Prediction Method and Apparatus for Video Image Components, and Computer Storage Medium

Abstract

An embodiment of the present application discloses a method and apparatus for predicting video image components, and a computer storage medium. The prediction method includes: obtaining a first image component adjacent reference value and a first image component reconstruction value corresponding to a current coding block; wherein, the first image component adjacent reference value is used to represent a first image component parameter corresponding to an adjacent reference point of the current coding block, and the first image component reconstruction value is used to represent a reconstruction parameter of one or more first image components corresponding to the current coding block; determining a correlation coefficient according to the first image component adjacent reference value and the first image component reconstruction value; wherein, the correlation coefficient is used to represent the degree of deviation of the image component between the current coding block and the adjacent reference point; inputting the correlation coefficient into a preset weight calculation model to obtain a weight coefficient corresponding to the adjacent reference point; determining a scaling factor according to the weight coefficient; and obtaining a second image component prediction value corresponding to the current coding block through the scaling factor.

Description

[0001] This application is a divisional application of Chinese Patent Application No. 201880096257.8, with the filing date of August 9, 2018 and the invention title of "Prediction Method and Apparatus for Video Image Components, and Computer Storage Medium". Technical Field

[0002] Embodiments of the present application relate to the intra prediction technology in the field of video coding, and in particular, to a prediction method and apparatus for video image components, and a computer storage medium. Background Art

[0003] In the next-generation video coding standard H.266 or Versatile Video Coding (VVC), prediction between luminance values and chrominance values or between chrominance values can be achieved through the Cross-component Linear Model Prediction (CCLM). Specifically, a linear model can be constructed for the adjacent luminance parameters and chrominance parameters corresponding to the current coding block by using the linear regression method, so that the chrominance prediction value can be calculated according to the linear model and the reconstructed luminance value.

[0004] However, when constructing the linear model, the currently considered factors are not comprehensive enough. For example, when predicting chrominance values with luminance values, mostly the factors of adjacent reference luminance values and adjacent reference chrominance values are considered; when predicting between chrominance values, mostly the factors of adjacent reference chrominance components are considered. Therefore, once there is a large deviation between the adjacent reference luminance value and the parameters corresponding to the current coding block, or there is a large deviation between the adjacent reference chrominance component and the parameters corresponding to the current coding block, it will cause the calculated linear model to deviate from the expected model, thereby reducing the prediction accuracy of the chrominance prediction value of the current coding block and causing a large deviation between the chrominance prediction value and the true chrominance value. Summary of the Invention

[0005] To solve the above technical problems, embodiments of the present application are expected to provide a prediction method and apparatus for video image components, and a computer storage medium, which can effectively improve the prediction accuracy of the second image component prediction value and make the second image component prediction value closer to the true component value.

[0006] The technical solution of the embodiments of the present application is implemented as follows:

[0007] A prediction method for video image components, the method includes:

[0008] Obtain the first image component adjacent reference value and the first image component reconstruction value corresponding to the current coding block; wherein, the first image component adjacent reference value is used to represent the first image component parameter corresponding to the adjacent reference point of the current coding block, and the first image component reconstruction value is used to represent the reconstruction parameter of one or more first image components corresponding to the current coding block;

[0009] Determine a correlation coefficient according to the first image component adjacent reference value and the first image component reconstruction value; wherein, the correlation coefficient is used to represent the deviation degree of the image components between the current coding block and the adjacent reference point;

[0010] Input the correlation coefficient into a preset weight calculation model to obtain the weight coefficient corresponding to the adjacent reference point;

[0011] Determine a scale factor according to the weight coefficient;

[0012] Obtain the predicted value of the second image component corresponding to the current coding block through the scale factor.

[0013] The embodiments of the present application provide a method and device for predicting video image components, and a computer storage medium. The prediction device obtains the first image component adjacent reference value and the first image component reconstruction value corresponding to the current coding block; wherein, the first image component adjacent reference value is used to represent the first image component parameter corresponding to the adjacent reference point of the current coding block, and the first image component reconstruction value is used to represent the reconstruction parameter of one or more first image components corresponding to the current coding block; determine a correlation coefficient according to the first image component adjacent reference value and the first image component reconstruction value; wherein, the correlation coefficient is used to represent the deviation degree of the image components between the current coding block and the adjacent reference point; input the correlation coefficient into a preset weight calculation model to obtain the weight coefficient corresponding to the adjacent reference point; determine a scale factor according to the weight coefficient; obtain the predicted value of the second image component corresponding to the current coding block through the scale factor. It can be seen that in the embodiments of the present application, the prediction device can determine the correlation coefficient based on the first image component adjacent reference value and the first image component reconstruction value corresponding to the current coding block, so as to realize the distribution of different weight coefficients for different adjacent reference points according to the correlation of the component parameters between the adjacent reference point and the current coding block, so as to construct a linear model that better conforms to the expected model, thereby effectively overcoming the defect that the linear model deviates from the expected model when there is a large deviation between the first image component adjacent reference value and the component parameters corresponding to the current coding block, or when there is a large deviation between the third image component adjacent reference value and the component parameters corresponding to the current coding block. Furthermore, when predicting the components of the current coding block according to this linear model, the prediction accuracy of the predicted value of the video image component can be greatly improved, and the predicted value of the video image component is closer to the true value of the video image component. Description of the Drawings

[0014] Figure 1 It is a schematic diagram of the video encoding process;

[0015] Figure 2 It is a schematic diagram of the video decoding process;

[0016] Figure 3 It is a schematic diagram of the positions of adjacent reference samples Figure 1 ;

[0017] Figure 4 It is a schematic diagram of the positions of adjacent reference samples Figure 2 ;

[0018] Figure 5 It is a schematic diagram of determining a linear model in the prior art Figure 1 ;

[0019] Figure 6 It is a schematic diagram of determining a linear model in the prior art Figure 2 ;

[0020] Figure 7 It is a schematic diagram of the implementation process of a method for predicting video image components proposed in an embodiment of the present application;

[0021] Figure 8 It is a schematic diagram of removing interference points in an embodiment of the present application;

[0022] Figure 9 It is a schematic diagram of the composition structure of a prediction device proposed in an embodiment of the present application Figure 1 ;

[0023] Figure 10 It is a schematic diagram of the composition structure of a prediction device proposed in an embodiment of the present application Figure 2 . Detailed Embodiments

[0024] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. It can be understood that the specific embodiments described herein are only used to explain the related application, rather than limiting the application. Additionally, it should be noted that for the sake of description, only the parts related to the relevant application are shown in the drawings.

[0025] In a video image, generally, the first image component, the second image component, and the third image component are used to represent an encoding block; among them, the first image component, the second image component, and the third image component may include a luminance component and two chrominance components. Specifically, the luminance component is usually represented by the symbol Y, and the chrominance components are usually represented by the symbols Cb and Cr, where Cb represents the blue chrominance component and Cr represents the red chrominance component.

[0026] It should be noted that in the embodiments of the present application, the first image component, the second image component, and the third image component may be the luminance component Y, the blue chrominance component Cb, and the red chrominance component Cr respectively. For example, the first image component may be the luminance component Y, the second image component may be the red chrominance component Cr, and the third image component may be the blue chrominance component Cb. The embodiments of the present application do not make specific limitations in this regard.

[0027] Furthermore, in the embodiments of the present application, the commonly used sampling format in which the luminance component and the chrominance component are separately represented is also referred to as the YCbCr format. Among them, the YCbCr format may include the 4:4:4 format, the 4:2:2 format, and the 4:2:0 format.

[0028] In the case where the video image adopts the YCbCr 4:2:0 format, if the luminance component of the video image is a coding block of size 2N×2N, the corresponding chrominance component is a coding block of size N×N, where N is the side length of the coding block. In the embodiments of the present invention, the 4:2:0 format will be taken as an example for description, but the technical solutions of the embodiments of the present invention are equally applicable to other sampling formats.

