can be implemented as follows:

In the first aspect, the embodiment of the present application provides a decoding method applied to a decoder, and the method includes:

Analyze the code stream to determine the absolute value of the reconstruction coefficient of the current block;

Scan the absolute value of the reconstruction coefficient of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

Make a sign prediction for the last coefficient group, and determine the sign prediction value for the last coefficient group.

In the second aspect, the embodiment of the present application provides an encoding method applied to an encoder, and the method includes:

determining transform coefficients for the current block;

Scan the transform coefficients of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

In the third aspect, the embodiment of the present application provides a code stream, which is generated by bit coding according to the information to be encoded; wherein the information to be encoded includes at least one of the following: the absolute value of the reconstruction coefficient and the symbol residual value .

In a fourth aspect, the embodiment of the present application provides an encoder, including a first determination unit and a first prediction unit; wherein,

A first determining unit configured to determine a transform coefficient of a current block;

The first determining unit is further configured to scan the transform coefficients of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

The first prediction unit is configured to perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.

In a fifth aspect, the embodiment of the present application provides an encoder, where the encoder includes a first memory and a first processor; wherein,

a first memory for storing a computer program capable of running on the first processor;

The first processor is configured to execute the method as described in the second aspect when running the computer program.

In a sixth aspect, the embodiment of the present application provides a decoder, including a decoding unit, a second determination unit, and a second prediction unit; wherein,

The decoding unit is configured to analyze the code stream and determine the absolute value of the reconstruction coefficient of the current block;

The second determining unit is configured to scan the absolute value of the reconstruction coefficient of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

In a seventh aspect, the embodiment of the present application provides a decoder, where the decoder includes a second memory and a second processor; wherein,

a second memory for storing a computer program capable of running on the second processor;

The second processor is configured to execute the method as described in the first aspect when running the computer program.

In an eighth aspect, an embodiment of the present application provides a computer storage medium, the computer storage medium stores a computer program, and when the computer program is executed, the method as described in the first aspect is implemented, or the method as described in the second aspect is implemented. . .

The embodiment of the present application provides a codec method, code stream, encoder, decoder, and storage medium. At the encoding end, the transformation coefficient of the current block is determined; the transformation coefficient of the current block is scanned according to the preset scanning order to determine The last coefficient group of the current block in the preset scanning order; perform sign prediction on the last coefficient group, and determine the sign prediction value of the last coefficient group. At the decoding end, analyze the code stream to determine the absolute value of the reconstruction coefficient of the current block; scan the absolute value of the reconstruction coefficient of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order; Sign prediction is performed for one coefficient group, and the sign prediction value for the last coefficient group is determined. In this way, both the encoding end and the decoding end can scan the transform coefficients of the current block according to the preset scanning order, and then determine the last coefficient group of the current block in the preset scanning order, so as to perform Sign prediction, which not only allows blocks of more shapes and sizes to perform sign prediction, but also expands the scope of application of sign prediction technology; and in the process of sign prediction, for the current block at the upper boundary or left of the image or coding tree unit For boundary conditions, the cost calculation method is also optimized, and at the same time, the energy calculation method for screening the transform coefficients to be predicted is also adaptively adjusted, thereby reducing the computational complexity and improving the accuracy of sign prediction Spend.

Description of drawings

Figure 1 is a schematic diagram of the composition of a hybrid coding framework;

Fig. 2 is a schematic diagram of a block including non-zero quantized coefficients;

Fig. 3 is a schematic diagram of a coefficient block to be inversely transformed;

Fig. 4 is a schematic diagram of a coefficient block of an inverse transform;

Fig. 5 is a schematic diagram of a normalized template;

Fig. 6 is a schematic diagram of a normalized template of inverse transformation;

FIG. 7 is a schematic diagram of an application for calculating a cost based on a hypothetical reconstruction value;

FIG. 8A is a schematic diagram of a system composition block diagram of an encoder provided by an embodiment of the present application;

FIG. 8B is a schematic diagram of a system composition block diagram of a decoder provided in an embodiment of the present application;

FIG. 9 is a first schematic flowchart of a decoding method provided by an embodiment of the present application;

FIG. 10 is a schematic diagram of a test result using symbol prediction for a block smaller than 4×4 provided by an embodiment of the present application;

FIG. 11 is a schematic diagram of a position where the current block is at the boundary of a coding tree unit provided by an embodiment of the present application;

FIG. 12 is a schematic diagram of a test result of using symbol prediction when the current block is at the coding tree unit boundary provided by the embodiment of the present application;

Fig. 13 is a schematic diagram of a position where the current block is at the image boundary provided by the embodiment of the present application;

Fig. 14 is a schematic diagram of a test result of using symbol prediction when the current block is at the image boundary provided by the embodiment of the present application;

Fig. 15 is a schematic diagram of the position of the current block in the upper left corner of the image provided by the embodiment of the present application;

Fig. 16 is a schematic diagram of a test result in which the current block is in the upper left corner of the image without using symbol prediction according to the embodiment of the present application;

FIG. 17 is a second schematic flow diagram of a decoding method provided by an embodiment of the present application;

FIG. 18 is a first schematic flowchart of an encoding method provided by an embodiment of the present application;

FIG. 19 is a schematic flow diagram II of an encoding method provided by an embodiment of the present application;

FIG. 20 is a schematic flow diagram III of an encoding method provided in an embodiment of the present application;

FIG. 21 is a schematic diagram of the composition and structure of an encoder provided by an embodiment of the present application;

FIG. 22 is a schematic diagram of a specific hardware structure of an encoder provided by an embodiment of the present application;

FIG. 23 is a schematic structural diagram of a decoder provided by an embodiment of the present application;

FIG. 24 is a schematic diagram of a specific hardware structure of a decoder provided in an embodiment of the present application;

FIG. 25 is a schematic diagram of the composition and structure of an encoding and decoding system provided by an embodiment of the present application.

Detailed ways

In order to understand the characteristics and technical contents of the embodiments of the present application in more detail, the implementation of the embodiments of the present application will be described in detail below in conjunction with the accompanying drawings. The attached drawings are only for reference and description, and are not intended to limit the embodiments of the present application.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application, and are not intended to limit the present application.

In the following description, references to "some embodiments" describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or a different subset of all possible embodiments, and Can be combined with each other without conflict. It should also be pointed out that the term "first\second\third" involved in the embodiment of the present application is only used to distinguish similar objects, and does not represent a specific ordering of objects. Understandably, "first\second\ The specific order or sequence of "third" may be interchanged where permitted so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described herein.

Before the embodiment of the present application is described in further detail, the nouns and terms involved in the embodiments of the present application are explained first, and the nouns and terms involved in the embodiments of the present application are applicable to the following explanations:

A new generation of video coding standard H.266/Versatile Video Coding (VVC)

Coding Unit (CU)

Coding Tree Unit (CTU)

Largest Coding Unit (LCU)

Prediction Unit (PU)

Transform Unit (Transform Unit, TU)

Discrete Cosine Transform (DCT)

VVC's reference software test model (VVC Test Model, VTM)

Sign Prediction (Sign Prediction, SP): a prediction technique for the positive and negative of non-zero transform coefficients

Enhanced Compression Model (ECM): Based on VTM-10.0, it integrates various new tools and can further explore the reference model of VVC performance

It can be understood that currently common video codec standards (such as VVC) all adopt a block-based hybrid coding framework. Each frame in the video image is divided into square LCUs of the same size (such as 128×128, 64×64, etc.), and each LCU can also be divided into rectangular CUs according to the rules; and the CU may also be divided into smaller ones PU, TU, etc. Specifically, as shown in Figure 1, the hybrid coding framework may include steps such as prediction, transformation, quantization, entropy coding, and loop filtering. Among them, prediction can be divided into intra prediction (Intra Prediction) and inter prediction (Inter Prediction), and inter prediction can include motion estimation (Motion Estimation) and motion compensation (Motion Compensation). Because there is a strong correlation between adjacent pixels in a video image, the use of intra-frame prediction in video coding and decoding technology can eliminate the spatial redundancy between adjacent pixels; but due to the adjacent pixels in the video image There is also a strong similarity between frames. In the video coding and decoding technology, the inter-frame prediction method is used to eliminate the temporal redundancy between adjacent frames, thereby improving the coding and decoding efficiency.

The basic flow of a video codec is as follows: In the encoder, a frame of image is divided into blocks, intra-frame prediction or inter-frame prediction is used for the current block to generate the prediction block of the current block, and the original block of the current block is subtracted from the prediction block to obtain The residual block is transformed and quantized to obtain a quantized coefficient matrix, and the quantized coefficient matrix is entropy encoded and output to the code stream. In the decoder, intra prediction or inter prediction is used for the current block to generate the prediction block of the current block. On the other hand, the quantization coefficient matrix is obtained by decoding the code stream, and the quantization coefficient matrix is dequantized and inversely transformed to obtain the residual block. The predicted block and the residual block are added to obtain the reconstructed block. The reconstructed blocks form a reconstructed image, and the decoded image is obtained by performing loop filtering on the reconstructed image based on the image or based on the block. The encoder also needs similar operations as the decoder to obtain the decoded image. The decoded image may serve as a reference frame for inter-frame prediction for subsequent frames. The block division information determined by the encoder, mode information or parameter information such as prediction, transformation, quantization, entropy coding, and loop filtering, if necessary, needs to be output to the code stream; then the decoder determines and The encoder has the same block division information, mode information or parameter information such as prediction, transformation, quantization, entropy coding, and loop filtering, so as to ensure that the decoded image obtained by the encoder is the same as that obtained by the decoder. The decoded image obtained by the encoder is usually also called the reconstructed image. During prediction, the current block may be divided into prediction units, and during transformation, the current block may be divided into transformation units, and the division of prediction units and transformation units may be different. The above is the basic process of the video encoder and decoder under the block-based hybrid coding framework. With the development of technology, some modules or steps of the framework or process may be optimized. The embodiments of this application are applicable to the block-based The basic flow of the video codec under the hybrid coding framework, but not limited to the framework and flow.

In this embodiment of the present application, the current block (Current Block) may be a current coding unit (CU), a current prediction unit (PU), or a current transform block (TU).

It should be noted that, similar to many codecs, the encoder in VVC makes a difference between the original value and the predicted value of the current block to obtain a residual block, and the residual block is transformed to obtain the transformed residual, and then the transformed The residual is quantized to obtain quantized coefficients, and then the quantized coefficients are encoded into the code stream. For a non-zero quantized coefficient obtained after transformation and quantization, since the probability of the non-zero quantized coefficient’s positive and negative occurrences is random, equal-probability coding is used in VVC and other standards (where the positive and negative probabilities are both 50 %).

Exemplarily, Fig. 2 shows a schematic diagram of a block containing non-zero quantized coefficients. Among them, the non-zero quantization coefficients are 5, -1, -2 and 1. Their plus and minus signs are +, -, -, + respectively. Positive and negative can be represented by a binary code, such as 1 for + and 0 for -. The Sign Prediction technology can be used to predict the positive and negative of the non-zero quantized coefficients in this block, assuming that the predicted positive and negative are +, -, -, -. Then according to the above content, the correct positive and negative values are 1, 0, 0, 1; the predicted values are 1, 0, 0, 0. Whether the prediction is correct or not can be called "residual error", and each "residual error" can also be represented by a binary code, for example, 0 is correct and 1 is wrong. At this time, the "residual" is 0, 0, 0, 1. It is not difficult to see that when the prediction is more accurate, the probability of 0 in the "residual" will be higher than the probability of 1. For such binary codes with obvious uneven distribution, using context model coding can greatly reduce the codeword. The Sign prediction technology is to use the information between the current block and the adjacent blocks to accurately predict the sign of the non-zero quantization coefficient in the current block, so that the "residual" coding cost of the sign is reduced.

It can also be understood that the transform coefficients are divisible. Among them, the transformation used in image/video compression is a linear transformation. Taking the most commonly used DCT-II transformation as an example, assuming that the transformation coefficient of the current block is shown in Figure 3, a coefficient block to be inversely transformed is shown here. Can be represented by block A.

After the block A in Fig. 3 is inversely transformed using DCT-II transformation, the results in Fig. 4 can be obtained (each value is the result of rounding after ×64), and a schematic diagram of a coefficient block of inverse transformation is shown here, It can be represented by InvA.

Since the transformation is linear, the embodiment of the present application can split the above inverse transformation. Specifically, two normalized templates shown in FIG. 5 can be used, including template B and template C. Then perform DCT-II inverse transformation on template B and template C respectively, and the result in Figure 6 can be obtained (each value is the result of rounding after ×64), which shows the two templates after inverse transformation, including template InvB and template InvC.

According to Figure 3 to Figure 6, it can be obtained that A=2×B-C, InvA=2×InvB-InvC. It can be seen that by constructing the inversely transformed template, the inversely transformed template can be accumulated in proportion to obtain any residual after inversely transformed.

It will also be appreciated that the sign of the transform coefficients can be predicted. Wherein, predicting the sign of the transformation coefficient should first be based on the known (decoded) basis of the absolute value of the quantization coefficient and the transformation mode used by the current block.

In ECM, Sign prediction only predicts the signs of non-zero transform coefficients in the upper left N×N area of the transform unit, and the number of predicted signs must be less than or equal to K. For example, N=4, K=8.

