Video decoding method and device for performing residual error prediction post-processing through switching of multiple schemes

Abstract

The invention provides setting of options of coding parameters and / or coding modes at least comprising a prediction mode and / or a transformation mode and / or a quantization mode of data. According to the data compression method and device, residual errors of one-to-one whole compression units of multiple sets of preset residual error coding schemes including entropy coding pre-processing are correspondingly selected for coding.

Description

[0001] This application is a divisional application of the following original application: Filing date of original application: 2020-03-01 Application number of the original application: 202010133384X The name of the invention of the original application: Data compression method and device for predictive residual entropy coding by switching multiple schemes Technical Field

[0002] The present invention relates to a coding and decoding (also called decoding) system for compressing data, and in particular to a method and device for entropy coding and decoding prediction residuals of data. Background Art

[0003] As human society enters the era of big data, artificial intelligence, cloud computing, mobile computing, cloud-mobile computing, ultra-high definition (4K) and ultra-high definition (8K) video image resolution, 4G / 5G communications, and virtual reality, ultra-high compression ratio and extremely high quality data compression for various data, including big data, image data, video data, audio data, voice data, neural network data, and various new forms of data, has become an indispensable technology. A new trend in the field of data compression is the diversification and differentiation of data, such as the increasing dynamic range and bit accuracy of its representation, and the increasing requirements for data compression, such as lossless, ultra-low distortion lossy, and low distortion lossy coding requirements.

[0004] A data set is a collection of data elements (e.g., bytes, bits, pixels, pixel components, spatial sampling points, transform domain coefficients). When encoding (and correspondingly decoding) a data set arranged in a certain spatial (one-dimensional, two-dimensional, or multi-dimensional) shape (e.g., a one-dimensional data queue, a two-dimensional data file, a frame of image, a video sequence, a transform domain, a transform block, multiple transform blocks, a three-dimensional scene, a sequence of continuously changing three-dimensional scenes), especially a two-dimensional or higher data set, the data set is usually divided into a number of whole compression units with a predetermined shape and size (i.e., the number of elements), and the whole compression units are used as units, and the encoding or decoding is performed one by one. At any time, the whole compression unit being encoded or decoded is called the current whole compression unit. The data element (sometimes also referred to as the element) being encoded or decoded is called the current data element, or simply the current element. An element consists of N components (usually 1≤N≤5), so both the data set and the whole compression unit also consist of N components.

[0005] Encoding in data compression usually consists of at least some or all of the following stages: 1) Prediction, which mainly includes adjacent element (such as pixel) prediction, string prediction, block prediction, etc., to generate prediction residual, referred to as residual; 2) Transformation, which mainly transforms the prediction residual to generate transformation coefficients, referred to as coefficients; when the transformation is an identity transformation, that is, no transformation is actually performed, the coefficients are completely equal to the residual; 3) Quantization, mainly quantizing the coefficients to produce quantized residuals; when the quantization is identical quantization, that is, no quantization is actually performed, the quantized residual is equal to the coefficient; when the transformation is an identical transformation, the quantized residual is the result of quantizing the prediction residual; when the quantization is identical quantization and the transformation is an identical transformation, the quantized residual is equal to the coefficient and also equal to the residual; 4) Entropy coding, which mainly includes entropy coding of the quantized residual, including at least binarization, to generate a compressed data code stream.

[0006] Decoding in data compression usually consists of at least some or all of the following stages: 1) Entropy decoding, which mainly involves parsing the compressed data bitstream and performing entropy decoding including at least debinarization to generate quantized residuals; 2) Dequantization, mainly dequantizing the quantization residual to generate reconstruction coefficients; when quantization is identical quantization, that is, no quantization is actually performed, dequantization is also identical dequantization, that is, no dequantization is actually performed, so that the reconstruction coefficient is equal to the quantization residual; 3) Inverse transformation, mainly inverse transformation of the reconstruction coefficients to generate reconstruction residuals; when the transformation is an identity transformation, that is, no transformation is actually performed, the inverse transformation is also an identity transformation, that is, no inverse transformation is actually performed, so that the reconstruction residual is completely equal to the reconstruction coefficient; when quantization is an identity quantization and the transformation is an identity transformation, the inverse quantization is also an identity inverse quantization and the inverse transformation is also an identity inverse transformation, so that the reconstruction residual is equal to the reconstruction coefficient and equal to the quantization residual; 4) Prediction compensation, which mainly includes adjacent element (such as pixel) prediction compensation, string prediction compensation, block prediction compensation, etc. The reconstructed data generated is also called reconstructed data or restored data.

