CN107396102A - A kind of inter-frame mode fast selecting method and device based on Merge technological movement vectors - Google Patents
A kind of inter-frame mode fast selecting method and device based on Merge technological movement vectors Download PDFInfo
- Publication number
- CN107396102A CN107396102A CN201710762301.1A CN201710762301A CN107396102A CN 107396102 A CN107396102 A CN 107396102A CN 201710762301 A CN201710762301 A CN 201710762301A CN 107396102 A CN107396102 A CN 107396102A
- Authority
- CN
- China
- Prior art keywords
- mrow
- mtd
- coded unit
- current coded
- merge
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/103—Selection of coding mode or of prediction mode
- H04N19/109—Selection of coding mode or of prediction mode among a plurality of temporal predictive coding modes
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/146—Data rate or code amount at the encoder output
- H04N19/147—Data rate or code amount at the encoder output according to rate distortion criteria
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses a kind of inter-frame mode fast selecting method and device based on Merge technological movement vectors, this method projects to reference frame by current coded unit CU motion vector MV, so as to find the projecting block that current CU is corresponded in reference frame, using correlation of both persons on predictive mode, decide whether to skip the inter-frame mode whether current CU skip Motion estimation and compensation using the feature of projecting block；Compared to prior art, by the information of known projecting block, to be predicted to current coded unit, the inter prediction computation complexity of video encoder is reduced, reduces the scramble time, improves code efficiency；And inventive algorithm is simple, amount of calculation is small, it is convenient to puts into practical application.
Description
Technical field
The invention belongs to field of video encoding, more particularly to a kind of inter-frame mode based on Merge technological movement vectors is fast
Fast system of selection and device.
Background technology
In coding framework, predictive coding is one of core technology of Video coding, and predictive coding is divided into infra-frame prediction again
And inter prediction.Infra-frame prediction is the spatial coherence according to video image, is predicted using neighborhood pixels encoded in image
Current pixel.Interframe encode is the temporal correlation according to video image, and image to be encoded is predicted using encoded image.By
In frame and inter prediction, encoder can eliminate the temporal correlation of video, to the residual error rather than original pixel value after prediction
Enter line translation, quantization, entropy code, thus greatly improve code efficiency.
Video encoding standard inter-predicted portions main at present all employ block-based motion compensation technique.Its is main
Principle is that a best matching blocks are found in encoded image before for each block of pixels of present image, and the process is referred to as transporting
Dynamic estimation.The image for being wherein used to predict is referred to as reference picture, and the displacement of reference block to current pixel block is referred to as motion vector, when
The difference of preceding piece and reference block is referred to as prediction residual.Due to the continuity of sequence of video images, usual motion vector in space and
There is also certain correlation on time, similarly, utilization space or time upper adjacent motion vector are to current block motion vector
It is predicted, only prediction residual is encoded, also can significantly saves the number of coded bits of motion vector.This predicted motion arrow
The technology of amount is referred to as Merge.
2013, ITU-T VCEG (Video Coding Experts Group) and ISO/IEC MPEG (dynamic image expert group) joints
It is proposed HEVC (efficient video coding) video compression scheme.Began from 2016, VCEG and MPEG begin one's study video of new generation
Encoder, and set up a panel of expert --- JVET (joint video research group), it is intended to further lifting HEVC pressure
Shrinkage.Video encoding standard of new generation is developed on the basis of HEVC, and the two is all employed in inter predication process
Merge technologies, unlike, the Merge patterns of video of new generation have three kinds：AffineMerge patterns based on affine transformation,
FRUC Merge patterns based on template matches and the 2Nx2N Merge patterns based on temporal correlation.These patterns are answered
With the compression performance for improving encoder, the scramble time is also considerably increased, have impact on the research and development speed and application value of standard.
Just there is motion to point out this drawback in the third session on video encoding standard of new generation, and ask to adopt its complexity
Take action.
Video encoding standard of new generation step when doing inter prediction is as follows：
Step 1：Affine Merge patterns, i.e. affine motion compensation prediction are first done, preserves its rate distortion costs and pre-
Measurement information, and current best mode is set to Affine Merge patterns,；
Step 2：2Nx2N Merge patterns, i.e. ordinary movement compensation prediction are done again, if the rate distortion costs of the pattern are small
In the rate distortion costs of Affine Merge patterns, then optimal mode is set to 2Nx2N Merge patterns and preserves its rate distortion
Cost and information of forecasting；
Step 3：Then FRUC Merge patterns are done, i.e., the motion vector generation based on template matches, if the rate of the pattern
Distortion cost is less than the rate distortion costs of current best mode, then optimal mode is set into FRUC Merge patterns and preserves its rate
Distortion cost and information of forecasting.Three of the above pattern belongs to Merge patterns；
Step 4：Then take exercises the inter-frame forecast mode of estimation and motion compensation, the pattern found out by motion search
Match block in reference frame draws motion vector and prediction residual, therefore time-consuming more.