In a specific embodiment, for a transformation coefficient containing L signs to be predicted, the prediction process is divided into the following steps:

Step 1: Construct the inverse-transformed templates of the possible sizes of each transformation block of each transformation type. The inverse-transformation template only needs to use the values of the uppermost row and the leftmost column, so only these values are saved.

Step 2: Calculate the absolute value of the coefficient after inverse quantization. If L is greater than K, select K coefficients according to a certain rule to predict the sign. 2 ^K possible positive and negative combinations, according to the pre-calculated template, calculate the hypothetical inverse transform residual value of the 2 ^K transform blocks in the first row and the first column.

Step 3: Using the inverse transformation residual values and predicted values of the 2 ^K hypotheses of the transform block, 2 ^K hypothesis reconstruction values (hypothesis reconstruction) can be obtained, and one of the 2 ^K hypothesis reconstruction values is selected to be consistent with the surrounding reconstructed The hypothetical reconstruction value with the least cost between the blocks of ; where the hypothetical reconstruction value is equal to the inverse transformation residual value plus the prediction value of the current transformation block. Here, as shown in FIG. 7 , the reconstructed adjacent pixel values of the current block may include the upper two rows and the left two columns of the current transform block, where the upper two rows are respectively represented by p _x,-1 and p _{x, -2} means, the two columns on the left are respectively represented by p _-1,y and p _-2,y , the range of x is 0,1,...,w-1, and the range of y is 0,1,...,h-1 , w represents the width of the current transformation block, and h represents the height of the current transformation block; at this time, the cost value is calculated according to the assumed reconstruction value, as shown below,

Step 4: The set of signs used by the inverse transformation residual value with the smallest cost value (expressed in cost) is the sign prediction made by Sign Prediction.

Step 5: For the selected group of predicted positive and negative values, specifically, the positive and negative values of the optimal quantization coefficient in the current mode selected by the current transform block at the encoding end through rate-distortion optimization are taken as the real positive and negative values value, and then determine the residual of the positive and negative values based on the true positive and negative values and the predicted positive and negative values. The context model is used in the subsequent encoding process to encode positive and negative residuals. In this way, the decoder decodes positive and negative residuals from the context model. In the follow-up process, the real positive and negative values can also be obtained according to the residual of the positive and negative values and the predicted positive and negative values, and the real positive and negative values can be directly assigned to the transformation coefficients to further complete subsequent decoding operations (such as inverse transformation, reconstruction, etc. ).

In the implementation of ECM, formula (1) can be further simplified as follows,

Due to the prediction value of the current block (denoted by pred), the surrounding neighboring pixel reconstruction values are fixed for the current block, i.e. (2×p _x,-1 -p _x,-2 -pred _x,0 ) and (2×p _-1,y -p _-2,y -pred _0,y ) values are fixed, only the assumed inverse transformation residual (denoted by resi) will be different due to different symbol combinations, using c _x to represent (2×p _x,-1 -p _x,-2 -pred _x,0 ), c _y represents (2×p _{-1, y} -p _{-2, y} -pred _{0, y} ), at this time the generation value ( Expressed by cost) can be written as the following formula,

However, in order to control the complexity of transform coefficient sign prediction, there are two main limitations of the prediction technique:

(1) Prediction only acts on the 4×4 region of the upper left corner of each 4×4 transform block that is greater than or equal to it. The reason is that the upper left corner of the transform block is where low-frequency coefficients are concentrated, and changes in the value of these coefficients have a greater impact on the entire transform block, so these coefficients are easier to predict than coefficients at other positions.

(2) In the above area, only predict the sign of less than or equal to 8 non-zero transform coefficients, when there are more than 8 non-zero transform coefficients in the 4×4 area, then only predict the first 8 raster scan order The sign of the nonzero transform coefficients on . The reason is that in order to control the complexity, predicting K coefficients requires finding the one with the smallest boundary cost from 2 ^K inversely transformed coefficients. The larger K is, the higher the complexity is; and the smaller K is, the more complex it is. Low.

Furthermore, in the implementation, for the situation that more than 8 non-zero transform coefficients may appear in the upper left 4×4 area, it is also proposed to select 8 non-zero transform coefficients according to certain criteria for sign prediction. Moreover, in the process of finding more than 8 non-zero transform coefficients, it is easier to predict the non-zero transform coefficients at positions that have a greater impact on the block boundary when inverse transform is assumed. One implementation method is to directly select the non-zero coefficient with a large absolute value from the 8 non-zero transformation coefficients; the other implementation method is to sort the non-zero transformation coefficients in the 4×4 area and reflect them on the edge of the transformation unit after inverse transformation Energy size, determine the 8 non-zero coefficients with the largest energy as the transform coefficients to be predicted positive and negative. where the energy value is calculated as follows,

Here, tc _i,j is the transformation coefficient in the upper left corner region of 4×4 (i,j represents coordinates), T _i,j (k) is the transformation coefficient at position (i,j) when it is 1, The k-th inverse transform value in the left edge and upper edge of the current transform block, w and h represent the width and height of the current transform block.

To put it simply, for a non-zero quantized coefficient obtained through transformation and quantization, these non-zero quantized coefficients can be positive or negative. At present, equal probability coding can be used, and sign prediction technology can also be used to predict the coefficients of these coefficients. The plus and minus signs make predictions. However, for the sign prediction of non-zero quantized coefficients at the image edge and block edge, due to the imperfection of related technologies, there may be defects such as limited transform block size and unreasonable cost calculation during sign prediction, resulting in the sign prediction results. Accuracy is low.

In this way, the embodiment of the present application provides a decoding method. By analyzing the code stream, the absolute value of the reconstruction coefficient of the current block is determined; the absolute value of the reconstruction coefficient of the current block is scanned according to the preset scanning order, and it is determined that the current block is within the preset scanning sequence. The last coefficient group in order; sign prediction is performed on the last coefficient group, and the sign prediction value of the last coefficient group is determined.

The embodiment of the present application also provides an encoding method, by determining the transformation coefficient of the current block; scanning the transformation coefficient of the current block according to the preset scanning order, and determining the last coefficient group of the current block in the preset scanning order; The last coefficient group performs sign prediction, and the sign prediction value of the last coefficient group is determined.

In this way, both the encoding end and the decoding end can scan the transform coefficients of the current block according to the preset scanning order, and then determine the last coefficient group of the current block in the preset scanning order, so as to perform Sign prediction, which not only allows blocks of more shapes and sizes to perform sign prediction, but also expands the scope of application of sign prediction technology; and in the process of sign prediction, for the current block at the upper boundary or left of the image or coding tree unit For boundary conditions, the cost calculation method is also optimized, and at the same time, the energy calculation method for screening the transform coefficients to be predicted is also adaptively adjusted, thereby reducing the computational complexity and improving the accuracy of sign prediction Spend.

Various embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

Referring to FIG. 8A , it shows a schematic diagram of a system composition block diagram of an encoder provided by an embodiment of the present application. As shown in FIG. 8A, the encoder 100 may include: a segmentation unit 101, a prediction unit 102, a first adder 107, a transform unit 108, a quantization unit 109, an inverse quantization unit 110, an inverse transform unit 111, a second adder 112, A filtering unit 113, a decoded picture buffer (Decoded Picture Buffer, DPB) unit 114 and an entropy encoding unit 115. Here, the input of the encoder 100 may be a video composed of a series of pictures or a static picture, and the output of the encoder 100 may be a bit stream (also called a "code stream") representing a compressed version of the input video. .

Wherein, the segmentation unit 101 divides the pictures in the input video into one or more coding tree units (Coding Tree Units, CTUs). The segmentation unit 101 divides the picture into multiple tiles (or called tiles, tiles), and can further divide a tile into one or more bricks (bricks). Here, a tile or a brick can include one or more complete and/or partial CTUs. In addition, the segmentation unit 101 may form one or more slices, wherein one slice may include one or more tiles arranged in a grid order in the picture, or one or more tiles covering a rectangular area in the picture. The segmentation unit 101 may also form one or more sub-pictures, wherein one sub-picture may include one or more slices, tiles or bricks.

During the encoding process of the encoder 100 , the division unit 101 transmits the CTU to the prediction unit 102 . Generally, the prediction unit 102 may be composed of a block division unit 103, a motion estimation (Motion Estimation, ME) unit 104, a motion compensation (Motion Compensation, MC) unit 105 and an intra prediction unit 106. Specifically, the block partitioning unit 103 iteratively uses quadtree partitioning, binary tree partitioning and ternary tree partitioning to further divide the input CTU into smaller coding units (Coding Units, CUs). Prediction unit 102 may use ME unit 104 and MC unit 105 to obtain an inter-prediction block for a CU. Intra prediction unit 106 may obtain an intra prediction block for a CU using various intra prediction modes including MIP mode. In an example, the rate-distortion optimized motion estimation approach can be invoked by the ME unit 104 and the MC unit 105 to obtain an inter prediction block, and the rate-distortion optimized mode determination approach can be invoked by the intra prediction unit 106 to obtain an intra prediction block .

The prediction unit 102 outputs the prediction block of the CU, and the first adder 107 calculates the difference between the CU in the output of the division unit 101 and the prediction block of the CU, that is, the residual CU. The transform unit 108 reads the residual CU and performs one or more transform operations on the residual CU to obtain coefficients. The quantization unit 109 quantizes the coefficients and outputs the quantized coefficients (ie levels). The inverse quantization unit 110 performs a scaling operation on the quantized coefficients to output reconstructed coefficients. The inverse transform unit 111 performs one or more inverse transforms corresponding to the transforms in the transform unit 108 and outputs a reconstruction residual. The second adder 112 calculates a reconstructed CU by adding the reconstruction residual to the prediction block of the CU from the prediction unit 102 . The second adder 112 also sends its output to the prediction unit 102 for use as an intra prediction reference. After all CUs in a picture or sub-picture are reconstructed, the filtering unit 113 performs loop filtering on the reconstructed picture or sub-picture. Here, the filtering unit 113 includes one or more filters, such as a deblocking filter, a sample adaptive offset (Sample Adaptive Offset, SAO) filter, an adaptive loop filter (Adaptive Loop Filter, ALF), a brightness map And chroma scaling (Luma Mapping with Chroma Scaling, LMCS) filters and filters based on neural networks, etc. Alternatively, when the filtering unit 113 determines that the CU is not used as a reference for encoding other CUs, the filtering unit 113 performs loop filtering on one or more target pixels in the CU.

The output of the filtering unit 113 is decoded pictures or sub-pictures, which are buffered into the DPB unit 114 . The DPB unit 114 outputs decoded pictures or sub-pictures according to timing and control information. Here, the pictures stored in the DPB unit 114 can also be used as a reference for the prediction unit 102 to perform inter prediction or intra prediction. Finally, the entropy encoding unit 115 converts the parameters (such as control parameters and supplementary information) necessary for decoding pictures from the encoder 100 into binary form, and writes such binary form into the code stream according to the syntax structure of each data unit , that is, the encoder 100 finally outputs a code stream.

Further, the encoder 100 may have a first processor and a first memory recording a computer program. When the first processor reads and runs the computer program, the encoder 100 reads the input video and generates a corresponding code stream. Additionally, encoder 100 may also be a computing device having one or more chips. These units, implemented as integrated circuits on the chip, have similar connection and data exchange functions as the corresponding units in FIG. 8A.

Referring to FIG. 8B , it shows a schematic diagram of a system composition block diagram of a decoder provided by an embodiment of the present application. As shown in FIG. 8B , the decoder 200 may include: an analysis unit 201 , a prediction unit 202 , an inverse quantization unit 205 , an inverse transformation unit 206 , an adder 207 , a filter unit 208 and a decoded picture buffer unit 209 . Here, the input of the decoder 200 is a bit stream representing a compressed version of a video or a still picture, and the output of the decoder 200 may be a decoded video composed of a series of pictures or a decoded still picture.

Wherein, the input code stream of the decoder 200 may be the code stream generated by the encoder 100 . The parsing unit 201 parses the input code stream and obtains the value of the syntax element from the input code stream. The parsing unit 201 converts the binary representation of the syntax elements into digital values and sends the digital values to units in the decoder 200 to obtain one or more decoded pictures. The parsing unit 201 may also parse one or more syntax elements from the input code stream to display the decoded picture.

During the decoding process of the decoder 200, the parsing unit 201 sends the value of the syntax element and one or more variables set or determined according to the value of the syntax element to obtain one or more decoded pictures to the decoder 200 unit.

The prediction unit 202 determines a prediction block for a currently decoded block (eg, CU). Here, the prediction unit 202 may include a motion compensation unit 203 and an intra prediction unit 204 . Specifically, when the inter-frame decoding mode is indicated for decoding the current decoding block, the prediction unit 202 passes the relevant parameters from the parsing unit 201 to the motion compensation unit 203 to obtain the inter-frame prediction block; when the intra-frame prediction mode ( When the MIP mode indicated based on the MIP mode index value) is used to decode the current decoding block, the prediction unit 202 transmits the relevant parameters from the parsing unit 201 to the intra prediction unit 204 to obtain the intra prediction block.

The dequantization unit 205 has the same function as the dequantization unit 110 in the encoder 100 . The inverse quantization unit 205 performs a scaling operation on the quantization coefficients (ie levels) from the parsing unit 201 to obtain reconstruction coefficients.