[0007] The prediction residual, transform coefficient, quantization residual in the above encoding stages and the quantization residual, reconstruction coefficient, and reconstruction residual in the decoding stages are often collectively referred to as residual or residual data. Generally, the context is used to determine whether the residual is a prediction residual, a transform coefficient, a quantization residual, a reconstruction coefficient, or a reconstruction residual. If it cannot be determined from the context, the residual refers to the prediction residual and / or the transform coefficient and / or the quantization residual and / or the reconstruction coefficient and / or the reconstruction residual.

[0008] The input data, original data, residual data and various intermediate data involved in each stage of encoding and decoding are called data samples, or samples or sample values ​​for short.

[0009] Both binarization and debinarization are about the one-to-one correspondence between the value Val and the binary symbol string representing the value Val, namely the binary codeword Code. This one-to-one correspondence is usually represented by a code table or an equivalent calculation formula or an equivalent pseudocode. The following are several commonly used binarization and debinarization methods and the code tables representing their corresponding relationships: 1) With the maximum value X = 2 n The corresponding relationship between the value Val and the codeword Code is shown in the following code table:

[0010] 2) Unary code, the correspondence between its value Val and codeword Code is shown in the following code table:

[0011] 3) The correspondence between the truncated unary code with the maximum value X and the codeword Code is shown in the following code table:

[0012] 4) The corresponding relationship between the value Val and the codeword Code of k-order exponential Golomb code is shown in the following code table: When k=0

[0013] When k = 1

[0014] When k = 2

[0015] Obviously, binarization is to obtain the corresponding codeword Code from the value Val, and debinarization is to obtain the corresponding value Val from the codeword Code.

[0016] In the prior art, no matter what prediction method is used, what transformation method is used, whether the transformation is an identical transformation, what quantization method is used, whether the quantization is an identical quantization, what is the dynamic range of the data, how many bits the data has a bit accuracy, what is the coding distortion requirement, and how to set the coding distortion control parameters, it is basically a quantization residual. When the transformation is an identical transformation and the quantization is an identical quantization, it is also a prediction residual, and a single binarization, debinarization, entropy coding and entropy decoding method is used.

[0017] In new application areas of data compression with new characteristics such as increasingly diversified and differentiated data, more diversified, differentiated and complex prediction, transformation and quantization technologies, large differences in data dynamic range and bit accuracy, such as ranging from 8 bits to 16 bits, and very different distortion requirements for data compression, single binarization, debinarization, entropy coding and entropy decoding methods can no longer meet the requirements of these new applications for coding efficiency. Summary of the invention

[0018] In order to further significantly improve the coding efficiency of data compression under new circumstances where data is becoming more diversified and differentiated, prediction, transformation and quantization technologies are also more diversified, differentiated and complicated, the dynamic range and bit accuracy of data vary greatly, and the distortion requirements of data compression are very different, the present invention provides a data compression method and device that selects one of multiple predetermined residual entropy coding schemes including binarization and / or entropy coding pre-processing to entropy code the residual of an entire compression unit according to the setting of coding parameters and / or coding mode options including at least data bit accuracy and / or data compression distortion requirements and / or transformation size and / or quantization parameter and / or prediction mode and / or transformation mode and / or quantization mode.

[0019] The primary technical feature of the present invention is that it has multiple sets of predetermined residual entropy encoding schemes and corresponding residual entropy decoding schemes; according to the setting of encoding parameters and / or encoding mode options including at least data bit accuracy and / or data compression distortion requirements and / or transform size and / or quantization parameter and / or prediction mode and / or transform mode and / or quantization mode, one of the multiple sets of predetermined residual entropy encoding schemes or one of the residual entropy decoding schemes is selected to perform entropy encoding or entropy decoding on the residual of an entire compression unit.

[0020] The most basic unique technical feature of the encoding method and device of the present invention is that there are A (2≤A≤5) sets of predetermined quantized residual entropy coding schemes; a plurality of predetermined coding parameters and / or coding methods, including data bit accuracy and / or data compression distortion requirements and / or transform size and / or quantization parameter and / or prediction mode and / or transform mode and / or quantization mode, each having a plurality of predetermined options; selecting one of the plurality of options of coding parameters and / or coding methods as a corresponding option corresponding to an entire compression unit; using the corresponding option to perform a coding operation on the entire compression unit including at least prediction and / or transform and / or quantization; according to a predetermined rule, determining a corresponding quantized residual entropy coding scheme from the corresponding option, using the corresponding quantized residual entropy coding scheme to perform entropy coding on the quantized residual of the entire compression unit, and generating a compressed data code stream containing at least part or all of the information required for the options representing the coding parameters and / or coding methods and / or the quantized residual entropy coding scheme. Figure 1It is a schematic diagram of the encoding method or device of the present invention.

