WO2011057557A1 - Method, apparatus and system for video image encoding and decoding - Google Patents

Abstract

A method, apparatus and encoding-decoding system for video image encoding processing and decoding processing are provided by the present invention. The encoding processing method includes: performing down-sampling disassembly on the macro block of a video image frame to obtain a plurality of blocks which at least include a first block, a second block and a third block; performing intra-frame predictive encoding processing on the first block to obtain the reconstructed image value of the first block; by using the reconstructed image value of the first block as a reference, performing predictive encoding processing on the second block to obtain the reconstructed image value of the second block; by using the reconstructed image value of the second block as a reference, performing predictive encoding processing on the third block. The embodiments of the present invention can sufficiently utilize the correlation among pixels, and improve prediction precision and the encoding compression efficiency of the video image.

Description

 The present invention claims to be submitted to the Chinese Patent Office on November 10, 2009, and the application number is 200910211704. 2 , and the invention name is "video image encoding processing, decoding processing method and device, and codec. The priority of the Chinese patent application of the system is also required to be submitted to the Chinese Patent Office on April 9, 2010, and the application number is 201010147573. 9. The invention name is "video image encoding processing, decoding processing method and device, and codec. The priority of the Chinese Patent Application, the entire disclosure of which is hereby incorporated by reference. Technical field

 The present invention relates to the field of video image processing technologies, and in particular, to a video image encoding process, a decoding processing method and apparatus, and a codec system. Background technique

In the future, the demand for the transmission and storage of high-definition video images will explode. However, the development of information network infrastructure such as the Internet, communication networks, and broadcast networks has lagged behind, and network bandwidth has become a bottleneck. Based on this fact, the next-generation video coding standard will focus on high-definition, ultra-high-definition video, and greatly improve the compression efficiency of video image coding to alleviate the contradiction between the growth of bandwidth demand for high-definition video image transmission and the development of information network infrastructure. How to improve the compression efficiency of video image coding is a core problem that needs to be solved in the new generation of video coding standards. Block-based hybrid video coding framework is one of the general video compression coding processing schemes, mainly including intra coding technology and inter coding technology. The intra-frame coding technique utilizes spatial correlation, such as the use of intra prediction (Intra prediction) technology to remove spatial redundancy information. The intra-frame coding can be completed by the current frame information itself, without reference to other frame information, ie No reference frames are required. The interframe coding technique utilizes time domain correlation, such as the use of motion compensated prediction techniques to remove time domain redundancy information, which requires predictive coding based on information of the reference frame. The above coding technique can achieve higher compression efficiency of video image coding.

 In order to obtain higher video coding compression efficiency, one prior art performs predictive coding by using correlation between coded neighboring blocks, and writes a prediction mode of the image component and its corresponding prediction residual information into the code. Streaming, so that the decoding end decodes according to the preset mode and its corresponding prediction residual information. Another prior art is: intra prediction coding using correlation between adjacent pixels, such as: downsampling a block of size 8 X 8 (in pixels) to obtain 4 sizes of 4 X The block of 4, intra-coded a block of size 4 X 4 to obtain a reconstructed image of the block, and predictively encodes other blocks of size 4×4 with reference to the reconstructed image of the block.

 The inventors found in the process of implementing the embodiments of the present invention that the prior art does not fully utilize the correlation between pixels, and the video encoding compression efficiency is relatively low. Summary of the invention

 The invention provides a video image encoding process, a decoding processing method and device, and a codec system to improve the compression efficiency of video image coding.

 The present invention provides a video image encoding processing method, including:

 Downsampling the macroblock of the video image frame to obtain a plurality of blocks, the plurality of blocks including at least: a first block, a second block, and a third block;

 Performing intra prediction encoding processing on the first block to obtain a reconstructed image value of the first block; performing prediction encoding processing on the second block with reference to the reconstructed image value of the first block; Obtaining a reconstructed image value of the second block;

The third block is subjected to predictive coding processing with reference to the reconstructed image value of the second block. The present invention also provides a video image decoding processing method, including: Obtaining coded bit information of a macroblock of a video image frame, the macroblock including at least: a first block, a second block, and a third block;

 Performing intra prediction decoding processing on the coded bit information of the first block to obtain a decoded image value of the first block;

 Referring to the decoded image value of the first block, performing prediction decoding processing on the encoded bit information of the second block to obtain a decoded image value of the second block;

 Deriving and decoding the coded bit information of the third block with reference to the decoded picture value of the second block to obtain a decoded picture value of the third block;

 The decoded image values of the blocks in the macroblock are subjected to upsampling synthesis processing to obtain decoded image values of the macroblocks.

 The present invention also provides a video image encoding processing apparatus, including:

 a downsampling and decoding module, configured to perform downsampling on a macroblock of a video image frame to obtain a plurality of blocks, where the plurality of blocks include at least: a first block, a second block, and a third block;

 An intra prediction coding module, configured to perform intra prediction coding processing on the first block to obtain a reconstructed image value of the first block;

 a first inter-pixel predictive coding module, configured to perform predictive coding processing on the second block with reference to a reconstructed image value of the first block, to obtain a reconstructed image value of the second block;

 The second inter-pixel predictive coding module is configured to perform predictive coding processing on the third block with reference to the reconstructed image value of the second block.

 The present invention also provides a video image decoding processing apparatus, including:

 a bit information acquiring module, configured to acquire coded bit information of a macroblock of a video image frame, where the macroblock includes at least: a first block, a second block, and a third block;

An intra prediction decoding module, configured to perform intra prediction decoding on the coded bit information of the first block Processing, obtaining a decoded image value of the first block;

 The first inter-pixel predictive decoding module is configured to perform predictive decoding processing on the encoded bit information of the second block with reference to the decoded image value of the first block to obtain a decoded image value of the second block;

 a second inter-pixel prediction decoding module, configured to perform predictive decoding processing on the encoded bit information of the third block with reference to the decoded image value of the second block, to obtain a decoded image value of the third block;

 And a decoded image generating module, configured to perform upsampling synthesis processing on the decoded image values of the blocks in the macroblock to obtain decoded image values of the macroblock.

 The invention also provides a video image coding and decoding system, comprising:

 a video image encoding processing apparatus, configured to downsample a macroblock of a video image frame to obtain a plurality of blocks, where the plurality of blocks include at least: a first block, a second block, and a third block; Performing intra prediction coding processing to obtain a reconstructed image value of the first block; performing prediction encoding processing on the second block with reference to the reconstructed image value of the first block to obtain the second block a reconstructed image value; performing predictive encoding processing on the third block with reference to the reconstructed image value of the second block;

 a video image decoding processing apparatus, configured to acquire coded bit information of the macroblock, where the macroblock includes at least: the first block, the second block, and the third block; and the first block The coded bit information is subjected to intra prediction decoding processing to obtain a decoded image value of the first block; and the coded bit information of the second block is subjected to predictive decoding processing with reference to the decoded image value of the first block. a decoded image value of the second block; performing predictive decoding processing on the encoded bit information of the third block with reference to the decoded image value of the second block to obtain a decoded image value of the third block; The decoded image values of the respective blocks are subjected to upsampling synthesis processing to obtain decoded image values of the macroblocks.

In the embodiment of the present invention, the macroblock is downsampled and decomposed to obtain a plurality of blocks, and the intra prediction technique and the interpixel prediction technique are combined in the process of predictive coding or decoding of the plurality of blocks, and at least two levels of cascade prediction are formed. Reference relationship, so you can make full use of the correlation between pixels, improve prediction accuracy and video The coding compression efficiency of the image. DRAWINGS

 In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present invention, and other drawings can be obtained from those skilled in the art without any inventive labor.

 1 is a flowchart of a video image encoding processing method according to a first embodiment of the present invention;

 2 is a flowchart of a video image decoding processing method according to a second embodiment of the present invention;

 FIG. 3 is a flowchart of a video image encoding processing method according to a third embodiment of the present invention; FIG. 3b is a block diagram of a logical implementation of a video image encoding processing method according to a third embodiment of the present invention; Schematic diagram of macroblock downsampling;

 FIG. 3 is a schematic diagram of an inter-pixel prediction coding interpolation process according to a third embodiment of the present invention; FIG. 4 is a flowchart of a video image decoding processing method according to a fourth embodiment of the present invention;

 5a is a flowchart of a video image encoding processing method according to a fifth embodiment of the present invention; FIG. 5b is a schematic diagram of a macroblock decomposition according to a fifth embodiment of the present invention;

 FIG. 6 is a flowchart of a video image decoding processing method according to a sixth embodiment of the present invention;

 FIG. 7 is a flowchart of a video image encoding processing method according to a seventh embodiment of the present invention; FIG. 7b is a schematic diagram of a macroblock down sampling decomposition according to a seventh embodiment of the present invention;

 8 is a flowchart of a video image decoding processing method according to an eighth embodiment of the present invention;

9 is a schematic structural diagram of a video image encoding processing apparatus according to a ninth embodiment of the present invention; FIG. 10 is a schematic structural diagram of a video image decoding processing apparatus according to a tenth embodiment of the present invention; Schematic diagram of video image codec system structure. detailed description

 BRIEF DESCRIPTION OF THE DRAWINGS The technical solutions in the embodiments of the present invention will be described in detail with reference to the accompanying drawings. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any creative work are within the scope of the present invention.

 FIG. 1 is a flowchart of a video image encoding processing method according to a first embodiment of the present invention. As shown in FIG. 1, the video image encoding processing method in this embodiment includes:

 Step 11. Down-sampling the macroblock of the video image frame to obtain a plurality of blocks, where the plurality of blocks include at least: a first block, a second block, and a third block.

 In an embodiment of the invention, the pixels in the first block include at least pixels in the lower right corner of the macroblock. The pixel in the lower right corner of the second block is in the horizontal or vertical direction from the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.

 Of course, in other optional embodiments, the pixels in the first block may include at least pixels in the upper left corner of the macroblock. The pixel in the upper left corner of the second block is in the horizontal or vertical direction from the pixel in the upper left corner of the first block. The pixel in the upper left corner of the third block and the pixel in the upper left corner of the first block are in the diagonal direction of the macro block.