The inverse transform unit 206 has the same function as the inverse transform unit 111 in the encoder 100 . The inverse transform unit 206 performs one or more transform operations (ie, the inverse of the one or more transform operations performed by the inverse transform unit 111 in the encoder 100 ) to obtain the reconstruction residual.

The adder 207 performs an addition operation on its inputs (the predicted block from the prediction unit 202 and the reconstructed residual from the inverse transform unit 206) to obtain the reconstructed block of the currently decoded block. The reconstructed block is also sent to the prediction unit 202 to be used as a reference for other blocks encoded in intra prediction mode.

After all CUs in a picture or sub-picture are reconstructed, filtering unit 208 performs loop filtering on the reconstructed picture or sub-picture. The filtering unit 208 includes one or more filters, such as a deblocking filter, a sampling adaptive compensation filter, an adaptive loop filter, a luma mapping and chroma scaling filter, and a neural network-based filter. Alternatively, when filtering unit 208 determines that the reconstructed block is not used as a reference when decoding other blocks, filtering unit 208 performs loop filtering on one or more target pixels in the reconstructed block. Here, the output of the filtering unit 208 is a decoded picture or sub-picture, and the decoded picture or sub-picture is cached in the DPB unit 209 . The DPB unit 209 outputs decoded pictures or sub-pictures according to timing and control information. The picture stored in the DPB unit 209 can also be used as a reference to perform inter prediction or intra prediction by the prediction unit 202 .

Further, the decoder 200 may have a second processor and a second memory recording a computer program. When the first processor reads and runs the computer program, the decoder 200 reads the input code stream and generates a corresponding decoded video. Additionally, decoder 200 may also be a computing device having one or more chips. These units implemented on-chip as integrated circuits have similar connectivity and data exchange functions as the corresponding units in Figure 8B.

It should also be noted that when the embodiment of the present application is applied to the encoder 100, the "current block" specifically refers to the block currently to be encoded in the video image; when the embodiment of the present application is applied to the decoder 200, the "current block "Specifically refers to the current block to be decoded in the video image. In addition, the current block here may be a current coding unit, a current prediction unit, or a current transformation block, etc., which is not limited in this embodiment of the present application.

In an embodiment of the present application, refer to FIG. 9 , which shows a schematic flowchart of a decoding method provided in an embodiment of the present application. As shown in Figure 9, the method may include:

S901: Analyze the code stream, and determine the absolute value of the reconstruction coefficient of the current block.

It should be noted that the decoding method in the embodiment of the present application is applied to a decoder, and may specifically be a sign prediction method, where the sign of the absolute value of the reconstruction coefficient in the current block is mainly predicted.

It should also be noted that, based on the composition structure of the decoder 200 shown in FIG. 8B , the decoding method of the embodiment of the present application is applied to the "analysis unit 201" part of the decoder 200. For the analysis unit 201, it can be decoded The code stream obtains the absolute value of the reconstruction coefficient of the current block, so that the absolute value of the reconstruction coefficient that needs to be predicted is selected according to the absolute value of the reconstruction coefficient, and then the sign of the reconstruction coefficient is predicted.

It can be understood that, in the implementation of ECM, for the current block, the block size is usually greater than or equal to 4×4. For a current block smaller than 4×4, since there is no 4×4 area in the upper left corner for sign prediction (Sign Prediction), such a current block is prohibited from using sign prediction. At the same time, sign prediction is also prohibited for intra blocks larger than 32×32 and inter blocks larger than 128×128. However, in the embodiment of the present application, by determining the last coefficient group (Coefficient Group, CG) in the preset scanning order, and then using sign prediction for the last coefficient group, the sign prediction can be applied to more shapes and more Multi-sized blocks extend the applicability of sign prediction techniques.

S902: Scan the absolute values of the reconstruction coefficients of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order.

S903: Perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.

It should be noted that, in the embodiment of the present application, it may be allowed to select non-zero transform coefficients in the last coefficient group in the preset scanning order to use sign prediction. In this way, firstly, the absolute value of the reconstruction coefficient of the current block can be scanned according to the preset scanning order to determine the last coefficient group of the current block in the preset scanning order; The sign prediction value of , which can allow blocks of more shapes and sizes to perform sign prediction, and expand the application range of sign prediction technology.

It should also be noted that, in this embodiment of the present application, the preset scanning order may be diagonal, Zigzag, horizontal, vertical, 4×4 sub-block scanning or any other scanning order, which is not limited herein.

It should also be noted that current blocks of different sizes may be divided into coefficient groups of different numbers. Therefore, for the division of coefficient groups, in some embodiments, the method may further include:

If the size of the current block is 1×N, and N is an integer greater than or equal to 16, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient group is 1×16;

If the size of the current block is N×1, and N is an integer greater than or equal to 16, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient group is 16×1;

If the size of the current block is 2×N, and N is an integer greater than or equal to 8, then determine that the absolute value of the reconstruction coefficient of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient group is 2×8;

If the size of the current block is N×2, and N is an integer greater than or equal to 8, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient group is 8×2;

If the size of the current block is M×N, and both M and N are integers greater than or equal to 4, it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (M×N/16) coefficient groups, and the size of the coefficient group for 4×4.

It should be noted that for the size of the current block, 1×N and 2×N blocks may also appear, and the shapes of these blocks and the corresponding coefficient group sizes are shown in Table 1.

Table 1

块尺寸block size	系数组的尺寸The size of the coefficient group
1×N,N≥161×N, N≥16	1×161×16
N×1,N≥16N×1,N≥16	16×116×1
2×N,N≥82×N,N≥8	2×82×8
N×2,N≥8N×2,N≥8	8×28×2
其它形状的当前块current block of other shapes	4×44×4

In the related art, for a side whose length is less than 4, ECM currently does not allow such a block to use sign prediction. However, in the embodiment of the present application, it is allowed to use sign prediction for non-zero coefficients in all regions or partial regions in the block (for example, the same as the 4×4 region at the upper left corner position in ECM).

Exemplarily, for a current block larger than 4×4 in the ECM, sign prediction is allowed for non-zero coefficients in the 4×4 area at the upper left corner (coefficient group at the upper left corner); similarly, for blocks smaller than 4×4 For the current block, it is also allowed to select non-zero coefficients in the last coefficient group in the preset scanning order to use sign prediction.

Referring to FIG. 10 , it shows a schematic diagram of a test result of using sign prediction for a block smaller than 4×4 provided by an embodiment of the present application. As shown in Figure 10, it is provided here that when a block smaller than 4×4 is allowed to use sign prediction for the last coefficient group in the preset scanning order, the coding effect improvement brought by the ECM reference software under full intra-frame coding is compared. Among them, under the three components of Y component, U component and V component under the YUV color space, the performance improvement of 0.02%, 0.1% and 0.07% was brought respectively, and the encoding time (expressed by EncT) and decoding time (expressed by DecT said) little changed.

In addition, in practical applications, the closer to the upper left corner, the later in the preset scanning order, that is, the last coefficient group is located at the upper left corner of the current block. Therefore, the embodiment of the present application may also cancel the restriction on the maximum size of a block using the sign prediction technology.

Exemplarily, in the ECM, the current intra block larger than 32×32 and the inter block larger than 128×128 cannot use sign prediction, but actually the maximum size block is currently allowed to be 256×256. So removing the restriction on the maximum block size means allowing up to 256x256 transformation blocks in part (for example, the 4x4 region with the upper left corner position in ECM) or the sign of non-zero coefficient signs in all regions predict.

Further, when the size of the current block is M×N, and both M and N are integers greater than or equal to 4, then the size of the coefficient group is 4×4. In some embodiments, the method may also include:

Determine the first area in the upper left corner of the current block;

The last coefficient group is determined according to the absolute value of the reconstruction coefficients in the first area; wherein, the number of absolute values of the reconstruction coefficients in the last coefficient group is 2 ^L , and L is the number greater than or equal to zero.

It should be noted that, generally, if the size of the current block is greater than or equal to 4×4, then the 4×4 area at the upper left corner can be directly used as the last coefficient group. In addition, considering that the number of non-zero coefficients used for symbol prediction is usually K, if the number of non-zero coefficients in the 4×4 area is less than K, the upper left area can be appropriately expanded at this time, for example, the 8 ×8 area, 16×16 area, etc. as the last coefficient group, so that the number of absolute value reconstruction coefficients in the last coefficient group is ^2L . In the embodiment of the present application, the first area may be a 4×4 area, but it is not specifically limited here.

It should also be noted that K is an integer greater than zero. Exemplarily, for the first transformation, the value of K may be 8; for the second transformation, the value of K may be 4, but this is not specifically limited.

In this way, the use range of the current block that can be predicted by using the sign of the transform coefficient can be expanded, allowing blocks with more shapes and sizes to use the sign prediction technology. Specifically, for example, a block smaller than 4×4, a block larger than 32×32 in an intra frame, and a block larger than 128×128 in an inter frame can all use this symbol prediction technology.

It can also be understood that, as shown in FIG. 11 , (a) shows that the current block is at the upper boundary of the coding tree unit, and (b) shows that the current block is at the left boundary of the coding tree unit. At this time, in the implementation of ECM, if the current block is at the edge of the coding tree unit, then the calculation of c _x and _cy will use the reconstruction value in the adjacent coding tree unit, see the aforementioned Figure 7 and formula (1) for details shown. Specifically, when the current block in the first row of the coding tree unit needs to use the reconstructed value in another coding tree unit above, more implementation complexity will be brought to the hardware.

In this way, in order to reduce the complexity and ensure the prediction accuracy of the sign, for the transformation units in the first row of the coding tree unit, the calculation method of the cost can be reasonably simplified.

In a possible implementation, the performing sign prediction on the last coefficient group and determining the sign prediction value of the last coefficient group may include:

If the current block is at the upper boundary of the object to which it belongs, then according to the value of the first adjacent pixel corresponding to the left side of the current block, the cost value of the last coefficient group under the combination of various candidate symbols is calculated;

According to the cost values under various candidate symbol combinations, the symbol prediction value corresponding to the last coefficient group is determined.

Here, the first adjacent pixel value may be composed of reference pixel values in two adjacent columns on the left side of the current block.

In another possible implementation manner, the performing sign prediction on the last coefficient group and determining the sign prediction value of the last coefficient group may include:

If the current block is at the left boundary of the object to which it belongs, then calculate the cost value of the last coefficient group under various candidate symbol combinations according to the second adjacent pixel value corresponding to the upper side of the current block;

Here, the second adjacent pixel values may be composed of reference pixel values of two adjacent rows on the upper side of the current block.

It should be noted that, in this embodiment of the present application, the belonging object may include at least one of the following: an image and a coding tree unit. That is to say, no matter whether the current block is at the upper boundary of the image or the upper boundary of the coding tree unit, at this time, the last coefficient group can be calculated according to the value of the first adjacent pixel corresponding to the left side of the current block among multiple candidate The cost value under the symbol combination; or, whether the current block is at the left boundary of the image or the left boundary of the coding tree unit, at this time, the last coefficient can be calculated according to the second adjacent pixel value corresponding to the upper side of the current block The cost value of the group under various candidate symbol combinations. It should be noted that the multiple candidate symbol combinations here may be determined based on different symbol prediction modes.

It should also be noted that, in the embodiment of the present application, the premise of realizing the technical solution of the embodiment of the present application is to meet some of the following limitations:

(1) The current sequence allows the use of sign prediction techniques;

(2) Sign prediction acts on a certain block area in the current block, for example, the N×M area at the upper left corner, where M and N are positive integers.

Specifically, for the current block in the first row of the coding tree unit, the current block needs to use the reconstructed value in the upper coding tree unit. In order to reduce the line cache, the calculation of the upper side cost can be omitted; similar to the cost of processing the edge of the image, the calculation of the reconstruction value in the upper coding tree unit is omitted, as shown below,

For the current block in the first column of the coding tree unit, the current block needs to use the reconstructed value in the left coding tree unit. In order to reduce the column cache, the calculation of the left cost can also be omitted; similar to the cost of processing the edge of the image, the calculation of the reconstruction value in the left coding tree unit is omitted, as shown below,

In this way, taking the current block at the upper boundary of the coding tree unit as an example, the full intra-frame coding configuration is used for testing under common test conditions, and the codec performance changes after this method (skipping the upper cost calculation) are shown in Figure 12. According to FIG. 12 , it can be obtained that under such simplification, the influence on the performance of encoding and decoding is not obvious.

Furthermore, in the implementation of ECM, the symbol prediction needs to refer to the reconstruction values of the two rows and two columns above and on the left, but in the above-mentioned embodiment, it is proposed that only the left The side way can reduce the complexity. In addition, an alternative solution is also proposed here, which is to modify the use of the upper two rows to only use the upper row, while still using two columns on the left.

