Decoding method, encoding method, apparatus and device

Abstract

This disclosure provides a decoding method, an encoding method, an apparatus, and a device that aim to improve chromaticity prediction accuracy and contribute to improved compression efficiency. The decoding method of an embodiment of this disclosure includes receiving encoding information corresponding to a target encoding block; determining a first model and a second model based on the index information and parameter information of the first model and the index information and parameter information of the second model; determining a target chromaticity prediction value for the first pixel through the first model based on the luminance downsampling reconstruction value of the first pixel; determining a target chromaticity prediction value for the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel; and determining a chromaticity reconstruction value for the target encoding block based on the target chromaticity prediction value and target difference value of the first pixel and target chromaticity prediction value and target difference value of the second pixel.

Description

[Technical Field]

[0001] This disclosure claims priority to a Chinese patent application filed with the China National Intellectual Property Office on 12 April 2023, application number 202310390303.8, with the title of the invention being a decoding method, encoding method, apparatus and device. The entire content of that priority claim is incorporated herein by reference. This application belongs to the field of video codec technology, and more particularly to decoding methods, encoding methods, apparatus, and devices. [Background technology]

[0002] In video codecs, there is a certain correlation between luminance and chromaticity. Therefore, to eliminate redundancy between different components of luminance and chromaticity, Versatile Video Coding (VVC) proposes an intra-predictive mode based on a Cross-Component Linear Model (CCLM).

[0003] CCLM assumes that a linear relationship exists between chromaticity pixel values ​​and the corresponding luminance pixel values ​​within the same coding unit (CU). Therefore, CCLM uses a linear model to generate predicted values ​​for the corresponding chromaticity pixels from the reconstructed values ​​of the luminance pixels. However, if a nonlinear relationship (or low linear correlation) exists between luminance pixels and chromaticity pixels within a CU, the prediction accuracy of CCLM decreases, resulting in larger prediction residuals and negatively impacting compression efficiency. [Overview of the project] [Problems that the invention aims to solve]

[0004] The embodiments of this application provide a decoding method, an encoding method, an apparatus, and a device that aim to improve chromaticity prediction accuracy and contribute to improved compression efficiency. [Means for solving the problem]

[0005] In the first embodiment, the embodiments of this application provide a decoding method. The decoding method receives encoding information corresponding to a target encoding block, and the encoding information includes index information and parameter information of a first model, index information and parameter information of a second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, wherein the first pixel is a pixel that satisfies a first condition in the target encoding block, the second pixel is a pixel that does not satisfy the first condition in the target encoding block, and the first condition is by the first model The process involves determining a first model and a second model based on the fact that the pixel chromaticity prediction error is within a first threshold range, the index information and parameter information of the first model, and the index information and parameter information of the second model; determining a chromaticity prediction value mapped from the luminance downsampling reconstruction value by the first model according to a linear mapping relationship, the second model determining a chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship; and determining a target chromaticity prediction value for the first pixel through the first model based on the luminance downsampling reconstruction value of the first pixel. The method includes determining a target chromaticity prediction value for the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel, and determining a chromaticity reconstruction value for the target encode block based on the target chromaticity prediction value and target difference value of the first pixel, and the target chromaticity prediction value and target difference value of the second pixel.

[0006] In a second embodiment, the embodiment of the present application provides an encoding method. The encoding method involves determining a first chromaticity prediction value for a target encoding block through a first model based on the luminance downsampling reconstruction value of the target encoding block, determining a chromaticity prediction value mapped from the luminance downsampling reconstruction value by the first model according to a linear mapping relationship, determining a first pixel from the target encoding block that satisfies a first condition, and determining a second pixel from the target encoding block that does not satisfy the first condition, the first condition being that the chromaticity prediction error of the pixel by the first model is within a first threshold range, determining a target chromaticity prediction value for the first pixel based on the first chromaticity prediction value of the first pixel, and determining a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel. The second model includes determining a predicted chromaticity value mapped from a luminance downsampling reconstruction value according to a nonlinear mapping relationship; determining a target difference value for the first pixel and a target difference value for the second pixel, respectively, based on the target predicted chromaticity value and the true chromaticity value of the first pixel, and the target predicted chromaticity value and the true chromaticity value of the second pixel; and generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value and the target difference value for the first pixel, and the luminance downsampling reconstruction value and the target difference value for the second pixel.

[0007] In a third embodiment, the embodiment of the present application provides a decoding device. The decoding device receives encoding information corresponding to the target encoding block, and the encoding information includes index information and parameter information of the first model, index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, the first pixel is a pixel that satisfies a first condition in the target encoding block, the second pixel is a pixel that does not satisfy the first condition in the target encoding block, and the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range. Based on this, the system includes a model determination module that determines a first model and a second model, the first model determining a chromaticity prediction value mapped from a luminance downsampling reconstruction value according to a linear mapping relationship, and the second model determining a chromaticity prediction value mapped from a luminance downsampling reconstruction value according to a nonlinear mapping relationship; a first chromaticity prediction module that determines a target chromaticity prediction value for the first pixel through the first model based on the luminance downsampling reconstruction value of the first pixel; a second chromaticity prediction module that determines a target chromaticity prediction value for the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel; and a reconstruction module that determines the chromaticity reconstruction value of the target encode block based on the target chromaticity prediction value and target difference value of the first pixel, and the target chromaticity prediction value and target difference value of the second pixel.

[0008] In a fourth embodiment, the embodiment of the present application provides an encoding device. The encoding device includes: a first determination module that determines a first chromaticity prediction value of a target encoding block through a first model based on the luminance downsampling reconstruction value of the target encoding block, and determines a chromaticity prediction value mapped from the luminance downsampling reconstruction value by the first model according to a linear mapping relationship; a second determination module that determines a first pixel from the target encoding block that satisfies a first condition, and a second pixel from the target encoding block that does not satisfy the first condition, wherein the chromaticity prediction error of the pixel by the first model is within a first threshold range; a third determination module that determines a target chromaticity prediction value of the first pixel based on the first chromaticity prediction value of the first pixel; and a second determination module that determines a target chromaticity prediction value of the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel. The second model includes: a fourth decision module that determines a predicted chromaticity value mapped from a luminance downsampling reconstruction value according to a nonlinear mapping relationship; a fifth decision module that determines a target difference value for the first pixel and a target difference value for the second pixel, based on the target predicted chromaticity value and the true chromaticity value of the first pixel, and the target predicted chromaticity value and the true chromaticity value of the second pixel, respectively; and a first generation module that generates encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value and the target difference value of the first pixel, and the luminance downsampling reconstruction value and the target difference value of the second pixel.

[0009] In a fifth embodiment, the embodiment of the present application provides a decoding device. The decoding device comprises a processor and memory, the memory storing a program or instruction executable on the processor, and when the program or instruction is executed by the processor, the decoding method described in the first embodiment is realized.

[0010] In a sixth embodiment, the embodiment of the present application provides an encoding device. The encoding device comprises a processor and memory, the memory storing a program or instruction executable on the processor, and when the program or instruction is executed by the processor, the encoding method described in the second embodiment is realized.

[0011] In a seventh embodiment, the embodiment of the present application provides a computing device. The computing device comprises a memory storing computer-readable code and one or more processors, and when the computer-readable code is executed by the one or more processors, the computing device executes the decoding method described in the first embodiment or the encoding method described in the second embodiment.

[0012] In the eighth embodiment, the embodiment of the present application provides a computer program. The computer program includes computer-readable code, and when the computer-readable code is executed on a computing device, the computing device is instructed to execute the decoding method described in the first embodiment or the encoding method described in the second embodiment.

[0013] In the ninth embodiment, the embodiment of the present application provides a computer-readable medium. The computer-readable medium stored the computer program described in the eighth aspect.

[0014] In a tenth embodiment, the embodiment of this application provides a codec system. The codec system includes an encoding device and a decoding device. The decoding device is configured to perform the decoding method described in the first embodiment, and the encoding device is configured to perform the encoding method described in the second embodiment.

[0015] In an eleventh embodiment, the embodiment of the present application provides a readable storage medium. A program or instructions are stored in a readable storage medium, and when the program or instructions are executed by a processor, the decoding method described in the first aspect or the encoding method described in the second aspect is realized.

[0016] In a twelfth aspect, an embodiment of the present application provides a chip. The chip includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor executes a program or instructions to realize the decoding method described in the first aspect or the encoding method described in the second aspect.

[0017] In a thirteenth aspect, an embodiment of the present application provides a computer program / program product. The computer program / program product is stored in a storage medium and, when executed by at least one processor, realizes the decoding method described in the first aspect or the encoding method described in the second aspect.

Advantages of the Invention

[0018] In the embodiments of the present application, pixels of a target encoding block are classified into two types according to a first condition, and target chromaticity prediction values of the two types of pixels are respectively determined using a first model and a second model, thereby improving the prediction accuracy of the entire target encoding block by an integrated model prediction method and improving the compression efficiency for the target encoding block. The above description is an overview of the technical solution of the present disclosure. In order to understand the technical means more clearly, it may also be implemented according to the content of the specification. In order to make the purpose, features and advantages of the present disclosure more clear, the following embodiments are specifically shown.

[0019] To more clearly illustrate the technical solutions in the embodiments of this disclosure or related technologies, the drawings necessary for describing the embodiments or related technologies are briefly described below. As is evident, the drawings in the following description are some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these without any creative effort. [Brief explanation of the drawing]

[0020] [Figure 1] This is a flowchart of the decoding method in the embodiment of the present invention. [Figure 2] This is a flowchart of the encoding method in the present embodiment. [Figure 3] This is a block diagram of the configuration of the decoding device in the present embodiment. [Figure 4] This is a block diagram of the encoding device configuration in the present embodiment. [Figure 5] This is a schematic block diagram of a computing device that implements the method disclosed herein. [Figure 6] This is a schematic block diagram of a storage unit that holds or carries the program code to implement the method described herein. [Modes for carrying out the invention]

[0021] To further clarify the objectives, technical solutions, and advantages of the embodiments of this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below in conjunction with the drawings of the embodiments of this disclosure. As is obvious, the embodiments described are only a selection of embodiments of this disclosure, not all embodiments. All other embodiments obtained based on the embodiments of this disclosure without the creative work of a person skilled in the art are all within the scope of this disclosure. The terms "first," "second," etc., in the specification and claims of this application are used to distinguish similar subjects and are not intended to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate, thereby allowing the embodiments of this application to be carried out in an order other than that illustrated or described herein. Furthermore, subjects distinguished by "first" and "second" are usually of one class and do not limit the number of subjects; for example, the first subject may be one or at least two. In addition, "and / or" in the specification and claims indicates at least one of the connected subjects, and the letter " / " generally indicates that the preceding and following related subjects are in an "or" relationship.