[0021] The most basic unique technical feature of the decoding method and device of the present invention is that they have A (2≤A≤5) sets of predetermined quantized residual entropy decoding schemes; a plurality of predetermined coding parameters and / or decoding methods including data bit accuracy and / or data compression distortion requirements and / or transform size and / or quantization parameter and / or prediction compensation method and / or inverse transform method and / or inverse quantization method each have a plurality of predetermined options; the compressed data bit stream is parsed to obtain the corresponding options of the coding parameters and / or decoding methods used by the entire compression unit; according to predetermined rules, the corresponding quantized residual entropy decoding scheme is determined from the corresponding options, entropy decoding is performed using the corresponding residual entropy decoding scheme, and the decoding operation of at least inverse quantization and / or inverse transform and / or prediction compensation is performed on the entire compression unit using the corresponding options to generate restored data. Figure 2 It is a schematic diagram of the decoding method or device of the present invention.

[0022] According to a first aspect of the present invention, there is provided a coding method for compressing a data set and its data, characterized in that: At least A (2≤A≤5) sets of predetermined quantized residual entropy coding schemes; when quantization is an identical quantization, the quantized residual is equal to the coefficient; when the transform is an identical transform, the quantized residual is the result of quantizing the predicted residual; when the quantization is an identical quantization and the transform is an identical transform, the quantized residual is equal to the predicted residual; A plurality of predetermined encoding parameters and / or encoding modes each have a plurality of predetermined options, wherein the encoding parameters at least include a bit accuracy of data and / or a distortion requirement of data compression and / or a transform size and / or a quantization parameter, and the encoding mode at least includes a prediction mode and / or a transform mode and / or a quantization mode; When encoding an entire compression unit, the steps involved include at least the following steps: 1) selecting one of the multiple options of the encoding parameter and / or the encoding mode as a corresponding option corresponding to the entire compression unit, 2) performing at least prediction and / or transform and / or quantization encoding operations on the entire compression unit using at least the corresponding options, 3) at least according to the corresponding option, according to a predetermined rule, selecting one of the A sets of quantized residual entropy coding schemes as the corresponding residual entropy coding scheme corresponding to the entire compression unit, 4) performing entropy coding pre-processing on the prediction residual of the entire compression unit and / or performing entropy coding on the quantized residual of the entire compression unit using at least the corresponding residual entropy coding scheme, 5) Writing the encoding result into a compressed data bitstream, wherein the compressed data bitstream at least includes part or all of the information required to indicate that the corresponding option is the encoding parameter and / or which option of the encoding method and / or which set of quantized residual entropy encoding scheme the corresponding residual entropy encoding scheme is.

[0023] According to a second aspect of the present invention, there is provided an encoding device for compressing a data set and its data, characterized in that: At least A (2≤A≤5) sets of predetermined quantized residual entropy coding schemes; when quantization is an identical quantization, the quantized residual is equal to the coefficient; when the transform is an identical transform, the quantized residual is the result of quantizing the predicted residual; when the quantization is an identical quantization and the transform is an identical transform, the quantized residual is equal to the predicted residual; A plurality of predetermined encoding parameters and / or encoding modes each have a plurality of predetermined options, wherein the encoding parameters at least include a bit accuracy of data and / or a distortion requirement of data compression and / or a transform size and / or a quantization parameter, and the encoding mode at least includes a prediction mode and / or a transform mode and / or a quantization mode; When encoding a whole compression unit, the modules involved include at least the following modules: A coding parameter and / or coding mode selection module: selects one of the multiple options of the coding parameter and / or the coding mode as a corresponding option corresponding to the entire compression unit, Encoding module: performing at least prediction and / or transformation and / or quantization encoding operations on the entire compression unit using at least the corresponding options, A residual entropy coding scheme selection module: at least according to the corresponding options and according to a predetermined rule, selects one of the A sets of quantized residual entropy coding schemes as the corresponding residual entropy coding scheme corresponding to the entire compression unit, an entropy coding pre-processing and / or entropy coding module: performing entropy coding pre-processing on the prediction residual of the entire compression unit and / or performing entropy coding on the quantized residual of the entire compression unit using at least the corresponding residual entropy coding scheme, A compressed data bitstream generation module: writes the encoding result into a compressed data bitstream, wherein the compressed data bitstream includes at least part or all of the information required to indicate that the corresponding option is the encoding parameter and / or which option of the encoding method and / or which set of quantized residual entropy encoding schemes the corresponding residual entropy encoding scheme is.