[0029] In the embodiments of the present application, the above-mentioned prediction method for video image components can be applied to the intra prediction part in the video coding hybrid framework. Specifically, the above-mentioned prediction method for video image components can act on both the coding end and the decoding end. For example, Figure 1 is a schematic diagram of the video coding process, as Figure 1 shown, video coding may include multiple specific steps such as intra estimation, intra prediction, and motion compensation. Among them, the prediction method for video image components proposed in the present application can be applied to the intra prediction part; Figure 2 is a schematic diagram of the video decoding process, as Figure 2 shown, video decoding may include multiple specific steps such as filtering, intra prediction, and motion compensation. Among them, the prediction method for video image components proposed in the present application can be applied to the intra prediction part.

[0030] In H.266, in order to further improve the coding performance and coding efficiency, the cross-component prediction (CCP) was extended and improved, and the cross-component linear model prediction (CCLM) was proposed. In H.266, CCLM realizes the prediction between the first image component and the second image component, the first image component and the third image component, and the second image component and the third image component. The following will take the prediction from the first image component to the second image component as an example for description, but the technical solutions of the embodiments of the present application can equally apply to the prediction of other image components.

[0031] Specifically, in the prior art, when the CCLM method realizes the prediction from the luminance component to the chrominance component, in order to reduce the redundancy between the luminance component and the chrominance component and between different chrominance components, an inter-component linear model prediction mode is used in the decoder of the next-generation video coding standard, such as the Joint Exploration Model (JEM) of H.266 / VVC or the VVC Test Model (VTM). For example, according to formula (1), the reconstructed luminance value of the same coding block is used to construct the predicted value of chrominance:

[0032] Pred C [i,j] = α·Rec L [i,j] + β (1)

[0033] where i and j represent the position coordinates of the sampling points in the coding block, i represents the horizontal direction, and j represents the vertical direction. Pred C [i,j] represents the predicted value of the second image component of the sampling point with the position coordinates [i,j] in the coding block, and Rec L [i,j] represents the reconstructed value of the first image component of the sampling point with the position coordinates [i,j] in the same coding block (after downsampling). α and β are the scale factors of the linear model and can be derived by minimizing the regression error between the adjacent reference values of the first image component and the adjacent reference values of the second image component, as shown in the following formula (2):

[0034]

[0035] where L(n) represents the adjacent reference values of the first image component after downsampling (such as the left and upper sides), C(n) represents the adjacent reference values of the second image component (such as the left and upper sides), and N is the number of adjacent reference values of the second image component. Figure 3 Schematic diagram of the positions of adjacent reference samplings Figure 1 , Figure 4 Schematic diagram of the positions of adjacent reference samplings Figure 2 , such as Figure 3 and Figure 4 shown. L(n) and C(n) are adjacent reference pixel points, and N is the number of adjacent reference values of the second image component. Among them, for a video image in the 4:2:0 format, for a first image component coding block of size 2N×2N, as Figure 3 shown, the corresponding size of the second image component is N×N, as Figure 4As shown. For square coding blocks, these two equations can be directly applied. For non-square coding blocks, the neighboring samples on the longer edge are first downsampled to obtain the same number of samples as those on the shorter edge. α and β do not need to be transmitted and are also calculated by formula (2) in the decoder; in the embodiments of the present application, no specific limitation is made in this regard.

[0036] Figure 5 Schematic diagram for determining the linear model in the prior art Figure 1 as Figure 5 shown, a, b, and c are adjacent reference values of the first image component, A, B, and C are adjacent reference values of the second image component, e is the reconstructed value of the first image component of the current coding block, and E is the predicted value of the second image component of the current coding block; among them, using all the adjacent reference values L(n) of the first image component and the adjacent reference values C(n) of the second image component of the current coding block, α and β can be calculated according to formula (2); then, substituting the reconstructed value e of the first image component of the current coding block into the linear model described in formula (1), the predicted value E of the second image component of the current coding block can be calculated. α and β; then, substituting the reconstructed value e of the first image component of the current coding block into the linear model described in formula (1), the predicted value E of the second image component of the current coding block can be calculated.

[0037] Specifically, in the prior art, in addition to the method of predicting the chrominance component with the luminance component in the CCLM prediction mode, that is, the method of predicting the second image component with the first image component or predicting the third image component with the first image component, it also includes the prediction between two chrominance components, that is, it also includes the prediction method between the second image component and the third image component. Among them, in the embodiments of the present application, the Cr component can be predicted from the Cb component, or the Cb component can be predicted from the Cr component.

[0038] It should be noted that in the embodiments of the present application, the prediction between chrominance components in the CCLM, that is, the prediction between the second image component and the third image component, can be applied to the residual domain. Taking the prediction of the Cr component as an example, the Cb residual can be used to predict the Cr residual. The final predicted value of the Cr component is obtained by adding a weighted reconstructed Cb residual to the traditional intra-frame predicted value of the Cr component, as shown in formula (3):

[0039] Pred* Cr [i,j] = γ·resi Cb '[i,j] + Pred Cr [i,j] (3)

[0040] Among them, represents the final predicted value of the Cr component at the sampling point with the position coordinates [i,j] in the above-mentioned current coding block, resi Cb'[i,j] is the prediction residual of the reconstructed Cb component. The calculation method of the scaling factor γ is the same as that of the prediction model parameters of the luminance component to the chrominance component in CCLM. The only difference is that a regression cost related to the default γ value in the error function is added, so that the obtained scaling factor γ tends to the default value of -0.5. Specifically, the scaling factor γ can be calculated by formula (4).

[0041]

[0042] Among them, Cb(n) represents the adjacent reference Cb value of the current coding block, Cr(n) represents the adjacent reference Cr value of the current coding block, and λ can be an empirical value. For example, λ = ∑(Cb(n)·Cb(n)) >> 9.

[0043] However, when calculating the linear model according to the first image component to predict the second image component or the third image component in the prior art, only the adjacent reference values of the first image component, the adjacent reference values of the second image component, or the adjacent reference values of the third image component are often considered. For example, as shown above Figure 5 shown, the scaling factors α and β are calculated by using all the adjacent reference values L(n) of the first image component and all the adjacent reference values C(n) of the second image component. Then, the reconstructed value of the first image component of the current coding block is substituted into the linear model to obtain the predicted value of the second image component of the current coding block. In this process, the correlation between the reconstructed value of the first image component of the current coding block and the adjacent reference values of the first image component is not considered. If the adjacent reference values of the first image component that deviate significantly from the current luminance value are used to construct the linear model, it may cause the predicted value of the second image component or the third image component of the current coding block to deviate far from the actual image component value, reducing the prediction accuracy. For example, Figure 6 is a schematic diagram of determining the linear model in the prior art Figure 2 , as Figure 6 shown, the pixel points b and c deviate from the reconstructed luminance value of the current coding block. Their corresponding relationship with the chrominance values deviates from the linear relationship between the reconstructed luminance value of the current coding block and the corresponding chrominance values. According to the calculation method of the prior art, the two pixel points b and c will also be used to calculate the parameters of the linear model, resulting in the derived linear model deviating from the linear model we expect to obtain.

[0044] In an embodiment of the present application, a prediction method for video image components proposes a calculation method for a linear model based on component correlation. Specifically, when calculating the scale factor of the linear model, not only the adjacent reference values of the first image component, the adjacent reference values of the second image component, or the adjacent reference values of the third image component are considered, but also the correlation and similarity degree between the reconstructed value of the first image component of the current coding block and the adjacent reference value of the first image component are considered. Then, the adjacent reference pixel points are weighted or screened according to the similarity degree, so that the linear model parameters are more suitable for the reconstructed value of the current first image component, and further the predicted value of the second image component or the third image component of the current coding block is more accurate.

[0045] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Among them, in the following embodiments, the first image component may be a luminance component Y, the second image component may be a red chrominance component Cr, and the third image component may be a blue chrominance component Cb. The embodiments of the present application do not make specific limitations on this.