[0022] In video codecs, there is a certain correlation between luminance and chromaticity. Therefore, to eliminate redundancy between different luminance and chromaticity components, VVC proposes an intra-predictive mode of CCLM based on CCLM. CCLM reduces crossover redundancy by generating predicted values ​​for corresponding chromaticity pixels from reconstructed luminance pixels using a linear model. Chromaticity sampling is calculated by luminance downsampling as follows:

number

number

[0023] CCLM assumes a local texture-linear correlation between luminance and chromaticity. However, linear models may not always accurately represent the precise relationship between luminance and chromaticity. Let's take the correspondence between luminance and chromaticity in a color space and RGB as an example: Y = 0.299R + 0.587G + 0.114B U = -0.1687R - 0.3313G + 0.5B + ​​128 V = 0.5R - 0.4187G - 0.0813B + 128 Here, Y represents luminance (i.e., grayscale value), U and V represent chromaticity and are used to represent the color and saturation of the image, and R, G, and B represent the three channels of color: red, green, and blue.

[0024] As can be seen from the above correspondence, there is a certain linear correlation and strong correlation redundancy between luminance and chromaticity, but the correlation is not perfectly linear. Therefore, when the mapping relationship between chromaticity and luminance in the CU is shown using luminance downsampling values ​​created with a single linear function, the chromaticity error predicted by the linear function becomes larger for points far from the line fitted by that linear function (there are many such far points for CUs with poor linear correlation), which increases the chromaticity prediction residual for the entire CU and reduces the compression ratio for the CU.

[0025] In response to problems existing in related technologies, this application proposes a general-purpose integrated model prediction-based codec method to achieve a wider range of applicability and higher compression ratios.

[0026] In the first embodiment, as shown in Figure 1, is a flowchart of a decoding method provided in an embodiment of the present application, which may include the following steps.

[0027] Step 101: Receive the encoding information corresponding to the target encoding block. Here, the encoding information includes index information and parameter information of the first model, index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, wherein the first pixel is a pixel that satisfies a first condition in the target encoding block, the second pixel is a pixel that does not satisfy the first condition in the target encoding block, and the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range. To make it understandable, the first condition is used to divide the pixels in the target encoding block into two categories: pixels corresponding to points close to the linear function fitted by the first model (i.e., first pixels where the chromaticity prediction error is within the first threshold range) and pixels corresponding to points far from the linear function fitted by the first model (i.e., second pixels where the chromaticity prediction error is outside the first threshold range).

[0028] Step 102: Determine the first model and the second model based on the index information and parameter information of the first model and the index information and parameter information of the second model. Here, the first model determines the predicted chromaticity value mapped from the luminance downsampling reconstruction value according to a linear mapping relationship, and the second model determines the predicted chromaticity value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship. In practice, the encoding end can determine a model for obtaining the target chromaticity prediction value of a pixel, and after determining the parameter values ​​of the model, generate index information (e.g., model index of the first model, model index of the second model) and parameter information for the model. Based on the received model index information and parameter information, the decoding end can then reconstruct the model used by the encoding end to determine the target chromaticity prediction value for the target encoding block.

[0029] Step 103: Determine the target chromaticity prediction value of the first pixel through the first model based on the luminance downsampling reconstruction value of the first pixel, and determine the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel. It can be understood that this first model is a model that performs chromaticity prediction using a linear function, similar to the linear model used in CCLM described above. In practice, the encoding end can determine the parameter values ​​in the first model (e.g., the linear model parameters α0 and α1 mentioned above) in advance based on a reference encoding block of the target encoding block (e.g., an adjacent sampling block, or a portion of the encoding blocks of the target encoding block). After determining the parameter values ​​in the first model, the first model can determine the subsequent target difference value by considering that it can achieve a better chromaticity prediction effect for pixels corresponding to points close to the linear function it has fitted. The first chromaticity prediction value output by the first model can be directly used as the target chromaticity prediction value corresponding to the first pixel, and the parameter values ​​in the first model can also be adjusted using the first pixel. By determining the target chromaticity prediction value of the first pixel through the first model after parameter adjustment, the straight line fitted by the first model after parameter adjustment can more accurately represent the mapping relationship between the chromaticity and luminance of the first pixel, thereby further reducing the chromaticity prediction error for the first pixel. Similarly, the encoding end can determine the parameter values ​​in the second model in advance based on the reference encoding block of the target encoding block, directly determine the parameter values ​​in the second model based on the second pixel, or adjust the determined parameter values ​​in the second model. After determining the parameter values ​​in the second model, the target chromaticity prediction value output by the second pixel can be obtained by inputting the luminance downsampling reconstruction value of the second pixel into the second model. To make it easier to understand, in codec processing, both integrated linear and nonlinear model approaches are used to determine the target chromaticity prediction value of the target encoding block. This allows for higher accuracy in chromaticity prediction for both target encoding blocks with relatively good or poor linear correlation between chromaticity and luminance, thereby expanding the applicability of the codec method.

[0030] Step 104: Determine the chromaticity reconstruction value of the target encoded block based on the target chromaticity prediction value and target difference value of the first pixel, and the target chromaticity prediction value and target difference value of the second pixel. In specific implementation, the encoding end determines the target chromaticity prediction values ​​for the first and second pixels, respectively, based on the steps described above, using the first and second models. Then, based on the target chromaticity prediction value and true chromaticity value of the first pixel, and the target chromaticity prediction value and true chromaticity value of the second pixel, it determines the target difference value for the first pixel and the target difference value for the second pixel, respectively. Furthermore, it generates encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, and the target difference value for the first and second pixels. The encoding end reduces the target difference value (e.g., residual value) of each pixel in the target encoding block as a whole using the integrated model prediction method. Therefore, the target difference value used by the encoding end to generate the encoding end information has a smaller data size, thereby improving the compression ratio for the target encoding block.

[0031] After receiving the encoding information corresponding to the target encoding block and the luminance downsampling reconstruction value of the target encoding block, the decoding terminal can determine the first model and the second model based on the index information and parameter information of the first model and the index information and parameter information of the second model, and determine the target chromaticity prediction value of the corresponding pixel through the first model and the second model based on the luminance downsampling reconstruction value of the first and second pixels, and finally determine the chromaticity reconstruction value of the target encoding block based on the target chromaticity prediction value and target difference value of the first pixel and the target chromaticity prediction value and target difference value of the second pixel.

[0032] In one possible embodiment, the above index information can also describe the positional distribution of pixels (e.g., first pixel, and second pixel) in the target encoding block corresponding to a model (e.g., first model, second model) in a method such as an index map, so that the decoding end can reconstruct the target chromaticity prediction value determined by the encoding end using the corresponding model for the pixels at the corresponding positions, based on the index information.

[0033] As can be seen from the steps above, the first condition classifies the pixels of the target encoding block into two types, and the first and second models are used to determine the target chromaticity prediction values ​​for each of these two types of pixels. This integrated model prediction method improves the prediction accuracy of the entire target encoding block and further improves the compression efficiency for the target encoding block. [Examples]

[0034] This embodiment describes an example of determining the target chromaticity prediction value of the first pixel. Based on the differences in the pixels used for model determination, it may be specifically classified into the following two cases.

[0035] Case 1: A first model is created based on the reference encoding block of the target encoding block, and the first chromaticity prediction value output by the first pixel according to the first model is directly determined as the target chromaticity prediction value of the first pixel. In practice, a linear model is selected beforehand, and a loss function corresponding to the linear model is created based on the luminance downsampled reconstruction value and chromaticity true value of the reference encoding block. For example, the loss function is determined based on the linear least mean squared error estimate of the linear model in which the chromaticity true value and luminance downsampled reconstruction value of the target encoding block are commonly mapped. Furthermore, the parameter values ​​in the linear model when the loss function is smallest are determined, and by substituting these parameter values ​​into the linear model, a first model corresponding to the target encoding block is obtained. Here, the linear model may be a model for showing a linear mapping relationship between luminance downsampling reconstruction values ​​such as CCLM and chromaticity prediction values. For example, the linear mapping relationship shown by the linear model (i.e., the linear mapping relationship used by the first model to determine the chromaticity prediction values) is as follows:

number

number

number

number

number

[0036] Case 2: The parameters of the first model (i.e., the initial first model) are adjusted based on the first chromaticity prediction value of the first pixel, and the chromaticity prediction value output by the first pixel by the first model after parameter adjustment is determined as the target chromaticity prediction value for that first pixel. In practice, the linear model used to create the first model is reused based on the luminance downsampling reconstruction value and true chromaticity value of the first pixel to obtain the corrected parameter values ​​(e.g., α0, α1). These corrected parameter values ​​are then substituted into the linear model to obtain the first model after parameter adjustment. By using this first model after parameter adjustment to determine the target chromaticity prediction value of the first pixel, errors can be effectively reduced. To make it clear, the process of determining the first pixel and the corrected parameter values ​​using the linear model is similar to the process of determining the parameter values ​​using the target encoding block and linear model described in Case 1, and will not be elaborated upon here. For Case 2, the encoding terminal can determine the modified parameter values ​​of the first model and then generate parameter information for the first model based on those modified parameter values. [Examples]

[0037] This embodiment describes an example of determining the target chromaticity prediction value for a second pixel. Specifically, a model can be selected in the following three cases to determine the target chromaticity prediction value for a second pixel.

