CN103929641A - Intra-frame encoding method based on virtual reference frame - Google Patents
Intra-frame encoding method based on virtual reference frame Download PDFInfo
- Publication number
- CN103929641A CN103929641A CN201410196016.4A CN201410196016A CN103929641A CN 103929641 A CN103929641 A CN 103929641A CN 201410196016 A CN201410196016 A CN 201410196016A CN 103929641 A CN103929641 A CN 103929641A
- Authority
- CN
- China
- Prior art keywords
- frame
- prediction
- virtual reference
- mode
- reference frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention provides an intra-frame encoding method based on a virtual reference frame. An HEVC standard prediction mode is used for encoding CUs in an intra-frame video boundary, and a mode that virtual reference frame prediction and standard prediction are combined is used for other CUs. The method comprises the steps that first, a virtual reference frame of a block to be encoded is constructed, the range of the virtual reference frame comprises reconstruction data with an encoded area and information with an unencoded area, and the data in the unencoded area are filled in a horizontal/vertical tiling mode; secondary, prediction based on the virtual reference frame is conducted on the CUs to be encoded, a quick diamond search method is used for the prediction mode, if the residual error of prediction is small enough, the block is marked to be in an INTER mode, and the value of the motion vector of the block is recorded; if the motion vector is zero, the block is marked as a SKIP block; if the residual error of prediction is big, an HEVC standard prediction mode is used, and the block is marked as an INTRA block. Through testing, the intra-frame encoding method based on the virtual reference frame can effectively improve the prediction precision of intra-frame encoding.
Description
Technical field
The present invention relates to a kind of new inner frame coding method, relate in particular to a kind of intraframe coding based on virtual reference frame, belong to technical field of computer vision.
Background technology
Improving constantly of computing power, the network bandwidth and mobile terminal disposal ability, has proposed the new demand of the aspect such as speed, performance to Video processing.In order to improve constantly compression performance, ITU and ISO tissue have been released a series of video encoding standards, comprise the H.26x series of ITU tissue and the MPEG-x series standard of ISO tissue, and the HEVC standard of up-to-date formulation.Up-to-date HEVC standard be devoted to be satisfied with user 1) high definition, 2) 3D, 3) mobile wireless, to meet the demand of the frontiers such as new home theater, remote monitoring, digital broadcasting, mobile flow medium, mobile imaging, medical imaging.In HEVC standard, there is various configurations pattern, comprise HE (High Efficiency) high-performance, the configuration of LC (Low Complexity) low complex degree.
The basic coding framework that these coding standards adopt is identical, adopts the basic framework of motion compensation+DCT more.Under this framework, frame of video is generally divided into l (Intra-frame), P (Predictive-frame), B (BidirectionaIIy predicted-frame) three types: l frame is in the time of prediction,, with reference to the information of other frame, do not complete coding by processes such as conversion, quantifications; P frame is taking the reconstruction image of forward direction coded frame as reference, carries out coded residual after motion compensation; Taking front and back, the reconstruction image of two-way coded frame is reference to B frame, carries out coded residual after motion compensation.
In the frame of video of three types, I frame is key frame, plays the fault-tolerant effect of code stream.In I frame cataloged procedure, generally only utilize spatial coherence, so compression ratio is generally lower.Although in a video sequence, the number of l frame is fewer, its every frame coding figure place is far above P, B frame.For example, in standard H.264, the Coding Rate of I frame is 3～5 times of P frame; In HEVC standard, along with the introducing of new inter prediction technology, the compression ratio of I frame and P frame further expands, and in some video, can reach 1: 10.For the video flowing of bit rate output constant (CBR), the code check of I frame increases suddenly, will directly cause the decline of the coding figure place that reduces follow-up P/B frame, and then has influence on the quality of Recovery image and predicted picture.The compression ratio that improves l frame, has not only played vital effect for stability and the continuity of video quality.