[0024] According to a third aspect of the present invention, there is provided a method for compressing a data set and its data, characterized in that: At least A (2≤A≤5) sets of predetermined quantized residual entropy decoding schemes are included; when the inverse quantization is an identical inverse quantization, the reconstruction coefficient is equal to the quantized residual; when the inverse transform is an identical inverse transform, the reconstruction residual is equal to the reconstruction coefficient; when the inverse quantization is an identical inverse quantization and the inverse transform is an identical inverse transform, the reconstruction residual is equal to the reconstruction coefficient and also equal to the quantized residual; A plurality of predetermined encoding parameters and / or decoding modes each have a plurality of predetermined options, wherein the encoding parameters at least include a bit accuracy of data and / or a distortion requirement of data compression and / or a transform size and / or a quantization parameter, and the decoding mode at least includes an inverse quantization mode and / or an inverse transform mode and / or a prediction compensation mode; When decoding an entire compression unit, the steps involved include at least the following steps: 1) parsing the compressed data code stream to obtain a corresponding option of the encoding parameter and / or the decoding method used by the entire compression unit, wherein the corresponding option is one of the multiple options of the encoding parameter and / or the decoding method, 2) at least according to the corresponding option, according to a predetermined rule, selecting one of the A sets of quantized residual entropy decoding schemes as the corresponding residual entropy decoding scheme corresponding to the entire compression unit, 3) performing entropy decoding on the entire compression unit using at least the corresponding residual entropy decoding scheme to obtain at least a quantized residual, and / or performing entropy decoding post-processing on the residual of the entire compression unit to be used for prediction compensation to obtain at least a reconstructed residual, 4) performing at least inverse quantization and / or inverse transformation and / or prediction compensation decoding operations on the entire compression unit using at least the corresponding options.

[0025] According to a fourth aspect of the present invention, there is provided a decoding device for compressing a data set and data thereof, characterized in that: At least A (2≤A≤5) sets of predetermined quantized residual entropy decoding schemes are included; when the inverse quantization is an identical inverse quantization, the reconstruction coefficient is equal to the quantized residual; when the inverse transform is an identical inverse transform, the reconstruction residual is equal to the reconstruction coefficient; when the inverse quantization is an identical inverse quantization and the inverse transform is an identical inverse transform, the reconstruction residual is equal to the reconstruction coefficient and also equal to the quantized residual; A plurality of predetermined encoding parameters and / or decoding modes each have a plurality of predetermined options, wherein the encoding parameters at least include a bit accuracy of data and / or a distortion requirement of data compression and / or a transform size and / or a quantization parameter, and the decoding mode at least includes an inverse quantization mode and / or an inverse transform mode and / or a prediction compensation mode; When decoding a whole compression unit, the modules involved include at least the following modules: A compressed data code stream parsing module: parses the compressed data code stream to obtain a corresponding option of the encoding parameter and / or the decoding method used by the entire compression unit, wherein the corresponding option is one of the multiple options of the encoding parameter and / or the decoding method. A residual entropy decoding scheme selection module: at least according to the corresponding options and according to a predetermined rule, selects one of the A sets of quantized residual entropy decoding schemes as the corresponding residual entropy decoding scheme corresponding to the entire compression unit, an entropy decoding and / or entropy decoding post-processing module: performing entropy decoding on the entire compression unit using at least the corresponding residual entropy decoding scheme to obtain at least a quantized residual, and / or performing entropy decoding post-processing on the residual of the entire compression unit to be used for prediction compensation to obtain at least a reconstructed residual, A decoding module: performs at least inverse quantization and / or inverse transformation and / or prediction compensation decoding operations on the entire compression unit using at least the corresponding options.

