CN100596202C - Fast mode selection method in frame - Google Patents
Fast mode selection method in frame Download PDFInfo
- Publication number
- CN100596202C CN100596202C CN 200810301864 CN200810301864A CN100596202C CN 100596202 C CN100596202 C CN 100596202C CN 200810301864 CN200810301864 CN 200810301864 CN 200810301864 A CN200810301864 A CN 200810301864A CN 100596202 C CN100596202 C CN 100596202C
- Authority
- CN
- China
- Prior art keywords
- pattern
- distortion value
- mode
- prediction modes
- optimal prediction
- 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.)
- Expired - Fee Related
Links
- 238000010187 selection method Methods 0.000 title claims description 6
- 238000000034 method Methods 0.000 claims abstract description 17
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000009826 distribution Methods 0.000 abstract description 2
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention relates to a video coding technology, in particular to an H.264 intraframe mode selecting method. The invention provides an intraframe mode selecting method that can raise the coding speed by a relatively wide margin when signal-to-noise ratio is ensured. The fast intraframe mode selecting method comprises the steps as follows: first, whether the distortion value of an MPM mode is less than or equal to a first threshold T1 is judged; if yes, the optimal predictive mode of a current block is the MPM mode; if not, the selection range of the predictive mode is gradually narrowed according to a second threshold T2 and the directivity of the predictive mode. The invention takes the directivity of the predictive mode into full consideration, and at the same time, the selection of the threshold meets the distribution characteristics of the distortion value of different modes, thus controlling the mode selection process well. Compared with traversal algorithm, the method of the invention raises the coding speed by a relatively wide margin, with the signal-to-noise ratio being hardly reduced, and is suitable for improving the H.264 coding method.
Description
Technical field
The present invention relates to video coding technique, be specifically related to a kind of mode selection method in frame H.264.
Background technology
H.264 be that (Joint Video Team JVT) formulates by joint video team.It is a video compression standard under face the future IP and the wireless environment, and it all increases significantly than present all video compression standards aspect video compression efficiency.H.264 used traditional scheme, continued to have continued to use the outstanding technology in the former coding standard based on macro block and motion compensation.Simultaneously it also provides a series of development of novel coding toolses, comprises the estimation of multimodal infra-frame prediction, FMO (flexible modular order), 1/4 pixel precision and motion compensation etc.H.264 in the coding standard, introduced intraframe prediction algorithm, utilized the information correlativity between adjacent macroblocks in the frame fully, only the interpolation of predicted value and actual value has been encoded for the efficient that improves frame coding in the frame.But intraframe prediction algorithm has high algorithm complex, is that its coding rate becomes slower just because of the high complexity of this calculating, can't satisfy the demand of practical application under normal conditions.So not changing H.264 the standard code flow structure and keeping under the situation of original code check, find the algorithm of simple possible to replace original complicated algorithm, improve its coding rate and just become current research focus.Common method has at present: full ergodic algorithm, sobel Operator Method etc.Full ergodic algorithm travels through all available predictive modes, obtains optimum predictive mode.This method can obtain optimum predictive mode, but the feasible needs that are difficult to satisfy practical application of the amount of calculation of the high complexity of introducing.The sobel Operator Method provides in JVT minutes document JVT-G013.doc.Its implementation method is with the pixel employing sobel operator of current 4x4 piece, calculates direction vector, makes it to compare with the prediction direction of various predictive modes, and the predictive mode that direction is the most close is an optimization model.But the effect behind the actual emulation is relatively poor, and signal to noise ratio descends more, not large increase on coding rate simultaneously.
Summary of the invention
Technical problem to be solved by this invention is to provide a kind of under the situation that guarantees signal to noise ratio, the mode selection method in frame of lifting coding rate by a relatively large margin.
