CN107087172A - Quick code check code-transferring method and its system based on HEVC SCC - Google Patents
Quick code check code-transferring method and its system based on HEVC SCC Download PDFInfo
- Publication number
- CN107087172A CN107087172A CN201710175153.3A CN201710175153A CN107087172A CN 107087172 A CN107087172 A CN 107087172A CN 201710175153 A CN201710175153 A CN 201710175153A CN 107087172 A CN107087172 A CN 107087172A
- Authority
- CN
- China
- Prior art keywords
- unit
- decoding
- patterns
- current coded
- intrabc
- 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
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000005457 optimization Methods 0.000 claims description 21
- 238000013139 quantization Methods 0.000 claims description 12
- 230000008859 change Effects 0.000 claims description 5
- CIWBSHSKHKDKBQ-JLAZNSOCSA-N Ascorbic acid Chemical compound OC[C@H](O)[C@H]1OC(=O)C(O)=C1O CIWBSHSKHKDKBQ-JLAZNSOCSA-N 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 description 14
- 238000012360 testing method Methods 0.000 description 11
- 230000033001 locomotion Effects 0.000 description 9
- 238000006243 chemical reaction Methods 0.000 description 8
- 238000011160 research Methods 0.000 description 7
- 230000003044 adaptive effect Effects 0.000 description 6
- 230000006835 compression Effects 0.000 description 4
- 238000007906 compression Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000013507 mapping Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 241000023320 Luma <angiosperm> Species 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 230000000368 destabilizing effect Effects 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- OSWPMRLSEDHDFF-UHFFFAOYSA-N methyl salicylate Chemical compound COC(=O)C1=CC=CC=C1O OSWPMRLSEDHDFF-UHFFFAOYSA-N 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000010181 polygamy Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- H04N19/134—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
-
- 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
- H04N19/157—Assigned coding mode, i.e. the coding mode being predefined or preselected to be further used for selection of another element or parameter
- H04N19/159—Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
-
- 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
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
-
- 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
- H04N19/70—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
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 quick code check code-transferring method based on HEVC SCC and its system, including building HEVC SCC code check transcoding frameworks, corresponding depth when obtaining depth and the decoding of current coded unit, the pattern information of the adjacent decoding unit of corresponding decoding unit when the pattern information of the adjacent encoded unit of statistics current coded unit and decoding;The depth of the number of IntraBC patterns and the number of Normal Intra patterns and adjacent encoded unit in the pattern of the adjacent encoded unit of current coded unit is recorded, the number of IntraBC patterns and Normal Intra patterns in the pattern of corresponding adjacent decoding unit when current coded unit is decoded is recorded;Calculate the mean depth of the adjacent encoded unit of current coded unit, during decoding in corresponding adjacent decoding unit IntraBC patterns probability;Decided whether to skip IntraBC patterns and/or Normal Intra patterns according to above parameter information.The present invention ensure that the quality of video image while largely accelerating the speed of transcoding.
Description
Technical field
The present invention relates to coding and decoding video field, more particularly to the quick code check transcoding algorithm based on HEVC-SCC.
Background technology
In HEVC coding standards, in order to improve compression ratio, a width input picture is first divided into the figure of predefined size
As block, abbreviation code tree unit (Coding Tree Unit, CTU).Each CTU can be by several coding unit (Coding
Unit, CU) composition, luma samples and correspondence chroma sample of the CU in the presence of 8 × 8,16 × 16,32 × 32,64 × 64.CU can be with
Be decomposed into smaller predicting unit (Prediction Unit, PU) and converter unit (Transform Unit, TU), make coding,
Prediction and conversion are preferably handled;Intra mode decision direction is 35 kinds by 9 kinds of mode expansions H.264/AVC.According to system
Meter, the time-consuming of HEVC model selections accounts for the 60~70% of the whole scramble time.HEVC also introduces advanced motion vector in advance simultaneously
Survey (Advance Motion Vector Prediction, AMVP), Merge patterns and Skip patterns based on Merge patterns,
The technologies such as sampling point adaptive equalization (Sample Adaptive Offset, SAO).Although these technologies can improve compression ratio and
Good effect is brought on video quality, but adds the complexity of HEVC encoders, the difficulty of practicality is increased.Whether
The HEVC standard at initial stage or initial H.264/AVC standard are all directed to compress the natural image video sequence of camera capture
Row.Although only having used several different cycle tests when these standards of research and development, natural image sequence has
Identical feature and such as all there is sensor noise, there is abundant translational motion etc..But have recently emerged some extensions
Application, such as Wireless Display, entertainment systems, remote desktop, remote game, Desktop Share, long-distance education, video conference should
With they need to use video equipment to show more contents in addition to natural image.These contents are included by natural image, meter
The mixed display contents such as calculation machine figure, Wireless Display, tablet personal computer and automobile are shown, Screen sharing.In the video of these applications
A large amount of static or motion computer graphic image and text of the type of appearance including comprising natural image content.However,
Different from natural image content, screen content does not include sensor noise generally, and this content may have a large amount of uniform flats
There is numerical value identical block or region in the color of region, the pattern repeated, high saturation or limited quantity, picture.In view of this
The development applied a bit, HEVC is also carrying out constantly extension and perfect, HEVC screen content coding (Screen content
Coding, SCC) extension (HEVC-SCC) is content that JCT-VC is being researched and developed recently, is assessing SCC motions in April, 2014
After multiple technologies, screen content coding standard is turned to HEVC extension, in July, 2014, HEVC screen contents encoded draft 1 just
Formula is promulgated.The main purpose of HEVC-SCC standardization efforts is can further to improve to render comprising a large amount of static or motion
Figure, the compression performance of video that the natural image of text (not being animation), also camera capture is mixed.HEVC-
SCC performance can save more bit rates compared with existing HEVC versions under identical distortion.HEVC-SCC is in original
Have and introduce many new technologies on the basis of HEVC standard, it is desirable to which be derived from encoding screen content is more preferable Encoding
Energy.These technologies mainly include:Intra block replicate (Intra block copy, IBC), pallet mode (Palette mode,
PLT), adaptive color space conversion (Adaptive colour transform, ACT) and adaptive motion vector parsing
(Adaptive motion vector resolution, AMVR).These technologies are in the technical extensions of former HEVC, very
The compression ratio of screen content video is improved in big degree, in still further aspect, the new technology of addition more adds coding
The complexity of device.So, in actual applications, in order to preferably promote the development and application that HEVC screen contents are encoded, research
It is inevitable, urgently to be resolved hurrily that fast algorithm, which reduces its algorithm complex, accelerates its coding rate, is significant.