[0026] In the present invention, data includes one or a combination of the following types of data: 1) One-dimensional data; 2) Two-dimensional data; 3) Multidimensional data; 4) Graphics; 5) Fractal graphics; 6) Image; 7) Sequence of images; 8) Video; 9) Audio; 10) Documents; 11) Bytes; 12) Bit; 13) Pixels; 14) Three-dimensional scene; 15) Sequences of continuously changing three-dimensional scenes; 16) Virtual reality scenes; 17) Sequence of continuously changing virtual reality scenes 18) Images in pixel form; 19) Transform domain data of the image; 20) A collection of bytes of two or more dimensions; 21) A set of bits of two or more dimensions; 22) A collection of pixels; 23) A set of three-component pixels (R, G, B) or their variants; 24) A set of three-component pixels (Y, U, V) or their variants; 25) A set of three-component pixels (Y, Cb, Cr) or their variants; 26) A set of three-component pixels (Y, Cg, Co) or their variants; 27) a set of four-component pixels (C, M, Y, K) or their variants; 28) A set of four-component pixels (R, G, B, A) or their variants; 29) a set of four-component pixels (Y, U, V, A) or their variants; 30) A set of four-component pixels (Y, Cb, Cr, A) or their variants; 31) A collection of four-component pixels (Y, Cg, Co, A) or their variations.

[0027] In the present invention, when the original data is an image, a sequence of images, a video, etc., the entire compression unit is a coding area or a decoding area of ​​the image or sequence, including the following cases: a sequence, a sequence of images, an image, a sub-image of an image, a slice, a tile, a macroblock, a maximum coding unit LCU, a coding tree unit CTU, a coding unit CU, a sub-area of ​​a CU, a sub-coding unit SubCU, a prediction block, a prediction unit PU, a sub-area of ​​a PU, a sub-prediction unit SubPU, a transform block, a transform unit TU, a sub-area of ​​a TU, and a sub-transform unit SubTU.

[0028] The technical features of the present invention are described above by using several specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the encoding method or device of the present invention.

[0030] Figure 2 It is a schematic diagram of the decoding method or device of the present invention. DETAILED DESCRIPTION

[0031] The following are more implementation details of the present invention.

[0032] Example 1 In the encoding method or device or the decoding method or device, the element of the data set is a byte or a bit or a pixel or N (1≤N≤5) component pixels or a component of N component pixels or a spatial sampling point or a transform domain coefficient.

[0033] Example 2 In the encoding method or device or the decoding method or device, the multiple options of the distortion requirement of the data compression are lossless compression and lossy compression.

[0034] Example 3 In the encoding method or device or the decoding method or device, the multiple options for the bit precision of the data include at least the following options: 8 bit, 10 bit, 12 bit, and 16 bit.

[0035] Example 4 In the encoding method or device or the decoding method or device, the multiple options of the encoding mode or the decoding mode are two options in which the transformation and quantization are both identical and the transformation and quantization are not identical, or the equivalent two options in which the inverse transformation and inverse quantization are both identical and the inverse transformation and inverse quantization are not identical.

[0036] Example 5 In the encoding method or device or the decoding method or device, the A set of quantized residual entropy encoding schemes or entropy decoding schemes include at least the following three schemes: Solution 1, at least including the following features: in entropy coding or entropy decoding, the syntax element coeff_level_minus1 called the absolute value of the non-zero quantized residual minus one is subjected to the first set of binarization methods or the inverse binarization methods; Solution 2, at least including the following features: in entropy coding or entropy decoding, a second set of binarization methods or inverse binarization methods are used for the syntax element coeff_level_minus1 called the absolute value of the non-zero quantized residual minus one; Solution 3 at least includes the following features: in entropy coding or entropy decoding, a third set of binarization methods or inverse binarization methods is used for the syntax element coeff_level_minus1, which is the absolute value of the non-zero quantized residual minus one.

[0037] Example 6 In the encoding method or device or the decoding method or device described in Embodiment 5, The first set of binarization or debinarization methods is: there is a predetermined threshold value K1, if coeff_level_minus1 is less than K1, a truncated unary code with a maximum value of K1-1 is used for binarization or debinarization, otherwise, a 0-order exponential Golomb code is used for binarization or debinarization; The second set of binarization or debinarization methods is: there is a predetermined threshold value K2, if coeff_level_minus1 is less than K2, a truncated unary code with a maximum value of K2-1 is used for binarization or debinarization, otherwise, a 0-order exponential Golomb code is used for binarization or debinarization; The second set of binarization methods or debinarization methods is: there is a predetermined threshold value K3=2 n If coeff_level_minus1 is less than K3, the n-bit specific length code is used for binarization or debinarization, otherwise, the BitDepth-bit specific length code is used for binarization or debinarization, where BitDepth is the bit precision of the data.