The present invention solves the problems of the technologies described above the technical scheme that is adopted to be, a kind of fast mode selection method in frame may further comprise the steps:
A, judge that whether the distortion value of MPM pattern is smaller or equal to first threshold T1; If the optimal prediction modes of current block is the MPM pattern; As not, enter step b;
B, calculate the distortion value of vertical pattern, horizontal pattern, DC pattern, judge in these three kinds of patterns whether a distortion value is arranged smaller or equal to the second threshold value T2; In this way, adopting this pattern is the optimal prediction modes of current block; As not, enter step c;
C, judge vertical pattern distortion value whether less than the distortion value of horizontal pattern; If enter steps d; Enter step e if not;
D, calculate the distortion value of right vertical mode and left vertical mode, and the distortion value of judging right vertical mode is whether less than the distortion value of left vertical mode; If, calculate the distortion value of bottom right diagonal model again, the pattern that distortion value in right vertical mode and the bottom right diagonal model is less is as the optimal prediction modes of current block; If not, calculate the distortion value of bottom left diagonal model again, the pattern that distortion value in left vertical mode and the bottom left diagonal model is less is as the optimal prediction modes of current block;
E, calculate the distortion value of horizontal pattern and last horizontal pattern down, and the distortion value of judging following horizontal pattern is whether less than the distortion value of last horizontal pattern; If, calculate the distortion value of bottom right diagonal model again, the pattern that distortion value in horizontal pattern and the bottom right diagonal model down is less is as the optimal prediction modes of current block; If not, calculate the distortion value of bottom left diagonal model again, the pattern that distortion value in last horizontal pattern and the bottom left diagonal model is less is as the optimal prediction modes of current block.
First threshold T1=totalT1/totalT1Num; Wherein, totalT1 is the distortion value summation that optimal prediction modes has adopted the piece of vertical pattern, horizontal pattern or DC pattern; TotalT1Num is the number that optimal prediction modes has adopted the piece of vertical pattern, horizontal pattern or DC pattern.The second threshold value T2=totalT2/totalT2Num; Wherein, totalT2: the distortion value summation of the optimal prediction modes of each encoding block; TotalT2Num is the number of piece in the cataloged procedure.
The invention has the beneficial effects as follows, taken into full account the directivity of predictive mode, simultaneously threshold value choose the distortion value characteristic distributions that also satisfies various patterns, can better controlled model selection process.Compare under the almost undiminished situation of signal to noise ratio with full ergodic algorithm, lifting by a relatively large margin coding rate.
Embodiment
Adopt JM (test model) that the test source coding commonly used back that ITU (international telegraph union) provides is found that intraframe coding finally adopts the situation of MPM pattern more, has accounted for about 60%～92%.Video source is level and smooth more, and employing MPM is that the situation of optimization model is more.MPM is meant the pattern that most probable adopts, and it is the actual prediction pattern according to two pieces employings of the fast left side of current 4x4 and top, by what calculate.If finally adopt this pattern, in coding, can represent so with littler data volume.Following predictive mode 0-8, each digital appointment be a kind of specific predictive mode, (http://www.itu.int/rec/T-REC-H.264/en) is described as follows in the ITU standard document.
Pattern 0:Intra_4x4_Vertical (vertically)
Pattern 1:Intra_4x4_Horizontal (level)
Pattern 2:Intra_4x4_DC
Mode 3: Intra_4x4_Diagonal_Down_Left (bottom left diagonal)
Pattern 4:Intra_4x4_Diagonal_Down_Right (bottom right diagonal)
Pattern 5:Intra_4x4_Vertical_Right (right vertical)
Pattern 6:Intra_4x4_Horizontal_Down (following level)
Mode 7: Intra_4x4_Vertical_Left (left side is vertical)
Pattern 8:Intra_4x4_Horizontal_Up (going up level)
Implementing procedure is as follows:
1, calculates the distortion value of MPM pattern earlier, judge C
MPMWhether smaller or equal to threshold value T1; If optimal prediction modes is the MPM pattern,, enter step 2 as not;
2, computation schema 0,1,2 distortion value, and whether in this three kind patterns have one distortion value smaller or equal to threshold value T2, in this way, adopt this pattern be optimal prediction modes, as not, enter step 3 if judging;
3, whether the distortion value of judgment model 0 is less than the distortion value of pattern 1, if enter step 4; If not, enter step 5;
Whether 4, the distortion value of computation schema 5 and mode 7, and the distortion value of judgment model 5 less than mode 7, if the distortion value of computation schema 4 again is made as optimal prediction modes with the pattern of distortion value minimum in pattern 5 and the pattern 4; If not, the distortion value of computation schema 3 again is made as optimal prediction modes with the pattern of distortion value minimum in mode 7 and the mode 3;
Whether 5, the distortion value of computation schema 6 and pattern 8, and the distortion value of judgment model 6 less than pattern 8, if the distortion value of computation schema 4 again is made as optimal prediction modes with the pattern of distortion value minimum in pattern 6 and the pattern 4; If not, the distortion value of computation schema 3 again is made as optimal prediction modes with the pattern of distortion value minimum in pattern 8 and the mode 3.