 The downsampling coefficient can be determined according to the actual coding requirement of the video image frame, and the macroblock is downsampled and decomposed according to the downsampling coefficient to obtain a plurality of blocks. If the downsampling coefficient is 丄, the macroblock is downsampled and decomposed.

 M

You can get a block. Grouping in A ² blocks, for example, dividing the blocks into N groups according to the positional relationship between the pixels constituting each block extracted in the downsampling process, each group consisting of one or more blocks, and N is greater than or equal to An integer of 3. The first group includes at least one first block, the second group includes at least one second block, the third group includes at least one third block, and so on, until packet processing of the blocks is completed. Optional In the process of grouping A ² blocks, grouping may be performed according to whether the positional relationship of the extracted pixels is adjacent, for example, dividing different blocks composed of pixels adjacent to the same pixel into the same group.

 Step 12: Perform intra prediction coding processing on the first block to obtain a reconstructed image value of the first block. In this step, the first block may be subjected to intra prediction encoding processing by using pixel values of other macroblocks encoded in the frame to obtain a reconstructed image value of the first block. The reconstructed image value consists of the predicted value of the first block and the predicted residual after the encoding process.

 Step 13. Perform a predictive coding process on the second block with reference to the reconstructed image value of the first block, and obtain a reconstructed image value of the second block.

 Optionally, during the predictive coding process on the second block, the reconstructed image value of the first block may be interpolated, and the pixel value of the corresponding position after interpolation is used as the predicted value of the second block, according to the second block. The predicted value and its prediction residual construct the reconstructed image value of the second block.

 Step 14. Perform prediction encoding processing on the third block with reference to the reconstructed image value of the second block. Optionally, during the predictive coding process on the second block, the reconstructed image value of the second block may be interpolated, and the pixel value of the corresponding position after interpolation is used as the predicted value of the third block, according to the third block. The predicted value and its prediction residual construct the reconstructed image value of the third block.

In this embodiment, the first block, the second block, and the third block are respectively composed of pixels extracted by using a preset extraction rule for all pixels of the macro block, such as every other or more in the horizontal and vertical directions or other directions. The pixels extracted by the pixels are respectively composed, and thus the pixels of the respective blocks are interlaced with each other in the spatial position. In the above steps 13 and 14, since the coded block (the 2/3th block) and the pixel of the reference block (the first block) are spatially interlaced with each other, the prediction between them can be fully utilized. Correlation between pixels, thereby improving prediction accuracy. It can be seen that this is an inter-pixel prediction method, which is different from the intra prediction method of step 12 (which is actually an inter-block prediction method). The inter-pixel predictive coding method is used to predictively encode other blocks than the first block, and the second block is the reconstructed image value of the first block. Referring to the prediction coding, the third block is predictive coding based on the reconstructed image value of the second block. Alternatively, a similar prediction reference relationship may be established between the plurality of blocks obtained by the macroblock downsampling decomposition. Thus, in the process of predictive coding of a plurality of blocks obtained by downsampling decomposition, at least two levels of cascaded prediction reference relationships are formed.

 Further, in addition to the above technical solution, further, coded bit information of each block obtained by encoding the prediction residual of each block in the macroblock may be written into the code stream. It should be noted that the case where the coded stream is written to the prediction residual information of each block will be described as an example. Actually, it is also possible to allow the prediction residual information of each block not to be written into the code stream, or the prediction residual information of the first block is written into the code stream and the prediction residual information of the second block and the third block is not. The case of writing a stream, or other combination.

 This embodiment is exemplified by encoding in the order of the first block, the second block, and the third block. Further, based on the same principle, encoding may be performed in the order of the third block, the second block, the first block, or the first block, the third block, and the second block. The encoding process will not be described again.

 In the video image coding processing method of the embodiment, the macroblock is downsampled and decomposed to obtain a plurality of blocks, and a cascaded reference relationship is formed between the plurality of blocks, so that correlation between pixels can be fully utilized to improve prediction accuracy, and at the same time, the implementation As an encoding technique of macroblock layer (MB level), the texture characteristics of different regions in the image can be better adapted on the macroblock layer, thereby improving coding compression efficiency.

 The size of the macro block in this embodiment may be 16 X 16, 32 X 32, 64 X 64 or 128 X 128, or the like. Macroblocks mentioned in the subsequent embodiments refer to this meaning unless otherwise stated.

 FIG. 2 is a flowchart of a video image decoding processing method according to a second embodiment of the present invention. This embodiment is a decoding end processing flow corresponding to the encoding processing method shown in Fig. 1. As shown in FIG. 2, the video image decoding processing method in this embodiment includes:

Step 21: Acquire coded bit information of a macroblock of a video image frame, where the macroblock includes at least: a first block, a second block, and a third block. In an embodiment of the invention, the pixels in the first block include at least pixels in the lower right corner of the macroblock. The pixel in the lower right corner of the second block is in the horizontal or vertical direction from the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.

 Of course, in other optional embodiments, the pixels in the first block may include at least pixels in the upper left corner of the macroblock. The pixel in the upper left corner of the second block is in the horizontal or vertical direction from the pixel in the upper left corner of the first block. The pixel in the upper left corner of the third block and the pixel in the upper left corner of the first block are in the diagonal direction of the macro block.

 The macroblock downsampling decomposition mode is shown in step 11 of the corresponding embodiment, and is not described here. The code side carries the coded bit information of each block of the macro block in the code stream sent to the decoding end, and the decoding end decodes the bit information of each block to obtain the residual decoding coefficient of each block.

 Step 22: Perform intra prediction decoding processing on the coded bit information of the first block to obtain a decoded image value of the first block.

 The intra-predictive decoding process of the coded bit information constituting the first block can be performed by using the pixel values of other macroblocks decoded in the frame, and the predicted value of the first block and its residual are obtained, and the predicted value of the first block is compared with The residuals are added to obtain the decoded image value of the first block. This is a traditional intra prediction decoding method.

 Step 23: Perform prediction decoding processing on the coded bit information of the second block with reference to the decoded image value of the first block to obtain a decoded image value of the second block.

 Optionally, in performing the predictive decoding process on the coded bit information of the second block, the method may include performing interpolation processing on the reconstructed image value of the first block, and the pixel value of the corresponding position after the interpolation is used as the predicted value of the second block. The decoded image value of the second block is constructed based on the predicted value of the second block and its residual decoding coefficient.

Step 24: Perform prediction decoding processing on the coded bit information of the third block with reference to the decoded image value of the second block to obtain a decoded image value of the third block. Optionally, in performing the predictive decoding process on the encoded bit information of the third block, the method may include performing interpolation processing on the reconstructed image value of the second block, and the pixel value of the corresponding position after the interpolation is used as the predicted value of the third block. The decoded image value of the third block is constructed based on the predicted value of the third block and its residual decoding coefficient.

 In the above steps 23 and 24, since the decoded block (the 2/3th block) and the pixel of the reference block (the first block) are spatially interlaced with each other, the prediction between them can be fully utilized. Correlation between each pixel, thereby improving prediction accuracy. It can be seen that this is an inter-pixel prediction method, which is different from the intra prediction method of step 22 (which is actually an inter-block prediction method). The inter-pixel prediction decoding method is used for predictive decoding on other blocks except the first block, the second block is predictive decoding based on the decoded image value of the first block, and the third block is based on the decoded image value of the second block. Perform predictive decoding. Optionally, a similar decoding prediction relationship may be established between multiple blocks included in the macroblock, so that at least two stages are formed in the prediction decoding process of the plurality of blocks obtained by downsampling decomposition. Cascaded predictive reference relationships.

 Step 25: Perform upsampling synthesis processing on the decoded image values of the blocks in the macroblock to obtain decoded image values of the macroblock.

 The upsampling synthesis process performed by the decoding end is an inverse process of the downsampling decomposition performed by the encoding end, that is, this step is the inverse process of the step 11 in the encoding process of the first embodiment of the present invention.

 It should be noted that the decoding process of each block according to the present embodiment is described by taking the case where the prediction residual coefficient exists in each block as an example. Actually, it is also possible to allow the case where the prediction residual coefficient is not present for each block, or the case where the prediction coefficient is present in the first block and the prediction residual coefficient is not present in the second block and the third block, or other combinations.

 This embodiment is exemplified by decoding in the order of the first block, the second block, and the third block. Further, based on the same principle, decoding may be performed in the order of the third block, the second block, the first block, or the first block, the third block, and the second block.

In this embodiment, the video image decoding processing method performs downsampling decomposition on the macroblock to obtain a plurality of blocks for pre-preprocessing. In the process of decoding and decoding, the intra-prediction decoding technology is combined with the inter-pixel prediction decoding technology, and at least two levels of cascaded prediction reference relationships are formed in the inter-pixel prediction decoding process, so that the pixel correlation can be fully utilized for encoding compression. Decoding of the video image afterwards. FIG. 3 is a flowchart of a video image encoding processing method according to a third embodiment of the present invention; FIG. 3b is a logic block diagram of a video image encoding processing method according to a third embodiment of the present invention. In this embodiment, a block-based hybrid video coding framework is used to perform a downsampling decomposition of a 16×16 macroblock as an example.

 2

Bright. As shown in FIG. 3a, the video image encoding processing method of this embodiment includes: Step 31: Input a video image frame to be encoded. Step 32: Taking 丄 as a downsampling coefficient, performing a 16×16 macroblock in the video image frame

 2

Decomposed into four blocks of 8 x 8 size. FIG. 3c is a schematic diagram of a macroblock downsampling decomposition according to a third embodiment of the present invention. As shown in Fig. 3c, using 丄 as the downsampling coefficient, downsampling the 16×16 macroblocks in the video image frame is: 2

Every 1 pixel is extracted in the horizontal and vertical directions, and the extracted pixels form a new block, which can form 4 blocks of 8×8 size. The pixels that make up the different blocks are distinguished in Figure 3c by different filling methods. To facilitate the description of the technical solution of the present embodiment, the blocks composed of the positional relationships of the pixels extracted from left to right and from top to bottom may be labeled as: blocks 0, 1, 2, and 3. According to the positional relationship of the extracted pixels in the macroblock downsampling process, the blocks are divided into multiple groups, for example: block 3 is a group, blocks 1 and 2 are a group, and block 0 is a group. In the following encoding process for each block, the predictive coding type of block 3 is intra prediction coding, and the prediction coding type of other blocks (such as blocks 0, 1, and 2) is inter-pixel predictive coding. The "I" and "P" in the figure respectively indicate that the prediction coding type of the corresponding block is: intra prediction coding or inter-pixel prediction type. Therefore, in this embodiment, the block 3 is the "first block" according to the embodiment of the present invention, and the blocks 1 and 2 are the "second block" according to the embodiment of the present invention, and the block 0 is the embodiment of the present invention. Said "third block". In the present invention In an embodiment, the pixel in the first block includes at least a pixel in a lower right corner of the macroblock. The pixel in the lower right corner of the second block is in the horizontal or vertical direction from the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.

