Method, apparatus, and computer storage medium for predicting image components

Abstract

To provide a method and a device for predicting an image component, and a computer storage medium, which integrate a process of deriving a model parameter without changing encoding / decoding prediction performance.SOLUTION: A method includes steps of: obtaining a first reference pixel set corresponding to a predicted image component of a coding block in a video image (S501); using a preset component value as a prediction value corresponding to the predicted image component if a number of effective pixel points is less than a preset number (S502); screening the first reference pixel set to obtain a second reference pixel set if the number of effective pixel points is greater than or equal to the preset number (S503); using the preset component value as a prediction value corresponding to a predicted image component (S504); determining a model parameter via a second reference pixel set, and obtaining the prediction model corresponding to the predicted image component (S505). The prediction model is used to implement prediction processing on the image component to be predicted to obtain a prediction value.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to the field of video coding and decoding, and more particularly to a method, apparatus, and computer storage medium for predicting image components. [Background technology]

[0002] With the increasing demand for higher video display quality, new forms of video applications have emerged, such as high-resolution and ultra-high-resolution video. H.265 / High Efficiency Video Coding (HEVC) could not meet the demands of the rapid development of video applications, so the Joint Video Exploration Team (JVET) proposed the next-generation video coding standard H.266 / Versatile Video Coding (VVC), and the corresponding test model is the VVC Reference Software Test Platform (VTM).

[0003] In VTM, we are currently integrating a predictive model-based method for predicting image components, which allows us to predict the chromaticity component from the luminance component of the current coding block (CB) via this predictive model. However, when constructing the predictive model, the number of adjacent reference pixel points used to derive the model parameters differs, which not only increases the additional processing but also increases the computational complexity. [Overview of the project] [Problems that the invention aims to solve]

[0004] This embodiment provides an image component prediction method, apparatus, and computer storage medium, which integrate the model parameter derivation process without changing the encoding / decoding prediction performance, and simultaneously reduces computational complexity by not increasing the number of additional processing modules when the number of effective pixel points in the adjacent reference pixel set is smaller than the preset number. [Means for solving the problem]

[0005] The technical solution of the present embodiment can be realized as follows.

[0006] In the first embodiment, the present invention provides a method for predicting image components, the method being: Obtaining a first reference pixel set corresponding to the predicted image components of an encoded block in a video image, If the number of effective pixel points in the first reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components; if the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set is screened to obtain a second reference pixel set, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets. The process includes: if the number of effective pixel points in the second reference pixel set is less than the number of presets, using the component values ​​of the presets as predicted values ​​corresponding to the predicted image components; and if the number of effective pixel points in the second reference pixel set is equal to the number of presets, determining model parameters via the second reference pixel set and obtaining a prediction model corresponding to the predicted image components according to the model parameters, wherein the prediction model is used to perform prediction processing on the predicted image components and obtain predicted values ​​corresponding to the predicted image components.

[0007] In a second aspect, an embodiment of the present application provides a prediction device for image components, comprising an acquisition unit, a prediction unit, and a screening unit, where the acquisition unit is configured to acquire a first set of reference pixel corresponding to a predicted image component of an encoded block in a video image; when the number of valid pixel points in the first set of reference pixels is smaller than a preset number, the prediction unit is configured to use the component value of the preset as a predicted value corresponding to the predicted image component; when the number of valid pixel points in the first set of reference pixels is greater than or equal to the preset number, the screening unit is configured to screen the first set of reference pixels to obtain a second set of reference pixels, where the number of valid pixel points in the second set of reference pixels is smaller than or equal to the preset number; when the number of valid pixel points in the second set of reference pixels is smaller than the preset number, the prediction unit is further configured to use the component value of the preset as a predicted value corresponding to the predicted image component, and when the number of valid pixel points in the second set of reference pixels is equal to the preset number, the prediction unit is configured to determine model parameters through the second set of reference pixels, and obtain a prediction model corresponding to the predicted image component according to the model parameters, where the prediction model is used to perform a prediction process on the predicted image component to obtain a predicted value corresponding to the predicted image component.

[0008] In a third aspect, an embodiment of the present application provides a prediction device for image components, comprising a memory and a processor, the memory is configured to store a computer program executable by the processor; when executing the computer program, the processor is configured to execute the method described in the first aspect.

[0009] In a fourth embodiment, the present invention provides a computer storage medium for storing an image component prediction program, wherein the image component prediction program is executed by at least one processor, thereby realizing the method described in the first embodiment. [Effects of the Invention]

[0010] The embodiments of the present application provide a method, an apparatus, and a computer storage medium for predicting an image component. By obtaining a first set of reference pixels corresponding to a predicted image component of an encoded block in a video image, when the number of valid pixel points in the first set of reference pixels is smaller than the number of presets, the component value of the preset is used as a predicted value corresponding to the predicted image component. When the number of valid pixel points in the first set of reference pixels is greater than or equal to the number of presets, the first set of reference pixels is screened to obtain a second set of reference pixels. The number of valid pixel points in the second set of reference pixels is less than or equal to the number of presets. When the number of valid pixel points in the second set of reference pixels is smaller than the number of presets, the component value of the preset is used as a predicted value corresponding to the predicted image component. When the number of valid pixel points in the second set of reference pixels is equal to the number of presets, model parameters are determined through the second set of reference pixels, and according to the model parameters, a prediction model corresponding to the predicted image component is obtained. The prediction model is used to implement a prediction process for the predicted image component to obtain a predicted value corresponding to the predicted image component. In this way, when the number of valid pixel points in the first set of reference pixels is smaller than the number of presets or the number of valid pixel points in the second set of reference pixels is smaller than the number of presets, the default value of the preset is directly used as the predicted value corresponding to the predicted image component. Only when the number of valid pixel points in the second set of reference pixels meets the number of presets, model parameters are determined according to the first set of reference pixels to establish a prediction model for the predicted image component, thereby integrating the derivation process of the model parameters. Furthermore, when the number of valid pixel points in the first set of reference pixels or the second set of reference pixels is smaller than the number of presets, especially when the number of valid pixel points is 0 or 2, since no additional processing module is added, the default value of the preset is directly used as the predicted value corresponding to the predicted image component, without the need for additional processing, reducing the computational complexity.

Brief Description of the Drawings

[0011] [Figure 1]This is a schematic diagram showing the distribution of effective adjacent regions according to the embodiment of the present invention. [Figure 2] This is a schematic diagram of the distribution of the selection region in the three modes according to the embodiment of the present invention. [Figure 3] This is a block diagram of an exemplary configuration of a video encoding system according to the present embodiment. [Figure 4] This is a block diagram of an exemplary configuration of a video decoding system according to the present embodiment. [Figure 5] This is an illustrative flowchart of the image component prediction method according to the present embodiment. [Figure 6A] This is an exemplary structural diagram of the selection of adjacent reference pixel points in INTRA_LT_CCLM mode according to the present embodiment. [Figure 6B] This is an illustrative structural diagram of the selection of adjacent reference pixel points in INTRA_L_CCLM mode according to the present embodiment. [Figure 6C] This is an illustrative structural diagram of the selection of adjacent reference pixel points in INTRA_T_CCLM mode according to the present embodiment. [Figure 7] This is an illustrative flowchart of another image component prediction method according to the embodiment of the present invention. [Figure 8A] This is an exemplary structural diagram illustrating the generation of 0 effective pixel points in INTRA_LT_CCLM mode according to the embodiment of the present invention. [Figure 8B] This is an exemplary structural diagram illustrating the generation of 0 effective pixel points in INTRA_L_CCLM mode according to the embodiment of the present invention. [Figure 8C] This is an exemplary structural diagram illustrating the generation of 0 effective pixel points in INTRA_T_CCLM mode according to the embodiment of the present invention. [Figure 9A] This is an exemplary structural diagram illustrating the generation of 2 effective pixel points in INTRA_LT_CCLM mode according to the present embodiment. [Figure 9B] This is an exemplary structural diagram illustrating the generation of 2 effective pixel points in INTRA_L_CCLM mode according to the embodiment of the present invention. [Figure 9C] This is an exemplary structural diagram illustrating the generation of 2 effective pixel points in INTRA_T_CCLM mode according to the present embodiment. [Figure 10] This is an exemplary flowchart for deriving the model parameters according to the embodiment of the present invention. [Figure 11] This is a simplified flowchart illustrating the derivation of model parameters according to the embodiment of the present invention. [Figure 12] This is another illustrative, simplified flowchart for deriving the model parameters according to the embodiment of the present invention. [Figure 13] This is an illustrative structural diagram of the configuration of an image component prediction device according to the present embodiment. [Figure 14] This is an illustrative diagram of the specific hardware structure of an image component prediction device according to the present embodiment. [Figure 15] This is an illustrative structural diagram of the encoder configuration according to the present embodiment. [Figure 16] This is an illustrative structural diagram of the decoder configuration according to the present embodiment. [Modes for carrying out the invention]

[0012] To gain a more detailed understanding of the features and technical content of the embodiments of this application, the embodiments will be described in detail below with reference to the drawings. The attached drawings are for reference purposes only and are not intended to limit the embodiments of this application.

[0013] In video images, an encoded block is typically represented using a first image component, a second image component, and a third image component, where these three image components are one luminance component, one blue chromaticity component, and one red chromaticity component, respectively. Specifically, the luminance component is usually represented by the code Y, the blue chromaticity component by the code Cb or U, and the red chromaticity component by the code Cr or V. In this way, a video image can be represented in YCbCr format or in YUV format.

[0014] In the embodiments of this application, the first image component may be a luminance component, the second image component may be a blue component, and the third image component may be a red component; however, the embodiments of this application are not particularly limited.