Therefore, in some embodiments, performing sign prediction on the last coefficient group and determining the sign prediction value of the last coefficient group may include:

If the current block is at the upper boundary of the object to which it belongs, calculate the last coefficient group in multiple The cost value under the combination of candidate symbols;

It should be noted that the first adjacent pixel value may be composed of the reference pixel values of two columns adjacent to the left side of the current block, and the second adjacent pixel value may be composed of the reference pixel values of the upper side of the current block adjacent to one row Composed of reference pixel values. In this case, the cost calculation is as follows,

It should also be noted that if the current block is at the left boundary of the object to which it belongs, it may also refer to the upper two rows and the left column. Therefore, in some embodiments, performing sign prediction on the last coefficient group and determining the sign prediction value of the last coefficient group may include:

If the current block is at the left boundary of the object to which it belongs, calculate the last coefficient group in multiple The cost value under the combination of candidate symbols;

It should be noted that the first adjacent pixel value may be composed of reference pixel values adjacent to the left side of the current block in one column, and the second adjacent pixel value may be composed of reference pixel values adjacent to the upper side of the current block in two rows. Composed of reference pixel values. In this case, the cost calculation is as follows,

It should also be noted that, in this embodiment of the present application, the belonging object may include at least one of the following: an image and a coding tree unit. In addition, for the current block at the left boundary or upper boundary of the image or coding tree unit, the embodiment of the present application can even use the reference pixel values of a column adjacent to the left side of the current block and the reference pixel values of a row adjacent to the upper side of the current block. The cost value is calculated with reference to the pixel value, which is not specifically limited here.

It should also be noted that, after determining the cost values under various candidate symbol combinations, determining the symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations may include:

Determining the minimum cost value from the cost values under a variety of candidate symbol combinations;

According to the candidate symbol combination corresponding to the minimum cost value, the symbol prediction value corresponding to the last coefficient group is determined.

That is to say, assuming that K coefficients are selected from the last coefficient group to predict the sign, 2 ^K possible combinations of signs and negatives can be obtained; then according to the pre-calculated template, 2 ^K kinds of transform blocks are calculated The hypothetical inverse transformation residual value of one row and the first column; and then using the inverse transformation residual value and predicted value of these 2 ^K hypotheses of the transformation block, 2 ^K hypothetical reconstruction values can be obtained, and reconstruction from 2 ^K hypothetical value to select a hypothetical reconstruction value with the smallest cost between the surrounding reconstructed blocks; finally, a set of signs used by the inverse transformation residual value with the smallest cost value is the sign prediction value obtained by sign prediction .

In this way, for the sign prediction of the sign of the transformation coefficient, if the current block is at the upper boundary of the coding tree unit, then the cost calculation method needs to be adjusted, for example, the upper side cost calculation is restricted, and the reconstruction value obtained across the coding tree unit is reduced In the case of , especially take the reconstructed value in the coding tree unit above. The embodiment of the present application proposes a simplified solution: directly skip the upper side cost calculation or only consider using one row.

It can also be understood that for the edge of the image, in the implementation of ECM, the cost calculation always skips the side that has no reconstruction value, for example, as shown in Figure 13, (a) shows that the current block is on the top of the image Border, (b) shows that the current block is at the left border of the image. If the current block is in the first row of the image (that is, the upper boundary), then the cost calculation skips the upper side and only calculates the left side, as shown in the above formula (5); if the current block is in the first column of the image (that is, the left boundary ), then the cost calculation skips the left side and only counts the upper side, as shown in the above formula (6).

For the case of only calculating the cost of one side, in one implementation, the total energy change of the upper side and the left side is still considered, so for the method of selecting non-zero transformation coefficients based on the energy size to predict the sign, the calculation is not reconstructed The energy on that side of the value is meaningless.

Based on this, the embodiment of the present application proposes another implementation manner. Firstly, the transform coefficients whose sign is to be predicted are selected based on the magnitude of the energy on the upper side and the left side of the current block. This way of selecting coefficients needs to satisfy the following restrictions:

(1) Sign prediction allows to predict the signs of no more than K non-zero coefficients in the region at most, for example, in ECM, K=8 for one transformation and K=4 for secondary transformation.

(2) When the non-zero coefficients in the current area exceed the maximum limit, the defined screening method can be used to filter out K coefficients with signs to be predicted.

Specifically, in some embodiments, performing sign prediction on the last coefficient group and determining a sign prediction value of the last coefficient group may include:

If the number of non-zero reconstruction coefficient absolute values in the last coefficient group is greater than K, then select K non-zero reconstruction coefficient absolute values from the last coefficient group; wherein, K is an integer greater than zero;

Using the absolute values of the K non-zero reconstruction coefficients for sign prediction, determine the sign prediction values of the K non-zero absolute values of the reconstruction coefficients.

Further, in some embodiments, the selection of K absolute values of non-zero reconstruction coefficients from the last coefficient group may include:

Scanning the last coefficient group by raster scanning order, and determining the absolute values of the first scanned K non-zero reconstruction coefficients as the selected K non-zero reconstruction coefficient absolute values;

or,

determine the absolute value of the non-zero reconstruction coefficients in the last coefficient group;

Determine the largest K absolute values from the absolute values of the non-zero reconstruction coefficients, and determine the selected K absolute values of the non-zero reconstruction coefficients according to the K absolute values;

or,

determine the energy values of the non-zero reconstruction coefficients in the last coefficient group;

The largest K energy values are determined from the energy values of the non-zero reconstruction coefficients, and the absolute values of K selected non-zero reconstruction coefficients are determined according to the K energy values.

That is to say, for the screening of non-zero coefficients used for sign prediction, the first K values scanned may be determined as the absolute values of the selected K non-zero reconstruction coefficients according to the raster scanning order in the ECM; Alternatively, it is also possible to determine the largest K absolute values according to the absolute value, and then determine the selected K absolute values of non-zero reconstruction coefficients according to the K absolute values; or, it is also possible to determine the largest K absolute values according to the energy value energy values, and then determine K selected absolute values of non-zero reconstruction coefficients according to the K energy values, which are not specifically limited in this embodiment of the present application.

Further, in the process of calculating the energy value, for the current block at the boundary of the coding tree unit or the boundary of the picture, the calculation of the cost may be referred to, and only the energy of one side is used to calculate the solution of skipping. Specifically, in some embodiments, the determining the energy value of the non-zero reconstruction coefficient in the last coefficient group may include:

determining the inverse transform value of the non-zero reconstruction coefficients in the last coefficient group on the left side of the current block and the inverse transform value on the upper side of the current block;

If the current block is at the upper boundary of the object to which it belongs, calculate the energy value of the non-zero reconstruction coefficient according to the inverse transformation value of the non-zero reconstruction coefficient on the left side of the current block;

If the current block is at the left boundary of the object to which it belongs, the energy value of the non-zero reconstruction coefficient is calculated according to the inverse transformation value of the non-zero reconstruction coefficient on the upper side of the current block.

It should be noted that, for the screening of non-zero coefficients used for sign prediction, when energy calculation is used, in ECM, for the current block at the image boundary, the cost is only calculated on the side with adjacent reconstruction blocks. The cost on one side of the image boundary implements skipping. For such cases, it is meaningless to calculate the energy beyond the side of the image boundary and should be skipped. Similarly, for other potential screening methods, calculations on the side beyond the image boundary in the screening method should be removed.

Exemplarily, when the current block is at the upper boundary of the image, the cost calculation on the upper side is skipped. At this time, only the left side is used for energy calculation. The energy calculation is as follows,

When the current block is at the left boundary of the image, the cost calculation on the left side is skipped. At this time, only the upper side is used for energy calculation. The energy calculation is as follows,

Among them, tc _i,j is the transformation coefficient in the area of the upper left corner of 4×4 (i,j represents coordinates), T _i,j (k) is the transformation coefficient at position (i,j) when it is 1, in The k-th inverse transform value in the left edge and upper edge of the current transform block, w and h represent the width and height of the current transform block.

Considering that in the subsequent implementation of ECM, there may be restrictions on the cost calculation of the current block in the first row of the coding tree unit, for example, for the current block of this row, the total cost of the upper side and the left side is no longer calculated at the same time Instead, only the left side is calculated and the upper side is skipped, and the energy screening standard can only refer to the left side instead of the upper side. Therefore, in the case of coding tree unit boundaries, the energy calculation should also skip the energy of a certain side, as shown in the above formulas (9) and (10).

That is to say, when using a certain side cost calculation to skip the solution, the criteria for screening the transformation coefficients of the sign to be predicted also need to be optimized accordingly. For example, when the cost calculation only refers to a certain side, the energy Calculations should also refer to that side only.

In this way, taking the above optimization applied to the current block energy calculation of the image boundary as an example, compared with the energy calculation scheme before optimization, the test is carried out in the full intra-frame coding configuration, and the codec performance changes are shown in Figure 14. According to FIG. 14 , it can be obtained that this optimization can provide a coding performance gain of 0.01% while simplifying the energy calculation of the image boundary.

It can also be understood that in the implementation of ECM, for the current block at the upper left corner of the image, as shown in Figure 15, the cost calculation formula at this time is as follows,

Wherein, default is an initial value, and when the pixel bit depth is n, default=2 ^n-1 . However, for the current block at the edge, since there is no reconstruction value of nearby blocks, such cost calculation is unreasonable.

In the embodiment of the present application, the method may further include: if the current block is at the upper left corner of the object to which it belongs, determining the absolute value of the reconstruction coefficient of the current block without performing sign prediction.

That is to say, since it is unreasonable to set the reference reconstruction value of the current block with no adjacent block at the upper left corner as the default value, the embodiment of the present application may directly skip the sign prediction of the transform coefficient for this block. Compared with the scheme of setting the default value, the test is carried out in the full intra-frame coding configuration, and the codec performance changes are shown in Figure 16. According to FIG. 16 , it can be obtained that if the transform block with no adjacent reference block on the upper left is skipped, a certain encoding performance gain can also be provided.

Further, in some embodiments, after the sign prediction value of the last coefficient group is determined, as shown in FIG. 17 , the method may further include:

S1701: Parse the code stream, and determine the sign residual value of the last coefficient group.

S1702: Determine the original sign value of the last coefficient group according to the sign prediction value of the last coefficient group and the sign residual value of the last coefficient group.

S1703: Determine the reconstruction coefficient of the current block according to the signed original value of the last coefficient group.

It should be noted that, in the embodiment of the present application, the sign residual value of the last coefficient group can be decoded based on the context model.

It should also be noted that, in a specific embodiment, for S1702, the original sign of the last coefficient group is determined according to the sign prediction value of the last coefficient group and the sign residual value of the last coefficient group. The value may include: performing an XOR operation on the predicted value of the sign of the last coefficient group and the residual value of the sign of the last coefficient group to determine the original value of the sign of the last coefficient group.

In another specific embodiment, for S1702, the determining the original sign value of the last coefficient group according to the sign prediction value of the last coefficient group and the sign residual value of the last coefficient group may include: An addition operation is performed on the predicted value according to the sign of the last coefficient group and the sign residual value of the last coefficient group to determine the original sign value of the last coefficient group.

That is to say, the decoder decodes the sign residual value of the sign from the context model, and then obtains the sign original value of the last coefficient group (that is, the real sign) according to the sign residual value and the sign prediction value, The actual positive and negative signs are directly assigned to the reconstruction coefficients, and then subsequent decoding operations (for example, inverse transformation, reconstruction, etc.) are further completed.

The embodiment of the present application provides a decoding method, which determines the absolute value of the reconstruction coefficient of the current block by analyzing the code stream; scans the absolute value of the reconstruction coefficient of the current block according to the preset scanning order, and determines that the current block is in the preset scanning order The last coefficient group of ; perform sign prediction on the last coefficient group, and determine the sign prediction value of the last coefficient group. In this way, the transform coefficients of the current block are scanned according to the preset scanning order, and then the last coefficient group of the current block in the preset scanning order is determined, so as to perform sign prediction on the last coefficient group, which not only allows more shapes Blocks of more sizes can be used for symbol prediction, which expands the applicable scope of symbol prediction technology; and in the process of symbol prediction, for the case where the current block is at the upper boundary or left boundary of the image or coding tree unit, the cost calculation method is also adjusted. Optimization is carried out, and at the same time, adaptive adjustments are made to the energy calculation method for screening the transform coefficients to be predicted, thereby reducing the computational complexity and improving the sign prediction accuracy of the sign.

In another embodiment of the present application, refer to FIG. 18 , which shows the first schematic flowchart of an encoding method provided by the embodiment of the present application. As shown in Figure 18, the method may include:

S1801: Determine the transformation coefficient of the current block.

It should be noted that the encoding method in the embodiment of the present application is applied to an encoder, and specifically may be a sign prediction method, where the sign of the transform coefficient in the current block is mainly predicted.

It should also be noted that, based on the composition structure of the encoder 100 shown in FIG. 8A, the encoding method of the embodiment of the present application is applied to the "entropy encoding unit 115" part of the encoder 100. For the entropy encoding unit 115, it can The residual value of the symbol is entropy coded by using the adaptive binary arithmetic coding mode based on the context model, and then written into the code stream. Wherein, the sign residual value is determined based on the sign original value and the sign prediction value.

It can be understood that, in the implementation of ECM, for the current block, the block size is usually greater than or equal to 4×4. For a current block smaller than 4×4, since there is no 4×4 area in the upper left corner for sign prediction (Sign Prediction), such a current block is prohibited from using sign prediction. At the same time, sign prediction is also prohibited for intra blocks larger than 32×32 and inter blocks larger than 128×128. However, in the embodiment of the present application, by determining the last coefficient group in the preset scanning order and then using sign prediction for the last coefficient group, the sign prediction can be applied to blocks of more shapes and sizes, extending The scope of applicability of sign prediction technology.