[0046] Embodiment 1

[0047] The embodiment of the present application provides a prediction method for video image components. Figure 7 It is a schematic flowchart of the implementation of a prediction method for video image components proposed in the embodiment of the present application. As Figure 7 shown, in the embodiment of the present application, the method for the above prediction device to predict video image components may include the following steps:

[0048] Step 101: Obtain the adjacent reference value of the first image component corresponding to the current coding block and the reconstructed value of the first image component; wherein, the adjacent reference value of the first image component is used to represent the first image component parameter corresponding to the adjacent reference point of the current coding block, and the reconstructed value of the first image component is used to represent the reconstruction parameter of one or more first image components corresponding to the current coding block.

[0049] In the embodiment of the present application, the above prediction device may first obtain the adjacent reference value of the first image component corresponding to the current coding block and the reconstructed value of the first image component. It should be noted that, in the embodiment of the present application, the above current coding block is a coding block divided by the above prediction device and including at least one pixel point.

[0050] It should be noted that, in the embodiments of the present application, the above-mentioned first image component adjacent reference value is used to represent the first image component parameter corresponding to the adjacent reference point of the above-mentioned current coding block. Among them, for the downsampling method, when the above-mentioned current coding block contains one pixel point, the corresponding adjacent reference points can be one pixel point above and to the left of this one pixel point; when the above-mentioned current coding block contains multiple pixel points, the corresponding adjacent reference points can be multiple pixel points above and to the left of these multiple pixel points.

[0051] Furthermore, in the embodiments of the present application, the above-mentioned adjacent reference points are pixel points adjacent to the above-mentioned current coding block. Specifically, in the embodiments of the present application, for the downsampling method, the above-mentioned adjacent reference points can be adjacent pixel points located to the left and above the above-mentioned current coding block.

[0052] It should be noted that, in the embodiments of the present application, when the above-mentioned prediction device obtains the above-mentioned first image component adjacent reference value, it can obtain all the first image component adjacent reference values corresponding to the above-mentioned adjacent reference points. Specifically, in the embodiments of the present application, one first image component adjacent reference value corresponding to one adjacent reference point, that is, when the above-mentioned adjacent reference points are multiple pixel points, the above-mentioned prediction device can obtain multiple first image component adjacent reference values corresponding to the above-mentioned multiple pixel points.

[0053] Furthermore, in the embodiments of the present application, the above-mentioned first image component reconstruction value is used to represent the reconstruction parameter of one or more first image components corresponding to the above-mentioned current coding block. Specifically, in the embodiments of the present application, when the above-mentioned current coding block contains one pixel point, the reconstruction parameter of the corresponding first image component is one; when the above-mentioned current coding block contains multiple pixel points, the reconstruction parameters of the corresponding first image components are multiple.

[0054] It should be noted that, in the embodiments of the present application, when the above-mentioned prediction device obtains the above-mentioned first image component reconstruction value, it can obtain all the first image component reconstruction values corresponding to the above-mentioned current coding block. Specifically, in the embodiments of the present application, one first image component reconstruction value corresponding to one pixel point in the above-mentioned current coding block.

[0055] Step 102: Determine a correlation coefficient according to the first image component adjacent reference value and the first image component reconstruction value; wherein, the correlation coefficient is used to represent the degree of deviation of the image components between the current coding block and the adjacent reference points.

[0056] In the embodiments of the present application, after the above-mentioned prediction device obtains the above-mentioned first image component adjacent reference value and the above-mentioned first image component reconstruction value corresponding to the above-mentioned current coding block, it can further determine a correlation coefficient according to the above-mentioned first image component adjacent reference value and the above-mentioned first image component reconstruction value.

[0057] It should be noted that, in the embodiments of the present application, the above correlation coefficient can be used to characterize the deviation degree of the image components between the above current coding block and the above adjacent reference points. Specifically, the above correlation coefficient can be used to characterize the correlation between the reconstructed value of the above first image component corresponding to the above current coding block and the adjacent reference value of the above first image component corresponding to the above adjacent reference point.

[0058] Furthermore, in the embodiments of the present application, when the above prediction device determines the above correlation coefficient according to the above adjacent reference value of the first image component and the above reconstructed value of the first image component, it can calculate the difference between the above adjacent reference value of the first image component and the above reconstructed value of the first image component, so as to further determine the above correlation coefficient according to the difference result; at the same time, when the above prediction device determines the above correlation coefficient according to the above adjacent reference value of the first image component and the above reconstructed value of the first image component, it can also perform matrix multiplication on the matrix corresponding to the above adjacent reference value of the first image component and the matrix corresponding to the above reconstructed value of the first image component, so as to further determine the above correlation coefficient according to the product result.

[0059] Step 103: Input the correlation coefficient into a preset weight calculation model to obtain the weight coefficient corresponding to the adjacent reference point.

[0060] In the embodiments of the present application, after the above prediction device determines the above correlation coefficient according to the above adjacent reference value of the first image component and the above reconstructed value of the first image component, it can input the above correlation coefficient into a preset weight calculation model, so as to obtain the weight coefficient corresponding to the above adjacent reference point.

[0061] It should be noted that, in the embodiments of the present application, the above preset weight calculation model is a calculation model preset by the above prediction device for weight allocation according to the above correlation coefficient.

[0062] Furthermore, in the embodiments of the present application, the above weight coefficient can be a weight value calculated according to the correlation between the above adjacent reference point and the above current coding block when predicting the predicted value of the second image component corresponding to the above current coding block.

[0063] It should be noted that, in the embodiments of the present application, the greater the correlation between one of the above adjacent reference points and the above current coding block, the greater the corresponding weight value of the above one reference point.

[0064] It should be noted that, in the embodiments of the present application, when calculating the predicted value of the second image component, any one of the above adjacent reference points has its corresponding above weight coefficient.

[0065] Furthermore, in the embodiments of the present application, when determining the above weight coefficients, the above prediction device may perform calculations through the above preset weight calculation model. Specifically, in the embodiments of the present application, the above weight calculation model is shown in the following formula (5):

[0066]

[0067] Wherein, L(n) represents the adjacent reference values of the first image component on the left and upper sides of the downsampling, L (i,j) is the reconstructed value of the current first image component, and σ is a value related to the quantization parameter in the encoding.

[0068] Step 104: Determine the scale factor according to the weight coefficient.

[0069] In the embodiments of the present application, after the above prediction device inputs the above correlation coefficient into the preset weight calculation model and obtains the above weight coefficient corresponding to the above adjacent reference point, it may determine the scale factor according to the above weight coefficient.

[0070] It should be noted that, in the embodiments of the present application, the above scale factor may be the coefficient in the linear model obtained when the above prediction device performs video image component prediction.

[0071] Furthermore, in the embodiments of the present application, for different prediction modes, the above scale factors are different coefficients. For example, if the above preset device predicts the second image component or the third image component through the first image component, the above scale factor may be the α and β in the above formula (1); if the above preset device performs prediction between the second image component and the third image component, the above scale factor may be γ in the above formula (3).

[0072] Step 105: Obtain the predicted value of the second image component corresponding to the current coding block through the scale factor.

[0073] In the embodiments of the present application, after the above prediction device determines the above scale factor according to the above weight coefficient, it may further obtain the predicted value of the second image component corresponding to the above current coding block through the above scale factor. Wherein, the above predicted value of the second image component is the prediction result obtained by the above prediction device for performing the second component prediction on the above current coding block.

[0074] It should be noted that, in the embodiments of the present application, when the above prediction device obtains the above predicted value of the second image component according to the above scale factor, it may first establish a linear model for video image component prediction according to the above scale factor.

[0075] Further, in the embodiments of the present application, for different prediction modes, the linear models established by the above prediction device are also different. That is, the linear model established when predicting the second image component or the third image component through the first image component is different from the linear model established when predicting between the second image component and the third image component.

[0076] It should be noted that, in the embodiments of the present application, the predicted value of the second image component obtained when predicting the second image component or the third image component through the first image component is the second image component or the third image component corresponding to the current coding block, that is, the Cb component or the Cr component; the predicted value of the second image component obtained when predicting between the second image component and the third image component is the second image component or the third image component corresponding to the current coding block, that is, the Cb component or the Cr component.