[0015] In current video image or video encoding / decoding processes, cross-component prediction techniques primarily include cross-component linear model prediction (CCLM) mode and multi-direction linear model prediction (MDLM) mode. Regardless of the model parameters derived according to the CCLM mode or the MDLM mode, all corresponding prediction models can achieve predictions between image components, such as from the first image component to the second, from the second to the first, from the first to the third, from the third to the first, from the second to the third, or from the third to the second.

[0016] Taking the prediction from the first image component to the second image component as an example, in order to reduce redundancy between the first and second image components, the CCLM mode is used for VVC. In this case, the first and second image components belong to the same coding block, that is, the predicted value of the second image component is constructed according to the reconstructed value of the first image component of the same coding block, as shown in equation (1) below.

number

[0017] JPEG0007901221000002.jpg49170

[0018] For a coding block, its adjacent regions may include the left adjacent region, the upper adjacent region, the lower left adjacent region, and the upper right adjacent region. VVC may include three cross-component linear model prediction modes, which are the left and upper adjacent intra CCLM mode (which can be represented as INTRA_LT_CCLM mode), the left and lower left adjacent intra CCLM mode (which can be represented as INTRA_L_CCLM mode), and the upper and upper right adjacent intra CCLM mode (which can be represented as INTRA_T_CCLM mode). In these three modes, each mode can select a preset number (e.g., four) of adjacent reference pixel points to use in deriving the model parameters α and β, and the main difference between these three modes is that the selected regions corresponding to the adjacent reference pixel points from which the model parameters α and β are derived are different.

[0019] JPEG0007901221000003.jpg98170

[0020] JPEG0007901221000004.jpg106170

[0021] Referring to Figure 1, a schematic diagram of the distribution of effective adjacent regions according to the embodiment of the present invention is shown. In Figure 1, the left adjacent region, the lower left adjacent region, the upper adjacent region, and the upper right adjacent region are all effective. Based on Figure 1, the selection regions for the three modes are as shown in Figure 2. Here, in Figure 2, (a) shows the selection region for the INTRA_LT_CCLM mode, including the left adjacent region and the upper adjacent region; (b) shows the selection region for the INTRA_L_CCLM mode, including the left adjacent region and the lower left adjacent region; and (c) shows the selection region for the INTRA_T_CCLM mode, including the upper adjacent region and the upper right adjacent region. After determining the selection regions for the three modes in this way, reference points for deriving model parameters can be selected in the selection regions. The reference points thus selected are called adjacent reference pixel points, and there are usually a maximum of four adjacent reference pixel points, and the positions of the adjacent reference pixel points are usually determined for a W×H coding block of a single size.

[0022] However, in some special cases, such as the boundaries of coded blocks, when they are unpredictable, or when adjacent reference pixel points cannot be obtained due to the coding order, and also when coding blocks are divided according to tiles and slices, adjacent regions may be invalid, resulting in fewer than four adjacent reference pixel points being selected from the adjacent region, meaning only 0 or 2 adjacent reference pixel points may be selected, thereby preventing the number of adjacent reference pixel points used in deriving model parameters from being consolidated, thereby increasing the computational complexity while simultaneously increasing the additional "duplicate" operations.

[0023] Assuming that the encoding and decoding prediction performance does not change, and in order to integrate the derivation process of model parameters and at the same time reduce computational complexity, the present embodiment provides a method for predicting image components by obtaining a first reference pixel set corresponding to the predicted image components of an encoded block in a video image, and if the number of effective pixel points in the first reference pixel set is less than the number of presets, the preset component values ​​are used as the predicted values ​​corresponding to the predicted image components, and if the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set is screened to obtain a second reference pixel set, and the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets, and if the number of effective pixel points in the second reference pixel set is less than the number of presets, the preset component values ​​are used as the predicted values ​​corresponding to the predicted image components, and if the number of effective pixel points in the second reference pixel set is greater than or equal to the number of presets If the number of resets is equal, the model parameters are determined via the first reference pixel set, and according to the model parameters, a prediction model corresponding to the predicted image components is obtained, where the prediction model is used to perform prediction processing on the predicted image components and obtain predicted values ​​corresponding to the predicted image components. In this way, if the number of effective pixel points in the first reference pixel set is less than the number of presets, or if the number of effective pixel points in the second reference pixel set is less than the number of presets, the CCLM mode is disabled, and the default values ​​of the presets are used directly as predicted values ​​corresponding to the predicted image components, without increasing the number of additional processing modules, thereby reducing computational complexity. Furthermore, the derivation of model parameters is performed only when the number of effective pixel points in the second reference pixel set is the number of presets, i.e., the CCLM mode is executed to unify the model parameter derivation process.

[0024] The embodiments of this application will be described in detail below with reference to the drawings.

[0025] Referring to Figure 3, an exemplary block diagram of the configuration of the video coding system according to this embodiment is shown. As shown in Figure 3, the video coding system 300 includes a transform and quantization unit 301, an intra-estimation unit 302, an intra-prediction unit 303, a motion compensation unit 304, a motion estimation unit 305, an inverse transform and inverse quantization unit 306, a filter control analysis unit 307, a filtering unit 308, a coding unit 309, and a decoded image cache unit 310, where the filtering unit 308 can implement deblocking filtering and sample adaptive offset (SAO) filtering, and the coding unit 309 can implement header coding and context-based adaptive binary arithmetic coding (CABAC). The input raw video signal can be used to obtain a single video coding block through the division of a coding tree unit (CTU), and then the video coding block is transformed via a transformation and quantization unit 301 with respect to the intra or interpretably predicted residual pixel information, the residual information is converted from a pixel field to a transformation field, and the acquired transformation coefficients are quantized to reduce the bit rate.The intra-estimation unit 302 and the intra-prediction unit 303 are configured to perform intra-prediction on the video coding block, more specifically, the intra-estimation unit 302 and the intra-prediction unit 303 are configured to determine the intra-prediction mode to be used to code the video coding block, the motion compensation unit 304 and the motion estimation unit 305 are configured to perform intra-prediction coding on the received video coding block corresponding to one or more blocks in one or more reference frames to provide time prediction information, the motion estimation performed by the motion estimation unit 305 is a process that generates motion vectors, which can estimate the motion of the video coding block, and then the motion compensation unit 304 performs motion compensation based on the motion vectors determined by the motion estimation unit 305. After determining the intra-prediction mode, the intra-prediction unit 303 is further configured to provide selected intra-prediction data to the coding unit 309, and the motion estimation unit 305 also transmits the calculated and determined motion vector data to the coding unit 309. Furthermore, the inverse transform and inverse quantization unit 306 is configured to reconstruct the video coding block and reconstruct the residual block in the pixel field. The reconstructed residual block is then removed from the frame by the filter control analysis unit 307 and the filtering unit 308 to remove block effect artifacts, and subsequently added to one prediction block in the frame of the decoding image cache unit 310 to generate the reconstructed video coding block.The encoding unit 309 is configured to encode various encoding parameters and quantized transformation coefficients, and in a CABAC-based encoding algorithm, the context content can be based on adjacent encoding blocks, encodes intra-predictive mode information determined by instructions, and outputs a bitstream of the video signal, and the decoding image cache unit 310 is configured to store video encoding blocks to be reconstructed in order to predict references. As video encoding is performed, new video encoding blocks to be reconstructed are continuously generated, and all of these reconstructed video encoding blocks are stored in the decoding image cache unit 310.

[0026] Referring to Figure 4, an exemplary block diagram of the configuration of a video decoding system according to an embodiment of the present invention is shown. As shown in Figure 4, the video decoding system 400 comprises a decoding unit 401, an inverse transform and inverse quantization unit 402, an intra prediction unit 403, a motion compensation unit 404, a filtering unit 405, and a decoded image cache unit 406, etc. Here, the decoding unit 401 can perform header decoding and CABAC decoding, and the filtering unit 405 can perform deblocking filtering and SAO filtering. The input video signal is subjected to the coding process shown in Figure 2, and then outputs a bitstream of the video signal. This bitstream is input to the video decoding system 400, first passing through the decoding unit 401 to obtain decoded conversion coefficients, and then processed via the inverse transform and inverse quantization unit 402 to generate residual blocks in the pixel field. The intra-prediction unit 403 is configured to generate prediction data for the current video decoding block based on the determined intra-prediction mode and data passing through the decoding block from the current frame or photograph. The motion compensation unit 404 is configured to determine prediction information for the video decoding block by analyzing motion vectors and other relevant grammatical elements, and to use this prediction information to generate a prediction block for the video decoding block being decoded. The residual block from the inverse transform and inverse quantization unit 402 is added to the corresponding prediction block generated by the intra-prediction unit 403 or the motion compensation unit 404 to form the video block to be decoded. The video signal of the decoded video signal passes through the filtering unit 405 to remove block effect artifacts and improve video quality. The video block to be decoded is then stored in the decoding image cache unit 406, which is configured to output the video signal while simultaneously storing a reference image for subsequent intra-prediction or motion compensation, i.e., obtaining the original video signal to be recovered.

[0027] The method for predicting image components in this embodiment is mainly applied to the portion of the intra-prediction unit 303 shown in Figure 3 and the portion of the intra-prediction unit 403 shown in Figure 4, and more specifically, to the CCLM prediction portion in intra-prediction. In other words, the method for predicting image components in this embodiment may be applied to a video encoding system, a video decoding system, or both simultaneously, but is not particularly limited in this embodiment. When the method is applied to the portion of the intra-prediction unit 303, "encoded block in the video image" specifically refers to the current encoded block in intra-prediction, and when the method is applied to the portion of the intra-prediction unit 403, "encoded block in the video image" specifically refers to the current decoded block in intra-prediction.