S1802: Scan the transform coefficients of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order.

S1803: Perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.

It should be noted that, in the embodiment of the present application, it may be allowed to select non-zero transform coefficients in the last coefficient group in the preset scanning order to use sign prediction. In this way, firstly, the transform coefficients of the current block can be scanned according to the preset scanning order to determine the last coefficient group of the current block in the preset scanning order; then the symbol of the last coefficient group can be predicted to determine the sign of the last coefficient group Prediction value, so that blocks with more shapes and sizes can be used for symbol prediction, which expands the applicable range of symbol prediction technology.

If the size of the current block is 1×N, and N is an integer greater than or equal to 16, it is determined that the transform coefficient of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient group is 1×16;

If the size of the current block is N×1, and N is an integer greater than or equal to 16, it is determined that the transform coefficient of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient group is 16×1;

If the size of the current block is 2×N, and N is an integer greater than or equal to 8, it is determined that the transform coefficient of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient group is 2×8;

If the size of the current block is N×2, and N is an integer greater than or equal to 8, it is determined that the transform coefficients of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient group is 8×2;

If the size of the current block is M×N, and both M and N are integers greater than or equal to 4, it is determined that the transform coefficients of the current block can be divided into (M×N/16) coefficient groups, and the size of the coefficient groups is 4 ×4.

It should be noted that for the size of the current block, 1×N and 2×N blocks may also appear, and the shapes of these blocks and the corresponding coefficient group sizes are shown in Table 1 above.

In the related art, for a side whose length is less than 4, ECM currently does not allow such a block to use sign prediction. However, in the embodiment of the present application, it is allowed to use sign prediction for non-zero transform coefficients in all regions or partial regions in the block (for example, the same as the 4×4 region at the upper left corner position in ECM).

Exemplarily, for a current block larger than 4×4 in the ECM, the non-zero transform coefficients in the 4×4 area (coefficient group at the upper left position) in the upper left corner position allow the use of sign prediction; similarly, for the current block smaller than 4×4 For the current block, it is also allowed to select non-zero transform coefficients in the last coefficient group in the preset scanning order to use sign prediction. Fig. 10 shows the encoding effect of the ECM reference software under full intra-frame encoding when the last coefficient group in the preset scanning order is allowed to use sign prediction for a block smaller than 4×4 provided by the embodiment of the present application. Raise sign. Among them, under the three components of Y component, U component and V component under the YUV color space, the performance improvement of 0.02%, 0.1% and 0.07% was brought respectively, and the encoding time (expressed by EncT) and decoding time (expressed by DecT said) little changed.

In addition, in practical applications, the closer to the upper left corner, the later in the preset scanning order, that is, the last coefficient group is located at the upper left corner of the current block. Therefore, the embodiment of the present application may also cancel the restriction on the maximum size of a block using the sign prediction technology. Exemplarily, in the ECM, the current intra block larger than 32×32 and the inter block larger than 128×128 cannot use sign prediction, but actually the maximum size block is currently allowed to be 256×256. So canceling the block maximum size restriction means allowing up to 256×256 transformation blocks in part (for example, the same as in the ECM 4×4 region for the upper left corner position) or the sign of non-zero coefficient signs in all regions predict.

Determine the first area in the upper left corner of the current block;

Determine the last coefficient group according to the transformation coefficients in the first area; wherein, the number of transformation coefficients in the last coefficient group is 2 ^L , and L is a number greater than or equal to zero.

It should be noted that, generally, if the size of the current block is greater than or equal to 4×4, then the 4×4 area at the upper left corner can be directly used as the last coefficient group. In addition, considering that the number of non-zero transform coefficients used for symbol prediction is usually K, if the number of non-zero transform coefficients in the 4×4 area is less than K, the upper left corner area can be appropriately expanded at this time, for example, the upper left corner position The 8×8 area, 16×16 area, etc. of the last coefficient group are used as the last coefficient group, so that the number of transformation coefficients in the last coefficient group is ^2L . In the embodiment of the present application, the first area may be a 4×4 area, but it is not specifically limited here.

It can also be understood that, in order to reduce the complexity and ensure the prediction accuracy of the sign, for the transformation units in the first row of the coding tree unit, the calculation method of the cost can be reasonably simplified. In a possible implementation manner, as shown in FIG. 19, for S1803, this step may include:

S1901: If the current block is at the upper boundary of the object to which it belongs, calculate the cost value of the last coefficient group under various candidate symbol combinations according to the first adjacent pixel value corresponding to the left side of the current block.

S1902: Determine the minimum cost value from the cost values under various candidate symbol combinations.

S1903: Determine the symbol prediction value corresponding to the last coefficient group according to the candidate symbol combination corresponding to the minimum cost value.

In the embodiment of the present application, the first adjacent pixel value may be composed of reference pixel values in two adjacent columns on the left side of the current block.

That is to say, if the current block is at the upper boundary of the object to which it belongs, then calculate the cost value of the last coefficient group under various candidate symbol combinations according to the first adjacent pixel value corresponding to the left side of the current block; The cost value under the candidate symbol combination determines the symbol prediction value corresponding to the last coefficient group. Here, the first adjacent pixel value may be composed of reference pixel values in two adjacent columns on the left side of the current block.

It should also be noted that the premise of the implementation of the technical solution of the embodiment of the present application is that the following restrictions are met: (1) the current sequence allows the use of symbol prediction technology; (2) the symbol prediction acts on a certain block area in the current block, for example The N×M area at the upper left corner, where M and N are positive integers.

Specifically, for the current block in the first row of the coding tree unit, the current block needs to use the reconstructed value in the upper coding tree unit. In order to reduce the line cache, the calculation of the upper side cost can be omitted; similar to the cost of processing the edge of the image, the calculation of the reconstruction value in the upper coding tree unit is omitted, as shown in the above formula (5). For the current block in the first column of the coding tree unit, the current block needs to use the reconstructed value in the left coding tree unit. In order to reduce the column cache, the calculation of the left cost can also be omitted; similar to the cost of processing the edge of the image, the calculation of the reconstruction value in the left coding tree unit is omitted, as shown in the above formula (6).

In this way, taking the current block at the upper boundary of the coding tree unit as an example, the full intra-frame coding configuration is used for testing under common test conditions, and the codec performance changes after this method (skipping the cost calculation on the upper side) are shown in Figure 12 . It can be obtained from this that under such simplification, the impact on the encoding and decoding performance is not obvious.

Furthermore, in the implementation of ECM, symbol prediction needs to refer to the reconstructed values of the two rows and two columns above and on the left. In the above-mentioned embodiment, it is proposed that only the left way to reduce complexity. An alternative would be to change from using the top two rows to just the top row, while still using two columns to the left.

In yet another possible implementation manner, the performing sign prediction on the last coefficient group and determining the sign prediction value of the last coefficient group may include:

Here, the first adjacent pixel value may be composed of reference pixel values adjacent to two columns on the left side of the current block, and the second adjacent pixel value may be composed of reference pixel values adjacent to one row on the upper side of the current block composed of values. In this case, the cost calculation is shown in Equation (7) above.

It should also be noted that if the current block is at the left boundary of the object to which it belongs, it may also refer to the upper two rows and the left column. Therefore, in yet another possible implementation manner, performing sign prediction on the last coefficient group and determining a sign prediction value of the last coefficient group may include:

Here, the first adjacent pixel value may be composed of reference pixel values adjacent to one column on the left side of the current block, and the second adjacent pixel value may be composed of reference pixel values adjacent to two rows on the upper side of the current block composed of values. In this case, the cost calculation is shown in equation (8) above.

That is to say, assuming that K transform coefficients are selected from the last coefficient group to predict the sign, 2 ^K possible sign combinations can be obtained; then calculate 2 ^K transform blocks in the first row and first column The hypothetical inverse transformation residual value of the transformation block, and then using the 2 ^K hypothetical inverse transformation residual values and predicted values of the transformation block, 2 ^K hypothetical reconstruction values can be obtained, and one of the hypothetical reconstruction values is selected from the 2 ^K hypothetical reconstruction values. The hypothetical reconstruction value with the least cost between the surrounding reconstructed blocks; finally, the set of signs used for the inverse transformation residual value with the least cost value is the sign prediction value obtained by sign prediction.

In this way, for the sign prediction of the sign of the transformation coefficient, if the current block is at the upper boundary of the coding tree unit, then the cost calculation method needs to be adjusted, for example, the upper side cost calculation is restricted, and the reconstruction value obtained across the coding tree unit is reduced In the case of , especially take the reconstructed value in the coding tree unit above. The embodiment of the present application proposes a simplified solution: directly skip the upper side cost calculation or only consider using one line.

It can also be understood that for image edges, in the implementation of ECM, the cost calculation always skips the side that has no reconstruction value. For the case of only calculating the cost of one side, in one implementation, the total energy change of the upper side and the left side is still considered, so for the method of selecting non-zero transformation coefficients based on the energy size to predict the sign, the calculation is not reconstructed The energy on that side of the value is meaningless.

Based on this, the embodiment of the present application proposes another implementation manner. Firstly, the transform coefficients whose sign is to be predicted are selected based on the magnitude of the energy on the upper side and the left side of the current block. This way of selecting coefficients needs to meet the following restrictions: (1) sign prediction allows to predict the signs of no more than K non-zero coefficients in this area at most, for example, in ECM, for one transformation K=8, the second transformation K=4. (2) When the non-zero coefficients in the current area exceed the maximum limit, the defined screening method can be used to filter out K coefficients with signs to be predicted.

Specifically, in some embodiments, performing sign prediction on the last coefficient group and determining the sign prediction value of the last coefficient group may include:

If the number of non-zero transformation coefficients in the last coefficient group is greater than K, then select K non-zero transformation coefficients from the last coefficient group; wherein, K is an integer greater than zero;

The K non-zero transform coefficients are used to perform sign prediction, and the sign prediction values of the K non-zero transform coefficients are determined.

Further, in some embodiments, the selection of K non-zero transform coefficients from the last coefficient group may include:

Scanning the last coefficient group by raster scanning order, and determining the first scanned K non-zero transformation coefficients as the selected K non-zero transformation coefficients;

or,

determine the absolute value of the non-zero transform coefficients in the last coefficient group;

Determine the largest K absolute values from the absolute values of the non-zero transformation coefficients, and determine the selected K non-zero transformation coefficients according to the K absolute values;

or,

determining the energy values of the non-zero transform coefficients in the last coefficient group;

The largest K energy values are determined from the energy values of the non-zero transformation coefficients, and K selected non-zero transformation coefficients are determined according to the K energy values.

That is to say, the screening of non-zero transform coefficients used for sign prediction may be to determine the first K values scanned as the selected K non-zero transform coefficients according to the raster scanning order in the ECM; or , or determine the largest K absolute values according to the absolute value, and then determine the selected K non-zero transformation coefficients according to the K absolute values; or, determine the largest K energy values according to the energy value, Then, K selected non-zero transformation coefficients are determined according to the K energy values, which is not specifically limited in this embodiment of the present application.

Further, in the process of calculating the energy value, for the current block at the boundary of the coding tree unit or the boundary of the picture, the calculation of the cost may be referred to, and only the energy of one side is used to calculate the solution of skipping. Specifically, in some embodiments, the determining the energy value of the non-zero transform coefficient in the last coefficient group may include:

determining the inverse transform value of the non-zero transform coefficients in the last coefficient group on the left side of the current block and the inverse transform value on the upper side of the current block;

If the current block is at the upper boundary of the object to which it belongs, calculate the energy value of the non-zero transform coefficient according to the inverse transform value of the non-zero transform coefficient on the left side of the current block;

If the current block is at the left boundary of the object to which it belongs, the energy value of the non-zero transform coefficient is calculated according to the inverse transform value of the non-zero transform coefficient on the upper side of the current block.

Exemplarily, when the current block is at the upper boundary of the image, the upper side cost calculation is skipped, and only the left side is used for energy calculation at this time, and the energy calculation is shown in the above formula (9). When the current block is at the left boundary of the image, the cost calculation on the left side is skipped. At this time, only the upper side is used for energy calculation, and the energy calculation is shown in the above formula (10).

Considering that in the subsequent implementation of ECM, there may be restrictions on the cost calculation of the current block in the first row of the coding tree unit, for example, for the current block of this row, the total cost of the upper side and the left side is no longer calculated at the same time Instead, only the left side is calculated and the upper side is skipped, and the energy screening standard can only refer to the left side instead of the upper side. Therefore, in the case of coding tree unit boundaries, the energy calculation should also skip the energy of a certain side, as shown in the above formulas (9) and (10). That is to say, when using a certain side cost calculation to skip the solution, the criteria for screening the transformation coefficients of the sign to be predicted also need to be optimized accordingly. For example, when the cost calculation only refers to a certain side, the energy Calculations should also refer to that side only.

In this way, taking the above optimization applied to the current block energy calculation of the image boundary as an example, compared with the energy calculation scheme before optimization, the test is carried out in the full intra-frame coding configuration, and the codec performance changes are shown in Figure 14. It can be obtained from this that this optimization can provide a coding performance gain of 0.01% while simplifying the energy calculation of the image boundary.

