CN102215407A - Bilaterally predicted block acquisition method, device and image decoding method - Google Patents
Bilaterally predicted block acquisition method, device and image decoding method Download PDFInfo
- Publication number
- CN102215407A CN102215407A CN2010101475847A CN201010147584A CN102215407A CN 102215407 A CN102215407 A CN 102215407A CN 2010101475847 A CN2010101475847 A CN 2010101475847A CN 201010147584 A CN201010147584 A CN 201010147584A CN 102215407 A CN102215407 A CN 102215407A
- Authority
- CN
- China
- Prior art keywords
- prediction
- template
- module
- matching error
- current block
- 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 66
- 230000015572 biosynthetic process Effects 0.000 claims description 37
- 150000001875 compounds Chemical class 0.000 claims 2
- 230000002146 bilateral effect Effects 0.000 abstract 1
- 239000000654 additive Substances 0.000 description 3
- 230000000996 additive effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 238000013139 quantization Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Images
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention discloses a bilaterally predicted block acquisition method, a bilaterally predicted block acquisition device and an image decoding method, which relate to a video image coding technology and aim to solve the problem of low coding efficiency caused by the writing of the guiding information used for acquiring a bilaterally predicted block into a code stream in the prior art. The bilaterally predicted block acquisition method provided by the embodiment of the invention comprises the following steps of: acquiring a reference template corresponding to a current template of a current block from a bilateral reference queue; acquiring a matching error between the current template and the reference template; selecting a prediction template from the reference template according to the matching error between the current template and the reference template; and acquiring the predicted block corresponding to the current block according to the prediction template. The embodiment of the invention can be applied to video image coding and decoding processes such as H.264/advanced video coding (AVC) and the like.
Description
Technical field
The present invention relates to the encoding video pictures technology, relate in particular to a kind of bi-directional predicted acquisition methods and device and picture decoding method.
Background technology
At present, mainly adopt the inter prediction technology of motion compensation that video image is encoded.Particularly, coding side is when encoding to the current block in the present frame, at first from the frame that video image has been encoded, choose an above reference frame arbitrarily, and from described reference frame, obtain the prediction piece of current block correspondence, calculate the residual values between prediction piece and the current block then, this residual values is carried out quantization encoding; Decoding end is at first obtained the prediction piece of current block correspondence when the current block in the present frame is decoded, obtain the residual values of prediction piece and current block then from the code stream that receives, according to this residual values and prediction piece decoding and reconstituting current block.
In the prior art, coding side is when encoding to bi-directional predicted video image, the instructive information such as movable information of the reference frame index information that the prediction piece need be belonged to, prediction direction information and prediction piece are enrolled code stream, so that decoding end can be obtained the prediction piece of current block correspondence according to described instructive information, according to this prediction piece decoding and reconstituting current block.
Yet prior art will predict that the instructive information such as movable information of reference frame index information, prediction direction information and the prediction piece of piece ownership enroll code stream, take the coding resource of video image, reduce the code efficiency of video image.
Summary of the invention
Embodiments of the invention provide a kind of bi-directional predicted acquisition methods and device and picture decoding method, can improve code efficiency.
On the one hand, provide a kind of bi-directional predicted acquisition methods, having comprised: from two-way with reference to obtain the formation current block when front template corresponding reference template; Obtain described matching error when front template and described reference template; According to described matching error, from described reference template, choose prediction module when front template and described reference template; Obtain the prediction piece of described current block correspondence according to described prediction module.
On the other hand, provide a kind of bi-directional predicted deriving means, having comprised:
First acquisition module, be used for from two-way with reference to formation obtain current block when front template corresponding reference template;
First matching module is used to obtain the matching error of the described reference template that obtains when front template and described first acquisition module;
First chooses module, is used for the matching error when front template and described reference template obtained according to described first matching module, chooses prediction module from the reference template that described first acquisition module obtains;
Second acquisition module is used for choosing the prediction piece that prediction module that module chooses obtains described current block correspondence according to described first.
On the one hand, provide a kind of picture decoding method again, be used for the B two field picture is decoded, having comprised: obtained prediction direction; From corresponding with prediction direction two-way with reference to obtain the formation current block when front template corresponding reference template; Obtain described matching error when front template and described reference template; According to described matching error, from described reference template, choose prediction module when front template and described reference template; Obtain the prediction piece of described current block correspondence according to described prediction module, and utilize this prediction piece that current block is decoded.