[0077] Further, in the embodiments of the present application, after the above prediction device establishes the linear model according to the above scale factor, it can perform video image component prediction on the current coding block according to the linear model, so as to obtain the predicted value of the second image component.

[0078] Further, in the embodiments of the present application, there are currently two prediction modes for CCLM: one is the prediction mode of single-model CCLM; the other is the prediction mode of multi-model CCLM (Multiple Model CCLM, MMLM), also known as the prediction mode of MMLM. As the name implies, the prediction mode of single-model CCLM is that there is only one linear model to predict the second image component or the third image component from the first image component; while the prediction mode of MMLM is that there are multiple linear models to predict the second image component or the third image component from the first image component. For example, in the prediction mode of MMLM, the adjacent reference value of the first image component and the adjacent reference value of the second image component of the current coding block are divided into two groups, and each group can be used as a training set for deriving the parameters of the linear model alone, that is, each group can derive a set of scale factors. Therefore, the prediction method of the video image component proposed in the embodiments of the present application can also be applied to the prediction mode of MMLM, and can also overcome the defect that the linear model deviates from the expected model when there is a large deviation between the adjacent reference value of the first image component and the parameters corresponding to the current coding block, or when there is a large deviation between the adjacent reference value of the third image component and the parameters corresponding to the current coding block. Furthermore, when predicting the second image component of the current coding block according to the linear model, the prediction accuracy of the predicted value of the second image component can be greatly improved.

[0079] It can be seen that in the embodiments of the present application, in the embodiments of the present application, the prediction device can determine the correlation coefficient based on the adjacent reference value of the first image component corresponding to the current coding block and the reconstructed value of the first image component. Thus, it is possible to assign different weight coefficients to different adjacent reference points according to the correlation between the adjacent reference points and the component parameters of the current coding block, so as to construct a linear model that better conforms to the expected model. Therefore, when there is a large deviation between the adjacent reference value of the first image component and the component parameters corresponding to the current coding block, or when there is a large deviation between the adjacent reference value of the third image component and the component parameters corresponding to the current coding block, the defect that the linear model deviates from the expected model can be effectively overcome. Furthermore, when predicting the components of the current coding block according to this linear model, the prediction accuracy of the predicted value of the video image component can be greatly improved, making the predicted value of the video image component closer to the true value of the video image component.

[0080] Embodiment 2

[0081] Based on the above Embodiment 1, in the embodiments of the present application, the method for the prediction device to determine the correlation coefficient according to the adjacent reference value of the first image component and the reconstructed value of the first image component may include the following steps:

[0082] Step 102a: Perform a difference operation on any one of the adjacent reference values of the first image component and each reconstructed value of the first image component to obtain the component difference corresponding to any one of the reference values; wherein, one adjacent reference value of the first image component and one reconstructed value of the first image component correspond to one difference.

[0083] In the embodiments of the present application, after the prediction device obtains the adjacent reference value of the first image component corresponding to the current coding block and the reconstructed value of the first image component, it may perform a difference operation on any one of the adjacent reference values of the first image component and each of the reconstructed values of the first image component to obtain the component difference corresponding to any one of the reference values.

[0084] It should be noted that in the embodiments of the present application, one reference value in the adjacent reference values of the first image component and one reconstructed value in the reconstructed values of the first image component correspond to one difference.

[0085] Furthermore, in the embodiments of the present application, the prediction device performing a difference operation on any one of the adjacent reference values of the first image component and each of the reconstructed values of the first image component can represent the correlation degree of the video image components between the adjacent reference points and the current coding module.

[0086] It should be noted that, in the embodiments of the present application, the larger the above-mentioned component difference is, the smaller the correlation degree between the above-mentioned adjacent reference point and the above-mentioned current coding module can be considered.

[0087] Step 102b: Determine the smallest difference in the component differences as the correlation coefficient.

[0088] In the embodiments of the present application, after the above-mentioned prediction device performs a difference operation on any one of the adjacent reference values of the above-mentioned first image component and each reconstructed value of the first image component to obtain the above-mentioned component difference corresponding to any one of the reference values, the prediction device may determine the smallest difference in the above-mentioned component differences as the above-mentioned correlation coefficient.

[0089] It should be noted that, in the embodiments of the present application, after the above-mentioned prediction device determines the smallest difference from the above-mentioned component differences, it may further determine the above-mentioned correlation coefficient according to the smallest difference.

[0090] Furthermore, in the embodiments of the present application, the smaller the above-mentioned component difference is, the greater the correlation degree between the above-mentioned adjacent reference point and the above-mentioned current coding module can be considered. Therefore, the above-mentioned prediction device may determine the above-mentioned correlation factor according to the smallest difference in the above-mentioned component differences.

[0091] Furthermore, in the embodiments of the present application, after the above-mentioned prediction device performs a difference operation on any one of the adjacent reference values of the above-mentioned first image component and each reconstructed value of the above-mentioned first image component to obtain the component difference corresponding to any one of the reference values, and before determining the scale factor according to the above-mentioned weight coefficient, the method for the above-mentioned prediction device to perform video image component prediction may further include the following steps:

[0092] Step 106: When the component difference corresponding to any one of the reference values is greater than a preset difference threshold, set the weight coefficient of the adjacent reference point corresponding to any one of the reference values to zero.

[0093] In the embodiments of the present application, after the above-mentioned prediction device performs a difference operation on any one of the adjacent reference values of the above-mentioned first image component and each reconstructed value of the above-mentioned first image component to obtain the above-mentioned component difference corresponding to any one of the reference values, and before determining the above-mentioned scale factor according to the above-mentioned weight coefficient, if the above-mentioned component difference corresponding to any one of the reference values is greater than the preset difference threshold, then the above-mentioned prediction device may set the above-mentioned weight coefficient of the adjacent reference point corresponding to any one of the reference values to zero.

[0094] It should be noted that in the embodiments of the invention, after the prediction device obtains the component difference corresponding to any one of the above reference values, the component difference can be compared with the preset difference threshold. If all the component differences corresponding to any one of the above reference values are greater than the preset difference threshold, it can be considered that there is a large deviation between the adjacent reference values of any one of the first image components and the current coding block. Therefore, the prediction device can remove the adjacent reference points corresponding to any one of the first image component adjacent reference values, that is, set the weight coefficient of the adjacent reference points corresponding to any one of the reference values to zero.

[0095] Furthermore, in the embodiments of the invention, before calculating the linear model, the prediction device can consider the correlation between the reconstructed value of the first image component of the current coding block and the adjacent reference value of the first image component. That is, the adjacent reference values of the first image component are screened according to the difference between the adjacent reference value of the first image component and the reconstructed value of the first image component.

[0096] Furthermore, in the embodiments of the invention, the prediction device can preset a threshold, that is, the preset difference threshold, and then traverse each reference value in the adjacent reference values of the first image component, and calculate the difference from each reconstructed value of the first image component in the current coding block respectively. If the minimum value of the differences between a certain adjacent reference value of the first image component adjacent reference values and each reconstructed value of the first image component is greater than the preset difference threshold, the prediction device can consider that this adjacent reference point is an interference point for calculating an accurate linear model. Therefore, it can be removed from the training samples of the linear model, so that the adjacent reference values of the first image component after removing the interference points can be used as the training samples of the linear model to calculate the model parameters.

[0097] In the embodiments of the invention, Figure 8 is a schematic diagram of removing interference points in the embodiments of the present application. As Figure 8 shown, if some of the first image component adjacent reference values of the current coding block deviate from the reconstructed value of the first image component of the current coding block, when calculating the linear model, this point is removed and not used as a training sample of the linear model. At this time, the calculated linear model is more accurate.