[0038] Example 7 In the encoding method or device or the decoding method or device described in Embodiment 6, The various options for the distortion requirement of the data compression are lossless compression and lossy compression. The multiple options of the bit depth of the data include at least the following options: 8 bit, 10 bit, 12 bit, 16 bit, The predetermined rule for selecting the corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme according to the corresponding option is: If the distortion requirement of the data compression is lossy compression, then, The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 1, otherwise, If the bit depth of the data is not 16 bits, then The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 2, otherwise, The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 3.

[0039] Example 8 In the encoding method or device or the decoding method or device described in Embodiment 6, The multiple options of the bit depth of the data include at least the following options: 8 bit, 10 bit, 12 bit, 16 bit, The multiple options of the encoding method or decoding method are two options in which the transformation and quantization are both identical and the transformation and quantization are not both identical, or the two options in which the inverse transformation and inverse quantization are both identical and the inverse transformation and inverse quantization are not both identical, The predetermined rule for selecting the corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme according to the corresponding option is: If the transform and quantization are not identical or equivalently the inverse transform and inverse quantization are not identical, then, The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 1, otherwise, If the bit depth of the data is not 16 bits, then The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 2, otherwise, The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 3.

[0040] Example 9 In the encoding method or device or decoding method or device described in Embodiment 7 or 8, K1=32, K2=2 BitDepth - 3 , K3=2 n =2048.

[0041] Example 10 In the encoding method or device or the decoding method or device, The multiple options of the encoding method or decoding method are the following two options: 1) The entire compression unit uses intra-frame prediction for encoding or decoding, 2) The entire compression unit does not use intra-frame prediction for encoding or decoding; The A set of quantized residual entropy coding schemes or entropy decoding schemes include at least the following two schemes: Solution 1, at least comprising the following features: the entropy coding pre-processing or the entropy decoding post-processing comprises multiplying by a predetermined constant proportional factor a or -a or dividing by a predetermined constant proportional factor a or -a, respectively; Solution 2, at least comprising the following features: the entropy encoding pre-processing or the entropy decoding post-processing comprises multiplying by a predetermined constant proportional factor a or -a or b or -b or dividing by a predetermined constant proportional factor a or -a or b or -b, respectively; Among them, a and b are both positive numbers; The predetermined rule for selecting the corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme according to the corresponding option is: If the entire compression unit is encoded or decoded using intra-frame prediction, then, The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 2, otherwise, The corresponding residual entropy encoding scheme or the corresponding residual entropy decoding scheme is the scheme 1.

[0042] Embodiment 11 In the encoding method or device or decoding method or device described in Example 10, a=1, b=2.

[0043] Example 12 In the encoding method or device or the decoding method or device described in Example 11, the division by 2 operation is a right shift by one binary bit, and the division by -2 operation is a right shift by one binary bit after multiplication by -1.

[0044] Embodiment 13 In the encoding method or device or the decoding method or device described in Embodiment 10, 11 or 12, information indicating the positive or negative sign of the predetermined constant proportional factor to be multiplied or divided exists at one or more of the following places in the compressed data code stream: 1) Image parameter set; 2) Sequence header; 3) Strip head; 4) Image header; 5) CTU header; 6) CU head; 7) Codec block header.

[0045] Embodiment 14 In the encoding method or device or the decoding method or device described in Example 13, the information is one or more directly existing syntax elements and / or one or more implicitly derived parameters; the directly existing syntax elements are one or more bit strings (bit strings) in the compressed data code stream or their predetermined default values; the implicitly derived parameters are parameters derived from one or more other encoding parameters and / or encoding and decoding variables and / or other syntax elements of the compressed data code stream or their predetermined default values.

[0046] Embodiment 15 In the decoding method or device described in Example 13, the information is the following syntax elements in a picture parameter set or a picture header: Residual_multiplier_sign_flag Its semantic description is: a binary variable; a value of '0' indicates that the sign of the predetermined constant scale factor processed after residual entropy decoding is positive; a value of '1' indicates that the sign of the predetermined constant scale factor processed after residual entropy decoding is negative; if the syntax element does not exist in the bitstream, it is equivalent to the value of the syntax element being the default value '0', that is, the sign of the predetermined constant scale factor processed after residual entropy decoding is positive.