Prediction is the method for raising compression ratio conventional in video, according to feature of image, utilizes between frame of video and inner correlation, as reference, retains less difference information taking coded object, reduces final coding figure place.How in the image of having encoded, to find the optimum Match source of current object to be predicted, remove to greatest extent redundant information, make difference information amount reach minimum value, this is the key point of prediction.Due to the effect of I frame in sequence, can not adopt the mode of P/B frame to carry out inter prediction, can only adopt infra-frame prediction, at the inner analogical object of finding of present image.Infra-frame prediction adopts " prediction of space consecutive points " more, and the also i.e. spatial correlation based on image thinks on image that each point and its space consecutive points have similitude, finds after similitude by search, carries out differential coding.
H.264 the macro block of the infra-frame prediction adopting in standard taking 16 × 16 and 4 × 4 piece, as fundamental forecasting unit, are predicted in 9 directions each point in piece.Taking 4 × 4 as example, as shown in Fig. 1 (a), a, b ..., p is current to be predicted, 17 some Q around, and l ..., P is encoded point.To a-p pointwise, along 0 shown in Fig. 1 (b), 1,, 8 and direction 2 (DC prediction) totally nine directions, get the point in P-Q-H, by suitable predictor formula calculating predicted value, do difference with crude sampling value, the pattern of difference minimum is final predictive mode, finally prediction residual coefficient is carried out to DCT coding.
In up-to-date HEVC standard, introduce CU (coding unit), PU (predicting unit), TU (converter unit), three new concepts.The macro block concept of coding unit in being similar to H.264/AVC, its big or small maximum can be 64 × 64; Predicting unit is the elementary cell of predicting; Converter unit is the elementary cell of carrying out transform and quantization.The separation of three unit, make conversion, predict and each processing links of encoding more flexible, also more meet the textural characteristics of video image.The size of predicting unit can be 4 × 4,8 × 8,16 × 16,32 × 32,64 × 64, and the varying in size of piece, the pattern of prediction is not identical yet, be respectively 17,34,34,34, and 3 kinds of patterns are optional. unified infra-frame prediction angle is: +/-[0,2,5,9,13,17,21,26,32]/32.
H.264/HEVC when standard is carried out intraframe coding, each picture element to be encoded is all adopted to multiple predictive mode, but just utilize the lastrow of to be predicted, the first from left column data, in this data area, adopt intimate exhaustive future position searching method to predict, if correlation is very poor between image consecutive points, this prediction not only can not make image data amount reduce also can introduce more redundant information because of factors such as logging mode information.
That is to say, the core of intraframe coding is: whether have enough similitudes, to improve the accuracy of prediction, directly determine the effect of prediction; How to find the position of similitude, it determines the computational complexity of whole forecasting process.
Summary of the invention
The object of this invention is to provide a kind of intra-frame predictive encoding method based on virtual reference frame.The method is by HEVC standard is analyzed, and the prediction mode while changing intraframe coding, provides the coding structure that is suitable for intraframe coding, and in this structure, will utilize more can reference data, improves the accuracy of predicting, the redundancy of minimizing data.
By HEVC standard is analyzed to discovery, in the time of intraframe coding, infra-frame prediction has only utilized the lastrow of current block, the first from left column data, and the broad regions data of its top and left are not all used.Inner frame coding method based on virtual reference frame of the present invention, utilizes the region that rebuild, adds the region of fabricating out one right-hand and below, as the reference block of current block to be encoded (or pixel to be encoded), completes infra-frame prediction.This method, in conjunction with current intra-frame prediction method, can obviously improve compression efficiency.
For achieving the above object, the present invention adopts following technical scheme.It is characterized in that comprising the following steps:
Step 1: adopt the predictive mode of HEVC standard to encode to being positioned at the coding unit CU (Coding Unit) on video border in frame;
Step 2:, remaining coding unit is carried out to prediction mode judgement, first treat prediction data and in virtual reference frame, search for coupling, the data of described virtual reference frame are by given data or the data reconstruction of coding region, and fabricate the right-hand and lower zone of out one and form, there is certain spatial coherence above-mentioned two regions and unit to be encoded, secondly, residual error between matched data and the data to be predicted that search coupling is obtained judges, if residual error is less than predetermined threshold, predict with regard to the mode that adopts virtual reference frame, and residual sum motion vector is encoded, otherwise, if residual error is greater than predetermined threshold, the predictive mode that further adopts HEVC standard is predicted, by comparing the residual error size of two kinds of prediction mode, choose a kind of mode that residual error is less and carry out infra-frame prediction, then the residual error data of prediction is quantized, entropy coding,
Step 3: before encoding code stream output, if adopt the predictive mode of HEVC standard to encode, CU type is INTRA piece; If encode based on virtual reference frame, CU type is INTER, and provides motion vector value.
In order to save the operand of decoding, a kind of special INTER Type C U is identified to the SKIP piece that motion vector is 0.
While utilizing not coded data constructing virtual reference frame, adopt the method for Horizontal Tile and vertical tile, Horizontal Tile is right side and the below to current block by the data Replica on the current block left side, and the mode of vertical tile is right side and the below to current block by the data Replica of current block top.
Intraframe prediction algorithm based on virtual reference frame provided by the present invention can overcome infra-frame prediction in original standard and only utilized the lastrow of current block, the first from left column data, the shortcoming that the broad regions data of its top and left are not all used, utilize the region having rebuild, add the region of fabricating the right-hand and below of out one, as the reference block of current block to be encoded, complete infra-frame prediction.This method, in conjunction with current intra-frame prediction method, can obviously improve compression efficiency.
Brief description of the drawings
Fig. 1 is the H.264 space consecutive points prediction in standard;
Fig. 2 is the infra-frame prediction in HEVC standard;
Fig. 3 is the relative position of each in virtual reference frame;
Fig. 4 is the coding framework based on virtual reference frame;
Fig. 5 is the intraframe coding flow process based on virtual reference frame;
Embodiment
Before address, the present invention, according to the inner frame coding method based on virtual reference frame, has realized and not having increased under the prerequisite of computational complexity, improves prediction effect, reduces the output code flow of I frame.
Below in conjunction with brief description of the drawings implementation of the present invention.
Step 1: the structure of virtual reference frame
Core of the present invention be utilize encoded and not coding region complete the infra-frame prediction to current block, the coding region that only utilizes being different from standard carries out infra-frame prediction.Be equivalent to intraframe coding algorithm and introduced the concept of estimation, from changing in essence the Forecasting Methodology of the infra-frame prediction of knowing clearly, do not adopt fixing interpolative mode and fixing angle to predict yet, but by the current to be predicted type of regarding interframe block as, in certain hunting zone, predict, if searched best matching blocks, this piece is exactly its prediction reference piece so, and then complete prediction, improve the code efficiency of I frame.Thereby the first step of the present invention is constructing virtual reference frame.
Want constructing virtual reference frame, need to utilize the region that rebuild, add the region of fabricating out one right-hand and below, as the reference frame of current block to be encoded.Fig. 3 has pointed out the position at each place, and the maximum magnitude shown in Fig. 3 is defined as the virtual reference frame of current block to be encoded Current.
The first row piece: UL is that UP Left piece is upper left piece; UP is upper piece; UR is that Up Right piece is upper right piece;
The second row piece: L is Left piece; Current is current block to be encoded; R is that Right is right;
The third line piece: DL is lower-left piece; Down is lower, and DR is bottom right piece.
BM piece is Best Match piece, be the best matching blocks that current block finds, BM1 in Fig. 3 and BM2 have represented the diverse location that best matching blocks may be located, if BM1 position, its data have been used encoded UL, UP, outside L block message, also will use the not information of encoding block Current; If BM2 position, except using encoded UP, UR, also will use the information of uncoded Current piece and R piece.
Except illustrated BM1 and two positions of BM2, the optional position in all right Fig. 3 of BM piece.Scope definition shown in Fig. 3 is the virtual reference frame of current block to be encoded Current, is also the scope that current block to be encoded will be predicted, searches for, in the hope of reaching the best infra-frame prediction effect of current block within the scope of this.Virtual reference frame refers to the size according to current to be predicted, be generally no more than ± 64 × 64 size of scope, and minimum scope can arrive ± and 8 × 8.In order to improve the precision of block search, adopt forward prediction+back forecast concept.
Forward prediction: refer to that the information that when prediction utilizes is to obtain the piece of rebuild content above, taking Fig. 3 as example, Current piece is at UL, UP, UR, finds the information of prediction of can being used in L, cannot use the content of R, Current, DL, Down, DR piece.
When back forecast refers to prediction, utilize below still uncoded information, Current piece can find the information for prediction in R, Current, DL, Down, DR piece.
The back forecast is here different from the back forecast concept of B frame, if used back forecast in the time of coding B frame, needing the order of decoded frame to adjust just in decoding end can be correctly decoded.In order not cause unnecessary trouble to decoding end, the order to encoding block is not made any adjustments.
How to utilize the still information of the piece position such as uncoded Current, R to carry out back forecast Current piece, this just becomes the emphasis of the present invention at constructing virtual reference frame.The key addressing this problem is how to construct non-existent data, and concrete grammar has 3 kinds below:
(1) in order to reduce amount of calculation, can adopt the method for Horizontal Tile and vertical tile.Horizontal Tile is exactly DL, Down, the DR piece that L piece is copied to Current, R piece and below; Vertical tile copies to UP piece Current, Down piece exactly, and by the information reproduction of UR piece to R and DR piece, by the information of the L piece DL piece that tiles.
(2) choosing appropriate information according to the texture statistics characteristic of piece copies.
(3) directly by Current, R, DL, Down, the information of DR piece fills in into 128 or other numerical value.
Above three kinds of methods respectively have pluses and minuses, and (1) kind numerical value make is simple, only relates to copying of data, but accurate not; (2) plants numerical value complex structure, need to carry out the statistics of characteristics of image, but data is relatively accurate; (3) plants numerical value simple structure, but correlation between data is weak.Consider the correlation of amount of calculation and data, the present invention will use (1) to construct not yet coded data.
Step 2: based on the forecasting process of virtual reference frame
Fig. 4 is the intraframe coding framework based on virtual reference frame, in the figure, is only that I frame is processed for frame type, and other frame type does not adopt this framework.Other frame type, still adopts the method for standard to encode.To the I frame of input, taking coding unit CU as unit, carry out virtual reference frame and predict that the mode combining with HEVC normative forecast predicts, get mode that residual error system is low as final predictive mode.To residual error coefficient convert, quantification, entropy coding exports, and carries out the output of code stream.
Fig. 5 is the intraframe coding flow process based on virtual reference frame, comprises the judgement of type of prediction and the grammer of block type mark in this flow process.
To Gather and input video sequence, determine frame type, current while being I frame, in cataloged procedure, be divided into N × N fritter, wherein i is capable, and the current block note of j row is Cur
i, j.
The first step:
to current block Cur
i, jjudge, judge that content is for being border? it is LCU level? if border or LCU level are all carried out infra-frame prediction by the method that adopts HEVC standard, record its Mode
i, j=INTRA.Judgment criterion is: if the UL piece of current block, UP piece, UR piece, L piece has one not exist, and illustrates in Cur
i, jfor boundary block; Otherwise Cur is described
i, jbe not that boundary block is not also LCU level, can search for prediction, enter next step.
Second step: create virtual reference frame, adopt in the present invention the mode of Horizontal Tile to create, the current current piece of Horizontal Tile and R piece.Setting search scope, if piece is of a size of 32 × 32, hunting zone is 64 × 64, when other block sizes, hunting zone is 32 × 32.
The 3rd step: to current block Cur
i, jpredict, the mode of prediction adopts the method for whole pixel and sub-pix estimation, and matching criterior adopts the method for sad value minimum, records its horizontal and vertical motion vector VectX, VectY;
The 4th step: judge its predictive mode according to motion vector type, predictive mode Mode
i, jbe 3 types, i.e. Inter, Intra, Skip.If sad value is less than the threshold value of setting, record VectX, VectY, with tense marker Mode
i, j=Inter.If VectX=VectY=0 is Mode
i, j=Skip piece is Mode with this block mode of tense marker
i, j=Inter.If sad value is greater than the threshold value of setting, explanation does not obtain best matching blocks, Mode
i, j=Intra, adopts the predictive mode of standard to predict.
The 5th step. enter different coding processing procedures according to different types
Mode
i, j=Skip only records its Mode in code stream
i, jinformation, need not write code stream.
Mode
i, j=Intra adopts the method for HEVC standard to predict, directly carries out conversion process, comprises Mode in code stream
i, jinformation and data message after treatment.
Mode
i, j=Inter, adopts the mode based on virtual reference frame of the present invention to predict, and residual error is carried out to conversion process, comprises Mode in code stream
i, jdata message after information and residual noise reduction and motion vector VectX, VectY information.
The 6th step: output squeezing code stream
performance evaluation
Adopt the inventive method the HM of HEVC standard is realized, tested and compare with canonical algorithm HM6.0 result.To Recovery image quality, the present invention adopts BDBR as measurement index.Configuration file: Low_delay_P_Main.cfg, low delay; Coded frame type: IPIPIP Coding frame number: totally 20 frames, 10 I frames, 10 P frames.Cycle tests is respectively CLASS A～E, wherein CLASS A be super clear video sequence " Traffic " (4096x2048p30fps), " PeopleOnStreet " (3840x2160p30fps). the sequence of the 2560x1600 of middle intercepting; CLASS B is 1920x1080p24fps: " ParkScene ", " Kimono ", 1920x108050-60fps: " Cactus ", " BasketballDrive ", " BQTerrace "; CLASS C is 832x480p30-60fps (WVGA): " BasketballDrill ", " BQMall ", " PartyScene ", " RaceHorses "; CLASS D is 416x240p30-60fps (WQVGA): " BasketballPaSS ", " BQSquare ", " BlowingBubbles ", " RaceHorses "; CLASS E is 1280x720p60fps video conferencing scenes: " Vidyo1 ", the video type resolution difference that " Vidyo3 " and " Vidyo4 " is different, frame frequency difference.
Table 1～table 3 has provided the test result in the situation of different hunting zones, and the hunting zone of table 1～3 is followed successively by 64,32, and 16.Taking table 1 as example, first classifies the other CLASS A～E of video class as, and the second～tetra-row, have on the basis of identical PSNR value luminance component Y, chromatic component U, chromatic component V, three bit rate percentages that component is saved with standard HEVC HM6.0.
Table 1 hunting zone 64, BDBR=-0.85%
| Sequence type | PSNR?Y | PSNR?U | PSNR?V |
| CLASS?A | -0.6％ | -0.9％ | -1.1％ |
| CLASS?B | -1.4％ | -2.0％ | -2.1％ |
| CLASS?C | -1.1％ | -1.6％ | -1.3％ |
| CLASS?D | -0.2％ | 0.4％ | 0.3％ |
| CLASS?E | -0.9％ | -1.6％ | -1.6％ |
Table 2 hunting zone 32, BDBR=-0.78%
| Sequence type | PSNR?Y | PSNR?U | PSNR?V |
| CLASS?A | -0.6％ | -0.9％ | -1.1％ |
| CLASS?B | -1.4％ | -2.0％ | -2.1％ |
| CLASS?C | -1.1％ | -1.6％ | -1.3％ |
| CLASS?D | -0.1％ | 0.4％ | 0.3％ |
| CLASS?E | -0.9％ | -1.5％ | -1.6％ |
Table 3 hunting zone 16, BDBR=-0.6%
| Sequence type | PSNR?Y | PSNR?U | PSNR?V |
| CLASS?A | -0.6％ | -0.9％ | -1.1％ |
| CLASS?B | -1.4％ | -2.0％ | -2.1％ |
| CLASS?C | -1.1％ | -1.6％ | -1.3％ |
| CLASSD | -0.2％ | 0.4％ | 0.3％ |
| CLASS?E | -0.9％ | -1.6％ | -1.6％ |
Table 1～table 3 has provided the BDBR value under different hunting zones, has represented under same objective quality, and the code check of two kinds of methods is saved situation.Hunting zone is larger, and BDBR value is more, has illustrated, herein, compared with the prediction algorithm of algorithm and standard, the code check of saving is more than 0.5%.
Claims (3)
1. the inner frame coding method based on virtual reference frame, is characterized in that:
Step 1: adopt the predictive mode of HEVC standard to encode to being positioned at the coding unit CU (Coding Unit) on video border in frame;
Step 2:, remaining coding unit is carried out to prediction mode judgement, first treat prediction data and in virtual reference frame, search for coupling, the data of described virtual reference frame are by given data or the data reconstruction of coding region, and fabricate the right-hand and lower zone of out one and form, there is certain spatial coherence above-mentioned two regions and unit to be encoded, secondly, residual error between matched data and the data to be predicted that search coupling is obtained judges, if residual error is less than predetermined threshold, predict with regard to the mode that adopts virtual reference frame, and residual sum motion vector is encoded, otherwise, if residual error is greater than predetermined threshold, the predictive mode that further adopts HEVC standard is predicted, by comparing the residual error size of two kinds of prediction mode, choose a kind of mode that residual error is less and carry out infra-frame prediction, then the residual error data of prediction is quantized, entropy coding,
Step 3: before encoding code stream output, if adopt the predictive mode of HEVC standard to encode, CU type is INTRA piece; If encode based on virtual reference frame, CU type is INTER, and provides motion vector value.
2. inner frame coding method as claimed in claim 1, is characterized in that: in order to save the operand of decoding, a kind of special INTER Type C U is identified to the SKIP piece that motion vector is 0.
3. inner frame coding method as claimed in claim 1, it is characterized in that: while utilizing not coded data constructing virtual reference frame, adopt the method for Horizontal Tile and vertical tile, Horizontal Tile is right side and the below to current block by the data Replica on the current block left side, and the mode of vertical tile is right side and the below to current block by the data Replica of current block top.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410196016.4A CN103929641A (en) | 2014-05-12 | 2014-05-12 | Intra-frame encoding method based on virtual reference frame |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410196016.4A CN103929641A (en) | 2014-05-12 | 2014-05-12 | Intra-frame encoding method based on virtual reference frame |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103929641A true CN103929641A (en) | 2014-07-16 |
Family
ID=51147699
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410196016.4A Pending CN103929641A (en) | 2014-05-12 | 2014-05-12 | Intra-frame encoding method based on virtual reference frame |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103929641A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016026457A1 (en) * | 2014-08-22 | 2016-02-25 | 中兴通讯股份有限公司 | Predictive coding/decoding method, corresponding coder/decoder, and electronic device |
| CN106375764A (en) * | 2016-08-31 | 2017-02-01 | 中国科学技术大学 | Directional intra prediction and block copy prediction combined video intra coding method |
| CN106791829A (en) * | 2016-11-18 | 2017-05-31 | 华为技术有限公司 | The method for building up and equipment of virtual reference frame |
| CN110072112A (en) * | 2019-03-12 | 2019-07-30 | 浙江大华技术股份有限公司 | Intra-frame prediction method, encoder and storage device |
| TWI690199B (en) * | 2016-08-31 | 2020-04-01 | 北京奇藝世紀科技有限公司 | Video coding intra prediction method and device |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105247569A (en) * | 2013-03-29 | 2016-01-13 | 谷歌公司 | Motion-compensated frame interpolation using smoothness constraints |
-
2014
- 2014-05-12 CN CN201410196016.4A patent/CN103929641A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105247569A (en) * | 2013-03-29 | 2016-01-13 | 谷歌公司 | Motion-compensated frame interpolation using smoothness constraints |
Non-Patent Citations (2)
| Title |
|---|
| SIU-LEONG YU AND CHRISTOS CHRYSAFIS: "New Intra Prediction using Intra-Macroblock Motion Compensation", 《JOINT VIDEO TEAM (JVT) OF ISO/IEC MPEG & ITU-T VCEG》 * |
| 叶迎宪等: "基于线性插值填充的帧内运动补偿", 《计算机应用》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016026457A1 (en) * | 2014-08-22 | 2016-02-25 | 中兴通讯股份有限公司 | Predictive coding/decoding method, corresponding coder/decoder, and electronic device |
| CN106375764A (en) * | 2016-08-31 | 2017-02-01 | 中国科学技术大学 | Directional intra prediction and block copy prediction combined video intra coding method |
| CN106375764B (en) * | 2016-08-31 | 2018-04-10 | 中国科学技术大学 | A kind of bonding position prediction and block replicate the video intra-frame encoding method of prediction |
| TWI690199B (en) * | 2016-08-31 | 2020-04-01 | 北京奇藝世紀科技有限公司 | Video coding intra prediction method and device |
| US10681367B2 (en) | 2016-08-31 | 2020-06-09 | Beijing Qiyi Century Science & Technology Co., Ltd. | Intra-prediction video coding method and device |
| CN106791829A (en) * | 2016-11-18 | 2017-05-31 | 华为技术有限公司 | The method for building up and equipment of virtual reference frame |
| CN106791829B (en) * | 2016-11-18 | 2020-01-21 | 华为技术有限公司 | Method and equipment for establishing virtual reference frame |
| CN110072112A (en) * | 2019-03-12 | 2019-07-30 | 浙江大华技术股份有限公司 | Intra-frame prediction method, encoder and storage device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112753218B (en) | Method and apparatus for video encoding and decoding | |
| CN105659602B (en) | Coder side option for the intra block duplication prediction mode that video and image encode | |
| CN103248893B (en) | From H.264/AVC standard to code-transferring method and transcoder thereof the fast frame of HEVC standard | |
| JP2019505144A (en) | Geometric transformation for filters for video coding | |
| JP7148612B2 (en) | Video data inter prediction method, apparatus, video encoder, video decoder and program | |
| KR100999091B1 (en) | Method and apparutus for video coding using arbitrary-size variable block | |
| CN103339938A (en) | Performing motion vector prediction for video coding | |
| CN103929641A (en) | Intra-frame encoding method based on virtual reference frame | |
| CN101888546B (en) | A kind of method of estimation and device | |
| CN113994701A (en) | Method and apparatus for motion field storage in video coding | |
| US20180131948A1 (en) | Reference image encoding method, reference image decoding method, reference image encoding device, and reference image decoding device | |
| CN113273188A (en) | Image encoding/decoding method and apparatus, and recording medium storing bit stream | |
| EP4024872A1 (en) | Video coding method and apparatus, video decoding method and apparatus, electronic device, and storage medium | |
| KR20090087767A (en) | Method for predictive intra coding for image data | |
| CN104639940A (en) | Quick HEVC (High Efficiency Video Coding) inter-frame prediction mode selection method | |
| CN117480778A (en) | Residual coding and video coding methods, devices, equipment and systems | |
| CN110913215B (en) | Method and device for selecting prediction mode and readable storage medium | |
| CN110913232B (en) | Selection method and device of TU division mode and readable storage medium | |
| CN116569549A (en) | Inter prediction method, encoder, decoder and storage medium | |
| CN113906754A (en) | Image encoding/decoding method and apparatus | |
| CN104219530A (en) | HEVC multiple-description coding scheme based on prediction mode copying | |
| CN110876058B (en) | Historical candidate list updating method and device | |
| US10764577B2 (en) | Non-MPM mode coding for intra prediction in video coding | |
| Chen et al. | CNN-based fast HEVC quantization parameter mode decision | |
| CN109302605B (en) | Image coding method and device based on multi-core processor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140716 |
|
| WD01 | Invention patent application deemed withdrawn after publication |