If the rate distortion costs of the pattern are less than the rate distortion costs of current best mode, optimal mode is set to motion
The inter-frame forecast mode of estimation and motion compensation simultaneously preserves its rate distortion costs and information of forecasting.
Wherein, the scramble time of the inter-frame mode of Motion estimation and compensation accounts for the 41% of total encoding time, therefore, such as
It is a kind of so as to jumping in three kinds of Merge patterns that fruit can just predict best inter mode before Motion estimation and compensation
A large amount of scramble times will then be reduced by crossing Motion estimation and compensation.
Although there are many interframe fast algorithms for being directed to HM video encoders at present, as T.Mallikarachchi scholar exists
Specific dimensions CU predictive coding, S.Ahn are skipped in proposition according to motion homogeneity in IEEE image procossing international conferences in 2014
It is proposed to use on Circuits and System for Video Technology, the IEEE Transactions of 2015
With the position CU current CU of pixel adaptive equalization parameter evaluation Texture complication, it is pre- that some interframe are skipped according to Texture complication
Survey pattern.But because video encoding standard of new generation employs QTBT (four fork y-bend divisions) coding structure and eliminates pre-
Unit PU concept is surveyed, so algorithm above is not particularly suited for video encoding standard of new generation.Other, such as based on variance
, method based on Bayes, be not particularly suited for practical application because computation complexity is too high.
The Geneva conference in May, 2016 proposes the test model JEM2.0 of video encoding standard of new generation, now
Average coding time of the JEM encoders under random arrangement is 5.3 times of HEVC encoders.Wherein, inter prediction is in total volume
Occupy for about 68% time in the code time, similarly, in conventional coding standard, when inter prediction also occupies a large amount of codings
Between, therefore inter prediction is the important module for reducing the scramble time, has very big room for improvement, if can be by inter prediction
Time, which is reduced, to greatly improve the efficiency of encoder.
The content of the invention
The purpose of the present invention is the defects of being directed to inter prediction encoding overlong time and the deficiencies in the prior art, proposes one
Inter-frame mode fast selecting method of the kind based on Merge technological movement vectors, shortens its scramble time, improves its practical application
Property, while also provided convenience for its further research and development.
A kind of inter-frame mode fast selecting method based on Merge technological movement vectors, comprises the following steps：
Step 1：Current coded unit is obtained under optimal inter-frame forecast mode, corresponding to the projecting block on reference frame；
After current coded unit CU finishes Affine Merge, 2Nx2N Merge and FRUC Merge patterns, according to rate
Distortion cost decision-making goes out current coded unit CU optimal inter-frame forecast mode；
Current coded unit CU motion vector MV is obtained based on optimal inter-frame forecast mode, by current coded unit CU
Each pixel translation MV after obtain with current coded unit CU size identical translational blocks, finally the translational block is projected to
In reference frame, obtain corresponding to current coded unit CU projecting block in reference frame；
Step 2：Inter-frame mode is Merge area in the projecting block that calculation procedure one obtains：
SM=∑ f (Mode (x, y)) (1)
Wherein, SMThe area for being Merge for inter-frame mode in projecting block, (x, y) are the coordinate of pixel in projecting block,
Mode (x, y) is the optimal inter-frame forecast mode for the pixel that coordinate is (x, y)；When coordinate is the optimal of the pixel of (x, y)
When pattern is Merge, Mode (x, y) takes 1, otherwise takes 0；
Step 3：Calculate the current coded unit CU gross area：
SC=∑ g (x1, y1) (3)
Wherein, SCFor the current coded unit CU gross area, Cur_CU represents current coded unit CU pixel coordinate model
Enclose；(x1, y1) is the coordinate of pixel in current frame image, when the coordinate of pixel (x1, y1) is in current coded unit CU models
When enclosing interior, g (x1, y1) takes 1, otherwise takes 0；
Step 4：By the current coded unit CU gross area meters in the Merge areas and step 3 of the projecting block of step 2
The area for calculating Merge patterns in projecting block accounts for the ratio γ of the gross area：
Step 5：When the ratio γ of step 4 is more than given threshold λ, step 6 is skipped, terminates current coded unit CU
Predictive coding；Otherwise, into step 6；
Wherein, λ can use any real number in [0,1]；
Step 6：The inter prediction of Motion estimation and compensation is carried out to current coded unit CU.
Further, the λ values are 0.85.