Bi-directional predicted acquisition methods and device that the embodiment of the invention provides, can be according to working as front template and two-way matching error with reference to reference template in the formation, from reference template, derive prediction module adaptively, and obtain the prediction piece of current block correspondence according to prediction module, the technical scheme that makes the embodiment of the invention provide need be in the process of coding, the instructive information that will be used to obtain bi-directional predicted is enrolled code stream, has saved coding resource, has improved code efficiency.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art, to do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below, apparently, accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.
The bi-directional predicted acquisition methods flow chart that Fig. 1 provides for the embodiment of the invention;
The flow chart of the bi-directional predicted acquisition methods step 103 that Fig. 2 provides for the embodiment of the invention shown in Figure 1;
The bi-directional predicted deriving means structural representation one that Fig. 3 provides for the embodiment of the invention;
Choose the structural representation of module in the bi-directional predicted deriving means that Fig. 4 provides for inventive embodiments shown in Figure 3;
The bi-directional predicted deriving means structural representation two that Fig. 5 provides for the embodiment of the invention;
The structural representation of determination module in the bi-directional predicted deriving means that Fig. 6 provides for inventive embodiments shown in Figure 5.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.
Enroll code stream in order to solve the instructive information that prior art will be used to obtain bi-directional predicted, cause the low problem of code efficiency, the embodiment of the invention provides a kind of bi-directional predicted acquisition methods and device and picture decoding method, is used for the B two field picture is carried out Code And Decode.
As shown in Figure 1, the bi-directional predicted acquisition methods that the embodiment of the invention provides comprises:
Wherein, two-way with reference to formation comprise forward direction with reference to formation (List0) and the back to reference formation (List1), each includes an above reference frame with reference to formation (List0 and List1); Current block is made up of the block of pixels of (perhaps decoding) to be encoded in the present frame; The block of pixels of in front template is block of pixels according to the template specification information obtained in advance (comprising block of pixels number that template comprises and the position relation between template and the current block etc.) encode at the current block periphery (perhaps decoding), choosing; Reference template is to be of similar shape and size in the reference frame (comprising the reference frame among List0 and/or the List1) with when the corresponding template of front template with when front template, a zone of being made up of the reconstructed pixels in the reference frame.
In the present embodiment, step 101 can comprise: the correlation of utilizing time and space is closed two-way the structure in reference to formation (List0 and List1) when the set of candidate motion vectors of front template correspondence; From two-way reference frame, obtain each the motion vector corresponding reference template in closing with described set of candidate motion vectors with reference to formation (List0 and List1).
In the present embodiment, step 102 can be obtained matching error when front template and reference template by several different methods, for example: can by calculate when front template with the absolute error of reference template and (SAD), the difference of two squares and (SSE), definitely transformed error and (sum of absolutetransformation differences) or absolute square difference and etc., obtain matching error when front template and reference template.
Certainly, more than only for for example, in the use of reality, step 102 also can be obtained matching error when front template and reference template by additive method, does not do herein and gives unnecessary details.
In the present embodiment, step 103 can be chosen prediction module in several ways according to the matching error when front template and reference template from reference template.
For example: step 103 can be chosen the reference template of a matching error minimum as prediction module according to the matching error when front template and reference template from two-way reference template with reference to formation (List0 and List1).
And for example: step 103 can be according to the matching error order from small to large when front template and reference template, two-way reference template with reference to formation (List0 and List1) is sorted, therefrom choose the individual reference template of preceding N (N is an integer, and N 〉=2) as prediction module.
For another example: step 103 can be according to the matching error order from small to large when front template and reference template, respectively the reference template among List0 and the List1 is sorted, (m is an integer to choose preceding m from List0, and m 〉=1) individual reference template, from List1, choose the individual reference template of preceding n (n is an integer, and n 〉=1) as prediction module.
For another example: step 103 can be according to the matching error when front template and reference template, from first the reference template, choose the reference template of a matching error minimum as the first optimum reference template with reference to formation (as List0), calculating second is with reference to the matching error of the reference template and the first optimum reference template of formation (as List1), (M is an integer to choose preceding M from second with reference to formation (as List1), and M 〉=1) reference template of the matching error minimum of individual and the first optimum reference template is as the second optimum reference template, with the first optimum reference template () and the second optimum reference template (M is individual) as prediction module.
Certainly, more than only be concrete for example, step 103 can also be chosen prediction module by other modes in the use of reality, every kind of situation is not given unnecessary details one by one herein.