[0038] Case 1: The target chromaticity prediction value for the second pixel is determined by directly using the second model. In this embodiment, since the correlation between luminance and chromaticity is not approximately linear, first, a first chromaticity prediction value for each pixel in the target encoding block is determined using a first model (e.g., CCLM). Next, an error calculation is performed based on the first chromaticity prediction value and the true chromaticity value. For the first pixel whose chromaticity prediction error is within the first threshold range, the chromaticity is predicted using a first-order linear model such as CCLM (i.e., the first model). For the second pixel whose chromaticity prediction error is outside the first threshold range, the chromaticity is predicted using a second-order linear model such as a curve model (i.e., the second model). This reduces the chromaticity prediction error of the first model through the second model and improves the prediction accuracy of the entire target encoding block.

[0039] Case 2: Select one model from the first and second models, and determine the target chromaticity prediction value for the second pixel. In this embodiment, the encoding terminal can determine a first prediction error (e.g., the sum of the mean squared errors of all second pixels) corresponding to the first model based on the first chromaticity prediction value and the true chromaticity value of the second pixel, determine a second chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel, and further determine a second prediction error corresponding to the second model based on the second chromaticity prediction value and the true chromaticity value of the second pixel. Subsequently, it is determined whether the first prediction error is smaller than the second prediction error, and when performing chromaticity prediction for the second pixel, it is determined whether the first model (the initial first model or the first model after parameter adjustment) has higher prediction accuracy than the second model. When the first prediction error is greater than or equal to the second prediction error, in order to explain why the prediction accuracy of the second model is higher than that of the first model, the second model is selected to determine the target chromaticity prediction value of the second pixel, and that is, the target chromaticity prediction value of the second pixel (second target pixel or fourth target pixel) is determined based on the second chromaticity prediction value of the second pixel. When the first prediction error is smaller than the second prediction error, it is explained that the prediction accuracy of the first model is higher than that of the second model. In this case, introducing the second model would not reduce the prediction error of the second pixel, so the first model is still selected to determine the target chromaticity prediction value of the second pixel. That is, the target chromaticity prediction value of the second pixel (i.e., the first target pixel or the third target pixel) is determined based on the first chromaticity prediction value of the second pixel.

[0040] In one embodiment, the decoding end can determine whether a second pixel is a first target pixel or a second target pixel based on the index information of a first model (i.e., an initial first model, in which case the encoding end uses the initial first model to determine the target chromaticity prediction value) and / or the index information of a second model. For example, the index information of the first model describes the model corresponding to each pixel in the target encoding block, and the index information of the second model describes the positional distribution of the first and second pixels in the target encoding block. In this case, the decoding end can determine the positional distribution of the second pixel and the corresponding model based on the received index information, and further determine whether the second pixel is a first target pixel or a second target pixel.

[0041] At the decoding end, if the second pixel is the first target pixel, the target chromaticity prediction value of the first target pixel is determined through the first model based on the luminance downsampling reconstruction value of the first target pixel. If the second pixel is the second target pixel, the target chromaticity prediction value of the second target pixel is determined through the second model based on the luminance downsampling reconstruction value of the second target pixel.

[0042] To make it understandable, the first target pixel is a second pixel whose first prediction error is smaller than the second prediction error, the second target pixel is a second pixel whose first prediction error is greater than or equal to the second prediction error, the first prediction error is determined based on the initial chromaticity prediction value of the second pixel by the first model and the true chromaticity value of the second pixel, and the second prediction error is determined based on the chromaticity prediction value of the second pixel by the second model and the true chromaticity value of the second pixel.

[0043] In other embodiments, when the second pixel is the third target pixel, the decoding end determines the target chromaticity prediction value of the third target pixel through the first model (i.e., the first model with adjusted parameters, in which case the encoding end determines the target chromaticity prediction value in accordance with the first model with adjusted parameters) based on the luminance downsampling reconstruction value of the third target pixel. When the second pixel is the fourth target pixel, the target chromaticity prediction value of the fourth target pixel is determined through the second model based on the luminance downsampling reconstruction value of the fourth target pixel.

[0044] To make it easier to understand, the third target pixel is the second pixel whose third prediction error is smaller than the fourth prediction error, the fourth target pixel is the second pixel whose third prediction error is greater than or equal to the fourth prediction error, the third prediction error is determined based on the chromaticity prediction value of the second pixel and the true chromaticity value of the second pixel by the first model with parameter adjustments, and the fourth prediction error is determined based on the chromaticity prediction value of the second pixel and the true chromaticity value of the second pixel by the second model.

[0045] Case 3: Select one model from the first and second models after parameter adjustment, and determine the target chromaticity prediction value for the second pixel.

[0046] In this embodiment, after performing parameter adjustments on the first model using the first pixel, the chromaticity of the second pixel can be predicted using the first model after parameter adjustment. Specifically, the third chromaticity prediction value of the second pixel is determined through the first model after parameter adjustment based on the luminance downsampling reconstruction value of the second pixel, and the third prediction error corresponding to the first model after parameter adjustment is determined based on the third chromaticity prediction value and the true chromaticity value of the second pixel.

[0047] Simultaneously, the chromaticity of the second pixel can be predicted using the created second model. Specifically, the fourth chromaticity prediction value of the second pixel is determined through the second model based on the luminance downsampling reconstruction value of the second pixel, and the fourth prediction error corresponding to the second model is determined based on the fourth chromaticity prediction value and the true chromaticity value of the second pixel.

[0048] The system determines whether the third prediction error is smaller than the fourth prediction error, that is, when performing chromaticity prediction for the second pixel, it determines whether the prediction accuracy of the first model after parameter adjustment is higher than the prediction accuracy of the second model. If the third prediction error is smaller than the fourth prediction error, it is explained that the prediction accuracy of the first model after parameter adjustment is higher than that of the second model, and in this case, the target chromaticity prediction value of the second pixel is determined based on the third chromaticity prediction value of the second pixel. If the third prediction error is greater than or equal to the fourth prediction error, in order to explain that the second model is higher than the prediction accuracy of the first model after parameter adjustment, the target chromaticity prediction value of the second pixel is determined based on the fourth chromaticity prediction value of the second pixel in order to reduce the chromaticity prediction error for the entire second pixel. [Examples]

[0049] This embodiment describes an example of a second model.

[0050] In this embodiment, the second model may be created based on a target encoding block (for example, a reference encoding block of the target encoding block), or it may be created based on a second pixel in order to improve the chromaticity prediction accuracy of the second model for the second pixel.

[0051] For example, a second model is created based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel point, using a target mapping relationship (e.g., a mapping relationship shown by a quadratic curve model) to show the nonlinear mapping relationship between the luminance downsampling reconstruction value and the chromaticity prediction value.

[0052] In practice, the values ​​of each parameter in the target equation, which shows the target mapping relationship, are determined based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel. For example, the values ​​of each parameter are determined based on the linear least mean squared error estimate of the target equation, to which the true chromaticity value and the luminance downsampling reconstruction value of the second pixel are commonly mapped. Furthermore, a second model is created based on the target equation and the values ​​of each parameter.

[0053] For example, the target expression is as follows:

number

number

[0054] In this case, the linear least mean squared error estimate of the target equation is determined by the following loss function:

number

[0055] Differentiating equation (5) with respect to parameter values ​​α2, α3, and α4, we obtain the following:

number

number

[0056] Substituting the determined α2, α3, and α4 into the target formula above yields a second model. As can be seen, the second model has 3 parameters, so at least 3 second pixels are needed to determine the parameter values ​​in the second model. Here, since the minimum size of a video image chromaticity block (i.e., a VVC video chromaticity block) is 4x4, that is, a chromaticity block in the target encoding block or an adjacent encoding block has at least 16 pixels, any first and second model described in this application can be completed using the target encoding block.

[0057] In one possible embodiment, to reduce computational load or ensure the successful construction of the second model, it may be determined whether to enable integrated chromaticity prediction based on the first and second models based on the number of second pixels. Specifically, it may be possible to detect whether the number of second pixels is greater than a quantitative threshold (e.g., 2), which can be determined based on the number of parameters in the second model and / or the number of pixels for which an acceptable prediction error is large.

[0058] If the number of second pixels is greater than or equal to a quantitative threshold, the target chromaticity prediction value of the target encoding block is determined using the first model directly, that is, the target chromaticity prediction value of the target encoding block is determined based on the first chromaticity prediction value of the target encoding block. If the number of second pixels exceeds the quantitative threshold, the integrated chromaticity prediction method is enabled, and for example, the target chromaticity prediction value of the first pixel is determined based on the first chromaticity prediction value of the first pixel, and the target chromaticity prediction value of the second pixel is determined through the second model based on the luminance downsampling reconstruction value of the second pixel.

[0059] In one embodiment, the encoding end determines the second pixels from the target encoding block that do not satisfy the first condition, and then detects whether the number of second pixels exceeds a quantity threshold. If the number of second pixels is less than or equal to a quantity threshold, the encoding terminal determines the target chromaticity prediction value of the target encoding block based on the first chromaticity prediction value of the target encoding block. Based on the target chromaticity prediction value and the true chromaticity value of the target encoding block, the target difference value of the target encoding block is determined. Based on the index information and parameter information of the first model, and the luminance downsampling reconstruction value and target difference value of the target encoding block, encoding information corresponding to the target encoding block is generated.

[0060] If the number of second pixels exceeds the aforementioned quantity threshold, the encoding end, Based on the first chromaticity prediction value of the first pixel, the target chromaticity prediction value of the first pixel is determined, If the number of the second pixels exceeds the quantity threshold, the target chromaticity prediction value of the second pixel is determined through the second model based on the luminance downsampling reconstruction value of the second pixel, Based on the predicted target chromaticity value and the true chromaticity value of the first pixel, and the predicted target chromaticity value and the true chromaticity value of the second pixel, the target difference value of the first pixel and the target difference value of the second pixel are determined, respectively. The operation of generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel is performed.

[0061] In one embodiment, the decoding end receives first encoding information corresponding to a target encoding block. This first encoding information is generated when the number of second pixels exceeds a quantitative threshold, and the first encoding information includes index information and parameter information of the first model, index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel.