Although the fast algorithm for studying Video coding is particularly significant come the complexity for reducing algorithm, HEVC-SCC standards
Application include the software that the electronic equipment such as palm PC, smart mobile phone is supported, particularly video conference, long-distance education etc. and apply
Client be in different bandwidth network or use different video standards.It is then desired to will be regarded using Video Transcoding Technology
Frequency circulation changes the video flowings of different code checks into, to meet the demand of different network and various users, so as to realize different nets
Seamless connection between network, different terminal equipment, this is accomplished by studying Video Transcoding Technology.Video Transcoding Technology is i.e. by video flowing
Be converted into another form from a kind of form, it is therefore an objective to make between different video form, between different access devices, it is all kinds of different
Energy slitless connection, communication are smooth between matter network.Video code conversion is broadly divided into two major classes:One class is turn between various criterion
Code, another kind of is the transcoding in standard of the same race.Video code conversion can be divided into spatial resolution transcoding, time point by function and purposes
The classes such as resolution transcoding, code check transcoding, syntax conversion, video code model conversion.Main research drop in code check transcoding, practical application
Code check transcoding, the purpose is to effectively reduce code check on the premise of low complex degree and high video quality is kept, makes in practical application
The code check of output meets the condition of network condition or user equipment.It is wide that code check transcoding is mainly used in Internet Streaming Media, TV
The transmission field such as broadcast.Drop main consideration in the research process of code check transcoding and reduce error drift and reduction transcoding complexity.But
Existing Video Transcoding Technology is mainly transcoding of the research based on natural image video, and most fast transcoding method is all uncomfortable
It is too poor for screen content video, or applied to screen content video rear video transcoding effect.Therefore research is based on HEVC screens
The curtain quick code check transcoding technology of audio content, the development to video technique has important practical significance.
At present, the transcoding efficiency of pixel domain code conversion is generally all than relatively low, and the resource of consumption is than larger, and equipment is relatively held high
It is expensive, tend not to be satisfied with actual requirement of engineering.And existing video code rate transcoding technology is mainly research and is based on nature
The transcoding of image/video, existing most of method does not apply to screen content video, and there is presently no based on HEVC screen contents
Quick code check transcoding in terms of research.
The content of the invention
Present invention aims at a kind of quick code check code-transferring method based on HEVC-SCC is provided, with the existing transcoding of solution
Method is not suitable for the prior art problem of screen content video code conversion.
To achieve the above object, the invention provides a kind of quick code check code-transferring method based on HEVC-SCC, including with
Lower step:
HEVC-SCC code check transcoding frameworks are built, when obtaining depth and the current coded unit decoding of current coded unit
The depth of corresponding decoding unit, counts the pattern information and present encoding list of the encoded unit adjacent with current coded unit
The pattern information of the adjacent decoding unit of corresponding decoding unit during member decoding;
The number and Normal of IntraBC patterns in the pattern of the record encoded unit adjacent with current coded unit
The depth of the number of Intra patterns and adjacent encoded unit, while corresponding solution when recording current coded unit decoding
The number and the number of Normal Intra patterns of IntraBC patterns in the pattern of the adjacent decoding unit of code unit;
The mean depth of the encoded unit adjacent with current coded unit is calculated, current coded unit is corresponding when decoding
The probability of IntraBC patterns in the adjacent decoding unit of decoding unit;
According to the depth of corresponding decoding unit, transcoding code when the depth of current coded unit, current coded unit decoding
Rate quantifies the adjacent encoded unit of the mean depth of the adjacent encoded unit of poor, current coded unit, current coded unit
Pattern in IntraBC patterns number and Normal Intra patterns number, current coded unit decoding when corresponding solution
The number of IntraBC patterns and the number of Normal Intra patterns and current in the pattern of the adjacent decoding unit of code unit
The probability of InrtaBC patterns decides whether to skip in the adjacent decoding unit of corresponding decoding unit when coding unit is decoded
InrtaBC patterns and/or Normal Intra patterns.
The above method is relied on, the invention also provides a kind of speed code based on HEVC-SCC corresponding with the above method
Rate trans-coding system, including with lower module:
Data acquisition module:For obtaining corresponding solution when the depth and current coded unit of current coded unit are decoded
The depth of code unit, pattern information and current coded unit decoding for counting the adjacent encoded unit of current coded unit
When corresponding decoding unit adjacent decoding unit pattern information, the encoded unit adjacent for recording current coded unit
Pattern in the number of IntraBC patterns and the number of Normal Intra patterns and adjacent encoded unit depth,
And for record current coded unit decoding when corresponding decoding unit adjacent decoding unit pattern in IntraBC patterns
Number and Normal Intra patterns number;
Computing module:Mean depth for calculating the adjacent encoded unit of current coded unit, current coded unit
During decoding in the adjacent decoding unit of corresponding decoding unit IntraBC patterns probability;
Decision-making module:Corresponding decoding unit during for the depth according to current coded unit, current coded unit decoding
Depth, that transcoding code check quantifies the mean depth of the adjacent encoded unit of poor, current coded unit, current coded unit is adjacent
Encoded unit pattern in IntraBC patterns number and number, the current coded unit solution of Normal Intra patterns
During code in the pattern of the adjacent decoding unit of corresponding decoding unit IntraBC patterns number and Normal Intra patterns
The probability of IntraBC patterns is determined in the adjacent decoding unit of corresponding decoding unit when number and current coded unit decoding
It is fixed whether to skip IntraBC patterns and/or Normal Intra patterns.
The invention has the advantages that:
The present invention is directed to HEVC-SCC code check transcoding frameworks, it is proposed that fast suitable for the infra-frame prediction of screen content screen
Fast code check transcoding algorithm, the speed of transcoding is largely accelerated by the fast transcoding algorithm of proposition, while multiple not increasing
The quality of video image is ensure that on the premise of polygamy.
Below with reference to accompanying drawings, the present invention is further detailed explanation.
Brief description of the drawings
The accompanying drawing for constituting the part of the application is used for providing a further understanding of the present invention, schematic reality of the invention
Apply example and its illustrate to be used to explain the present invention, do not constitute inappropriate limitation of the present invention.In the accompanying drawings:
Fig. 1 is a kind of quick code check transcoding algorithm flow chart based on HEVC-SCC of the preferred embodiment of the present invention;
Fig. 2 is the HEVC-SCC code check transcoding architecture figures of the preferred embodiment of the present invention;
Fig. 3 is the CU mapping graphs of the preferred embodiment of the present invention;
Fig. 4 is preferred embodiment of the present invention coding unit model selection fast transcoding algorithm flow chart;
Fig. 5 is preferred embodiment of the present invention predicting unit model selection fast transcoding algorithm flow chart.
Embodiment
Embodiments of the invention are described in detail below in conjunction with accompanying drawing, but the present invention can be defined by the claims
Implement with the multitude of different ways of covering.
Embodiment 1:
Code check transcoding can be divided into CBR (constant bitrate, cbr (constant bit rate)) transcodings and VBR (variable
Bitrate, variable bit rate).CBR transcodings refer to a cbr (constant bit rate) (Ri) input video stream be transcoded into separately by transcoder
One cbr (constant bit rate) (Ro) target video stream output.Wherein, RiAnd RoAll set in advance, encoder is exported mesh
The video flowing of coding rate, needs to open Rate Control switch in the encoder.VBR transcodings be then by change quantization parameter QP come
Realize code check transcoding, it is not necessary to open Rate Control.When QP is smaller, the video stream bit rate of output is larger, defeated when QP is larger
The video code rate gone out is smaller.Therefore, when carrying out VBR transcodings, the QP of video flowing is commonly enterediIt is less than outputting video streams
QPo, can quantify difference Δ QP to control the probable ranges of code check by both transcoding code checks of adjustment, Δ QP can be defined as
Δ QP=QPo-QPi, what the present invention was mainly studied is VBR transcodings." IntraBC " pattern refers to Intra Block in the present embodiment
Copy, frame in copy block pattern;" Normal Intra patterns " refers to traditional frame mode.