In the present embodiment, step 104 can be obtained the corresponding prediction piece of current block according to prediction module and the template specification of obtaining in advance information (comprising position relation between block of pixels number that template comprises and template and the current block etc.).
Further, if step 103 has been obtained two above prediction modules, then step 104 can be obtained the alternative prediction piece of each prediction module correspondence respectively, then described alternative prediction piece is made up the prediction piece that (as calculating each mean value of predicting piece or weighted average etc.) generates the current block correspondence.
The bi-directional predicted acquisition methods that the embodiment of the invention provides both can be applied in coding side also can be applied in decoding end.After bi-directional predicted acquisition methods that coding side adopts present embodiment to provide obtains the prediction piece of current block correspondence, can adopt this prediction piece that current block is predicted, obtain the residual values of prediction piece and current block, residual values is carried out quantization encoding write code stream; Decoding end can adopt the method identical with coding side, the bi-directional predicted acquisition methods that utilizes the embodiment of the invention to provide obtains the prediction piece, and from the code stream that receives, obtain the residual values of predicting piece and current block, according to this residual values and prediction piece decoding and reconstituting current block.
The bi-directional predicted acquisition methods that the embodiment of the invention provides, can be according to working as front template and two-way matching error with reference to reference template in the formation, from reference template, derive prediction module adaptively, and obtain the prediction piece of current block correspondence according to prediction module, the technical scheme that makes the embodiment of the invention provide need be in the process of coding, the instructive information that will be used to obtain bi-directional predicted is enrolled code stream, has saved coding resource, has improved code efficiency.
Another embodiment of the present invention also provides a kind of bi-directional predicted acquisition methods, this method basic identical with as shown in Figure 1, and difference is that as shown in Figure 2, in the present embodiment, step 103 shown in Figure 1 further can comprise:
In the present embodiment, the gang form of alternative prediction module can be combined to form by the alternative prediction module of any number, for example: when step 1031 has been obtained four alternative prediction modules, can from described four alternative prediction modules, choose two arbitrarily and make up, generate six
Gang form also can be chosen three arbitrarily and make up, and generates four
Gang form.
In the present embodiment, can in several ways alternative prediction module be made up, generate the gang form of alternative prediction module.
For example: if the alternative prediction module that step 1031 is chosen is for (N is an integer from the two-way preceding N that chooses with reference to formation (List0 and List1) according to matching error order from small to large, and N 〉=2) individual reference template, step 1032 can be chosen x arbitrarily from this N reference template (x＜N) individual reference template makes up; Further, if comprise a first reference template and b second reference template in N the reference template with reference to formation (as List1) with reference to formation (as List0), wherein, a+b=N, step 1032 can (the individual reference template of c＜a) be with (d＜b) individual reference template makes up from second with reference to the d that chooses arbitrarily the formation (as List1) from first with reference to choosing c the formation (as List0) arbitrarily.
And for example: if the alternative prediction module that step 1031 is chosen is that (m is an integer to the preceding m that chooses from List0 according to matching error order from small to large, and m 〉=1) (n is an integer for individual reference template and the preceding n that chooses from List1, and n 〉=1) individual reference template, step 1032 can be chosen the individual reference template of e (e＜(m+n)) arbitrarily and make up from m+n reference template, ((g＜m) individual reference template makes up the individual reference template of f＜m) with the g that chooses arbitrarily from List1 also can to choose f arbitrarily from the individual reference template of m (m is an integer, and m 〉=1) of List0;
For another example: if the alternative prediction module that step 1031 is chosen is for from the first first optimum reference template of choosing with reference to formation (as List0) () with from (M of the second second optimum reference template of choosing with reference to formation (as List1), M is an integer, and M 〉=1), step 1032 can be chosen h arbitrarily (the individual second optimum reference template of h＜M) and the first optimum reference template make up from the second optimum reference template.
Certainly, more than only be that concrete in the use of reality, step 1032 also can pass-through mode make up prediction module for example, the gang form of generation forecast template is not given unnecessary details every kind of situation herein one by one.
In the present embodiment, the gang form of alternative prediction module can be by the arithmetic average structure to alternative each pixel of prediction module of being used for making up.For example: when gang form is combined to form by two alternative prediction module TM0 and TM1, can by following formula (1) obtain gang form (x, y) pixel value of position:
comTM(x，y)＝[TM
0(x，y)+TM
1(x，y)]/2 (1)
Wherein, TM
0(x is that alternative prediction module TM0 is at (x, the y) pixel value of position, TM y)
1(x is that alternative prediction module TM1 is at (x, y) pixel value of position y).
