CN107749984A - Multiple view video coding complexity control method based on mode map - Google Patents
Multiple view video coding complexity control method based on mode map Download PDFInfo
- Publication number
- CN107749984A CN107749984A CN201710997433.2A CN201710997433A CN107749984A CN 107749984 A CN107749984 A CN 107749984A CN 201710997433 A CN201710997433 A CN 201710997433A CN 107749984 A CN107749984 A CN 107749984A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- viewpoint
- complexity
- mode
- 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
- H04N19/134—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
- H04N19/10—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
-
- 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
- H04N19/136—Incoming video signal characteristics or properties
- H04N19/14—Coding unit complexity, e.g. amount of activity or edge presence estimation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The present invention relates to a kind of multiple view video coding complexity control method based on mode map, it is therefore intended that solves on the premise of encoder complexity is controlled, realizes the problem of encoding rate distortion costs minimum.For the present invention by obtaining the rate distortion dependence between multiple views and in viewpoint, the rate distortion for building multiple views relies on model；Based on the principle that multi-vision-point encoding totality rate distortion costs are minimum, the allocative decision of computation complexity is realized, Optimal Complexity is distributed to each viewpoint and each frame of video；Model prediction and mode selection algorithm based on cost function are proposed, is realized in the case where controlling encoder complexity, the purpose of forced coding quality.A kind of multiple view video coding complexity control method based on mode map proposed by the present invention, on the premise of encoder complexity is controlled, the minimum effect of coding rate distortion costs is reached.
Description
Technical field
The present invention relates to technical field of video coding, particularly a kind of multiple view video coding based on mode map is complicated
Spend control method.
Background technology
Most encoder complexity control method is only applicable to handle the vision signal that single camera shooting forms, i.e.,
The Video Applications scene of single view.And as raising of the people to video-see quality, the video of single view planarization can not
Meet the needs of people, there is high-resolution, fine definition, the video of third dimension and interactivity is only becoming for video development from now on
Gesture.Multi-view point video refers to shooting one group of video that Same Scene obtains with different view by multiple video cameras of different points of view
Signal, it is a kind of effective 3D representation of video shot methods, being capable of more vivo reconstruction of scenes, there is provided third dimension and interactive function.
Multi-view point video is compared with single-view video, and the data volume of multi-view point video is linearly increasing with the number increase of video camera,
Therefore the optimization method of single-view video can not be continued to use, it is necessary to multiple views are regarded using more efficient computation complexity control algolithm
Frequency is according to being stored and transmitted.Based on background above, the present invention proposes a kind of multiple view video coding based on mode map
Complexity control method.
The content of the invention
It is an object of the invention to provide a kind of multiple view video coding complexity control method based on mode map, with
Overcome defect present in prior art.
To achieve the above object, the technical scheme is that：A kind of multiple view video coding based on mode map is answered
Miscellaneous degree control method, is realized in accordance with the following steps：
Step S1：Obtain and establish in multi-vision-point encoding rate distortion dependence between viewpoint；
Step S2：Obtain and establish in multi-vision-point encoding rate distortion dependence in viewpoint；
Step S3：Based on the principle that multi-vision-point encoding totality rate distortion costs are minimum, complexity allocative decision is obtained, and be
Each viewpoint and each frame of video distribution Optimal Complexity；
Step S4：Model selection is carried out, and pattern to be predicted is obtained to the pass between candidate pattern by a cost function
System, complete mode map；
Step S5：Current CU optimum prediction mode is determined according to domain prediction pattern and the cost function；
Step S6：Based on model prediction, centered on optimum prediction mode, using the coefficient related to complexity as radius,
It is determined that current CU model selection hunting zone, complexity corresponding to completion control.