[0047] Example 16 In the encoding method or device or the decoding method or device described in Embodiment 10, 11 or 12, information representing the absolute value of the predetermined constant scaling factor to be multiplied or divided exists at one or more of the following places in the compressed data code stream: 1) Image parameter set; 2) Sequence header; 3) Strip head; 4) Image header; 5) CTU header; 6) CU head; 7) Codec block header.

[0048] Embodiment 17 In the encoding method or device or the decoding method or device described in Example 16, the information is one or more directly existing syntax elements and / or one or more implicitly derived parameters; the directly existing syntax elements are one or more bit strings (bit strings) in the compressed data code stream or their predetermined default values; the implicitly derived parameters are parameters derived from one or more other encoding parameters and / or encoding and decoding variables and / or other syntax elements of the compressed data code stream or their predetermined default values.

[0049] Embodiment 18 In the decoding method or device described in Example 16, the information is a parameter ResidualMultiplierMode derived from other coding parameters and / or codec variables and / or other syntax elements of the compressed data code stream: when ResidualMultiplierMode takes a value of 2, the absolute value of the predetermined constant proportional factor for residual entropy decoding post-processing is a; when ResidualMultiplierMode takes a value of 1 and 3, the absolute value of the predetermined constant proportional factor for residual entropy decoding post-processing is b; when ResidualMultiplierMode takes a value of 0, residual entropy decoding post-processing is not performed on the residual, which is equivalent to the value of the predetermined constant proportional factor for residual entropy decoding post-processing being 1.

[0050] Embodiment 19 In the encoding method or device or the decoding method or device, part or all of the information required to indicate that the corresponding option is the encoding parameter and / or which option of the encoding mode or decoding mode and / or which set of quantized residual entropy encoding schemes or entropy decoding schemes the corresponding residual entropy encoding scheme or decoding scheme is exists at one or more of the following places in the compressed data code stream: 1) Sequence parameter set; 2) Image parameter set; 3) Sequence header; 4) Strip head; 5) Image header; 6) CTU header; 7) CU head; 8) Codec block header.

[0051] Embodiment 20 In the encoding method or device or the decoding method or device described in Example 19, the part or all of the information is one or more directly existing syntax elements and / or one or more implicitly derived parameters; the directly existing syntax elements are one or more bit strings (bit strings) in the compressed data code stream or their predetermined default values; the implicitly derived parameters are parameters derived from one or more other encoding parameters and / or encoding and decoding variables and / or other syntax elements of the compressed data code stream or their predetermined default values.

[0052] Embodiment 21 In the encoding method or device or the decoding method or device, the entire compression unit is a sequence or a sequence of images; part or all of the information required to indicate which option of the encoding parameter and / or the encoding method or decoding method the corresponding option is exists in a sequence parameter set or a sequence header of the compressed data code stream.

[0053] Embodiment 22 In the encoding method or device or the decoding method or device, the entire compression unit is a sequence or a sequence of images; part or all of the information required to indicate that the corresponding option is the encoding parameter and / or which option of the encoding method or decoding method is the profile label profile_id present in the sequence parameter set or sequence header of the compressed data code stream.

[0054] Embodiment 23 In the encoding method or device or the decoding method or device described in Example 22, the profile label profile_id represents two options of the distortion requirement of the data compression. If profile_id is equal to a predetermined value K, the distortion requirement of the data compression is lossless compression, otherwise, the distortion requirement of the data compression is lossy compression.

[0055] Embodiment 24 In the encoding method or device or the decoding method or device, the entire compression unit is a CTU or a CU; and part or all of the information required to indicate which option of the corresponding option is the encoding parameter and / or the encoding method or the decoding method is a flag bit cu_transquant_bypass_flag present in the CTU header or the CU header of the compressed data bitstream.

[0056] Embodiment 25 In the encoding method or device or the decoding method or device described in Example 24, the flag cu_transquant_bypass_flag represents two options of the distortion requirement of the data compression. If cu_transquant_bypass_flag is equal to 1, the distortion requirement of the data compression is lossless compression, otherwise, the distortion requirement of the data compression is lossy compression.