 Of course, in other optional embodiments, the pixels in the first block may include at least pixels in the upper left corner of the macroblock. The pixel in the upper left corner of the second block is in the horizontal or vertical direction from the pixel in the upper left corner of the first block. The pixel in the upper left corner of the third block and the pixel in the upper left corner of the first block are in the diagonal direction of the macro block.

 An aliasing effect may result during downsampling decomposition. In order to further improve the coding performance, optionally, the embodiment may perform anti-aliasing filtering on some or all of the pixels of the macroblock in the downsampling decomposition process:

 For example one: anti-aliasing filtering can be performed on all pixels of a macroblock, such as anti-aliasing filtering using a low-pass filter to filter out high-frequency components, thereby reducing aliasing effects on encoding during downsampling decomposition The impact.

 For example two: Anti-aliasing filtering can be used in the downsampling decomposition of blocks 0, 1 and 2, and anti-aliasing filtering is performed in the downsampling process of block 3, such as using a low-pass filter. The anti-aliasing filtering process is performed as a downsampling filter to filter out high frequency components in the block 3. Since block 3 has a large influence on the coding quality of other blocks (blocks 0, 1, and 2), only block 3 is subjected to anti-aliasing filtering processing, and the aliasing effect in the process of reducing the downsampling decomposition is caused by coding. Based on the technical effects of the impact, it also helps to reduce the complexity of the implementation.

 Step 33: Perform intra prediction coding processing on the block 3 to obtain the reconstructed image value of the block 3 and its prediction residual.

This step can perform intra prediction encoding on block 3 by using the pixel values of other macroblocks encoded in the frame. A predicted value is obtained, and the predicted value is compared with the original pixel of the macroblock to obtain a prediction residual of the block 3, and the reconstructed image value of the block 3 is constructed based on the predicted value of the block 3 and the encoded residual after the encoding. Step 34: Perform prediction encoding processing on Block 1 and Block 2 with reference to the reconstructed image values of the block 3, and obtain reconstructed image values of the blocks 1 and 2 and their prediction residuals, respectively. Step 35: Perform prediction encoding processing on block 0 with reference to the reconstructed image values of block 1 and/or block 2, to obtain reconstructed image values of block 0 and prediction residuals thereof. Optionally, in the foregoing step 34 and step 35, during the predictive coding process on the blocks 0, 1 and 2, in order to improve the accuracy of the predictive coding, the reconstructed image values of the referenced block may be interpolated, and the interpolation is performed. The pixel value of the position is the predicted value. The referenced reconstructed image values may be interpolated using a multi-tap interpolation filter. FIG. 3 is a schematic exploded view of inter-pixel predictive coding interpolation processing according to a third embodiment of the present invention. As shown in Fig. 3d, a 4-tap (4-tap) interpolation filter can be used to interpolate the referenced reconstructed image values, where: Interpolation processing for block 1 and block 2 is as follows: 4-tap interpolation The filter performs interpolation of the reconstructed image values of the block 3 in the horizontal direction or the vertical direction, and interpolates the vertical direction of the reconstructed image values of the block 3.

 2

The pixel value corresponding to the position is a predicted value of the block 1 pixel, and the pixel value corresponding to the interpolation position in the horizontal direction of the reconstructed image value of the block 3 is the predicted value of the block 2 pixel. The interpolation processing for block 0 is, for example, finding an average value of pixel values of a plurality of pixels around the pixel constituting the block 0, which is a predicted value of the pixel of the block 0. Specifically, the average of the pixel values of the plurality of pixels around the pixel of the block 0 can be averaged, and FIG. 3d only illustrates the case where the four pixels are from the block 1 and the block 2, respectively. It can be understood that if block 0 is referenced to the reconstructed image value of block 1, the four pixels may come from block 1 or from blocks 1 and 3, if block 0 is referenced to the reconstructed image value of block 2; The number of pixels may be from block 2 or from blocks 2 and 3; the number of pixels for averaging is not limited to four, for example, the average of the pixel values of 8 pixels around the pixel of block 0 may be averaged. , these 8 pixels can come from blocks 1, 2 And 3, etc.

 Optionally, the reconstructed image values of the macroblocks may be constructed according to the reconstructed image values of the blocks obtained in the above steps 33-35, and the reconstructed image values of the macroblocks are stored.

 Step 36: Write coded bit information of each block obtained by encoding the prediction residual of each block in the macroblock into the code stream.

 It should be noted that the present embodiment will be described by taking an example in which the prediction residual information of each block is written into the code stream. Actually, it is also possible to allow the prediction residual information of each block not to be written into the code stream, or the prediction residual information of the first block is written into the code stream and the prediction residual information of the second block and the third block is not The case of writing a stream, or other combination.

 In the block-based hybrid video coding framework, the predictive coding process for any block includes processes such as prediction, transform, quantization, and entropy coding to implement coding compression of the video image. FIG. 3b is a logic block diagram of a video image encoding processing method according to a third embodiment of the present invention. The following describes the processing procedure of the blocks included in the macroblock in the application scenario of the block-based hybrid video coding framework in this embodiment, which may specifically include:

 (1) Performing intra prediction on block 3 to obtain a predicted value and a prediction residual thereof, and performing transform processing, quantization processing, and entropy encoding processing on the prediction residual of block 3, and constructing according to the prediction residual of block 3 and its predicted value. The reconstructed image value of block 3.

 (2) Taking the reconstructed image value of block 3 as a reference, predicting blocks 1 and 2 respectively to obtain predicted values and their prediction residuals, and performing transform processing, quantization processing, and entropy coding on the prediction residuals of blocks 1 and 2, respectively. Processing, reconstructed image values of blocks 1 and 2 are constructed based on respective prediction residuals of blocks 1 and 2 and their predicted values, respectively.

(3) Taking the reconstructed image values of block 1 and/or 2 blocks as reference, predicting block 0 to obtain the predicted value and its prediction residual, and performing transform processing, quantization processing and entropy coding on the prediction residual of block 0 in sequence. deal with, The reconstructed image value of block 0 is constructed from the predicted value of block 0 and its prediction residual.

 Optionally, if the macroblock further includes other blocks, the other blocks are separately predicted, and the prediction residuals of the other blocks are sequentially subjected to transform processing, quantization processing, and entropy encoding processing. The prediction residuals of the blocks included in the macroblock are written into the code stream by the entropy coding process, and the reconstructed image values of the respective blocks are combined and processed to obtain reconstructed image values of the macroblocks.

 Due to the introduction of cascading predictive coding, the above-mentioned block can have two blocks of different predictive coding types, such as intra prediction and inter-pixel prediction, so that the block can be optimized in the process of quantization processing, as described below. :

 A block-level quantization mapping mechanism can be established within the macroblock, that is, according to the importance of different blocks within the macroblock, divided into multiple groups, and different quantization parameters (Quinization Parameter, QP for short) are assigned to each block.

 Since the quality of the referenced block affects the quality of the block with which it is referenced, in order to obtain better Rate-Distortion (RD) performance, the principle of QP value allocation between blocks is: The relationship determines the QP value. In the above encoding process, there are the following reference relationships: Block 0 refers to block 1 and block 2, and block 1 and block 2 refer to block 3. The QP value of the block being referenced is given a smaller QP value.

 According to the above principle, the reference relationship between the blocks shown in Figure 3c, the QP value of the block can be satisfied as follows:

 QP (block 3) ^QP (block 1/2) ^QP (block 0)

 In practical applications, the QP value of each block can be determined experimentally. For example, the following relationship can be sampled to determine the QP value of each block:

 QP (block 1/2) = QP (block 0) = QP (Slice)

QP (block 3) = QP (Slice) -6 Where QP (Slice) represents the QP value corresponding to the slice of the video image, and the QP value corresponding to the slice can be obtained in advance.

 After determining the QP value of each block, the prediction residual of each block is subjected to corresponding quantization processing according to the QP value of each block. In order to obtain better quantization and coding performance, different quantization and optimization processes can be performed on different blocks according to the different QP values of different blocks in the quantization process as follows:

 Example 1: In the implementation of Rate-Distortion Optimization (RDO), it is necessary to consider the different QP values of different blocks within a macroblock. When calculating the rate distortion overhead (rdcost) for different blocks, determine according to the QP value. The lambda coefficient corresponding to different QP values is different from the corresponding lambda coefficient.

 Example 2: In the process of quantization rounding, consider different prediction coding types of each block in the macroblock, and use different dead zones for different blocks to match intra prediction and inter-pixel prediction. The coefficients of the block prediction residual of the prediction coding type are transformed to have different distribution laws. In the process of performing quantization and rounding, the selection principle of the dead zone used by each block is: a smaller dead zone is used for the intra prediction coding type block, and a larger dead zone is used for the inter-pixel prediction coding type block. That is, a smaller dead zone is used for block 3, and a larger dead zone is used for blocks 2, 1 and 0, for example, the dead zone used for block 3 is A, and the dead zone used for block 2, 1 or 0 is B, Then A is less than B.

Other coded bit information may be carried in the code stream outputted by the coding end, such as entropy coding information of a syntax element "Coded Block Pattern" (CBP) in the video coding standard. Due to the introduction of cascading predictive coding, there may be two blocks of different predictive coding types, intra prediction and inter-pixel prediction, and the prediction residuals of different prediction coding types have different distributions after quantization. Regularity, taking this into consideration, the entropy coding mode of the syntax element CBP can be optimized accordingly. Specifically, CBP is used to indicate whether there is a non-zero coefficient for the block corresponding to each bit (Bit). For the Luma component, there are 4 bits in the CBP indicating 4 8 X 8 large in the macroblock. Whether small blocks have non-zero coefficients. In the entropy coding process, the entropy coding process is performed on each block according to the distribution rule of the quantized coefficients of the prediction residuals of each block to obtain better performance, and the results of these entropy coding processes can be written. Code stream.

 For example: In the Context-based Adaptive Variable Length Coding (CAVLC) method, it is necessary to re-design the combination of intra prediction and inter-pixel according to different prediction coding types in different macroblocks. A code table of CBPs that predict block characteristics for two different predictive coding types. The probability that the CBP bit of the intra prediction coding type block has a "1" is large, and the CBP bit of the inter-preferential coding type block has a large "0" rate, and the support: CBP is used to indicate the block. Whether there are 4 bits of non-zero coefficient in sequence, which are used to indicate whether there are non-zero coefficients in block 3, block 2, block 1 and block 0, then the value of 4 bits in the CBP of the macro block is "1000". " can be regarded as a large probability event, assigning shorter codewords to each block; and the value of 4 bits in the CBP of the macroblock can be regarded as a small probability event in the order of "0111", and each block is allocated longer. Codeword. The prediction residual of the corresponding block is subjected to entropy coding processing according to the codeword corresponding to the code table of the CBP of each block.

 For example 2: In the Context-based Adaptive Binary Arithmetic Coding (CAB AC) method, it is necessary to redesign the probabilistic model initialization and the model update method according to different prediction coding types of different blocks within the macroblock. Specifically, different intra-prediction coding types of intra- and inter-pixel blocks within a macroblock are respectively applied with different initialization probability models, and different update methods are applied to each block. For example: in the CBP of the block of the intra prediction coding type, the above bit value is "1", which means a More Probable Symbol (MPS), and the bit value of the block of the inter-pixel predictive coding type is "1". "Less Probable Symbol (LPS)", when performing the selection of the probabilistic model and the model update, the model selection and update can be performed based only on the position of each bit of the CBP without using the context information.

In the video image encoding processing method of this embodiment, block 0 is predicted by using block 1 and/or block 2 as a reference. Coding, and block 1 and block 2 are both predictively coded with reference to block 3, which is equivalent to forming a cascaded prediction reference relationship between blocks; and since this embodiment encodes at the macroblock layer (MB level) Processing can thus be referred to as "MB-level cascade predictive coding." Different from the frame level downsampling coding method (such as the method of dividing the Intra frame into two fields for field coding), this embodiment is used as a macroblock layer coding processing technology, and can be combined with other existing macroblock types. Participate in RDO decision coding together, so as to better adapt to the texture characteristics of different regions in the image on the macroblock layer.

 It should be noted that although the predictive coding of the macroblock of the size of X 16 in this embodiment is described, it can be understood that other sizes of blocks, such as 32 X 32, 64 X 64 or The macroblocks of size 128 X 128 are used for predictive coding. The downsampling coefficients in the predictive coding process can be divided into four groups or more than four groups according to actual needs, thereby forming three or more cascades. The implementation mechanism of the prediction reference relationship is similar to that of the embodiment, and details are not described herein again.

 FIG. 4 is a flowchart of a video image decoding processing method according to a fourth embodiment of the present invention. This embodiment is a decoding end processing flow corresponding to the encoding processing method provided by the third embodiment of the present invention. As shown in FIG. 4, the video image decoding processing method in this embodiment includes:

 Step 41: Acquire a code stream of coded bit information of blocks 0, 1, 2, and 3 included in the macro block of the video image frame.

 Step 42: Decode the coded bit information of blocks 0, 1, 2, and 3, respectively, to obtain a residual decoding coefficient of each block.

 In the block-based hybrid video coding framework, the decoding process of the encoded bit information of any block includes processes such as entropy decoding, inverse quantization, inverse transform, and prediction to implement decoding of the encoded compressed video image. Optionally, in the application scenario of the block-based hybrid video coding framework, step 42 may include step 421 - step 423.

Step 421: Perform entropy decoding processing on the coded bit information of each block. In the entropy decoding process, according to the distribution rule of the residual decoding coefficients of each block, the encoding bit information of each block is subjected to corresponding entropy decoding processing, and the block using the intra prediction coding mode for predictive coding has non-zero coefficients. The probability of occurrence is greater than the probability that a block that is predictively coded using inter-pixel predictive coding mode exhibits non-zero coefficients, such as: the probability that block 3 exhibits a non-zero coefficient, which is greater than the probability that blocks 0, 1, and 2 have non-zero coefficients.

 Step 422: Perform inverse quantization processing on the coefficients after entropy decoding processing of each block.

 Since the quality of the referenced block affects the quality of the block with which it is referenced, in order to obtain better RD performance, the QP value assignment principle between blocks is: Determine the QP value based on the reference relationship. The reference relationship existing in the subsequent decoding process can be determined in advance, for example, there is a reference relationship as follows: block 0 reference block 1/2, block 1/2 reference sub-block 3. The QP value of the block to be referenced is given a smaller QP value.

 According to the above principle, the QP value of the available block satisfies the following formula:

 QP (block 3) ^QP (block 1/2) ^QP (block 0)

 In practical applications, the QP value of each block can be determined experimentally. For example, the following relationship can be sampled to determine the QP value of each block:

 QP (block 1/2) = QP (block 0) = QP (Slice)

 QP (block 3) = QP (Slice) -6

 Among them, QP (Slice) represents the QP value corresponding to the slice of the video image, and the QP value corresponding to the slice can be obtained in advance.

 After determining the QP value of each block, the entropy-decoded coefficients of the coded bit information of each block are subjected to corresponding inverse quantization processing according to the QP value of each block.

 Step 423: Perform inverse transform processing on the coefficients of each block subjected to the inverse quantization process to obtain a residual.

Step 43: Perform intra prediction decoding processing on the block 3, and add the predicted value to the residual of the block 3 to obtain the decoded image value of the block 3. Step 44: Perform prediction decoding processing on Block 1 and Block 2 with reference to the decoded image value of the block 3, and add the predicted values of the block 1 and the block 2 to their respective residuals to obtain the blocks 1 and 2 respectively. Decode image values.

 Step 45: Perform prediction decoding processing on block 0 with reference to the decoded image values of block 1 and/or block 2, and add the predicted value of block 0 to its residual to obtain a decoded image value of block 0.

 The foregoing step 43 to step 45 are equivalent to establishing a cascading prediction reference relationship in each block prediction decoding process, and the implementation manner thereof is similar to the method in which the coding end establishes a cascading prediction reference relationship in each block predictive coding process. For details, see step 32. - the description of step 34, the method for performing interpolation processing on blocks 0, 1 and 2 in steps 33 and 34 is also applicable to the decoding steps 44 and 45 to improve the accuracy of the prediction, and the specific implementation is here. No longer.

 Step 46: Perform up-sampling synthesis processing on the decoded image values of the blocks to obtain decoded image values of the macroblocks.

 It should be noted that the decoding process of each block according to the present embodiment is described by taking the case where the prediction residual coefficient exists in each block as an example. Actually, it is also possible to allow the case where the prediction residual coefficient is not present for each block, or the case where the prediction coefficient is present in the first block and the prediction residual coefficient is not present in the second block and the third block, or other combinations.

 In the video image decoding processing method of this embodiment, both block 1 and block 2 perform prediction decoding with reference to block 3, and block 0 performs prediction decoding with reference to block 1 and/or block 2, which is equivalent to forming between blocks. The cascaded prediction reference relationship enables decoding of a video image that is encoded and compressed to fully utilize pixel correlation.

FIG. 5a is a flowchart of a video image encoding processing method according to a fifth embodiment of the present invention. The difference between this embodiment and the third embodiment of the present invention is that the block 3 is further divided into a plurality of sub-blocks, and the embodiment can be applied to the encoding process of a video image with rich texture details. As shown in Figure 5a, this is The embodiment video image encoding processing method includes:

 Step 51: Input a video image frame to be encoded.

 Step 52: Perform downsampling on the 16×16 macroblock in the video image frame to obtain four 8×8 size blocks, which are respectively labeled as: blocks 0, 1, 2, and 3, and divide the block 3 into 4 blocks. 4 X 4 size sub-blocks, labeled as: Sub-blocks 30, 31, 32, and 33, respectively.

 In this step, the method of decomposing the 16×16 macroblocks into four 8×8-sized blocks is similar to the step 32, and details are not described herein again.

 FIG. 5b is a schematic diagram of macroblock decomposition according to a fifth embodiment of the present invention. As shown in Fig. 5b, this embodiment divides the block 3 into four sub-blocks of size 4 x 4, i.e., sub-blocks 30, 31, 32, and 33. The decomposition process of block 3 in this embodiment does not involve downsampling. In the figure, the prediction coding types indicated by "I" in each sub-block are: intra prediction coding.

 Step 53: Perform intra prediction encoding processing on the sub-blocks 30, 31, 32, and 33, respectively, obtain reconstructed image values of the respective sub-blocks, perform synthesis processing on the reconstructed image values of the respective sub-blocks, and obtain reconstruction of the block 3. Image value.

 In this step, the prediction value may be obtained by performing intra prediction on the constituent sub-block 30 by using the pixel values of other macroblocks encoded in the frame, and comparing the predicted value with the original pixel of the macroblock to obtain a prediction residual of the sub-block 30, according to The predicted value of the sub-block 30 and the reconstructed image value of the prediction residual constructed sub-block 30 after the encoding process. In a similar manner, the predicted values of the sub-blocks 31, 32, and 33 and their corresponding prediction residuals are obtained, respectively, and reconstructed image values of the sub-blocks 31, 32, and 33 are constructed, respectively. The reconstructed image values of the sub-blocks 30, 31, 32, and 33 are subjected to synthesis processing to obtain reconstructed image values of the block 3.

 Steps 54-55 are the same as steps 34-35, and are not described here.

Step 56: Write coded bit information of each sub-block and each block obtained by encoding the prediction residual of each sub-block and each block in the macro block into the code stream. The encoding processing of the prediction residuals of the sub-blocks 30, 31, 32, and 33 in this step is the same as the encoding processing of the prediction residual of the block (such as block 3) in which the prediction encoding type is intra-predictive encoding in step 36. The prediction residuals for blocks 0, 1, and 2 are the same as the encoding process for the prediction residuals of the blocks whose prediction coding type is inter-pixel prediction coding (such as blocks 0, 1, and 2), and are not described herein again.

 The video image encoding processing method of this embodiment implements the technical effects of the third embodiment of the present invention, and divides the block 3 into a plurality of sub-blocks (ie, sub-blocks 30, 31, 32, and 33), and performs sub-blocks on each sub-block. In intra prediction coding, the reconstructed image values of the block 3 obtained by combining the reconstructed image values of the respective sub-blocks are used as references of the blocks 1 and 2. The difference between this embodiment and the third embodiment of the present invention is that one 8×8 intra prediction code in the third embodiment of the present invention is replaced by four 4×4 intra prediction codes. 4 X 4 intra prediction coding is more suitable for encoding video images with rich texture details.

 FIG. 6 is a flowchart of a video image decoding processing method according to a sixth embodiment of the present invention. This embodiment is a decoding end processing flow corresponding to the encoding processing method provided by the fifth embodiment of the present invention. As shown in FIG. 6, the video image decoding processing method in this embodiment includes:

 Step 61: The code stream of the encoded bit information of the sub-blocks 30, 31, 32 and 33 and the blocks 0, 1 and 2 of the acquired macroblock of the video image frame.

 Step 62: Decode the coded bit information of the sub-blocks 30, 31, 32, and 33 and the blocks 0, 1, and 2, respectively, to obtain the residual of each block.

 The decoding process of the coded bit information of the sub-blocks 30, 31, 32 and 33 in this step is the same as the decoding process of the coded bit information of the block (such as block 3) of the predictive coding type being intra-predictive coded in step 42. The coded bit information for the blocks 0, 1 and 2 is the same as the method for decoding the bit information of the block (for example, blocks 0, 1 and 2) whose prediction coding type is inter-pixel predictive coding, and will not be described herein.

Step 63: Perform intra prediction decoding processing on the sub-blocks 30, 31, 32, and 33, and add the predicted value to the residual corresponding to each sub-block to obtain a decoded image value of each sub-block, and decode each sub-block. Image value The line synthesis processing obtains the decoded image value of the block 3. In this step, the prediction value may be obtained by performing intra prediction on the constituent sub-block 30 by using the pixel values of other macroblocks decoded in the frame, and the decoded image values of the sub-block 30 are constructed according to the prediction value of the sub-block 30 and its residual. Using a similar method, the decoded image values of the sub-blocks 31, 32, and 33 are obtained, respectively. The decoded image values of the sub-blocks 30, 31, 32, and 33 are subjected to synthesis processing to obtain decoded image values of the block 3. Step 64-step 66 is the same as step 44-step 46, and details are not described herein again. In the video image decoding processing method of the embodiment, the video image decoding method provided by the fourth embodiment of the present invention achieves the technical effect, and the plurality of sub-blocks divided by the intra-predictive coding block are rich in texture details. In the application scenario of the encoding process of the video image, decoding of the video image after the encoding and compression using the pixel correlation can be realized. FIG. 7a is a flowchart of a video image encoding processing method according to a seventh embodiment of the present invention. The difference between this embodiment and the third embodiment of the present invention is that the present embodiment performs the downsampling decomposition again in block 3 to obtain a plurality of sub-blocks. The difference between this embodiment and the fifth embodiment of the present invention is that the block is in this embodiment. 3 The downsampling decomposition is performed again to obtain a plurality of sub-blocks, and the decomposition process in the fifth embodiment does not involve downsampling. This embodiment can be applied to the encoding process of a video image with less texture details. As shown in FIG. 7a, the video image encoding processing method of this embodiment includes: Step 71: Input a video image frame to be encoded. Step 72: Taking 丄 as a downsampling coefficient, performing a 16×16 macroblock in the video image frame

 2

Decomposed into four blocks of 8 x 8 size, labeled as: blocks 0, 1, 2, and 3;

2 sample coefficients, downsampling decomposition of block 8 of 8×8 size, and obtaining 4 4×4 size sub-blocks, respectively labeled as: sub-blocks 30, 31, 32 and 33. The block of the intra prediction coding type of the prediction coding type may be downsampled again according to the actual coding requirement of the video image frame to obtain multiple sub-blocks, and the prediction coding type of the plurality of sub-blocks is intra prediction. The sub-block of the code is subjected to downsampling decomposition again... and so on, multi-level downsampling decomposition is performed, and the decomposed block or sub-block is encoded. This idea can be called Hierarchical coding method. .

 Assuming that the downsampling coefficient is 丄, the type of predictive coding that can be obtained by downsampling the macroblock is intraframe.

The block of the predictive coding is subjected to downsampling decomposition again, and ^two sub-blocks are obtained. Grouping in ² blocks, for example, according to the positional relationship between the sub-block pixels extracted in the downsampling process, the m ² sub-blocks are divided into groups, each group consisting of one or more sub-blocks, each group consisting of One or more blocks consisting of "integers greater than or equal to 3. The first group includes at least one first sub-block, the second group includes at least one second sub-block, the third group includes at least one third sub-block, and so on, until packet processing of ² sub-blocks is completed . Optionally, in the process of grouping m ² sub-blocks, grouping may be performed according to whether the positional relationship of the extracted pixels is adjacent, for example, dividing different sub-blocks composed of pixels adjacent to the same pixel into the same group.

 The following description will be made with reference to examples. FIG. 7b is a schematic diagram of a macroblock downsampling decomposition according to a seventh embodiment of the present invention. As shown in Fig. 7b, this embodiment is equivalent to dividing the block 3 into four sub-blocks of 4 x 4 size, i.e., sub-blocks 30, 31, 32 and 33, on the basis of that shown in Fig. 3c. The 8×8 size block 3 is downsampled to obtain 4 4×4 size sub-block implementation methods, and the 16×16 size macroblock is downsampled into 4 8×8 size blocks. The method is the same. For details, refer to the description in step 32, and details are not described herein again.

 In the sub-blocks 30, 31, 32, and 33, the sub-blocks may be divided into groups according to the positional relationship of the extracted pixels in the macroblock downsampling process, for example, the sub-blocks 33 are a group, and the sub-blocks 31 and 32 are one. Group, sub-block 30 is a group.

In the following encoding process for each sub-block or block, sub-block 33 encodes the predictive coding type as intra-prediction, and the predictive coding types for sub-blocks 30, 31 and 32 and blocks 0, 1 and 2 are between pixels. Predictive coding. The "I" and "P" in the figure respectively indicate that the prediction coding type of the corresponding block is: intra prediction coding or inter-pixel prediction coding type. Therefore, in this embodiment, the sub-block 33 is the first embodiment of the present invention. 1 sub-blocks, sub-blocks 31 and 32 are "second sub-blocks" according to an embodiment of the present invention, and sub-block 30 is a "third sub-block" according to an embodiment of the present invention.

 In an embodiment of the invention, the pixels in the first block include at least pixels in the lower right corner of the macroblock. The pixel in the lower right corner of the second block is in the horizontal or vertical direction from the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.

 Of course, in other optional embodiments, the pixels in the first block may include at least pixels in the upper left corner of the macroblock. The pixel in the upper left corner of the second block is in the horizontal or vertical direction from the pixel in the upper left corner of the first block. The pixel in the upper left corner of the third block and the pixel in the upper left corner of the first block are in the diagonal direction of the macro block.

 Step 73: Perform intra prediction encoding processing on the sub-block 33 to obtain a reconstructed image value of the sub-block 33. The intra prediction encoding processing method of the sub-block 33 in this step is similar to the method for performing intra prediction encoding processing on the block 3 in step 33, and details are not described herein again.

 Step 74: Perform prediction coding processing on the sub-blocks 31 and 32 with reference to the reconstructed image values of the sub-blocks 33, and obtain reconstructed image values of the sub-blocks 31 and 32, respectively.

 The determination of the inter-pixel predictive coding and the prediction reference relationship of the sub-blocks 31 and 32 in this step is similar to the method for predictive coding processing of the blocks 1 and 2 in step 34, and details are not described herein again. Optionally, in performing the predictive coding process on the sub-block 31 or 32, the reconstructed image value of the sub-block 33 may be interpolated, and the pixel value of the corresponding position after interpolation is used as the predicted value of the sub-block 31 or 32. The reconstructed image value of the sub-block 31 or 32 is constructed according to the predicted value of the sub-block 31 or 32 and its prediction residual, and the specific implementation thereof is similar to the method for the interpolation processing of the blocks 1 and 2 in step 34, and is no longer Narration.

Step 75: Perform prediction encoding processing on the sub-block 30 with reference to the reconstructed image value of the sub-block 31 and/or the sub-block 32 to obtain a reconstructed image value of the sub-block 30. The determination of the inter-pixel predictive coding and the prediction reference relationship of the sub-block 30 in this step is similar to the method of predictive coding processing for the block 0 in the step 35, and details are not described herein again. Optionally, in performing the predictive coding process on the sub-block 30, the reconstructed image values of the sub-blocks 31 and/or 32 may be subjected to interpolation processing, and the pixel values of the corresponding positions after interpolation are used as the predicted values of the sub-block 30, The method for constructing the reconstructed image of the sub-block 30 according to the predicted value of the sub-block 30 and its prediction residual is similar to the method for the interpolation processing of the block 0 in step 35, and details are not described herein again.

 Step 76: Synthesize the reconstructed image values of the sub-blocks 30, 31, 32, and 33 to obtain the reconstructed image values of the block 3.

 Step 77-Step 78 is the same as Step 34-Step 35, and details are not described herein again.

 Step 79: Write coded bit information of each block obtained by encoding the prediction residuals of the sub-blocks 30, 31, 32, and 33, and blocks 0, 1, and 2 into the code stream.

 When the present embodiment is applied to a scenario of a block-based hybrid video coding framework, the predictive coding process for each sub-block or each block includes performing prediction, transform, quantization, and entropy coding, etc., according to steps 71-79 above. It can be seen that each sub-block and each block form a cascade prediction reference relationship. Due to the introduction of cascading predictive coding, there may be two different predictive coding types in the macroblock, such as intra prediction and inter-pixel prediction. There may be two different pre-J codes in the intra-pre and inter-pixel prediction. The sub-blocks of the type can be optimized during the quantization process of the block. For example, the basic principle of assigning QP values to each sub-block or block is: Determine the QP value according to the reference relationship. In the above encoding process, there are the following reference relationships: block 0 refers to block 1/2, block 1/2 refers to block 3, in block 3, sub-block 30 refers to sub-block 31/32, sub-block 31/32 refers to sub-block 33. Similar to step 362, that is: the QP value of the block being referenced is given a smaller QP value.

According to the above principle, as shown in the sub-block and the reference relationship between the blocks as shown in FIG. 7b, the QP values of the sub-blocks and the blocks satisfy the following formula: QP (sub-block 33) ^QP (sub-block 31/32) ^QP (sub-block 30) ^QP (block 1/2) ^QP (block 0)

 In practical applications, the QP value of each block can be determined experimentally. For example, the following relationship can be sampled to determine the sub-block and the QP value of the block:

 QP (block 1/2) = QP (block 0) = QP (Slice)

 QP (subblock 31/32) = QP (subblock 30) = QP (Slice) -6

 QP (subblock 33) = QP (Slice) -10

 Among them, QP (Slice) represents the QP value corresponding to the slice of the video image, and the QP value corresponding to the slice can be obtained in advance.

 After determining the QP values of the sub-blocks 30, 31, 32, and 33 and the blocks 0, 1, and 2, the respective prediction residuals are subjected to corresponding quantization processing based on the QP values of the sub-blocks or blocks. A method for performing a corresponding quantization optimization process on different sub-blocks or blocks is similar to the description of the third embodiment of the present invention, and details are not described herein again.

 Due to the introduction of cascading predictive coding, there may be two blocks of different predictive coding types, such as intra prediction and inter-pixel prediction, within the macroblock, and there may be two sub-blocks of different prediction coding types, such as intra prediction and inter-pixel prediction. Therefore, the entropy coding of the macroblocks may be optimized. The entropy coding mode of the syntax element CBP is optimized. For details, refer to the description of the third embodiment of the present invention, and details are not described herein.

The syntax element "cbp_block_flag" in the video coding standard indicates whether four 4x4 sub-blocks in one 8x8 block contain non-zero coefficients. Due to the introduction of cascading predictive coding, a block of 8×8 size may have two sub-blocks of different predictive coding types within and between pixels, and the prediction residuals of sub-blocks of different predictive coding types are quantized. The coefficients have different distribution rules. In view of this, the entropy coding mode of the syntax element "cbp_block_flag" needs to be designed accordingly. The design method is similar to the optimization design method of the entropy coding mode of the syntax element CBP in step 363. This will not be repeated here. In the video image encoding processing method of this embodiment, on the basis of implementing the technical effects of the third embodiment of the present invention, a plurality of sub-blocks (ie, sub-blocks 30, 31, 32, and 33) are obtained by sub-sampling the block 3 again, and the sub-blocks are obtained. Different layers such as layers and block layers respectively establish respective corresponding cascading prediction reference relationships, which is equivalent to forming two levels of cascading coding processing. In an application scenario of encoding processing of a video image with less texture details, the embodiment can more fully utilize the correlation between pixels, thereby improving prediction accuracy and encoding compression efficiency of the video image.

 It should be noted that, although the macroblock of the size of 16 16 in this embodiment performs two levels of downsampling decomposition as an example, it can be understood that it can be based on the actual needs of video image encoding, such as less rich in texture details. In the application scenario of the encoding process of the video image, three or more levels of downsampling decomposition may be performed on the basis of the following, for example: sub-blocking decomposition in the sub-block 33 of 4×4 size, thereby forming three levels of The cascading coding process is similar to the embodiment, and will not be described again.

 FIG. 8 is a flowchart of a video image decoding processing method according to an eighth embodiment of the present invention. This embodiment is a decoding end processing flow corresponding to the encoding processing method provided by the seventh embodiment of the present invention. As shown in FIG. 8, the video image decoding processing method in this embodiment includes:

 Step 81: Acquire a code stream of the sub-blocks 30, 31, 32, and 33 included in the macro block of the video image frame and the coded bit information of the blocks 0, 1, and 2.

 Step 82: Decode the coded bit information of the sub-blocks 30, 31, 32, and 33 and the blocks 0, 1, and 2, respectively, to obtain residual decoding coefficients of each sub-block or block.

 The decoding process of the encoded bit information of any sub-block or block may include: entropy decoding, inverse quantization, and inverse transform, etc., which is equivalent to the inverse processing of step 79. Optionally, step 82 may include step 821 - step 823.

Step 821, performing entropy decoding processing on the coded bit information of the sub-blocks 30, 31, 32, and 33 and the blocks 0, 1, and 2. In the entropy decoding process, according to the distribution rule of the residual decoding coefficients of each block, the encoding bit information of each block is subjected to corresponding entropy decoding processing, and the block using the intra prediction coding mode for predictive coding has non-zero coefficients. The probability of the block larger than the predictive coding using the inter-plot predictive coding mode, such as: the probability that the sub-block 33 residual decoding coefficient has a non-zero coefficient is greater than other sub-blocks 30, 31, and 32 or block 0. The residual decoding coefficients of 1 and 2 have a probability of non-zero coefficients. The entropy decoding implementation mechanism is similar to the steps 421 to 423, and details are not described herein.

 Step 822: Perform inverse quantization processing on the coefficients of the sub-blocks 30, 31, 32, and 33 and the coded bit information of the blocks 0, 1, and 2 subjected to entropy decoding processing.

 Since the quality of the referenced block affects the quality of the block with which it is referenced, in order to obtain better RD performance, the QP value assignment principle for each sub-block and each block is: Determine the QP value based on the reference relationship. The reference relationship existing in the subsequent decoding process may be determined in advance, for example, there is a following reference relationship: block 0 refers to block 1/2, block 1/2 refers to block 3, and in block 3, subblock 0 refers to sub-block 31/32, sub-block 31/32 refers to sub-block 33. The QP value of the block to be referenced is given a smaller QP value.

 According to the above principle, the QP value of the available block satisfies the following formula:

 QP (sub-block 33) ^QP (sub-block 31/32) ^QP (sub-block 30) ^QP (block 1/2) ^QP

(block 0)

 In practical applications, the QP value of each block can be determined experimentally. For example, the following relationship can be sampled to determine the QP value of each block:

 QP (block 1/2) = QP (block 0) = QP (Slice)

 QP (subblock 31/32) = QP (subblock 30) = QP (Slice) -6

 QP (subblock 33) = QP (Slice) -10

Where QP (Slice) represents the QP value corresponding to the slice of the video image, and the QP value corresponding to the slice can be obtained in advance. After determining the QP value of each block, according to the QP value of each block, the coded bit information of each block is subjected to corresponding inverse quantization processing by the entropy-decoded coefficients.

 Step 823: Perform inverse transform processing on the coefficients of the sub-blocks 30, 31, 32, and 33 and the blocks 0, 1, and 2 after the inverse quantization to obtain a residual.

 Step 83: Perform intra prediction decoding processing on the sub-block 33, and add the predicted value to the residual of the sub-block 33 to obtain a decoded image value of the sub-block 33.

 Step 84: Perform prediction decoding processing on the sub-blocks 31 and 32 respectively with reference to the decoded image values of the sub-blocks 33, and obtain predicted values of the sub-blocks 31 and 32, and predict the values of the sub-blocks 31 and 32 and their respective residuals. The difference is added to obtain decoded image values of the sub-blocks 31 and 32.

 Step 85: Perform prediction decoding processing on the sub-block 30 with reference to the decoded image values of the sub-blocks 31 and/or 32, obtain a prediction value of the sub-block 30, and add the predicted value of the sub-block 30 and its residual to obtain a sub- The decoded image value of block 30.

 Step 86: Perform upsampling synthesis processing on the decoded image values of the respective sub-blocks 30, 31, 32, and 33 to obtain decoded image values of the block 3.

 Steps 87 to 89 are the same as steps 44 to 46, and are not described here.

The video image encoding processing method of this embodiment implements the technical effects of the fourth embodiment of the present invention, and the encoding end desamples the block 3 again to obtain a plurality of sub-blocks (ie, sub-blocks 30, 31, 32, and 33), and the decoding end. The respective cascading prediction reference relationships are established for different layers such as the sub-block layer and the block layer, which is equivalent to forming two levels of cascading decoding processing. In an application scenario of video image processing with less texture details, this embodiment can implement decoding of a video image after encoding and compressing fully utilizing pixel correlation. If the encoding end performs three or more levels of downsampling decomposition on the macroblock, the decoding end performs the implementation of the cascading decoding processing of three or more levels, which is similar to the embodiment, and is not Let me repeat. FIG. 9 is a schematic structural diagram of a video image encoding processing apparatus according to a ninth embodiment of the present invention. As shown in FIG. 9, the video image encoding processing apparatus of this embodiment includes: a downsampling decomposition module 91, an intra prediction encoding module 92, a first inter-pixel predictive encoding module 93, and a second inter-pixel predictive encoding module 94.

 The downsampling decomposition module 91 is configured to downsample the macroblock of the video image frame to obtain a plurality of blocks, and the plurality of blocks include at least: a first block, a second block, and a third block. The pixel in the first block contains at least the pixel in the lowermost corner of the macroblock. The pixel in the lowermost corner of the second block is in the horizontal or vertical direction from the pixel in the lowermost corner of the first block. The pixel in the lowermost corner of the third block is diagonal to the pixel in the lowermost corner of the first block.

 The intra prediction encoding module 92 is configured to perform intra prediction encoding processing on the first block to obtain a reconstructed image value of the first block.

 The first inter-pixel predictive coding module 93 is configured to perform predictive coding on the second block with reference to the reconstructed image value of the first block to obtain a reconstructed image value of the second block.

 The second inter-pixel predictive coding module 94 is configured to perform predictive coding processing on the third block with reference to the reconstructed image value of the second block.

 Based on the above technical solutions, in order to improve the prediction accuracy, an optional:

 The first inter-pixel predictive coding module 93 may be further configured to perform interpolation processing on the reconstructed image values of the first block, such as: performing interpolation processing on the reconstructed image values of the first block, and interpolating the pixel values of the corresponding positions as the second block. The predicted value constructs the reconstructed image value of the second block from the predicted value of the second block and its prediction residual.

 The second inter-pixel predictive coding module 94 may be further configured to perform interpolation processing on the reconstructed image values of the second block, such as performing interpolation processing on the reconstructed image values of the second block, and interpolating the pixel values of the corresponding positions as the third block. The predicted value constructs the reconstructed image value of the third block from the predicted value of the third block and its prediction residual.

In this embodiment, the video image coding processing apparatus performs down-sampling decomposition on the macroblock to obtain a plurality of blocks, and combines the intra-prediction coding technology and the inter-pixel prediction coding technology in performing predictive coding on the plurality of blocks. At least two levels of cascading prediction reference relationships are formed in the inter-pixel predictive coding process, so that correlation between pixels can be fully utilized, and prediction accuracy and coding compression efficiency of video images are improved. For the implementation mechanism of the video image encoding processing apparatus of this embodiment, refer to the description of the foregoing video image encoding processing method embodiment, and details are not described herein again.

 FIG. 10 is a schematic structural diagram of a video image decoding processing apparatus according to a tenth embodiment of the present invention. As shown

As shown in FIG. 10, the video image decoding processing apparatus of this embodiment includes: a bit information acquiring module 101, an intra prediction decoding module 102, a first inter-pixel predictive decoding module 103, a second inter-pixel predictive decoding module 104, and a decoded image generating module 105. .

 The bit information obtaining module 101 is configured to obtain coded bit information of a macroblock of a video image frame, where the macroblock includes at least: a first block, a second block, and a third block.

 In an embodiment of the invention, the pixels in the first block include at least pixels in the lower right corner of the macroblock. The pixel in the lower right corner of the second block is in the horizontal or vertical direction from the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.

 Of course, in other optional embodiments, the pixels in the first block may include at least pixels in the upper left corner of the macroblock. The pixel in the upper left corner of the second block is in the horizontal or vertical direction from the pixel in the upper left corner of the first block. The pixel in the upper left corner of the third block and the pixel in the upper left corner of the first block are in the diagonal direction of the macro block.

 The intra prediction decoding module 102 is configured to perform intra prediction decoding processing on the coded bit information of the first block to obtain a decoded image value of the first block.

 The first inter-pixel predictive decoding module 103 is configured to perform predictive decoding processing on the coded bit information of the second block with reference to the decoded picture value of the first block to obtain a decoded picture value of the second block.

The second inter-pixel prediction decoding module 104 is configured to use the decoded image value of the second block as a reference, for the third The coded bit information of the block is subjected to predictive decoding processing to obtain a decoded image value of the third block.

 The decoded image generation module 105 is configured to perform upsampling synthesis processing on the decoded image values of the blocks in the macroblock to obtain decoded image values of the macroblock.

 Based on the above technical solutions, in order to improve the prediction accuracy, an optional:

 The first inter-pixel prediction decoding module 103 is further configured to perform interpolation processing on the decoded image values of the first block, for example, performing interpolation processing on the reconstructed image values of the first block, and the pixel values of the corresponding positions after interpolation are used as the second block. The predicted value constructs the decoded image value of the second block from the predicted value of the second block and its residual decoding coefficient.

 The second inter-pixel predictive decoding module 104 can also be used to interpolate the decoded image values of the second block. For example, the reconstructed image value of the second block is subjected to interpolation processing, and the pixel value of the corresponding position after interpolation is used as the predicted value of the third block, and the decoded image of the third block is constructed based on the predicted value of the third block and its residual decoding coefficient. value.

 In this embodiment, the video image decoding processing apparatus combines the intra prediction decoding technology with the inter-pixel prediction decoding technology by performing downsampling decomposition on the macroblock to obtain a plurality of blocks, and forms at least two in the inter-pixel prediction decoding process. The hierarchical cascading predicts the reference relationship, thus facilitating the full utilization of pixel correlation between blocks to achieve reconstruction of the encoded compressed video image. For the implementation mechanism of the video image decoding processing apparatus of this embodiment, reference may be made to the above description of the video image decoding processing method embodiment, and details are not described herein again.

 FIG. 11 is a schematic structural diagram of a video image codec system according to an eleventh embodiment of the present invention. As shown in FIG. 11, the video image encoding and decoding system of this embodiment includes: a video image encoding processing device 111 and a video image decoding processing device 112.

The video image encoding processing device 111 is configured to downsample the macroblock of the video image frame to obtain a plurality of blocks, where the plurality of blocks include at least: a first block, a second block, and a third block; performing intraframe on the first block Predicting the encoding process to obtain the reconstructed image value of the first block; performing the predictive encoding process on the second block with reference to the reconstructed image value of the first block to obtain the reconstructed image value of the second block; The reconstructed image value is a reference, and the third block is subjected to predictive coding processing.

 The video image decoding processing device 112 is configured to acquire coded bit information of the macro block, wherein the macro block includes at least: the first block, the second block, and the third block; and performing intra prediction decoding on the coded bit information of the first block. Processing, obtaining a decoded image value of the first block; performing predictive decoding processing on the encoded bit information of the second block with reference to the decoded image value of the first block to obtain a decoded image value of the second block; decoding by the second block The image value is a reference, and the coded bit information of the third block is subjected to predictive decoding processing to obtain a decoded image value of the third block; and the decoded image value of each block in the macro block is subjected to upsampling synthesis processing to obtain a decoded image of the macroblock. value.

 In an embodiment of the invention, the pixels in the first block include at least pixels in the lower right corner of the macroblock. The pixel in the lower right corner of the second block is in the horizontal or vertical direction from the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.

 Of course, in other optional embodiments, the pixels in the first block may include at least pixels in the upper left corner of the macroblock. The pixel in the upper left corner of the second block is in the horizontal or vertical direction from the pixel in the upper left corner of the first block. The pixel in the upper left corner of the third block and the pixel in the upper left corner of the first block are in the diagonal direction of the macro block.

In the video image coding and decoding system of the embodiment of the present invention, the macroblock is downsampled and decomposed to obtain a plurality of blocks, and the intra prediction technique and the interpixel prediction technology are combined in the process of predictive coding or decoding of the plurality of blocks to form at least The second-level cascading prediction reference relationship can make full use of the correlation between pixels, improve the prediction accuracy and the coding compression efficiency of video images. For the detailed structure of the video image encoding processing apparatus of this embodiment, refer to the description of the corresponding embodiment of FIG. 9, the video image decoding processing apparatus refines the structure. For the description of the corresponding embodiment of FIG. 10, the mechanism for implementing video image coding and decoding can be referred to the description of the corresponding embodiment in FIG. 1 to FIG. 8 , and details are not described herein again.

 It will be understood by those of ordinary skill in the art that the drawings are only a schematic representation of one embodiment, and the modules or processes in the drawings are not necessarily required to practice the invention.

 It will be understood by those skilled in the art that the modules in the apparatus in the embodiments may be distributed in the apparatus of the embodiment according to the embodiment, or may be correspondingly changed in one or more apparatuses different from the embodiment. The modules of the above embodiments may be combined into one module, or may be further split into a plurality of sub-modules.

 The serial numbers of the embodiments of the present invention are merely for the description, and do not represent the advantages and disadvantages of the embodiments.

 A person skilled in the art can understand that all or part of the steps of implementing the above method embodiments may be completed by using hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed. The foregoing steps include the steps of the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

 It should be noted that the above embodiments are only for explaining the technical solutions of the present invention, and are not intended to be limiting; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that: The technical solutions described in the foregoing embodiments are modified, or the equivalents of the technical features are replaced by the equivalents of the technical solutions of the embodiments of the present invention.