In the present invention shows an embodiment, in the step S1, rate distortion relies between viewpoint in the multi-vision-point encoding
Relation is：
Wherein, the RDcost changing values of n-th of viewpoint are Δ Cn, then (n+1)th viewpoint as caused by n-th of viewpoint
RDcost knots modification is Δ Cn+1, the relation of n-th of viewpoint and (n+1)th viewpoint is εn；
Make RD dependent coefficients between viewpoint：Then：
ΔCtot,n=βn·ΔCn。
In the present invention shows an embodiment, in the step S2, rate is lost in viewpoint in the multi-vision-point encoding
True dependence relation is：
ΔCtot,n=αlΔcn,l
Wherein, Δ cn,lFor n-th of viewpoint l layers RDcost changing value, αlFor RD dependences coefficient in viewpoint, andγ is linear coefficient, l=0,1 ... L.
In the present invention shows an embodiment, in the step S3, the given overall complexity of note user is Ψ, n-th
The complexity controlling elements of viewpoint l layers are θn,l, and：
According to the principle minimum based on multi-vision-point encoding totality rate distortion costs, then：
Solved with reference to above formula, and by using MatLab linear programming problems, obtain complexity allocative decision；
Wherein, ωn,lThe ratio of totalframes, β are accounted for for the video frame number of n-th of viewpoint l layernThe RD dependent coefficients between viewpoint,
αlFor RD dependences coefficient in viewpoint, and：
Wherein, ε1For the relation of the 1st viewpoint and the 2nd viewpoint, ε2For the relation of the 2nd viewpoint and the 3rd viewpoint, γ
For linear coefficient, l=0,1 ... L.
In the present invention shows an embodiment, in the step S5, the cost function C (mi,M*) represent to work as predictive mode
For M*, predictive mode M*To candidate pattern miCost：And by cost function C (mi,M*) be designated as：
Wherein, J (M*) and J (mi) represent predictive mode M*With candidate pattern miRDcost, C ∈ [0,1].
In the present invention shows an embodiment, in the step S5, according to the correlation between space-time and viewpoint, according to having compiled
The field block of code, including：Time domain, between spatial domain and viewpoint, optimal segmentation pattern is chosen as predictive mode；Remember final checked
Optimum prediction mode isWith reference to the cost function, it is by optimum prediction modeIt is expressed as：
Wherein, mkThe forced coding pattern that the CU of same position chooses between time domain, spatial domain and viewpoint.
In the present invention shows an embodiment, in the step S6, the radius r is：
R=(1- θn,l)·rmin+θn,l·rmax
Wherein, θn,lComplexity controlling elements for n-th of viewpoint l layer are θn,lComplexity controlling elements, also it is to distribute
To viewpoint n video layers l computation complexity, θn,l∈ [0,1], rminFor the minimum value of hunting zone, rmaxFor hunting zone most
Big value.
Compared to prior art, the invention has the advantages that：The present invention provides a kind of based on the more of mode map
Viewpoint video encoder complexity control method, reach on the premise of encoding computational complexity is controlled, realized encoding rate distortion
Least-cost, the optimal technique effect of coding quality.
Brief description of the drawings
Fig. 1 is the flow chart of the multiple view video coding complexity control method based on mode map in the present invention.
Fig. 2 is the prediction structural representation of three viewpoints in one embodiment of the invention.
Fig. 3 is model selection schematic diagram in one embodiment of the invention.
Embodiment
Below in conjunction with the accompanying drawings, technical scheme is specifically described.
A kind of multiple view video coding complexity control method based on mode map of the present invention, as shown in figure 1, including such as
Lower step：
Step S1：Rate distortion dependence between viewpoint in structure multi-vision-point encoding.
Step S2：Rate distortion dependence in viewpoint in structure multi-vision-point encoding.
Step S3：Based on the principle that multi-vision-point encoding totality rate distortion costs are minimum, the distribution side of computation complexity is realized
Case, to each viewpoint and each frame of video distribution Optimal Complexity.
Step S4：All Fractionation regimens are calculated to the cost between remaining pattern, and are defined as cost function.
Relation between some predictive mode and other all patterns, i.e. mode map are obtained by cost function.
Step S5：Current CU optimum prediction mode is determined according to domain prediction pattern and cost function.
Step S6：Based on model prediction, centered on optimum prediction mode, the coefficient related to complexity is radius, really
Settled preceding CU model selection hunting zone.CU ignores all Fractionation regimens in hunting zone outside searched radius
Pattern.
Further, in the present embodiment, in step sl, rate distortion dependence between viewpoint is built in multi-vision-point encoding
Method is specific as follows：
As shown in Fig. 2 being the pre- geodesic structure of three viewpoints, according to the model of RD dependences between multiple views, can obtain each
RD performances are to the influence relation of overall RD performances between viewpoint：
Wherein, the RDcost changing values of n-th of viewpoint are Δ Cn, then (n+1)th viewpoint as caused by n-th of viewpoint
RDcost knots modification is Δ Cn+1, the relation of n-th of viewpoint and (n+1)th viewpoint is εn。
OrderIt can obtain：
ΔCtot,n=βn·ΔCn. (2)
Further, in the present embodiment, it is as follows using rate distortion dependence in viewpoint in step S2：
ΔCtot,n=αlΔcn,l (3)
Wherein,γ is linear coefficient, l=0,1 ... L.
Further, in the present embodiment, in step S3, the computation complexity distribution principle and scheme tool of multiple views are realized
Body is as follows：According to the RD dependences of each interlayer between each viewpoint and in viewpoint, and convolution (2) and (3), can be overall by multiple views
RDcost incrementss are described as：
ΔCtot,{n,l}=βn·αl·Δcn,l, (4)
Wherein, n-th of viewpoint l layers RDcost changing value Δ cn,lThe knots modification for causing overall RDcost is Δ
Ctot,{n,l}, αlFor RD dependences coefficient, β in viewpointnThe RD dependent coefficients between viewpoint.In order that coding quality reaches best,
It is minimum to encode overall rate distortion costs RDcost values, exists：
In the present embodiment, overall complexity user given is defined as Ψ.When Ψ is smaller, encoder complexity is represented
Lower, then the complexity controlling elements of n-th of viewpoint l layer are represented by θN, l, and θN, lIt is as follows with Ψ relation：
Wherein ωn,lThe ratio of totalframes is accounted for for the video frame number of n-th of viewpoint l layer.
Further, as the θ of distributionN, lWhen bigger, the coding quality of the frame is better, overall rate distortion costs Δ cn,l
With regard to smaller.In the present embodiment, by complexity controlling elements θn,lWith RDcost knots modification Δs cn,lBetween, closed substantially there is linear
Formula (5), can be adjusted to by system：
In summary, in order on the premise of encoding computational complexity is controlled so that the rate distortion of coding reaches minimum,
I.e.：
Wherein,
Further, in the present embodiment, convolution (8), and solved by using matlab linear programming problems, obtain
The complexity allocative decision of different points of view different layers.Table 1 gives some of complex degree allocative decision of different points of view different layers.
Table 1
Further, can be in the case where giving overall computation complexity Ψ, to different points of view different layers using table 1
Secondary frame of video distributes different encoder complexities.
Further, in the present embodiment, used in CU levels, the present embodiment the mode map scheme based on multiple views with
Solves the mode selection problem in multi-vision-point encoding, as shown in Figure 3.Wherein, a kind of coding of candidate point is represented each origin
Cut pattern, such as Skip, 2N*2N, 2N*N etc..Center origin represents current CU predictive mode, in order to avoid coding efficiency declines,
Need to centered on current prediction mode, other Fractionation regimens in the range of radius r are tested.rminFor hunting zone
Minimum value, rmaxFor the maximum of hunting zone.When r is bigger, it is necessary to which the Fractionation regimen of test is also more, encoder complexity
Also improve therewith.Search radius r and encoder complexity relation are expressed as：
R=(1- θn,l)·rmin+θn,l·rmax, (9)
Wherein, θn,lTo be assigned to the computation complexity of n-th of viewpoint l layer, θn,l∈[0,1]。
Further, in the present embodiment, in step s 4, the scheme that mode map is completed by cost function is as follows：
Current CU can test RDcost caused by several segmentation candidates patterns when carrying out model selection, final to choose
Pattern minimum RDcost is as coding mode.When RDcost is smaller, it is selected and is also got over for the probability of forced coding pattern
Greatly.Cost function C (m are used in the present embodimenti,M*) represent when predictive mode is M*, M*To some candidate pattern miCost.
Cost function C is defined as：
Wherein, J (M*) and J (mi) represent predictive mode and some candidate pattern miRDcost, C ∈ [0,1].Pass through generation
Valency function C and formula (10), we are with regard to that can obtain the relation between some predictive mode and other all patterns, as shown in figure 3, should
Process is mode map process.
Further, in the present embodiment, an experiment is devised to obtain mapping relations between each pattern.Testing
In, travel through each segmentation candidates pattern, while record they RDcost and encoder final choice optimal mode, finally
The cost relation between each optimal mode and remaining candidate pattern is calculated using formula (10), and result is arranged in cost letter
In number result table, as shown in table 2.
Table 2
Further, in the present embodiment, in step S5, determine current CU's according to domain prediction pattern and cost function
The principle and method of optimum prediction mode be specially：
, can be according to encoded field block, including time domain using the correlation between space-time and viewpoint, spatial domain, between viewpoint, choosing
In certain optimal segmentation pattern as predictive mode.It is assumed that the predictive mode of final checked isConvolution (10) can be by its table
It is shown as：
Wherein, mkThe forced coding pattern that the CU of same position chooses between expression time domain, spatial domain and viewpoint.Utilize cost letter
Number result table, any one pattern can be calculated to the cost of some reference model, and chosen into all reference models
Average cost minimum is used as optimum prediction mode
Further, with reference to optimum prediction mode and cost function table, it can be deduced that the distribution relation signal of each pattern
Figure, and then in the case of the computation complexity of given distribution, complete the process of CU model selections.
In order to allow those skilled in the art to further appreciate that method proposed by the invention, enter with reference to specific embodiment
Row illustrates.
In order to verify the validity of inventive algorithm, by anti-error pass-algorithm and multi-vision-point encoding mode selection algorithm phase
With reference to, and applied in 3D-HEVC standard testing softwares HTM.The test uses CTC standard 3D test sets, including：”
Newspaper”,”GhostTownFly”,”balloons”,”kendo”,”PoznanHall2”,”PoznanStreet”,”
Shark ", " the standard testing video of Undo_Dancer " this eight different accuracies.Experimental result is as shown in table 3.
Table 3
Above is presently preferred embodiments of the present invention, all changes made according to technical solution of the present invention, caused function are made
During with scope without departing from technical solution of the present invention, protection scope of the present invention is belonged to.
Claims (7)
1. a kind of multiple view video coding complexity control method based on mode map, it is characterised in that in accordance with the following steps
Realize：
Step S1：Obtain and establish in multi-vision-point encoding rate distortion dependence between viewpoint；
Step S2：Obtain and establish in multi-vision-point encoding rate distortion dependence in viewpoint；
Step S3：Based on the principle that multi-vision-point encoding totality rate distortion costs are minimum, complexity allocative decision is obtained, and is each
Viewpoint and each frame of video distribution Optimal Complexity；
Step S4：Model selection is carried out, and pattern to be predicted is obtained to the relation between candidate pattern by a cost function, it is complete
Become the mode mapping；
Step S5：Current CU optimum prediction mode is determined according to domain prediction pattern and the cost function；
Step S6：Based on model prediction, centered on optimum prediction mode, using the coefficient related to complexity as radius, it is determined that
Current CU model selection hunting zone, complexity corresponding to completion control.
2. a kind of multiple view video coding complexity control method based on mode map according to claim 1, it is special
Sign is, in the step S1, rate distortion dependence is between viewpoint in the multi-vision-point encoding：
<mrow>
<msub>
<mi>&Delta;C</mi>
<mrow>
<mi>t</mi>
<mi>o</mi>
<mi>t</mi>
<mo>,</mo>
<mi>n</mi>
</mrow>
</msub>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mo>(</mo>
<mn>1</mn>
<mo>+</mo>
<msub>
<mi>&epsiv;</mi>
<mn>1</mn>
</msub>
<mo>+</mo>
<msub>
<mi>&epsiv;</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
<mi>&Delta;</mi>
<msub>
<mi>C</mi>
<mi>n</mi>
</msub>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<mi>&Delta;</mi>
<msub>
<mi>C</mi>
<mi>n</mi>
</msub>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
<mi>e</mi>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>,</mo>
</mrow>
Wherein, the RDcost changing values of n-th of viewpoint are Δ Cn, then as caused by n-th of viewpoint (n+1)th viewpoint RDcost
Knots modification is Δ Cn+1, the relation of n-th of viewpoint and (n+1)th viewpoint is εn；
Make RD dependent coefficients between viewpoint：Then：
ΔCtot,n=βn·ΔCn。
3. a kind of multiple view video coding complexity control method based on mode map according to claim 1, it is special
Sign is, in the step S2, rate distortion dependence is in viewpoint in the multi-vision-point encoding：
ΔCtot,n=αlΔcn,l
Wherein, Δ cn,lFor n-th of viewpoint l layers RDcost changing value, αlFor RD dependences coefficient in viewpoint, andγ is linear coefficient, l=0,1 ... L.
4. a kind of multiple view video coding complexity control method based on mode map according to claim 1, it is special
Sign is, in the step S3, the given overall complexity of note user is Ψ, the complexity controlling elements of n-th of viewpoint l layer
For θn,l, and：
<mrow>
<mi>s</mi>
<mo>.</mo>
<mi>t</mi>
<mo>.</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>N</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>L</mi>
</munderover>
<msub>
<mi>&omega;</mi>
<mrow>
<mi>n</mi>
<mo>,</mo>
<mi>l</mi>
</mrow>
</msub>
<msub>
<mi>&theta;</mi>
<mrow>
<mi>n</mi>
<mo>,</mo>
<mi>l</mi>
</mrow>
</msub>
<mo>=</mo>
<mi>&Psi;</mi>
</mrow>
According to the principle minimum based on multi-vision-point encoding totality rate distortion costs, then：
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mo>{</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>N</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mi>L</mi>
</munderover>
<msub>
<mi>&beta;</mi>
<mi>n</mi>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>&alpha;</mi>
<mi>l</mi>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>&theta;</mi>
<mrow>
<mi>n</mi>
<mo>,</mo>
<mi>l</mi>
</mrow>
</msub>
<mo>}</mo>
</mrow>
Solved with reference to above formula, and by using MatLab linear programming problems, obtain complexity allocative decision；
Wherein, ωn,lThe ratio of totalframes, β are accounted for for the video frame number of n-th of viewpoint l layernThe RD dependent coefficients between viewpoint, αlFor
RD dependences coefficient in viewpoint, and：
<mrow>
<msub>
<mi>&beta;</mi>
<mi>n</mi>
</msub>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mn>1</mn>
<mo>+</mo>
<msub>
<mi>&epsiv;</mi>
<mn>1</mn>
</msub>
<mo>+</mo>
<msub>
<mi>&epsiv;</mi>
<mn>2</mn>
</msub>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>1</mn>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
<mi>e</mi>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
<mrow>
<msub>
<mi>&alpha;</mi>
<mi>l</mi>
</msub>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mfrac>
<mn>1</mn>
<mrow>
<mn>1</mn>
<mo>-</mo>
<mi>&gamma;</mi>
</mrow>
</mfrac>
<msup>
<mrow>
<mo>(</mo>
<mrow>
<mn>1</mn>
<mo>+</mo>
<mn>2</mn>
<mi>&gamma;</mi>
</mrow>
<mo>)</mo>
</mrow>
<mi>L</mi>
</msup>
<mo>,</mo>
<mi>l</mi>
<mo>=</mo>
<mn>0</mn>
</mtd>
</mtr>
<mtr>
<mtd>
<msup>
<mrow>
<mo>(</mo>
<mrow>
<mn>1</mn>
<mo>+</mo>
<mn>2</mn>
<mi>&gamma;</mi>
</mrow>
<mo>)</mo>
</mrow>
<mrow>
<mi>L</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<mo>,</mo>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
<mi>e</mi>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, ε1For the relation of the 1st viewpoint and the 2nd viewpoint, ε2For the relation of the 2nd viewpoint and the 3rd viewpoint, γ is line
Property coefficient, l=0,1 ... L.
5. a kind of multiple view video coding complexity control method based on mode map according to claim 1, it is special
Sign is, in the step S4, the cost function C (mi,M*) represent when predictive mode is M*, predictive mode M*To candidate
Mode miCost：And by cost function C (mi,M*) be designated as：
<mrow>
<mi>C</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msup>
<mi>M</mi>
<mo>*</mo>
</msup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>J</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>J</mi>
<mrow>
<mo>(</mo>
<msup>
<mi>M</mi>
<mo>*</mo>
</msup>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>J</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>J</mi>
<mrow>
<mo>(</mo>
<msup>
<mi>M</mi>
<mo>*</mo>
</msup>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
Wherein, J (M*) and J (mi) represent predictive mode M*With candidate pattern miRDcost, C ∈ [0,1].
6. a kind of multiple view video coding complexity control method based on mode map according to claim 5, it is special
Sign is, in the step S5, according to the correlation between space-time and viewpoint, and according to encoded field block, including：Time domain,
Between spatial domain and viewpoint, optimal segmentation pattern is chosen as predictive mode；Note final checked optimum prediction mode beKnot
The cost function is closed, is by optimum prediction modeIt is expressed as：
<mrow>
<mover>
<mi>M</mi>
<mo>^</mo>
</mover>
<mo>=</mo>
<mi>arg</mi>
<munder>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
</munder>
<mrow>
<mo>(</mo>
<mi>&Sigma;</mi>
<mi>c</mi>
<mo>(</mo>
<mrow>
<msub>
<mi>m</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>m</mi>
<mi>k</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
Wherein, mkThe forced coding pattern that the CU of same position chooses between time domain, spatial domain and viewpoint.
7. a kind of multiple view video coding complexity control method based on mode map according to claim 1, it is special
Sign is, in the step S6, the radius r is：
R=(1- θn,l)·rmin+θn,l·rmax
Wherein, θn,lTo be assigned to the computation complexity of n-th of viewpoint l layer, θn,l∈ [0,1], rminFor the minimum of hunting zone
Value, rmaxFor the maximum of hunting zone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710997433.2A CN107749984B (en) | 2017-10-24 | 2017-10-24 | Multi-view video coding complexity control method based on mode mapping |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710997433.2A CN107749984B (en) | 2017-10-24 | 2017-10-24 | Multi-view video coding complexity control method based on mode mapping |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107749984A true CN107749984A (en) | 2018-03-02 |
| CN107749984B CN107749984B (en) | 2019-12-17 |
Family
ID=61253133
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710997433.2A Active CN107749984B (en) | 2017-10-24 | 2017-10-24 | Multi-view video coding complexity control method based on mode mapping |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107749984B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108989801A (en) * | 2018-07-27 | 2018-12-11 | 福州大学至诚学院 | A kind of adaptive model selection fast method based on space-time characterisation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101035261A (en) * | 2007-04-11 | 2007-09-12 | 宁波大学 | Image signal processing method of the interactive multi-view video system |
| CN101569197A (en) * | 2006-12-21 | 2009-10-28 | 汤姆森许可贸易公司 | Methods and apparatus for improved signaling using high level syntax for multi-view video coding and decoding |
| CN101917618A (en) * | 2010-08-20 | 2010-12-15 | 浙江大学 | Calculation complexity control method for multi-viewpoint video coding and layering B frame predicting structure |
| CN105247863A (en) * | 2013-07-08 | 2016-01-13 | 联发科技（新加坡）私人有限公司 | Method of simplified cabac coding in 3d video coding |
-
2017
- 2017-10-24 CN CN201710997433.2A patent/CN107749984B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101569197A (en) * | 2006-12-21 | 2009-10-28 | 汤姆森许可贸易公司 | Methods and apparatus for improved signaling using high level syntax for multi-view video coding and decoding |
| CN101035261A (en) * | 2007-04-11 | 2007-09-12 | 宁波大学 | Image signal processing method of the interactive multi-view video system |
| CN101917618A (en) * | 2010-08-20 | 2010-12-15 | 浙江大学 | Calculation complexity control method for multi-viewpoint video coding and layering B frame predicting structure |
| CN105247863A (en) * | 2013-07-08 | 2016-01-13 | 联发科技（新加坡）私人有限公司 | Method of simplified cabac coding in 3d video coding |
Non-Patent Citations (1)
| Title |
|---|
| TIESONG ZHAO: "Flexible Mode Selection and Complexity Allocation in High Efficiency Video Coding", 《IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108989801A (en) * | 2018-07-27 | 2018-12-11 | 福州大学至诚学院 | A kind of adaptive model selection fast method based on space-time characterisation |
| CN108989801B (en) * | 2018-07-27 | 2021-10-08 | 福州大学至诚学院 | Space-time characteristic-based adaptive mode selection rapid method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107749984B (en) | 2019-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104754357B (en) | Intraframe coding optimization method and device based on convolutional neural networks | |
| CN104717513B (en) | A kind of bidirectional interframe predictive method and device | |
| CN105160678A (en) | Convolutional-neural-network-based reference-free three-dimensional image quality evaluation method | |
| CN107909150B (en) | Method and system for on-line training CNN based on block-by-block random gradient descent method | |
| CN101668202A (en) | Method and device for selecting intra-frame prediction mode | |
| CN104010193B (en) | Video steganographic method based on macro block partition mode disturbance | |
| CN107734333A (en) | A kind of method for improving video error concealing effect using network is generated | |
| CN108765512B (en) | Confrontation image generation method based on multi-level features | |
| CN106157339A (en) | The animated Mesh sequence compaction algorithm extracted based on low-rank vertex trajectories subspace | |
| CN104064193B (en) | The information concealing method of a kind of linear prediction voice coding and extracting method | |
| CN106507116A (en) | A kind of 3D HEVC coding methods that is predicted based on 3D conspicuousnesses information and View Synthesis | |
| CN114463209A (en) | Image restoration method based on deep multi-feature collaborative learning | |
| CN109584340A (en) | New Century Planned Textbook synthetic method based on depth convolutional neural networks | |
| CN112651360B (en) | Skeleton action recognition method under small sample | |
| CN107749984A (en) | Multiple view video coding complexity control method based on mode map | |
| CN107690069A (en) | A kind of cascade method for video coding of data-driven | |
| CN115830575A (en) | Transformer and cross-dimension attention-based traffic sign detection method | |
| CN107426570A (en) | A kind of adaptive Qp Cascading Methods of low latency Video coding | |
| CN103702131B (en) | Pattern-preprocessing-based intraframe coding optimization method and system | |
| CN112770120B (en) | 3D video depth map intra-frame rapid coding method based on depth neural network | |
| CN113068041B (en) | Intelligent affine motion compensation coding method | |
| CN109448039B (en) | Monocular vision depth estimation method based on deep convolutional neural network | |
| Brahimi et al. | Improved very deep recurrent convolutional neural network for object recognition | |
| CN110147733A (en) | A kind of cross-domain a wide range of scene generating method | |
| CN112862946B (en) | Gray rock core image three-dimensional reconstruction method for generating countermeasure network based on cascade condition |
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 |