[0062] If the first encoding information is not received, but the second encoding information corresponding to the target encoding block is received, the encoding end determines the target chromaticity prediction value of the target encoding block through the first model based on the luminance downsampling reconstruction value of the target encoding block. Based on the target chromaticity prediction value and target difference value of the target encoding block, the chromaticity reconstruction value of the target encoding block is determined.

[0063] Here, the second encoding information is generated when the number of the second pixels is less than or equal to the quantity threshold, and the second encoding information includes the index information and parameter information of the first model, and the luminance downsampling reconstruction value and target difference value of the target encoding block.

[0064] In the above embodiment 3, the second model may be any model other than a first-order linear model that can realize chromaticity prediction. For example, the second model may be a curve model, or a linear model with added nonlinear attributes (such as a linear model combining angles, a piecewise linear model, etc.).

[0065] Therefore, in addition to the second model created by the target equation described above, this second model is determined by the parameters of one of the following models: Cross-Component General Model (CCGM), Slope adjustment of CCLM, Multi-Model Linear Model (MM-CCLM), Multi-sampling Filter linear model (MF-CCLM), Angular Prediction (LM-LAP), Convolutional Cross-Component Model (CCCM), and Gradient Linear Model (GLM).

[0066] Here, CCGM is a chromaticity prediction model that uses an activation function to correct errors. For example, CCGM represents the following two types of nonlinear mapping relationships. Nonlinear mapping relationship 1:

number

number

number

[0067] This embodiment illustrates an example of the first condition. Based on the difference in the method for determining the chromaticity prediction error, it may be classified into the following two cases.

[0068] Case 1: The chromaticity prediction error of a pixel by the first model is determined based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value. For example, the pixel chromaticity prediction error by the first model is determined through equation 5, and the said equation 5 is,

number

[0069] Case 2: The chromaticity prediction error of the pixel by the first model is determined based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel. For example, the chromaticity prediction error of the pixel by the first model is determined through equation 7, where this equation 7 is:

number

[0070] In a second embodiment, as shown in Figure 2, the embodiment of the present application provides an encoding method. This encoding method includes the following steps:

[0071] Step 201: Based on the luminance downsampling reconstruction value of the target encoding block, a first chromaticity prediction value of the target encoding block is determined through a first model, and the first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a linear mapping relationship.

[0072] Step 202: Determine a first pixel from the target encoding block that satisfies the first condition, and determine a second pixel from the target encoding block that does not satisfy the first condition, wherein the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range.

[0073] Step 203: Based on the first chromaticity prediction value of the first pixel, the target chromaticity prediction value of the first pixel is determined, and based on the luminance downsampling reconstruction value of the second pixel, the target chromaticity prediction value of the second pixel is determined through the second model. Here, the second model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship.

[0074] Step 204: Based on the target chromaticity prediction value and true chromaticity value of the first pixel, and the target chromaticity prediction value and true chromaticity value of the second pixel, the target difference value of the first pixel and the target difference value of the second pixel are determined, respectively.

[0075] Step 205: Based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, encoding information corresponding to the target encoding block is generated.

[0076] From the steps described above, the pixels of the target encoding block are classified into two types according to the first condition, and the target chromaticity prediction values ​​for the two types of pixels are determined using the first and second models, thereby improving the prediction accuracy of the entire target encoding block and improving the compression efficiency for the target encoding block using the integrated model prediction method.

[0077] In one possible embodiment, determining a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel includes determining a first prediction error corresponding to the first model based on a first chromaticity prediction value and the true chromaticity value of the second pixel; determining a second chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel, and determining a second prediction error corresponding to the second model based on the second chromaticity prediction value and the true chromaticity value of the second pixel; determining whether the first prediction error is less than the second prediction error; and, if the first prediction error is greater than or equal to the second prediction error, determining a target chromaticity prediction value for the second pixel based on the second chromaticity prediction value of the second pixel. The encoding method further includes determining a target chromaticity prediction value for the second pixel based on a first chromaticity prediction value for the second pixel when the first prediction error is smaller than the second prediction error.

[0078] In one possible embodiment, determining a target chromaticity prediction value for the first pixel based on a first chromaticity prediction value for the first pixel includes adjusting the parameters of the first model based on the first chromaticity prediction value for the first pixel, and determining the target chromaticity prediction value for the first pixel through the parameter-adjusted first model based on the luminance downsampling reconstruction value for the first pixel.

[0079] In one possible embodiment, determining a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel includes: determining a third chromaticity prediction value for the second pixel through a parameter-adjusted first model based on the luminance downsampling reconstruction value of the second pixel, and determining a third prediction error corresponding to the parameter-adjusted first model based on the third chromaticity prediction value and the true chromaticity value of the second pixel; determining a fourth chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel, and determining a fourth prediction error corresponding to the second model based on the fourth chromaticity prediction value and the true chromaticity value of the second pixel; determining whether the third prediction error is less than the fourth prediction error; determining a target chromaticity prediction value for the second pixel based on the third chromaticity prediction value if the third prediction error is less than the fourth prediction error; and determining a target chromaticity prediction value for the second pixel based on the fourth chromaticity prediction value if the third prediction error is greater than or equal to the fourth prediction error.

[0080] In one possible embodiment, after determining the second pixels that do not satisfy the first condition from the target encoding block, the encoding method detects whether the number of the second pixels exceeds a quantity threshold, and if the number of the second pixels is less than or equal to the quantity threshold, determines a target chromaticity prediction value for the target encoding block based on the first chromaticity prediction value of the target encoding block, determines a target difference value for the target encoding block based on the target chromaticity prediction value and the true chromaticity value of the target encoding block, and performs luminance downsampling reconstruction of the target encoding block using the index information and parameter information of the first model and the luminance downsampling reconstruction of the target encoding block. The process further includes generating encoding information corresponding to the target encoding block based on a value and a target difference value, wherein determining a target chromaticity prediction value for the first pixel based on a first chromaticity prediction value for the first pixel includes determining a target chromaticity prediction value for the first pixel based on a first chromaticity prediction value for the first pixel when the number of second pixels exceeds the quantity threshold, and determining a target chromaticity prediction value for the second pixel through a second model based on a luminance downsampling reconstruction value for the second pixel includes determining a target chromaticity prediction value for the second pixel through a second model based on a luminance downsampling reconstruction value for the second pixel when the number of second pixels exceeds the quantity threshold. Determining the target difference value of the first pixel and the target difference value of the second pixel, respectively, based on the target chromaticity prediction value and true chromaticity value of the first pixel, and the target chromaticity prediction value and true chromaticity value of the second pixel, includes, when the number of second pixels exceeds the quantity threshold, determining the target difference value of the first pixel and the target difference value of the second pixel, respectively, based on the target chromaticity prediction value and true chromaticity value of the first pixel, and the target chromaticity prediction value and true chromaticity value of the second pixel, respectively, the index information and parameter information of the first model, the index information and parameter information of the second model, and the luminance downsample of the first pixel Generating encoding information corresponding to the target encoding block based on the luminance downsampling reconstruction value and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, includes, when the number of second pixels exceeds the quantity threshold, generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value and the target difference value of the second pixel.

[0081] In one possible embodiment, before determining a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampled reconstructed value of the second pixel, the encoding method further includes constructing the second model through a target mapping relationship based on the luminance downsampled reconstructed value and the true chromaticity value of the second pixel, wherein the target mapping relationship exhibits a nonlinear mapping relationship between the luminance downsampled reconstructed value and the chromaticity prediction value.

[0082] In one possible embodiment, creating the second model through a target mapping relationship based on the luminance downsampled reconstruction value and the chromaticity true value of the second pixel includes determining the values ​​of each parameter in the target equation based on the luminance downsampled reconstruction value and the chromaticity true value of the second pixel, such that the target equation represents the target mapping relationship, and creating the second model based on the target equation and the values ​​of each parameter.

[0083] In one possible embodiment, the target expression is:

number

number

[0084] In one possible embodiment, the chromaticity prediction error of the pixel by the first model is determined based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value, or The chromaticity prediction error of the pixel by the first model is determined based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel.

[0085] In one possible embodiment, if the pixel chromaticity prediction error by the first model is determined based on the ratio of the first chromaticity prediction value to the true chromaticity value, then the first condition is that the pixel chromaticity prediction error by the first model satisfies equations 1 and 2, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 3, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 4, where equation 1 is: A(i,j) <= threadhold1, The second equation is, A(i,j)>=-threadhold1, The third equation is, A(i,j) > threadhold1, The fourth equation is, A(i,j) <- threadhold1, Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold1 is the first threshold.

[0086] In one possible embodiment, determining the chromaticity prediction error of a pixel by the first model based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value includes determining the chromaticity prediction error of the pixel by the first model through equation 5, Here, the fifth equation is,

number

[0087] In one possible embodiment, when the chromaticity prediction error of a pixel by the first model is determined based on the difference between a first chromaticity prediction value and the true chromaticity value of the pixel, the first condition is that the chromaticity prediction error of the pixel by the first model satisfies equation 6. Here, the sixth equation is, Abs(A(i,j))>=threadhold2 Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold2 is the second threshold.

[0088] As one possible implementation, determining the chrominance prediction error of the pixel by the first model based on the difference between the first chrominance prediction value and the true chrominance value of the pixel includes determining the chrominance prediction error of the pixel by the first model through Equation 7, where the Equation 7 is [Equation] where A(i, j) is the chrominance prediction error of the pixel at coordinates (i, j) by the first model, pred C (i, j) is the chrominance prediction value of the pixel at coordinates (i, j), and C(i, j) is the true chrominance value of the pixel at coordinates (i, j).

[0089] As one possible implementation, the first model determines the chrominance prediction value mapped from the luminance downsampling reconstruction value according to the linear mapping relationship of the following formula [Equation] where pred C (i, j) is the chrominance prediction value of the pixel at coordinates (i, j), [Equation] is the luminance downsampling reconstruction value of the pixel at coordinates (i, j), and α0, α1 are the values of the parameters of the first model.