Further, the gang form of alternative prediction module also can be by the alternative prediction module difference weighting structure to being used to make up.For example: when gang form is combined to form by two alternative prediction module TM0 and TM1, can by following formula (2) obtain gang form (x, y) pixel value of position:
comTM(x，y)＝TM
0/(TM
0+TM
1)×TM
0(x，y)+TM
1/(TM
0+TM
1)×TM
1(x，y) (2)
Wherein, TM
0Be the mean value of among the alternative prediction module TM0 pixel, TM
1Be the mean value of among the alternative prediction module TM1 pixel, TM
0(x is that alternative prediction module TM0 is at (x, the y) pixel value of position, TM y)
1(x is that alternative prediction module TM1 is at (x, y) pixel value of position y).
In the present embodiment, step 1033 can be obtained matching error when front template and gang form by several different methods, for example: can by calculate when front template with the absolute error of gang form and (SAD), the difference of two squares and (SSE), definitely transformed error and (sum of absolutetransformation differences) or absolute square difference and etc., obtain matching error when front template and gang form.
Certainly, more than only for for example, in the use of reality, step 1033 also can be obtained matching error when front template and gang form by additive method, does not do herein and gives unnecessary details.
The bi-directional predicted acquisition methods that the embodiment of the invention provides, can be according to working as front template and two-way matching error with reference to reference template in the formation, from reference template, derive prediction module adaptively, and obtain the prediction piece of current block correspondence according to prediction module, the technical scheme that makes the embodiment of the invention provide need be in the process of coding, the instructive information that will be used to obtain bi-directional predicted is enrolled code stream, has saved coding resource, has improved code efficiency; Further, prediction module can be a gang form optimum in the gang form of a plurality of alternative prediction modules, the prediction piece territory and the error between the current block that obtain according to this optimum combination template are littler, have improved the accuracy of current block encoding and decoding, can reach the effect of high-fidelity.
Further embodiment of this invention also provides a kind of bi-directional predicted acquisition methods, this method basic identical with as shown in Figure 1, difference is, the bi-directional predicted acquisition methods that present embodiment provides can also comprise the step of determining prediction direction, then at this moment, step 104 as shown in Figure 1 is specifically as follows according to the prediction module in the formation of described prediction direction corresponding reference, obtains the prediction piece of described current block correspondence.
Present embodiment can be determined prediction direction by three kinds of methods: a kind ofly be, obtain prediction direction information from the code stream that receives, determine prediction direction according to described prediction direction information; Another kind is, chooses strategy according to the prediction direction that sets in advance, and chooses prediction direction; The third is according to the matching error when front template and prediction module, to choose prediction direction.
Method to above three kinds of definite prediction direction is elaborated below:
One, from the code stream that receives, obtains prediction direction information, determine prediction direction according to described prediction direction information.
The bi-directional predicted acquisition methods that the embodiment of the invention provides both can be applied in coding side also can be applied in decoding end, when the bi-directional predicted acquisition methods that provides when the embodiment of the invention is applied in decoding end, decoding end can be after receiving the coding side transmitted stream, from code stream, decode prediction direction information, determine prediction direction according to this prediction direction information.
In the present embodiment, decoding end determines that according to prediction direction information the step of prediction direction can finish before step 101 as shown in Figure 1, and then this moment, step 101-104 as shown in Figure 1 all carries out in the formation of prediction direction corresponding reference.
Two, choose strategy according to the prediction direction that sets in advance, choose prediction direction.
In the present embodiment, coding side and decoding end all can be chosen strategy according to the prediction direction that sets in advance, choose prediction direction, this step can be finished before step 104 as shown in Figure 1, and wherein, prediction direction is chosen strategy and can be comprised as optimization criterias such as rate distortion principles, certainly, in the use of reality, prediction direction is chosen strategy can also be other strategies, every kind of situation is not given unnecessary details one by one herein.
Three, according to matching error, choose prediction direction when front template and prediction module.
Coding side and decoding end all can be according to the matching errors when front template and prediction module, choose prediction direction, at this moment, coding side does not need the prediction direction information of prediction direction correspondence is encoded in the code stream, can reach the saving coding resource, improve the effect of code efficiency.
In the present embodiment, according to the matching error when front template and prediction module, the step of choosing prediction direction can be after step 103 as shown in Figure 1, finishes before the step 104.
Present embodiment can be chosen prediction direction by several different methods according to the matching error when front template and prediction module:
For example: if the prediction module that step 103 is as shown in Figure 1 chosen is for (N is an integer from the two-way preceding N that chooses with reference to formation (List0 and List1) according to matching error order from small to large, and N 〉=2) individual reference template, wherein, comprise that (m is an integer to m, and m 〉=1) (n is an integer for the reference template of individual List0 and n, and the n 〉=1) reference template of individual List1, then the reference template of m List0 can be combined into the gang form of a List0, the reference template of n List1 is combined into the gang form of a List1, obtain the gang form and first matching error of working as front template of List0, and the gang form of List1 and second matching error of working as front template, the comparison value of first matching error and second matching error is compared with the threshold value that sets in advance, determine prediction direction.As, the threshold value that sets in advance is N, first matching error is cost0, second matching error is cost1, if cost0-cost1＞N determines that then prediction direction is List1, if cost1-cost0＞N, determine that then prediction direction is List0, if | cost0-cost1|≤N, determine that then prediction direction is two-way.
Certainly, more than only for for example, in the use of reality, can also choose prediction direction by additive method, every kind of situation is not given unnecessary details one by one herein.
The bi-directional predicted acquisition methods that the embodiment of the invention provides, can be according to working as front template and two-way matching error with reference to reference template in the formation, from reference template, derive prediction module adaptively, and obtain the prediction piece of current block correspondence according to prediction module, the technical scheme that makes the embodiment of the invention provide need be in the process of coding, the instructive information that will be used to obtain bi-directional predicted is enrolled code stream, has saved coding resource, has improved code efficiency; Further, owing to can determine prediction direction adaptively, make coding side (decoding end) can on prediction direction, obtain the prediction piece, improved the accuracy of obtaining the prediction piece, and then improved accuracy to the current block encoding and decoding, can reach the effect of high-fidelity.
As shown in Figure 3, the embodiment of the invention also provides a kind of bi-directional predicted deriving means, comprising:
Choose module 303, be used for the matching error obtained according to matching module 302, from the reference template that first acquisition module 301 obtains, choose prediction module when front template and reference template;
Further, as shown in Figure 4, choosing module 303 can comprise:
First chooses submodule 3031, is used for selecting from reference template the alternative prediction module more than two of matching error minimum;
First obtains submodule 3032, is used to obtain described a plurality of gang forms of alternative prediction module more than two;
Second obtains submodule 3033, is used to obtain each described gang form and the matching error of working as front template;
Second chooses submodule 3034, is used for selecting the gang form of matching error minimum as prediction module from described gang form.
Further, as shown in Figure 5, the bi-directional predicted deriving means that the embodiment of the invention provides can also comprise:
Further, as shown in Figure 6, determination module 305 can comprise:
First determines submodule 3051, is used for obtaining prediction direction information from the code stream that receives, and determines prediction direction according to prediction direction information; Or,
Second determines submodule 3052, is used for choosing strategy according to the prediction direction that sets in advance, and chooses prediction direction; Or,
The 3rd determines submodule 3053, is used for choosing prediction direction according to the matching error when front template and prediction module.
Need to prove, in the use of reality, more than as Fig. 3-embodiment of the invention shown in Figure 6 provides a plurality of modules that bi-directional predicted deriving means divides, can not give unnecessary details by a module or a functional module realization that has similar functions with described a plurality of modules herein yet.
The specific implementation method of the bi-directional predicted deriving means that the embodiment of the invention provides can be described referring to the bi-directional predicted acquisition methods that the embodiment of the invention provides, and repeats no more herein.
The bi-directional predicted deriving means that the embodiment of the invention provides, can be according to working as front template and two-way matching error with reference to reference template in the formation, from reference template, derive prediction module adaptively, and obtain the prediction piece of current block correspondence according to prediction module, the technical scheme that makes the embodiment of the invention provide need be in the process of coding, the instructive information that will be used to obtain bi-directional predicted is enrolled code stream, has saved coding resource, has improved code efficiency.
Bi-directional predicted acquisition methods that the embodiment of the invention provides and device and picture decoding method can expanded application as H.264/AVC wait in the video image encoding and decoding process.