As shown in figure 1, present embodiment discloses a kind of quick code check code-transferring method based on HEVC-SCC, including:
Step S1, build HEVC-SCC code check transcoding frameworks, obtain the depth and present encoding list of current coded unit
The depth of corresponding decoding unit during member decoding, counts the pattern information of the encoded unit adjacent with current coded unit and works as
The pattern information of the adjacent decoding unit of corresponding decoding unit during preceding coding unit decoding.Traditional transcoding architecture be by
HEVC-SCC decoders and encoder are in series, and high bit-rate video code stream first passes around decoder decoding and obtains unpressed regard
Frequency image sequence, subsequently into encoder encode obtaining low bit-rate video code stream.It is this simple for structure and flexible, no
Need the basic structure of change decoder and encoder, it is ensured that the independence and integrality of decoder and encoder.But this
The complexity planted when transcoding architecture is realized is higher.In order to reduce its complexity, the present invention proposes HEVC- as shown in Figure 2
SCC code check transcoding architecture figures, encoder using decode come pattern, MV, CU (coding unit) depth and PU (predicting unit)
The information such as pattern simplify cataloged procedure, lift the efficiency of transcoding.The present invention needs to obtain the depth D of current coded unitcurWith
And current coded unit decoding when corresponding decoding unit depth Ddec, the adjacent encoded unit of statistics current coded unit
Pattern information and the pattern information of the adjacent decoding unit of corresponding decoding unit is follow-up step during current coded unit decoding
It is rapid used.Fig. 3 gives the mapping relations figure of current coded unit and current coded unit in decoding end in present frame.Left side is
Decoding frame, right side is present frame.
In the pattern of step S2, the record encoded unit adjacent with current coded unit the number of IntraBC patterns and
The depth of the number of Normal Intra patterns and adjacent encoded unit, at the same record current coded unit decoding when pair
The number of the number of IntraBC patterns and Normal Intra patterns in the pattern of the adjacent decoding unit for the decoding unit answered.
Record the number N of IntraBC patterns in the pattern of the adjacent encoded unit of current coded unitcurIBCWith Normal Intra
The number N of patterncurIntraAnd the depth of adjacent encoded unit, while corresponding solution when recording current coded unit decoding
The number N of IntraBC patterns in the pattern of the adjacent decoding unit of code unitdecIBCWith the number of Normal Intra patterns
NdecIntra, and define flag1 and flag0.
The mean depth of step S3, the calculating encoded unit adjacent with current coded unit, current coded unit decoding
When corresponding decoding unit adjacent decoding unit in IntraBC patterns probability.Calculate the adjacent volume of current coded unit
The mean depth D of code unitcurAve, IntraBC in the adjacent decoding unit of corresponding decoding unit when current coded unit is decoded
The probability R of patternIBC。
Step S4, the depth according to current coded unit, current coded unit decoding when corresponding decoding unit depth,
Transcoding code check quantifies the adjacent volume of the mean depth of the adjacent encoded unit of poor, current coded unit, current coded unit
Code unit pattern in IntraBC patterns number and Normal Intra patterns number, current coded unit decoding when pair
In the pattern of the adjacent decoding unit for the decoding unit answered the number of the number of IntraBC patterns and Normal Intra patterns with
And the probability of InrtaBC patterns decides whether in the adjacent decoding unit of corresponding decoding unit during current coded unit decoding
Skip InrtaBC patterns and/or Normal Intra patterns.Referring to Fig. 4 (a) and Fig. 4 (b), decide whether to skip IntraBC moulds
Formula and/or Normal Intra patterns follow following rule:
(1) if decoding when decoding unit depth Ddec=3, then current coded unit depth bounds is Dcur={ 2,3 },
All pattern rate-distortion optimizations that depth is 0 and 1 layer are skipped when encoding to calculate;
(2) if decoding when decoding unit depth Ddec=0, as Δ QP < λ, then current coded unit depth bounds
For Dcur={ 0,1 }, skips all pattern rate-distortion optimizations that depth is 2 and 3 layers and calculates during coding;As Δ QP >=λ, and
The mean depth D of current coded unit adjacent encoder unitcurAve< 0.5, skips all patterns that depth is 2 and 3 layers during coding
Rate-distortion optimization is calculated, and wherein Δ QP is that the quantization of transcoding code check is poor, and λ is constant;
(3) if decoding when decoding unit depth Ddec=1, as Δ QP < λ, then current coded unit depth bounds
For Dcur={ 0,1,2 }, all pattern rate-distortion optimizations that the layer that depth is 3 is skipped during coding are calculated;As Δ QP >=λ, and
The mean depth D of current coded unit adjacent encoder unitcurAve< 1.5, skips all patterns for the layer that depth is 3 during coding
Rate-distortion optimization is calculated;
(4) if decoding when Ddec> Dcur, and the pattern of correspondence decoding unit is during the decoding of current coded unit
IntraBC, meeting following condition, one of those then skips Normal Intra patterns:
1. (Δ QP < λ &&N are metdecIBC≥2)||NcurIBC≥3||NdecIBC>=3, Normal Intra patterns are skipped, it is no
Then perform 2.;
2. Δ QP >=λ &&N are metcurIBC>=2, Normal Intra patterns are skipped, are otherwise performed 3.;
3. R is metIBC> 0.6 | | NcurIBC+NdecIBC(the N of < 0.8curIntra+NdecIntra), skip Normal Intra moulds
4. formula, otherwise perform;
4. N is metcurIBC≥1&&NdecIBC>=1, Normal Intra patterns are skipped, otherwise terminate wherein NdecIBCTo be current
Coding unit decode when corresponding decoding unit adjacent decoding unit pattern in IntraBC patterns number, NdecIntraFor
Current coded unit decode when corresponding decoding unit adjacent decoding unit pattern in Normal Intra patterns number,
NcurIBCFor the number of IntraBC patterns in the pattern of the adjacent encoded unit of current coded unit, NcurIntraFor present encoding
The number of Normal Intra patterns in the pattern of the adjacent encoded unit of unit;
(5) D is worked ascur< 3&&RIBCDuring < 0.05, IntraBC patterns are skipped, otherwise depth D upon decodingdec> Dcur,
And pattern is Normal Intra during the decoding of current coded unit, meeting following condition, one of those then skips IntraBC
Pattern, otherwise performs step (6):
1. (Δ QP < λ &&N are metdecIntra≥3)||(ΔQP≥λ&&NcurIntra>=3), and IntraBC patterns are skipped, it is no
Then perform 2.;
2. N is metcurIntra≥1&&NdecIntra>=1, IntraBC patterns are skipped, are otherwise performed 3.;
3. R is metIBC< 0.2 | | NcurIBC+NdecIBC(the N of > 1.2curIntra+NdecIntra), IntraBC patterns are skipped, otherwise
Perform (6);
(6) when meeting Ddec=Dcur+ 2 and current coded unit decoding when pattern be Normal Intra, skip
IntraBC patterns, otherwise perform (7);
(7) encoded coding unit selection IntraBC and Normal Intra average rate distortion value is calculated, will
The rate distortion value of optimal mode after Normal Intra model selections as Rule of judgment, if:
BestCU_Cost < min (AvgNorCost, AvgIntraBCCost)
Then enter following condition judgment, otherwise do not enter fast algorithm, be normally carried out IntraBC model selections, wherein
AvgNorCost selects IntraBC average rate distortion value for encoded coding unit, and AvgIntraBCCost is encoded
Coding unit selection Normal Intra average rate distortion value, BestCU_Cost be Normal Intra model selections it
The rate distortion value of optimal mode afterwards:
If 1. RIBC< 0.1, then skip IntraBC patterns, otherwise performs 2.;
If 2. RIBCThe & & N of < 0.5curIntra>=3, then IntraBC patterns are skipped, otherwise do not enter fast algorithm, normally
Carry out IntraBC model selections.