Wherein step 1 and step 2 have been used two threshold value T1 and T2.T1 is used for judging whether the distortion value of MPM pattern is enough little.Whether T2 is used for the distortion value of judgment model 0, pattern 1, mode 3 enough little.
The evaluation process of T1 and T2 is as follows:
T1:T1=totalT1/totalT1Num; Wherein, totalT1 is the distortion value summation that optimal prediction modes has adopted the piece of pattern 0, pattern 1 or mode 3; TotalT1Num is the number that optimal prediction modes has adopted the piece of pattern 0, pattern 1 or mode 3.In the practical operation, behind the optimal prediction modes that obtains current 4x4 piece, if predictive mode is wherein a kind of in pattern 0, pattern 1, the mode 3, just the value of temporary variable totalT1 is added the distortion value of this optimal prediction modes, value with totalT1Num adds 1 simultaneously, otherwise does not revise the value of above-mentioned variable.
T2:T2=totalT2/totalT2Num; Wherein, totalT2 is the distortion value summation of the optimal prediction modes of each encoding block; TotalT2Num is the number of piece in the cataloged procedure.In the practical operation, behind the optimal prediction modes that obtains current 4x4 piece, just the value of temporary variable totalT2 is added the distortion value of optimization model, the value with totalT2Num adds 1 simultaneously.
Claims (1)
1, a kind of fast mode selection method in frame is characterized in that, may further comprise the steps:
A, judge that whether the distortion value of MPM pattern is smaller or equal to first threshold T1; If the optimal prediction modes of current block is the MPM pattern; As not, enter step b;
B, calculate the distortion value of vertical pattern, horizontal pattern, DC pattern, judge in these three kinds of patterns whether a distortion value is arranged smaller or equal to the second threshold value T2; In this way, adopting this pattern is the optimal prediction modes of current block; As not, enter step c;
C, judge vertical pattern distortion value whether less than the distortion value of horizontal pattern; If enter steps d; Enter step e if not;
D, calculate the distortion value of right vertical mode and left vertical mode, and the distortion value of judging right vertical mode is whether less than the distortion value of left vertical mode; If, calculate the distortion value of bottom right diagonal model again, the pattern that distortion value in right vertical mode and the bottom right diagonal model is less is as the optimal prediction modes of current block; If not, calculate the distortion value of bottom left diagonal model again, the pattern that distortion value in left vertical mode and the bottom left diagonal model is less is as the optimal prediction modes of current block;
E, calculate the distortion value of horizontal pattern and last horizontal pattern down, and the distortion value of judging following horizontal pattern is whether less than the distortion value of last horizontal pattern; If, calculate the distortion value of bottom right diagonal model again, the pattern that distortion value in horizontal pattern and the bottom right diagonal model down is less is as the optimal prediction modes of current block; If not, calculate the distortion value of bottom left diagonal model again, the pattern that distortion value in last horizontal pattern and the bottom left diagonal model is less is as the optimal prediction modes of current block,
Described first threshold T1=totalT1/totalT1Num; Wherein, totalT1 is the distortion value summation that optimal prediction modes has adopted the piece of vertical pattern, horizontal pattern or DC pattern; TotalT1Num is the number that optimal prediction modes has adopted the piece of vertical pattern, horizontal pattern or DC pattern; The described second threshold value T2=totalT2/totalT2Num; Wherein, totalT2: the distortion value summation of the optimal prediction modes of each encoding block; TotalT2Num is the number of piece in the cataloged procedure.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200810301864 CN100596202C (en) | 2008-05-30 | 2008-05-30 | Fast mode selection method in frame |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200810301864 CN100596202C (en) | 2008-05-30 | 2008-05-30 | Fast mode selection method in frame |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101287125A CN101287125A (en) | 2008-10-15 |
| CN100596202C true CN100596202C (en) | 2010-03-24 |
Family
ID=40059057
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200810301864 Expired - Fee Related CN100596202C (en) | 2008-05-30 | 2008-05-30 | Fast mode selection method in frame |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100596202C (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20120070479A (en) | 2010-12-21 | 2012-06-29 | 한국전자통신연구원 | Method and apparatus for encoding and decoding of intra prediction mode information |