Claims

Rights request  A video image encoding processing method, comprising:  Downsampling the macroblock of the video image frame to obtain a plurality of blocks, the plurality of blocks including at least: a first block, a second block, and a third block;  Performing intra prediction encoding processing on the first block to obtain a reconstructed image value of the first block; performing prediction encoding processing on the second block with reference to the reconstructed image value of the first block; Obtaining a reconstructed image value of the second block;  The third block is subjected to predictive coding processing with reference to the reconstructed image value of the second block. 2. The video image encoding processing method according to claim 1, wherein the predictive encoding processing on the second block is performed by using the reconstructed image value of the first block, and the method includes: 1 block of reconstructed image values are interpolated;  The predictive coding process is performed on the third block with reference to the reconstructed image value of the second block, and includes: performing interpolation processing on the reconstructed image value of the second block.  The video image encoding processing method according to claim 1, wherein the predictive encoding process is performed on any one of the first block, the second block, and the third block, including:  Transforming the prediction residuals of each block;  Encoding the transformed residuals of the prediction residuals of the respective blocks according to the quantization parameter values of the blocks, wherein the quantization parameter values of the first block are less than or equal to the quantization parameter values of the second block, and the The quantization parameter value of the second block is less than or equal to the quantization parameter value of the third block;  Entropy coding processing is performed on the quantized coefficients of each block. The video image encoding processing method according to claim 1, further comprising: performing, according to different distribution rules of prediction residual coefficients of each of the plurality of blocks, each of the plurality of blocks Whether or not the indication information including the non-zero coefficient is subjected to entropy coding processing, and the first block includes non-zero coefficients The probability of being greater than the probability that the second or third block contains non-zero coefficients.  The video image encoding processing method according to claim 1, wherein the performing the intra prediction encoding process on the first block to obtain the reconstructed image value of the first block comprises:  Dividing the pixels of the first block into a plurality of sub-blocks, performing intra prediction encoding processing on each sub-block, obtaining reconstructed image values of the respective sub-blocks, and synthesizing the reconstructed image values of the respective sub-blocks to obtain a The reconstructed image value of the first block.  The video image encoding processing method according to claim 1, wherein the performing the intra prediction encoding process on the first block to obtain the reconstructed image value of the first block comprises:  Downsampling the first block to obtain a plurality of sub-blocks, where the plurality of sub-blocks at least include: a first sub-block, a second sub-block, and a third sub-block;  Performing intra prediction encoding processing on the first sub-block to obtain a reconstructed image value of the first sub-block; and performing the second sub-block on the reference to the reconstructed image value of the first sub-block Predicting an encoding process to obtain a reconstructed image value of the second sub-block;  And performing predictive coding processing on the third sub-block with reference to the reconstructed image value of the second sub-block to obtain a reconstructed image value of the third sub-block;  Performing upsampling synthesis processing on the reconstructed image values of the sub-blocks of the plurality of sub-blocks to obtain reconstructed image values of the first block.  The video image encoding processing method according to claim 6, wherein performing predictive encoding processing on any one of the first sub-block, the second sub-block, and the third sub-block includes:  Performing transformation processing on prediction residuals of each sub-block; Encoding the transformed residuals of the prediction sub-blocks according to the quantization parameter values of the sub-blocks, and the quantization parameter values of the first sub-block are less than or equal to the quantization parameter values of the second sub-block And the quantization parameter value of the second sub-block is less than or equal to the quantization parameter value of the third sub-block; Entropy coding processing is performed on the quantized coefficients of each sub-block.  The video image encoding processing method according to claim 6, further comprising: predicting a residual residual for each sub-block according to a distribution rule in which the prediction residuals of the respective sub-blocks are quantized by different coefficients After the quantization process, the corresponding entropy coding process is performed, and the probability of occurrence of the non-zero coefficient of the first sub-block is greater than the probability of occurrence of the non-zero coefficient of the second sub-block or the third sub-block.  The video image encoding processing method according to claim 1, wherein, in the process of downsampling the macroblock, the method further comprises: anti-aliasing part or all of the pixels of the macroblock Filter processing.  10. The video image encoding method according to claim 1, wherein the pixels in the first block include at least pixels in a lower right corner of the macroblock. The pixel in the lower right corner of the second block is in a horizontal or vertical direction with the pixel in the lower right corner of the first block. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.  A video image decoding processing method, comprising:  Obtaining coded bit information of a macroblock of a video image frame, the macroblock including at least: a first block, a second block, and a third block;  Performing intra prediction decoding processing on the coded bit information of the first block to obtain a decoded image value of the first block;  Referring to the decoded image value of the first block, performing prediction decoding processing on the encoded bit information of the second block to obtain a decoded image value of the second block;  Deriving and decoding the coded bit information of the third block with reference to the decoded picture value of the second block to obtain a decoded picture value of the third block; The decoded image values of the blocks in the macroblock are subjected to upsampling synthesis processing to obtain decoded image values of the macroblocks. The video image decoding processing method according to claim 11, wherein the predicting and decoding the second block by referring to the decoded image value of the first block includes: The decoded image values of the block are subjected to interpolation processing;  The predictive decoding process is performed on the third block with reference to the decoded image value of the second block, and includes: performing interpolation processing on the decoded image value of the second block.  The video image decoding processing method according to claim 11, wherein the predictive decoding process is performed on any one of the first block, the second block, and the third block, including:  Entropy decoding processing on coded bit information of each block;  Performing inverse quantization processing on the coefficients after the entropy decoding process according to the quantization parameter values of the blocks, wherein the quantization parameter value of the first block is less than or equal to the quantization parameter value of the second block, and the quantization of the second block The parameter value is less than or equal to the quantization parameter value of the third block;  The inverse transform processing is performed on the inverse quantization process.  The video image decoding processing method according to claim 11, further comprising: performing corresponding entropy decoding on whether each block includes non-zero coefficient indication information according to different distribution rules of residual coefficients of each block Processing, the probability that the first block has a non-zero coefficient is greater than the probability that the second block or the third block has a non-zero coefficient.  The video image decoding processing method according to claim 11, wherein the first block includes a plurality of sub-blocks, and the intra-prediction decoding process is performed on the coded bit information of the first block, including: The coded bit information of the sub-block is decoded to obtain a decoded image value of each sub-block; and the decoded image values of the respective sub-blocks are combined to obtain a decoded image value of the first block. The video image decoding processing method according to claim 11, wherein the first block includes at least: a first sub-block, a second sub-block, and a third sub-block; and the encoding of the first block The bit information is subjected to an intra prediction decoding process, and the decoded image value of the first block is obtained, including: Performing intra prediction decoding processing on the first sub-block to obtain a decoded image value of the first sub-block; performing intra-prediction decoding processing on the decoded image value of the first sub-block to obtain a second sub-block The decoded image value of the block;  Performing intra prediction decoding processing with reference to the decoded image value of the second sub-block to obtain a decoded image value of the third sub-block;  The decoded image values of the respective sub-blocks are subjected to up-sampling synthesis processing to obtain decoded image values of the first block.  The video image decoding processing method according to claim 16, wherein performing predictive decoding processing on any one of the first sub-block, the second sub-block, and the third sub-block includes:  Entropy decoding processing on coded bit information of each sub-block;  Performing inverse quantization processing on the coefficients after the entropy decoding process according to the quantization parameter values of the respective sub-blocks, wherein the quantization parameter value of the first sub-block is less than or equal to the quantization parameter value of the second sub-block, and the second parameter The quantization parameter value of the sub-block is less than or equal to the quantization parameter value of the third sub-block;  The inverse transform processing is performed on the inverse quantization process.  The video image decoding processing method according to claim 16, further comprising: performing corresponding information on whether each sub-block includes non-zero coefficients according to different distribution rules of residual coefficients of each sub-block In the entropy decoding process, a probability that the first sub-block exhibits a non-zero coefficient is greater than a probability that the second sub-block or the third sub-block exhibits a non-zero coefficient.  The video image decoding processing method according to claim 11, wherein the pixels in the first block include at least pixels in a lower right corner of the macroblock. The pixel in the lower right corner of the second block and the pixel in the lower right corner of the first block are in a horizontal or vertical direction. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block. 20. A video image encoding processing apparatus, comprising: a downsampling decomposition module, configured to perform downsampling on a macroblock of a video image frame to obtain a plurality of blocks, where the plurality of blocks include at least: a first block, a second block, and a third block;  An intra prediction coding module, configured to perform intra prediction coding processing on the first block to obtain a reconstructed image value of the first block;  a first inter-pixel predictive coding module, configured to perform predictive coding processing on the second block with reference to a reconstructed image value of the first block, to obtain a reconstructed image value of the second block;  The second inter-pixel predictive coding module is configured to perform predictive coding processing on the third block with reference to the reconstructed image value of the second block.  The video image encoding processing device according to claim 20, wherein the first inter-pixel predictive encoding module is further configured to perform interpolation processing on the reconstructed image values of the first block;  The second inter-pixel predictive coding module is further configured to perform interpolation processing on the reconstructed image values of the second block.  The video image encoding processing device according to claim 20, wherein the pixels in the first block include at least pixels in a lower right corner of the macroblock. The pixel in the lower right corner of the second block and the pixel in the lower right corner of the first block are in a horizontal or vertical direction. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.  A video image decoding processing apparatus, comprising:  a bit information acquiring module, configured to acquire coded bit information of a macroblock of a video image frame, where the macroblock includes at least: a first block, a second block, and a third block;  An intra prediction decoding module, configured to perform intra prediction decoding processing on the encoded bit information of the first block to obtain a decoded image value of the first block; a first inter-pixel prediction decoding module, configured to use the decoded image value of the first block as a reference, The coded bit information of the second block is subjected to predictive decoding processing to obtain a decoded image value of the second block; and the second inter-pixel predictive decoding module is configured to reference the decoded image value of the second block to the third The coded bit information of the block is subjected to predictive decoding processing to obtain a decoded image value of the third block;  And a decoded image generating module, configured to perform upsampling synthesis processing on the decoded image values of the blocks in the macroblock to obtain decoded image values of the macroblock.  The video image decoding processing device according to claim 23, wherein the first inter-pixel predictive decoding module is further configured to perform interpolation processing on the decoded image values of the first block;  The second inter-pixel predictive decoding module is further configured to perform interpolation processing on the decoded image values of the second block.  The video image decoding processing device according to claim 23, wherein the pixels in the first block include at least pixels in a lower right corner of the macroblock. The pixel in the lower right corner of the second block and the pixel in the lower right corner of the first block are in a horizontal or vertical direction. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.  26. A video image encoding and decoding system, comprising:  a video image encoding processing apparatus, configured to downsample a macroblock of a video image frame to obtain a plurality of blocks, where the plurality of blocks include at least: a first block, a second block, and a third block; Performing intra prediction coding processing to obtain a reconstructed image value of the first block; performing prediction encoding processing on the second block with reference to the reconstructed image value of the first block to obtain the second block a reconstructed image value; performing predictive encoding processing on the third block with reference to the reconstructed image value of the second block; a video image decoding processing apparatus, configured to acquire coded bit information of the macroblock, where the macroblock includes at least: the first block, the second block, and the third block; and the first block Encoding bit information is subjected to intra prediction decoding processing to obtain a decoded image value of the first block; and a solution to the first block The code image value is a reference, and the coded bit information of the second block is subjected to predictive decoding processing to obtain a decoded image value of the second block; and the decoded image value of the second block is referred to, and the third block is referred to The coded bit information is subjected to predictive decoding processing to obtain a decoded image value of the third block; and the decoded image value of each block in the macro block is subjected to upsampling synthesis processing to obtain a decoded image value of the macroblock.  27. The video image encoding and decoding system according to claim 26, wherein the pixels in the first block include at least pixels in a lower right corner of the macroblock. The pixel in the lower right corner of the second block and the pixel in the lower right corner of the first block are in a horizontal or vertical direction. The pixel in the lower right corner of the third block and the pixel in the lower right corner of the first block are in the diagonal direction of the macro block.