In addition, in some embodiments, the method may further include: if the current block is at the upper left corner of the object to which it belongs, determining that the transform coefficients of the current block do not perform sign prediction.

That is to say, in the implementation of ECM, for the current block at the upper left corner of the image (as shown in Figure 15), it is unreasonable to set it as the default value because the current block at the upper left corner without adjacent blocks refers to the reconstruction value , then in this embodiment of the present application, the sign prediction of the transform coefficient for this block can be directly skipped. Compared with the scheme of setting the default value, the test is carried out in the full intra-frame coding configuration, and the codec performance changes are shown in Figure 16. It can be obtained from this that if the transform block with no adjacent reference block on the upper left is skipped, a certain coding performance gain can also be provided.

Further, in some embodiments, after the sign prediction value of the last coefficient group is determined, as shown in FIG. 20 , the method may further include:

S2001: Determine the sign residual value of the last coefficient group according to the sign prediction value of the last coefficient group.

S2002: Encode the symbol residual value, and write the obtained encoded bits into the code stream.

It should be noted that, in some embodiments, the determining the sign residual value of the last coefficient group according to the sign prediction value of the last coefficient group may include:

Determine the signed original value of the last coefficient group according to the transform coefficients of the current block;

A signed residual value of the last coefficient group is determined based on the signed original value of the last coefficient group and the signed predicted value of the last coefficient group.

In a specific embodiment, the determining the sign residual value of the last coefficient group according to the sign original value of the last coefficient group and the sign predicted value of the last coefficient group may include: The signed original value and the signed predicted value of the last coefficient group are XORed to determine the signed residual value of the last coefficient group.

In another specific embodiment, the determining the sign residual value of the last coefficient group according to the sign original value of the last coefficient group and the sign predicted value of the last coefficient group may include: for the last coefficient group Subtract the signed original value of , and the signed predicted value of the last coefficient group to determine the signed residual value of the last coefficient group.

It should also be noted that, in the embodiment of the present application, encoding the sign residual value may be based on a context model.

That is to say, in the encoder, firstly, the positive and negative values of the optimal quantization coefficient in the current mode selected by the current block through rate-distortion optimization are used as the original value of the sign, and then it can be determined according to the original value of the sign and the obtained predicted value of the sign The signed residual value of the sign is extracted, and then the signed residual value of the sign is encoded using the context model.

In short, in the embodiment of the present application, on the one hand, the range of use of the transform block predicted using the sign of the transform coefficient can be extended, allowing blocks of more shapes and sizes to use this technology, specifically, for example, less than This technology can be used for 4×4 transformation blocks, transformation blocks larger than 32×32 in a frame, and larger than 128×128 in an interframe. On the other hand, for existing transform coefficient sign prediction schemes, when the transform block is at the upper boundary of the coding tree unit, the cost calculation method should be adjusted, such as limiting the cost calculation on the upper side, and reducing the cost across coding tree units. In the case of taking reconstructed values, especially those in the upper coding tree unit; here a simplified solution is proposed: directly skip the upper side cost calculation or consider only using the upper side row. On the other hand, when using a certain side cost calculation skip solution, the scheme of screening the transformation coefficient of the sign to be predicted also needs to be adjusted accordingly. For example, when the calculation of the cost only refers to a certain side, the calculation of the energy Also only that side should be referenced etc.

The embodiment of the present application also provides an encoding method, by determining the transformation coefficient of the current block; scanning the transformation coefficient of the current block according to the preset scanning order, and determining the last coefficient group of the current block in the preset scanning order; The last coefficient group performs sign prediction, and the sign prediction value of the last coefficient group is determined. In this way, the transform coefficients of the current block are scanned according to the preset scanning order, and then the last coefficient group of the current block in the preset scanning order is determined, so as to perform sign prediction on the last coefficient group, which not only allows more shapes Blocks of more sizes can be used for symbol prediction, which expands the applicable scope of symbol prediction technology; and in the process of symbol prediction, for the case where the current block is at the upper boundary or left boundary of the image or coding tree unit, the cost calculation method is also adjusted. Optimization is carried out, and at the same time, adaptive adjustments are made to the energy calculation method for screening the transform coefficients to be predicted, thereby reducing the computational complexity and improving the sign prediction accuracy of the sign.

In yet another embodiment of the present application, the embodiment of the present application provides a code stream, which is generated by performing bit coding according to the information to be encoded; wherein, the information to be encoded may include at least one of the following: reconstruction coefficient absolute value and signed residual value.

In this way, in the embodiment of the present application, the encoder can transmit the absolute value of the reconstruction coefficient and the sign residual value to the decoder through the code stream, and then scan the absolute value of the reconstruction coefficient of the current block in the decoder according to the preset scanning order , and then determine the last coefficient group of the current block in the preset scanning order, so as to perform sign prediction on the last coefficient group, which not only allows more shapes and sizes of blocks to perform sign prediction, but also expands the sign prediction technology Scope of application; and in the sign prediction process, in view of the current block at the upper boundary or left boundary of the image or coding tree unit, the cost calculation method is also optimized, and at the same time, the energy of the positive and negative transform coefficients to be predicted is screened The calculation method is also adjusted adaptively, thereby reducing the calculation complexity and improving the sign prediction accuracy of the positive and negative signs.

In yet another embodiment of the present application, based on the same inventive concept as the foregoing embodiments, refer to FIG. 21 , which shows a schematic structural diagram of an encoder 210 provided in the embodiment of the present application. As shown in Figure 21, the encoder 210 may include: a first determination unit 2101 and a first prediction unit 2102; wherein,

The first determining unit 2101 is configured to determine the transform coefficient of the current block;

The first determining unit 2101 is further configured to scan the transform coefficients of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

The first prediction unit 2102 is configured to perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.

In some embodiments, referring to FIG. 21 , the encoder 210 may further include an encoding unit 2103;

The first determining unit 2101 is further configured to determine the sign residual value of the last coefficient group according to the sign prediction value of the last coefficient group;

The encoding unit 2103 is configured to encode the symbol residual value, and write the obtained encoded bits into a code stream.

In some embodiments, the first determining unit 2101 is further configured to determine the original sign value of the last coefficient group according to the transform coefficient of the current block; and determine the original sign value of the last coefficient group according to the original sign value of the last coefficient group and the The sign prediction value of the last coefficient group is determined, and the sign residual value of the last coefficient group is determined.

In some embodiments, the first determining unit 2101 is further configured to perform an XOR operation on the original sign value of the last coefficient group and the predicted sign value of the last coefficient group to determine the sign residual value of the last coefficient group.

In some embodiments, if the size of the current block is 1×N, and N is an integer greater than or equal to 16, it is determined that the transform coefficients of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient groups is 1×16; if the size of the current block is N×1, and N is an integer greater than or equal to 16, it is determined that the transform coefficient of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient group is 16× 1; if the size of the current block is 2×N, and N is an integer greater than or equal to 8, then it is determined that the transform coefficients of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient group is 2×8; If the size of the current block is N×2, and N is an integer greater than or equal to 8, it is determined that the transform coefficients of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient group is 8×2; if the current The size of the block is M×N, and both M and N are integers greater than or equal to 4, then it is determined that the transform coefficients of the current block can be divided into (M×N/16) coefficient groups, and the size of the coefficient group is 4×4 .

In some embodiments, when the size of the current block is M×N, and both M and N are integers greater than or equal to 4, the first determining unit 2101 is further configured to determine the first region; and determine the last coefficient group according to the transformation coefficients in the first region; wherein, the number of transformation coefficients in the last coefficient group is 2 ^L , and L is the number greater than or equal to zero.

In some embodiments, the first prediction unit 2102 is further configured to calculate the last coefficient group in multiple The cost value under the candidate symbol combination; and according to the cost value under the various candidate symbol combinations, determine the symbol prediction value corresponding to the last coefficient group; wherein, the first adjacent pixel value is determined by the two adjacent left sides of the current block Column reference pixel values.

In some embodiments, the first prediction unit 2102 is further configured to calculate the last coefficient group in multiple The cost value under the combination of candidate symbols; and according to the cost values under various candidate symbol combinations, determine the symbol prediction value corresponding to the last coefficient group; wherein, the second adjacent pixel value is determined by the upper side of the current block. The reference pixel value of the row.

In some embodiments, the first prediction unit 2102 is further configured to: if the current block is at the upper boundary of the object to which it belongs, according to the value of the first adjacent pixel corresponding to the left side of the current block and the value corresponding to the upper side of the current block For the second adjacent pixel value, calculate the cost value of the last coefficient group under multiple candidate symbol combinations; and determine the symbol prediction value corresponding to the last coefficient group according to the cost values under multiple candidate symbol combinations; wherein, the first The adjacent pixel values are composed of reference pixel values in two adjacent columns on the left side of the current block, and the second adjacent pixel value is composed of reference pixel values in a row adjacent to the upper side of the current block.

In some embodiments, the belonging object includes at least one of: a picture and a coding tree unit.

In some embodiments, the first determination unit 2101 is further configured to determine the minimum cost value from the cost values under various candidate symbol combinations; and determine the symbol corresponding to the last coefficient group according to the candidate symbol combination corresponding to the minimum cost value Predictive value.

In some embodiments, the first prediction unit 2102 is further configured to select K non-zero transform coefficients from the last coefficient group if the number of non-zero transform coefficients in the last coefficient group is greater than K; where K is an integer greater than zero; and perform sign prediction using K non-zero transform coefficients, and determine sign prediction values of the K non-zero transform coefficients.

In some embodiments, the first determining unit 2101 is further configured to scan the last coefficient group in a raster scan order, and determine the first scanned K non-zero transform coefficients as K non-zero transform coefficients; or, Determine the absolute values of the non-zero transform coefficients in the last coefficient group, determine the largest K absolute values from the absolute values of the non-zero transform coefficients, determine the K non-zero transform coefficients according to the K absolute values; or, determine the last coefficient The energy values of the non-zero transform coefficients in the group, the largest K energy values are determined from the energy values of the non-zero transform coefficients, and the K non-zero transform coefficients are determined according to the K energy values.

In some embodiments, the first determining unit 2101 is further configured to determine the inverse transform value of the non-zero transform coefficients in the last coefficient group on the left side of the current block and the inverse transform value on the upper side of the current block; and If the current block is on the upper boundary of the object to which it belongs, calculate the energy value of the non-zero transform coefficient according to the inverse transform value of the non-zero transform coefficient on the left side of the current block; The inverse transform value of the zero transform coefficient on the upper side of the current block, calculates the energy value of the non-zero transform coefficient.

In some embodiments, the first determining unit 2101 is further configured to determine that the transform coefficients of the current block do not perform sign prediction if the current block is at the upper left corner of the object to which it belongs.

It can be understood that in the embodiments of the present application, a "unit" may be a part of a circuit, a part of a processor, a part of a program or software, etc., of course it may also be a module, or it may be non-modular. Moreover, each component in this embodiment may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software function modules.

If the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment is essentially or It is said that the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product, the computer software product is stored in a storage medium, and includes several instructions to make a computer device (which can It is a personal computer, a server, or a network device, etc.) or a processor (processor) that executes all or part of the steps of the method described in this embodiment. The aforementioned storage medium includes: U disk, mobile hard disk, read only memory (Read Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other various media that can store program codes.

Therefore, an embodiment of the present application provides a computer storage medium, which is applied to the encoder 210, and the computer storage medium stores a computer program, and when the computer program is executed by the first processor, it implements any one of the preceding embodiments. Methods.

Based on the above composition of the encoder 210 and the computer storage medium, refer to FIG. 22 , which shows a schematic diagram of a specific hardware structure of the encoder 210 provided by the embodiment of the present application. As shown in FIG. 22 , the encoder 210 may include: a first communication interface 2201 , a first memory 2202 and a first processor 2203 ; each component is coupled together through a first bus system 2204 . It can be understood that the first bus system 2204 is used to realize connection and communication between these components. In addition to the data bus, the first bus system 2204 also includes a power bus, a control bus and a status signal bus. However, for clarity of illustration, the various buses are labeled as first bus system 2204 in FIG. 22 . in,

The first communication interface 2201 is used for receiving and sending signals during the process of sending and receiving information with other external network elements;

The first memory 2202 is used to store computer programs that can run on the first processor 2203;

The first processor 2203 is configured to, when running the computer program, execute:

determining transform coefficients for the current block;

Scan the transform coefficients of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

Sign prediction is performed on the last coefficient group, and a sign prediction value of the last coefficient group is determined.

It can be understood that the first memory 2202 in the embodiment of the present application may be a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memories. Among them, the non-volatile memory can be read-only memory (Read-Only Memory, ROM), programmable read-only memory (Programmable ROM, PROM), erasable programmable read-only memory (Erasable PROM, EPROM), electronically programmable Erase Programmable Read-Only Memory (Electrically EPROM, EEPROM) or Flash. The volatile memory can be Random Access Memory (RAM), which acts as external cache memory. By way of illustration and not limitation, many forms of RAM are available, such as Static Random Access Memory (Static RAM, SRAM), Dynamic Random Access Memory (Dynamic RAM, DRAM), Synchronous Dynamic Random Access Memory (Synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (Double Data Rate SDRAM, DDRSDRAM), enhanced synchronous dynamic random access memory (Enhanced SDRAM, ESDRAM), synchronous connection dynamic random access memory (Synchlink DRAM, SLDRAM) And Direct Memory Bus Random Access Memory (Direct Rambus RAM, DRRAM). The first memory 2202 of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

The first processor 2203 may be an integrated circuit chip, which has a signal processing capability. In the implementation process, each step of the above method may be completed by an integrated logic circuit of hardware in the first processor 2203 or instructions in the form of software. The above-mentioned first processor 2203 may be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a ready-made programmable gate array (Field Programmable Gate Array, FPGA) Or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. Various methods, steps, and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like. The steps of the method disclosed in connection with the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, register. The storage medium is located in the first memory 2202, and the first processor 2203 reads the information in the first memory 2202, and completes the steps of the above method in combination with its hardware.

It should be understood that the embodiments described in this application may be implemented by hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application specific integrated circuits (Application Specific Integrated Circuits, ASIC), digital signal processor (Digital Signal Processing, DSP), digital signal processing device (DSP Device, DSPD), programmable Logic device (Programmable Logic Device, PLD), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA), general-purpose processor, controller, microcontroller, microprocessor, other devices used to perform the functions described in this application electronic unit or its combination. For software implementation, the techniques described herein can be implemented through modules (eg, procedures, functions, and so on) that perform the functions described herein. Software codes can be stored in memory and executed by a processor. Memory can be implemented within the processor or external to the processor.

Optionally, as another embodiment, the first processor 2203 is further configured to execute the method described in any one of the foregoing embodiments when running the computer program.

This embodiment provides an encoder, and the encoder may include a first determination unit and a first prediction unit. In this way, the transform coefficients of the current block are scanned according to the preset scanning order, and the last coefficient group of the current block in the preset scanning order is determined, so as to perform sign prediction on the last coefficient group, which not only allows more shapes and more Multi-size blocks can be used for sign prediction, which expands the scope of application of sign prediction technology; and in the process of sign prediction, for the case where the current block is at the upper boundary or left boundary of the image or coding tree unit, the cost calculation method is also adjusted. Optimization, and at the same time, adaptive adjustments are made to the energy calculation method for screening the transform coefficients to be predicted, which also reduces the computational complexity and improves the sign prediction accuracy of the sign.

In yet another embodiment of the present application, based on the same inventive concept as the preceding embodiments, refer to FIG. 23 , which shows a schematic diagram of the composition and structure of a decoder 230 provided in the embodiment of the present application. As shown in FIG. 23, the decoder 230 may include: a decoding unit 2301, a second determination unit 2302, and a second prediction unit 2303; wherein,

The decoding unit 2301 is configured to analyze the code stream and determine the absolute value of the reconstruction coefficient of the current block;

The second determining unit 2302 is configured to scan the absolute value of the reconstruction coefficient of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

The second prediction unit 2303 is configured to perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.

In some embodiments, the decoding unit 2301 is further configured to analyze the code stream, determine the sign residual value of the last coefficient group; and determine according to the sign prediction value of the last coefficient group and the sign residual value of the last coefficient group The signed original value of the last coefficient group; and determining the reconstruction coefficient of the current block according to the signed original value of the last coefficient group.

In some embodiments, the second determining unit 2302 is further configured to perform an XOR operation on the predicted sign value of the last coefficient group and the sign residual value of the last coefficient group to determine the original sign value of the last coefficient group.

In some embodiments, the second determining unit 2302 is further configured to determine that the absolute value of the reconstruction coefficient of the current block can be divided into (N/16 ) coefficient groups, and the size of the coefficient group is 1×16; if the size of the current block is N×1, and N is an integer greater than or equal to 16, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/16 ) coefficient groups, and the size of the coefficient group is 16×1; if the size of the current block is 2×N, and N is an integer greater than or equal to 8, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/8 ) coefficient groups, and the size of the coefficient group is 2×8; if the size of the current block is N×2, and N is an integer greater than or equal to 8, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/8 ) coefficient groups, and the size of the coefficient group is 8×2; if the size of the current block is M×N, and both M and N are integers greater than or equal to 4, it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into ( M×N/16) coefficient groups, and the size of the coefficient group is 4×4.

In some embodiments, when the size of the current block is M×N, and both M and N are integers greater than or equal to 4, the second determining unit 2302 is further configured to determine the first area; and determine the last coefficient group according to the absolute value of the reconstruction coefficient in the first area; wherein, the number of absolute values of the reconstruction coefficient in the last coefficient group is 2 ^L , and L is a number greater than or equal to zero.

In some embodiments, the second prediction unit 2303 is further configured to calculate the last coefficient group in multiple The cost value under the candidate symbol combination; and according to the cost value under the various candidate symbol combinations, determine the symbol prediction value corresponding to the last coefficient group; wherein, the first adjacent pixel value is determined by the two adjacent left sides of the current block Column reference pixel values.

In some embodiments, the second prediction unit 2303 is further configured to calculate the last coefficient group in multiple The cost value under the combination of candidate symbols; and according to the cost values under various candidate symbol combinations, determine the symbol prediction value corresponding to the last coefficient group; wherein, the second adjacent pixel value is determined by the upper side of the current block. The reference pixel value of the row.

In some embodiments, the second prediction unit 2303 is further configured to: if the current block is at the upper boundary of the object to which it belongs, according to the value of the first adjacent pixel corresponding to the left side of the current block and the value corresponding to the upper side of the current block For the second adjacent pixel value, calculate the cost value of the last coefficient group under multiple candidate symbol combinations; and determine the symbol prediction value corresponding to the last coefficient group according to the cost values under multiple candidate symbol combinations; wherein, the first The adjacent pixel values are composed of reference pixel values in two adjacent columns on the left side of the current block, and the second adjacent pixel value is composed of reference pixel values in a row adjacent to the upper side of the current block.

In some embodiments, the second determination unit 2302 is further configured to determine the minimum cost value from the cost values under various candidate symbol combinations; and determine the symbol corresponding to the last coefficient group according to the candidate symbol combination corresponding to the minimum cost value Predictive value.

In some embodiments, the second prediction unit 2303 is further configured to select K absolute values of non-zero reconstruction coefficients from the last coefficient group if the number of absolute values of non-zero reconstruction coefficients in the last coefficient group is greater than K; Wherein, K is an integer greater than zero; and performing sign prediction by using the absolute values of the K non-zero reconstruction coefficients, and determining the sign prediction values of the K absolute values of the non-zero reconstruction coefficients.

In some embodiments, the second determining unit 2302 is further configured to scan the last coefficient group in raster scan order, and determine the first scanned absolute values of the K non-zero reconstruction coefficients as the K non-zero reconstruction coefficient absolute values value; or, determine the absolute value of the non-zero reconstruction coefficient in the last coefficient group; determine the largest K absolute values from the absolute values of the non-zero reconstruction coefficients, and determine the K absolute values of the non-zero reconstruction coefficients according to the K absolute values; Or, determine the energy values of the non-zero reconstruction coefficients in the last coefficient group, determine the largest K energy values from the energy values of the non-zero reconstruction coefficients, and determine the K absolute values of the non-zero reconstruction coefficients according to the K energy values.

In some embodiments, the second determining unit 2302 is further configured to determine the inverse transform value of the non-zero reconstruction coefficients in the last coefficient group on the left side of the current block and the inverse transform value on the upper side of the current block; and If the current block is at the upper boundary of the object, calculate the energy value of the non-zero reconstruction coefficient according to the inverse transformation value of the non-zero reconstruction coefficient on the left side of the current block; if the current block is at the left boundary of the object, calculate the energy value of the non-zero reconstruction coefficient according to the non-zero reconstruction coefficient The inverse transform value of the zero reconstruction coefficient on the upper side of the current block, calculates the energy value of the non-zero reconstruction coefficient.

In some embodiments, the second determination unit 2302 is further configured to determine the absolute value of the reconstruction coefficient of the current block without performing sign prediction if the current block is at the upper left corner of the object to which it belongs.

It can be understood that, in this embodiment, a "unit" may be a part of a circuit, a part of a processor, a part of a program or software, etc., of course it may also be a module, or it may be non-modular. Moreover, each component in this embodiment may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software function modules.

If the integrated units are implemented in the form of software function modules and are not sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, this embodiment provides a computer storage medium, which is applied to the decoder 230, and the computer storage medium stores a computer program, and when the computer program is executed by the second processor, any one of the preceding embodiments is implemented. the method described.

Based on the above composition of the decoder 230 and the computer storage medium, refer to FIG. 24 , which shows a schematic diagram of a specific hardware structure of the decoder 230 provided by the embodiment of the present application. As shown in FIG. 24 , the decoder 230 may include: a second communication interface 2401 , a second memory 2402 and a second processor 2403 ; each component is coupled together through a second bus system 2404 . It can be understood that the second bus system 2404 is used to realize connection and communication between these components. In addition to the data bus, the second bus system 2404 also includes a power bus, a control bus and a status signal bus. However, for clarity of illustration, the various buses are labeled as the second bus system 2404 in FIG. 24 . in,

The second communication interface 2401 is used for receiving and sending signals during the process of sending and receiving information with other external network elements;

The second memory 2402 is used to store computer programs that can run on the second processor 2403;

The second processor 2403 is configured to, when running the computer program, execute:

Optionally, as another embodiment, the second processor 2403 is further configured to execute the method described in any one of the foregoing embodiments when running the computer program.

It can be understood that the hardware function of the second memory 2402 is similar to that of the first memory 2202, and the hardware function of the second processor 2403 is similar to that of the first processor 2203; details will not be described here.

This embodiment provides a decoder, which may include a decoding unit, a second determining unit, and a second predicting unit. In this way, the transform coefficients of the current block are scanned according to the preset scanning order, and the last coefficient group of the current block in the preset scanning order is determined, so as to perform sign prediction on the last coefficient group, which not only allows more shapes and more Multi-size blocks can be used for sign prediction, which expands the scope of application of sign prediction technology; and in the process of sign prediction, for the case where the current block is at the upper boundary or left boundary of the image or coding tree unit, the cost calculation method is also adjusted. Optimization, and at the same time, adaptive adjustments are made to the energy calculation method for screening the transform coefficients to be predicted, which also reduces the computational complexity and improves the sign prediction accuracy of the sign.

In yet another embodiment of the present application, refer to FIG. 25 , which shows a schematic diagram of the composition and structure of a codec system provided by the embodiment of the present application. As shown in FIG. 25 , the codec system 250 may include an encoding device 2501 and a decoding device 2502 . Wherein, the encoding device 2501 may be the encoder described in any one of the foregoing embodiments, and the decoding device 2502 may be the decoder described in any one of the foregoing embodiments.

In the embodiment of the present application, the codec system 250 can scan the transform coefficients of the current block according to the preset scanning order, and then determine the last coefficient group of the current block in the preset scanning order, so that the last coefficient group Perform sign prediction, which not only allows blocks of more shapes and sizes to perform sign prediction, expanding the scope of application of sign prediction technology; but also in the process of sign prediction, for the current block at the upper boundary of the image or coding tree unit or In the case of the left boundary, the cost calculation method is also optimized, and the energy calculation method for screening the transform coefficients to be predicted is also adaptively adjusted, thereby reducing the computational complexity and improving the sign prediction of the sign Accuracy.

It should be noted that in this application, the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article or apparatus comprising a set of elements includes not only those elements , but also includes other elements not expressly listed, or also includes elements inherent in such a process, method, article, or device. Without further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus comprising that element.

The serial numbers of the above embodiments of the present application are for description only, and do not represent the advantages and disadvantages of the embodiments.

The methods disclosed in several method embodiments provided in this application can be combined arbitrarily to obtain new method embodiments under the condition of no conflict.

The features disclosed in several product embodiments provided in this application can be combined arbitrarily without conflict to obtain new product embodiments.

The features disclosed in several method or device embodiments provided in this application can be combined arbitrarily without conflict to obtain new method embodiments or device embodiments.

The above is only a specific implementation of the application, but the scope of protection of the application is not limited thereto. Anyone familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the application. Should be covered within the protection scope of this application. Therefore, the protection scope of the present application should be determined by the protection scope of the claims.

Industrial Applicability

In the embodiment of the present application, at the coding end, the transformation coefficient of the current block is determined; the transformation coefficient of the current block is scanned according to the preset scanning order, and the last coefficient group of the current block in the preset scanning order is determined; for the last coefficient sign prediction for the last coefficient group, determining the sign prediction value for the last coefficient group. At the decoding end, analyze the code stream to determine the absolute value of the reconstruction coefficient of the current block; scan the absolute value of the reconstruction coefficient of the current block according to the preset scanning order, and determine the last coefficient group of the current block in the preset scanning order; Sign prediction is performed for one coefficient group, and the sign prediction value for the last coefficient group is determined. In this way, both the encoding end and the decoding end can scan the transform coefficients of the current block according to the preset scanning order, and then determine the last coefficient group of the current block in the preset scanning order, so as to perform Sign prediction, which not only allows blocks of more shapes and sizes to perform sign prediction, but also expands the scope of application of sign prediction technology; and in the process of sign prediction, for the current block at the upper boundary or left of the image or coding tree unit For boundary conditions, the cost calculation method is also optimized, and at the same time, the energy calculation method for screening the transform coefficients to be predicted is also adaptively adjusted, thereby reducing the computational complexity and improving the accuracy of sign prediction Spend.