[0098] It can be seen that in the embodiments of the present application, the prediction device can determine the correlation coefficient based on the adjacent reference value of the first image component corresponding to the current coding block and the reconstructed value of the first image component. Thus, it is possible to assign different weight coefficients to different adjacent reference points according to the correlation between the adjacent reference points and the component parameters of the current coding block, so as to construct a linear model that better conforms to the expected model. Therefore, when there is a large deviation between the adjacent reference value of the first image component and the component parameters corresponding to the current coding block, or when there is a large deviation between the adjacent reference value of the third image component and the component parameters corresponding to the current coding block, the defect that the linear model deviates from the expected model can be effectively overcome. Furthermore, when predicting the components of the current coding block according to this linear model, the prediction accuracy of the predicted value of the video image component can be greatly improved, and the predicted value of the video image component can be closer to the true value of the video image component.

[0099] Embodiment 3

[0100] Based on the above Embodiment 1, in the embodiments of the present application, before the above prediction device determines the scale factor according to the above weight coefficient, that is, before step 104, the method for the above prediction device to perform video image component prediction may further include the following steps:

[0101] Step 107: Obtain the adjacent reference value of the second image component corresponding to the current coding block; wherein, the adjacent reference value of the second image component is the second component parameter corresponding to the adjacent reference point and different from the first image component parameter.

[0102] In the embodiments of the present application, before the above prediction device determines the above scale factor according to the above weight coefficient, it may first obtain the adjacent reference value of the second image component corresponding to the current coding block.

[0103] It should be noted that in the embodiments of the present application, the adjacent reference value of the second image component is the second component parameter corresponding to the adjacent reference point and different from the first image component parameter.

[0104] In the embodiments of the present application, the scale factor includes a first scale parameter and a second scale parameter. The method for the above prediction device to determine the scale factor according to the above weight coefficient may include the following steps:

[0105] Step 104a: Input the weight coefficient, the adjacent reference value of the first image component, and the adjacent reference value of the second image component into the first preset factor calculation model to obtain the first scale parameter.

[0106] In an embodiment of the present invention, after the prediction device inputs the correlation coefficient into a preset weight calculation model and obtains the weight coefficient corresponding to the adjacent reference value of the first image component, the weight coefficient, the adjacent reference value of the first image component, and the adjacent reference value of the second image component can be input into a first preset factor calculation model to obtain the first proportionality parameter.

[0107] It should be noted that, in an embodiment of the present invention, when the preset device predicts the second image component through the first image component, the proportionality factor may include the first proportionality parameter and the second proportionality parameter, wherein the first proportionality parameter and the second proportionality parameter are used to construct the predicted value of the second image component according to the reconstructed value of the first image component.

[0108] Furthermore, in an embodiment of the present application, when the prediction device further determines the calculation of the first proportionality parameter according to the adjacent reference value of the first image component and the adjacent reference value of the second image component, the first preset factor calculation model that can be used is as shown in the following formula (6):

[0109]

[0110] where α is the first proportionality parameter, L(n) represents the adjacent reference values of the first image component on the left and upper sides after downsampling, C(n) represents the adjacent reference values of the second image component on the left and upper sides, and w(n) is the weight coefficient corresponding to each adjacent reference value of the first image component.

[0111] Step 104b: Input the weight coefficient, the first proportionality parameter, the adjacent reference value of the first image component, and the adjacent reference value of the second image component into a second preset factor calculation model to obtain the second proportionality parameter.

[0112] In an embodiment of the present application, after the prediction device inputs the weight coefficient, the adjacent reference value of the first image component, and the adjacent reference value of the second image component into the first preset factor calculation model and obtains the first proportionality parameter, the weight coefficient, the first proportionality parameter, the adjacent reference value of the first image component, and the adjacent reference value of the second image component can be continuously input into a second preset factor calculation model, so as to obtain the second proportionality parameter.

[0113] Furthermore, in an embodiment of the present application, when the prediction device further determines the calculation of the second proportionality parameter according to the adjacent reference value of the first image component and the adjacent reference value of the second image component, the second preset factor calculation model that can be used is as shown in the following formula (7):

[0114]

[0115] Wherein, β is the second proportional parameter, L(n) represents the adjacent reference values of the first image component on the left and upper sides of the downsampling, C(n) represents the adjacent reference values of the second image component on the left and upper sides, and w(n) is the weight coefficient corresponding to each adjacent reference point.

[0116] In an embodiment of the present application, further, before the prediction device determines the proportional factor according to the weight coefficient, that is, before step 104, the method for the prediction device to perform video image component prediction may further include the following steps:

[0117] Step 108: Obtain the adjacent reference values of the second image component and the adjacent reference values of the third image component corresponding to the current coding block; wherein, the adjacent reference values of the second image component and the adjacent reference values of the third image component respectively represent the second image component parameters and the third image component parameters of the adjacent reference points.

[0118] In an embodiment of the present application, before the prediction device determines the proportional factor according to the weight coefficient, it may first obtain the adjacent reference values of the second image component and the adjacent reference values of the third image component corresponding to the current coding block.

[0119] It should be noted that, in an embodiment of the present application, the adjacent reference values of the second image component and the adjacent reference values of the third image component respectively represent the second image component parameters and the third image component parameters of the adjacent reference points. Specifically, the adjacent reference values of the third image component and the adjacent reference values of the second image component can be used for prediction between the same components.

[0120] Further, in an embodiment of the present application, the adjacent reference values of the third image component and the adjacent reference values of the second image component may be the adjacent reference Cb values and the adjacent reference Cr values corresponding to the adjacent reference points, respectively.

[0121] In an embodiment of the present application, further, before the prediction device obtains the predicted value of the second image component corresponding to the current coding block through the proportional factor, that is, before step 105, the method for the prediction device to perform video image component prediction may further include the following steps:

[0122] Step 109: Obtain the estimated value of the second image component and the reconstruction residual of the third image component corresponding to the current coding block.

[0123] In an embodiment of the present application, before the prediction device obtains the predicted value of the second image component corresponding to the current coding block according to the proportional factor, it may first obtain the estimated value of the second image component and the reconstruction residual of the third image component corresponding to the current coding block.

[0124] It should be noted that, in the embodiments of the present application, the above second image component estimation value is obtained by performing traditional component prediction based on the second image component corresponding to the current coding block; the above third image component reconstruction residual is used to characterize the prediction residual of the third image component corresponding to the current coding block.

[0125] In the embodiments of the present application, further, the scale factor includes a third scale parameter, and the method for the prediction device to determine the scale factor according to the weight coefficient may include the following steps:

[0126] Step 104c: Input the weight coefficient, the third image component adjacent reference value, and the second image component adjacent reference value into a third preset factor calculation model to obtain the third scale parameter.

[0127] In the embodiments of the present invention, after the prediction device inputs the above correlation coefficient into the preset weight calculation model to obtain the weight coefficient corresponding to the above adjacent reference point, the weight coefficient, the above third image component adjacent reference value, and the above second image component adjacent reference value may be input into a third preset factor calculation model to obtain the above third scale parameter.

[0128] It should be noted that, in the embodiments of the present invention, if the above preset device performs prediction between the second image component and the third image component, the scale factor may include the above third scale parameter.

[0129] Further, in the embodiments of the present application, if the prediction device further determines the calculation of the above third scale parameter according to the above third image component adjacent reference value and the above second image component adjacent reference value, the above third preset factor calculation model that can be used is shown in the following formula (8):

[0130]

[0131] Where γ is the third scale parameter, Cb(n) represents the adjacent reference Cb value of the current coding block, that is, the third image component adjacent reference value; Cr(n) represents the adjacent reference Cr value of the current coding block, that is, the second image component adjacent reference value; λ may be an empirical value, and w(n) is the weight coefficient corresponding to each adjacent reference point.