| WO2013062198A1 (en) * | 2011-10-24 | 2013-05-02 | (주)인터앱 | Image decoding apparatus |
| SI2942954T1 (en) * | 2011-10-24 | 2020-10-30 | Innotive Ltd | Image decoding apparatus |
| ES2816567T3 (en) * | 2011-10-24 | 2021-04-05 | Innotive Ltd | Method and apparatus for decoding intra-prediction mode |
| KR20130049523A (en) * | 2011-11-04 | 2013-05-14 | 오수미 | Apparatus for generating intra prediction block |
| KR20130049522A (en) * | 2011-11-04 | 2013-05-14 | 오수미 | Method for generating intra prediction block |
| KR20130050404A (en) * | 2011-11-07 | 2013-05-16 | 오수미 | Method for generating reconstructed block in inter prediction mode |
| CN104639939B (en) * | 2015-02-04 | 2018-02-06 | 四川虹电数字家庭产业技术研究院有限公司 | A kind of optimization method of infra-frame prediction MPM mechanism |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1535027A (en) * | 2004-01-16 | 2004-10-06 | 北京工业大学 | A Method of Intra-frame Prediction for Video Coding |
| CN1736103A (en) * | 2003-01-10 | 2006-02-15 | 汤姆森许可贸易公司 | Fast mode decision making for interframe encoding |
| CN101060626A (en) * | 2007-05-10 | 2007-10-24 | 武汉大学 | A rapid selection method for video transcoding in frame mode |
| CN101141649A (en) * | 2007-07-31 | 2008-03-12 | 北京大学 | Method and device for selecting optimal mode of intra-frame prediction coding for video coding |
| CN101179728A (en) * | 2007-12-13 | 2008-05-14 | 北京中星微电子有限公司 | Method and apparatus for determining interframe encoding mode |
-
2008
- 2008-05-30 CN CN 200810301864 patent/CN100596202C/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1736103A (en) * | 2003-01-10 | 2006-02-15 | 汤姆森许可贸易公司 | Fast mode decision making for interframe encoding |
| CN1535027A (en) * | 2004-01-16 | 2004-10-06 | 北京工业大学 | A Method of Intra-frame Prediction for Video Coding |
| CN101060626A (en) * | 2007-05-10 | 2007-10-24 | 武汉大学 | A rapid selection method for video transcoding in frame mode |
| CN101141649A (en) * | 2007-07-31 | 2008-03-12 | 北京大学 | Method and device for selecting optimal mode of intra-frame prediction coding for video coding |
| CN101179728A (en) * | 2007-12-13 | 2008-05-14 | 北京中星微电子有限公司 | Method and apparatus for determining interframe encoding mode |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101287125A (en) | 2008-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100596202C (en) | Fast mode selection method in frame | |
| CN103517069B (en) | A kind of HEVC intra-frame prediction quick mode selection method based on texture analysis | |
| CN102984521B (en) | High-efficiency video coding inter-frame mode judging method based on temporal relativity | |
| CN103188496B (en) | Based on the method for coding quick movement estimation video of motion vector distribution prediction | |
| CN106101707A (en) | Block elimination filtering | |
| CN104601993A (en) | Video coding method and device | |
| CN100574447C (en) | Fast intraframe predicting mode selecting method based on the AVS video coding | |
| CN101815218A (en) | Method for coding quick movement estimation video based on macro block characteristics | |
| KR101630871B1 (en) | Method and apparatus for deciding intra prediction mode | |
| CN102364950B (en) | H.264/advanced video coding (AVC)-standard-based intra-frame prediction mode rapid selection method and device | |
| CN105141954A (en) | HEVC interframe coding quick mode selection method | |
| CN104853191B (en) | A kind of HEVC fast encoding method | |
| CN104902271A (en) | Prediction mode selection method and device | |
| CN107566846A (en) | Video coding skip mode decision-making technique, device, equipment and storage medium | |
| CN101304529A (en) | Method and device for selecting macro block mode | |
| CN101389028A (en) | A Video Intra-Frame Coding Method Based on Spatial Domain Decomposition | |
| CN110365975A (en) | A kind of AVS2 video encoding and decoding standard prioritization scheme | |
| CN101895761B (en) | Quick intraframe prediction algorithm | |
| CN109151467B (en) | Fast selection of inter-frame mode for screen content coding based on image block activity | |
| CN102547257B (en) | Method for obtaining optimal prediction mode and device | |
| CN101867818B (en) | Selection method and device of macroblock mode | |
| CN109302616A (en) | A kind of HEVC Inter Prediction High-speed Algorithm based on RC prediction | |
| CN100558169C (en) | Quick selection method of inter-frame mode in video transcoding | |
| CN102196255B (en) | A Formation Method of Video Coding Complexity Control Model | |
| CN103051896B (en) | Mode skipping-based video frequency coding method and mode skipping-based video frequency coding system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100324 Termination date: 20160530 |