[0028] Referring to Figure 5, based on the example application scenario in Figure 3 or Figure 4 above, an exemplary flowchart of the method for predicting image components according to the embodiment of the present invention is shown. As shown in Figure 5, the method may include the following steps:

[0029] In S501, a first set of reference pixels corresponding to the predicted image components of the encoded blocks in the video image is obtained.

[0030] A video image can be divided into multiple encoding blocks, each encoding block may include a first image component, a second image component, and a third image component. Note that in this embodiment, the encoding block is the current block in the video image being encoded. If the first image component needs to be predicted via the prediction model, the predicted image component is the first image component; if the second image component needs to be predicted via the prediction model, the predicted image component is the second image component; and if the third image component needs to be predicted via the prediction model, the predicted image component is the third image component.

[0031] Furthermore, it should be noted that when the left adjacent region, the lower left adjacent region, the upper adjacent region, and the upper right adjacent region are all valid regions, for INTRA_LT_CCLM mode, the first reference pixel set is composed of adjacent reference pixel points in the left adjacent region and the upper adjacent region of the encoded block, as shown in Figure 2(a); for INTRA_L_CCLM mode, the first reference pixel set is composed of adjacent reference pixel points in the left adjacent region and the lower left adjacent region of the encoded block, as shown in Figure 2(b); and for INTRA_T_CCLM mode, the first reference pixel set is formed by adjacent reference pixel points in the upper adjacent region and the upper right adjacent region of the encoded block, as shown in Figure 2(c).

[0032] In some embodiments, for example, obtaining a first set of reference pixels corresponding to the predicted image components of the encoded blocks in the video image for S501 may include the following steps:

[0033] In S501a-1, a reference pixel point adjacent to at least one side of the coding block is obtained, where the at least one side includes the left side and / or the upper side of the coding block.

[0034] In S501a-2, a first set of reference pixels corresponding to the predicted image components is formed based on the reference pixel points.

[0035] At least one side of the coding block may include the left side and / or the upper side of the coding block, that is, at least one side of the coding block may point to the upper side of the coding block, or to the left side of the coding block, or to both the upper and left sides of the coding block, and it should be noted that this is not particularly limited in the embodiments of this application.

[0036] In this way, for INTRA_LT_CCLM mode, if the left adjacent region and the upper adjacent region are all valid regions, the first reference pixel set can be formed by reference pixel points adjacent to the left edge of the coding block and reference pixel points adjacent to the upper edge of the coding block; if the left adjacent region is a valid region and the upper adjacent region is an invalid region, the first reference pixel set can be formed by reference pixel points adjacent to the left edge of the coding block; and if the left adjacent region is an invalid region and the upper adjacent region is a valid region, the first reference pixel set can be formed by reference pixel points adjacent to the upper edge of the coding block.

[0037] In some embodiments, for example, obtaining a first set of reference pixels corresponding to the predicted image components of the encoded blocks in the video image for S501 may include the following steps:

[0038] In S501b-1, a reference pixel point in a reference row or reference column adjacent to the coding block is obtained, where the reference row is formed by rows adjacent to the upper and upper right sides of the coding block, and the reference column is formed by columns adjacent to the left and lower left sides of the coding block.

[0039] In S501b-2, a first set of reference pixels corresponding to the predicted image components is formed based on the reference pixel points.

[0040] Reference rows adjacent to an encoding block can be formed by rows adjacent to the upper and upper right edges of the encoding block, and reference columns adjacent to an encoding block can be formed by columns adjacent to the left and lower left edges of the encoding block. Reference rows or columns adjacent to an encoding block may refer to reference rows adjacent to the upper edge of the encoding block, reference columns adjacent to the left edge of the encoding block, and reference rows or columns adjacent to other edges of the encoding block, and it should be noted that this is not particularly limited in the embodiments of this application. For the sake of explanation, in the embodiments of this application, reference rows adjacent to an encoding block will be described using reference rows adjacent to the upper edge as an example, and reference columns adjacent to an encoding block will be described using reference columns adjacent to the left edge as an example.

[0041] Here, the reference pixel points in a reference row adjacent to the coding block may include reference pixel points adjacent to the upper edge and the upper right edge (which may also be referred to as adjacent reference pixel points corresponding to the upper edge and the upper right edge), where the upper edge represents the upper edge of the coding block, and the upper right edge represents the length of the edge that extends horizontally to the right from the upper edge of the coding block and is the same height as the current coding block. The reference pixel points in a reference column adjacent to the coding block may further include reference pixel points adjacent to the left edge and the lower left edge (which may also be referred to as adjacent reference pixel points corresponding to the left edge and the lower left edge), where the left edge represents the left edge of the coding block, and the lower left edge represents the length of the edge that extends vertically downward from the left edge of the coding block and is the same width as the current decoding block, but this is not particularly limited in the present embodiment.

[0042] In this way, for INTRA_L_CCLM mode, if the left adjacent region and the lower left adjacent region are valid regions, the first reference pixel set can be formed by reference pixel points in the reference column adjacent to the coded block, and for INTRA_T_CCLM mode, if the upper adjacent region and the upper right adjacent region are valid regions, the first reference pixel set can be formed by reference pixel points in the reference row adjacent to the coded block.

[0043] In S502, if the number of effective pixel points in the first reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components.

[0044] It should be noted that the number of effective pixel points can be determined according to the effectiveness of adjacent regions and according to the number of effective pixel points within the selected region. In some special cases, such as the boundaries of coded blocks, when unpredictable and when adjacent reference pixel points cannot be obtained due to the coding order, and when coding blocks are divided according to tile or slice, the left adjacent region, the lower left adjacent region, the upper adjacent region, and the upper right adjacent region are not all effective regions and may contain invalid regions, which can result in the number of effective pixel points within the selected region being less than the preset number, and thus the number of effective pixel points in the first reference pixel set being less than the preset number.

[0045] Furthermore, the number of presets is a determination of the number of pre-set effective pixel points, used to determine whether the predicted image components perform the steps of deriving model parameters and constructing a predictive model, where the number of presets may be four, but is not particularly limited in this embodiment. Thus, assuming the number of presets is four, that is, when the number of effective pixel points in the first reference pixel set is 0 or 2, the component values ​​of the presets are used directly as predicted values ​​corresponding to the predicted image components to reduce computational complexity.

[0046] Furthermore, the preset component values ​​are used to represent fixed values ​​(which may also be called default values) corresponding to pre-set predicted image components. Here, the preset component values ​​mainly relate to the bit information of the current video image. Therefore, in some embodiments, if the number of effective pixel points in the first reference pixel set is less than the number of presets for S502, using the preset component values ​​as predicted values ​​corresponding to the predicted image components may include the following steps.

[0047] In S502a, the preset component range corresponding to the predicted image component is determined based on the bit information of the video image.

[0048] In S502b, according to the component range of the preset, the intermediate value of the component range of the preset is determined, and the intermediate value is used as the predicted value corresponding to the predicted image component, where the intermediate value is shown in the component value of the preset.

[0049] In the embodiment of this application, it should be noted that the intermediate value of the preset component range corresponding to the predicted image component can be used as the preset component value, and then used as the predicted value corresponding to the predicted image component. Here, assuming that the bit depth of the predicted image component is represented by BitDepthC, the calculation method for the intermediate value of the predicted image component can be obtained as 1 << (BitDepthC-1), and this calculation method can be set according to the actual case and is not particularly limited in the embodiment of this application.

[0050] For example, assuming that the predicted image component is the chromaticity component, if we assume that the current video image is an 8-bit video, the component range corresponding to the chromaticity component is 0 to 255, where the intermediate value is 128. In this case, the preset component value can be 128, i.e., the default value can be 128. If we assume that the current video image is a 10-bit video, the component range corresponding to the chromaticity component is 0 to 1023, where the intermediate value is 512. In this case, the preset component value can be 512, i.e., the default value is 512. In the embodiment of this application, the bit information of the video image is 10 bits, for example, meaning the preset component value is 512.

[0051] Furthermore, in some embodiments, after using preset component values ​​as predicted values ​​corresponding to the predicted image components for S502, the method may further include the following steps.

[0052] In S502c, for each pixel point within the encoding block, the preset component values ​​are used to fill in the predicted image components of each pixel point with predicted values.

[0053] Note that if the number of effective pixel points in the first reference pixel set is less than the preset number, the predictive values ​​will be filled into the predicted image components within the coding block using fixed default values ​​directly, without the need to increase the number of additional processing modules.

[0054] For example, assuming a preset component value of 512 and the predicted image component is the chromaticity component, the chromaticity prediction value corresponding to each pixel point in the coding block can be populated directly using 512.

[0055] In S503, if the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set is screened to obtain the second reference pixel set.

[0056] Within the first parameter pixel set, there may be some unimportant reference pixel points (e.g., poorly related reference pixel points) or some anomalous reference pixel points. To ensure the accuracy of the prediction model, these reference pixel points must be removed to obtain a second reference pixel set, where the number of effective pixel points is less than or equal to the number of presets. In practical applications, the number of effective pixel points in the second reference pixel set is typically selected as four, but is not limited to these presets in this embodiment.

[0057] Furthermore, if the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set can be screened to obtain the second reference pixel set. Note that after obtaining the second reference pixel set, decisions must still be made according to the number of effective pixel points and the number of presets in the second reference pixel set. Here, if the number of effective pixel points in the second reference pixel set is less than the number of presets, the component values ​​of the presets can be adopted and used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the second reference pixel set is equal to the number of presets, the model parameters can be derived according to the second reference pixel set.

[0058] Furthermore, in some embodiments, screening the first reference pixel set for S503 to obtain the second reference pixel set is performed as follows: The selected pixel point position is determined based on the pixel position and / or image component intensity corresponding to each adjacent reference pixel point in the first set of reference pixels, This may include selecting effective pixel points from the first reference pixel set corresponding to the selected pixel point positions according to the determined selected pixel point positions, and forming a second reference pixel set with the selected effective pixel points, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets.