A kind of quick selection device of inter-frame mode based on Merge technological movement vectors, including：
Projecting block acquiring unit：Current coded unit is obtained under optimal inter-frame forecast mode, corresponding on reference frame
Projecting block；
After current coded unit CU finishes Affine Merge, 2Nx2N Merge and FRUC Merge patterns, according to rate
Distortion cost decision-making goes out current coded unit CU optimal inter-frame forecast mode；
Current coded unit CU motion vector MV is obtained based on optimal inter-frame forecast mode, by current coded unit CU
Each pixel translation MV after obtain with current coded unit CU size identical translational blocks, finally the translational block is projected to
In reference frame, obtain corresponding to current coded unit CU projecting block in reference frame；
Inter-frame mode Merge areal calculation unit：According to the inter-frame mode of each pixel in projecting block, interframe mould is calculated
Formula is Merge area：
SM=∑ f (Mode (x, y))
Wherein, SMThe area for being Merge for inter-frame mode in projecting block, (x, y) are the coordinate of pixel in projecting block,
Mode (x, y) is the optimal inter-frame forecast mode for the pixel that coordinate is (x, y)；When coordinate is the optimal of the pixel of (x, y)
When pattern is Merge, Mode (x, y) takes 1, otherwise takes 0；
Current coded unit CU gross area computing unit：Whether belong to current volume according to each pixel in current frame image
Code unit CU, calculate the current coded unit CU gross area：
SC=∑ g (x1, y1)
Wherein, SCFor the current coded unit CU gross area, Cur_CU represents current coded unit CU pixel coordinate model
Enclose；(x1, y1) is the coordinate of pixel in current frame image, when the coordinate of pixel (x1, y1) is in current coded unit CU models
When enclosing interior, g (x1, y1) takes 1, otherwise takes 0；
Projecting block Merge pattern ratio computing units：By the Merge areas and the current coded unit CU gross areas of projecting block
The area for calculating Merge patterns in projecting block accounts for the ratio γ of the gross area：
Skip unit：When ratio γ is more than given threshold λ, skips and estimation and fortune are carried out to current coded unit CU
The inter prediction of dynamic compensation, terminates current coded unit CU predictive coding；
Wherein, λ can use any real number in [0,1].
Further, the threshold value λ values skipped in unit are 0.85.
Beneficial effect
, should the invention provides a kind of inter-frame mode fast selecting method and device based on Merge technological movement vectors
Method projects to reference frame by current coded unit CU motion vector MV, so as to find the throwing that current CU is corresponded in reference frame
Shadow block, using correlation of both persons on predictive mode, decide whether to skip whether current CU jumps using the feature of projecting block
Cross the inter-frame mode of Motion estimation and compensation；Compared to prior art, by the information of known projecting block, to current
Coding unit is predicted, and is reduced the inter prediction computation complexity of video encoder, is reduced the scramble time, improves volume
Code efficiency；And inventive algorithm is simple, amount of calculation is small, it is convenient to puts into practical application.
Brief description of the drawings
Fig. 1 is current coded unit and its projecting block corresponding relation and motion vector schematic diagram, wherein, (a) closes to be corresponding
System, (b) is motion vector schematic diagram；
Fig. 2 is CU information storage means；
Fig. 3 is the flow chart of the method for the invention.
Embodiment
Technical scheme is described in detail with preferred embodiment below in conjunction with the accompanying drawings.Selected implementation
Test model --- the JEM4.0 that encoder used in example is issued for video encoding standard expert group of future generation, specific coding parameter
Configuration selection JEM standard configuration files：Encoder_randomaccess_jvet10.cfg, and corresponding cycle tests
Standard configuration file.
To reduce the scramble time, improve operating efficiency, the technical scheme that specifically uses of the present invention for：(compiled by current CU
Code block) motion vector MV project to reference frame, in theory can be approximate so as to find the projecting block that current CU is corresponded in reference frame
Think that the projecting block has arrived the present encoding CU of present frame position (left see figure one) by motion vector MV displacement movement.Cause
The properties of this projecting block and present encoding CU properties should be identical, such as pixel distribution situation, interframe
Predictive mode etc..Just the similarity both on predictive mode, one threshold value of setting (are designated as skip thresholds to the present invention below
Value), decided whether to skip the inter-frame mode of Motion estimation and compensation according to the threshold value.
As shown in figure 3, the specific method of the present invention is as follows：
Step 1：JEM encoders finish Affine Merge, 2Nx2N Merge and FRUC Merge patterns in current CU
Afterwards, an optimal mode can be now gone out according to rate distortion costs decision-making.Motion vector MV first in optimal mode ought
Preceding CU is translated (right see figure one).Wherein, motion vector includes horizontal displacement components MVx and vertical displacement components MVy, translation
Method be first to record current CU apex coordinate, the wide of CU and high, be designated as (x, y), width and height respectively, then translate
Block summit is (x+MVx, y+MVy) in the coordinate of reference frame, and the size of translational block is consistent with present encoding CU, then again will translation
Block is projected in reference frame (as shown in Figure 1a).
Step 2：Count the area that inter-frame mode in the projecting block obtained by step 1 is finally Merge.Due to new one
Pattern information storage mode for video encoding standard be in units of the block of pixel 4x4 sizes storage rather than using pixel as
Unit (see Fig. 2), so statistical is each fritter for traveling through projecting block, area can be calculated with following formula.