The method of describing in conjunction with embodiment disclosed herein or the step of algorithm can directly use the software module of hardware, processor execution, and perhaps the combination of the two is implemented.Software module can place the storage medium of any other form known in random asccess memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or the technical field.
Method in the embodiment of the invention and device can be applied in the various electronic installations or with it and be associated, for example: mobile phone, wireless device, personal digital assistant (PDA), hand-hold type or portable computer, GPS receiver/omniselector, camera, audio/video player, video camera, video tape recorder, watch-dog etc.
The above; only be the specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; can expect easily changing or replacing, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion by described protection range with claim.
Claims (16)
1. one kind bi-directional predicted acquisition methods is characterized in that, comprising:
From two-way with reference to obtain the formation current block when front template corresponding reference template;
Obtain described matching error when front template and described reference template;
According to described matching error, from described reference template, choose prediction module when front template and described reference template;
Obtain the prediction piece of described current block correspondence according to described prediction module.
2. method according to claim 1 is characterized in that, and is described according to described matching error when front template and described reference template, chooses prediction module and comprise from described reference template:
From described reference template, choose at least one prediction module of matching error minimum.
3. method according to claim 1 is characterized in that, and is described according to described matching error when front template and described reference template, chooses prediction module and comprise from described reference template:
From described reference template, select the alternative prediction module more than two of matching error minimum;
Obtain described a plurality of gang forms of alternative prediction module more than two;
Obtain each described gang form and the matching error of working as front template;
From described gang form, select the gang form of matching error minimum as prediction module.
4. method according to claim 3 is characterized in that, the described prediction piece that obtains described current block correspondence according to described prediction module comprises:
Obtain the alternative prediction piece of each template correspondence in the gang form respectively, the compound mode according to gang form makes up described alternative prediction piece then, generates the prediction piece of current block correspondence.
5. method according to claim 1 is characterized in that, also comprises:
Determine prediction direction;
The described prediction piece that obtains described current block correspondence according to described prediction module is:
According to the prediction module in the formation of described prediction direction corresponding reference, obtain the prediction piece of described current block correspondence.
6. method according to claim 5 is characterized in that, described definite prediction direction comprises:
From the code stream that receives, obtain prediction direction information, determine prediction direction according to described prediction direction information; Or,
Choose strategy according to the prediction direction that sets in advance, choose prediction direction; Or,
According to described matching error, choose prediction direction when front template and described prediction module.
7. one kind bi-directional predicted deriving means is characterized in that, comprising:
First acquisition module, be used for from two-way with reference to formation obtain current block when front template corresponding reference template;
Matching module is used to obtain the matching error of the described reference template that obtains when front template and described first acquisition module;
Choose module, be used for the matching error obtained according to described matching module, from the reference template that described first acquisition module obtains, choose prediction module when front template and described reference template;
Second acquisition module is used for choosing the prediction piece that prediction module that module chooses obtains described current block correspondence according to described.
8. device according to claim 7 is characterized in that, the described module of choosing comprises:
First chooses submodule, is used for selecting from described reference template the alternative prediction module more than two of matching error minimum;
First obtains submodule, is used to obtain described a plurality of gang forms of alternative prediction module more than two;
Second obtains submodule, is used to obtain each described gang form and the matching error of working as front template;
Second chooses submodule, is used for selecting the gang form of matching error minimum as prediction module from described gang form.
9. device according to claim 7 is characterized in that, also comprises:
Determination module is used for determining prediction direction;
Described second acquisition module also is used for the prediction module according to the definite prediction direction corresponding reference formation of described determination module, obtains the prediction piece of described current block correspondence.
10. device according to claim 9 is characterized in that, described determination module comprises:
First determines submodule, is used for obtaining prediction direction information from the code stream that receives, and determines prediction direction according to described prediction direction information; Or,
Second determines submodule, is used for choosing strategy according to the prediction direction that sets in advance, and chooses prediction direction; Or,
The 3rd determines submodule, is used for choosing prediction direction according to described matching error when front template and described prediction module.
11. a picture decoding method is used for the B two field picture is decoded, and it is characterized in that, obtains the prediction piece of current block when decoding B two field picture as follows:
Obtain prediction direction;
From with the formation of prediction direction corresponding reference obtain current block when front template corresponding reference template;
Obtain described matching error when front template and described reference template;
According to described matching error, from described reference template, choose prediction module when front template and described reference template;
Obtain the prediction piece of described current block correspondence according to described prediction module, and utilize this prediction piece that current block is decoded.