[0090] As one possible implementation, the value of each parameter of the first model and / or the value of each parameter of the second model are determined based on the linear least squares error estimation value jointly mapped by the true chrominance value and the luminance downsampling reconstruction value of the target encoding block.

[0091] To ensure understanding, the entities that execute the encoding method and decoding method provided in the embodiment of the present application may be an encoding device and a decoding device. In this embodiment, the encoding device and decoding device provided in the present application will be described as an example in which the encoding device and decoding device each execute the encoding method and decoding method.

[0092] In a third embodiment, the embodiment of the present application provides a decoding device. As shown in Figure 3, the decoding device 100 receives encoding information corresponding to a target encoding block, and the encoding information includes index information and parameter information of a first model, index information and parameter information of a second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, wherein the first pixel is a pixel that satisfies a first condition in the target encoding block, and the second pixel is a pixel that does not satisfy the first condition in the target encoding block, and the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range. Information receiving module 101, index information and parameter information of the first model, index information and parameter information of the second model The system includes: a model determination module 102 that determines a first model and a second model based on the report, the first model determining a chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a linear mapping relationship, and the second model determining a chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship; a first chromaticity prediction module 103 that determines a target chromaticity prediction value for the first pixel through the first model based on the luminance downsampling reconstruction value of the first pixel; a second chromaticity prediction module 104 that determines a target chromaticity prediction value for the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel; and a reconstruction module 105 that determines the chromaticity reconstruction value of the target encode block based on the target chromaticity prediction value and target difference value of the first pixel, and the target chromaticity prediction value and target difference value of the second pixel.

[0093] Selectively, the decoding device further includes a third chromaticity prediction module that, when the second pixel is a first target pixel, determines a target chromaticity prediction value for the first target pixel through the first model based on the luminance downsampling reconstruction value of the first target pixel. The second chromaticity prediction module 104 includes a fourth chromaticity prediction module that, when the second pixel is a second target pixel, determines a target chromaticity prediction value for the second target pixel through the second model based on the luminance downsampling reconstruction value of the second target pixel. Here, the first target pixel is a second pixel whose first prediction error is smaller than the second prediction error, the second target pixel is a second pixel whose first prediction error is greater than or equal to the second prediction error, the first prediction error is determined based on the chromaticity prediction value of the second pixel by the first model and the true chromaticity value of the second pixel, and the second prediction error is determined based on the chromaticity prediction value of the second pixel by the second model and the true chromaticity value of the second pixel.

[0094] Selectively, the first model is obtained after parameter adjustments have been made to the initial first model and the chromaticity prediction value of the first pixel.

[0095] Selectively, the decoding device further includes a fifth chromaticity prediction module that, when the second pixel is a third target pixel, determines a target chromaticity prediction value for the third target pixel through the first model based on the luminance downsampling reconstruction value of the third target pixel. The second chromaticity prediction module 104 includes a sixth chromaticity prediction module that, when the second pixel is a fourth target pixel, determines a target chromaticity prediction value for the fourth target pixel through the second model based on the luminance downsampling reconstruction value of the fourth target pixel. Here, the third target pixel is the second pixel whose third prediction error is smaller than the fourth prediction error, the fourth target pixel is the second pixel whose third prediction error is greater than or equal to the fourth prediction error, the third prediction error is determined based on the chromaticity prediction value of the second pixel by the first model and the true chromaticity value of the second pixel, and the fourth prediction error is determined based on the chromaticity prediction value of the second pixel by the second model and the true chromaticity value of the second pixel.

[0096] Selectively, the information receiving module 101 receives first encoding information corresponding to a target encoding block, and the first encoding information is generated when the number of second pixels exceeds a quantity threshold, and the first information receiving submodule includes index information and parameter information of a first model, index information and parameter information of a second model, a luminance downsampling reconstruction value of the first pixel and a target difference value of the first pixel, and a luminance downsampling reconstruction value of the second pixel and a target difference value of the second pixel. The decoding device is The system further includes a second information receiving submodule that, when the first encoding information is not received but the second encoding information corresponding to the target encoding block is received, determines a target chromaticity prediction value for the target encoding block through the first model based on the luminance downsampling reconstruction value of the target encoding block, and a first chromaticity reconstruction module that determines a chromaticity reconstruction value for the target encoding block based on the target chromaticity prediction value and target difference value of the target encoding block. Here, the second encoding information is generated when the number of the second pixels is less than or equal to the quantity threshold, and the second encoding information includes the index information and parameter information of the first model, and the luminance downsampling reconstruction value and target difference value of the target encoding block.

[0097] Selectively, the second model is constructed through a target mapping relationship based on the luminance downsampled reconstruction value and the true chromaticity value of the second pixel, and the target mapping relationship represents a nonlinear mapping relationship between the luminance downsampled reconstruction value and the chromaticity prediction value.

[0098] The second model is constructed based on a target formula and the values ​​of each parameter in the target formula, wherein the values ​​of each parameter in the target formula are determined based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel, and the target formula represents the target mapping relationship.

[0099] Selectively, the target expression is

number

number

[0100] Selectively, the chromaticity prediction error of the pixel by the first model is determined based on the ratio of the first chromaticity prediction value of the pixel by the first model to the true chromaticity value, or the chromaticity prediction error of the pixel by the first model is determined based on the difference between the first chromaticity prediction value of the pixel by the first model and the true chromaticity value.

[0101] If, selectively, the pixel chromaticity prediction error by the first model is determined based on the ratio of the first chromaticity prediction value to the true chromaticity value, then the first condition is that the pixel chromaticity prediction error by the first model satisfies equations 1 and 2, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 3, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 4, where equation 1 is, A(i,j) <= threadhold1, The second equation is, A(i,j)>=-threadhold1, The third equation is, A(i,j) > threadhold1, The fourth equation is, A(i,j) <- threadhold1, Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold1 is the first threshold.

[0102] Selectively determining the chromaticity prediction error of a pixel by the first model based on the ratio of the first chromaticity prediction value of the pixel by the first model to the true chromaticity value includes determining the chromaticity prediction error of the pixel by the first model through equation 5, where equation 5 is:

number

[0103] When selectively determining the chromaticity prediction error of a pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel by the first model, the first condition is that the chromaticity prediction error of the pixel by the first model satisfies equation 6. Here, the sixth equation is, Abs(A(i,j))>=threadhold2 Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold2 is the second threshold.

[0104] Selectively determining the chromaticity prediction error of a pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel by the first model includes determining the chromaticity prediction error of the pixel by the first model through equation 7, where equation 7 is:

number

[0105] Selectively, the first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to the linear mapping relationship of the following equation:

number

number

[0106] Selectively, the values ​​of each parameter of the first model and / or each parameter of the second model are determined based on a linear least-squares error estimate jointly mapped between the true chromaticity value and the luminance downsampling reconstruction value of the target encoded block.

[0107] The decoding device provided in the embodiment of the present application can implement each process realized by the embodiment of the decoding method described in the first aspect, and can achieve the same technical effects. To avoid redundant explanation, a detailed explanation is omitted here.

[0108] In a fourth embodiment, the embodiment of the present application provides an encoding device. As shown in Figure 4, the encoding device 200 includes: a first determination module 201 that determines a first chromaticity prediction value of a target encoding block through a first model based on the luminance downsampling reconstruction value of the target encoding block, and determines a chromaticity prediction value mapped from the luminance downsampling reconstruction value by the first model according to a linear mapping relationship; a second determination module 202 that determines a first pixel from the target encoding block that satisfies a first condition and a second pixel from the target encoding block that does not satisfy the first condition, wherein the chromaticity prediction error of the pixel by the first model is within a first threshold range; a third determination module 203 that determines a target chromaticity prediction value of the first pixel based on the first chromaticity prediction value of the first pixel; and a target color of the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel. The system includes: a fourth determination module 204 that determines a chromaticity prediction value and determines a chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship for the second model; a fifth determination module 205 that determines the target difference value for the first pixel and the target difference value for the second pixel, respectively, based on the target chromaticity prediction value and the true chromaticity value of the first pixel, and the target chromaticity prediction value and the true chromaticity value of the second pixel; and a first generation module 206 that generates encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value and the target difference value of the first pixel, and the luminance downsampling reconstruction value and the target difference value of the second pixel.

[0109] Selectively, the fourth decision module 204 includes: a first decision submodule that determines a first prediction error corresponding to the first model based on a first chromaticity prediction value and a true chromaticity value of the second pixel; a second decision submodule that determines a second chromaticity prediction value of the second pixel through a second model based on a luminance downsampling reconstruction value of the second pixel, and determines a second prediction error corresponding to the second model based on the second chromaticity prediction value and a true chromaticity value of the second pixel; a third decision submodule that determines whether the first prediction error is less than the second prediction error; and a fourth decision submodule that, if the first prediction error is greater than or equal to the second prediction error, determines a target chromaticity prediction value of the second pixel based on the second chromaticity prediction value of the second pixel. The encoding method further includes a sixth determination module that determines a target chromaticity prediction value for the second pixel based on a first chromaticity prediction value for the second pixel when the first prediction error is smaller than the second prediction error.

[0110] Selectively, the third decision module 203 includes a fifth decision submodule that adjusts the parameters of the first model based on the first chromaticity prediction value of the first pixel, and a sixth decision submodule that determines the target chromaticity prediction value of the first pixel through the parameterized first model based on the luminance downsampling reconstruction value of the first pixel.

[0111] Selectively, the fourth decision module 204 includes: a seventh decision submodule that determines a third chromaticity prediction value of the second pixel through the parameter-adjusted first model based on the luminance downsampling reconstruction value of the second pixel, and determines a third prediction error corresponding to the parameter-adjusted first model based on the third chromaticity prediction value and the true chromaticity value of the second pixel; an eighth decision submodule that determines a fourth chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel, and determines a fourth prediction error corresponding to the second model based on the fourth chromaticity prediction value and the true chromaticity value of the second pixel; a ninth decision submodule that determines whether the third prediction error is smaller than the fourth prediction error; a tenth decision submodule that determines a target chromaticity prediction value of the second pixel based on the third chromaticity prediction value of the second pixel if the third prediction error is smaller than the fourth prediction error; and an eleventh decision submodule that determines a target chromaticity prediction value of the second pixel based on the fourth chromaticity prediction value of the second pixel if the third prediction error is greater than or equal to the fourth prediction error.

[0112] After selectively determining second pixels from the target encoding block that do not satisfy the first condition, the encoding device further includes: a first detection module that detects whether the number of second pixels exceeds a quantity threshold; a seventh determination module that, if the number of second pixels is less than or equal to the quantity threshold, determines a target chromaticity prediction value for the target encoding block based on a first chromaticity prediction value for the target encoding block; an eighth determination module that determines a target difference value for the target encoding block based on the target chromaticity prediction value and the true chromaticity value for the target encoding block; and a ninth determination module that generates encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the luminance downsampling reconstruction value and the target difference value for the target encoding block. The third decision module 203 includes a twelfth decision submodule that determines a target chromaticity prediction value for the first pixel based on the first chromaticity prediction value of the first pixel when the number of the second pixels exceeds the quantity threshold. The fourth decision module 204 includes a thirteenth decision submodule that, when the number of second pixels exceeds the quantity threshold, determines a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel. The fifth decision module 205 includes a fourteenth decision submodule that, when the number of second pixels exceeds the quantity threshold, determines the target difference value of the first pixel and the target difference value of the second pixel, respectively, based on the target chromaticity prediction value and the true chromaticity value of the first pixel, and the target chromaticity prediction value and the true chromaticity value of the second pixel. The first generation module 206 includes a first generation submodule that, when the number of second pixels exceeds the quantity threshold, generates encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value and the target difference value of the second pixel.

[0113] Selectively, before determining a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampled reconstruction value of the second pixel, the encoding device constructs the second model through a target mapping relationship based on the luminance downsampled reconstruction value and the true chromaticity value of the second pixel, further comprising a first construction module in which the target mapping relationship shows a nonlinear mapping relationship between the luminance downsampled reconstruction value and the chromaticity prediction value.

[0114] Selectively, the first construction module includes a first construction submodule that determines the values ​​of each parameter in a target equation based on the luminance downsampling reconstruction value and the chromaticity true value of the second pixel, so that the target equation represents the target mapping relationship, and a second construction submodule that constructs the second model based on the target equation and the values ​​of each parameter.

[0115] Selectively, the target expression is

number

number

[0116] Selectively, the chromaticity prediction error of the pixel by the first model is determined based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value, or the chromaticity prediction error of the pixel by the first model is determined based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel.

[0117] If, selectively, the pixel chromaticity prediction error by the first model is determined based on the ratio of the first chromaticity prediction value to the true chromaticity value, then the first condition is that the pixel chromaticity prediction error by the first model satisfies equations 1 and 2, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 3, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 4, where equation 1 is, A(i,j) <= threadhold1, The second equation is, A(i,j)>=-threadhold1, The third equation is, A(i,j) > threadhold1, The fourth equation is, A(i,j) <- threadhold1, Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold1 is the first threshold.

[0118] Selectively determining the chromaticity prediction error of a pixel by the first model based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value includes determining the chromaticity prediction error of the pixel by the first model through equation 5. Here, the fifth equation is,

number

[0119] When selectively determining the chromaticity prediction error of a pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel, the first condition is that the chromaticity prediction error of the pixel by the first model satisfies equation 6. Here, the sixth equation is, Abs(A(i,j))>=threadhold2 Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold2 is the second threshold.

[0120] Selectively determining the chromaticity prediction error of the pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel includes determining the chromaticity prediction error of the pixel by the first model through Equation 7. Here, equation 7 is,

number

[0121] Selectively, the first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to the linear mapping relationship of the following equation:

number

number

[0122] Selectively, the values ​​of each parameter of the first model and / or each parameter of the second model are determined based on a linear least-squares error estimate jointly mapped between the true chromaticity value and the luminance downsampling reconstruction value of the target encoded block.

[0123] The encoding apparatus provided in the embodiment of the present application can realize each process realized by the embodiment of the encoding method described in the second aspect, and can achieve the same technical effects. To avoid redundant explanation, a detailed explanation is omitted here.

[0124] The embodiments of the apparatus described above are merely illustrative. Units described as separate components may or may not be physically separate. Components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Depending on actual needs, some or all of these modules can be selected to achieve the objectives of this embodiment. Those skilled in the art will understand and implement this without requiring any creative effort.

[0125] Embodiments of each component of the present disclosure may be implemented in hardware, by software modules running on one or more processors, or by a combination thereof. As those skilled in the art will understand, in practical applications, a microprocessor or digital signal processor (DSP) can be used to implement some or all of the functions of some or all components of a computing device relating to embodiments of the present disclosure. The present disclosure may also be implemented as a program for a device or apparatus (e.g., a computer program and a computer program product) to perform some or all of the methods described herein. Such a program implementing the present disclosure may be stored on a computer-readable medium or may be in the form of one or more signals. Such signals may be obtained by downloading from a website on the Internet, may be provided on a carrier signal, or may be provided in any other form.

[0126] For example, Figure 5 shows a computing device capable of implementing the method according to this disclosure. The computing device includes a conventional processor 1010 and a memory 1020 in the form of a computer program product or a computer-readable medium. The memory 1020 may be an electronic storage device such as flash memory, EEPROM (electrically erasable programmable read-only memory), EPROM, hard disk, or ROM. The memory 1020 has a storage area 1030 that stores program code 1031 for performing any of the method steps described above. For example, the storage area 1030 that stores the program code may each contain individual program code 1031 for performing various steps in the method described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards, and floppy disks. Such computer program products are typically portable or fixed storage units, as described with reference to Figure 6. The storage unit may have storage segments, storage areas, etc., arranged similarly to the memory 1020 of the computing device in Figure 5. The program code may be compressed in an appropriate format, for example. Typically, the storage unit contains computer-readable code 1031', i.e., code readable by a processor such as 1010, and when these codes are executed by the computing device, they cause the computing device to perform each step of the method described above.

[0127] Embodiments of the present invention further provide a decoding device, which includes a processor and memory, the memory storing a program or instruction executable on the processor. When the program or instruction is executed by the processor, each step of the embodiment of the decoding method described above is realized and the same technical effect is achieved. To avoid redundant explanation, a detailed explanation is omitted here.

[0128] Embodiments of the present application further provide an encoding device, the encoding device comprising a processor and memory, the memory storing a program or instruction executable on the processor. When the program or instruction is executed by the processor, each process of the above-described embodiment of the encoding method is realized and the same technical effect is achieved. To avoid redundant explanation, a detailed explanation is omitted here.

[0129] Embodiments of the present invention further provide a readable storage medium. The readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, each process of the above-described embodiment of the decoding method or the above-described embodiment of the encoding method is realized and the same technical effects are achieved. To avoid redundant explanation, a detailed explanation is omitted here.

[0130] Here, the processor is the processor of the terminal device described in the above embodiment. The readable storage medium includes computer-readable storage media such as computer read-only memory ROM, random access memory RAM, magnetic disk, and optical disk.

[0131] Embodiments of the present invention further provide a chip comprising a processor and a communication interface, the communication interface being coupled with the processor, the processor being used to execute a program or instructions, and realizing each step of the embodiment of the decoding method or the embodiment of the encoding method described above, and achieving the same technical effects. To avoid redundant explanation, a detailed explanation is omitted here.

[0132] To ensure clarity, the chips referred to in the embodiments of this application may also be called system-on-a-chip, system chip, chip system, or on-chip system, etc.

[0133] Embodiments of the present application further provide a computer program / program product. The computer program / program product is stored on a storage medium and, when executed by one or more processors, realizes each of the processes of the above-described embodiment of the decoding method or the above-described embodiment of the encoding method, and achieves the same technical effects. To avoid redundant explanation, a detailed explanation is omitted here.

[0134] Embodiments of the present application further provide a codec system comprising an encoding device and a decoding device, wherein the decoding device is used to perform the steps of the decoding method described in the first embodiment, and the encoding device is used to perform the steps of the encoding method described in the second embodiment.

[0135] In this specification, the terms “include,” “incorporate,” or any other variations thereof are intended to be non-exclusive, so that a process, method, article, or apparatus containing a set of elements includes not only those elements but also other elements not explicitly listed, or further elements specific to that process, method, article, or apparatus. Unless further limited, an element limited by “one…” does not preclude the existence of other identical elements in a process, method, article, or apparatus containing that element. It should also be noted that the scope of methods and apparatus in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may include performing functions in a nearly simultaneous or reverse order depending on the relevant functions. For example, a described method may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described by reference to one example may be combined in other examples.

[0136] Through the above description of the implementation, those skilled in the art will clearly understand that the methods of the embodiments described above can be implemented by adding a general-purpose hardware platform necessary for the software. Of course, it can also be implemented by hardware, but in many cases the former is a more preferred implementation. Based on this understanding, the essence of the technical solution of the present application, or the part that contributes to existing technology, can be embodied in the form of a computer software product. The computer software product is stored in a storage medium (e.g., ROM / RAM, magnetic disk, optical disk), contains a number of instructions, and causes a single terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0137] Although embodiments of the present application have been described above by combining drawings, the present application is not limited to the specific embodiments described above. The specific embodiments described above are merely schematic and not limiting. Under the disclosure of the present application, a person skilled in the art can devise many forms without departing from the spirit and claims of the present application, all of which fall within the scope of protection of the present application.

[0138] In this specification, the terms "one embodiment," "embodiment," or "one or more embodiments" mean that a particular feature, structure, or property described in connection with an embodiment is included in one embodiment of this disclosure. Furthermore, it should be noted that examples of the expression "in one embodiment" as used herein do not necessarily all refer to the same embodiment.

[0139] This specification provides a great deal of specific detail. However, it can be understood that the embodiments of this disclosure may be carried out without these specific details. In some examples, well-known methods, structures, and techniques are not described in detail in order to avoid obscuring the understanding of this specification.