[0059] Specifically, the screening of the first set of reference pixels may be performed according to the position of the selected reference pixel points, or according to the intensity of the image components (luminance value, chromaticity value, etc.), thereby forming a second set of reference pixels with the screened selected reference pixel points. The following explanation uses the position of the selected reference pixel points as an example.

[0060] Assuming that numSampT is the number of effective pixel point samples in the upper and upper right regions adjacent to the current coded block, and numSampL is the number of effective pixel point samples in the left and lower left regions adjacent to the current coded block, the screening process is as follows (where availT represents the effectiveness of the rows adjacent to the current coded block above, availL represents the effectiveness of the columns adjacent to the left of the current coded block, nTbW represents the width of the current coded block, and nTbH represents the height of the current coded block).

[0061] If the current block's intra prediction mode is INTRA_LT_CCLM mode, numSampT=availT?nTbW:0 numSampL = availL ? nTbH : 0, Otherwise, numSampT=(availT&&predModeIntra==INTRA_T_CCLM)?(nTbW+Min(numTopRight,nTbH)):0 numSampL=(availL&&predModeIntra==INTRA_L_CCLM)?(nTbH+Min(numLeftBelow,nTbW)):0.

[0062] Here, numTopRight represents the number of effective pixel points within the upper right nTbW range, and numLeftBelow represents the number of effective pixel points within the lower left nTbH range. The number of pixel points screened at each edge is denoted by cntN, the starting point position is denoted by startPosN, the selection interval is denoted by pickStepN, and the selected pixel point position is denoted by pickPosN[pos]. The derivation process is as follows:

[0063] The variable numIs4N indicates whether or not to screen pixel points on one side: numIs4N=((availT&&availL&&predModeIntra==INTRA_LT_CCLM)?0:1) The variable startPosN represents the starting point position: startPosN=numSampN>>(2+numIs4N) The variable pickStepN represents the selection interval: pickStepN=Max(1,numSampN>>(1+numIs4N)) Here, N is replaced by T and L respectively, representing the cases of screening pixel points on the upper side and the left side, respectively; that is, the N side here represents either the T side or the L side. Here, if the validity of the N side, availN, is TRUE, and the selected intra-mode, predModeIntra, is INTRA_LT_CCLM mode or INTRA_N_CCLM mode, then the number of pixel points screened on the N side, cntN, and the selected pixel point positions, pickPosN[pos], are as follows (it should be noted that the total number of screened pixel points should be cntT + cntL).

[0064] cntN=Min(numSampN,(1+numIs4N)<<1) pickPosN[pos]=(startPosN+pos*pickStepN),with pos=0...cntN-1 Otherwise, cntN is set to 0, meaning the number of screened pixel points is 0.

[0065] Assuming that the prediction samples for the current coded block are predSamples[x][y] with x=0...nTbW-1,y=0...nTbH-1, the derivation is as follows:

[0066] If both numSampL and numSampT are disabled, the preset values ​​will be used, as follows:

[0067] predSamples[x][y]=1<<(BitDepth C -1) Otherwise, In the first step, obtain the luminance reconstruction sample pY[x][y] of the isotopic luminance block with x=0...nTbW*2-1, y=0...nTbH*2-1, In the second step, obtain the adjacent luminance reconstruction sample pY[x][y], In the third step, obtain the downsampled luminance reconstruction sample pDsY[x][y] with x=0...nTbW-1, y=0...nTbH-1, In step 4, if numSampL is greater than 0, the chromaticity value pSelC[idx] of the point selected by the left side is set to p[-1][pickPosL[idx]] with idx=0...cntL-1, and the downsampled reconstructed luminance value pSelDsY[idx] with idx=0...cntL-1 of the point selected by the left side is obtained.

[0068] In step 5, if numSampT is greater than 0, the chromaticity value pSelC[idx] of the point selected by the upper side is set to p[pickPosT[idx-cntL]][-1] with idx=cntL...cntL+cntT-1, and the downsampled reconstructed luminance value pSelDsY[idx] with idx=0...cntL+cntT-1 of the point selected by the upper side is obtained.

[0069] In step 6, if cntT + cntL is not equal to 0, the derivation of the variables minY, maxY, minC, and maxC is as follows:

[0070] If cntT + cntL is equal to 2, then pSelComp[3] is set to pSelComp[0], pSelComp[2] to pSelComp[1], pSelComp[0] to pSelComp[1], and pSelComp[1] to pSelComp[3], where Comp is replaced by DsY and C respectively to represent the reconstructed luminance and chromaticity of the selected adjacent sample.

[0071] The derivation of array minGrpIdx and maxGrpIdx is as follows:

[0072] minGrpIdx[0]=0 minGrpIdx[1]=2 maxGrpIdx[0]=1 maxGrpIdx[1]=3 If pSelDsY[minGrpIdx[0]] is greater than pSelDsY[minGrpIdx[1]], then minGrpIdx[0] is swapped with minGrpIdx[1]. (minGrpIdx[0],minGrpIdx[1])=Swap(minGrpIdx[0],minGrpIdx[1]) If pSelDsY[maxGrpIdx[0]] is greater than pSelDsY[maxGrpIdx[1]], then maxGrpIdx[0] is swapped with maxGrpIdx[1]. (maxGrpIdx[0],maxGrpIdx[1])=Swap(maxGrpIdx[0],maxGrpIdx[1]) If pSelDsY[minGrpIdx[0]] is greater than pSelDsY[maxGrpIdx[1]], then array minGrpIdx is swapped with maxGrpIdx. (minGrpIdx,maxGrpIdx)=Swap(minGrpIdx,maxGrpIdx) If pSelDsY[minGrpIdx[1]] is greater than pSelDsY[maxGrpIdx[0]], then minGrpIdx[1] swaps with maxGrpIdx[0]. (minGrpIdx[1],maxGrpIdx[0])=Swap(minGrpIdx[1],maxGrpIdx[0]) The calculations for the variables maxY, maxC, minY, and minC are as follows (each representing the average value of two groups):

[0073] maxY=(pSelDsY[maxGrpIdx[0]]+pSelDsY[maxGrpIdx[1]]+1)>>1 maxC=(pSelC[maxGrpIdx[0]]+pSelC[maxGrpIdx[1]]+1)>>1 minY=(pSelDsY[minGrpIdx[0]]+pSelDsY[minGrpIdx[1]]+1)>>1 minC=(pSelC[minGrpIdx[0]]+pSelC[minGrpIdx[1]]+1)>>1 In step 7, the derivation process for the linear model parameters a, b, and k is as follows (where a is the slope (the ratio of the chromaticity difference to the luminance difference), b is the interception, k is the shift of a, and a is stored as an integer):

[0074] If numSampL is equal to 0 and numSampT is equal to 0, k=0 a=0 b=1<<(BitDepth C -1) Otherwise, diff = maxY - minY If diff is different from 0, diffC = maxC - minC x = Floor(Log2(diff)) normDiff=((diff<<4)>>x)&15 x += (normDiff != 0) ? 1 : 0 y = Floor(Log2(Abs(diffC))) + 1 a=(diffC*(divSigTable[normDiff]|8)+2y-1)>>y k=((3+xy)<1)?1:3+xy a=((3+xy)<1)?Sign(a)*15:a b = minC - ((a*minY)>>k) Here, divSigTable[ ] is divSigTable[ ]={0,7,6,5,5,4,4,3,3,2,2,1,1,1,1,0}, Otherwise (if diff is equal to 0), k=0 a=0 b = min C In step 8, the chromaticity prediction sample predSamples[x][y] with x=0...nTbW-1, y=0...nTbH-1 is obtained through the following calculation (where Clip1C restricts the predicted value to between 0 and 1023): predSamples[x][y]=Clip1C(((pDsY[x][y]*a)>>k)+b)

[0075] JPEG0007901221000005.jpg49170

[0076] JPEG0007901221000006.jpg99170

[0077] JPEG0007901221000007.jpg85170

[0078] JPEG0007901221000008.jpg86170

[0079] In this way, the second reference pixel set can be obtained by screening the first reference pixel set for cases where the number of effective pixel points in the first reference pixel set is greater than or equal to the preset number, and the second reference pixel set contains four effective pixel points.

[0080] In S504, if the number of effective pixel points in the second reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components.

[0081] In S505, if the number of effective pixel points in the second reference pixel set is equal to the number of presets, the model parameters are determined via the second reference pixel set, and a prediction model corresponding to the predicted image component is obtained according to the model parameters.

[0082] Note that after screening the first set of reference pixels, a second set of reference pixels is obtained. The number of effective pixel points in the second set of reference pixels may be less than the number of presets, or greater than or equal to the number of presets. If the number of effective pixel points in the second set of reference pixels is less than the number of presets, the component values ​​of the presets are used directly as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the second set of reference pixels is greater than or equal to the number of presets, the model parameters are determined through the second set of reference pixels, and a predictive model corresponding to the predicted image components is obtained according to the model parameters. Note that since there are usually four reference pixel points used to derive the model parameters, the second set of reference pixels obtained by screening will either be less than the number of presets (the second set of reference pixels is less than four effective pixel points) or equal to the number of presets (the second set of reference pixels contains four effective pixel points).

[0083] Furthermore, it should be noted that the prediction model may be a linear model or a nonlinear model. Here, the nonlinear model may be a nonlinear shape such as a quadratic curve, or a nonlinear shape formed by multiple linear models. For example, the cross-component prediction technique of Multi-Model CCLM (MMLM) is a nonlinear shape formed by multiple linear models, and is not particularly limited in the embodiments of this application. Here, the prediction model can be used to perform prediction processing on the predicted image components and obtain predicted values ​​corresponding to the predicted image components.

[0084] After obtaining the second reference pixel set, if the number of effective pixel points in the second reference pixel set is equal to the number of presets, the model parameters can be determined according to the second reference pixel set. After deriving the model parameters, a prediction model corresponding to the chromaticity component can be obtained according to the model parameters, as shown in equation (1). Then, using this prediction model, a prediction process is performed on the chromaticity component to obtain the predicted value corresponding to the chromaticity component.

[0085] Furthermore, in some embodiments, after obtaining a prediction model corresponding to the predicted image component according to the model parameters for S505, the method proceeds as follows: Based on the prediction model, this may further include performing a prediction process on the predicted image components of each pixel point in the coding block to obtain a predicted value corresponding to the predicted image component of each pixel point.

[0086] Note that when the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, it is necessary to determine model parameters (e.g., α and β) via the first reference pixel set in order to obtain predicted values ​​corresponding to the predicted image components of each pixel point in the coding block, and then obtain a prediction model corresponding to the predicted image components according to the model parameters. For example, assuming that the predicted image component is the chromaticity component, a prediction model corresponding to the chromaticity component shown in equation (1) can be obtained according to the model parameters α and β, and then the prediction process can be performed on the chromaticity component of each pixel point in the coding block using the prediction model shown in equation (1), and in this way, a predicted value corresponding to the chromaticity component of each pixel point can be obtained.

[0087] In the embodiments of this application, there may be effective regions and ineffective regions in the left adjacent region, the lower left adjacent region, the upper adjacent region, and the upper right adjacent region, and as a result, the number of effective pixel points selected from the adjacent region may be less than the number of presets. Therefore, in some embodiments, with reference to Figure 7, an exemplary flowchart of a different image component prediction method according to the embodiments of this application is shown. As shown in Figure 7, after S501, the method may further include the following steps.

[0088] In S701, the number of effective pixel points in the first reference pixel set is determined, and it is determined whether the number of effective pixel points is less than the preset number.

[0089] Furthermore, after S503, the method may further include the following steps.

[0090] In S702, it is determined whether the number of effective pixel points in the second reference pixel set is less than the number of presets.

[0091] It should be noted that the number of effective pixel points can be determined according to the effectiveness of adjacent regions. After determining the number of effective pixel points in this way, the number of effective pixel points is compared with the number of presets. If the number of effective pixel points in the first reference pixel set is less than the number of presets, step S502 is executed. If the number of effective pixel points in the second reference pixel set is less than the number of presets, step S504 is executed. If the number of effective pixel points in the second reference pixel set is greater than or equal to the number of presets, step S505 is executed.

[0092] Furthermore, please note that the number of presets can be as many as four. The following provides a detailed explanation using the example of having four presets.

[0093] In one possible embodiment, if the number of effective pixel points in the first reference pixel set is greater than or equal to the preset number, since there are typically four reference pixel points used to derive the model parameters, the first reference pixel set can be screened to have four effective pixel points, then the model parameters can be derived according to these four effective pixel points, a prediction model corresponding to the predicted image components can be obtained according to these model parameters, and a prediction value corresponding to the predicted image components can be obtained.

[0094] Specifically, we assume that the predicted image component is the chromaticity component, and that we predict the chromaticity component via the luminance component. We assume that the numbers of the four effective pixel points selected by screening are 0, 1, 2, and 3, respectively. By comparing these four selected effective pixel points, we can further select two pixel points with relatively large luminance values ​​(which may include the pixel point with the largest luminance value and the pixel point with the next largest luminance value) and two pixel points with relatively small luminance values ​​(which may include the pixel point with the smallest luminance value and the pixel point with the next smallest luminance value). Furthermore, we set up two arrays, minIdx[2] and maxIdx[2], to store two groups of pixel points, respectively, and initially, we first put the effective pixel points numbered 0 and 2 into minIdx[2] and the effective pixel points numbered 1 and 3 into maxIdx[2], as follows.

[0095] Init:minIdx[2]={0,2},maxIdx[2]={1,3} Subsequently, through four comparisons, it is possible to ensure that minIdx[2] contains two pixel points with relatively small luminance values, and maxIdx[2] contains two pixel points with relatively large luminance values, specifically as follows:

[0096] Step 1: if (L[minIdx[0]]>L[minIdx[1]],swap(minIdx[0],minIdx[1]) Step 2: if (L[maxIdx[0]]>L[maxIdx[1]],swap(maxIdx[0],maxIdx[1]) Step 3: if (L[minIdx[0]]>L[maxIdx[1]],swap(minIdx,maxIdx) Step 4: if (L[minIdx[1]]>L[maxIdx[0]],swap(minIdx[1],maxIdx[0]) In this way, two pixel points with relatively small luminance values ​​can be obtained, and the corresponding luminance values ​​are given by luma 0 min and luma 1 minas shown, the corresponding chroma values are respectively chroma 0 min and chroma 1 min as shown. At the same time, two pixel points with relatively large luminance values can be obtained, and the corresponding luminance values are respectively luma 0 max and luma 1 max as shown, and the corresponding chroma values are respectively chroma 0 max and chroma 1 max as shown. Further, an average value calculation can be performed on the luminance values corresponding to two relatively small pixel points to obtain the luminance value corresponding to the first average value point, which is shown as luma min as shown. An average value calculation can be performed on the luminance values corresponding to two relatively large pixel points to obtain the luminance value corresponding to the second average value point, which is shown as luma max as shown. Similarly, the chroma values corresponding to the two average value points can also be obtained, which are respectively chroma min and chroma max as shown. Specifically, it is as follows.

[0097] luma min =(luma 0 min +luma 1 min +1)>>1 luma max =(luma 0 max +luma 1 max +1)>>1 chroma min =(chroma 0 min +chroma 1 min +1)>>1 chroma max =(chroma 0 max +chroma 1 max +1)>>1 That is, the two average value points (luma min , chroma min) and (luma max chroma max After obtaining the points (), the model parameters can be obtained using a calculation method in which "two points determine a straight line". Specifically, the model parameters α and β can be obtained by calculating using equation (2).

number

[0098] Here, the model parameter α is the slope within the prediction model, and the model parameter β is the interception within the prediction model. After deriving the model parameters in this way, a prediction model corresponding to the chromaticity component can be obtained according to the model parameters, as shown in equation (1). Then, using this prediction model, a prediction process is performed on the chromaticity component to obtain the predicted value corresponding to the chromaticity component.

[0099] In another possible embodiment, for some special cases, such as the boundaries of an encoded block, when unpredictable and when adjacent reference pixel points cannot be obtained by the encoding order, and furthermore, when the encoding block is divided according to tile, slice, the left adjacent region, the lower left adjacent region, the upper adjacent region, and the upper right adjacent region are not all valid regions and may contain invalid regions, thereby reducing the number of valid pixel points in the first reference pixel set to less than the preset number.

[0100] Thus, since there are four presets, in some embodiments, for S502, if the number of effective pixel points in the first reference pixel set is less than the number of presets, the component values ​​of the presets can be used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the first reference pixel set is 0 or 2, this may include using preset component values ​​as predicted values ​​corresponding to the predicted image components.

[0101] In some embodiments, for S504, if the number of effective pixel points in the current second reference pixel set is less than the number of presets, the preset component values ​​may be used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the second reference pixel set is 0 or 2, this may include using preset component values ​​as predicted values ​​corresponding to the predicted image components.

[0102] In other words, if there are four presets, the number of effective pixel points is 0 or 2, even if the number of effective pixel points in the first reference pixel set is less than the number of presets, or even if the number of effective pixel points in the second reference pixel set is less than the number of presets.

[0103] Specifically, if the sum of adjacent reference pixel points within the selected region used for encoding a block is 0, then 0 effective pixel points are selected. In the following three special cases, 0 effective pixel points are obtained:

[0104] JPEG0007901221000010.jpg33170

[0105] JPEG0007901221000011.jpg34170

[0106] JPEG0007901221000012.jpg34170

[0107] It should be noted that the number of effective pixel points being 0 is determined according to the effectiveness of the adjacent regions, that is, the number of effective pixel points in the first reference pixel set can be determined according to the effectiveness of the adjacent regions. If the number of effective pixel points is 0, the model parameter α can be set to 0, and the model parameter β can be set to a preset component value corresponding to the predicted image component.

[0108] JPEG0007901221000013.jpg80170

[0109] JPEG0007901221000014.jpg59170

[0110] JPEG0007901221000015.jpg46170

[0111] JPEG0007901221000016.jpg40170

[0112] JPEG0007901221000017.jpg41170

[0113] It should be noted that the presence of two effective pixel points can be determined according to the effectiveness of adjacent regions, according to the number of effective pixel points in a selected region, and according to other determination conditions, and is not particularly limited in the embodiments of this application. In this way, the number of effective pixel points in the first reference pixel set can be determined according to the effectiveness of adjacent regions.

[0114] In the prior art solution, if there are two effective pixel points, it is necessary to obtain four pixel points by duplicating those two effective pixel points. Exemplarily, assuming that the numbers of the four pixel points are 0, 1, 2, and 3, then number 0 is the selected second effective pixel point, number 1 is the selected first effective pixel point, number 2 is the selected second effective pixel point, and number 3 is the selected first effective pixel point. By determining the model parameters α and β according to the four pixel points numbered 0, 1, 2, and 3, a prediction model shown in equation (1) can be established, and through this prediction model, predicted values ​​corresponding to the predicted image components can be obtained.

[0115] JPEG0007901221000018.jpg75170

[0116] Therefore, in the prior art's technical solution, when the number of effective pixel points is two, an additional "duplication" operation is required to use the same processing module. This involves acquiring four pixel points and thereby deriving the model parameters according to the same operational process as when the number of effective pixel points is four, increasing the additional "duplication" operation. The acquired four pixel points still require four comparisons and four averaging calculations, thereby increasing the computational complexity. In the embodiment of the present invention, the processing when the number of effective pixel points is two is aligned with the processing when the number of effective pixel points is zero. In this case, the computational complexity can be reduced by directly using the same processing module without increasing the additional operation.

[0117] Referring to Figure 10, an exemplary flowchart of the model parameter derivation according to the embodiment of the present invention is shown. In Figure 10, assuming that the predicted image component is the chromaticity component, first, adjacent reference pixel points are obtained from the selected region to form a first adjacent reference pixel set, then the number of effective pixel points in the first adjacent reference pixel set is determined, and if the number of effective pixel points is greater than or equal to 4, a screening process is performed on the first reference pixel set to obtain a second reference pixel set, then the number of effective pixel points in the second adjacent reference pixel set is determined, and if the number of effective pixel points in the first or second reference pixel set is 0, the model parameter α is set to 0 and the model parameter β is set to its default value, in which case the predicted value corresponding to the chromaticity component is populated with the default value. If the number of effective pixel points in the first or second reference pixel set is 2, the processing steps are the same as the processing steps when the number of effective pixel points is 0. In the case where there are four effective pixel points in the second reference pixel set, first, two pixel points with relatively large chromaticity components and two pixel points with relatively small chromaticity components are obtained through four comparisons. Then, two mean points are obtained, and based on these two mean points, the model parameters α and β are derived, and the chromaticity component prediction process is performed according to the constructed prediction model. Therefore, the model parameters in CCLM mode can be derived only for coding blocks that satisfy the condition that there are four effective pixel points in the second reference pixel set. For coding blocks with fewer than four effective pixel points, the method of filling with default values ​​can be used directly to reduce the computational complexity when the number of reference pixel points in the selected region is four or less, and furthermore, the coding and decoding performance can be kept essentially unchanged.

[0118] In this embodiment, the model parameter derivation process is integrated, namely, regarding the number of effective pixel points used for deriving the model parameters in the first reference pixel set, if the number of effective pixel points in the first reference pixel set is greater than or equal to the preset number, a screening of effective pixel points is performed on the first reference pixel set to obtain the second reference pixel set, and then the number of effective pixel points in the second adjacent reference pixel set is determined. If the number of effective pixel points in the second reference pixel set satisfies the preset number, the current coding block must perform the steps of deriving the model parameters and building the predictive model in CCLM; if the number of effective pixel points in the first or second reference pixel set is less than the preset number, the current coding block can use default values ​​to populate the predicted values ​​corresponding to the predicted image components of the coding block. Thus, as shown in Figure 11, this embodiment can further provide a simplified process for model parameter derivation.

[0119] Compared to Figure 10, the derivation process of the model parameters shown in Figure 11 is more streamlined. In Figure 11, assuming that the predicted image component is the chromaticity component and the preset component value is 512, first, adjacent reference pixel points are obtained from the selected region to form a first adjacent reference pixel set. Then, the number of effective pixel points in the first adjacent reference pixel set is determined. If the number of effective pixel points in the first reference pixel set is greater than or equal to the preset number, a screening process is performed on the first reference pixel set to obtain a second reference pixel set. Then, the number of effective pixel points in the second adjacent reference pixel set is determined. If the number of effective pixel points in the first or second reference pixel set is less than the preset number, the model parameter α is set to 0 and the model parameter β is set to 512. In this case, the predicted value corresponding to the chromaticity component is populated with 512. If the number of effective pixel points in the second reference pixel set satisfies the preset number, first, two pixel points representing relatively large chromaticity components and two pixel points representing relatively small chromaticity components are obtained through four comparisons. Then, two mean points are obtained, and based on these two mean points, model parameters α and β are derived. Finally, the chromaticity component prediction process is performed according to the constructed prediction model. It should be noted that the derivation of model parameters in CCLM mode can only be performed for coding blocks where the number of effective pixel points in the second reference pixel set satisfies the preset number. For coding blocks where the number of effective pixel points is less than the preset number, the method of directly filling with default values ​​can be used to reduce the complexity of calculations when the number of reference pixel points in the selected region is less than or equal to the preset number, and furthermore, the coding and decoding performance can be kept essentially unchanged. Typically, the number of presets in this embodiment may be four.

[0120] Furthermore, in the embodiment of the present invention, the number of effective pixel points in the first reference pixel set can be determined according to the number of effective pixel points in the selected region. Thus, as shown in Figure 12, the embodiment of the present invention can further provide a simplified process for deriving other model parameters.

[0121] In Figure 12, assuming that the predicted image component is the chromaticity component and the preset component value is 512, first, a selection region is determined to obtain the first reference pixel set. Then, the number of effective pixel points in the first reference pixel set is determined. If the number of effective pixel points in the first reference pixel set is greater than or equal to the preset number, a screening process is performed on the first reference pixel set to obtain the second reference pixel set. Next, the number of effective pixel points in the second adjacent reference pixel set is determined. If the number of effective pixel points in the first or second reference pixel set is less than the preset number, the predicted value corresponding to the chromaticity component is filled with 512. If the number of effective pixel points in the second reference pixel set satisfies the preset number, first, two pixel points representing a relatively large chromaticity component and two pixel points representing a relatively small chromaticity component are obtained through four comparisons. Then, two mean points are obtained, and based on the two mean points, model parameters α and β are derived, and the chromaticity component prediction process is performed according to the constructed prediction model. Therefore, the derivation of model parameters in CCLM mode can be performed only for coding blocks where the number of effective pixel points in the second reference pixel set satisfies the number of presets, and the method of directly filling with default values ​​for coding blocks where the number of effective pixel points is less than or equal to the number of presets can be used to reduce the complexity of calculations when the number of reference pixel points in the selected region is less than or equal to the number of presets, and furthermore, the coding and decoding performance can be kept essentially unchanged. Typically, the number of presets in this embodiment may be four.

[0122] JPEG0007901221000019.jpg123170

[0123] Furthermore, VVC defines the case where both variables numSampL and numSampT are 0 (in which case the number of effective pixel points screened and used to derive the model parameters can be set to 0), and the predicted values ​​corresponding to the chromaticity component are directly set to their default values; otherwise, the model parameters must be derived, specifically as follows:

[0124] if(numSampL==0&&numSampT==0) Set the predicted value corresponding to the chromaticity component to the default value. else Model parameters are derived, and the constructed predictive model is used to predict the chromaticity component and obtain the predicted value corresponding to the chromaticity component.

[0125] JPEG0007901221000020.jpg55170

[0126] if(numSampL+numSampT<number of presets) Set the predicted value corresponding to the chromaticity component to the default value. else Model parameters are derived, and the constructed predictive model is used to predict the chromaticity component and obtain the predicted value corresponding to the chromaticity component.

[0127] In this embodiment, by obtaining a first reference pixel set corresponding to the predicted image components of an encoded block in a video image, if the number of effective pixel points in the first reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set is screened to obtain a second reference pixel set. If the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets, and if the number of effective pixel points in the second reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the prime set is equal to the number of presets, model parameters are determined via a second reference pixel set, and a prediction model corresponding to the predicted image component is obtained according to the model parameters. This prediction model is used to perform prediction processing on the predicted image component and obtain the predicted value corresponding to the predicted image component. In this way, the derivation process of the model parameters is integrated, assuming that the encoding / decoding prediction performance is not changed. At the same time, in cases where the number of effective pixel points in the adjacent reference pixel set is less than the number of presets, or sometimes when there are 0 or 2 effective pixel points, no additional processing is required as no additional processing modules are added, thus reducing computational complexity.

[0128] In another embodiment of the present invention, if the number of effective pixel points in the first reference pixel set is less than the number of presets, the CCLM mode can be directly disabled and the preset component values ​​can be used as predicted values ​​corresponding to the predicted image components. Therefore, in some embodiments, after screening the first reference pixel set to obtain the second reference pixel set, the method can be used. If the number of effective pixel points in the first reference pixel set is less than the number of presets, or if the number of effective pixel points in the second reference pixel set is less than the number of presets, the component values ​​of the presets are used as the predicted values ​​corresponding to the predicted image components. The further includes, if the number of effective pixel points in the second reference pixel set is greater than or equal to the number of presets, adopting CCLM mode to perform predictive processing on the predicted image components.

[0129] Note that if the number of effective pixel points in the first or second reference pixel set is less than the preset number, the CCLM mode can be disabled. For example, the identifier for whether to use CCLM mode can be set to "Disable CCLM mode". In this case, the predicted values ​​corresponding to the predicted image components are directly populated with default values. The CCLM mode can only be used if the number of effective pixel points in the second reference pixel set is greater than or equal to the preset number. For example, the identifier for whether to use CCLM mode can be set to "Enable CCLM mode". In this case, prediction processing for the predicted image components can be performed via CCLM mode.

[0130] Furthermore, assuming that the predicted image component is the chromaticity component and there are four presets, it should be noted that for all cases where two effective pixel points can be generated (where the method for determining whether to generate two effective pixel points is not particularly limited in this embodiment), the model parameter α can be set to 0 and the model parameter β can be set to an intermediate value (which may also be called a default value) corresponding to the chromaticity component, thereby filling the predicted values ​​corresponding to the chromaticity component of all pixel points in the coding block with the default value. Additionally, for all cases where two effective pixel points can be generated, both numSampL and numSampT can be set to 0 to fill the predicted values ​​corresponding to the chromaticity component of all pixel points in the coding block with the default value.

[0131] Furthermore, for all cases where two effective pixel points may be generated, the predicted values ​​corresponding to the chromaticity component can be directly populated with default values, or the CCLM mode can be disabled for all cases where two effective pixel points may be generated, or the CCLM mode can be disabled for all cases where two or zero effective pixel points may be generated, or the predicted values ​​corresponding to the chromaticity component can be directly populated with default values ​​for all cases where two or zero effective pixel points may be generated.

[0132] Therefore, this approach unifies the model parameter derivation process when the number of effective pixel points used in model parameter derivation differs. Specifically, when there are two effective pixel points, no additional processing is required; the existing processing module is directly read (i.e., the processing for two effective pixel points is matched to the processing for zero effective pixel points), thereby reducing computational complexity.

[0133] The image component prediction method in this embodiment is based on the latest VVC reference software VTM5.0. Under all intra conditions, the test sequence requested from JVET, according to general test conditions, showed average BD-rate changes of 0.00%, 0.02%, and 0.02% in the Y, Cb, and Cr components, respectively. This demonstrates that this invention has virtually no impact on encoding and decoding performance.

[0134] Assuming that it does not affect encoding and decoding performance, this invention can have the following beneficial effects.

[0135] First, the present invention can integrate the process of deriving model parameters in CCLM mode. In the prior art solution, when there are two effective pixel points, it is necessary to perform an additional "duplication" operation to generate four usable pixel points, thereby performing the same operation as when there are four effective pixel points and further deriving the model parameters. However, the present invention reduces the additional "duplication" operation and, at the same time, can make the processing for the number of effective pixel points two the same as the processing for the number of effective pixel points zero, so that the same processing module can be used directly without increasing the number of additional operations, thereby integrating the process of deriving linear model parameters.

[0136] JPEG0007901221000021.jpg93170

[0137] This embodiment provides a method for predicting image components, and through the technical solution of this embodiment, the number of effective pixel points in a first reference pixel set is compared to a preset number. If the number of effective pixel points in the first or second reference pixel set is less than the preset number, the default value of the preset is directly used as the predicted value corresponding to the predicted image component. If the number of effective pixel points in the second reference pixel set is greater than or equal to the preset number, model parameters are determined according to the first reference pixel set to construct a prediction model for the predicted image component, thereby integrating the derivation process of the model parameters. Furthermore, the computational complexity is reduced because no additional processing modules are added for the case where the number of effective pixel points in the first reference pixel set is less than the preset number.

[0138] Referring to Figure 13, based on the same inventive concept as the above embodiment, an exemplary structural diagram of the image component prediction device 130 according to the present embodiment is shown. The image component prediction device 130 may comprise an acquisition unit 1301, a prediction unit 1302, and a screening unit 1303, where, The acquisition unit 1301 is configured to acquire a first set of reference pixels corresponding to the predicted image components of the encoded blocks in the video image. The prediction unit 1302 is configured to use the preset component values ​​as predicted values ​​corresponding to the predicted image components if the number of effective pixel points in the first reference pixel set is less than the number of presets. The screening unit 1303 is configured to screen the first reference pixel set to obtain a second reference pixel set if the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, where the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets. The prediction unit 1302 is further configured to use the component values ​​of a preset as predicted values ​​corresponding to the predicted image component if the number of effective pixel points in the second reference pixel set is less than the number of presets, and to determine model parameters via the second reference pixel set and obtain a prediction model corresponding to the predicted image component according to the model parameters, where the prediction model is used to perform prediction processing on the predicted image component and obtain predicted values ​​corresponding to the predicted image component.

[0139] In the above proposed technology, the acquisition unit 1301 specifically acquires reference pixel points adjacent to at least one side of the coding block, where the at least one side includes the left side and / or the upper side of the coding block, and is configured to form a first set of reference pixels corresponding to the predicted image components based on the reference pixel points.

[0140] In the above proposed technology, the acquisition unit 1301 specifically acquires reference pixel points in a reference row or reference column adjacent to the coding block, where the reference row is formed by rows adjacent to the upper and upper right sides of the coding block, and the reference column is formed by columns adjacent to the left and lower left sides of the coding block, and is configured to form a first set of reference pixels corresponding to the predicted image components based on the reference pixel points.

[0141] In the above proposed technology, the screening unit 1303 is configured to determine a selected pixel point position based on the pixel position and / or image component intensity corresponding to each adjacent reference pixel point in the first reference pixel set, to select an effective pixel point from the first reference pixel set corresponding to the determined selected pixel point position, and to form a second reference pixel set with the selected effective pixel points, where the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets.

[0142] Referring to Figure 13 in the above proposed technology, the image component prediction device 130 further comprises a determination unit 1304 configured to determine a preset component range corresponding to the predicted image component based on the bit information of the video image, determine an intermediate value of the preset component range according to the preset component range, and use the intermediate value as a predicted value corresponding to the predicted image component, where the intermediate value is shown in the preset component value.

[0143] Referring to Figure 13, the image component prediction device 130 further includes a filling unit 1305 configured to fill the predicted image component of each pixel point in the coding block with predicted values ​​using the preset component values.

[0144] In the above proposed technology, the prediction unit 1302 is further configured to perform a prediction process on the predicted image components of each pixel point in the coding block based on the prediction model, and to obtain a predicted value corresponding to the predicted image component of each pixel point.

[0145] In the above proposed technology, the value of the number of presets is 4; the prediction unit 1302 is further configured to use the preset component values ​​as predicted values ​​corresponding to the predicted image components when the number of effective pixel points in the first reference pixel set is 0 or 2. Accordingly, the prediction unit 1302 is further configured to use a preset component value as a predicted value corresponding to the predicted image component if the number of effective pixel points in the second reference pixel set is 0 or 2.

[0146] Referring to Figure 13 in the above technical proposal, the image component prediction device 130 further includes a determination unit 1306 configured to use a preset component value as a predicted value corresponding to the predicted image component if the number of effective pixel points in the first reference pixel set is less than the number of presets, or if the number of effective pixel points in the second reference pixel set is less than the number of presets, and to adopt CCLM mode to perform prediction processing for the predicted image component if the number of effective pixel points in the second reference pixel set is greater than or equal to the number of presets.

[0147] In this embodiment, the "unit" may be part of a circuit, part of a processor, part of a program or software, etc., and may be modular or non-modular. Furthermore, each component in this embodiment may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated unit can be implemented in hardware form or in the form of a software function module.

[0148] Based on the understanding that the integrated unit can be stored in a single computer-readable storage medium if it is implemented in the form of a software function module and is not sold or used as an independent product, the technical solution of this embodiment, in essence or in part with respect to the prior art, or all or part of the technical solution, can be embodied in the form of a computer software product, which is stored in a single storage medium and contains several instructions to cause a single computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the method described in this embodiment. The aforementioned storage medium includes various media capable of storing program code, such as U disks, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0149] Accordingly, this embodiment provides a computer storage medium that stores an image component prediction program, and when the image component prediction program is executed by at least one processor, it realizes the steps of the method described in any one of the above embodiments.

[0150] Referring to Figure 14, based on the above combination of image component prediction device 130 and computer storage medium, the specific hardware structure of the image component prediction device 130 according to this embodiment of the application is shown, comprising a network interface 1401, memory 1402 and processor 1403, each component being coupled together via a bus system 1404. Understandably, the bus system 1404 is used to embody connection communication between these components. In addition to the data bus, the bus system 1404 includes a power bus, a control bus and a status signal bus. However, for clarity of explanation, various buses are represented as the bus system 1404 in Figure 14. Here, the network interface 1401 is configured to send and receive signals in the process of sending and receiving information with another external network element, Memory 1402 is configured to store computer programs that can be executed by processor 1403. The processor 1403 is configured to perform the following steps by executing the computer program, the steps being: Obtaining a first reference pixel set corresponding to the predicted image components of an encoded block in a video image, If the number of effective pixel points in the first reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components; if the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set is screened to obtain a second reference pixel set, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets. The process includes: if the number of effective pixel points in the second reference pixel set is less than the number of presets, using the component values ​​of the presets as predicted values ​​corresponding to the predicted image components; and if the number of effective pixel points in the second reference pixel set is equal to the number of presets, determining model parameters via the second reference pixel set and obtaining a prediction model corresponding to the predicted image components according to the model parameters, wherein the prediction model is used to perform prediction processing on the predicted image components and obtain predicted values ​​corresponding to the predicted image components.

[0151] It should be understood that the memory 1402 in this embodiment may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Here, non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may be random access memory (RAM) used as an external cache. While illustrative and not limiting, various forms of RAM can be used, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM). The memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0152] The processor 1403 may be an integrated circuit chip having signal processing capabilities. In the implementation process, each step of the above method can be completed by instructions in the form of hardware integrated logic circuits or software within the processor 1403. The processor 1403 may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates, or transistor logic devices, discrete hardware components, etc. Each method, step, and logic block diagram disclosed in the present embodiment can be implemented or executed. The general-purpose processor may be a microprocessor, or the processor may be any conventional processor, etc. The steps of the method disclosed in the present embodiment may be performed directly by a hardware decoding processor, or by a combination of hardware and software modules within the decoding processor. The software modules may be located in conventional storage media such as random-access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers, etc. The storage medium is placed in memory 1402, and the processor 1403 reads the information in memory 1402 and combines it with its hardware to complete the steps of the method.

[0153] It can be understood that these embodiments described herein can be implemented in hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementations, the processing unit may be implemented in one or more application-specific integrated circuits (ASICs), digital signal processing (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers, microprocessors, other electronic units configured to perform the functions described herein, or a combination thereof.