[0132] It can be seen that in the embodiments of the present application, the prediction device can determine the correlation coefficient based on the adjacent reference value of the first image component corresponding to the current coding block and the reconstructed value of the first image component, so that the correlation between the component parameters between the adjacent reference points and the current coding block can be realized, and different weight coefficients can be assigned to different adjacent reference points to construct a linear model that better conforms to the expected model. Therefore, when there is a large deviation between the adjacent reference value of the first image component and the component parameters corresponding to the current coding block, or when there is a large deviation between the adjacent reference value of the third image component and the component parameters corresponding to the current coding block, the defect that the linear model deviates from the expected model can be effectively overcome. Furthermore, when predicting the components of the current coding block according to the linear model, the prediction accuracy of the predicted value of the video image component can be greatly improved, and the predicted value of the video image component is closer to the true value of the video image component.

[0133] Embodiment 4

[0134] Based on the above Embodiment 3, in the embodiments of the present application, the method by which the above prediction device obtains the predicted value of the second image component corresponding to the current coding block through the above scaling factor may include the following steps:

[0135] Step 105a: Obtain the predicted value of the second image component according to the first scaling parameter, the second scaling parameter, and the reconstructed value of the first image component.

[0136] In the embodiments of the present application, after the above prediction device determines the above scaling factor according to the above weight coefficient, it may further obtain the above predicted value of the second image component according to the scaling factor including the above first scaling parameter and the above second scaling parameter, and the above reconstructed value of the first image component.

[0137] It should be noted that in the embodiments of the present application, after the above prediction device obtains the scaling factor including the above first scaling parameter and the above second scaling parameter, it may establish a linear model for video image component prediction according to the above first scaling parameter, the above second scaling parameter, and the above reconstructed value of the first image component, so that the second image component of the above current coding block can be constructed according to the linear model to obtain the above predicted value of the second image component.

[0138] Furthermore, in the embodiments of the present application, the above prediction device may determine the above predicted value of the second image component through the above formula (1), that is, the above prediction device may calculate and obtain the above predicted value of the second image component according to the above first scaling parameter, the above second scaling parameter, and the above reconstructed value of the first image component.

[0139] In an embodiment of the present application, the method for the prediction device to obtain the predicted value of the second image component corresponding to the current coding block through the scaling factor may further include the following steps:

[0140] Step 105b: Obtain the predicted value of the second image component according to a third scaling parameter, an estimated value of the second image component, and a reconstruction residual of the third image component.

[0141] In an embodiment of the present application, after the prediction device determines the scaling factor according to the weight coefficient, it may further obtain the predicted value of the second image component according to the scaling factor including the third scaling parameter, the estimated value of the second image component, and the reconstruction residual of the third image component.

[0142] It should be noted that in an embodiment of the present application, after the prediction device obtains the scaling factor including the third scaling parameter, it may establish a linear model for video image component prediction according to the third scaling parameter, the estimated value of the second image component, and the reconstruction residual of the third image component, so that the second image component of the current coding block can be constructed according to this linear model to obtain the predicted value of the second image component.

[0143] Furthermore, in an embodiment of the present application, the prediction device may determine the predicted value of the second image component through the above formula (3), that is, the prediction device may calculate and obtain the predicted value of the second image component according to the third scaling parameter, the estimated value of the second image component, and the reconstruction residual of the third image component.

[0144] Thus, it can be seen that in an embodiment of the present application, the prediction device can determine the correlation coefficient based on the adjacent reference value of the first image component and the reconstructed value of the first image component corresponding to the current coding block, so as to realize the assignment of different weight coefficients to different adjacent reference points according to the correlation between the adjacent reference points and the component parameters of the current coding block, in order to construct a linear model that better conforms to the expected model, thereby effectively overcoming the defect that the linear model deviates from the expected model when there is a large deviation between the adjacent reference value of the first image component and the component parameters corresponding to the current coding block, or when there is a large deviation between the adjacent reference value of the third image component and the component parameters corresponding to the current coding block. Furthermore, when predicting the components of the current coding block according to this linear model, the prediction accuracy of the predicted value of the video image component can be greatly improved, and the predicted value of the video image component is closer to the true value of the video image component.

[0145] Embodiment Five

[0146] Based on the same inventive concept of the above Embodiments 1 to 4, Figure 9Structural schematic diagram of the prediction device proposed in the embodiment of the present application Figure 1 , as Figure 9 shown, the prediction device 1 proposed in the embodiment of the present application may include an acquisition part 11, a determination part 12, and a setting part 13.

[0147] The acquisition part 11 is configured to acquire the adjacent reference value of the first image component corresponding to the current coding block and the reconstruction value of the first image component; wherein, the adjacent reference value of the first image component is used to represent the first image component parameter corresponding to the adjacent reference point of the current coding block, and the reconstruction value of the first image component is used to represent the reconstruction parameter of one or more first image components corresponding to the current coding block.

[0148] The determination part 12 is configured to determine a correlation coefficient according to the adjacent reference value of the first image component and the reconstruction value of the first image component after the acquisition part 11 acquires the adjacent reference value of the first image component corresponding to the current coding block and the reconstruction value of the first image component; wherein, the correlation coefficient is used to represent the deviation degree of the image component between the current coding block and the adjacent reference point.

[0149] The acquisition part 11 is further configured to input the correlation coefficient into a preset weight calculation model after the determination part 12 determines the correlation coefficient according to the adjacent reference value of the first image component and the reconstruction value of the first image component, so as to obtain the weight coefficient corresponding to the adjacent reference point.

[0150] The determination part 12 is further configured to determine a scale factor according to the weight coefficient after the acquisition part 11 inputs the correlation coefficient into a preset weight calculation model and obtains the weight coefficient corresponding to the adjacent reference point.

[0151] The acquisition part 11 is further configured to obtain the predicted value of the second image component corresponding to the current coding block through the scale factor after the determination part 12 determines the scale factor according to the weight coefficient.

[0152] Further, in the embodiment of the present application, the determination part 12 is specifically configured to perform a difference operation on any one of the adjacent reference values of the first image component and each of the reconstruction values of the first image component to obtain the component difference corresponding to the any one of the reference values; wherein, one adjacent reference value of the first image component and one reconstruction value of the first image component correspond to one difference; and determine the smallest difference among the component differences as the correlation coefficient.

[0153] Further, in the embodiments of the present application, before the obtaining part 11 is further configured to determine a scaling factor according to the weight coefficient, it obtains a second image component adjacent reference value corresponding to the current coding block; wherein, the second image component adjacent reference value is a second image component parameter corresponding to the adjacent reference point and different from the first image component parameter.

[0154] Further, in the embodiments of the present application, the scaling factor includes a first scaling parameter and a second scaling parameter. The determining part 12 is further specifically configured to input the weight coefficient, the first image component adjacent reference value, and the second image component adjacent reference value into a first preset factor calculation model to obtain the first scaling parameter; and input the weight coefficient, the first scaling parameter, the first image component adjacent reference value, and the second image component adjacent reference value into a second preset factor calculation model to obtain the second scaling parameter.

[0155] Further, in the embodiments of the present application, before the obtaining part 11 is further configured to determine a scaling factor according to the weight coefficient, it obtains a third image component adjacent reference value and a second image component adjacent reference value corresponding to the current coding block; wherein, the second image component adjacent reference value and the third image component adjacent reference value respectively represent the second image component parameter and the third image component parameter of the adjacent reference point.

[0156] Further, in the embodiments of the present application, before the obtaining part 11 is further configured to obtain a second image component prediction value corresponding to the current coding block through the scaling factor, it obtains a second image component estimated value and a third image component reconstruction residual corresponding to the current coding block; wherein, the second image component estimated value is obtained by performing component prediction according to the second image component corresponding to the current coding block; the third image component reconstruction residual is used to represent the prediction residual corresponding to the current coding block.

[0157] Further, in the embodiments of the present application, the scaling factor includes a third scaling parameter. The determining part 12 is further specifically configured to input the weight coefficient, the second image component adjacent reference value, and the third image component adjacent reference value into a third preset factor calculation model to obtain the third scaling parameter.

[0158] Further, in the embodiments of the present application, the obtaining part 11 is specifically configured to obtain the second image component prediction value according to the first scaling parameter, the second scaling parameter, and the first image component reconstruction value.