[0140] In the claims, no reference numerals placed between parentheses should be construed as limiting the claims. The word “inclusion” does not preclude the existence of elements or steps not enumerated in the claims. The word “one” or “1” preceding an element does not preclude the existence of multiple such elements. The disclosure can be realized by hardware comprising multiple different elements and by a appropriately programmed computer. In a unit claim enumerating multiple devices, multiple of these devices may be concretely embodied by the same hardware item. The use of words such as “first,” “second,” “third,” etc., does not indicate any order. These words may be construed as names.

[0141] Finally, it should be noted that the above embodiments are merely for illustrative purposes and not to limit the technical solutions of the present disclosure. Although the present disclosure has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the above embodiments, or that some of the technical features therein can be replaced with equivalents. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure. [Explanation of symbols]

[0142] 100 Decoders 101 Information receiving module 102 Model Decision Module 103 Chromaticity Prediction Module 104 Second chromaticity prediction module 105 Reconfiguration Module 200 Encoding Devices 201 First Decision Module 202 Second Decision Module 203 Third Decision Module 204 Decision 4 Module 205 Fifth Decision Module 206 First Generation Module 1010 Processor 1020 memory 1030 storage area 1031 Program Code 1031' Computer-readable code

Claims

1. A decoding method, The system receives encoding information corresponding to a target encoding block, wherein the encoding information includes index information and parameter information of a first model, index information and parameter information of a second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, the first pixel is a pixel that satisfies a first condition in the target encoding block, the second pixel is a pixel that does not satisfy the first condition in the target encoding block, and the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range. Based on the index information and parameter information of the first model and the index information and parameter information of the second model, the first model and the second model are determined, the first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a linear mapping relationship, and the second model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship. Based on the luminance downsampling reconstruction value of the first pixel, the target chromaticity prediction value of the first pixel is determined through the first model, Based on the luminance downsampling reconstruction value of the second pixel, the target chromaticity prediction value of the second pixel is determined through the second model, This includes determining the chromaticity reconstruction value of the target encoded block based on the target chromaticity prediction value and target difference value of the first pixel, and the target chromaticity prediction value and target difference value of the second pixel, A decoding method characterized by the following features.

2. The decoding method further includes, when the second pixel is a first target pixel, determining a target chromaticity prediction value for the first target pixel through the first model based on the luminance downsampling reconstruction value of the first target pixel, Determining the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel is: If the second pixel is a second target pixel, the method includes determining a target chromaticity prediction value for the second target pixel through the second model based on the luminance downsampling reconstruction value of the second target pixel. Here, the first target pixel is a second pixel whose first prediction error is smaller than the second prediction error, the second target pixel is a second pixel whose first prediction error is greater than or equal to the second prediction error, the first prediction error is determined based on the chromaticity prediction value of the second pixel by the first model and the true chromaticity value of the second pixel, and the second prediction error is determined based on the chromaticity prediction value of the second pixel by the second model and the true chromaticity value of the second pixel. The decoding method according to feature 1.

3. The first model is obtained after parameter adjustments have been made to the initial first model and the chromaticity prediction value of the first pixel obtained by the initial first model. The decoding method according to feature 1.

4. The decoding method is, If the second pixel is a third target pixel, the method further includes determining a target chromaticity prediction value for the third target pixel through the first model based on the luminance downsampling reconstruction value of the third target pixel. Determining the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel is: If the second pixel is a fourth target pixel, the method includes determining a target chromaticity prediction value for the fourth target pixel through the second model based on the luminance downsampling reconstruction value of the fourth target pixel. Here, the third target pixel is the second pixel whose third prediction error is smaller than the fourth prediction error, the fourth target pixel is the second pixel whose third prediction error is greater than or equal to the fourth prediction error, the third prediction error is determined based on the chromaticity prediction value of the second pixel by the first model and the true chromaticity value of the second pixel, and the fourth prediction error is determined based on the chromaticity prediction value of the second pixel by the second model and the true chromaticity value of the second pixel. The decoding method according to feature 3.

5. Receiving encoding information corresponding to the target encoding block means The method includes receiving first encoding information corresponding to a target encoding block, generating the first encoding information when the number of second pixels exceeds a quantity threshold, and comprising index information and parameter information of a first model, index information and parameter information of a second model, a luminance downsampling reconstruction value of the first pixel and a target difference value of the first pixel, and a luminance downsampling reconstruction value of the second pixel and a target difference value of the second pixel. The decoding method is, If the first encoding information is not received, but the second encoding information corresponding to the target encoding block is received, the target chromaticity prediction value of the target encoding block is determined through the first model based on the luminance downsampling reconstruction value of the target encoding block. The method further includes determining the chromaticity reconstruction value of the target encode block based on the target chromaticity prediction value and target difference value of the target encode block, Here, the second encoding information is generated when the number of the second pixels is less than or equal to the quantity threshold, and the second encoding information includes the index information and parameter information of the first model, and the luminance downsampling reconstruction value and target difference value of the target encoding block. The decoding method according to feature 1.

6. The second model is constructed through a target mapping relationship based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel, and the target mapping relationship represents a nonlinear mapping relationship between the luminance downsampling reconstruction value and the chromaticity prediction value. The decoding method according to feature 1.

7. The second model is constructed based on a target formula and the values ​​of each parameter in the target formula, wherein the values ​​of each parameter in the target formula are determined based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel, and the target formula represents the target mapping relationship. The decoding method according to feature 6.

8. The aforementioned target expression is, [Math 1] Here, bitDepth is the bit depth, midValue is the midpoint of the bit depth, and pred Ｃ (i,j) is the predicted chromaticity value of the pixel at coordinate (i,j), [Math 2] is the luminance downsampling reconstruction value of the pixel at coordinate (i, j), and α ２ , α ３ , α ４ is the parameter value of the second model, and >> is a shift operation. The decoding method according to feature 7.

9. Based on the ratio of the first chromaticity prediction value of the pixel by the first model to the true chromaticity value, the chromaticity prediction error of the pixel by the first model is determined, or, Based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel according to the first model, the chromaticity prediction error of the pixel according to the first model is determined. The decoding method according to feature 1.

10. When the pixel chromaticity prediction error according to the first model is determined based on the ratio of the first chromaticity prediction value to the true chromaticity value, The first condition is that the pixel chromaticity prediction error by the first model satisfies equations 1 and 2, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 3, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 4. Here, the first equation is, A(i,j) <= threereadhold1, The second equation is, A(i,j) >= -threadhold1, The third equation above is, A(i,j) > threadhold1, The fourth equation is, A(i,j) < -threadhold1, Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold1 is the first threshold. The decoding method according to feature 9.

11. Determining the chromaticity prediction error of a pixel by the first model based on the ratio of the first chromaticity prediction value of the pixel by the first model to the true chromaticity value is: This includes determining the chromaticity prediction error of the pixel according to the first model through the fifth equation, Here, the fifth equation is, [Math 3] Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and pred Ｃ (i,j) is the predicted chromaticity of the pixel at coordinate (i,j), and C(i,j) is the true chromaticity of the pixel at coordinate (i,j). The decoding method according to feature 9.

12. When determining the chromaticity prediction error of a pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel by the first model, the first condition is that the chromaticity prediction error of the pixel by the first model satisfies equation 6. Here, equation 6 is, Abs(A(i,j))>=threadhold2 Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold2 is the second threshold. The decoding method according to feature 9.

13. Determining the chromaticity prediction error of a pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel by the first model is: This includes determining the chromaticity prediction error of the pixel according to the first model through equation 7, Here, the seventh equation is, [Math 4] Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and pred Ｃ (i,j) is the predicted chromaticity of the pixel at coordinate (i,j), and C(i,j) is the true chromaticity of the pixel at coordinate (i,j). The decoding method according to feature 9.

14. The first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to the linear mapping relationship of the following equation: [Math 5] And here, pred Ｃ (i,j) is the predicted chromaticity value of the pixel at coordinate (i,j), [Math 6] is the luminance downsampling reconstruction value of the pixel at the coordinates (i, j), and α ０ , α １ is the value of the parameter of the first model, The decoding method according to any one of claims 1 to 13.

15. The values ​​of each parameter in the first model and / or each parameter in the second model are determined based on a linear least-squares error estimate obtained by jointly mapping the true chromaticity value and the luminance downsampling reconstruction value of the target encoding block. The decoding method according to any one of claims 1 to 13.

16. It is an encoding method, Based on the luminance downsampling reconstruction value of the target encoding block, a first chromaticity prediction value of the target encoding block is determined through a first model, and the first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a linear mapping relationship, A first pixel that satisfies a first condition is determined from the target encoding block, and a second pixel that does not satisfy the first condition is determined from the target encoding block, wherein the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range, Based on the first chromaticity prediction value of the first pixel, the target chromaticity prediction value of the first pixel is determined, Based on the luminance downsampling reconstruction value of the second pixel, the target chromaticity prediction value of the second pixel is determined through the second model, and the second model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship. Based on the predicted target chromaticity value and the true chromaticity value of the first pixel, and the predicted target chromaticity value and the true chromaticity value of the second pixel, the target difference value of the first pixel and the target difference value of the second pixel are determined, respectively. The method includes generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel. An encoding method characterized by the following.

17. Determining the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel is: Based on the first predicted chromaticity value and the true chromaticity value of the second pixel, a first prediction error corresponding to the first model is determined, Based on the luminance downsampling reconstruction value of the second pixel, the second chromaticity prediction value of the second pixel is determined through the second model, and based on the second chromaticity prediction value and the true chromaticity value of the second pixel, the second prediction error corresponding to the second model is determined. To determine whether the first prediction error is smaller than the second prediction error, If the first prediction error is greater than or equal to the second prediction error, the target chromaticity prediction value of the second pixel is determined based on the second chromaticity prediction value of the second pixel, and the above includes: The aforementioned encoding method is The further step is to determine a target chromaticity prediction value for the second pixel based on a first chromaticity prediction value for the second pixel, if the first prediction error is smaller than the second prediction error. The encoding method according to claim 16.