Claims

1. A method for compressing a residual of an image, a sequence of images or video data, characterized in that: At least 2 residual decoding schemes are included: Solution 1, at least including the following features: residual decoding post-processing includes multiplying by a predetermined constant scaling factor a or -a, Solution 2, at least including the following features: residual decoding post-processing includes multiplying by a predetermined constant scaling factor a or -a or b or -b, Among them, a=1, b=1 / 2; The multiplication by 1 / 2 is a right shift by one binary bit, and the multiplication by -1 / 2 is a right shift by one binary bit after multiplication by -1. A plurality of predetermined encoding parameters and / or decoding modes each have a predetermined plurality of options; The multiple options of the decoding method include at least: Option 1, intra-frame prediction, Option 2: non-intra prediction; The step of decoding an entire compression unit includes at least the steps of performing the following functions and / or operations: Parse the compressed data code stream to obtain the decoding method options used by the entire compression unit, and select the residual decoding scheme according to at least the following rules: If the entire compression unit uses option 1, then, The residual decoding scheme is the scheme 2, If the entire compression unit uses option 2, then, The residual decoding scheme is the scheme 1; The entire compression unit is subjected to residual decoding post-processing using at least the selected residual decoding scheme.

2. The decoding method according to claim 1, characterized in that: The entire compression unit is an image, a sequence of images, or a decoding area of ​​a video, including the following situations: macroblock, maximum coding unit LCU, coding tree unit CTU, coding unit CU, sub-area of ​​CU, sub-coding unit SubCU, prediction block, prediction unit PU, sub-area of ​​PU, sub-prediction unit SubPU, transform block, transform unit TU, sub-area of ​​TU, ​​sub-transform unit SubTU.

3. The decoding method according to claim 1, characterized in that: Information indicating the sign of the predetermined constant proportional factor to be multiplied exists at one or more of the following places in the compressed data code stream: 1) Image parameter set; 2) Sequence header; 3) Strip head; 4) Image header; 5) CTU header; 6) CU head; 7) Codec block header.

4. The decoding method according to claim 3, characterized in that: The information is one or more directly existing syntax elements and / or one or more implicitly derived parameters; the directly existing syntax elements are one or more bit strings in the compressed data code stream or their predetermined default values; the implicitly derived parameters are parameters derived from one or more other coding parameters and / or codec variables and / or other syntax elements of the compressed data code stream or their predetermined default values.

5. The decoding method according to claim 3, characterized in that: The information is the following syntax elements in a picture parameter set or a picture header: Residual_multiplier_sign_flag Its semantic description is: a binary variable; a value of '0' indicates that the sign of the predetermined constant scale factor of the residual post-decoding processing is positive; a value of '1' indicates that the sign of the predetermined constant scale factor of the residual post-decoding processing is negative; if the syntax element does not exist in the bitstream, it is equivalent to the value of the syntax element being the default value '0', that is, the sign of the predetermined constant scale factor of the residual post-decoding processing is positive.

6. The decoding method according to claim 1, characterized in that: Information indicating the absolute value of the predetermined constant proportional factor to be multiplied exists at one or more of the following places in the compressed data code stream: 1) Image parameter set; 2) Sequence header; 3) Strip head; 4) Image header; 5) CTU header; 6) CU head; 7) Codec block header.

7. The decoding method according to claim 6, characterized in that: The information is a parameter ResidualMultiplierMode derived from other coding parameters and / or codec variables and / or other syntax elements of the compressed data code stream: when ResidualMultiplierMode takes a value of 2, the absolute value of the predetermined constant proportional factor for residual post-decoding processing is 1; when ResidualMultiplierMode takes a value of 1 and 3, the absolute value of the predetermined constant proportional factor for residual post-decoding processing is 1 / 2; when ResidualMultiplierMode takes a value of 0, residual post-decoding processing is not performed on the residual, which is equivalent to the value of the predetermined constant proportional factor for residual post-decoding processing being 1.

8. The decoding method according to claim 1, characterized in that: Part or all of the information required to indicate which option of the encoding parameter and / or the decoding mode the option is and / or which set of residual decoding schemes the residual decoding scheme is exists in one or more of the following places in the compressed data code stream: 1) Sequence parameter set; 2) Image parameter set; 3) Sequence header; 4) Strip head; 5) Image header; 6) CTU header; 7) CU head; 8) Codec block header.

9. The decoding method according to claim 8, characterized in that: The part or all of the information is one or more directly existing syntax elements and / or one or more implicitly derived parameters; the directly existing syntax elements are one or more bit strings in the compressed data code stream or their predetermined default values; the implicitly derived parameters are parameters derived from one or more other encoding parameters and / or decoding variables and / or other syntax elements of the compressed data code stream or their predetermined default values.

10. A decoding device for compressing a residual of an image, a sequence of images, or video data, characterized in that: It comprises a module for implementing a decoding method for compressing an image, a sequence of images, or a residual of video data as described in any one of claims 1 to 9.

Images