[0159] Further, in the embodiments of the present application, the obtaining part 11 is further specifically configured to obtain the predicted value of the second image component according to the third proportional parameter, the estimated value of the second image component, and the reconstruction residual of the third image component.

[0160] Further, in the embodiments of the present application, the setting part 13 is configured to, after performing a difference operation on any one of the adjacent reference values of the first image component and each of the reconstruction values of the first image component to obtain the component difference corresponding to the any one of the reference values, and before determining the proportional factor according to the weight coefficient, when the component differences corresponding to any one of the reference values are all greater than a preset difference threshold, set the weight coefficient of the adjacent reference point corresponding to the any one of the reference values to zero.

[0161] Figure 10 Schematic diagram of the composition structure of the prediction device proposed in the embodiments of the present application Figure 2 , such as Figure 10 shown, the prediction device 1 proposed in the embodiments of the present application may further include a processor 14, a memory 15 storing executable instructions of the processor 14, a communication interface 16, and a bus 17 for connecting the processor 14, the memory 15, and the communication interface 16.

[0162] In the embodiments of the present application, the above-mentioned processor 14 may be at least one of an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, and a microprocessor. It can be understood that for different devices, the electronic devices for implementing the above-mentioned processor functions may be others, and the embodiments of the present application do not make specific limitations. The device 1 may further include a memory 15, and the memory 15 may be connected to the processor 14. Among them, the memory 15 is used to store executable program codes, and the program codes include computer operation instructions. The memory 15 may include a high-speed RAM memory and may also include non-volatile memory, for example, at least two disk memories.

[0163] In an embodiment of the present application, the bus 17 is used to connect the communication interface 16, the processor 14, and the memory 15 and enable mutual communication among these components.

[0164] In an embodiment of the present application, the memory 15 is used to store instructions and data.

[0165] Further, in an embodiment of the present application, the above-mentioned processor 14 is used to obtain the adjacent reference value of the first image component and the reconstructed value of the first image component corresponding to the current coding block; wherein, the adjacent reference value of the first image component is used to represent the first image component parameters corresponding to the adjacent reference points of the current coding block, and the reconstructed value of the first image component is used to represent the reconstruction parameters of one or more first image components corresponding to the current coding block; determine the correlation coefficient according to the adjacent reference value of the first image component and the reconstructed value of the first image component; wherein, the correlation coefficient is used to represent the deviation degree of the image components between the current coding block and the adjacent reference points; input the correlation coefficient into a preset weight calculation model to obtain the weight coefficient corresponding to the adjacent reference points; determine the scaling factor according to the weight coefficient; and obtain the predicted value of the second image component corresponding to the current coding block through the scaling factor.

[0166] In practical applications, the above-mentioned memory 15 may be a volatile first memory (volatile memory), such as a random access first memory (Random-Access Memory, RAM); or a non-volatile first memory (non-volatile memory), such as a read-only first memory (Read-Only Memory, ROM), a flash first memory (flash memory), a hard disk (Hard Disk Drive, HDD), or a solid-state drive (Solid-State Drive, SSD); or a combination of the above types of first memories, and provide instructions and data to the processor 14.

[0167] In addition, in this embodiment, each functional module may be integrated in a processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional module.

[0168] When the integrated unit is implemented in the form of a software functional module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method of this embodiment. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs.

[0169] A device proposed in an embodiment of this application. This prediction device can determine the correlation coefficient based on the adjacent reference value of the first image component corresponding to the current coding block and the reconstruction value of the first image component, so as to realize the assignment of different weight coefficients to different adjacent reference points according to the correlation between the adjacent reference points and the component parameters of the current coding block, in order to construct a linear model that better conforms to the expected model. Thus, when there is a large deviation between the adjacent reference value of the first image component and the component parameters corresponding to the current coding block, or when there is a large deviation between the adjacent reference value of the third image component and the component parameters corresponding to the current coding block, the defect that the linear model deviates from the expected model can be effectively overcome. Furthermore, when predicting the components of the current coding block according to this linear model, the prediction accuracy of the predicted value of the video image component can be greatly improved, making the predicted value of the video image component closer to the true value of the video image component.

[0170] An embodiment of this application provides a first computer-readable storage medium, on which a program is stored. When the program is executed by a processor, it implements the methods of Embodiments 1 to 4.

[0171] Specifically, the program instructions corresponding to a method for predicting a video image component in this embodiment can be stored on storage media such as optical discs, hard disks, and USB flash drives. When the program instructions corresponding to a method for predicting a video image component in the storage medium are read or executed by an electronic device, the following steps are included:

[0172] Obtain the adjacent reference value of the first image component corresponding to the current coding block and the reconstruction value of the first image component; wherein, the adjacent reference value of the first image component is used to represent the first image component parameters corresponding to the adjacent reference points of the current coding block, and the reconstruction value of the first image component is used to represent the reconstruction parameters of one or more first image components corresponding to the current coding block;

[0173] Determine a correlation coefficient according to the adjacent reference value and the reconstructed value of the first image component; wherein, the correlation coefficient is used to characterize the deviation degree of the image component between the current coding block and the adjacent reference point;

[0174] Input the correlation coefficient into a preset weight calculation model to obtain the weight coefficient corresponding to the adjacent reference point;

[0175] Determine a scaling factor according to the weight coefficient;

[0176] Obtain a predicted value of the second image component corresponding to the current coding block through the scaling factor.

[0177] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories and optical memories, etc.) containing computer-usable program codes.

[0178] The present application is described with reference to the implementation flow diagrams and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the implementation flow diagrams and / or block diagrams, as well as the combination of the processes and / or blocks in the implementation flow diagrams and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, so that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in one or more processes or multiple processes and / or blocks Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0179] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in one or more processes or multiple processes and / or blocks Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0180] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are executed on the computer or other programmable apparatus to generate a computer-implemented process, thereby providing instructions for implementing the functions specified in one or more of the flowcharts and / or boxes shown in the flowchart. Figure 1 in the flowchart of one or more of the processes and / or boxes Figure 1 steps for implementing the functions specified in one or more of the boxes.

[0181] As mentioned above, the foregoing are only preferred embodiments of the present application and are not intended to limit the scope of protection of the present application.

Claims

1. A method for predicting video image components, characterized in that, The method is applied to an encoder, and the method includes: Obtaining a first image component adjacent reference value and a first image component reconstruction value corresponding to a current coding block; wherein, the first image component adjacent reference value is used to represent a first image component parameter corresponding to an adjacent reference point of the current coding block, and the first image component reconstruction value is used to represent a reconstruction value of one or more first image components corresponding to the current coding block; Determining a weight coefficient corresponding to the adjacent reference point according to the current coding block and the adjacent reference point; Determining a scaling factor according to the first image component adjacent reference value, the first image component reconstruction value, and the weight coefficient by using a preset factor calculation model; Obtaining a second image component prediction value corresponding to the current coding block based on a linear model corresponding to the scaling factor.

2. The method according to claim 1, characterized in that, The determining the weight coefficient corresponding to the adjacent reference point according to the current coding block and the adjacent reference point includes: Determining the weight coefficient corresponding to the adjacent reference point according to the correlation between the current coding block and the adjacent reference point; wherein, the correlation represents the position correlation between the current coding block and the adjacent reference point.

3. The method according to claim 1, characterized in that, Before determining the scaling factor according to the first image component adjacent reference value, the first image component reconstruction value, and the weight coefficient, the method further includes: Obtaining a second image component adjacent reference value corresponding to the current coding block; wherein, the second image component adjacent reference value is a second component parameter corresponding to the adjacent reference point and different from the first image component parameter.

4. The method according to claim 3, characterized in that, The determining the scaling factor according to the first image component adjacent reference value, the first image component reconstruction value, and the weight coefficient by using a preset factor calculation model includes: Inputting the weight coefficient, the first image component adjacent reference value, and the second image component adjacent reference value into the preset factor calculation model to obtain the scaling factor.

5. The method according to claim 3, wherein The first image component adjacent reference value and the first image component reconstruction value corresponding to the current coding block include at least two sets of first image component adjacent reference values and first image component reconstruction values; the method further includes: Respectively determining weight coefficients for each set of first image component adjacent reference values and first image component reconstruction values; Respectively inputting the weight coefficient, the first image component adjacent reference value, and the second image component adjacent reference value into the preset factor calculation model to obtain at least two sets of scaling factors corresponding to the at least two sets of first image component adjacent reference values and first image component reconstruction values; Selecting a set of scaling factors from the at least two sets of scaling factors.

6. The method according to claim 5, wherein The method further includes: Calculating a rate-distortion cost result of each set of scaling factors in the at least two sets of scaling factors; Selecting a set of scaling factors corresponding to the minimum rate-distortion cost result.

7. The method according to claim 3, wherein Before obtaining the second image component prediction value corresponding to the current coding block based on the linear model corresponding to the scaling factor, the method further includes: Obtain a second image component estimation value corresponding to the current coding block; wherein, the second image component estimation value is obtained by component prediction based on the second image component corresponding to the current coding block.

8. The method according to claim 4 or 5, characterized in that, The obtaining of the second image component prediction value corresponding to the current coding block based on the linear model corresponding to the scaling factor includes: Obtain the second image component prediction value according to the linear model corresponding to the scaling factor and the first image component reconstruction value.

9. The method according to claim 7, wherein The obtaining of the second image component prediction value corresponding to the current coding block based on the linear model corresponding to the scaling factor includes: Obtain the second image component prediction value according to the scaling factor and the second image component estimation value.

10. The method according to claim 2, characterized in that, The method further includes: Preprocess the adjacent reference points.

11. The method according to claim 10, wherein The method further includes: Set the weight coefficients of some reference points corresponding to some adjacent reference values to zero; wherein, the position correlation between the some reference points and the current coding block does not meet a preset threshold.

12. The method according to claim 10, characterized in that, The method further includes: Eliminate some reference points from the adjacent reference points; wherein, the position correlation between the some reference points and the current coding block does not meet a preset threshold.

13. The method according to claim 1, wherein The adjacent reference points at least include one of the following: the upper pixel points of the current coding block, the upper left pixel points of the current coding block, and the left pixel points of the current coding block.

14. The method according to claim 13, wherein The adjacent reference points further at least include one of the following: the upper right pixel points of the current coding block, the lower left pixel points of the current coding block.

15. The method according to claim 1, wherein The scaling factor is a coefficient of the linear model; wherein, the linear model is used to predict the second image component based on the first image component.

16. A method for predicting video image components, characterized in that The method is applied to a decoder, and the method includes: Obtain a first image component adjacent reference value and a first image component reconstruction value corresponding to the current block; wherein, the first image component adjacent reference value is used to represent the first image component parameters corresponding to the adjacent reference points of the current block, and the first image component reconstruction value is used to represent the reconstruction values of one or more first image components corresponding to the current block; Determine the weight coefficients corresponding to the adjacent reference points according to the current block and the adjacent reference points; Determine a scaling factor according to the first image component adjacent reference value, the first image component reconstruction value, and the weight coefficients by using a preset factor calculation model; Obtain a second image component prediction value corresponding to the current block based on the linear model corresponding to the scaling factor.

17. The method according to claim 16, wherein The determining of the weight coefficients corresponding to the adjacent reference points according to the current block and the adjacent reference points includes: Determine the weight coefficients corresponding to the adjacent reference points according to the correlation between the current block and the adjacent reference points; wherein, the correlation represents the position correlation between the current block and the adjacent reference points.

18. The method according to claim 16, wherein Before determining the scale factor according to the adjacent reference value of the first image component, the reconstruction value of the first image component, and the weight coefficient, the method further includes: Obtaining the adjacent reference value of the second image component corresponding to the current block; wherein, the adjacent reference value of the second image component is the second component parameter corresponding to the adjacent reference point and different from the parameter of the first image component.

19. The method according to claim 18, wherein Determining the scale factor according to the adjacent reference value of the first image component, the reconstruction value of the first image component, and the weight coefficient by using a preset factor calculation model, includes: Inputting the weight coefficient, the adjacent reference value of the first image component, and the adjacent reference value of the second image component into the preset factor calculation model to obtain the scale factor.

20. The method according to claim 18, wherein Before obtaining the predicted value of the second image component corresponding to the current block based on the linear model corresponding to the scale factor, the method further includes: Obtaining the estimated value of the second image component corresponding to the current block; wherein, the estimated value of the second image component is obtained by component prediction according to the second image component corresponding to the current block.

21. The method according to claim 19, wherein Obtaining the predicted value of the second image component corresponding to the current block based on the linear model corresponding to the scale factor, includes: Obtaining the predicted value of the second image component according to the linear model corresponding to the scale factor and the reconstruction value of the first image component.

22. The method according to claim 20, wherein Obtaining the predicted value of the second image component corresponding to the current block based on the linear model corresponding to the scale factor, includes: Obtaining the predicted value of the second image component according to the scale factor and the estimated value of the second image component.

23. The method according to claim 17, wherein The method further includes: Preprocessing the adjacent reference points.

24. The method according to claim 23, wherein The method further includes: Setting the weight coefficients of some reference points corresponding to some adjacent reference values to zero; wherein, the position correlation between the some reference points and the current block does not meet a preset threshold.

25. The method according to claim 23, wherein The method further includes: Removing some reference points from the adjacent reference points; wherein, the position correlation between the some reference points and the current block does not meet a preset threshold.

26. The method according to claim 16, wherein The adjacent reference points at least include one of the following: the upper pixel point of the current block, the upper left pixel point of the current block, and the left pixel point of the current block.

27. The method according to claim 26, wherein The adjacent reference points further at least include one of the following: the upper right pixel point of the current block, the lower left pixel point of the current block.

28. The method according to claim 16, wherein The scale factor is the coefficient of the linear model; wherein, the linear model is used to predict the second image component according to the first image component.

29. An encoder, characterized in that, The encoder includes: an acquisition unit and a determination unit, The obtaining unit is configured to obtain a first image component adjacent reference value and a first image component reconstruction value corresponding to a current coding block; wherein, the first image component adjacent reference value is used to represent a first image component parameter corresponding to an adjacent reference point of the current coding block, and the first image component reconstruction value is used to represent a reconstruction value of one or more first image components corresponding to the current coding block; The determining unit is configured to determine a weight coefficient corresponding to the adjacent reference point according to the current coding block and the adjacent reference point; The determining unit is further configured to determine a scaling factor by using a preset factor calculation model according to the first image component adjacent reference value, the first image component reconstruction value, and the weight coefficient; The obtaining unit is further configured to obtain a second image component prediction value corresponding to the current coding block based on a linear model corresponding to the scaling factor.

30. An encoder, characterized in that, The encoder includes a processor and a memory storing instructions executable by the processor. When the instructions are executed, the processor implements the method according to any one of claims 1-15.

31. A decoder, characterized in that, The decoder includes: an obtaining unit and a determining unit, The obtaining unit is configured to obtain a first image component adjacent reference value and a first image component reconstruction value corresponding to a current block; wherein, the first image component adjacent reference value is used to represent a first image component parameter corresponding to an adjacent reference point of the current block, and the first image component reconstruction value is used to represent a reconstruction value of one or more first image components corresponding to the current block; The determining unit is configured to determine a weight coefficient corresponding to the adjacent reference point according to the current block and the adjacent reference point; The determining unit is further configured to determine a scaling factor by using a preset factor calculation model according to the first image component adjacent reference value, the first image component reconstruction value, and the weight coefficient; The obtaining unit is further configured to obtain a second image component prediction value corresponding to the current block based on a linear model corresponding to the scaling factor.

32. A decoder, characterized in that, The decoder includes a processor and a memory storing instructions executable by the processor. When the instructions are executed, the processor implements the method according to any one of claims 16-28.

33. A computer-readable storage medium, characterized in that, A program is stored thereon, and when the program is executed by a processor, the method according to any one of claims 1-15 or 16-28 is implemented.

Images