18. Determining the target chromaticity prediction value of the first pixel based on the first chromaticity prediction value of the first pixel is: Based on the predicted first chromaticity value of the first pixel, the parameters of the first model are adjusted. This includes determining a target chromaticity prediction value for the first pixel through the first model, which has been parameter-adjusted based on the luminance downsampling reconstruction value of the first pixel, The encoding method according to feature 17.

19. Determining the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel is: Based on the luminance downsampling reconstruction value of the second pixel, the third chromaticity prediction value of the second pixel is determined through the parameter-adjusted first model, and based on the third chromaticity prediction value and the true chromaticity value of the second pixel, the third prediction error corresponding to the parameter-adjusted first model is determined. Based on the luminance downsampling reconstruction value of the second pixel, the fourth chromaticity prediction value of the second pixel is determined through the second model, and based on the fourth chromaticity prediction value and the true chromaticity value of the second pixel, the fourth prediction error corresponding to the second model is determined. To determine whether the third prediction error is smaller than the fourth prediction error, When the third prediction error is smaller than the fourth prediction error, the target chromaticity prediction value of the second pixel is determined based on the third chromaticity prediction value of the second pixel. If the third prediction error is greater than or equal to the fourth prediction error, the target chromaticity prediction value of the second pixel is determined based on the fourth chromaticity prediction value of the second pixel, including: The encoding method according to feature 18.

20. After determining a second pixel that does not satisfy the first condition from the target encoding block, the encoding method is as follows: To detect whether the number of the second pixels exceeds a quantity threshold, If the number of the second pixels is less than or equal to the quantity threshold, the target chromaticity prediction value of the target encoding block is determined based on the first chromaticity prediction value of the target encoding block. Based on the target chromaticity prediction value and the true chromaticity value of the target encoding block, the target difference value of the target encoding block is determined, The method further includes generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, and the luminance downsampling reconstruction value and target difference value of the target encoding block, Determining the target chromaticity prediction value of the first pixel based on the first chromaticity prediction value of the first pixel is: The process includes determining a target chromaticity prediction value for the first pixel based on the first chromaticity prediction value of the first pixel when the number of the second pixels exceeds the quantity threshold, Determining the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel is: If the number of the second pixels exceeds the quantity threshold, the method includes determining a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel. Determining the target difference value for the first pixel and the target difference value for the second pixel, respectively, based on the target chromaticity prediction value and the true chromaticity value of the first pixel, and the target chromaticity prediction value and the true chromaticity value of the second pixel, is as follows: If the number of the second pixels exceeds the quantity threshold, the system includes determining the target difference value for the first pixel and the target difference value for the second pixel, respectively, based on the target chromaticity prediction value and the true chromaticity value of the first pixel, and the target chromaticity prediction value and the true chromaticity value of the second pixel. Generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel is: When the number of the second pixels exceeds the quantity threshold, the method includes generating encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel. The encoding method according to claim 16.

21. Before determining the target chromaticity prediction value of the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel, the encoding method: Based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel, the second model is constructed through a target mapping relationship, further comprising the fact that the target mapping relationship shows a nonlinear mapping relationship between the luminance downsampling reconstruction value and the chromaticity prediction value. The encoding method according to claim 16.

22. Creating the second model through a target mapping relationship based on the luminance downsampling reconstruction value and chromaticity true value of the second pixel is: Based on the luminance downsampling reconstruction value and the true chromaticity value of the second pixel, the values ​​of each parameter in the target equation are determined, and the target equation represents the target mapping relationship. This includes creating the second model based on the target formula and the values ​​of each parameter. The encoding method according to feature 21.

23. The aforementioned target expression is, [Number 7] Here, bitDepth is the bit depth, midValue is the midpoint of the bit depth, and pred Ｃ (i,j) is the predicted chromaticity value of the pixel at coordinate (i,j), [Number 8] is the luminance downsampling reconstruction value of the pixel at coordinate (i, j), and α ２ , α ３ , α ４ is the parameter value of the second model, and >> is a shift operation. The encoding method according to feature 22.

24. Based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value, the chromaticity prediction error of the pixel by the first model is determined, or, Based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel, the chromaticity prediction error of the pixel by the first model is determined. The encoding method according to claim 16.

25. When the pixel chromaticity prediction error according to the first model is determined based on the ratio of the first chromaticity prediction value to the true chromaticity value, The first condition is that the pixel chromaticity prediction error by the first model satisfies equations 1 and 2, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 3, or the first condition is that the pixel chromaticity prediction error by the first model satisfies equation 4. Here, the first equation is, A(i,j) <= threereadhold1, The second equation is, A(i,j) >= -threadhold1, The third equation above is, A(i,j) > threadhold1, The fourth equation is, A(i,j) < -threadhold1, Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold1 is the first threshold. The encoding method according to feature 24.

26. Determining the chromaticity prediction error of a pixel by the first model based on the ratio of the first chromaticity prediction value of the pixel to the true chromaticity value is: This includes determining the chromaticity prediction error of the pixel according to the first model through the fifth equation, Here, the fifth equation is, [Number 9] Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and pred Ｃ (i,j) is the predicted chromaticity of the pixel at coordinate (i,j), and C(i,j) is the true chromaticity of the pixel at coordinate (i,j). The encoding method according to feature 24.

27. When determining the chromaticity prediction error of a pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel, the first condition is that the chromaticity prediction error of the pixel by the first model satisfies equation 6. Here, equation 6 is, Abs(A(i,j))>=threadhold2 Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and threadhold2 is the second threshold. The encoding method according to feature 24.

28. Determining the chromaticity prediction error of the pixel by the first model based on the difference between the first chromaticity prediction value and the true chromaticity value of the pixel is: This includes determining the chromaticity prediction error of the pixel according to the first model through equation 7, Here, the seventh equation is, [Number 10] Here, A(i,j) is the chromaticity prediction error of the pixel at coordinate (i,j) according to the first model, and pred Ｃ (i,j) is the predicted chromaticity of the pixel at coordinate (i,j), and C(i,j) is the true chromaticity of the pixel at coordinate (i,j). The encoding method according to feature 24.

29. The first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to the linear mapping relationship of the following equation: [Math 11] And here, pred Ｃ (i,j) is the predicted chromaticity value of the pixel at coordinate (i,j), [Math 12] is the luminance downsampling reconstruction value of the pixel at coordinate (i, j), and α ０ , α １ These are the parameter values ​​of the first model. The encoding method according to any one of claims 16 to 28.

30. The values ​​of each parameter in the first model and / or each parameter in the second model are determined based on a linear least-squares error estimate obtained by jointly mapping the true chromaticity value and the luminance downsampling reconstruction value of the target encoding block. The encoding method according to any one of claims 16 to 28.

31. A decoding device, An information receiving module receives encoding information corresponding to a target encoding block, wherein the encoding information includes index information and parameter information of a first model, index information and parameter information of a second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, wherein the first pixel is a pixel that satisfies a first condition in the target encoding block, and the second pixel is a pixel that does not satisfy the first condition in the target encoding block, and the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range. A model determination module that determines the first model and the second model based on the index information and parameter information of the first model and the index information and parameter information of the second model, the first model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a linear mapping relationship, and the second model determines the chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship, A first chromaticity prediction module that determines a target chromaticity prediction value for the first pixel through the first model based on the luminance downsampling reconstruction value of the first pixel, A second chromaticity prediction module that determines a target chromaticity prediction value for the second pixel through the second model based on the luminance downsampling reconstruction value of the second pixel, A reconstruction module that determines the chromaticity reconstruction value of the target encoding block based on the target chromaticity prediction value and target difference value of the first pixel, and the target chromaticity prediction value and target difference value of the second pixel, is included. A decoding device characterized by the following features.

32. An encoding device, A first determination module that determines a first chromaticity prediction value of a target encoding block through a first model based on the luminance downsampling reconstruction value of the target encoding block, and determines a chromaticity prediction value mapped from the luminance downsampling reconstruction value by the first model according to a linear mapping relationship, A first pixel that satisfies a first condition is determined from the target encoding block, and a second pixel that does not satisfy the first condition is determined from the target encoding block, wherein the first condition is that the chromaticity prediction error of the pixel by the first model is within a first threshold range for the second determination module, A third determination module that determines a target chromaticity prediction value for the first pixel based on the first chromaticity prediction value of the first pixel, A fourth decision module determines a target chromaticity prediction value for the second pixel through a second model based on the luminance downsampling reconstruction value of the second pixel, and the second model determines a chromaticity prediction value mapped from the luminance downsampling reconstruction value according to a nonlinear mapping relationship. A fifth determination module that determines the target difference value of the first pixel and the target difference value of the second pixel, respectively, based on the target chromaticity prediction value and the true chromaticity value of the first pixel, and the target chromaticity prediction value and the true chromaticity value of the second pixel, The system includes a first generation module that generates encoding information corresponding to the target encoding block based on the index information and parameter information of the first model, the index information and parameter information of the second model, the luminance downsampling reconstruction value of the first pixel and the target difference value of the first pixel, and the luminance downsampling reconstruction value of the second pixel and the target difference value of the second pixel, An encoding device characterized by the following.

33. A decoding device, Equipped with a processor and memory, The memory stores a program or instruction that can be executed on the processor, and when the program or instruction is executed by the processor, the decoding method described in any one of claims 1 to 15 is realized. A decoding device characterized by the following features.

34. An encoding device, Equipped with a processor and memory, The memory stores a program or instruction that can be executed on the processor, and when the program or instruction is executed by the processor, the encoding method described in any one of claims 16 to 30 is realized. An encoding device characterized by the following features.

35. A computing device, Memory containing computer-readable code, Equipped with one or more processors, When the computer-readable code is executed by one or more processors, the computing device performs the decoding method according to any one of claims 1 to 15, or the encoding method according to any one of claims 16 to 30. A computing device characterized by the following features.

36. It is a computer program, Includes computer-readable code, When the computer-readable code is executed on a computing device, the computing device is instructed to execute the decoding method described in any one of claims 1 to 15, or the encoding method described in any one of claims 16 to 30. A computer program characterized by the following features.

37. Computer-readable medium, A computer program according to claim 36 is stored in A computer-readable medium characterized by the following features.

Images