Claims

A decoding method applied to a decoder, the method comprising:

Analyze the code stream to determine the absolute value of the reconstruction coefficient of the current block;

Scan the absolute value of the reconstruction coefficient of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

Sign prediction is performed on the last coefficient group, and a sign prediction value of the last coefficient group is determined.
The method according to claim 1, wherein the method further comprises:

Parsing the code stream to determine the sign residual value of the last coefficient group;

determining an original sign value of the last coefficient group based on the sign prediction value of the last coefficient group and the sign residual value of the last coefficient group;

Determining the reconstruction coefficients of the current block according to the signed original value of the last coefficient group.
The method according to claim 2, wherein said determining the original sign value of the last coefficient group according to the sign prediction value of the last coefficient group and the sign residual value of the last coefficient group comprises :

Exclusive OR operation is performed on the predicted value according to the sign of the last coefficient group and the sign residual value of the last coefficient group to determine the original sign value of the last coefficient group.
The method according to claim 1, wherein the method further comprises:

If the size of the current block is 1×N, and N is an integer greater than or equal to 16, it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/16) coefficient groups, and the coefficient groups of the Dimensions are 1×16;

If the size of the current block is N×1, and N is an integer greater than or equal to 16, it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/16) coefficient groups, and the coefficient groups of the Dimensions are 16×1;

If the size of the current block is 2×N, and N is an integer greater than or equal to 8, it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/8) coefficient groups, and the coefficient groups of the Dimensions are 2×8;

If the size of the current block is N×2, and N is an integer greater than or equal to 8, it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (N/8) coefficient groups, and the coefficient groups of the Dimensions are 8×2;

If the size of the current block is M×N, and both M and N are integers greater than or equal to 4, then it is determined that the absolute value of the reconstruction coefficient of the current block can be divided into (M×N/16) coefficient groups, and The size of the coefficient group is 4x4.
The method according to claim 4, wherein, when the size of the current block is M×N, and both M and N are integers greater than or equal to 4, the method further comprises:

determining the first area in the upper left corner of the current block;

The last coefficient group is determined according to the absolute values of reconstruction coefficients in the first area; wherein, the number of absolute values of reconstruction coefficients in the last coefficient group is 2L, and L is a number greater than or equal to zero.
The method according to claim 1, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the current block is at the upper boundary of the object to which it belongs, calculating the cost value of the last coefficient group under various candidate symbol combinations according to the first adjacent pixel value corresponding to the left side of the current block;

determining a symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations;

Wherein, the first adjacent pixel value is composed of reference pixel values of two adjacent columns on the left side of the current block.
The method according to claim 1, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the current block is at the left boundary of the object to which it belongs, calculating the cost value of the last coefficient group under various candidate symbol combinations according to the second adjacent pixel value corresponding to the upper side of the current block;

determining a symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations;

Wherein, the second adjacent pixel values are composed of reference pixel values of two adjacent rows on the upper side of the current block.
The method according to claim 1, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the current block is at the upper boundary of the object to which it belongs, the calculated value is calculated according to the first adjacent pixel value corresponding to the left side of the current block and the second adjacent pixel value corresponding to the upper side of the current block. Describe the cost value of the last coefficient group under various candidate symbol combinations;

determining a symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations;

Wherein, the first adjacent pixel value is composed of the reference pixel values of the two columns adjacent to the left side of the current block, and the second adjacent pixel value is composed of the upper side of the current block. It is composed of the reference pixel values of the adjacent row.
The method according to any one of claims 6 to 8, wherein the belonging object includes at least one of the following: a picture and a coding tree unit.
The method according to any one of claims 6 to 8, wherein, according to the cost values under the various candidate symbol combinations, determining the symbol prediction value corresponding to the last coefficient group includes:

determining a minimum cost value from the cost values under the plurality of candidate symbol combinations;

A symbol prediction value corresponding to the last coefficient group is determined according to the candidate symbol combination corresponding to the minimum cost value.
The method according to claim 1, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the number of absolute values of non-zero reconstruction coefficients in the last coefficient group is greater than K, then select K absolute values of non-zero reconstruction coefficients from the last coefficient group; wherein, K is an integer greater than zero;

Sign prediction is performed by using the K absolute values of non-zero reconstruction coefficients, and a sign prediction value of the K absolute values of non-zero reconstruction coefficients is determined.
The method according to claim 11, wherein said selecting K absolute values of non-zero reconstruction coefficients from said last coefficient group comprises:

Scanning the last coefficient group in a raster scanning order, and determining the absolute values of K non-zero reconstruction coefficients scanned first as the absolute values of the K non-zero reconstruction coefficients;

or,

determining the absolute value of the non-zero reconstruction coefficients in said last coefficient group;

determining the largest K absolute values from the absolute values of the non-zero reconstruction coefficients, and determining the K absolute values of the K non-zero reconstruction coefficients according to the K absolute values;

or,

determining energy values for non-zero reconstruction coefficients in said last coefficient group;

The largest K energy values are determined from the energy values of the non-zero reconstruction coefficients, and the absolute values of the K non-zero reconstruction coefficients are determined according to the K energy values.
The method of claim 12, wherein said determining the energy values of the non-zero reconstruction coefficients in said last coefficient group comprises:

determining an inverse transform value of the non-zero reconstruction coefficients in the last coefficient group on the left side of the current block and an inverse transform value on the upper side of the current block;

If the current block is at the upper boundary of the object to which it belongs, calculate the energy value of the non-zero reconstruction coefficient according to the inverse transformation value of the non-zero reconstruction coefficient on the left side of the current block;

If the current block is at the left boundary of the object to which it belongs, calculating an energy value of the non-zero reconstruction coefficient according to an inverse transform value of the non-zero reconstruction coefficient on the upper side of the current block.
The method according to claim 1, wherein the method further comprises:

If the current block is at the upper left corner of the object to which it belongs, it is determined that the absolute value of the reconstruction coefficient of the current block does not perform sign prediction.
An encoding method applied to an encoder, the method comprising:

determining transform coefficients for the current block;

Scan the transform coefficients of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

Sign prediction is performed on the last coefficient group, and a sign prediction value of the last coefficient group is determined.
The method according to claim 15, wherein said method further comprises:

determining a sign residual value of the last coefficient group based on the sign prediction value of the last coefficient group;

Encoding the symbol residual value, and writing the obtained coded bits into a code stream.
The method according to claim 16, wherein said determining the sign residual value of the last coefficient group according to the sign prediction value of the last coefficient group comprises:

Determining the signed original value of the last coefficient group according to the transform coefficient of the current block;

A sign residual value of the last coefficient group is determined based on the sign original value of the last coefficient group and the sign predicted value of the last coefficient group.
The method according to claim 17, wherein said determining the sign residual value of the last coefficient group according to the sign original value of the last coefficient group and the sign predicted value of the last coefficient group comprises :

Performing an XOR operation on the original sign value of the last coefficient group and the predicted sign value of the last coefficient group to determine a sign residual value of the last coefficient group.
The method according to claim 15, wherein said method further comprises:

If the size of the current block is 1×N, and N is an integer greater than or equal to 16, then it is determined that the transform coefficients of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient groups is 1×16;

If the size of the current block is N×1, and N is an integer greater than or equal to 16, then it is determined that the transform coefficients of the current block can be divided into (N/16) coefficient groups, and the size of the coefficient groups is 16×1;

If the size of the current block is 2×N, and N is an integer greater than or equal to 8, it is determined that the transform coefficients of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient groups is 2×8;

If the size of the current block is N×2, and N is an integer greater than or equal to 8, it is determined that the transform coefficients of the current block can be divided into (N/8) coefficient groups, and the size of the coefficient groups is 8×2;

If the size of the current block is M×N, and both M and N are integers greater than or equal to 4, it is determined that the transform coefficients of the current block can be divided into (M×N/16) coefficient groups, and the The size of the coefficient group is 4×4.
The method according to claim 19, wherein, when the size of the current block is M×N, and both M and N are integers greater than or equal to 4, the method further comprises:

determining the first area in the upper left corner of the current block;

The last coefficient group is determined according to the transformation coefficients in the first area; wherein, the number of transformation coefficients in the last coefficient group is 2 L , and L is a number greater than or equal to zero.
The method according to claim 15, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the current block is at the upper boundary of the object to which it belongs, calculating the cost value of the last coefficient group under various candidate symbol combinations according to the first adjacent pixel value corresponding to the left side of the current block;

determining a symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations;

Wherein, the first adjacent pixel value is composed of reference pixel values of two adjacent columns on the left side of the current block.
The method according to claim 15, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the current block is at the left boundary of the object to which it belongs, calculating the cost value of the last coefficient group under various candidate symbol combinations according to the second adjacent pixel value corresponding to the upper side of the current block;

determining a symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations;

Wherein, the second adjacent pixel values are composed of reference pixel values of two adjacent rows on the upper side of the current block.
The method according to claim 15, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the current block is at the upper boundary of the object to which it belongs, the calculated value is calculated according to the first adjacent pixel value corresponding to the left side of the current block and the second adjacent pixel value corresponding to the upper side of the current block. Describe the cost value of the last coefficient group under various candidate symbol combinations;

determining a symbol prediction value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations;

Wherein, the first adjacent pixel value is composed of the reference pixel values of the two columns adjacent to the left side of the current block, and the second adjacent pixel value is composed of the upper side of the current block. It is composed of the reference pixel values of the adjacent row.
The method according to any one of claims 21 to 23, wherein the belonging object includes at least one of the following: a picture and a coding tree unit.
The method according to any one of claims 21 to 23, wherein the determining the predicted symbol value corresponding to the last coefficient group according to the cost values under the various candidate symbol combinations includes:

determining a minimum cost value from the cost values under the plurality of candidate symbol combinations;

A symbol prediction value corresponding to the last coefficient group is determined according to the candidate symbol combination corresponding to the minimum cost value.
The method according to claim 1, wherein said performing sign prediction on said last coefficient group and determining a sign prediction value of said last coefficient group comprises:

If the number of non-zero transform coefficients in the last coefficient group is greater than K, select K non-zero transform coefficients from the last coefficient group; where K is an integer greater than zero;

Perform sign prediction by using the K non-zero transform coefficients, and determine sign prediction values of the K non-zero transform coefficients.
The method according to claim 26, wherein said selecting K non-zero transform coefficients from said last coefficient group comprises:

Scanning the last coefficient group in a raster scanning order, and determining the first scanned K non-zero transform coefficients as the K non-zero transform coefficients;

or,

determining the absolute values of the non-zero transform coefficients in said last coefficient group;

determining the largest K absolute values from the absolute values of the non-zero transform coefficients, and determining the K non-zero transform coefficients according to the K absolute values;

or,

determining energy values for non-zero transform coefficients in said last coefficient group;

The largest K energy values are determined from the energy values of the non-zero transform coefficients, and the K non-zero transform coefficients are determined according to the K energy values.
The method of claim 27, wherein said determining energy values of non-zero transform coefficients in said last coefficient group comprises:

determining an inverse transform value of the non-zero transform coefficients in the last coefficient group on the left side of the current block and an inverse transform value on the upper side of the current block;

If the current block is at the upper boundary of the object to which it belongs, calculate the energy value of the non-zero transform coefficient according to the inverse transform value of the non-zero transform coefficient on the left side of the current block;

If the current block is at the left boundary of the object to which it belongs, calculating an energy value of the non-zero transform coefficient according to an inverse transform value of the non-zero transform coefficient on the upper side of the current block.
The method according to claim 15, wherein said method further comprises:

If the current block is at the upper left corner of the object to which it belongs, it is determined that no sign prediction is performed on the transform coefficients of the current block.
A bit stream is generated by performing bit coding according to the information to be encoded; wherein the information to be encoded includes at least one of the following: an absolute value of a reconstruction coefficient and a symbol residual value.
An encoder, comprising a first determination unit and a first prediction unit; wherein,

The first determining unit is configured to determine a transform coefficient of a current block;

The first determining unit is further configured to scan the transform coefficients of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

The first prediction unit is configured to perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.
An encoder comprising a first memory and a first processor; wherein,

The first memory is used to store a computer program capable of running on the first processor;

The first processor is configured to execute the method according to any one of claims 15 to 29 when running the computer program.
A decoder, including a decoding unit, a second determination unit, and a second prediction unit; wherein,

The decoding unit is configured to analyze the code stream and determine the absolute value of the reconstruction coefficient of the current block;

The second determining unit is configured to scan the absolute value of the reconstruction coefficient of the current block according to a preset scanning order, and determine the last coefficient group of the current block in the preset scanning order;

The first prediction unit is configured to perform sign prediction on the last coefficient group, and determine a sign prediction value of the last coefficient group.
A decoder comprising a second memory and a second processor; wherein,

The second memory is used to store a computer program capable of running on the second processor;

The second processor is configured to execute the method according to any one of claims 1 to 14 when running the computer program.
A computer storage medium, wherein the computer storage medium stores a computer program, and when the computer program is executed, the method according to any one of claims 1 to 14 is realized, or the method according to any one of claims 15 to 29 is realized. method described in the item.