Further, the pattern information and/or current coded unit of the adjacent encoded unit of the current coded unit
The pattern information of the adjacent decoding unit of corresponding decoding unit is specially encoded unit and/or current coded unit solution during decoding
The pattern information of the adjacent decoding unit top of corresponding decoding unit during code, upper left side, upper right side and Left prediction unit.Choosing
It is that, because close from current coded unit, spatial coherence is strong, and standard is predicted using these information to take these predicting units
True rate is higher.
Further, referring to Fig. 5, the transcoding algorithm of predicting unit is described as follows:
(1) decoding schema S when current prediction unit is decoded in the information that decoding end is preserved is obtaineddec;
(2) if Sdec=2Nx2N, then coding side predicting unit selection 2Nx2N patterns, skip other patterns;Otherwise perform
(3);
(3) if Sdec=NxN, skips 2NxN patterns;If Sdec=NxN, and transcoding code check quantization difference is less than λ and adjacent four
Predicting unit pattern is NxN number NNxN>=3, then 2NxN patterns are skipped, otherwise (4) are performed;
(4) if Sdec=Nx2N, then skip 2NxN patterns, and any pattern is not otherwise skipped and is normally carried out model selection.
Each pattern of PU 2Nx2N, 2NxN, Nx2N, NxN RDcost can be calculated in HEVC-SCC video coding process, so
Compare minimum using the RDcost of which mode computation afterwards, the final minimum pattern of loss of choosing is used as optimal mode.The algorithm
The main difference according to QP is come the PU patterns or the adjacent PU patterns of current block with reference to the corresponding solution code block of current block come directly
Some PU patterns RDcost calculating is skipped quickly to select current PU pattern.
Further, the poor value of transcoding quantization is λ=5.As shown in table 1, wherein DcurWhat is represented is current CU depth,
DdecRefer to corresponding CU depth during current CU decodings.Corresponding CU depth when current CU depth is decoded with it is reflected in table
Correlation, i.e., depth D upon decodingdecDuring={ 0,1,2,3 }, current CU depth DcurSelected respectively in Δ QP=2 and Δ QP=4
Select accuracy rate when { 0,1,2,3 }.As shown in table 2, PcurWhat is represented is the pattern of current PU (predicting unit), PdecRepresent be
Corresponding PU patterns when current block is decoded.The correlation of corresponding PU patterns when current PU patterns are decoded with it is reflected in table,
PU patterns i.e. upon decoding are respectively 2Nx2N, 2NxN, Nx2N, during NxN, current block selection 2Nx2N, 2NxN, Nx2N, during NxN
Probability, that is, when doing PU model selection, current block entirely by reference to decoded information progress PU model selections when standard
True rate.By Tables 1 and 2 analysis it is recognised that when the YUV to decoding is re-encoded, may be referred to the letter decoded
Breath carries out CU depth prediction, model selection.However, it is also possible to the difference according to transcoding Δ QP is found, reference decoder information
Accuracy rate is different, and Δ QP is bigger, and the correlation of coding information and decoding is smaller, therefore the letter referring only to decoding of blindness
Breath, which carries out coding, may bring larger error.In order on the premise of the subjective quality for ensureing the video that transcoding goes out is constant,
The quickening transcoding speed of most maximal efficiency.Fast mode decision transcoding algorithm based on QP and CU depth is the difference according to Δ QP
To select reference decoder CU information or current CU adjacent C U information, or with reference to decoding CU information and current CU adjacent Cs U
Information is come to carrying out fast mode decision.I.e. Normal Intra patterns or IntraBC patterns are skipped in selection.According to data in table
Analysis is drawn:Δ QP is bigger, and the correlation of coding information and decoding is smaller, now can be adjacent with reference to current coded unit emphatically
CU information be predicted can be more accurate.Therefore Δ QP < λ are worked as, emphatically reference decoder frame information;As Δ QP >=λ, join emphatically
The coding information of present frame is examined, takes λ=5 the most suitable by analysis of experiments.
Table 1
Table 2
The above method is relied on, the invention also provides one kind corresponding with the above method is based on the quick code checks of HEVC-SCC
Trans-coding system, including with lower module:
Data acquisition module:For obtaining corresponding solution when the depth and current coded unit of current coded unit are decoded
The depth of code unit, pattern information and current coded unit decoding for counting the adjacent encoded unit of current coded unit
When corresponding decoding unit adjacent decoding unit pattern information, the encoded unit adjacent for recording current coded unit
Pattern in the number of IntraBC patterns and the number of Normal Intra patterns and adjacent encoded unit depth,
And for record current coded unit decoding when corresponding decoding unit adjacent decoding unit pattern in IntraBC patterns
Number and Normal Intra patterns number.
Computing module:Mean depth for calculating the adjacent encoded unit of current coded unit, current coded unit
During decoding in the adjacent decoding unit of corresponding decoding unit IntraBC patterns probability.
Decision-making module:Corresponding decoding unit during for the depth according to current coded unit, current coded unit decoding
Depth, that transcoding code check quantifies the mean depth of the adjacent encoded unit of poor, current coded unit, current coded unit is adjacent
Encoded unit pattern in IntraBC patterns number and number, the current coded unit solution of Normal Intra patterns
During code in the pattern of the adjacent decoding unit of corresponding decoding unit IntraBC patterns number and Normal Intra patterns
The probability of IntraBC patterns is determined in the adjacent decoding unit of corresponding decoding unit when number and current coded unit decoding
It is fixed whether to skip IntraBC patterns and/or Normal Intra patterns.
Further, decision-making module decides whether that skipping IntraBC patterns and/or Normal Intra patterns follows following
Rule:
(1) if decoding when decoding unit depth Ddec=3, then current coded unit depth bounds is Dcur={ 2,3 },
All pattern rate-distortion optimizations that depth is 0 and 1 layer are skipped when encoding to calculate;
(2) if decoding when decoding unit depth Ddec=0, as Δ QP < λ, then current coded unit depth bounds
For Dcur={ 0,1 }, skips all pattern rate-distortion optimizations that depth is 2 and 3 layers and calculates during coding;As Δ QP >=λ, and
The mean depth D of current coded unit adjacent encoder unitcurAve< 0.5, skips all patterns that depth is 2 and 3 layers during coding
Rate-distortion optimization is calculated, and wherein Δ QP is that the quantization of transcoding code check is poor, and λ is constant;
(3) if decoding when decoding unit depth Ddec=1, as Δ QP < λ, then current coded unit depth bounds
For Dcur={ 0,1,2 }, all pattern rate-distortion optimizations that the layer that depth is 3 is skipped during coding are calculated;As Δ QP >=λ, and
The mean depth D of current coded unit adjacent encoder unitcurAve< 1.5, skips all patterns for the layer that depth is 3 during coding
Rate-distortion optimization is calculated;
(4) D upon decodingdec> Dcur, and the pattern of decoding unit is IntraBC during the decoding of current coded unit, it is full
Being enough lower condition, one of those then skips Normal Intra patterns:
1. (Δ QP < λ &&N are metdecIBC≥2)||NcurIBC≥3||NdecIBC>=3, Normal Intra patterns are skipped, it is no
Then perform 2.;
2. Δ QP >=λ & &N are metcurIBC>=2, Normal Intra patterns are skipped, are otherwise performed 3.;
3. R is metIBC> 0.6 | | NcurIBC+NdecIBC(the N of < 0.8curIntra+NdecIntra), skip Normal Intra moulds
4. formula, otherwise perform;
4. N is metcurIBC≥1&&NdecIBC>=1, Normal Intra patterns are skipped, are otherwise terminated, wherein NdecIBCTo work as
During the decoding of preceding coding unit in the pattern of the adjacent decoding unit of corresponding decoding unit IntraBC patterns number, NdecIntra
When being decoded for current coded unit in the pattern of the adjacent decoding unit of corresponding decoding unit Normal Intra patterns
Number, NcurIBCFor the number of IntraBC patterns in the pattern of the adjacent encoded unit of current coded unit, NcurIntraTo be current
The number of Normal Intra patterns in the pattern of the adjacent encoded unit of coding unit;
(5) D is worked ascur< 3&&RIBCDuring < 0.05, IntraBC patterns are skipped, otherwise depth D upon decodingdec> Dcur,
And pattern is Normal Intra during the decoding of current coded unit, meeting following condition, one of those then skips IntraBC
Pattern, otherwise performs step (6):
1. (Δ QP < λ &&N are metdecIntra≥3)||(ΔQP≥λ&&NcurIntra>=3), and IntraBC patterns are skipped, it is no
Then perform 2.;
2. N is metcurIntra≥1&&NdecIntra>=1, IntraBC patterns are skipped, are otherwise performed 3.;
3. R is metIBC< 0.2 | | NcurIBC+NdecIBC(the N of > 1.2curIntra+NdecIntra), IntraBC patterns are skipped, otherwise
Perform (6), wherein RIBCIntraBC patterns in the adjacent decoding unit of corresponding decoding unit when being decoded for current coded unit
Probability;
(6) when meeting Ddec=Dcur+ 2 and current coded unit decoding when pattern be Normal Intra, skip
IntraBC patterns, otherwise perform (7);
(7) encoded coding unit selection IntraBC and Normal Intra average rate distortion value is calculated, will
The rate distortion value of optimal mode after Normal Intra model selections as Rule of judgment, if:
BestCU_Cost < min (AvgNorCost, AvgIntraBCCost)
Then enter following condition judgment, otherwise do not enter fast algorithm, be normally carried out IntraBC model selections, wherein
AvgNorCost selects IntraBC average rate distortion value for encoded coding unit, and AvgIntraBCCost is encoded
Coding unit selection Normal Intra average rate distortion value, BestCU_Cost be Normal Intra model selections it
The rate distortion value of optimal mode afterwards:
If 1. RIBC< 0.1, then skip IntraBC patterns, otherwise performs 2.;
If 2. RIBC< 0.5&&NcurIntra>=3, then IntraBC patterns are skipped, does not otherwise enter fast algorithm, normally enters
Row IntraBC model selections.
Further, the pattern information of the adjacent encoded unit of the current coded unit of data acquisition module acquisition
And/or the pattern information of the adjacent decoding unit of corresponding decoding unit is specially encoded unit during current coded unit decoding
And/or during current coded unit decoding above the adjacent decoding unit of corresponding decoding unit, upper left side, upper right side and left are pre-
Survey the pattern information of unit.
Further, the transcoding algorithm for the predicting unit that data acquisition module is obtained is as follows:
(1) decoding schema S when current prediction unit is decoded in the information that decoding end is preserved is obtaineddec;
(2) if Sdec=2Nx2N, then coding side predicting unit selection 2Nx2N patterns, skip other patterns, otherwise perform
(3);
(3) if Sdec=NxN, skips 2NxN patterns;If Sdec=NxN, and transcoding quantization difference is less than λ and adjacent four predictions
Unit mode is NxN number NNxN>=3, then 2NxN patterns are skipped, otherwise (4) are performed;
(4) if Sdec=Nx2N, then skip 2NxN patterns, and any pattern is not otherwise skipped and is normally carried out model selection.
Further, the poor value of transcoding quantization is λ=5.
In summary, the present embodiment proposes the full structural framing for compiling complete solution, Yi Jishi for HEVC-SCC code check transcodings
The quick code check transcoding algorithm of infra-frame prediction for screen content screen, is largely added by the fast transcoding algorithm of proposition
The speed of fast transcoding, while ensure that the quality of video image on the premise of complexity is not increased.
Embodiment 2:
In order to verify the correctness and validity of quick code check transcoding algorithm proposed by the invention, the present invention is based on
HM-16.2+SCM-3.0 has built pixel domain spatial resolution transcoder as shown in Figure 1, and is based in the transcoder
Visual studio 2013 realize quick code check transcoding algorithm.Finally, in collection experimental result data link, it is contemplated that pen
Remember that the destabilizing factor table of this computer is more, therefore in order to ensure the real reliability of experimental result, all experiments are in height
Complied executing in performance calculating platform.The platform employs hybrid-type cluster (Cluster) framework, and its calculating network is used
Infinband high speed switch, is effective to ensure that the stability and reliability of experimental result.Therefore, all realities of the invention
It is to carry out emulation testing in the environment in the absence of the stabilization of any interference to obtain to test data, with real reliability.This
The coding parameter of literary grace is consistent with the SCC standard test environments (Common Test Condition, CTC) that JCTVC is issued.
Experiment test is configured to full I frames, lossy coding, using 4 different quantization parameters, respectively 22,27,32 and 37 during coding.
The present invention organizes HEVC-SCC standards, the representative screen content cycle tests of issue using JCTVC
[62], 4 classes are divided into:The first kind is the text and figure of motion, represents typical remote desktop application content, including
The sequences such as WebBrowsing, Map and Programming;Equations of The Second Kind screen mixes content, including text, map and natural image
Deng screen content common in daily life, such as MissionControlClip3 sequences.3rd class is animated content, such as
Robot, represents the application contents such as high in the clouds game.4th class is natural image content, such as Kimonol.Cycle tests has RGB respectively
With YUV (4:4:4) form.The resolution ratio of cycle tests has two kinds of 1080p and 720p, and frame per second has 60,30,24,20 four kinds, such as table
Shown in 3.
The cycle tests sample table of table 3
Test result after final emulation reflects the property of algorithm using two parameters of BDBR and Δ Time as shown in table 4
Energy.Wherein BDBR is the performance indications for reflecting picture quality, using optimized algorithm the encoder obtained video of coding relative to
Situation of change of the video that original encoding device is obtained on code check and PSNR, bigger represent of BDBR is used after the algorithm, video image
Mass loss it is bigger, i.e., algorithm performance is poorer.Δ Time represents that the scramble time T of optimized algorithm is not usedorgWith using being carried
The scramble time T gone out after algorithmproDifference, computational methods are as follows, and the bigger effect of optimization for representing algorithm of the difference is more
Good, i.e. the performance of algorithm is better.
As can be seen from Table 4, BDBR average values are 1.78%, and Δ Time average values are 45.09%, therefore can draw knot
By:The fast algorithm of the present invention is evaluated from objective quality, it may be said that bright fast transcoding algorithm of the invention is achieved
Good effect, i.e., under the premise of the subjective quality of video image is constant, time reduction 45.09%, and quality objective quality
Loss almost can be ignored.
Table 4
To sum up, this implementation row are disclosed is based on the quick code check code-transferring methods of HEVC-SCC and its system, has built HEVC-
SCC code check transcoding frameworks, based on HEVC-SCC code check transcoding frameworks, it is proposed that the quick code check transcoding algorithm of infra-frame prediction, pass through
The fast transcoding algorithm of proposition largely accelerates the speed of transcoding, and video figure is also assures that while transcoding speed is accelerated
The quality of picture.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, for the skill of this area
For art personnel, the present invention can have various modifications and variations.Within the spirit and principles of the invention, that is made any repaiies
Change, equivalent substitution, improvement etc., should be included in the scope of the protection.
Claims (10)
1. a kind of quick code check code-transferring method based on HEVC-SCC, it is characterised in that comprise the following steps:
HEVC-SCC code check transcoding frameworks are built, correspondence when obtaining depth and the current coded unit decoding of current coded unit
Decoding unit depth, count the pattern information and current coded unit solution of the encoded unit adjacent with current coded unit
The pattern information of the adjacent decoding unit of corresponding decoding unit during code;
The number and Normal Intra of IntraBC patterns in the pattern of the record encoded unit adjacent with current coded unit
The depth of the number of pattern and adjacent encoded unit, while corresponding decoding unit when recording current coded unit decoding
Adjacent decoding unit pattern in IntraBC patterns number and the number of Normal Intra patterns;
Calculate the mean depth of the encoded unit adjacent with current coded unit, corresponding decoding when current coded unit is decoded
The probability of IntraBC patterns in the adjacent decoding unit of unit;
According to depth, the transcoding code check amount of corresponding decoding unit when the depth of current coded unit, current coded unit decoding
The mould of the adjacent encoded unit of the mean depth of the adjacent encoded unit of the poor, current coded unit of change, current coded unit
Corresponding decoding is single when the number of the number of IntraBC patterns and Normal Intra patterns, current coded unit decoding in formula
The number and the number and present encoding of Normal Intra patterns of IntraBC patterns in the pattern of the adjacent decoding unit of member
The probability of InrtaBC patterns decides whether to skip in the adjacent decoding unit of corresponding decoding unit when unit is decoded
InrtaBC patterns and/or Normal Intra patterns.
2. the quick code check code-transferring method according to claim 1 based on HEVC-SCC, it is characterised in that the decision is
It is no to skip IntraBC patterns and/or Normal Intra patterns follow following rule:
(1) if decoding when decoding unit depth Ddec=3, then current coded unit depth bounds is Dcur={ 2,3 }, that is, compile
All pattern rate-distortion optimizations that depth is 0 and 1 layer are skipped during code to calculate;
(2) if decoding when decoding unit depth Ddec=0, as Δ QP < λ, then current coded unit depth bounds is Dcur
={ 0,1 }, skips all pattern rate-distortion optimizations that depth is 2 and 3 layers and calculates during coding;As Δ QP >=λ, and currently compile
The mean depth D of code unit adjacent encoder unitcurAve< 0.5, skips all pattern rate distortions that depth is 2 and 3 layers during coding
Optimization is calculated, and wherein Δ QP is that the quantization of transcoding code check is poor, and λ is constant;
(3) if decoding when decoding unit depth Ddec=1, as Δ QP < λ, then current coded unit depth bounds is Dcur
={ 0,1,2 }, all pattern rate-distortion optimizations that the layer that depth is 3 is skipped during coding are calculated;As Δ QP >=λ, and currently
The mean depth D of coding unit adjacent encoder unitcurAve< 15, all pattern rates that the layer that depth is 3 is skipped during coding are lost
True optimization is calculated;
(4) if decoding when Ddec> Dcur, and the pattern of correspondence decoding unit is IntraBC during the decoding of current coded unit,
Meeting following condition, one of those then skips Normal Intra patterns:
1. (Δ QP < λ &&N are metdecIBC≥2)||NcurIBC≥3||NdecIBC>=3, Normal Intra patterns are skipped, are otherwise held
Row is 2.;
2. Δ QP >=λ &&N are metcurIBC>=2, Normal Intra patterns are skipped, are otherwise performed 3.;
3. R is metIBC> 0.6 | | NcurIBC+NdecIBC(the N of < 0.8curIntra+NdecIntra), Normal Intra patterns are skipped, otherwise
Perform 4.;
4. N is metcurIBC≥1&&NdecIBC>=1, Normal Intra patterns are skipped, are otherwise terminated, wherein NdecIBCFor current volume
Code unit decoding when corresponding decoding unit adjacent decoding unit pattern in IntraBC patterns number, NdecIntraTo work as
During the decoding of preceding coding unit in the pattern of the adjacent decoding unit of corresponding decoding unit Normal Intra patterns number,
NcurIBCFor the number of IntraBC patterns in the pattern of the adjacent encoded unit of current coded unit, NcurIntraFor present encoding
The number of Normal Intra patterns in the pattern of the adjacent encoded unit of unit;
(5) D is worked ascur< 3&&RIBCDuring < 0.05, IntraBC patterns are skipped, otherwise depth D upon decodingdec> Dcur, and work as
Pattern is Normal Intra during the decoding of preceding coding unit, and meeting following condition, one of those then skips IntraBC patterns,
Otherwise step (6) is performed:
1. (Δ QP < λ &&N are metdecIntra≥3)||(ΔQP≥λ&&NcurIntra>=3), IntraBC patterns are skipped, are otherwise performed
②;
2. N is metcurIntra≥1&&NdecIntra>=1, IntraBC patterns are skipped, are otherwise performed 3.;
3. R is metIBC< 0.2 | | NcurIBC+NdecIBC(the N of > 1.2curIntra+NdecIntra), IntraBC patterns are skipped, are otherwise performed
(6), wherein RIBCWhen being decoded for current coded unit in the adjacent decoding unit of corresponding decoding unit IntraBC patterns it is general
Rate;
(6) D is worked asdec=Dcur+ 2, and current coded unit decoding when pattern be Normal Intra, skip IntraBC patterns,
Otherwise (7) are performed;
(7) encoded coding unit selection IntraBC and Normal Intra average rate distortion value is calculated, by Normal
The rate distortion value of optimal mode after Intra model selections as Rule of judgment, if:
BestCU_Cost < min (AvgNorCost, AvgIntraBCCost)
Then enter following condition judgment, otherwise do not enter fast algorithm, be normally carried out IntraBC model selections, wherein
AvgNorCost selects IntraBC average rate distortion value for encoded coding unit, and AvgIntraBCCost is encoded
Coding unit selection Normal Intra average rate distortion value, BestCU_Cost be Normal Intra model selections it
The rate distortion value of optimal mode afterwards:
If 1. RIBC< 0.1, then skip IntraBC patterns, otherwise performs 2.;
If 2. RIBC< 0.5& & NcurIntra>=3, then IntraBC patterns are skipped, otherwise do not enter fast algorithm, are normally carried out
IntraBC model selections.
3. the quick code check code-transferring method according to claim 1 based on HEVC-SCC, it is characterised in that the current volume
Corresponding decoding unit adjacent decoding when the pattern information of the adjacent encoded unit of code unit and/or current coded unit decoding
The adjacent decoding of corresponding decoding unit when the pattern information of unit is specially encoded unit and/or current coded unit decoding
The pattern information of unit top, upper left side, upper right side and Left prediction unit.
4. the quick code check code-transferring method according to claim 3 based on HEVC-SCC, it is characterised in that the prediction list
The transcoding algorithm of member is as follows:
(1) decoding schema S when current prediction unit is decoded in the information that decoding end is preserved is obtaineddec;
(2) if Sdec=2Nx2N, then coding side predicting unit selection 2Nx2N patterns, skip other patterns, otherwise perform (3);
(3) if Sdec=NxN, skips 2NxN patterns;If Sdec=NxN, transcoding code check quantifies difference and is less than λ and adjacent four predictions list
Meta schema is NxN number NNxN>=3, then 2NxN patterns are skipped, otherwise (4) are performed;
(4) if Sdec=Nx2N, then skip 2NxN patterns, and any pattern is not otherwise skipped and is normally carried out model selection.
5. according to any described quick code check code-transferring methods based on HEVC-SCC of claim 1-4, it is characterised in that described
It is λ=5 that transcoding code check, which quantifies poor value,.
6. a kind of quick code check trans-coding system based on HEVC-SCC, it is characterised in that including with lower module:
Data acquisition module:It is single for obtaining corresponding decoding when the depth and current coded unit of current coded unit are decoded
Member depth, for count the adjacent encoded unit of current coded unit pattern information and current coded unit decoding when pair
The pattern information of the adjacent decoding unit for the decoding unit answered, the mould for recording the adjacent encoded unit of current coded unit
The depth of the number of the number of IntraBC patterns and Normal Intra patterns and adjacent encoded unit in formula, and
For recording IntraBC patterns in the pattern of the adjacent decoding unit of corresponding decoding unit when current coded unit is decoded
The number of number and NormalIntra patterns;
Computing module:Mean depth for calculating the adjacent encoded unit of current coded unit, current coded unit decoding
When corresponding decoding unit adjacent decoding unit in IntraBC patterns probability;
Decision-making module:The depth of corresponding decoding unit during for the depth according to current coded unit, current coded unit decoding
It is adjacent that degree, transcoding code check quantify the mean depth of the adjacent encoded unit of poor, current coded unit, current coded unit
In the pattern of coding unit when the number of the number of IntraBC patterns and Normal Intra patterns, current coded unit decoding
The number of the number of IntraBC patterns and Normal Intra patterns in the pattern of the adjacent decoding unit of corresponding decoding unit
And the probability of IntraBC patterns determines to be in the adjacent decoding unit of corresponding decoding unit during current coded unit decoding
It is no to skip IntraBC patterns and/or Normal Intra patterns.
7. the quick code check trans-coding system according to claim 6 based on HEVC-SCC, it is characterised in that decision-making module is determined
Determine whether to skip IntraBC patterns and/or Normal Intra patterns follow following rule:
(1) if decoding when decoding unit depth Ddec=3, then current coded unit depth bounds is Dcur={ 2,3 }, that is, compile
All pattern rate-distortion optimizations that depth is 0 and 1 layer are skipped during code to calculate;
(2) if decoding when decoding unit depth Ddec=0, as Δ QP < λ, then current coded unit depth bounds is Dcur
={ 0,1 }, skips all pattern rate-distortion optimizations that depth is 2 and 3 layers and calculates during coding;As Δ QP >=λ, and currently compile
The mean depth D of code unit adjacent encoder unitcurAve< 0.5, skips all pattern rate distortions that depth is 2 and 3 layers during coding
Optimization is calculated, and wherein Δ QP is that the quantization of transcoding code check is poor, and λ is constant;
(3) if decoding when decoding unit depth Ddec=1, as Δ QP < λ, then current coded unit depth bounds is Dcur
={ 0,1,2 }, all pattern rate-distortion optimizations that the layer that depth is 3 is skipped during coding are calculated;As Δ QP >=λ, and currently
The mean depth D of coding unit adjacent encoder unitcurAve< 1.5, all pattern rates that the layer that depth is 3 is skipped during coding are lost
True optimization is calculated;
(4) D upon decodingdec> Dcur, and the pattern of decoding unit is IntraBC during the decoding of current coded unit, meet with
Lower condition one of those then skip Normal Intra patterns:
1. (Δ QP < λ &&N are metdecIBC≥2)||NcurIBC≥3||NdecIBC>=3, Normal Intra patterns are skipped, are otherwise held
Row is 2.;
2. Δ QP >=λ &&N are metcurIBC>=2, Normal Intra patterns are skipped, are otherwise performed 3.;
3. R is metIBC> 0.6 | | NcurIBC+NdecIBC(the N of < 0.8curIntra+NdecIntra), Normal Intra patterns are skipped, otherwise
Perform 4.;
4. N is metcurIBC≥1&&NdecIBC>=1, Normal Intra patterns are skipped, are otherwise terminated, wherein NdecIBCFor current volume
Code unit decoding when corresponding decoding unit adjacent decoding unit pattern in IntraBC patterns number, NdecIntraTo work as
During the decoding of preceding coding unit in the pattern of the adjacent decoding unit of corresponding decoding unit Normal Intra patterns number,
NcurIBCFor the number of IntraBC patterns in the pattern of the adjacent encoded unit of current coded unit, NcurIntraFor present encoding
The number of Normal Intra patterns in the pattern of the adjacent encoded unit of unit;
(5) D is worked ascur< 3&&RIBCDuring < 0.05, IntraBC patterns are skipped, otherwise depth D upon decodingdec> Dcur, and work as
Pattern is Normal Intra during the decoding of preceding coding unit, and meeting following condition, one of those then skips IntraBC patterns,
Otherwise step (6) is performed:
1. (Δ QP < λ &&N are metdecIntra≥3)||(ΔQP≥λ&&NcurIntra>=3), IntraBC patterns are skipped, are otherwise performed
②;
2. N is metcurIntra≥1&&NdecIntra>=1, IntraBC patterns are skipped, are otherwise performed 3.;
3. R is metIBC< 0.2 | | NcurIBC+NdecIBC(the N of > 1.2curIntra+NdecIntra), IntraBC patterns are skipped, are otherwise performed
(6), wherein RIBCWhen being decoded for current coded unit in the adjacent decoding unit of corresponding decoding unit IntraBC patterns it is general
Rate;
(6) when meeting Ddec=Dcur+ 2 and current coded unit decoding when pattern be NormalIntra, skip IntraBC moulds
Formula, otherwise performs (7);
(7) encoded coding unit selection IntraBC and Normal Intra average rate distortion value is calculated, by Normal
The rate distortion value of optimal mode after Intra model selections as Rule of judgment, if:
BestCU_Cost < min (AvgNorCost, AvgIntraBCCost)
Then enter following condition judgment, otherwise do not enter fast algorithm, be normally carried out IntraBC model selections, wherein
AvgNorCost selects IntraBC average rate distortion value for encoded coding unit, and AvgIntraBCCost is encoded
Coding unit selection Normal Intra average rate distortion value, BestCU_Cost be Normal Intra model selections it
The rate distortion value of optimal mode afterwards:
If 1. RIBC< 0.1, then skip IntraBC patterns, otherwise performs 2.;
If 2. RIBC< 0.5&&NcurIntra>=3, then IntraBC patterns are skipped, otherwise do not enter fast algorithm, are normally carried out
IntraBC model selections.
8. the quick code check trans-coding system according to claim 6 based on HEVC-SCC, it is characterised in that data acquisition mould
The correspondence when pattern information for the adjacent encoded unit of the current coded unit that block is obtained and/or current coded unit decoding
The adjacent decoding unit of decoding unit pattern information be specially encoded unit and/or current coded unit decoding when it is corresponding
The pattern information of the adjacent decoding unit top of decoding unit, upper left side, upper right side and Left prediction unit.
9. the quick code check trans-coding system according to claim 8 based on HEVC-SCC, it is characterised in that the data are obtained
The transcoding algorithm for the predicting unit that modulus block is obtained is as follows:
(1) decoding schema S when current prediction unit is decoded in the information that decoding end is preserved is obtaineddec;
(2) if Sdec=2Nx2N, then coding side predicting unit selection 2Nx2N patterns, skip other patterns, otherwise perform (3);
(3) if Sdec=NxN, skips 2NxN patterns;If Sdec=NxN, transcoding quantization difference is less than λ and adjacent four predicting unit moulds
Formula is NxN number NNxN>=3, then 2NxN patterns are skipped, otherwise (4) are performed;
(4) if Sdec=Nx2N, then skip 2NxN patterns, and any pattern is not otherwise skipped and is normally carried out model selection.
10. according to any described quick code check trans-coding systems based on HEVC-SCC of claim 6-9, it is characterised in that institute
It is λ=5 to state the poor value of transcoding quantization.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710175153.3A CN107087172B (en) | 2017-03-22 | 2017-03-22 | Quick code check code-transferring method based on HEVC-SCC and its system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710175153.3A CN107087172B (en) | 2017-03-22 | 2017-03-22 | Quick code check code-transferring method based on HEVC-SCC and its system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107087172A true CN107087172A (en) | 2017-08-22 |
CN107087172B CN107087172B (en) | 2018-08-07 |
Family
ID=59614633
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710175153.3A Expired - Fee Related CN107087172B (en) | 2017-03-22 | 2017-03-22 | Quick code check code-transferring method based on HEVC-SCC and its system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107087172B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115580723A (en) * | 2022-12-09 | 2023-01-06 | 中南大学 | Method, system, device and medium for encoding and optimizing screen content image |
CN116600107A (en) * | 2023-07-20 | 2023-08-15 | 华侨大学 | HEVC-SCC quick coding method and device based on IPMS-CNN and spatial neighboring CU coding modes |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104125469A (en) * | 2014-07-10 | 2014-10-29 | 中山大学 | Fast coding method for high efficiency video coding (HEVC) |
WO2015054813A1 (en) * | 2013-10-14 | 2015-04-23 | Microsoft Technology Licensing, Llc | Encoder-side options for intra block copy prediction mode for video and image coding |
CN104581170A (en) * | 2015-01-23 | 2015-04-29 | 四川大学 | Rapid inter-frame transcoding method for reducing video resolution based on HEVC |
CN104702958A (en) * | 2013-12-05 | 2015-06-10 | 中国科学院深圳先进技术研究院 | HEVC intraframe coding method and system based on spatial correlation |
CN106131573A (en) * | 2016-06-27 | 2016-11-16 | 中南大学 | A kind of HEVC spatial resolution code-transferring method |
CN106210721A (en) * | 2016-07-05 | 2016-12-07 | 中南大学 | A kind of HEVC quick code check code-transferring method |
-
2017
- 2017-03-22 CN CN201710175153.3A patent/CN107087172B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015054813A1 (en) * | 2013-10-14 | 2015-04-23 | Microsoft Technology Licensing, Llc | Encoder-side options for intra block copy prediction mode for video and image coding |
CN104702958A (en) * | 2013-12-05 | 2015-06-10 | 中国科学院深圳先进技术研究院 | HEVC intraframe coding method and system based on spatial correlation |
CN104125469A (en) * | 2014-07-10 | 2014-10-29 | 中山大学 | Fast coding method for high efficiency video coding (HEVC) |
CN104581170A (en) * | 2015-01-23 | 2015-04-29 | 四川大学 | Rapid inter-frame transcoding method for reducing video resolution based on HEVC |
CN106131573A (en) * | 2016-06-27 | 2016-11-16 | 中南大学 | A kind of HEVC spatial resolution code-transferring method |
CN106210721A (en) * | 2016-07-05 | 2016-12-07 | 中南大学 | A kind of HEVC quick code check code-transferring method |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115580723A (en) * | 2022-12-09 | 2023-01-06 | 中南大学 | Method, system, device and medium for encoding and optimizing screen content image |
CN116600107A (en) * | 2023-07-20 | 2023-08-15 | 华侨大学 | HEVC-SCC quick coding method and device based on IPMS-CNN and spatial neighboring CU coding modes |
CN116600107B (en) * | 2023-07-20 | 2023-11-21 | 华侨大学 | HEVC-SCC quick coding method and device based on IPMS-CNN and spatial neighboring CU coding modes |
Also Published As
Publication number | Publication date |
---|---|
CN107087172B (en) | 2018-08-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Pan et al. | A low-complexity screen compression scheme for interactive screen sharing | |
CN111801945A (en) | Hybrid motion compensated neural network with side information based video coding | |
CN111277826B (en) | Video data processing method and device and storage medium | |
CN105900425A (en) | Representing Motion Vectors in an Encoded Bitstream | |
Saldanha et al. | Configurable fast block partitioning for VVC intra coding using light gradient boosting machine | |
CN107566848A (en) | The method and device of encoding and decoding | |
CN108495135A (en) | A kind of fast encoding method of screen content Video coding | |
Shen et al. | Ultra fast H. 264/AVC to HEVC transcoder | |
CN103491334B (en) | Video transcode method from H264 to HEVC based on region feature analysis | |
CN109862356B (en) | Video coding method and system based on region of interest | |
CN104853191B (en) | A kind of HEVC fast encoding method | |
CN106961603A (en) | Intracoded frame code rate allocation method and device | |
CN105900419A (en) | Video encoding of screen content data | |
CN107431816A (en) | Low complex degree sample adaptively offsets (SAO) decoding | |
CN109379594A (en) | Video coding compression method, device, equipment and medium | |
Fernandez-Escribano et al. | An MPEG-2 to H. 264 video transcoder in the baseline profile | |
CN104811729B (en) | A kind of video multi-reference frame coding method | |
CN107404648A (en) | A kind of multi-channel video code-transferring method based on HEVC | |
CN107071422A (en) | Low complex degree HEVC rate adaption transformation coding methods based on image correlation model | |
CN108769696A (en) | A kind of DVC-HEVC video transcoding methods based on Fisher discriminates | |
Wang et al. | UHD video coding: A light-weight learning-based fast super-block approach | |
CN106131573B (en) | A kind of HEVC spatial resolutions code-transferring method | |
CN107087172B (en) | Quick code check code-transferring method based on HEVC-SCC and its system | |
CN115118964A (en) | Video encoding method, video encoding device, electronic equipment and computer-readable storage medium | |
CN111770334B (en) | Data encoding method and device, and data decoding 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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180807 |