12. method according to claim 11 is characterized in that, according to described matching error when front template and described reference template, chooses prediction module and comprise from described reference template:
From described reference template, select at least one prediction module of matching error minimum.
13. method according to claim 11 is characterized in that, according to described matching error when front template and described reference template, chooses prediction module and comprise from described reference template:
From described reference template, select the alternative prediction module more than two of matching error minimum;
Obtain described a plurality of gang forms of alternative prediction module more than two;
Obtain each described gang form and the matching error of working as front template;
From described gang form, select the gang form of matching error minimum as prediction module.
14. method according to claim 13 is characterized in that, obtains the prediction piece of described current block correspondence according to described prediction module, comprising:
Obtain the alternative prediction piece of each template correspondence in the gang form respectively, the compound mode according to gang form makes up described alternative prediction piece then, generates the prediction piece of current block correspondence.
15. method according to claim 11 is characterized in that, obtains the prediction piece of described current block correspondence according to described prediction module, comprising:
Obtain the alternative prediction piece of each template correspondence in a plurality of prediction modules respectively, described alternative prediction piece is made up, generate the prediction piece of current block correspondence.
16. method according to claim 11 is characterized in that, described definite prediction direction comprises:
From the code stream that receives, obtain prediction direction information, determine prediction direction according to described prediction direction information; Or
Choose strategy according to the prediction direction that sets in advance, choose prediction direction; Or
According to described matching error, choose prediction direction when front template and described prediction module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010147584 CN102215407B (en) | 2010-04-09 | 2010-04-09 | Bilaterally predicted block acquisition method, device and image decoding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010147584 CN102215407B (en) | 2010-04-09 | 2010-04-09 | Bilaterally predicted block acquisition method, device and image decoding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102215407A true CN102215407A (en) | 2011-10-12 |
| CN102215407B CN102215407B (en) | 2013-07-10 |
Family
ID=44746501
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010147584 Active CN102215407B (en) | 2010-04-09 | 2010-04-09 | Bilaterally predicted block acquisition method, device and image decoding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102215407B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104253997A (en) * | 2013-06-26 | 2014-12-31 | 鸿富锦精密工业（深圳）有限公司 | Intra prediction method and system |
| CN105338351A (en) * | 2014-05-28 | 2016-02-17 | 华为技术有限公司 | Template coupling based intraframe prediction encoding and decoding method and apparatus, and array scanning method and apparatus |
| WO2018120290A1 (en) * | 2016-12-26 | 2018-07-05 | 华为技术有限公司 | Prediction method and device based on template matching |
| WO2018120420A1 (en) * | 2016-12-26 | 2018-07-05 | 华为技术有限公司 | Prediction method and device based on template matching |
| CN110121073A (en) * | 2018-02-06 | 2019-08-13 | 浙江大学 | A kind of bidirectional interframe predictive method and device |
| CN113170176A (en) * | 2019-06-24 | 2021-07-23 | 华为技术有限公司 | Video encoder, video decoder and corresponding methods |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1925614A (en) * | 2005-08-30 | 2007-03-07 | 深圳清华大学研究院 | Rapid movement estimating method |
| CN101557514A (en) * | 2008-04-11 | 2009-10-14 | 华为技术有限公司 | Method, device and system for inter-frame predicting encoding and decoding |
| CN101605256A (en) * | 2008-06-12 | 2009-12-16 | 华为技术有限公司 | A kind of method of coding and decoding video and device |
| CN101686393A (en) * | 2008-09-28 | 2010-03-31 | 华为技术有限公司 | Fast-motion searching method and fast-motion searching device applied to template matching |
-
2010
- 2010-04-09 CN CN 201010147584 patent/CN102215407B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1925614A (en) * | 2005-08-30 | 2007-03-07 | 深圳清华大学研究院 | Rapid movement estimating method |
| CN101557514A (en) * | 2008-04-11 | 2009-10-14 | 华为技术有限公司 | Method, device and system for inter-frame predicting encoding and decoding |