SM=∑ f (Mode (x, y))
Wherein, SMIt is finally Merge area for inter-frame mode in projecting block, (x, y) is the seat of pixel in projecting block
Mark.Mode (x, y) is that coordinate is the optimal inter-frame forecast mode of the pixel of (x, y) in projecting block.When pixel in projecting block
When the optimal mode of coordinate (x, y) is Merge, Mode (x, y) takes 1, otherwise takes 0.
Step 3：Calculate the current coded unit CU gross area.Same step 2, the statistical of the CU gross areas is also traversal
Current coded unit CU each fritter, obtain the quantity of pixel in current coded unit CU.Specific calculation is as follows：
SC=∑ g (x1, y1)
Wherein, SCFor the current coded unit CU gross area, Cur_CU is current coded unit CU coordinate range, (x1,
Y1) be pixel in image coordinate, when the coordinate of pixel is in the range of current coded unit CU, g (x1, y1) takes 1, no
Then take 0.
Step 4：By the current coded unit CU gross area meters in the Merge areas and step 3 of the projecting block of step 2
The area for calculating Merge patterns in projecting block accounts for the ratio γ of the gross area.
The ratio that the area of the Merge patterns of any one projecting block accounts for the gross area can be obtained by following formula：
Step 5：When the ratio γ of step 4 is more than threshold value λ, illustrate current coded unit CU best inter mode
It is very likely Merge, therefore skips step 6, terminates current CU predictive coding.Wherein, γ can use any in [0,1]
Real number, when higher to video quality demands, when the scramble time requires not strict, λ can use value larger in scope, on the contrary it is desirable compared with
Small value.Counted by many experiments, preferable balance can be obtained between video quality and scramble time when λ takes 0.85.
Step 6：Carry out the inter prediction of Motion estimation and compensation.
A kind of quick selection device of inter-frame mode based on Merge technological movement vectors, including：
Projecting block acquiring unit：Current coded unit is obtained under optimal inter-frame forecast mode, corresponding on reference frame
Projecting block；
After current coded unit CU finishes Affine Merge, 2Nx2N Merge and FRUC Merge patterns, according to rate
Distortion cost decision-making goes out current coded unit CU optimal inter-frame forecast mode；
Current coded unit CU motion vector MV is obtained based on optimal inter-frame forecast mode, by current coded unit CU
Each pixel translation MV after obtain with current coded unit CU size identical translational blocks, finally the translational block is projected to
In reference frame, obtain corresponding to current coded unit CU projecting block in reference frame；
Inter-frame mode Merge areal calculation unit：According to the inter-frame mode of each pixel in projecting block, interframe mould is calculated
Formula is Merge area：
SM=∑ f (Mode (x, y))
Wherein, SMThe area for being Merge for inter-frame mode in projecting block, (x, y) are the coordinate of pixel in projecting block,
Mode (x, y) is the optimal inter-frame forecast mode for the pixel that coordinate is (x, y)；When coordinate is the optimal of the pixel of (x, y)
When pattern is Merge, Mode (x, y) takes 1, otherwise takes 0；
Current coded unit CU gross area computing unit：Whether belong to current volume according to each pixel in current frame image
Code unit CU, calculate the current coded unit CU gross area：
SC=∑ g (x1, y1)
Wherein, SCFor the current coded unit CU gross area, Cur_CU represents current coded unit CU pixel coordinate model
Enclose；(x1, y1) is the coordinate of pixel in current frame image, when the coordinate of pixel (x1, y1) is in current coded unit CU models
When enclosing interior, g (x1, y1) takes 1, otherwise takes 0；
Projecting block Merge pattern ratio computing units：By the Merge areas and the current coded unit CU gross areas of projecting block
The area for calculating Merge patterns in projecting block accounts for the ratio γ of the gross area：
Skip unit：When ratio γ is more than given threshold λ, skips and estimation and fortune are carried out to current coded unit CU
The inter prediction of dynamic compensation, terminates current coded unit CU predictive coding；
Wherein, λ can use any real number in [0,1].
The threshold value λ values skipped in unit are 0.85.
In order to verify the feasibility and validity of proposed interframe fast algorithm, based on video encoding standard of new generation
Test model JEM4.0 realizes fast algorithm mentioned above.And last data are all in the high performance platform of school
Operation obtains, and ensures the true and accurate of experimental data.The configuration selection JEM standard configurations of the specific coding parameter of all experiments
File：Encoder_randomaccess_jvet10.cfg, and the standard configuration file of corresponding cycle tests.
Experimental result is as shown in table 1.Wherein, QP is quantization parameter, and Δ Bits% is the bit compared with traditional encoder
Rate changes percentage, and Δ PSNR/dB is that Y-PSNR changes compared with traditional encoder, and TS/% is and traditional coding
Device compares saved percentage of time.Δ BDBR is illustrated under same objective quality, conventional codec and improved volume
The code check of code device saves situation.The smaller explanation algorithm effects of Δ BDBR are better.
The experimental result of table 1
By the way that in experiment simulation, the experimental result of the quick inter-frame algorithm proposed in the present invention is as shown in table 1.By table 1
Understand, the algorithm has reached the purpose on the premise of the quality of video is ensured, improving the efficiency of coding.
Specific embodiment described herein is only to spirit explanation for example of the invention.Technology belonging to the present invention is led
The technical staff in domain can be made various modifications or supplement to described specific embodiment or be replaced using similar mode
Generation, but without departing from the spiritual of the present invention or surmount scope defined in appended claims.
Claims (4)
1. a kind of inter-frame mode fast selecting method based on Merge technological movement vectors, it is characterised in that including following step
Suddenly：
Step 1：Current coded unit is obtained under optimal inter-frame forecast mode, corresponding to the projecting block on reference frame；
After current coded unit CU finishes Affine Merge, 2Nx2N Merge and FRUC Merge patterns, according to rate distortion
Cost decision-making goes out current coded unit CU optimal inter-frame forecast mode；
Current coded unit CU motion vector MV is obtained based on optimal inter-frame forecast mode, will be every in current coded unit CU
Obtained after individual pixel translation MV with current coded unit CU size identical translational blocks, the translational block is finally projected into reference
In frame, obtain corresponding to current coded unit CU projecting block in reference frame；
Step 2：Inter-frame mode is Merge area in the projecting block that calculation procedure one obtains：
SM=∑ f (Mode (x, y)) (1)
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mi> </mi>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>M</mi>
<mi>e</mi>
<mi>r</mi>
<mi>g</mi>
<mi>e</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mi> </mi>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>!</mo>
<mo>=</mo>
<mi>M</mi>
<mi>e</mi>
<mi>r</mi>
<mi>g</mi>
<mi>e</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, SMThe area for being Merge for inter-frame mode in projecting block, (x, y) be projecting block in pixel coordinate, Mode (x,
Y) it is optimal inter-frame forecast mode of the coordinate for the pixel of (x, y)；When coordinate is for the optimal mode of the pixel of (x, y)
During Merge, Mode (x, y) takes 1, otherwise takes 0；
Step 3：Calculate the current coded unit CU gross area：
SC=∑ g (x1, y1) (3)
<mrow>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mn>1</mn>
<mo>,</mo>
<mi>y</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mn>1</mn>
<mo>,</mo>
<mi>y</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>&Element;</mo>
<mi>C</mi>
<mi>u</mi>
<mi>r</mi>
<mo>_</mo>
<mi>C</mi>
<mi>U</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mn>1</mn>
<mo>,</mo>
<mi>y</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>&NotElement;</mo>
<mi>C</mi>
<mi>u</mi>
<mi>r</mi>
<mo>_</mo>
<mi>C</mi>
<mi>U</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, SCFor the current coded unit CU gross area, Cur_CU represents current coded unit CU pixel coordinate scope；
(x1, y1) is the coordinate of pixel in current frame image, when the coordinate of pixel (x1, y1) is in the range of current coded unit CU
When, g (x1, y1) takes 1, otherwise takes 0；
Step 4：Calculated and thrown by the current coded unit CU gross areas in the Merge areas and step 3 of the projecting block of step 2
The area of Merge patterns accounts for the ratio γ of the gross area in shadow block：
<mrow>
<mi>&gamma;</mi>
<mo>=</mo>
<mfrac>
<msub>
<mi>S</mi>
<mi>M</mi>
</msub>
<msub>
<mi>S</mi>
<mi>C</mi>
</msub>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 5：When the ratio γ of step 4 is more than given threshold λ, step 6 is skipped, terminates the pre- of current coded unit CU
Survey coding；Otherwise, into step 6；
Wherein, λ can use any real number in [0,1]；
Step 6：The inter prediction of Motion estimation and compensation is carried out to current coded unit CU.
2. according to the method for claim 1, it is characterised in that the λ values are 0.85.
A kind of 3. quick selection device of inter-frame mode based on Merge technological movement vectors, it is characterised in that including：
Projecting block acquiring unit：Current coded unit is obtained under optimal inter-frame forecast mode, corresponding to the projection on reference frame
Block；
After current coded unit CU finishes Affine Merge, 2Nx2N Merge and FRUC Merge patterns, according to rate distortion
Cost decision-making goes out current coded unit CU optimal inter-frame forecast mode；
Current coded unit CU motion vector MV is obtained based on optimal inter-frame forecast mode, will be every in current coded unit CU
Obtained after individual pixel translation MV with current coded unit CU size identical translational blocks, the translational block is finally projected into reference
In frame, obtain corresponding to current coded unit CU projecting block in reference frame；
Inter-frame mode Merge areal calculation unit：According to the inter-frame mode of each pixel in projecting block, calculating inter-frame mode is
Merge area：
SM=∑ f (Mode (x, y))
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mi> </mi>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>M</mi>
<mi>e</mi>
<mi>r</mi>
<mi>g</mi>
<mi>e</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mi> </mi>
<mi>M</mi>
<mi>o</mi>
<mi>d</mi>
<mi>e</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>!</mo>
<mo>=</mo>
<mi>M</mi>
<mi>e</mi>
<mi>r</mi>
<mi>g</mi>
<mi>e</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, SMThe area for being Merge for inter-frame mode in projecting block, (x, y) be projecting block in pixel coordinate, Mode (x,
Y) it is optimal inter-frame forecast mode of the coordinate for the pixel of (x, y)；When coordinate is for the optimal mode of the pixel of (x, y)
During Merge, Mode (x, y) takes 1, otherwise takes 0；
Current coded unit CU gross area computing unit：Whether belong to present encoding list according to each pixel in current frame image
First CU, calculate the current coded unit CU gross area：
SC=∑ g (x1, y1)
<mrow>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mn>1</mn>
<mo>,</mo>
<mi>y</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mn>1</mn>
<mo>,</mo>
<mi>y</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>&Element;</mo>
<mi>C</mi>
<mi>u</mi>
<mi>r</mi>
<mo>_</mo>
<mi>C</mi>
<mi>U</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mn>1</mn>
<mo>,</mo>
<mi>y</mi>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>&NotElement;</mo>
<mi>C</mi>
<mi>u</mi>
<mi>r</mi>
<mo>_</mo>
<mi>C</mi>
<mi>U</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, SCFor the current coded unit CU gross area, Cur_CU represents current coded unit CU pixel coordinate scope；
(x1, y1) is the coordinate of pixel in current frame image, when the coordinate of pixel (x1, y1) is in the range of current coded unit CU
When, g (x1, y1) takes 1, otherwise takes 0；
Projecting block Merge pattern ratio computing units：Calculated by the Merge areas and the current coded unit CU gross areas of projecting block
The area of Merge patterns accounts for the ratio γ of the gross area in projecting block：
<mrow>
<mi>&gamma;</mi>
<mo>=</mo>
<mfrac>
<msub>
<mi>S</mi>
<mi>M</mi>
</msub>
<msub>
<mi>S</mi>
<mi>C</mi>
</msub>
</mfrac>
</mrow>
Skip unit：When ratio γ is more than given threshold λ, skips and estimation and motion benefit are carried out to current coded unit CU
The inter prediction repaid, terminate current coded unit CU predictive coding；
Wherein, λ can use any real number in [0,1].
4. device according to claim 1, it is characterised in that the threshold value λ values skipped in unit are 0.85.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710762301.1A CN107396102B (en) | 2017-08-30 | 2017-08-30 | A kind of inter-frame mode fast selecting method and device based on Merge technological movement vector |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710762301.1A CN107396102B (en) | 2017-08-30 | 2017-08-30 | A kind of inter-frame mode fast selecting method and device based on Merge technological movement vector |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107396102A true CN107396102A (en) | 2017-11-24 |
| CN107396102B CN107396102B (en) | 2019-10-08 |
Family
ID=60348165
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710762301.1A Active CN107396102B (en) | 2017-08-30 | 2017-08-30 | A kind of inter-frame mode fast selecting method and device based on Merge technological movement vector |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107396102B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108174204A (en) * | 2018-03-06 | 2018-06-15 | 中南大学 | A kind of interframe fast schema selection method based on decision tree |
| CN108347616A (en) * | 2018-03-09 | 2018-07-31 | 中南大学 | A kind of depth prediction approach and device based on optional time domain motion-vector prediction |
| CN110662041A (en) * | 2018-06-29 | 2020-01-07 | 北京字节跳动网络技术有限公司 | Extending interactions between Merge modes and other video coding tools |
| CN110809156A (en) * | 2018-08-04 | 2020-02-18 | 北京字节跳动网络技术有限公司 | Interaction between different decoder-side motion vector derivation modes |
| CN111698502A (en) * | 2020-06-19 | 2020-09-22 | 中南大学 | VVC (variable visual code) -based affine motion estimation acceleration method and device and storage medium |
| CN112637592A (en) * | 2020-12-11 | 2021-04-09 | 百果园技术(新加坡)有限公司 | Method and device for video predictive coding |
| CN112839224A (en) * | 2019-11-22 | 2021-05-25 | 腾讯科技（深圳）有限公司 | Prediction mode selection method and device, video coding equipment and storage medium |
| CN114339231A (en) * | 2021-12-27 | 2022-04-12 | 杭州当虹科技股份有限公司 | Method for fast jumping Cu-level mode selection by utilizing motion vector |
| US11778170B2 (en) | 2018-10-06 | 2023-10-03 | Beijing Bytedance Network Technology Co., Ltd | Temporal gradient calculations in bio |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080165855A1 (en) * | 2007-01-08 | 2008-07-10 | Nokia Corporation | inter-layer prediction for extended spatial scalability in video coding |
| CN103338372A (en) * | 2013-06-15 | 2013-10-02 | 浙江大学 | Method and device for processing video |
| CN103379324A (en) * | 2012-04-16 | 2013-10-30 | 乐金电子(中国)研究开发中心有限公司 | Parallel realization method, device and system for advanced motion vector prediction AMVP |
| CN104038764A (en) * | 2014-06-27 | 2014-09-10 | 华中师范大学 | H.264-to-H.265 video transcoding method and transcoder |
| CN104601988A (en) * | 2014-06-10 | 2015-05-06 | 腾讯科技（北京）有限公司 | Video coder, method and device and inter-frame mode selection method and device thereof |
| US20150222904A1 (en) * | 2011-03-08 | 2015-08-06 | Texas Instruments Incorporated | Parsing friendly and error resilient merge flag coding in video coding |
| CN105959611A (en) * | 2016-07-14 | 2016-09-21 | 同观科技（深圳）有限公司 | Adaptive H264-to-HEVC (High Efficiency Video Coding) inter-frame fast transcoding method and apparatus |
| TW201637449A (en) * | 2015-01-29 | 2016-10-16 | Ｖｉｄ衡器股份有限公司 | Intra-block copy searching |
| US20160373766A1 (en) * | 2015-06-22 | 2016-12-22 | Cisco Technology, Inc. | Block-based video coding using a mixture of square and rectangular blocks |
-
2017
- 2017-08-30 CN CN201710762301.1A patent/CN107396102B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080165855A1 (en) * | 2007-01-08 | 2008-07-10 | Nokia Corporation | inter-layer prediction for extended spatial scalability in video coding |
| US20150222904A1 (en) * | 2011-03-08 | 2015-08-06 | Texas Instruments Incorporated | Parsing friendly and error resilient merge flag coding in video coding |
| CN103379324A (en) * | 2012-04-16 | 2013-10-30 | 乐金电子(中国)研究开发中心有限公司 | Parallel realization method, device and system for advanced motion vector prediction AMVP |
| CN103338372A (en) * | 2013-06-15 | 2013-10-02 | 浙江大学 | Method and device for processing video |
| CN104601988A (en) * | 2014-06-10 | 2015-05-06 | 腾讯科技（北京）有限公司 | Video coder, method and device and inter-frame mode selection method and device thereof |
| CN104038764A (en) * | 2014-06-27 | 2014-09-10 | 华中师范大学 | H.264-to-H.265 video transcoding method and transcoder |
| TW201637449A (en) * | 2015-01-29 | 2016-10-16 | Ｖｉｄ衡器股份有限公司 | Intra-block copy searching |
| US20160373766A1 (en) * | 2015-06-22 | 2016-12-22 | Cisco Technology, Inc. | Block-based video coding using a mixture of square and rectangular blocks |
| CN105959611A (en) * | 2016-07-14 | 2016-09-21 | 同观科技（深圳）有限公司 | Adaptive H264-to-HEVC (High Efficiency Video Coding) inter-frame fast transcoding method and apparatus |
Non-Patent Citations (1)
| Title |
|---|
| 黄晗: "《HEVC帧间帧内预测及优化技术研究》", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108174204A (en) * | 2018-03-06 | 2018-06-15 | 中南大学 | A kind of interframe fast schema selection method based on decision tree |
| CN108347616A (en) * | 2018-03-09 | 2018-07-31 | 中南大学 | A kind of depth prediction approach and device based on optional time domain motion-vector prediction |
| CN108347616B (en) * | 2018-03-09 | 2020-02-14 | 中南大学 | Depth prediction method and device based on optional time domain motion vector prediction |
| CN110662041B (en) * | 2018-06-29 | 2022-07-29 | 北京字节跳动网络技术有限公司 | Method and apparatus for video bitstream processing, method of storing video bitstream, and non-transitory computer-readable recording medium |
| CN110662041A (en) * | 2018-06-29 | 2020-01-07 | 北京字节跳动网络技术有限公司 | Extending interactions between Merge modes and other video coding tools |
| US11451819B2 (en) | 2018-08-04 | 2022-09-20 | Beijing Bytedance Network Technology Co., Ltd. | Clipping of updated MV or derived MV |
| CN110809156B (en) * | 2018-08-04 | 2022-08-12 | 北京字节跳动网络技术有限公司 | Interaction between different decoder-side motion vector derivation modes |
| US12120340B2 (en) | 2018-08-04 | 2024-10-15 | Beijing Bytedance Network Technology Co., Ltd | Constraints for usage of updated motion information |
| US11109055B2 (en) | 2018-08-04 | 2021-08-31 | Beijing Bytedance Network Technology Co., Ltd. | MVD precision for affine |
| US11470341B2 (en) | 2018-08-04 | 2022-10-11 | Beijing Bytedance Network Technology Co., Ltd. | Interaction between different DMVD models |
| US11330288B2 (en) | 2018-08-04 | 2022-05-10 | Beijing Bytedance Network Technology Co., Ltd. | Constraints for usage of updated motion information |
| CN110809156A (en) * | 2018-08-04 | 2020-02-18 | 北京字节跳动网络技术有限公司 | Interaction between different decoder-side motion vector derivation modes |
| CN110809155A (en) * | 2018-08-04 | 2020-02-18 | 北京字节跳动网络技术有限公司 | Restriction using updated motion information |
| US11778170B2 (en) | 2018-10-06 | 2023-10-03 | Beijing Bytedance Network Technology Co., Ltd | Temporal gradient calculations in bio |
| CN112839224B (en) * | 2019-11-22 | 2023-10-10 | 腾讯科技（深圳）有限公司 | Prediction mode selection method and device, video coding equipment and storage medium |
| CN112839224A (en) * | 2019-11-22 | 2021-05-25 | 腾讯科技（深圳）有限公司 | Prediction mode selection method and device, video coding equipment and storage medium |
| CN111698502A (en) * | 2020-06-19 | 2020-09-22 | 中南大学 | VVC (variable visual code) -based affine motion estimation acceleration method and device and storage medium |
| WO2022121786A1 (en) * | 2020-12-11 | 2022-06-16 | 百果园技术(新加坡)有限公司 | Video predictive coding method and apparatus |
| CN112637592B (en) * | 2020-12-11 | 2024-07-05 | 百果园技术(新加坡)有限公司 | Video predictive coding method and device |
| CN112637592A (en) * | 2020-12-11 | 2021-04-09 | 百果园技术(新加坡)有限公司 | Method and device for video predictive coding |
| CN114339231A (en) * | 2021-12-27 | 2022-04-12 | 杭州当虹科技股份有限公司 | Method for fast jumping Cu-level mode selection by utilizing motion vector |
| CN114339231B (en) * | 2021-12-27 | 2023-10-27 | 杭州当虹科技股份有限公司 | Method for rapidly jumping Cu-level mode selection by utilizing motion vector |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107396102B (en) | 2019-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107396102B (en) | A kind of inter-frame mode fast selecting method and device based on Merge technological movement vector | |
| CN107147911A (en) | LIC quick interframe coding mode selection method and device is compensated based on local luminance | |
| CN102137263B (en) | Distributed video coding and decoding methods based on classification of key frames of correlation noise model (CNM) | |
| CN110519600B (en) | Intra-frame and inter-frame joint prediction method and device, coder and decoder and storage device | |
| CN108781284A (en) | The method and device of coding and decoding video with affine motion compensation | |
| CN110087087A (en) | VVC interframe encode unit prediction mode shifts to an earlier date decision and block divides and shifts to an earlier date terminating method | |
| CN102025995B (en) | Spatial enhancement layer rapid mode selection method of scalable video coding | |
| CN108347616A (en) | A kind of depth prediction approach and device based on optional time domain motion-vector prediction | |
| CN107222742B (en) | Video coding Merge mode quick selecting method and device based on time-space domain correlation | |
| CN110062239B (en) | Reference frame selection method and device for video coding | |
| CN107079165A (en) | Use the method for video coding and device of prediction residual | |
| CN107197297A (en) | A kind of video steganalysis method of the detection based on DCT coefficient steganography | |
| CN102932642A (en) | Interframe coding quick mode selection method | |
| CN102065298A (en) | High-performance macroblock coding implementation method | |
| CN104811729B (en) | A kind of video multi-reference frame coding method | |
| CN107318016A (en) | A kind of HEVC inter-frame forecast mode method for rapidly judging based on zero piece of distribution | |
| CN105120290A (en) | Fast coding method for depth video | |
| CN101888546A (en) | Motion estimation method and device | |
| CN105898308A (en) | Resolution-variable coding mode prediction method and device | |
| CN108769696A (en) | A kind of DVC-HEVC video transcoding methods based on Fisher discriminates | |
| CN101102492A (en) | Conversion method from compression domain MPEG-2 based on interest area to H.264 video | |
| CN110519591A (en) | A kind of prediction mode fast selecting method based on intraframe coding in multipurpose coding | |
| CN106331700A (en) | Coding and decoding methods of reference image, coding device, and decoding device | |
| CN106131573B (en) | A kind of HEVC spatial resolutions code-transferring method | |
| CN109688411B (en) | Video coding rate distortion cost estimation method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |