WO2020258053A1

WO2020258053A1 - Image component prediction method and apparatus, and computer storage medium

Info

Publication number: WO2020258053A1
Application number: PCT/CN2019/092859
Authority: WO
Inventors: 霍俊彦; 马彦卓; 万帅; 杨付正; 李新伟
Original assignee: Oppo广东移动通信有限公司
Priority date: 2019-06-25
Filing date: 2019-06-25
Publication date: 2020-12-30
Also published as: CN113196770A

Abstract

An image component prediction method and apparatus, and a computer storage medium. The method comprises: obtaining N adjacent reference pixel points corresponding to an image component to be predicted of a coding block in a video image (S701), wherein N is a preset integer value; comparing N first image component values corresponding to the N adjacent reference pixel points, and determining a first reference pixel subset and a second reference pixel subset (S702), wherein the first reference pixel subset comprises two adjacent reference pixel points corresponding to a minimum first image component value and a secondary minimum first image component value, and the second reference pixel subset comprises two adjacent reference pixel points corresponding to a maximum first image component value and a secondary maximum first image component value; performing comparison of the first image component value on the first reference pixel subset and/or the second reference pixel subset by means of a first mean value point, and determining two fitting points (S704); and determining a model parameter according to the two fitting points, and according to the model parameter, obtaining a prediction model corresponding to the image component to be predicted.

Description

Image component prediction method, device and computer storage medium

Technical field

The embodiments of the present application relate to the field of video coding and decoding technologies, and in particular to an image component prediction method, device, and computer storage medium.

Background technique

As people's requirements for video display quality increase, new video applications such as high-definition and ultra-high-definition video have emerged. H.265/High Efficiency Video Coding (HEVC) has been unable to meet the needs of the rapid development of video applications. The Joint Video Exploration Team (JVET) proposed the next-generation video coding standard H.266/multiple Functional Video Coding (Versatile Video Coding, VVC), and its corresponding test model is the VVC Reference Software Test Platform (VVC Test Model, VTM).

In VTM, an image component prediction method based on a prediction model has been integrated, through which the chrominance component can be predicted from the luminance component of the current coding block (CB). However, when constructing a prediction model, the selection of adjacent reference pixels used for model parameter derivation is unreasonable, resulting in inaccurate prediction models and reducing the coding and decoding prediction performance of video images.

Summary of the invention

The embodiments of the application provide an image component prediction method, device, and computer storage medium. By optimizing the fitting points used in the derivation of model parameters, the robustness of prediction is improved, so that the constructed prediction model is more accurate, and Improved the codec prediction performance of video images.

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

In the first aspect, an embodiment of the present application provides an image component prediction method, the method includes:

Acquire N adjacent reference pixels corresponding to the to-be-predicted image components of the encoding block in the video image; wherein, the N adjacent reference pixels are reference pixels adjacent to the encoding block, and N is a preset Integer value

The N first image component values corresponding to the N adjacent reference pixels are compared to determine the first reference pixel subset and the second reference pixel subset; wherein the preset number of adjacent reference pixels corresponds to the preset number The first image component value of the first reference pixel subset includes: two adjacent references corresponding to the smallest first image component value and the second smallest first image component value among the preset number of first image component values Pixels, the second reference pixel subset includes: two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value among the preset number of first image component values;

Calculating the average value of the N adjacent reference pixel points to obtain the first average value point;

Comparing the values of the first image component to the first reference pixel subset and/or the second reference pixel subset by using the first average point to determine two fitting points;

Based on the two fitting points, the model parameters are determined, and the prediction model corresponding to the image component to be predicted is obtained according to the model parameters; wherein, the prediction model is used to realize the prediction processing of the image component to be predicted, To obtain the predicted value corresponding to the image component to be predicted.

In a second aspect, an embodiment of the present application provides an image component prediction device. The image component prediction device includes: an acquisition unit, a comparison unit, a calculation unit, and a prediction unit, wherein:

The acquiring unit is configured to acquire N adjacent reference pixels corresponding to the image components to be predicted of the encoding block in the video image; wherein the N adjacent reference pixels are reference pixels adjacent to the encoding block Point, N is a preset integer value;

The comparison unit is configured to compare the N first image component values corresponding to the N adjacent reference pixels to determine a first reference pixel subset and a second reference pixel subset; wherein, a preset number of adjacent The reference pixel points correspond to a preset number of first image component values, and the first reference pixel subset includes: among the preset number of first image component values, the smallest first image component value and the second smallest first image component value are Corresponding to two adjacent reference pixel points, the second reference pixel subset includes: among the preset number of first image component values, two corresponding to the largest first image component value and the second largest first image component value Adjacent reference pixels;

The calculation unit is configured to calculate the average value of the N adjacent reference pixel points to obtain a first average value point;

The comparison unit is further configured to compare the values of the first image component of the first reference pixel subset and/or the second reference pixel subset through the first average point, and determine two fitting points;

The prediction unit is configured to determine model parameters based on the two fitting points, and obtain a prediction model corresponding to the image component to be predicted according to the model parameters; wherein, the prediction model is used to implement Prediction processing of the predicted image component to obtain the predicted value corresponding to the image component to be predicted.

In a third aspect, an embodiment of the present application provides an image component prediction device, the image component prediction device including: a memory and a processor;

The memory is used to store a computer program that can run on the processor;

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

In a fourth aspect, an embodiment of the present application provides a computer storage medium, the computer storage medium stores an image component prediction program, and when the image component prediction program is executed by at least one processor, the method described in the first aspect is implemented. method.

The embodiments of the present application provide an image component prediction method, device, and computer storage medium. First, for the image component to be predicted of a coding block in a video image, N adjacent reference pixels are obtained, where the N adjacent reference pixels are Is the reference pixel adjacent to the coding block, and N is a preset integer value; then the N first image component values corresponding to the N adjacent reference pixels are compared to determine the first reference pixel subset and A second reference pixel subset, which includes: two adjacent reference pixels corresponding to the smallest first image component value and the second smallest first image component value among the preset number of first image component values Point, the second reference pixel subset includes: among the preset number of first image component values, two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value are calculated; The average value of N adjacent reference pixels is obtained to obtain the first average value point; the first image component value is compared with the first reference pixel subset and/or the second reference pixel subset through the first average value point to determine the two pseudo Combining point; then determine the model parameters based on the two fitting points, and obtain the prediction model corresponding to the image component to be predicted according to the model parameters to obtain the predicted value corresponding to the image component to be predicted; in this way, after the second comparison process, the prediction Assuming that the number of adjacent reference pixels is divided into multiple situations, using two different fitting points to construct the prediction model for these multiple situations can improve the robustness of CCLM prediction; that is, by deriving all the model parameters The fitting points used are optimized, so that the constructed prediction model is more accurate, and the coding and decoding prediction performance of the video image is improved.

Description of the drawings

Figure 1 is a schematic diagram of the distribution of an effective adjacent area provided by a related technical solution;

Figure 2 is a schematic diagram of the distribution of selected areas in one of three modes provided by related technical solutions;

Figure 3 is a schematic flow chart of a traditional solution for deriving model parameters provided by related technical solutions;

Fig. 4A is a schematic diagram of a prediction model under a traditional scheme provided by a related technical scheme;

4B is a schematic diagram of a prediction model under another traditional solution provided by the related technical solution;

FIG. 5 is a schematic diagram of the composition of a video encoding system provided by an embodiment of the application;

FIG. 6 is a schematic diagram of the composition of a video decoding system provided by an embodiment of the application;

FIG. 7 is a schematic flowchart of an image component prediction method provided by an embodiment of the application;

8A is a schematic structural diagram of adjacent reference pixel selection in INTRA_LT_CCLM mode according to an embodiment of the application;

8B is a schematic structural diagram of adjacent reference pixel selection in INTRA_L_CCLM mode according to an embodiment of the application;

FIG. 8C is a schematic structural diagram of adjacent reference pixel selection in the INTRA_T_CCLM mode according to an embodiment of the application;

9 is a schematic flowchart of another image component prediction method provided by an embodiment of the application;

10 is a schematic flowchart of a model parameter derivation solution provided by an embodiment of the application;

FIG. 11 is a schematic diagram of comparison between a prediction model provided by an embodiment of the present application and a traditional solution;

FIG. 12 is a schematic diagram of comparison between another solution of the application provided by an embodiment of the application and a prediction model under the traditional solution;

FIG. 13 is a schematic diagram of comparison between another solution of the application provided by an embodiment of the application and a prediction model under a traditional solution;

FIG. 14 is a schematic diagram of a structure for determining a fitting point provided by an embodiment of the application;

FIG. 15 is a schematic structural diagram of a prediction model provided by an embodiment of this application;

16 is a schematic diagram of the composition structure of an image component prediction apparatus provided by an embodiment of the application;

FIG. 17 is a schematic diagram of a specific hardware structure of an image component prediction apparatus provided by an embodiment of the application;

18 is a schematic diagram of the composition structure of an encoder provided by an embodiment of the application;

FIG. 19 is a schematic diagram of the composition structure of a decoder provided by an embodiment of the application.

Detailed ways

In order to have a more detailed understanding of the features and technical content of the embodiments of the present application, the implementation of the embodiments of the present application will be described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference and explanation purposes only and are not used to limit the embodiments of the present application.

In video images, the first image component, the second image component, and the third image component are generally used to characterize the coding block; among them, the three image components are a luminance component, a blue chrominance component, and a red chrominance component. Specifically, the luminance component is usually represented by the symbol Y, the blue chrominance component is usually represented by the symbol Cb or U, and the red chrominance component is usually represented by the symbol Cr or V; in this way, the video image can be represented in YCbCr format or YUV. Format representation.

In the embodiment of the present application, the first image component may be a luminance component, the second image component may be a blue chrominance component, and the third image component may be a red chrominance component, but the embodiment of the present application does not specifically limit it.

In the current video image or video encoding and decoding process, the cross-component prediction technology mainly includes the cross-component linear model prediction (CCLM) mode and the multi-directional linear model prediction (Multi-Directional Linear Model Prediction, MDLM) mode, whether it is based on the model parameters derived from the CCLM mode or the model parameters derived from the MDLM mode, its corresponding prediction model can realize the first image component to the second image component, and the second image component to the first image component , Prediction between image components such as the first image component to the third image component, the third image component to the first image component, the second image component to the third image component, or the third image component to the second image component.

Taking the prediction from the first image component to the second image component as an example, in order to reduce the redundancy between the first image component and the second image component, the CCLM mode is used in VVC. At this time, the first image component and the second image component For the same coding block, that is, the predicted value of the second image component is constructed according to the reconstruction value of the first image component of the same coding block, as shown in equation (1),

Pred _C [i,j]=α·Rec _L [i,j]+β (1)

Among them, i,j represent the position coordinates of the pixel in the coding block, i represents the horizontal direction, j represents the vertical direction, and Pred _C [i,j] represents the pixel corresponding to the location coordinate [i,j] in the coding block The predicted value of the second image component, Pred _L [i,j] represents the reconstructed value of the first image component corresponding to the pixel with the position coordinate [i,j] in the same coding block (downsampled), and α and β represent the model parameter.

For a coding block, its adjacent areas may include a left adjacent area, an upper adjacent area, a lower left adjacent area, and an upper right adjacent area. In VVC, three cross-component linear model prediction modes can be included, which are: the intra-CCLM mode adjacent to the left and upper side (can be represented by the INTRA_LT_CCLM mode), and the intra-CCLM mode adjacent to the left and lower left side (It can be represented by INTRA_L_CCLM mode) and the intra-frame CCLM mode adjacent to the upper side and the upper right side (can be represented by INTRA_T_CCLM mode). In these three modes, each mode can select a preset number (such as 4) of adjacent reference pixels for the derivation of model parameters α and β, and the biggest difference between these three modes is that they are used to derive the model The selected regions corresponding to the adjacent reference pixels of the parameters α and β are different.

Specifically, for the size of the coding block corresponding to the second image component is W×H, it is assumed that the upper selection area corresponding to the adjacent reference pixel is W', and the left selection area corresponding to the adjacent reference pixel is H'; ,

For the INTRA_LT_CCLM mode, adjacent reference pixels can be selected in the upper adjacent area and the left adjacent area, that is, W'=W, H'=H;

For INTRA_L_CCLM mode, adjacent reference pixels can be selected in the adjacent area on the left side and the adjacent area on the lower left side, that is, H'=W+H, and set W'=0;

For the INTRA_T_CCLM mode, adjacent reference pixels can be selected in the upper adjacent area and the upper right adjacent area, that is, W'=W+H, and set H'=0.

It should be noted that in the latest VVC reference software VTM5.0, for the upper right adjacent area, only pixels in the W range at most are stored, and for the lower left adjacent area, only pixels in the H range at most are stored; therefore Although the selection area of INTRA_L_CCLM mode and INTRA_T_CCLM mode is defined as W+H, in practical applications, the selection area of INTRA_L_CCLM mode will be limited to H+H, and the selection area of INTRA_T_CCLM mode will be limited to W+W. Inside; like this,

For INTRA_L_CCLM mode, adjacent reference pixels can be selected in the adjacent area on the left and the adjacent area on the lower left, H'=min{W+H,H+H};

For the INTRA_T_CCLM mode, adjacent reference pixels can be selected in the upper adjacent area and the upper right adjacent area, W'=min{W+H,W+W}.

Refer to FIG. 1, which shows a schematic diagram of the distribution of effective adjacent areas provided by related technical solutions. In FIG. 1, the left side adjacent area, the lower left side adjacent area, the upper side adjacent area, and the upper right side adjacent area are all valid. On the basis of Figure 1, the selection areas for the three modes are shown in Figure 2. Among them, in Figure 2, (a) shows the selection area of INTRA_LT_CCLM mode, including the left adjacent area and upper side adjacent area; (b) shows the selection area of INTRA_L_CCLM mode, including the left adjacent area And the adjacent area on the lower left side; (c) shows the selection area of INTRA_T_CCLM mode, including the adjacent area on the upper side and the adjacent area on the upper right side. In this way, after the selection areas of the three modes are determined, reference points for deriving model parameters can be selected in the selection area. The reference points selected in this way can be called adjacent reference pixels, and usually the number of adjacent reference pixels is at most 4; and for a W×H code block with a certain size, the positions of the adjacent reference pixels Generally certain.

After obtaining the adjacent reference pixels, according to the number of effective pixels, the model parameters are currently determined according to the schematic flow diagram of the traditional solution for deriving the model parameters shown in FIG. 3. According to the process shown in FIG. 3, assuming that the image component to be predicted is a chrominance component, the process may include:

S301: Obtain adjacent reference pixels from the selected area;

S302: Determine the number of effective pixel points in adjacent reference pixels;

S303: Set the first model parameter to 0 and the second model parameter to the default value;

S304: Fill the chromaticity prediction value as the default value;

S305: Obtain four effective pixels through "copy";

S306: Obtain two pixels with a larger value and two pixels with a smaller brightness component after 4 comparisons;

S307: Calculate two mean points;

S308: Use two mean points as fitting points to derive the first model parameter and the second model parameter;

S309: Perform prediction processing of the chrominance component according to the constructed prediction model.

It should be noted that when the number of effective pixels is 0, steps S303 and S304 will be executed; when the number of effective pixels is 2, step S305 will be executed, and then steps S306-S309 will be executed sequentially; When the number of points is 4, steps S306-S309 will be executed sequentially.

It should also be noted that when the number of effective pixels is 0, the predicted values Pred _C [i,j] corresponding to the chrominance components of all pixels in the current coding block can be filled with the default values of the chrominance components; In the embodiment of the present application, the default value is the middle value of the chrominance component. Assuming that the bit depth of the chrominance component is represented by BitDepthC, the intermediate value can be calculated as 1<<(BitDepthC-1). Exemplarily, assuming that the current video image is an 8-bit video, the component range corresponding to the chrominance component is 0-255, and the intermediate value is 128 at this time. At this time, the preset component value can be 128; assuming that the current video image is 10 bits For video, the component range corresponding to the chrominance component is 0-1023, and the intermediate value is 512 at this time. At this time, the preset component value can be 512.

Understandably, the principle of "two points determine a straight line" is used to construct a prediction model; the two points here can be called fitting points. In the current traditional scheme, after four effective pixels are obtained, two adjacent reference pixels with a larger value and two adjacent reference pixels with a smaller value are obtained after four comparisons; then according to the brightness Two adjacent reference pixels with larger components, find a mean point (which can be represented by mean _max ), and find another mean point based on two adjacent reference pixels with a smaller brightness component (you can use mean _min ) to obtain two mean points mean _max and mean _min ; then mean _max and mean _{min are} used as two fitting points to derive the model parameters (which can be expressed by α and β), and finally build according to the model parameters Predict model, and perform chroma component prediction processing based on the predict model.

In other words, when deriving the model parameters, taking 4 adjacent reference pixels as an example, currently the average value of 2 larger adjacent reference pixels and 2 smaller adjacent reference pixels will be used The obtained mean values are used as two fitting points to derive model parameters. When the four adjacent reference pixels are more evenly distributed, the schematic diagram of the prediction model under the traditional scheme shown in Figure 4A; where the abscissa of the coordinate axis represents the brightness value (which can be represented by Luma), and the vertical of the coordinate axis The coordinates represent the chromaticity value (it can be represented by chroma), the 4 black dots are the 4 adjacent reference pixels, and the 2 gray dots are the 2 larger adjacent reference pixels of the 4 adjacent reference pixels. The mean point corresponding to the pixel point and the mean point corresponding to the two smaller adjacent reference pixels (that is, two fitting points), the gray diagonal line indicates the prediction model constructed according to the two fitting points; the gray dotted line The prediction model fitted by the Least Mean Square (LMS) algorithm for these 4 adjacent reference pixels; it can be seen from Figure 4A that the gray diagonal line is relatively close to the gray dotted line, that is, the traditional scheme The constructed prediction model can more accurately fit the distribution of these four adjacent reference pixels.

However, when the four adjacent reference pixels are non-uniformly distributed, the prediction model constructed by the traditional scheme cannot accurately fit the distribution of the four adjacent reference pixels, as shown in Figure 4B. A schematic diagram of the prediction model under a traditional scheme; as can be seen from Figure 4B, the gray slanted line and the gray dotted line have a certain deviation, that is, the prediction model constructed by the traditional scheme cannot be compared with the distribution of these 4 adjacent reference pixels. Fits well. In other words, the current traditional scheme lacks robustness in the process of model parameter derivation, and cannot fit well when the four adjacent reference pixels are in an uneven distribution.

In order to improve the robustness of CCLM prediction and at the same time improve the coding and decoding performance, an embodiment of the present application provides an image component prediction method, which obtains N adjacent reference pixels for the image components to be predicted of the coding block in the video image. Here The N adjacent reference pixels are reference pixels adjacent to the encoding block, and N is a preset integer value; the N first image component values corresponding to the N adjacent reference pixels are compared to determine A first reference pixel subset and a second reference pixel subset, the first reference pixel subset includes: among the preset number of first image component values, the smallest first image component value and the second smallest first image component value correspond to The second reference pixel subset includes two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value among the preset number of first image component values Reference pixel; calculate the average value of the N adjacent reference pixels to obtain the first average point; perform the first image component value on the first reference pixel subset and/or the second reference pixel subset through the first average point Determine the two fitting points; determine the model parameters based on the two fitting points, and obtain the prediction model corresponding to the image component to be predicted according to the model parameters to obtain the predicted value corresponding to the image component to be predicted; in this way, after the second comparison After processing, the preset number of adjacent reference pixels can be divided into multiple situations, and two different fitting points are used to construct the prediction model for these multiple situations, which can improve the robustness of CCLM prediction; that is, By optimizing the fitting points used in the model parameter derivation, the constructed prediction model is more accurate, and the coding and decoding prediction performance of the video image is improved.

The embodiments of the present application will be described in detail below in conjunction with the drawings.

Refer to FIG. 5, which shows an example of a block diagram of a video encoding system provided by an embodiment of the present application; as shown in FIG. 7, the video encoding system 500 includes a transform and quantization unit 501, an intra-frame estimation unit 502, and an intra-frame The prediction unit 503, the motion compensation unit 504, the motion estimation unit 505, the inverse transform and inverse quantization unit 506, the filter control analysis unit 507, the filtering unit 508, the encoding unit 509, and the decoded image buffer unit 510, etc., wherein the filtering unit 508 may Implementing deblocking filtering and Sample Adaptive Offset (SAO) filtering, the encoding unit 509 can implement header information encoding and context-based adaptive binary arithmetic coding (Context-based Adaptive Binary Arithmatic Coding, CABAC). For the input original video signal, a video coding block can be obtained by dividing the coding tree block (Coding Tree Unit, CTU), and then the residual pixel information obtained after intra or inter prediction is paired by the transform and quantization unit 501 The video coding block is transformed, including transforming the residual information from the pixel domain to the transform domain, and quantizing the resulting transform coefficients to further reduce the bit rate; the intra-frame estimation unit 502 and the intra-frame prediction unit 503 are used for Perform intra prediction on the video encoding block; specifically, the intra estimation unit 502 and the intra prediction unit 503 are used to determine the intra prediction mode to be used to encode the video encoding block; the motion compensation unit 504 and the motion estimation unit 505 is used to perform inter-frame predictive coding of the received video coding block relative to one or more blocks in one or more reference frames to provide temporal prediction information; the motion estimation performed by the motion estimation unit 505 is for generating motion vectors In the process, the motion vector can estimate the motion of the video coding block, and then the motion compensation unit 504 performs motion compensation based on the motion vector determined by the motion estimation unit 505; after determining the intra prediction mode, the intra prediction unit 503 also It is used to provide the selected intra prediction data to the encoding unit 509, and the motion estimation unit 505 also sends the calculated motion vector data to the encoding unit 509; in addition, the inverse transform and inverse quantization unit 506 is used for the video The reconstruction of the coding block, the reconstruction of the residual block in the pixel domain, the reconstruction of the residual block through the filter control analysis unit 507 and the filtering unit 508 to remove the blocking artifacts, and then the reconstructed residual block is added to the decoding A predictive block in the frame of the image buffer unit 510 is used to generate reconstructed video coding blocks; the coding unit 509 is used to encode various coding parameters and quantized transform coefficients. In the coding algorithm based on CABAC, The context content can be based on adjacent coding blocks, can be used to encode information indicating the determined intra prediction mode, and output the code stream of the video signal; and the decoded image buffer unit 510 is used to store reconstructed video coding blocks for Forecast reference. As the encoding of the video image progresses, new reconstructed video encoding blocks will be continuously generated, and these reconstructed video encoding blocks will be stored in the decoded image buffer unit 510.

Refer to FIG. 6, which shows an example of a block diagram of a video decoding system provided by an embodiment of the present application; as shown in FIG. 6, the video decoding system 600 includes a decoding unit 601, an inverse transform and inverse quantization unit 602, and an intra-frame The prediction unit 603, the motion compensation unit 604, the filtering unit 605, and the decoded image buffer unit 606, etc., wherein the decoding unit 601 can implement header information decoding and CABAC decoding, and the filtering unit 605 can implement deblocking filtering and SAO filtering. After the input video signal is subjected to the encoding process of FIG. 7, the code stream of the video signal is output; the code stream is input into the video decoding system 600, and first passes through the decoding unit 601 to obtain the decoded transform coefficient; The inverse transform and inverse quantization unit 602 performs processing to generate residual blocks in the pixel domain; the intra prediction unit 603 can be used to generate data based on the determined intra prediction mode and the data from the previous decoded block of the current frame or picture The prediction data of the current video decoding block; the motion compensation unit 604 determines the prediction information for the video decoding block by analyzing the motion vector and other associated syntax elements, and uses the prediction information to generate the predictability of the video decoding block being decoded Block; by summing the residual block from the inverse transform and inverse quantization unit 602 and the corresponding predictive block generated by the intra prediction unit 603 or the motion compensation unit 604 to form a decoded video block; the decoded video signal Through the filtering unit 605 in order to remove the block effect artifacts, the video quality can be improved; then the decoded video blocks are stored in the decoded image buffer unit 606, and the decoded image buffer unit 606 stores reference images for subsequent intra prediction or motion compensation , It is also used for the output of the video signal, that is, the restored original video signal is obtained.

The image component prediction method in the embodiment of this application is mainly applied to the part of the intra prediction unit 503 shown in FIG. 5 and the part of the intra prediction unit 603 shown in FIG. 6, and is specifically applied to CCLM prediction in intra prediction section. That is to say, the image component prediction method in the embodiment of this application can be applied to a video encoding system, a video decoding system, or even a video encoding system and a video decoding system at the same time. However, the embodiment of this application There is no specific limitation. When the method is applied to the intra prediction unit 503 part, the "coding block in the video image" specifically refers to the current coding block in the intra prediction; when the method is applied to the intra prediction unit 603 part, "coding in the video image "Block" specifically refers to the current decoded block in intra prediction.

Based on the application scenario example of FIG. 5 or FIG. 6, refer to FIG. 7, which shows a schematic flowchart of an image component prediction method provided by an embodiment of the present application. As shown in Figure 7, the method may include:

S701: Acquire N adjacent reference pixels corresponding to the image component to be predicted of the coding block in the video image;

It should be noted that the N adjacent reference pixels are reference pixels adjacent to the coding block, and N is a preset integer value, which may also be referred to as a preset number. The video image can be divided into multiple coding blocks, and each coding block can include a first image component, a second image component, and a third image component. The coding block in this embodiment of the application is a video image to be encoded. The current block. When the first image component needs to be predicted by the prediction model, the image component to be predicted is the first image component; when the second image component needs to be predicted by the prediction model, the image component to be predicted is the second image component; When the third image component is predicted by the prediction model, the image component to be predicted is the third image component.

In this way, a preset number of adjacent reference pixels can be obtained from the reference pixels adjacent to the coding block to determine the fitting points for deriving the model parameters. In addition, the value of N may generally be 4, but the embodiment of the present application does not specifically limit it.

In some embodiments, for S701, the obtaining N adjacent reference pixels corresponding to the image component to be predicted of the coding block in the video image may include:

S701-1: Obtain a first reference pixel set corresponding to the image component to be predicted of the coding block in the video image;

It should be noted that when the left adjacent area, the lower left adjacent area, the upper adjacent area, and the upper right adjacent area are all valid areas, for the INTRA_LT_CCLM mode, the first reference pixel set is determined by the left side of the coding block. The adjacent reference pixels in the side adjacent area and the upper adjacent area are composed of adjacent reference pixels, as shown in Figure 2 (a); for the INTRA_L_CCLM mode, the first reference pixel set is composed of the left adjacent area and It is composed of adjacent reference pixels in the adjacent area on the lower left side, as shown in Figure 2 (b); for the INTRA_T_CCLM mode, the first reference pixel set is composed of the adjacent area on the upper side of the coding block and the adjacent area on the upper right side. It is composed of adjacent reference pixels in Figure 2 (c).

In some embodiments, optionally, for S701-1, the acquiring the first reference pixel set corresponding to the image component to be predicted of the coding block in the video image may include:

Acquiring reference pixels adjacent to at least one side of the coding block; wherein the at least one side includes the left side of the coding block and/or the upper side of the coding block;

Based on the reference pixel points, a first reference pixel set corresponding to the image component to be predicted is formed.

It should be noted that at least one side of the coding block can refer to the upper side of the coding block, or the left side of the coding block, or even the upper and left sides of the coding block. The embodiments are not specifically limited.

In this way, for the INTRA_LT_CCLM mode, when the adjacent area on the left and the adjacent area on the upper side are all valid areas, the first reference pixel set at this time can be the sum of the reference pixels adjacent to the left side of the encoding block and the encoding The upper side of the block is composed of adjacent reference pixels. When the adjacent area on the left is the effective area and the adjacent area on the upper side is the ineffective area, then the first reference pixel set can be composed of the adjacent pixels on the left side of the encoding block. It is composed of reference pixels adjacent to the side; when the adjacent area on the left is an invalid area and the adjacent area on the upper side is an effective area, then the first reference pixel set can be composed of the upper side of the coding block. It is composed of adjacent reference pixels.

Obtain a reference row or reference pixel point in a reference column adjacent to the encoding block; wherein the reference row is composed of the upper side of the encoding block and the row adjacent to the upper right side, the The reference column is composed of columns adjacent to the left side and the lower left side of the coding block;

It should be noted that the reference row or reference column adjacent to the encoding block can refer to the reference row adjacent to the upper side of the encoding block, or the reference column adjacent to the left side of the encoding block, or even It refers to a reference row or a reference column adjacent to other sides of the coding block, which is not specifically limited in the embodiment of the present application. For the convenience of description, in the embodiments of the present application, the adjacent reference rows of the coding blocks will describe the reference behaviors with adjacent sides above, and the reference columns adjacent to the coding blocks will take the reference columns adjacent to the left as an example. description.

Wherein, the reference pixels in the reference row adjacent to the coding block may include reference pixels adjacent to the upper side and the upper right side (also referred to as adjacent reference pixels corresponding to the upper side and the upper right side) Point), where the upper side represents the upper side of the coding block, and the upper right side represents the side length of the upper side of the coding block that is horizontally extended to the right and the same height as the current coding block; in the reference column adjacent to the coding block The reference pixels of may also include reference pixels adjacent to the left side and the lower left side (also referred to as the adjacent reference pixels corresponding to the left side and the lower left side), where the left side represents the code The left side of the block and the lower left side represent the side length that is the same width as the current decoded block, which is vertically extended downward from the left side of the encoding block; however, the embodiment of the present application does not specifically limit it.

In this way, for the INTRA_L_CCLM mode, when the left adjacent area and the lower left adjacent area are valid areas, the first reference pixel set at this time may be composed of reference pixels in the reference column adjacent to the coding block; In the INTRA_T_CCLM mode, when the upper adjacent area and the upper right adjacent area are valid areas, the first reference pixel set at this time may be composed of reference pixels in the reference row adjacent to the coding block.

S701-2: Perform screening processing on the first reference pixel set to obtain a second reference pixel set; wherein, the second reference pixel set includes N adjacent reference pixels;

It should be noted that in the first parameter pixel set, there may be some unimportant reference pixels (for example, these reference pixels have poor correlation) or some abnormal reference pixels, in order to ensure the accuracy of the prediction model , These reference pixels need to be removed to obtain the second reference pixel set; wherein, the number of effective pixels contained in the second reference pixel set is usually selected as 4 in practical applications, but the embodiment of the application There is no specific limitation.

In some embodiments, for S701-2, the filtering process on the first reference pixel set to obtain the second reference pixel set may include:

Determine the position of the pixel to be selected based on the pixel position and/or image component intensity corresponding to each adjacent reference pixel in the first reference pixel set;

According to the determined position of the pixel to be selected, adjacent reference pixels corresponding to the position of the pixel to be selected are selected from the first reference pixel set, and the selected adjacent reference pixels form a second reference pixel set ; Wherein, the second reference pixel set includes N adjacent reference pixels.

It should be noted that the image component intensity can be represented by image component values, such as brightness value, chroma value, etc.; here, the larger the image component value, the higher the image component intensity. In this way, the screening of the first reference pixel set can be based on the position of the reference pixel to be selected, or based on the intensity of the image component (such as luminance value, chrominance value, etc.), so that The filtered reference pixels to be selected form a second reference pixel set. The following will describe the position of the reference pixel to be selected as an example.

Exemplarily, assuming that the positions of the reference pixels in the upper selection area W'are S[0,-1],...,S[W'-1,-1], the positions of the reference pixels in the left selection area H' The positions are S[-1,0],..., S[-1,H'-1]; in this way, the selection method of selecting at most 4 adjacent reference pixels is as follows:

For the INTRA_LT_CCLM mode, when the upper adjacent area and the left adjacent area are both valid, at this time, 2 adjacent reference pixels to be selected can be filtered out in the upper selection area W', and their corresponding positions are S[ W'/4,-1] and S[3W'/4,-1]; 2 adjacent reference pixels to be selected can be screened out in the left selection area H', and their corresponding positions are respectively S[-1, H'/4] and S[-1,3H'/4]; these 4 adjacent reference pixels to be selected form a second reference pixel set, as shown in FIG. 8A. In Fig. 8A, the adjacent area on the left side and the adjacent area on the upper side of the coding block are both effective, and in order to maintain the same resolution of the luminance component and the chrominance component, the luminance component needs to be down-sampled, so that The down-sampled luminance component and chrominance component have the same resolution.

For the INTRA_L_CCLM mode, when only the adjacent area on the left and the adjacent area on the lower left are valid, 4 adjacent reference pixels to be selected can be filtered out in the left selection area H', and their corresponding positions are S[ -1,H'/8], S[-1,3H'/8], S[-1,5H'/8] and S[-1,7H'/8]; put these 4 to be selected adjacent The reference pixels constitute a second reference pixel set, as shown in FIG. 8B. In Fig. 8B, the adjacent area on the left side and the adjacent area on the lower left side of the coding block are both effective, and in order to maintain the same resolution of the luminance component and the chrominance component, the luminance component still needs to be down-sampled, so The down-sampled luminance component and chrominance component have the same resolution.

For the INTRA_T_CCLM mode, when only the upper adjacent area and the upper right adjacent area are valid, 4 adjacent reference pixels to be selected can be filtered out in the upper selection area W', and their corresponding positions are S[ W'/8,-1], S[3W'/8,-1], S[5W'/8,-1] and S[7W'/8,-1]; put these 4 to be selected adjacent The reference pixels constitute a second reference pixel set, as shown in FIG. 8C. In Fig. 8C, the upper adjacent area and the upper right adjacent area of the coding block are both effective, and in order to maintain the same resolution of the luminance component and the chrominance component, the luminance component still needs to be down-sampled, so that The down-sampled luminance component and chrominance component have the same resolution.

In this way, by filtering the first reference pixel set, the second reference pixel set can be obtained, and the second reference pixel set generally includes a preset number of adjacent reference pixels, such as 4 adjacent reference pixels; The preset number of adjacent reference pixels are grouped by comparison, so that the two fitting points determined subsequently are more accurate.

S702: Compare the N first image component values corresponding to the N adjacent reference pixels, and determine a first reference pixel subset and a second reference pixel subset;

It should be noted that each adjacent reference pixel corresponds to the first image component value and the second image component value; in this way, the preset number of adjacent reference pixels corresponds to the preset number of first image component values, and the After a comparison process, the first reference pixel subset and the second reference pixel subset can be determined according to the value of the first image component.

In some embodiments, for S702, the comparing the N first image component values corresponding to the N adjacent reference pixels to determine the first reference pixel subset and the second reference pixel subset may include :

Obtaining a preset number of first image component values based on the preset number of adjacent reference pixels;

The preset number of first image component values are compared multiple times to obtain a first array composed of the smallest first image component value and the second smallest first image component value, and the largest first image component value and the second largest first image component value. A second array of image component values;

Putting two adjacent reference pixels corresponding to the first array into a first reference pixel subset to obtain the first reference pixel subset;

Put two adjacent reference pixels corresponding to the second array into a second reference pixel subset to obtain the second reference pixel subset.

It should be noted that after obtaining the preset number of first image component values, multiple comparisons are performed on the preset number of first image component values (for example, four adjacent reference pixels need to be compared four times) , Obtain the first reference pixel subset and the second reference pixel subset; wherein the first reference pixel subset includes the smallest first image component value and the second smallest first image component value among the preset number of first image component values Corresponding two adjacent reference pixels, the second reference pixel subset includes two adjacent ones corresponding to the largest first image component value and the second largest first image component value among the preset number of first image component values Reference pixels.

Specifically, it can be that the preset number of adjacent reference pixels are numbered first, and then the preset number of adjacent reference pixels are initially grouped, and the first image corresponding to the i-th number is stored in the first array Component value, the first image component value corresponding to the i+1 number is stored in the second array, i is a positive integer greater than or equal to 0; then the first image component value in the first array and the first image component value in the second array An image component value is compared for multiple times, so that the smallest first image component value and the second smallest first image component value are stored in the first array, and the largest first image component value and the second largest first image component value are stored in the second array. Thus, the first reference pixel subset and the second reference pixel subset are obtained.

Further, after the first reference pixel subset and the second reference pixel subset are obtained, the second average point of the first reference pixel subset and the third average point of the second reference pixel subset can also be obtained. Therefore, in some embodiments, refer to FIG. 9, which shows a schematic flowchart of another image component prediction method provided by an embodiment of the present application. As shown in FIG. 9, after S702, the method may further include:

S901: Based on the first image component value and the second image component value corresponding to each adjacent reference pixel point in the first reference pixel subset, obtain the second mean value corresponding to the first image component and the first image component corresponding to the second image component. Two averages, get the second average point;

S902: Based on the first image component value and the second image component value corresponding to each adjacent reference pixel point in the second reference pixel subset, obtain a third mean value corresponding to the first image component and a first image component corresponding to the second image component. Three averages, get the third average point.

That is to say, the average value of the first image component value corresponding to each adjacent reference pixel in the first reference pixel subset is calculated to obtain the average value of the first image component corresponding to the multiple first image components, which can be called the first image The second mean value of the component (which can be represented by luma _min ); the mean value of the second image component value corresponding to each adjacent reference pixel in the first reference pixel subset is calculated to obtain the second image corresponding to the multiple second image components The component average value can be called the second average value of the second image component (it can be represented by chroma _min ). Here, the second average value point can be represented by mean _min ; here, the first image component of the second average value point is luma _min , The second image component of the two-mean point is chroma _min .

In the same way, the average value of the first image component value corresponding to each adjacent reference pixel in the second reference pixel subset is calculated to obtain the average value of the first image component corresponding to the multiple first image components, which can be called the first image component The third mean value of (can be represented by luma _max ); the mean value of the second image component values corresponding to each adjacent reference pixel in the second reference pixel subset is calculated to obtain the second image components corresponding to multiple second image components The mean value can be called the third mean value of the second image component (it can be expressed by chroma _max ). Here, the third mean point can be expressed by mean _max ; here, the first image component of the third mean point is luma _max and the third The second image component of the mean point is chroma _max .

S703: Calculate the average value of the N adjacent reference pixel points to obtain a first average value point;

It should be noted that the first average point is obtained by averaging the N adjacent reference pixel points. Therefore, in some embodiments, as shown in FIG. 9, for S703, this step may include:

S903: Obtain the first average value and the second image corresponding to the first image component based on the first image component value and the second image component value corresponding to each adjacent reference pixel in the preset number of adjacent reference pixels The first mean value corresponding to the component, obtain the first mean value point;

In other words, the average value of the first image component corresponding to each of the N adjacent reference pixels is calculated to obtain the average value of the first image component corresponding to the N first image components, which can be called the first image component value. The first average value of an image component (which can be represented by meanL); the average value of the second image component value corresponding to each adjacent reference pixel in the preset number of adjacent reference pixels is calculated to obtain N first image components The corresponding mean value of the first image component can be referred to as the first mean value of the first image component (which can be represented by meanC). Here, the first mean point can be expressed by mean; here, the first image component of the first mean point is meanL , The second image component of the first mean point is meanC. It should be noted that the calculation of the first mean point can also be obtained by averaging the second mean point and the third mean point. For example, for meanL, it can be obtained by averaging luma _min and luma _max ; For meanC, it can be obtained by averaging chroma _min and chroma _max .

S704: Perform a comparison of first image component values on the first reference pixel subset and/or the second reference pixel subset through the first average point, and determine two fitting points;

It should be noted that the two fitting points include the first fitting point and the second fitting point. The second comparison process is performed on the first reference pixel subset and/or the second reference pixel subset through the first average point. There can be three situations. For example, the first image component corresponding to the first average point is smaller than the first reference pixel subset. The second smallest first image component value in the set, or the first image component corresponding to the first average point is greater than the second largest first image component value in the second reference pixel subset, or the first image component corresponding to the first average point is less than or equal to the first image component The second largest first image component value in the second reference pixel subset. For these three cases, the determined first fitting point and the second fitting point are different, so that the robustness of CCLM prediction can be improved.

Further, in some embodiments, before S704, the method may further include:

Comparing the first image component values corresponding to two adjacent reference pixels in the first reference pixel subset;

According to the result of the comparison, determine the smallest first image component value and the second smallest first image component value of the two first image component values, and use the pixel corresponding to the smallest first image component value as the first adjacent reference pixel , The pixel corresponding to the second smallest first image component value is used as the second adjacent reference pixel.

Further, in some embodiments, the method may further include:

Comparing the first image component values corresponding to two adjacent reference pixels in the second reference pixel subset;

According to the result of the comparison, determine the second largest first image component value and the largest first image component value among the two first image component values, and use the pixel corresponding to the second largest first image component value as the third adjacent reference pixel Point, the pixel point corresponding to the largest first image component value is used as the fourth adjacent reference pixel point.

In this way, by comparing the first image component values corresponding to two adjacent reference pixels in the first reference pixel subset, the smallest first image component value and the second smallest first image component value can be determined, and the smallest first image component value can be determined. The pixel corresponding to an image component value is taken as the first adjacent reference pixel, and the pixel corresponding to the second smallest first image component value is taken as the second adjacent reference pixel; by comparing two adjacent reference pixels in the second reference pixel subset The first image component value corresponding to each pixel is compared, the largest first image component value and the second largest first image component value can be determined, and the pixel corresponding to the second largest first image component value is used as the third neighbor reference Pixel, the pixel corresponding to the largest first image component value is used as the fourth adjacent reference pixel.

Further, in some embodiments, as shown in FIG. 9, for S704, the first image component is performed on the first reference pixel subset and/or the second reference pixel subset through the first mean point. Comparison of values to determine two fitting points can include:

S904: Compare the first image component of the first average point with the second smallest first image component value in the first reference pixel subset;

S905: When the first image component of the first average point is greater than or equal to the second smallest first image component value in the first reference pixel subset, combine the first image component of the first average point with the second reference pixel Concentrate the second largest first image component value for comparison;

S906: When the first image component of the first average point is less than or equal to the second largest first image component value in the second reference pixel subset, use the second average point as the first fitting point and the third average point as the The second fitting point; wherein the two fitting points include a first fitting point and a second fitting point.

Further, after S904, the method may further include:

S907: When the first image component of the first average point is smaller than the second smallest first image component value, the first adjacent reference pixel point is used as the first fitting point, and the third average point is used as the second fitting point.

Further, after S905, the method may further include:

S908: When the first image component of the first average point is greater than the second largest first image component value, the second average point is used as the first fitting point, and the fourth adjacent reference pixel point is used as the second fitting point.

It should be noted that for steps S904 and S905, the first image component of the first mean point may be compared with the second smallest first image component value in the first reference pixel subset, and then the value of the first mean point The first image component is compared with the second largest first image component value in the second reference pixel subset; it can also be the first image component of the first mean point with the second largest first image component in the second reference pixel subset. The image component values are compared, and then the first image component of the first mean point is compared with the second smallest first image component value in the first reference pixel subset; the order of comparison is not specifically limited in this embodiment of the application.

It should also be noted that the first image component of the first average point is the first average value corresponding to the first image component. For step S904, if the first image component of the first mean point is greater than or equal to the second smallest first image component value in the first reference pixel subset, then step S905 will be executed; otherwise, step S907 will be executed. One fitting point is the first adjacent reference pixel point, and the second fitting point is the third mean point; for step S905, if the first image component of the first mean point is less than or equal to the second reference pixel subset The largest first image component value, then step S906 will be executed. At this time, the first fitting point is the second mean point, and the second fitting point is the third mean point; otherwise, step S908 will be executed. The fitting point is the second mean point, and the second fitting point is the fourth adjacent reference pixel point. It can be seen that after the second comparison process, the preset number of adjacent reference pixels can be divided into three cases. For these three cases, two different fitting points are used to construct the prediction model, which can improve the CCLM prediction. Robustness.

It should also be noted that there is no restriction on the judgment of "equal to". For example, when the first image component of the first average point is greater than the second smallest first image component value in the first reference pixel subset, step S905 is executed, and when the first image component of the first average point is less than or equal to the first image component. When the second smallest first image component value in the reference pixel subset, step S907 is executed; or when the first image component of the first mean point is less than the second largest first image component value in the second reference pixel subset, step S906 is executed, and when the first When the first image component of an average point is greater than or equal to the second largest first image component value in the second reference pixel subset, step S908 is executed.

S705: Determine model parameters based on the two fitting points, and obtain a prediction model corresponding to the image component to be predicted according to the model parameters; wherein, the prediction model is used to predict the image component to be predicted Processing to obtain the predicted value corresponding to the image component to be predicted;

It should be noted that after the first fitting point and the second fitting point are obtained, the model parameters can be determined according to the first fitting point and the second fitting point; here, the model parameters include the first model parameters (can Denoted by α) and the second model parameter (denoted by β). Assuming that the image component to be predicted is a chrominance component, according to the model parameters α and β, the prediction model corresponding to the chrominance component shown in formula (1) can be obtained.

In some embodiments, for S705, the model parameters include a first model parameter and a second model parameter, and the determining the model parameter based on the two fitting points may include:

S705-1: Based on the first fitting point and the second fitting point, obtain the first model parameter through a first preset factor calculation model;

S705-2: Based on the first model parameter and the first fitting point, obtain the second model parameter through a second preset factor calculation model;

Further, in some embodiments, after S705-1, the method may further include:

S705-3: Based on the first model parameter and the first average point, obtain the second model parameter through a second preset factor calculation model.

It should be noted that after the first fitting point (luma ₁ , chroma ₁ ) and the second fitting point (luma ₂ , chroma ₂ ) are obtained, the first model parameter can be calculated according to the first preset factor calculation model α, as shown in formula (2),

After the first model parameter α is obtained, the second model parameter β can be calculated according to the calculation model combining the first fitting point (luma ₁ , chroma ₁ ) and the second preset factor, as shown in formula (3),

β=chroma ₁ -α×luma ₁ (3)

The second model parameter β is calculated according to the combination of the first mean point (meanL, meanC) and the second preset factor calculation model, as shown in formula (4),

β=meanC-α×meanL (4)

In this way, after the first model parameter α and the first model parameter β are obtained, a preset model can be constructed. Assuming that the image component to be predicted is a chrominance component, the prediction model corresponding to the chrominance component can be obtained according to the model parameters (α and β), as shown in equation (1); then the prediction model is used to predict the chrominance component, To get the predicted value corresponding to the chrominance component.

Further, in some embodiments, after S705, the method may further include:

Performing prediction processing on the image component to be predicted for each pixel in the coding block based on the prediction model to obtain a predicted value corresponding to the image component to be predicted for each pixel.

It should be noted that after the second comparison process is performed on the first reference pixel subset and/or the second reference pixel subset through the first mean point, two fitting points can be determined; in this way, one can be determined according to the "two points The principle of "straight line" can determine the slope of the line (ie the first model parameter) and the intercept of the line (ie the second model parameter), so that the prediction model corresponding to the image component to be predicted can be obtained based on these two model parameters , In order to obtain the predicted value corresponding to the image component to be predicted for each pixel in the coding block. For example, assuming that the image component to be predicted is a chrominance component, according to the first model parameter α and the second model parameter β, the prediction model corresponding to the chrominance component shown in equation (1) can be obtained; then, the equation (1) The prediction model shown in) performs prediction processing on the chrominance component of each pixel in the coding block, so that the predicted value corresponding to the chrominance component of each pixel can be obtained.

This embodiment provides a method for predicting image components. For the image components to be predicted of the coding block in the video image, and for the image components to be predicted of the coding block in the video image, N adjacent reference pixels are obtained, where N The adjacent reference pixels are reference pixels adjacent to the encoding block, and N is a preset integer value; then, the N first image component values corresponding to the N adjacent reference pixels are compared to determine the first The reference pixel subset and the second reference pixel subset, the first reference pixel subset includes: among the preset number of first image component values, two corresponding to the smallest first image component value and the second smallest first image component value Adjacent reference pixels, the second reference pixel subset includes: two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value among the preset number of first image component values Point; Calculate the average value of the N adjacent reference pixel points to obtain the first average value point; compare the first image component value of the first reference pixel subset and/or the second reference pixel subset through the first average value point , Determine two fitting points; determine model parameters based on the two fitting points, and obtain the prediction model corresponding to the image component to be predicted according to the model parameters to obtain the predicted value corresponding to the image component to be predicted; in this way, after the second comparison process , The preset number of adjacent reference pixels can be divided into multiple situations, and two different fitting points are used to construct the prediction model for these multiple situations, thereby improving the robustness of CCLM prediction; that is, through Optimizing the fitting points used in the model parameter derivation can make the constructed prediction model more accurate and improve the coding and decoding prediction performance of the video image.

In another embodiment of the present application, in practical applications, the number of adjacent reference pixels used for derivation of the model parameters is generally 4; that is, the preset number may be 4. A detailed description will be given below taking the preset number equal to 4 as an example.

As shown in the process shown in Figure 3, assuming that the image component to be predicted is a chrominance component, when 0 adjacent reference pixels are obtained, the first model parameter α can be directly set to 0, and the second model parameter β is set to the default Value (it can also be the median value of chrominance component); when two adjacent reference pixels are obtained, the two adjacent reference pixels can be copied to obtain 4 adjacent reference pixels, and then according to The operation of 4 adjacent reference pixels is used to construct the prediction model. In other words, after obtaining 4 adjacent reference pixels, two fitting points can be determined by 4 comparisons and average points, so as to use the principle of "two points to determine a straight line" to derive model parameters ; In this way, a prediction model corresponding to the image component to be predicted can be constructed according to the model parameters to obtain the predicted value corresponding to the image component to be predicted.

Specifically, it is assumed that the image component to be predicted is a chrominance component, and the chrominance component is predicted by the luminance component. It is assumed that the numbers of the obtained 4 adjacent reference pixels are 0, 1, 2, and 3 respectively. By comparing these four adjacent reference pixels, based on four comparisons, two adjacent reference pixels with larger brightness values can be further selected (including the pixel with the largest brightness value and the pixel with the next largest brightness value). Pixels) and 2 adjacent reference pixels with smaller brightness values (may include the pixel with the smallest brightness value and the pixel with the next smallest brightness value). Further, two arrays of minIdx[2] and maxIdx[2] can be set to store two sets of adjacent reference pixels respectively. Initially, the adjacent reference pixels numbered 0 and 2 are put into minIdx[2], and The adjacent reference pixels numbered 1 and 3 are put into maxIdx[2], as shown below,

Init:minIdx[2]={0,2},maxIdx[2]={1,3}

After that, through four comparisons, the two adjacent reference pixels with smaller brightness values can be stored in minIdx[2], and the two adjacent reference pixels with larger brightness values are stored in maxIdx[2] Pixels, as shown below,

Step1: if(L[minIdx[0]]>L[minIdx[1]],swap(minIdx[0],minIdx[1])

Step2: if(L[maxIdx[0]]>L[maxIdx[1]],swap(maxIdx[0],maxIdx[1])

Step3: if(L[minIdx[0]]>L[maxIdx[1]],swap(minIdx,maxIdx)

Step4: if(L[minIdx[1]]>L[maxIdx[0]],swap(minIdx[1],maxIdx[0])

In this way, two adjacent reference pixels with smaller brightness values can be obtained, and their corresponding brightness values are represented by luma ⁰ _min and luma ¹ _min respectively, and the corresponding chromaticity values are represented by chroma ⁰ _min and chroma ¹ _{min respectively} ; at the same time Two adjacent reference pixels with larger brightness values can also be obtained, and the corresponding brightness values are represented by luma ⁰ _max and luma ¹ _max respectively, and the corresponding chromaticity values are represented by chroma ⁰ _max and chroma ¹ _{max respectively} . Further, for the two smaller averaging the neighboring reference pixels can be obtained is represented by the mean point _min Mean, Mean _min the mean point corresponding to the luminance value LUMA _min, _min Chroma color value; for 2 The average value of a larger adjacent reference pixel point can be obtained, and the second average value point can be represented by mean _max . The luminance value of this mean value point mean _max is luma _max and the chroma value is chroma _max ; the details are as follows,

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

In other words, after obtaining two mean points mean _min (luma _min , chroma _min ) and mean _max (luma _max , chroma _max ), these two mean points can be used as two fitting points at this time, and the model parameters can be These two fitting points are obtained through the calculation method of "two points determine a straight line". Specifically, the model parameters α and β can be calculated by formula (5),

Among them, the model parameter α is the slope in the prediction model, and the model parameter β is the intercept in the prediction model. In this way, after the model parameters are derived, the prediction model corresponding to the chrominance component can be obtained according to the model parameters, as shown in equation (1); then the prediction model is used to predict the chrominance component to obtain the corresponding chrominance component Predictive value.

In this way, because the current traditional scheme uniformly uses the average point of the larger two adjacent reference pixels and the average point of the smaller two adjacent reference pixels among the four adjacent reference pixels as two fitting points To build a prediction model, the prediction model lacks robustness, so when the distribution of 4 adjacent reference pixels is not uniform, the built prediction model will not be able to accurately fit these 4 adjacent reference pixels. The distribution of reference pixels, such as the comparison schematic diagram of the prediction model shown in FIG. 4B.

In the embodiment of this application, on the basis of deriving the two mean points mean _min and mean _max by retaining the traditional scheme, first obtain the luminance mean value of these 4 adjacent reference pixels, and after obtaining 4 adjacent reference pixels After the brightness average value, the first reference pixel subset and/or the second reference pixel subset may be subjected to a second comparison process through the brightness average value, so as to divide the preset number of adjacent reference pixels into multiple situations. In many cases, two different fitting points are used to construct a predictive model.

Refer to FIG. 10, which shows a schematic flowchart of a model parameter derivation solution provided by an embodiment of the present application. As shown in Figure 10, the process may include:

S1001: Obtain 4 adjacent reference pixels;

S1002: Obtain two pixels with a larger value and two pixels with a smaller brightness component after 4 comparisons;

S1003: Calculate two mean points mean _min and mean _max ;

S1004: Calculate the brightness mean value meanL corresponding to 4 adjacent reference pixels;

S1005: Through a comparison, determine the first pixel corresponding to the minimum brightness value and the second pixel corresponding to the second minimum brightness value;

S1006: Determine whether meanL is greater than or equal to the second minimum brightness value;

S1007: When meanL is less than the second minimum brightness value, use the first pixel point corresponding to the minimum brightness value and the mean point mean _max as two fitting points to construct a prediction model;

S1008: When meanL is greater than or equal to the second minimum brightness value, determine the third pixel point corresponding to the second maximum brightness value and the fourth pixel point corresponding to the maximum brightness value through a comparison;

S1009: Determine whether meanL is less than or equal to the second maximum brightness value;

S1010: when meanL is greater than the second maximum brightness value, use the fourth pixel point corresponding to the maximum brightness value and the mean point mean _min as two fitting points to construct a prediction model;

S1011: When meanL is less than or equal to the second maximum brightness value, use the mean points mean _min and mean _max as two fitting points to construct a prediction model;

S1012: Perform chroma component prediction processing according to the constructed prediction model.

It should be noted that for steps S1006 and S1009, the comparison of these two steps is not in order; that is, the comparison of step S1006 can be performed first, and then the comparison of step S1009 can be performed; or the comparison of step S1009 can be performed first. Compare, and then perform the comparison in step S1006; the embodiment of the present application does not specifically limit it. For example, if the comparison of step S1009 is performed first, the flow at this time is modified as follows. After the third pixel corresponding to the second maximum brightness value and the fourth pixel corresponding to the maximum brightness value are determined through a comparison, it is determined whether meanL is less than or equal to The second maximum brightness value; when meanL is greater than the second maximum brightness value, the fourth pixel point corresponding to the maximum brightness value and the mean point mean _{min are} used as two fitting points to construct the prediction model; when meanL is less than or equal to the second maximum brightness value, Determine the first pixel corresponding to the minimum brightness value and the second pixel corresponding to the second minimum brightness value through a comparison; then determine whether meanL is greater than or equal to the second minimum brightness value; when meanL is less than the second minimum brightness value, the minimum brightness value The corresponding first pixel point and mean point mean _{max are} used as two fitting points to construct a prediction model; when meanL is greater than or equal to the second minimum brightness value, the mean points mean _min and mean _{max are} used as two fitting points to construct a prediction model; Finally, the chrominance component is predicted according to the constructed prediction model.

It should also be noted that the principle of "two points determine a straight line" is used to construct the prediction model; the two points here can be called fitting points. After obtaining 4 adjacent reference pixels, you can calculate the average value of these 4 adjacent reference pixels (which can be represented by mean), and then use the average value for the second comparison process to determine the fitting points in different situations .

Specifically, it is assumed that the image component to be predicted is a chrominance component, and the chrominance component is predicted by the luminance component. The following steps will be performed,

In the first step, assuming that the numbers of the four adjacent reference pixels obtained are 0, 1, 2, and 3 respectively, the four adjacent reference pixels are compared four times, and the 2 with the larger brightness value can be selected. Two adjacent reference pixels (can include the pixel with the largest brightness value and the pixel with the next largest brightness value) and 2 adjacent reference pixels with a smaller brightness value (can include the pixel with the smallest brightness value and the second brightness value The smallest pixel). Further, two arrays of minIdx[2] and maxIdx[2] can be set to store two sets of adjacent reference pixels respectively. Initially, the adjacent reference pixels numbered 0 and 2 are put into minIdx[2], and The adjacent reference pixels numbered 1 and 3 are put into maxIdx[2], as shown below,

Init:minIdx[2]={0,2},maxIdx[2]={1,3}

In the second step, through four comparisons, the two adjacent reference pixels with smaller brightness values can be stored in minIdx[2], and the two adjacent reference pixels with larger brightness values are stored in maxIdx[2] Pixels, as shown below,

Step1: if(L[minIdx[0]]>L[minIdx[1]],swap(minIdx[0],minIdx[1])

Step2: if(L[maxIdx[0]]>L[maxIdx[1]],swap(maxIdx[0],maxIdx[1])

Step3: if(L[minIdx[0]]>L[maxIdx[1]],swap(minIdx,maxIdx)

Step4: if(L[minIdx[1]]>L[maxIdx[0]],swap(minIdx[1],maxIdx[0])

In the third step, two adjacent reference pixels with smaller brightness values can be obtained. The corresponding brightness values are represented by luma ⁰ _min and luma ¹ _min , and the corresponding chroma values are represented by chroma ⁰ _min and chroma ¹ _min . ; At the same time, two adjacent reference pixels with larger brightness values can be obtained, and the corresponding brightness values are represented by luma ⁰ _max and luma ¹ _max respectively, and the corresponding chromaticity values are represented by chroma ⁰ _max and chroma ¹ _{max respectively} . Further, for the two smaller averaging the neighboring reference pixels can be obtained is represented by the mean point _min Mean, Mean _min the mean point corresponding to the luminance value LUMA _min, _min Chroma color value; for 2 The average value of a larger adjacent reference pixel point can be obtained, and the second average value point can be represented by mean _max . The luminance value of this mean value point mean _max is luma _max and the chroma value is chroma _max ; the details are as follows,

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

In the fourth step, after obtaining the two mean points mean _min (luma _min , chroma _min ) and mean _max (luma _max , chroma _max ), you can continue to execute the following modules:

Ⅰ. Obtain the meanL of the brightness of the four adjacent reference pixels with

numbers

0, 1, 2, and 3 respectively. It should be noted that it can be obtained by adding the brightness values corresponding to four adjacent reference pixels and dividing by four, or by (luma _min +luma _max +1)>>1;

Ⅱ, compare luma ⁰ _min and luma ¹ _min to determine the minimum point and the second minimum point (first comparison);

Ⅲ. Compare the average brightness value meanL of four adjacent reference pixels with the brightness value of the second smallest point (second comparison):

If meanL is greater than or equal to the brightness value of the second smallest point;

Skip to Ⅳ;

otherwise

Determine that the first fitting point is the minimum point, and the second fitting point is mean _max (luma _max , chroma _max ), exit this module at the same time, and skip to step 5;

Ⅳ, compare luma ⁰ _max and luma ¹ _max to determine the maximum point and the second maximum point (the third comparison);

Ⅴ, compare the brightness meanL of the four adjacent reference pixels with the brightness value of the second largest point (fourth comparison):

If meanL is less than or equal to the brightness value of the next maximum point;

Determine that the first fitting point is mean _min (luma _min , chroma _min ), and the second fitting point is mean _max (luma _max , chroma _max ), exit this module at the same time, and skip to step 5;

otherwise

Determine that the first fitting point is mean _min (luma _min , chroma _min ), and the second fitting point is the maximum point. At the same time, exit this module and skip to step 5;

In the fifth step, the model parameters can be derived from these two fitting points using "two points to determine a straight line". Assuming the first fitting point (luma ₁ , chroma ₁ ) and the second fitting point (luma ₂ , chroma ₂ ), the model parameters α and β can be calculated by formula (6),

In the sixth step, after the model parameters are derived, the chrominance value of the current coding block can be predicted through the prediction model.

In other words, after deriving the two model parameters (α and β), the prediction model corresponding to the chrominance component can be obtained according to the model parameters, as shown in equation (1); then the prediction model is used to predict the chrominance component Processing to obtain the predicted value corresponding to the chrominance component.

Before deriving the two fitting points, compared with the traditional solution, the embodiment of the present application adds one averaging operation and at least two and up to four comparison operations, so that the four adjacent reference The pixel points are divided into three cases, and two different fitting points are used to construct the prediction model for these three cases, thereby improving the robustness of CCLM prediction.

In another embodiment of the present application, the preset number is 4, so there may be 3 situations for these 4 adjacent reference pixels, such as the brightness value of meanL at the second maximum point and the brightness value of the second minimum point. Between, or the mean brightness value meanL is less than the brightness value of the second smallest point, or the mean brightness value meanL is greater than the brightness value of the second largest point; that is, the four adjacent reference pixels can show a uniform distribution trend or a non-uniform distribution trend, The following will specifically describe these three situations.

Optionally, in some embodiments, when the mean brightness value meanL is between the brightness value of the second-maximum point and the brightness value of the second-minimum point, the four adjacent reference pixel points are uniformly distributed. Refer to FIG. 11, which shows a schematic diagram of a comparison between a prediction model provided by an embodiment of the present application and a traditional solution provided by the present application. As shown in Figure 11, the 4 black dots are the 4 adjacent reference pixels, and the 2 gray dots are the average points corresponding to the two larger adjacent reference pixels among the 4 adjacent reference pixels. Mean points corresponding to 2 smaller adjacent reference pixels (that is, the two fitting points obtained by the traditional scheme), so the gray diagonal line is the prediction model constructed according to the traditional scheme; the bold black dashed line is 4 The brightness average value of adjacent reference pixels, the two bold black circles are two fitting points, so the bold black diagonal line is the prediction model constructed according to the scheme of the embodiment of the application; the gray dotted line is the four phases The prediction model fitted by the adjacent reference pixel using the LMS algorithm; as can be seen from Figure 11, the two fitted points are mean _min (luma _min , chroma _min ) and mean _max (luma _max , chroma _max ), so gray The oblique line and the bold black oblique line overlap, and are relatively close to the gray dotted line, that is, the scheme of the embodiment of the application is the same as the prediction model constructed by the traditional scheme, and both can accurately fit these 4 adjacent reference pixels Distribution.

Optionally, in some embodiments, when the mean brightness value meanL is less than the brightness value of the second smallest point, the four adjacent reference pixel points are non-uniformly distributed. Refer to FIG. 12, which shows a schematic diagram of comparison between another prediction model provided by the present application and the traditional scheme provided by an embodiment of the present application. As shown in Figure 12, the 4 black dots are the 4 adjacent reference pixels, and the 2 gray dots are the average points corresponding to the two larger adjacent reference pixels among the 4 adjacent reference pixels. Mean points corresponding to 2 smaller adjacent reference pixels (that is, the two fitting points obtained by the traditional scheme), so the gray diagonal line is the prediction model constructed according to the traditional scheme; the bold black dashed line is 4 The brightness average value of adjacent reference pixels, the two bold black circles are two fitting points, so the bold black diagonal line is the prediction model constructed according to the scheme of the embodiment of the application; the gray dotted line is the four phases The prediction model fitted by the adjacent reference pixel using the LMS algorithm; as can be seen from Figure 12, the two fitted points are the minimum point and the mean _max (luma _max , chroma _max ), so the black diagonal line and the gray point are bolded The line is closer, that is, the solution of the embodiment of the present application is more in line with the distribution of the four adjacent reference pixels than the prediction model constructed by the traditional solution.

Optionally, in some embodiments, when the mean brightness value meanL is greater than the brightness value of the second largest point, the four adjacent reference pixel points are non-uniformly distributed. Refer to FIG. 13, which shows a schematic diagram of a comparison between another prediction model provided by the present application and the traditional solution provided by an embodiment of the present application. As shown in Figure 13, the 4 black dots are the 4 adjacent reference pixels, and the 2 gray dots are the average points corresponding to the two larger adjacent reference pixels among the 4 adjacent reference pixels. Mean points corresponding to 2 smaller adjacent reference pixels (that is, the two fitting points obtained by the traditional scheme), so the gray diagonal line is the prediction model constructed according to the traditional scheme; the bold black dashed line is 4 The brightness average value of adjacent reference pixels, the two bold black circles are two fitting points, so the bold black diagonal line is the prediction model constructed according to the scheme of the embodiment of the application; the gray dotted line is the four phases The prediction model fitted by the adjacent reference pixel using the LMS algorithm; as can be seen from Figure 13, the two fitted points are the mean _min (luma _min , chroma _min ) and the maximum point, so the black diagonal line and the gray point are thickened The line is closer, that is, the solution of the embodiment of the present application is more in line with the distribution of the four adjacent reference pixels than the prediction model constructed by the traditional solution.

It can be seen from Figure 12 or Figure 13 that when the 4 adjacent reference pixels are unevenly distributed, the linear model constructed by the current traditional scheme lacks robustness and cannot fit these 4 adjacent references well. Pixels, thus making the prediction performance inaccurate. In the embodiment of the present application, by optimizing the fitting points used in the derivation of the model parameters, the CCLM prediction can be made more robust; specifically, the embodiment of the present application has undergone a mean value calculation with minimum Under the premise that the complexity of two comparisons and up to four comparisons increases, the distribution of 4 adjacent reference points is divided into three cases, and different fitting points are used to construct the prediction model for these three cases, thereby improving the CCLM Robustness of prediction; and in the case of different distributions of these 4 adjacent reference pixels, using different fitting points to derive model parameters can avoid the "divide by 3" operation and other division operations, such as "Division by 2" and "division by 4", etc., can be completely replaced by shifting, which can also save computational complexity. For example, based on the latest VVC reference software VTM5.0, under All intra conditions, the test sequence required by JVET is subject to the general test conditions, so that the average change of BD-rate on the Y component, Cb component and Cr component is -0.03%. , -0.18%, -0.17%, it also shows that the solution of the embodiment of the present application brings a certain improvement in prediction performance under the premise of a small increase in complexity.

In still another embodiment of the present application, after obtaining 4 adjacent reference pixels according to the latest VVC reference software VTM5.0, the 4 adjacent reference pixels can also be calculated according to the average brightness of the 4 adjacent reference pixels Divided into two categories, for example, the left one is represented by L, and the right one is represented by R. After classification, when the number of pixels in the L or R class is 3, the operation of "dividing by 3" will occur when the average value of the class is calculated. In the embodiment of the application, the fitting points of this class may not be Then there is the mean point of 3 pixels, but the mean point of 2 pixels is arbitrarily selected from the 3 pixels. This operation can remove the "divide" operation in the process of determining the fitting point (to be precise, the "divide by 3" operation can be avoided, and the "divide by 2" and "divide by 4" operations can be replaced by shifting) , Which can save computational complexity. Therefore, in some embodiments, after S701, the method may further include:

Grouping the preset number of adjacent reference pixel points by the first average point to obtain a third reference pixel subset and a fourth reference pixel subset;

Determine the first fitting point based on the third reference pixel subset; determine the second fitting point based on the fourth reference pixel subset;

Based on the first fitting point and the second fitting point, the model parameters are determined, and the prediction model corresponding to the image component to be predicted is obtained according to the model parameters; wherein, the prediction model is used to realize the Prediction processing of the image component to be predicted to obtain the predicted value corresponding to the image component to be predicted.

Further, the determining the first fitting point based on the third reference pixel subset; and determining the second fitting point based on the fourth reference pixel subset may include:

Selecting part of adjacent reference pixels from the third reference pixel subset, performing average calculation on the part of adjacent reference pixels, and using the calculated average point as the first fitting point;

A part of adjacent reference pixels is selected from the fourth reference pixel subset, the average value of the part of adjacent reference pixels is calculated, and the calculated average point is used as the second fitting point.

Selecting one of the adjacent reference pixel points from the third reference pixel subset as the first fitting point;

One of the adjacent reference pixel points is selected from the fourth reference pixel subset as the second fitting point.

Further, assuming that the value of the preset number is 4, determining the first fitting point based on the third reference pixel subset; and determining the second fitting point based on the fourth reference pixel subset includes:

If the third reference pixel subset includes three adjacent reference pixels and the fourth reference pixel subset includes one adjacent reference pixel, then two adjacent reference pixels are selected from the third reference pixel subset For reference pixels, perform average calculation on the selected two adjacent reference pixels, use the calculated average point as the first fitting point, and use 1 adjacent reference pixel in the fourth reference pixel subset As the second fitting point;

If the third reference pixel subset includes 1 adjacent reference pixel, and the fourth reference pixel subset includes 3 adjacent reference pixels, select 2 adjacent reference pixels from the fourth reference pixel subset For reference pixels, perform average calculation on the selected two adjacent reference pixels, use the calculated average point as the second fitting point, and use 1 adjacent reference pixel in the third reference pixel subset As the first fitting point.

It should be noted that some of the reference pixels selected here may be reference pixels of a power of two. For example, after dividing 4 adjacent reference pixels into two categories (such as L and R) according to the average brightness value, if the number of pixels in the L or R category is 3, then from 3 pixels In the operation of arbitrarily selecting 2 pixels in the points to find the average point, it can also be that if there are 3 pixels in the L class, then 2 pixels with the smaller brightness value are selected; if there are 3 pixels in the R class, Then select 2 pixels with larger brightness value.

In addition, after dividing the 4 adjacent reference pixels into two categories (such as the L category and the R category) according to the average brightness, one pixel point from the L category can be selected as the first fitting point. One pixel point is arbitrarily selected from the R category as the second fitting point. In this way, the two fitting points can be used to construct the model parameters α and β; the two fitting points can also be used to derive the first model parameter α, and the average point of 4 pixels can be used to derive the second model parameter β; In the process, it is possible to omit the operation of averaging points for L and R respectively.

Refer to FIG. 14, which shows a schematic structural diagram for determining a fitting point provided by an embodiment of the present application. As shown in Figure 14, after all the adjacent reference pixels are obtained, the luminance mean value L _{mean of} these adjacent reference pixels can be obtained, and then the adjacent reference pixels can be divided into L type and R type using the L _mean , The mean points (L _Lmean , C _Lmean ) and (L _Rmean , C _Rmean ) in each category can be used as two fitting points, and the first model parameter α is derived from these two fitting points, and the mean point ( L _mean , C _mean ) to derive the second model parameter β:

It should also be noted that the above method can also be applied to the case of at most 4 valid adjacent reference pixels in VTM5.0. For example, after obtaining 4 adjacent reference pixels through VTM5.0, first obtain the average brightness of the 4 pixels, use the average to divide the 4 pixels into two categories, and then use the average points of the two categories as two Fitting points to derive the model parameters α and β. When these 4 pixels cannot be divided into two categories according to the average brightness value, that is, when the number of pixels in the L type is 0 or the number of pixels in the R type is 0, the chromaticity average value of these 4 pixels can be directly used as Chromaticity prediction value. In addition, the two types of mean points can be used as two fitting points to derive the first model parameter α, and the mean point of 4 pixels can be used to derive the second model parameter β.

In another embodiment of the present application, on the basis of deriving the first model parameter α based on the two fitting points that have been obtained, the second model parameter β may be derived directly using the mean point of the two mean points that have been obtained. . Based on VTM5.0, the two average points are rounded down to an integer. At this time, the average value of these two average points is calculated again, and the result obtained by directly calculating the average value of 4 adjacent reference pixels may be different. , Which makes the prediction performance slightly different.

Refer to FIG. 15, which shows a schematic structural diagram of a prediction model provided by an embodiment of the present application. As shown in Figure 15, the gray oblique line represents the slope calculated based on the two fitting points, that is, the first model parameter α; then the gray oblique line is translated, and the amount of translation is the second model parameter β. The thick black diagonal line is the final prediction model. Specifically, based on the JVET-N0524 proposal, on the basis of the CCLM technology in VTM4.0, the two points with the largest and smallest brightness values among all adjacent reference points corresponding to the coding block can be used as the two fitting points to derive the first A model parameter α, and a second model parameter β is derived using the mean value of all adjacent reference pixels at the same time. In addition, the JVET-N0524 proposal can also be applied to the case of up to 4 valid adjacent reference pixels in VTM5.0. At this time, the average point and two of the two larger points of the 4 pixels can be used. The mean point of the two smaller points is used as two fitting points to derive the first model parameter α, and then the mean point of these 4 pixels is used to derive the second model parameter β.

This embodiment provides an image component prediction method. With the technical solution of this embodiment, after the second comparison processing, a preset number of adjacent reference pixels can be divided into 3 cases, and the three cases Two different fitting points are used to construct the prediction model, thereby improving the robustness of CCLM prediction; that is to say, by optimizing the fitting points used in the model parameter derivation, the constructed prediction model can be made more accurate , Can also improve the codec prediction performance of video images. .

Based on the same inventive concept as the foregoing embodiment, refer to FIG. 16, which shows a schematic diagram of the composition structure of an image component prediction apparatus 160 provided by an embodiment of the present application. The image component prediction device 160 may include: an acquisition unit 1601, a comparison unit 1602, a calculation unit 1603, and a prediction unit 1604, where

The acquiring unit 1601 is configured to acquire N adjacent reference pixels corresponding to the image components to be predicted of the encoding block in the video image; wherein, the N adjacent reference pixels are reference adjacent to the encoding block Pixels, N is a preset integer value;

The comparing unit 1602 is configured to compare the N first image component values corresponding to the N adjacent reference pixels to determine a first reference pixel subset and a second reference pixel subset; wherein, a preset number of phases The adjacent reference pixel points correspond to a preset number of first image component values, and the first reference pixel subset includes: among the preset number of first image component values, the smallest first image component value and the second smallest first image component value Corresponding to two adjacent reference pixels, the second reference pixel subset includes: among the preset number of first image component values, the two corresponding to the largest first image component value and the second largest first image component value Adjacent reference pixels;

The calculation unit 1603 is configured to calculate the average value of the N adjacent reference pixel points to obtain a first average value point;

The comparison unit 1602 is further configured to compare the values of the first image component of the first reference pixel subset and/or the second reference pixel subset through the first average point, and determine two fitting points;

The prediction unit 1604 is configured to determine model parameters based on the two fitting points, and obtain a prediction model corresponding to the image component to be predicted according to the model parameters; wherein, the prediction model is used to Prediction processing of the image component to be predicted to obtain the predicted value corresponding to the image component to be predicted.

In the above solution, referring to FIG. 16, the image component prediction device 160 may further include a screening unit 1605,

The obtaining unit 1601 is further configured to obtain a first reference pixel set corresponding to the image component to be predicted of the coding block in the video image;

The screening unit 1605 is configured to perform screening processing on the first reference pixel set to obtain a second reference pixel set; wherein, the second reference pixel set includes N adjacent reference pixels.

In the above solution, the acquiring unit 1601 is specifically configured to acquire reference pixels adjacent to at least one side of the encoding block; wherein, the at least one side includes the left side of the encoding block and/or the The upper side of the coding block; and based on the reference pixels, a first reference pixel set corresponding to the image component to be predicted is formed.

In the above solution, the acquiring unit 1601 is specifically configured to acquire reference pixels in a reference row or a reference column adjacent to the encoding block; wherein, the reference row is defined by the upper side of the encoding block And the rows adjacent to the upper right side, the reference column is composed of the columns adjacent to the left side and the lower left side of the coding block; and based on the reference pixels, the waiting Predict the first reference pixel set corresponding to the image component.

In the above solution, the screening unit 1605 is specifically configured to determine the position of the pixel to be selected based on the pixel position and/or image component intensity corresponding to each adjacent reference pixel in the first reference pixel set; and Determining the position of the pixel to be selected, selecting adjacent reference pixels corresponding to the position of the pixel to be selected from the first reference pixel set, and composing the selected adjacent reference pixels into a second reference pixel set; Wherein, the second reference pixel set includes N adjacent reference pixels.

In the above solution, the acquiring unit 1601 is further configured to acquire a preset number of first image component values based on the preset number of adjacent reference pixels;

The comparing unit 1602 is specifically configured to perform multiple comparisons on a preset number of first image component values to obtain a first array composed of the smallest first image component value and the second smallest first image component value and the largest first image A second array composed of component values and the next largest first image component value; and placing two adjacent reference pixels corresponding to the first array into the first reference pixel subset to obtain the first reference pixel subset And placing two adjacent reference pixels corresponding to the second array into a second reference pixel subset to obtain the second reference pixel subset.

In the above solution, the obtaining unit 1601 is further configured to obtain the Nth image component value and the second image component value corresponding to each of the N adjacent reference pixels. The average value of the first image component corresponding to one image component and the average value of the second image components corresponding to the N second image components are obtained to obtain the first average point.

In the above solution, the obtaining unit 1601 is further configured to obtain the first reference pixel based on the first image component value and the second image component value corresponding to each adjacent reference pixel in the first reference pixel subset. The mean values of the first image components corresponding to the multiple first image components in the pixel subset and the mean values of the second image components corresponding to the multiple second image components to obtain the second mean point; and based on each phase in the second reference pixel subset The first image component value and the second image component value corresponding to the adjacent reference pixels are obtained, and the first image component average value corresponding to the multiple first image components in the second reference pixel subset and the first image component corresponding to the multiple second image components are obtained. The two image components are averaged to obtain the third average point.

In the above solution, the comparing unit 1602 is specifically configured to compare the first image component of the first mean point with the second smallest first image component value in the first reference pixel subset; and when the first mean point is When the first image component is greater than or equal to the second smallest first image component value in the first reference pixel subset, the first image component of the first mean point is combined with the second largest first image component value in the second reference pixel subset Compare; and when the first image component of the first average point is less than or equal to the second largest first image component value in the second reference pixel subset, the second average point is used as the first fitting point, and the third average Point as the second fitting point; wherein, the two fitting points include a first fitting point and a second fitting point.

In the above solution, the comparison unit 1602 is further configured to compare the first image component values corresponding to two adjacent reference pixel points in the first reference pixel subset; and determine two values according to the comparison result. The smallest first image component value and the second smallest first image component value among the first image component values, and the pixel corresponding to the smallest first image component value is used as the first adjacent reference pixel, and the second smallest first image component value The corresponding pixel is used as the second adjacent reference pixel;

The comparing unit 1602 is further configured to compare the first image component values corresponding to two adjacent reference pixels in the second reference pixel subset; and determine the two first image component values according to the comparison result The second largest first image component value and the largest first image component value in the, and the pixel corresponding to the second largest first image component value is taken as the third adjacent reference pixel, and the pixel corresponding to the largest first image component value is taken as The fourth adjacent reference pixel.

In the above solution, the comparing unit 1602 is further configured to use the first adjacent reference pixel as the first fitting point when the first image component of the first mean point is less than the second smallest first image component value, and The third mean point is used as the second fitting point.

In the above solution, the comparing unit 1602 is further configured to use the second average point as the first fitting point when the first image component of the first average point is greater than the second largest first image component value, and the fourth phase The adjacent reference pixel is used as the second fitting point.

In the above solution, the calculation unit 1603 is further configured to obtain the first model parameter based on the first fitting point and the second fitting point, by calculating a model with a first preset factor; and For the first model parameter and the first fitting point, the second model parameter is obtained through a second preset factor calculation model.

In the above solution, the calculation unit 1603 is further configured to obtain the second model parameter through a second preset factor calculation model based on the first model parameter and the first average point.

In the above solution, referring to FIG. 16, the image component prediction device 160 may further include a grouping unit 1606 and a determining unit 1607, where:

The grouping unit 1606 is configured to perform grouping processing on the N adjacent reference pixels through a first average point to obtain a third reference pixel subset and a fourth reference pixel subset;

The determining unit 1607 is configured to determine the first fitting point based on the third reference pixel subset; determine the second fitting point based on the fourth reference pixel subset;

The prediction unit 1604 is specifically configured to determine model parameters based on the first fitting point and the second fitting point, and obtain a prediction model corresponding to the image component to be predicted according to the model parameters; wherein, The prediction model is used to implement prediction processing on the image component to be predicted, so as to obtain the predicted value corresponding to the image component to be predicted.

In the above solution, the determining unit 1607 is specifically configured to select a part of adjacent reference pixels from the third reference pixel subset, perform average calculation on the part of adjacent reference pixels, and calculate the average value obtained by the calculation. As the first fitting point; and selecting a part of adjacent reference pixel points from the fourth reference pixel subset, performing average calculation on the part of adjacent reference pixels, and using the calculated average point as the first Two fitting points.

In the above solution, the determining unit 1607 is specifically configured to select one of the adjacent reference pixel points from the third reference pixel subset as the first fitting point; and from the fourth reference pixel subset One of the adjacent reference pixels is selected as the second fitting point.

In the above solution, the value of N is 4; the determining unit 1607 is specifically configured to: if the third reference pixel subset includes 3 adjacent reference pixels, the fourth reference pixel subset includes 1 Adjacent reference pixels, select two adjacent reference pixels from the third reference pixel subset, perform average calculation on the selected two adjacent reference pixels, and use the calculated average point as the first Fitting point, using one adjacent reference pixel in the fourth reference pixel subset as the second fitting point; and if the third reference pixel subset includes one adjacent reference pixel, the first The four reference pixel subset includes 3 adjacent reference pixels, then 2 adjacent reference pixels are selected from the fourth reference pixel subset, and the average value of the selected 2 adjacent reference pixels is calculated to obtain The mean point of is used as the second fitting point, and an adjacent reference pixel point in the third reference pixel subset is used as the first fitting point.

In the above solution, the prediction unit 1604 is specifically configured to perform prediction processing on the image component to be predicted for each pixel in the coding block based on the prediction model to obtain the image component to be predicted for each pixel. Predictive value.

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

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

Therefore, this embodiment provides a computer storage medium that stores an image component prediction program that implements the method described in any one of the foregoing embodiments when the image component prediction program is executed by at least one processor.

Based on the composition of the image component prediction device 160 and the computer storage medium described above, refer to FIG. 17, which shows the specific hardware structure of the image component prediction device 160 provided by the embodiment of the present application, which may include: a network interface 1701, a memory 1702, and a processor 1703: The components are coupled together through the bus system 1704. It can be understood that the bus system 1704 is used to implement connection and communication between these components. In addition to the data bus, the bus system 1704 also includes a power bus, a control bus, and a status signal bus. However, for clarity of description, various buses are marked as the bus system 1704 in FIG. 19. Among them, the network interface 1701 is used to receive and send signals in the process of sending and receiving information with other external network elements;

The memory 1702 is configured to store computer programs that can run on the processor 1703;

The processor 1703 is configured to execute: when the computer program is running:

It can be understood that the memory 1702 in the embodiment of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory. Among them, the non-volatile memory can be read-only memory (Read-Only Memory, ROM), programmable read-only memory (Programmable ROM, PROM), erasable programmable read-only memory (Erasable PROM, EPROM), and electrically available Erase programmable read-only memory (Electrically EPROM, EEPROM) or flash memory. The volatile memory may be a random access memory (Random Access Memory, RAM), which is used as an external cache. By way of exemplary but not restrictive description, many forms of RAM are available, such as static random access memory (Static RAM, SRAM), dynamic random access memory (Dynamic RAM, DRAM), synchronous dynamic random access memory (Synchronous DRAM, SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (Double Data Rate SDRAM, DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (Enhanced SDRAM, ESDRAM), Synchronous Link Dynamic Random Access Memory (Synchlink DRAM, SLDRAM) And Direct Rambus RAM (DRRAM). The memory 1702 of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

The processor 1703 may be an integrated circuit chip with signal processing capabilities. In the implementation process, the steps of the foregoing method can be completed by hardware integrated logic circuits in the processor 1703 or instructions in the form of software. The aforementioned processor 1703 may be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (ASIC), a ready-made programmable gate array (Field Programmable Gate Array, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like. The steps of the method disclosed in the embodiments of the present application may be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers. The storage medium is located in the memory 1702, and the processor 1703 reads the information in the memory 1702, and completes the steps of the foregoing method in combination with its hardware.

It can be understood that the embodiments described herein can be implemented by hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more Application Specific Integrated Circuits (ASIC), Digital Signal Processing (DSP), Digital Signal Processing Equipment (DSP Device, DSPD), programmable Logic device (Programmable Logic Device, PLD), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA), general-purpose processors, controllers, microcontrollers, microprocessors, and others for performing the functions described in this application Electronic unit or its combination.

For software implementation, the technology described herein can be implemented through modules (such as procedures, functions, etc.) that perform the functions described herein. The software codes can be stored in the memory and executed by the processor. The memory can be implemented in the processor or external to the processor.

Optionally, as another embodiment, the processor 1703 is further configured to execute the method described in any one of the foregoing embodiments when the computer program is running.

Refer to FIG. 18, which shows a schematic diagram of the composition structure of an encoder provided by an embodiment of the present application. As shown in FIG. 18, the encoder 180 may at least include the image component prediction device 160 described in any of the foregoing embodiments.

Refer to FIG. 19, which shows a schematic diagram of the composition structure of a decoder provided by an embodiment of the present application. As shown in FIG. 19, the decoder 190 may at least include the image component prediction device 190 described in any of the foregoing embodiments.

It should be noted that in this application, the terms "include", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements , And also includes other elements not explicitly listed, or elements inherent to the process, method, article, or device. If there are no more restrictions, the element defined by the sentence "including a..." does not exclude the existence of other identical elements in the process, method, article or device that includes the element.

The serial numbers of the foregoing embodiments of the present application are for description only, and do not represent the superiority of the embodiments.

The methods disclosed in the several method embodiments provided in this application can be combined arbitrarily without conflict to obtain new method embodiments.

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

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

The above are only specific implementations of this application, but the protection scope of this application is not limited to this. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application. Should be covered within the scope of protection of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Industrial applicability

In the embodiment of the present application, first, for the image components to be predicted of the coding block in the video image, N adjacent reference pixels are obtained, where the N adjacent reference pixels are reference pixels adjacent to the coding block. N is a preset integer value; then, the N first image component values corresponding to the N adjacent reference pixels are compared to determine the first reference pixel subset and the second reference pixel subset, the first reference pixel The set includes: two adjacent reference pixel points corresponding to the smallest first image component value and the second smallest first image component value among the preset number of first image component values, and the second reference pixel subset includes: Set the two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value in the number of first image component values; calculate the average value of the N adjacent reference pixels to obtain the first Mean point; compare the first image component values of the first reference pixel subset and/or the second reference pixel subset through the first mean point to determine two fitting points; then determine the model parameters based on the two fitting points , Obtain the prediction model corresponding to the image component to be predicted according to the model parameters to obtain the predicted value corresponding to the image component to be predicted; in this way, after the second comparison process, the preset number of adjacent reference pixels can be divided into multiple situations, For these various situations, using two different fitting points to construct the prediction model can improve the robustness of CCLM prediction; that is to say, by optimizing the fitting points used in the derivation of model parameters, the constructed prediction The model is more accurate and improves the coding and decoding prediction performance of video images.

Claims

An image component prediction method, the method includes:

Acquire N adjacent reference pixels corresponding to the to-be-predicted image components of the encoding block in the video image; wherein, the N adjacent reference pixels are reference pixels adjacent to the encoding block, and N is a preset Integer value

Compare the N first image component values corresponding to the N adjacent reference pixels to determine a first reference pixel subset and a second reference pixel subset; wherein, the first reference pixel subset includes: The two adjacent reference pixel points corresponding to the smallest first image component value and the second smallest first image component value in the number of first image component values, the second reference pixel subset includes: a preset number of first Two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value in the image component values;

Calculating the average value of the N adjacent reference pixel points to obtain the first average value point;

Comparing the values of the first image component to the first reference pixel subset and/or the second reference pixel subset by using the first average point to determine two fitting points;

Based on the two fitting points, the model parameters are determined, and the prediction model corresponding to the image component to be predicted is obtained according to the model parameters; wherein, the prediction model is used to realize the prediction processing of the image component to be predicted, To obtain the predicted value corresponding to the image component to be predicted.
The method according to claim 1, wherein the obtaining N adjacent reference pixels corresponding to the image component to be predicted of the coding block in the video image comprises:

Acquiring the first reference pixel set corresponding to the image component to be predicted of the coding block in the video image;

The first reference pixel set is screened to obtain a second reference pixel set; wherein, the second reference pixel set includes N adjacent reference pixels.
The method according to claim 2, wherein said obtaining the first reference pixel set corresponding to the image component to be predicted of the coding block in the video image comprises:

Acquiring reference pixels adjacent to at least one side of the coding block; wherein the at least one side includes the left side of the coding block and/or the upper side of the coding block;

Based on the reference pixel points, a first reference pixel set corresponding to the image component to be predicted is formed.
The method according to claim 2, wherein said obtaining the first reference pixel set corresponding to the image component to be predicted of the coding block in the video image comprises:

Obtain a reference row or reference pixel point in a reference column adjacent to the encoding block; wherein the reference row is composed of the upper side of the encoding block and the row adjacent to the upper right side, the The reference column is composed of columns adjacent to the left side and the lower left side of the coding block;

Based on the reference pixel points, a first reference pixel set corresponding to the image component to be predicted is formed.
The method according to any one of claims 2 to 4, wherein the filtering process on the first reference pixel set to obtain a second reference pixel set comprises:

Determine the position of the pixel to be selected based on the pixel position and/or image component intensity corresponding to each adjacent reference pixel in the first reference pixel set;

According to the determined position of the pixel to be selected, adjacent reference pixels corresponding to the position of the pixel to be selected are selected from the first reference pixel set, and the selected adjacent reference pixels form a second reference pixel set ; Wherein, the second reference pixel set includes N adjacent reference pixels.
The method according to claim 1, wherein the comparing the N first image component values corresponding to the N adjacent reference pixels to determine the first reference pixel subset and the second reference pixel subset comprises:

Obtaining a preset number of first image component values based on the preset number of adjacent reference pixels;

The preset number of first image component values are compared multiple times to obtain a first array composed of the smallest first image component value and the second smallest first image component value, and the largest first image component value and the second largest first image component value. A second array of image component values;

Putting two adjacent reference pixels corresponding to the first array into a first reference pixel subset to obtain the first reference pixel subset;

Put two adjacent reference pixels corresponding to the second array into a second reference pixel subset to obtain the second reference pixel subset.
The method according to claim 1, wherein the calculating the average value of the N adjacent reference pixel points to obtain the first average value point comprises:

Based on the first image component value and the second image component value corresponding to each adjacent reference pixel in the N adjacent reference pixels, the first image component average value and the Nth image component corresponding to the N first image components are obtained. The average value of the second image component corresponding to the two image components is obtained to obtain the first average value point.
The method according to claim 1, wherein after said comparing the N-to-first image component values corresponding to the N adjacent reference pixels to determine the first reference pixel subset and the second reference pixel subset, The method also includes:

Based on the first image component value and the second image component value corresponding to each adjacent reference pixel point in the first reference pixel subset, first images corresponding to the multiple first image components in the first reference pixel subset are obtained A component average value and a second image component average value corresponding to a plurality of second image components to obtain a second average value point;

Based on the first image component value and the second image component value corresponding to each adjacent reference pixel point in the second reference pixel subset, first images corresponding to the multiple first image components in the second reference pixel subset are obtained The component average value and the second image component average value corresponding to the multiple second image components are obtained to obtain the third average value point.
8. The method according to claim 8, wherein the first image component value is compared on the first reference pixel subset and/or the second reference pixel subset through the first mean point to determine two fittings Points, including:

Comparing the first image component of the first mean point with the second smallest first image component value in the first reference pixel subset;

When the first image component of the first average point is greater than or equal to the second smallest first image component value in the first reference pixel subset, the first image component of the first average point and the second reference pixel subset are divided into the second The largest first image component value for comparison;

When the first image component of the first mean point is less than or equal to the second largest first image component value in the second reference pixel subset, the second mean point is taken as the first fitting point, and the third mean point is taken as the second Fitting points; wherein the two fitting points include a first fitting point and a second fitting point.
The method according to claim 9, wherein before said comparing the first image component of the first mean point with the second smallest first image component value in the first reference pixel subset, the method further comprises:

Comparing the first image component values corresponding to two adjacent reference pixels in the first reference pixel subset;

According to the result of the comparison, determine the smallest first image component value and the second smallest first image component value of the two first image component values, and use the pixel corresponding to the smallest first image component value as the first adjacent reference pixel , The pixel corresponding to the second smallest first image component value is used as the second adjacent reference pixel;

Correspondingly, before the comparing the first image component of the first average point with the second largest first image component value in the second reference pixel subset, the method further includes:

Comparing the first image component values corresponding to two adjacent reference pixels in the second reference pixel subset;

According to the result of the comparison, determine the second largest first image component value and the largest first image component value among the two first image component values, and use the pixel corresponding to the second largest first image component value as the third adjacent reference pixel Point, the pixel point corresponding to the largest first image component value is used as the fourth adjacent reference pixel point.
The method according to claim 10, wherein after the comparing the first image component of the first mean point with the second smallest first image component value in the first reference pixel subset, the method further comprises:

When the first image component of the first average point is smaller than the second smallest first image component value, the first adjacent reference pixel point is used as the first fitting point, and the third average point is used as the second fitting point.
The method according to claim 10, wherein after the comparing the first image component of the first mean point with the second largest first image component value in the second reference pixel subset, the method further comprises:

When the first image component of the first average point is greater than the second largest first image component value, the second average point is taken as the first fitting point, and the fourth adjacent reference pixel point is taken as the second fitting point.
The method according to any one of claims 9 to 12, wherein the determining model parameters based on the two fitting points comprises:

Based on the first fitting point and the second fitting point, obtaining the first model parameter through a first preset factor calculation model;

Based on the first model parameter and the first fitting point, the second model parameter is obtained through a second preset factor calculation model.
The method according to claim 13, wherein, after the first model parameter is obtained through the first preset factor calculation model, the method further comprises:

Based on the first model parameter and the first average point, the second model parameter is obtained through a second preset factor calculation model.
The method according to claim 1, wherein after said obtaining a preset number of adjacent reference pixels, the method further comprises:

Grouping the N adjacent reference pixel points by the first average point to obtain a third reference pixel subset and a fourth reference pixel subset;

Determine the first fitting point based on the third reference pixel subset; determine the second fitting point based on the fourth reference pixel subset;

Based on the first fitting point and the second fitting point, the model parameters are determined, and the prediction model corresponding to the image component to be predicted is obtained according to the model parameters; wherein, the prediction model is used to realize the Prediction processing of the image component to be predicted to obtain the predicted value corresponding to the image component to be predicted.
The method according to claim 15, wherein the determining the first fitting point based on the third reference pixel subset; and determining the second fitting point based on the fourth reference pixel subset comprises:

Selecting part of adjacent reference pixels from the third reference pixel subset, performing average calculation on the part of adjacent reference pixels, and using the calculated average point as the first fitting point;

A part of adjacent reference pixels is selected from the fourth reference pixel subset, the average value of the part of adjacent reference pixels is calculated, and the calculated average point is used as the second fitting point.
The method according to claim 15, wherein the determining the first fitting point based on the third reference pixel subset; and determining the second fitting point based on the fourth reference pixel subset comprises:

Selecting one of the adjacent reference pixel points from the third reference pixel subset as the first fitting point;

One of the adjacent reference pixel points is selected from the fourth reference pixel subset as the second fitting point.
The method according to claim 16, wherein the value of N is 4; the determining the first fitting point based on the third reference pixel subset; determining the second fitting point based on the fourth reference pixel subset includes :

If the third reference pixel subset includes three adjacent reference pixels and the fourth reference pixel subset includes one adjacent reference pixel, then two adjacent reference pixels are selected from the third reference pixel subset For reference pixels, perform average calculation on the selected two adjacent reference pixels, use the calculated average point as the first fitting point, and use 1 adjacent reference pixel in the fourth reference pixel subset As the second fitting point;

If the third reference pixel subset includes 1 adjacent reference pixel, and the fourth reference pixel subset includes 3 adjacent reference pixels, select 2 adjacent reference pixels from the fourth reference pixel subset For reference pixels, perform average calculation on the selected two adjacent reference pixels, use the calculated average point as the second fitting point, and use 1 adjacent reference pixel in the third reference pixel subset As the first fitting point.
The method according to any one of claims 1 to 18, wherein, after obtaining the prediction model corresponding to the image component to be predicted according to the model parameters, the method further comprises:

Performing prediction processing on the image component to be predicted for each pixel in the coding block based on the prediction model to obtain a predicted value corresponding to the image component to be predicted for each pixel.
An image component prediction device. The image component prediction device includes: an acquisition unit, a comparison unit, a calculation unit, and a prediction unit, wherein,

The acquiring unit is configured to acquire N adjacent reference pixels corresponding to the image components to be predicted of the encoding block in the video image; wherein the N adjacent reference pixels are reference pixels adjacent to the encoding block Point, N is a preset integer value;

The comparing unit is configured to compare the N-to-first image component values corresponding to the N adjacent reference pixels to determine a first reference pixel subset and a second reference pixel subset; wherein, the first reference pixel The subset includes: two adjacent reference pixel points corresponding to the smallest first image component value and the second smallest first image component value among the preset number of first image component values, and the second reference pixel subset includes: Two adjacent reference pixels corresponding to the largest first image component value and the second largest first image component value among the preset number of first image component values;

The calculation unit is configured to calculate the average value of the N adjacent reference pixel points to obtain a first average value point;

The comparison unit is further configured to perform a second comparison process of the values of the first image component on the first reference pixel subset and/or the second reference pixel subset through the first average point, and determine two fittings point;

The prediction unit is configured to determine model parameters based on the two fitting points, and obtain a prediction model corresponding to the image component to be predicted according to the model parameters; wherein, the prediction model is used to implement Prediction processing of the predicted image component to obtain the predicted value corresponding to the image component to be predicted.
An image component prediction device, wherein the image component prediction device includes: a memory and a processor;

The memory is used to store a computer program that can run on the processor;

The processor is configured to execute the method according to any one of claims 1 to 19 when running the computer program.
A computer storage medium, wherein the computer storage medium stores an image component prediction program that implements the method according to any one of claims 1 to 19 when the image component prediction program is executed by at least one processor.