[0154] In the case of execution using software, the technologies described herein can be realized through modules (processes, functions, etc.) that perform the functions described herein. The software code can be stored in memory and executed by a processor. Memory can be implemented within or outside the processor.

[0155] Exemplary, in another embodiment, the processor 1403 is further configured to perform the steps of the method described in any one of the above embodiments by executing the computer program.

[0156] Referring to Figure 15, an exemplary structural diagram of the encoder configuration according to the embodiment of the present invention is shown. As shown in Figure 15, the encoder 150 may include at least the image component prediction device 130 described in any one of the embodiments above.

[0157] Referring to Figure 16, an exemplary structural diagram of the decoder configuration according to the embodiment of the present invention is shown. As shown in Figure 16, the decoder 160 may include at least the image component prediction device 130 described in any one of the embodiments above.

[0158] In this specification, the terms “compose,” “include,” or any other variation thereof are intended to cover non-exclusive implications, so that a process, method, item, or apparatus containing a set of elements may include not only those elements but also other elements not expressly enumerated, or any elements inherent to those processes, methods, items, or apparatus. Without further restriction, an element limited by the phrase “includes one…” does not preclude the presence of other related elements in a process, method, item, or apparatus containing that element.

[0159] The numbers in the above-mentioned embodiments of this application do not indicate any ranking of the embodiments, but are provided for the convenience of explanation.

[0160] The methods disclosed in some embodiments of the present application can be arbitrarily combined without conflict to obtain new embodiments of the method.

[0161] The features disclosed in some product embodiments of this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0162] The features disclosed in some of the embodiments of the methods or apparatus of this application can be arbitrarily combined without conflict to obtain new embodiments of the methods or apparatus.

[0163] The foregoing describes only specific embodiments of the present application, and the scope of protection of this application is not limited thereto. Any modification or substitution that a person skilled in the art can easily conceive within the technical scope disclosed herein should be included within the scope of protection of this application. Accordingly, the scope of protection of this application shall be subject to the scope of protection of the claims. [Industrial applicability]

[0164] In this embodiment, by obtaining a first reference pixel set corresponding to the predicted image components of an encoded block in a video image, if the number of effective pixel points in the first reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the first reference pixel set is greater than or equal to the number of presets, the first reference pixel set is screened to obtain a second reference pixel set, the number of effective pixel points in the second reference pixel set is less than or equal to the number of presets, and if the number of effective pixel points in the second reference pixel set is less than the number of presets, the component values ​​of the presets are used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the second reference pixel set is equal to the number of presets, model parameters are determined via the second reference pixel set, and a prediction model corresponding to the predicted image components is obtained according to the model parameters, and the prediction model corresponds to the predicted image components. This method implements prediction processing to obtain predicted values ​​corresponding to the predicted image components. In this way, if the number of effective pixel points in the first reference pixel set is less than the preset number, or if the number of effective pixel points in the second reference pixel set is less than the preset number, the default values ​​of the presets are used directly as the predicted values ​​corresponding to the predicted image components. Only when the number of effective pixel points in the second reference pixel set satisfies the preset number, the model parameters are determined according to the first reference pixel set to establish a prediction model for the predicted image components, thereby integrating the derivation process of the model parameters. Furthermore, if the number of effective pixel points in the first or second reference pixel set is less than the preset number, especially when the number of effective pixel points is 0 or 2, no additional processing modules are required, so the default values ​​of the presets are used directly as the predicted values ​​corresponding to the predicted image components, reducing computational complexity.

Claims

1. A method for predicting image components, which is applied to a decoder, Obtaining a first set of reference pixels corresponding to the predicted image components of the decoded blocks in the video image, If the number of effective pixel points in the first reference pixel set is equal to 0, the preset component values ​​are used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the first reference pixel set is greater than or equal to a preset number, the first reference pixel set is screened to obtain a second reference pixel set, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the preset number. The process includes determining model parameters via the second reference pixel set if the number of effective pixel points in the second reference pixel set is equal to the preset number, and obtaining a prediction model corresponding to the predicted image component according to the model parameters, wherein the prediction model is used to perform prediction processing on the predicted image component and obtain a predicted value corresponding to the predicted image component. Screening the first set of reference pixels to obtain a second set of reference pixels is: The selected pixel point position is determined based on the pixel position corresponding to the adjacent reference pixel point in the first set of reference pixels, A method for predicting image components, comprising: selecting effective pixel points from a first reference pixel set corresponding to the selected pixel point positions according to the determined selected pixel point positions; and determining a second reference pixel set based on the selected effective pixel points, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the preset number.

2. Obtaining a first reference pixel set corresponding to the predicted image components of the decoding block in the video image is: Obtaining a reference pixel point adjacent to at least one side of the decoding block, wherein the at least one side includes the left side and / or the upper side of the decoding block. This includes determining a first set of reference pixels corresponding to the predicted image components based on the aforementioned reference pixel points, The method for predicting image components according to claim 1.

3. Obtaining a first reference pixel set corresponding to the predicted image components of the decoding block in the video image is: Obtaining reference pixel points in reference rows and / or reference columns adjacent to the decoding block, wherein the reference rows are formed by rows adjacent to the upper and upper right sides of the decoding block, and the reference columns are formed by columns adjacent to the left and lower left sides of the decoding block. This includes determining a first set of reference pixels corresponding to the predicted image components based on the aforementioned reference pixel points, The method for predicting image components according to claim 1.

4. Using the preset component values ​​as predicted values ​​corresponding to the predicted image components is, The process includes determining the preset component values ​​based on the bit depth of the video image, The method for predicting image components according to claim 1.

5. The preset component value is 1 << (BitDepth - 1), where BitDepth is the bit depth of the predicted image component. The method for predicting image components according to claim 4.

6. The method for predicting the aforementioned image components is: The further step includes, for the pixel points in the decoding block, performing predicted value filling for the predicted image components of the pixel points using the preset component values, The method for predicting image components according to claim 1.

7. After obtaining a prediction model corresponding to the predicted image component according to the model parameters, the method for predicting the image component is: The process further includes performing a prediction process on the predicted image components of the pixel points in the decoding block based on the prediction model, and obtaining predicted values ​​corresponding to the predicted image components of the pixel points. The method for predicting image components according to claim 1.

8. A method for predicting image components, which is applied to an encoder, Obtaining a first set of reference pixels corresponding to the predicted image components of an encoded block in a video image, If the number of effective pixel points in the first reference pixel set is equal to 0, the preset component values ​​are used as predicted values ​​corresponding to the predicted image components. If the number of effective pixel points in the first reference pixel set is greater than or equal to a preset number, the first reference pixel set is screened to obtain a second reference pixel set, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the preset number. The process includes determining model parameters via the second reference pixel set if the number of effective pixel points in the second reference pixel set is equal to the preset number, and obtaining a prediction model corresponding to the predicted image component according to the model parameters, wherein the prediction model is used to perform prediction processing on the predicted image component and obtain a predicted value corresponding to the predicted image component. Screening the first set of reference pixels to obtain a second set of reference pixels is: The selected pixel point position is determined based on the pixel position corresponding to the adjacent reference pixel point in the first set of reference pixels, A method for predicting image components, comprising: selecting effective pixel points from a first reference pixel set corresponding to the selected pixel point positions according to the determined selected pixel point positions; and determining a second reference pixel set based on the selected effective pixel points, wherein the number of effective pixel points in the second reference pixel set is less than or equal to the preset number.

9. Obtaining a first reference pixel set corresponding to the predicted image components of the encoded block in the video image is: Obtaining a reference pixel point adjacent to at least one side of the coding block, wherein the at least one side includes the left side and / or the upper side of the coding block. This includes determining a first set of reference pixels corresponding to the predicted image components based on the aforementioned reference pixel points, The method for predicting image components according to claim 8.

10. Obtaining a first reference pixel set corresponding to the predicted image components of the encoded block in the video image is: Obtaining reference pixel points in reference rows and / or reference columns adjacent to the coding block, wherein the reference rows are formed by rows adjacent to the upper and upper right sides of the coding block, and the reference columns are formed by columns adjacent to the left and lower left sides of the coding block. This includes determining a first set of reference pixels corresponding to the predicted image components based on the aforementioned reference pixel points, The method for predicting image components according to claim 8.

11. Using the preset component values ​​as predicted values ​​corresponding to the predicted image components is, The process includes determining the preset component values ​​based on the bit depth of the video image, The method for predicting image components according to claim 8.

12. The preset component value is 1 << (BitDepth - 1), where BitDepth is the bit depth of the predicted image component. The method for predicting image components according to claim 11.

13. The method for predicting the aforementioned image components is: For pixel points within the coding block, the further includes performing predicted value filling for the predicted image components of the pixel points using the preset component values. The method for predicting image components according to claim 8.

14. After obtaining a prediction model corresponding to the predicted image component according to the model parameters, the method for predicting the image component is: The process further includes performing a prediction process on the predicted image components of the pixel points in the coding block based on the prediction model, and obtaining a predicted value corresponding to the predicted image component of the pixel point. The method for predicting image components according to claim 8.

15. A decoder comprising memory and a processor, The memory is configured to store computer programs that can be executed by the processor, The decoder is configured to perform the image component prediction method described in any one of claims 1 to 7 when the processor executes the computer program.

16. An encoder comprising memory and a processor, The memory is configured to store computer programs that can be executed by the processor, The encoder is configured to perform the image component prediction method described in any one of claims 8 to 14 when the processor executes the computer program.

17. A computer-readable storage medium, A computer-readable storage medium storing a computer program that, when executed by at least one processor, causes the processor to perform the image component prediction method described in any one of claims 1 to 7.

18. A method for generating a bitstream, wherein the encoder generates and outputs the bitstream by performing the image component prediction method described in any one of claims 8 to 14.

Images