| CN101605256A (en) * | 2008-06-12 | 2009-12-16 | 华为技术有限公司 | A kind of method of coding and decoding video and device |
| CN101686393A (en) * | 2008-09-28 | 2010-03-31 | 华为技术有限公司 | Fast-motion searching method and fast-motion searching device applied to template matching |
Non-Patent Citations (1)
| Title |
|---|
| STEFFEN KAMP ET AL.: "Decoder side motion vector derivation for inter frame video coding", 《IMAGE PROCESSING,2008.ICIP2008.15TH IEEE INTERNATIONAL CONFERENCE ON》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104253997A (en) * | 2013-06-26 | 2014-12-31 | 鸿富锦精密工业（深圳）有限公司 | Intra prediction method and system |
| CN105338351A (en) * | 2014-05-28 | 2016-02-17 | 华为技术有限公司 | Template coupling based intraframe prediction encoding and decoding method and apparatus, and array scanning method and apparatus |
| US10219003B2 (en) | 2014-05-28 | 2019-02-26 | Huawei Technologies Co., Ltd. | Intra-frame predictive coding and decoding methods based on template matching, array scanning method and apparatus, and apparatus |
| CN105338351B (en) * | 2014-05-28 | 2019-11-12 | 华为技术有限公司 | Intra prediction coding and decoding, array scanning method and device based on template matching |
| CN109565601B (en) * | 2016-12-26 | 2020-10-23 | 华为技术有限公司 | Template matching-based prediction method and device |
| WO2018120290A1 (en) * | 2016-12-26 | 2018-07-05 | 华为技术有限公司 | Prediction method and device based on template matching |
| WO2018120420A1 (en) * | 2016-12-26 | 2018-07-05 | 华为技术有限公司 | Prediction method and device based on template matching |
| CN109565601A (en) * | 2016-12-26 | 2019-04-02 | 华为技术有限公司 | A kind of prediction technique and device based on template matching |
| CN110121073A (en) * | 2018-02-06 | 2019-08-13 | 浙江大学 | A kind of bidirectional interframe predictive method and device |
| CN110121073B (en) * | 2018-02-06 | 2021-07-09 | 浙江大学 | A bidirectional inter-frame prediction method and device |
| CN113170176A (en) * | 2019-06-24 | 2021-07-23 | 华为技术有限公司 | Video encoder, video decoder and corresponding methods |
| CN113170176B (en) * | 2019-06-24 | 2022-04-26 | 华为技术有限公司 | Video encoder, video decoder and corresponding methods |
| US12028527B2 (en) | 2019-06-24 | 2024-07-02 | Huawei Technologies Co., Ltd. | Video encoder, video decoder, and corresponding method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102215407B (en) | 2013-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104363451B (en) | Image prediction method and relevant apparatus | |
| AU2019275631B2 (en) | Motion vector field coding and decoding method, coding apparatus, and decoding apparatus | |
| US20200244986A1 (en) | Picture prediction method and related apparatus | |
| US9781443B2 (en) | Motion vector encoding/decoding method and device and image encoding/decoding method and device using same | |
| CN102215407B (en) | Bilaterally predicted block acquisition method, device and image decoding method | |
| CN101816183B (en) | Method and apparatus for inter prediction encoding/decoding an image using sub-pixel motion estimation | |
| CN102316317B (en) | Method and device for generating predicted value of picture | |
| WO2016050051A1 (en) | Image prediction method and relevant device | |
| CN111385569A (en) | Coding and decoding method and equipment thereof | |
| US8229233B2 (en) | Method and apparatus for estimating and compensating spatiotemporal motion of image | |
| CN110933426B (en) | Decoding and encoding method and device thereof | |
| US20190098312A1 (en) | Image prediction method and related device | |
| CN102598670A (en) | Method and apparatus for encoding/decoding image with reference to a plurality of frames | |
| CN102710934B (en) | Motion predicting or compensating method | |
| CN100591136C (en) | A Video Intra-Frame Coding Method Based on Spatial Domain Decomposition | |
| JP2021512568A (en) | Less complex frame rate up conversion | |
| CN102017626B (en) | Method of coding, decoding, coder and decoder | |
| CN110062243B (en) | Light field video motion estimation method based on neighbor optimization | |
| CN102377992A (en) | Method and device for obtaining predicted value of motion vector | |
| CN104333758B (en) | The method and relevant apparatus of prediction technique and the detection pixel point of depth map | |
| CN102843554A (en) | Interframe image prediction encoding and decoding methods and video encoding and decoding device | |
| CN102595110A (en) | Video coding method, decoding method and terminal | |
| CN102215391B (en) | Video data encoding and decoding method and device as well as transform processing method and device | |
| CN114007078A (en) | Method, device and equipment for constructing motion information candidate list | |
| CN104662905A (en) | Inter prediction method using multiple hypothesis estimators and device therefor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |