CN104104955B - The decoding method and device of a kind of image block - Google Patents

The decoding method and device of a kind of image block Download PDF

Info

Publication number
CN104104955B
CN104104955B CN201310123316.5A CN201310123316A CN104104955B CN 104104955 B CN104104955 B CN 104104955B CN 201310123316 A CN201310123316 A CN 201310123316A CN 104104955 B CN104104955 B CN 104104955B
Authority
CN
China
Prior art keywords
block
division
sampled point
movable information
motion
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.)
Active
Application number
CN201310123316.5A
Other languages
Chinese (zh)
Other versions
CN104104955A (en
Inventor
杨海涛
昝劲文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
MIGU Culture Technology Co Ltd
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN201310123316.5A priority Critical patent/CN104104955B/en
Publication of CN104104955A publication Critical patent/CN104104955A/en
Application granted granted Critical
Publication of CN104104955B publication Critical patent/CN104104955B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The present invention provides a kind of decoding method and device of image block.Coding method determines the reference block of object block and the movable information of reference block including encoder；Sampled point in the reference block corresponds with the sampled point in the object block；The encoder carries out motion compensation using the movable information of the reference block to the reference block, obtains residual block；The value of each sampled point in the residual block is carried out motion division by the encoder compared with the division threshold value that the encoder determines, and according to the comparative result of each sampled point to the object block；The encoder encodes according to the result of the motion division to the object block.Compared with current Video Codec, on the premise of same video quality is ensured, the embodiment of the present invention can improve compression performance.

Description

The decoding method and device of a kind of image block
Technical field
The present invention relates to the decoding method and dress of video signal encoding and decoding technique, more particularly to a kind of image block Put.
Background technology
Intrinsic redundancy is divided into spatial redundancy information and time redundancy information to vision signal in itself.Spatial redundancy is believed Breath can be removed by way of orthogonal transform coding, and time redundancy information can be removed by the method for motion compensation.Working as In preceding general movement compensating algorithm, to a video block, one or more of dividing modes can be pre-defined.
Dividing mode of the prior art is fixed, for example, H.264/AVC the dividing mode defined in standard is symmetrical , the video macro block such as to a 16x16,16x16,8x16,16x8 or 8x8 etc. can be divided into；Defined in HEVC standard Dividing mode is outside 2Nx2N, Nx2N, 2NxN and NxN symmetrical division, in addition to asymmetrical division, as except above-mentioned Outside four kinds of dividing modes, 2N × nU, 2N × nD, nL × 2N or nR × 2N etc. can also be included.
Using fixed dividing mode, can not accurate description moving object border, encoder or decoder also can not just obtain Accurate movable information is taken, so as to which accurate motion compensated prediction can not be carried out, causes compression performance to be damaged.
The content of the invention
In view of this, it is existing to solve the embodiments of the invention provide a kind of decoding method of image block and device Present in technology because compression performance is impaired caused by image block division can not accurately embody object of which movement the problem of.
First aspect, there is provided a kind of image block coding method, methods described include：
Encoder determines the reference block of object block and the movable information of reference block, sampled point and institute in the reference block The sampled point stated in object block corresponds；
The encoder carries out motion compensation using the movable information of the reference block to the reference block, obtains residual error Block；
The division threshold value that the encoder determines the value of each sampled point in the residual block and the encoder It is compared, and motion division is carried out to the object block according to the comparative result of each sampled point；
The encoder encodes according to the result of the motion division to the object block.
The image block coding method of embodiment according to a first aspect of the present invention, the precision of motion compensation can be effectively improved And performance, and then lift the compression effectiveness of vision signal.
With reference in a first aspect, in the first possible implementation of first aspect, the division threshold value is described The average of whole residual values in residual block；Or, the intermediate value that the division threshold value is whole residual values in the residual block.
In second of mode in the cards of first aspect present invention embodiment with reference to above-mentioned one of any embodiment, institute The encoder is stated to be compared the division threshold value that the value of each sampled point in the residual block determines with the encoder Compared with, and according to comparative result to the object block carry out move division include：
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block second motion division.
With reference in above-mentioned one of any the third mode in the cards of the first aspect present invention embodiment of embodiment, institute The encoder is stated to be compared the division threshold value that the value of each sampled point in the residual block determines with the encoder Compared with, and according to comparative result to the object block carry out move division include：
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block first motion division；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the second motion division of the object block.
In the 4th kind of mode in the cards of first aspect present invention embodiment with reference to above-mentioned one of any embodiment, institute The movable information that encoder determines reference block is stated, including：
When the encoder encodes according to non-layered coded system to the former video block, obtain in following item At least one is as the reference block movable information：
The movable information of the video block adjacent with the former video block space；
The movable information of temporally adjacent video block with the former video block.
In the 5th kind of mode in the cards of first aspect present invention embodiment with reference to above-mentioned one of any embodiment, institute The reference block for determining object block is stated, including：
When the encoder encodes according to hierarchical coding mode to the former video block, weight is layered according to low quality The correspondence image block for building object block described in image determines the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than described The reconstruction image of layering where object block.
In the 6th kind of mode in the cards of first aspect present invention embodiment with reference to above-mentioned one of any embodiment, institute The movable information for determining reference block is stated, including：
When the encoder encodes according to hierarchical coding mode to the former video block, obtain in following item extremely Movable information of the one item missing as the reference block：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
In the 7th kind of mode in the cards of first aspect present invention embodiment with reference to above-mentioned one of any embodiment, institute The reference block for determining object block is stated, including：
When the encoder encodes according to non-layered coded system to the object block, by the reference frame of storage The 3rd image block be defined as reference block, the 3rd image block refers under the distortion metrics of setting, the reference of the storage There is the image block minimum with the distortion metrics of the object block in frame.
A kind of coding method of image block according to embodiments of the present invention, can effectively improve motion compensation precision and Performance, and then lift the compression effectiveness of vision signal；The efficiency of motion compensation can be improved simultaneously, reduce the computing thus brought Complexity requirement.
Second aspect, there is provided a kind of image block coding/decoding method, methods described include：
Decoder determines the reference block of object block and the movable information of reference block, sampled point and institute in the reference block The sampled point stated in object block corresponds；
The decoder carries out motion compensation using the movable information of the reference block to the reference block, obtains residual error Block；
The division threshold value that the decoder determines the value of each sampled point in the residual block and the decoder It is compared, and motion division is carried out to the object block according to the comparative result of each sampled point；
The decoder decodes according to the result of the motion division to the object block.
With reference to second aspect, in the first possible implementation of second aspect, the division threshold value is described The average of whole residual values in residual block；Or, the intermediate value that the division threshold value is whole residual values in the residual block.
With reference to the possible implementation of the first of second aspect or second aspect, second in second aspect is possible In implementation, division that the decoder determines the value of each sampled point in the residual block and the decoder Threshold value is compared, and according to comparative result to the object block carry out move division include：
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block second motion division.
With reference to the first of second aspect or second aspect to second of any possible implementation, in second aspect The third possible implementation in, the decoder is by the value of each sampled point in the residual block and the solution Code device determine division threshold value be compared, and according to comparative result to the object block carry out move division include：
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block first motion division；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the second motion division of the object block.
In the 4th kind of mode in the cards of second aspect of the present invention embodiment with reference to above-mentioned one of any embodiment, institute The reference block for determining object block is stated, including：
When the encoder encodes according to hierarchical coding mode to the former video block, the decoder is according to low The correspondence image block of object block described in mass segregation reconstruction image determines the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than described The reconstruction image of layering where object block.
In the 5th kind of mode in the cards of second aspect of the present invention embodiment with reference to above-mentioned one of any embodiment, institute The movable information for determining reference block is stated, including：
When the encoder encodes according to hierarchical coding mode to the former video block, the decoder obtains such as At least one in lower item is as the reference block movable information：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
In the 6th kind of mode in the cards of second aspect of the present invention embodiment with reference to above-mentioned one of any embodiment, institute The reference block for determining object block is stated, including：
When the encoder encodes according to non-layered coded system to the object block, the decoder will store Reference frame in the 3rd image block be defined as reference block, the 3rd image block refers under the distortion metrics of setting, described There is the image block minimum with the distortion metrics of the object block in the reference frame of storage.
The embodiment of third aspect present invention provides a kind of division device of image block, and described device is located at encoder In, described device includes：
Acquisition module, for obtaining the reference block of object block and the movable information of reference block, adopting in the reference block Sampling point corresponds with the sampled point in the object block；
Motion compensating module, the movable information for being obtained using the acquisition module are obtained to the acquisition module The reference block carry out motion compensation, obtain residual block；
Division module is moved, for the division for determining the value of each sampled point in the residual block and the encoder Threshold value is compared, and carries out motion division to the object block according to the comparative result of each sampled point；
Coding module, for being encoded according to the result of the motion division to the object block.
With reference to the third aspect, in the first possible implementation of the third aspect, the acquisition module is specifically used for：
Obtain the reference block of the former video block；
When the encoder encodes according to hierarchical coding mode to the former video block, obtain in following item extremely Movable information of the one item missing as the reference block：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
With reference to the third aspect, in second of possible implementation of the third aspect, the acquisition module is specifically used for：
When the encoder encodes according to hierarchical coding mode to the former video block, weight is layered according to low quality The correspondence image block for building object block described in image determines the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than described The reconstruction image of layering where object block；
Obtain the movable information of the reference block.
With reference to the first of the third aspect or the third aspect to second of any possible implementation, in the third aspect The third possible implementation in, it is described motion division module be specifically used for：
By the value of each sampled point in the residual block compared with the division threshold value that the encoder determines；
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block second motion division.
With reference to the first of the third aspect or the third aspect to second of any possible implementation, in the third aspect The 4th kind of possible implementation in, it is described motion division module be specifically used for：
By the value of each sampled point in the residual block compared with the division threshold value that the encoder determines；
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block first motion division；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the second motion division of the object block.
A kind of fourth aspect, there is provided the division device of image block, it is characterised in that described device is located in decoder, Described device includes：
Acquisition module, for obtaining the reference block of object block and the movable information of reference block, adopting in the reference block Sampling point corresponds with the sampled point in the object block；
Motion compensating module, the movable information for being obtained using the acquisition module are obtained to the acquisition module The reference block carry out motion compensation, obtain residual block；
Division module is moved, for the division for determining the value of each sampled point in the residual block and the decoder Threshold value is compared, and carries out motion division to the object block according to the comparative result of each sampled point；
Decoder module, for being decoded according to the result of the motion division to the object block.
With reference to fourth aspect, in the first possible implementation of fourth aspect, the acquisition module is specifically used for：
Obtain the reference block of the former video block；
When the encoder encodes according to hierarchical coding mode to the former video block, obtain in following item extremely Movable information of the one item missing as the reference block：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
With reference to the possible implementation of the first of fourth aspect or fourth aspect, second in fourth aspect is possible In implementation, the acquisition module is specifically used for：
When the encoder encodes according to hierarchical coding mode to the former video block, weight is layered according to low quality The correspondence image block for building object block described in image determines the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than described The reconstruction image of layering where object block；
Obtain the movable information of the reference block.
With reference to the first of fourth aspect or fourth aspect to second of any possible implementation, in fourth aspect The third possible implementation in, it is described motion division module be specifically used for：
By the value of each sampled point in the residual block compared with the division threshold value that the encoder determines；
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block first motion division；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the second motion division of the object block.
In the 4th kind of mode in the cards of fourth aspect present invention embodiment with reference to above-mentioned one of any embodiment, institute Motion division module is stated to be specifically used for：
By the value of each sampled point in the residual block compared with the division threshold value that the encoder determines；
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that right in the object block The sampled point answered belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that corresponding in the object block Sampled point belong to the object block second motion division.
Pass through above-mentioned technical proposal, motion compensation is carried out to reference block using available motion information, and motion compensation is obtained The residual block arrived carries out statistical analysis, and reference block is divided according to the result of statistical analysis, can be completed to reference block Adaptive division, can obtain more accurately movable information, and then can carry out accurate motion compensation by adaptively dividing Prediction, improve compression performance.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, make required in being described below to embodiment Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are some embodiments of the present invention, for this For the those of ordinary skill of field, without having to pay creative labor, other can also be obtained according to these accompanying drawings Accompanying drawing.
Fig. 1 is a kind of schematic flow sheet of image block coding method provided in an embodiment of the present invention；
Fig. 2 is the schematic flow sheet of another image block coding method provided in an embodiment of the present invention；
Fig. 3 is the generation schematic diagram adaptively divided in the embodiment of the present invention；
Fig. 4 is to obtain the schematic flow sheet of rate distortion costs value according to dividing mode in the embodiment of the present invention；
Fig. 5 is to complete video division with being completed using the adaptive division of the embodiment of the present invention using HEVC in the prior art The comparison schematic diagram of video division；
Fig. 6 is a kind of schematic flow sheet of coding method provided in an embodiment of the present invention；
Fig. 7 is the structural representation of encoder corresponding to Fig. 6；
Fig. 8 is a kind of schematic flow sheet of coding/decoding method provided in an embodiment of the present invention；
Fig. 9 is the structural representation of decoder corresponding to Fig. 8；
Figure 10 is a kind of structural representation of the division device of image block of encoder provided in an embodiment of the present invention.
Figure 11 is a kind of structural representation of the division device of image block of decoder provided in an embodiment of the present invention.
Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is Part of the embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
Fig. 1 is a kind of schematic flow sheet of the division methods of video block provided in an embodiment of the present invention, including：
11：Encoder determines the reference block of object block and the movable information of reference block, the sampled point in the reference block Corresponded with the sampled point in the object block.
In one embodiment of the invention, for example, can be to input (can be with for the current coded unit in enhancement layer It is regular shape or irregular shape), determine available motion information
To the current coded unit in enhancement layer, find corresponding Primary layer coding unit and obtain Primary layer coding unit Correlative coding information, including but not limited to movable information etc..
Determine the movable information in available motion information list.Movable information in movable information list is essentially from three Aspect：The adjacent movable information in the space of current coded unit in enhancement layer；Current coded unit is temporally adjacent in enhancement layer Movable information；The movable information of Primary layer coding unit corresponding to current coded unit in enhancement layer.Available motion packet In terms of including but being not limited to three above, such as can also add can for the room and time movable information of corresponding basic layer unit With movable information list.Movable information quantity in movable information list can be the number made an appointment or be taken from code stream , movable information therein can have identical or must be different.
In one embodiment of the invention, when corresponding blocks are identical with object block resolution ratio, using corresponding blocks as reference Block, corresponding blocks are otherwise subjected to resampling, using the image block after resampling as reference block.The purpose of resampling is to change to correspond to The resolution ratio of block, make it identical with object block.
In one embodiment of the invention, the coordinate of corresponding blocks exists corresponding with the coordinate of size and object block with size Relation.
In one embodiment of the invention, the video of input can be divided, without loss of generality, it is assumed that obtain big It is small be 64 × 64 multiple nonoverlapping video blocks, former video block is properly termed as to the video block of one 64 × 64 to be encoded.
In coding side, encoder can be encoded to obtain bit stream to the former video block, and in decoding end, decoder can be with The bit stream of reception is decoded to recover the former video block.
In addition, the video block of the 64x64 sizes is only a kind of example, the current video block in practical application can also be it His size.
Available motion information refers to currently available movable information, for example, in coding side, available motion information is included but simultaneously It is not limited to the movable information of the adjacent encoded video block of former video block.
Specifically, for encoder, in the application of non-layered Video coding, available motion information is including but not limited to as follows At least one of in：
The movable information of the video block adjacent with former video block space；
The movable information of temporally adjacent video block with former video block.
In layered video coding application, available motion information includes but is not limited at least one in following item：
The movable information of the video block adjacent with former video block space；
The movable information of temporally adjacent video block with former video block；
The movable information of first video block；
The movable information of the video block adjacent with the first video block space；
The movable information of temporally adjacent video block with the first video block；
Wherein, the first video block refers to video block corresponding with former video block in first layer, first layer refer to the time, The resolution ratio of at least one in space, signal to noise ratio is less than the layering being layered where former video block.For example, where former video block Enhancement layer is layered as, first layer refers to Primary layer.
In the embodiment of the present invention, when time layering and/or signal to noise ratio are layered, " corresponding " regards in Primary layer and enhancement layer Frequency block refers to position identical video block, such as when being layered the time, frame that Primary layer and enhancement layer use when being encoded Time interval is different, but the video block used is at same location, and the frame number as Primary layer encodes is respectively 0,2,4,6..., The frame number of enhancement layer coding is respectively 0,1,2,3..., and the first video block is located at the 2nd frame of Primary layer, and position for (x0~xn, Y0~ym), former video block is located at the 1st frame of enhancement layer and position is also (x0~xn, y0~ym).In space delamination, substantially " corresponding " video block refers to that the correlation between pixel quantity and position meets regarding for spatial resolution in layer and enhancement layer Frequency block, for example, the ratio between spatial resolution of enhancement layer and Primary layer be 2: 1, then assuming that the position of the first video block for (x0~ Xn, y0~ym), then the position of former video block is (x0/2~xn/2, y0/2~ym/2).
It is understood that above-mentioned available motion information is a kind of example, the motion that other manner obtains can also be Information, obtained for example, the movable information of video block that can be adjacent to the time of former video block and/or space is weighted processing etc. The movable information arrived, it might even be possible to be independently of the movable information of time and/or the space adjacent video block of former video block.
Above-mentioned movable information includes but is not limited at least one in following item：Motion vector, such as motion vector is in x Direction and the component values in y directions；Predictive mode, such as single directional prediction or bi-directional predicted；And reference image frame is referring to Index in image list etc..
For decoder, available motion information can be obtained in the bit stream of reception.For example, encoder and decoder can So that above-mentioned available motion information is formed into available motion information list, each available motion information arranges in the available motion information There is an index, encoder, which exports, can include the available motion information that encoder uses in the bit stream to decoder in table Index, decoder, by searching available motion information list, obtains available motion letter according to the index of the available motion information Breath.Or it can also pass through in the bit stream of decoder reception directly comprising the bit after being encoded to available motion information Decoded bits obtain available motion information.
I.e., optionally, for decoder, the available motion information for obtaining former video block, including：
The decoder obtains available motion information from the bit stream；
Wherein, the bit stream includes the available motion information index, with the rope according to the available motion information Draw, and pre-establish with encoder identical available motion information list, search and determine the available motion information；Or Person,
The bit stream includes the bit after the coding of the available motion information, and obtaining described can use by decoding transports Dynamic information.
Above-mentioned reference block is that coding side and decoding end are available so that the reference block that encoder uses is adopted with decoder Reference block is identical.
In hierarchical coding, reference block can be pre-defined in the encoder and the decoder, specifically can will be with former video The video block that block has same spatial resolution is defined as reference block.
When non-layered encodes, obtained in the reference frame that reference block can store.For encoder, the former video block is can , therefore the immediate video block of division in the reference frame of storage with former video block can be defined as reference block. Afterwards, the relative displacement information between former video block and reference block can be included in the bit stream that coding obtains so that decoding Device is according to the relative displacement acquisition of information reference block.For example, in non-layered coding application, because coding side original video block can , divide as motion compensation to obtain division one directly using former video block, using a certain video block in frame memory as joining Division two can be obtained by examining block and making motion compensation division, encoder selection optimal reference block be exactly under a certain distortion criterion, it is right The division two of this reference block is minimum relative to the distortion of division one, the relative displacement between optimal reference block and current video block with And reference block deposits Zhu Qizhong position in frame and is defined as dividing vector (partition vector) or shape vector together (shape vector)；In hierarchical coding application, due to the strong correlation of Primary layer and enhancement layer, the second video block (second The definition of video block may refer to subsequent descriptions) can be reference block the selection of one, certainly, the now selection of reference block It can use and be similar in non-layered coding application, a reference block is found in frame memory.
That is, for encoder, optionally, the reference block of former video block is obtained, including：
When the encoder is in hierarchical coding, the second video block is defined as reference block, second video block and institute Stating former video block has identical spatial resolution, wherein, second video block refer in the second layering with former video block pair The video block answered, second layering refer to what the resolution ratio of time and/or signal to noise ratio was layered less than the former video block place Layering, or, second video block refers to the video block corresponding with the former video block up-sampled in the 3rd layering, described 3rd layering refers to layering of the spatial resolution less than layering where the former video block；Or
When the encoder encodes in non-layered, the 3rd video block in the reference frame of storage is defined as reference block, 3rd video block refers under the division distortion metrics of setting, with the closest division of division with the former video block Video block.
For decoder, optionally, the reference block of former video block is obtained, including：
In hierarchical coding, the reference block that the decoder makes an appointment encoder and decoder is defined as the decoding The reference block that device uses (reference block now is deduced by codec, and do not need extra bit to spend)；Or
Included when non-layered encodes, in the bit stream and be used for the configured information for indicating the reference block that encoder uses, The decoder determines the reference block of the decoder use according to the configured information, wherein, if the reference block is by one Individual video block composition, location index of the reference frame in frame memory where the configured information includes the reference block, and The relative displacement of the reference block and the former video block；Or if the reference block is formed by the weighting of multiple video blocks, institute Stating configured information includes location index of each video block place reference frame in frame memory, each video block and former video block Relative displacement, and the weight of each video block.
After available motion information and reference block is obtained, first equipment can be right according to the available motion information The reference block is adaptively divided.
Wherein, adaptive division refers to be not limited to fixed partition mode in the prior art, such as is not limited to H.264/AVC Symmetrical dividing mode, HEVC asymmetric dividing mode is also not necessarily limited to, the adaptive division that the embodiment of the present invention proposes is simultaneously unlimited Can be any division shape in rectangular partition.
Above-mentioned adaptive division can be carried out according to motion compensation, i.e. it is optional, fortune can be used described in the basis Dynamic information, is adaptively divided to the reference block, including：
According to the available motion information, using motion compensation approach, the reference block is adaptively divided.
Optionally, it is described according to the available motion information, using motion compensation approach, the reference block is carried out adaptive It should divide, including：
12：The encoder carries out motion compensation using the movable information of the reference block to the reference block, obtains residual Poor block.
13：The encoder determines the value of each sampled point in the residual block and the encoder
Division threshold value is compared, and according to the comparative result of each sampled point to the target
Block carries out motion division.
Wherein, the intermediate value of residual block refers to the numerical value of all residual blocks being ranked up, and then takes centre
Data.
In one embodiment of the invention, the pixel that the value of residual block is less than to the intermediate value of residual block forms
Be divided into division 0, otherwise be division 1.
In one embodiment of the invention, the value of residual block is less than or equal to the pixel of the intermediate value of residual block
Composition is divided into division 0, is otherwise division 1.
14：The encoder encodes according to the result of the motion division to the object block.
Wherein, each or every group of available motion information can correspond to a kind of dividing mode.It is understood that every group can use Movable information refers to include at least two available motion information；A kind of dividing mode can divide to obtain two or at least three Video block.
In one embodiment of the invention, object block is encoded according to motion division, including：
The movable information of described two motion divisions is determined by estimation；
According to identified movable information, motion compensation is carried out to described two motion divisions and obtains the pre- of the object block Survey, and the residual error of the object block is calculated according to the prediction；
The residual error of the movable information of described two motion divisions and the object block is write into code stream.
Can be according to an available motion letter for example, when each available motion information corresponds to a kind of dividing mode Breath, obtains a residual block, and according to the statistical analysis of a residual block, division obtains two video blocks, or at least three Video block.Or
When every group of available motion information corresponds to a kind of dividing mode, it can be according at least two movable informations, obtain At least two residual blocks, according to the statistical analysis of at least two residual block, division obtains at least three video blocks.Above-mentioned division When to each residual block carry out statistical analysis when, one or at least two threshold values can be set.
Because motion compensating module is the module that is generally used in existing Video Codec, therefore on the one hand the present invention is real Prior art can be made full use of by applying the implementation of example proposition, and strengthen the embodiment of the present invention applies range and maturity； On the other hand realize that adaptive division can reduce computational complexity using motion compensating module, motion complexity is limited to can The scope of receiving.
Optionally, when for encoder when, can be with after to reference block adaptively divide according to available motion information Including：
The encoder determines the dividing mode of the former video block according to the adaptive division to the reference block；
The encoder is divided and encoded to the former video block according to the dividing mode of the former video block, Bit stream after being encoded simultaneously is exported to decoder.
Adaptive division of the above-mentioned basis to reference block, it is determined that the dividing mode of former video can refer to：Due to each Available motion information or its combination can correspond to a kind of adaptive division to reference block, it is determined that after a kind of adaptive division, The adaptive division can be applied in former video block, former video block is divided and encoded using the division again afterwards, A kind of evaluation of estimate is obtained, the optimal dividing mode of evaluation of estimate is defined as to the dividing mode of former video block.The evaluation of estimate include but It is not limited to rate distortion costs value.
After the dividing mode of former video block is determined, former video block can be divided using the dividing mode and Encode, can include in the bit stream after coding：Bit corresponding to dividing mode, the dividing mode are shown to be adaptive division； Bit corresponding to movable information, the bit can be index of the movable information in available motion information list, and e.g., MV0 is can It is 0 with the index in movable information list, then bit corresponding to the movable information is 0, or, compare corresponding to the movable information Spy can be exactly the bit after directly being encoded to MV0.It is understood that this sentences an available motion information (MV0) Exemplified by a kind of corresponding dividing mode, optionally, when dividing mode is corresponding at least two available motion information, then include at least Bit corresponding to two available motion information, such as including at least two available operation informations in available motion information list Index, or the bit after directly being encoded to available motion information.
The above-mentioned dividing mode given when the first equipment is encoder, in encoder-side due to needing from a variety of possibility Dividing mode in determine a kind of final dividing mode, therefore above-mentioned cost value comparison procedure be present.And for decoding Device, because encoder can include dividing mode and movable information in the bit stream of output, decoder can be directly from the ratio Dividing mode and movable information are determined in spy's stream, is decoded using these information, to recover former video block.
I.e., optionally, also include for decoder, methods described：
The decoder receives bit stream；
The decoder decodes, to recover according to the adaptive division to the reference block to the bit stream State former video block.
The present embodiment completes the adaptive division to reference block by using available motion information, in coding side to reference block Adaptive division may apply in former video block so that former video block completes adaptive division, can basis in decoding end Adaptive division to reference block recovers former video block, and more accurately movable information can be obtained by adaptively dividing, and then Accurate motion compensated prediction can be carried out, improves compression performance.
Referring to Fig. 2, the flow signal of the dividing mode of former video block is determined the embodiments of the invention provide a kind of encoder Figure, without loss of generality, assumes that former video block has two divisions in the present embodiment.By the value of each sampled point in residual block with The division threshold value that encoder determines is compared, and object block is transported according to the comparative result of each sampled point Dynamic division.
Wherein, the intermediate value of residual block refers to the numerical value of all residual blocks being ranked up, and then takes the data of centre；.
Or to the Quadtree Partition that former video block is iterated, then make two divisions on video block after singulation.This Embodiment can include：
21：Corresponding each available motion information, motion compensation is carried out to reference block using the available motion information, obtained residual Poor block；
For example, the available motion information used is MV0, then reference block can be transported using the MV0 first Dynamic compensation, obtains corresponding residual block.
Referring to Fig. 3, by taking hierarchical coding as an example, it is above-mentioned use available motion information to carry out motion compensation to reference block can be with Including：
0th step, determine that the available motion information of the former video block of enhancement layer (shows 3 movable informations, MV0, MV1 in Fig. 3 And MV2)；
1st step, it is assumed that reference block be Primary layer the first video block, the first video block refer in Primary layer with former video block Corresponding video block；
The picture material strong correlation that the picture material that first video block is included is included with former video, this strong correlation are Strong correlation based on encoded video in hierarchical coding.
In one embodiment of the invention, the content of the first video block and the content of former video are closely similar.Example Such as all it is that a same part for an object is encoded.
2nd step, using an available motion information (such as MV0), motion compensation is carried out to reference block and obtains prediction block；
3rd step, prediction block subtract each other to obtain residual block with reference block.
22：Statistical analysis is carried out to the residual block, obtains dividing mode corresponding to the available motion information of reference block.
The present embodiment is so that each available motion information corresponds to a kind of dividing mode as an example, it is to be understood that can also be One group of available motion information corresponds to a kind of dividing mode.
Referring to Fig. 3,22 particular content can include：
4th step, statistical analysis is carried out to residual block；
5th step, the dividing mode according to corresponding to the result of statistical analysis obtains the available motion information.
The statistical analysis is the picture material based on video block, is adhered to separately not according to the picture material with different motion information With the principle of zoning, for example, picture material includes A and B, for a good partitioning algorithm, when A and B has identical fortune It belongs to same divided block during dynamic information, and when A and B has different motion information, it is belonging respectively to different divided blocks.
Because above-mentioned statistical analysis is the movable information of image content-based, therefore when picture material has different fortune During dynamic information, will there is the characteristics of adaptive with different dividing modes, i.e. division.
Specifically, statistical analysis includes but is not limited to：Calculate residual block pixel absolute value average, and using this average as Threshold value, dividing mode is determined by the comparison of residual block pixel absolute value and this threshold value；Or by calculating residual block pixel Grad determines dividing mode；The intermediate value of residual block pixel is calculated, or calculates the average value of residual block pixel.
For example, when the absolute value of some residual block pixel exceedes the average threshold value being calculated, then the pixel is not with surpassing The pixel for crossing threshold value belongs to different divisions.Or the Grad of some residual block pixel exceedes default threshold value, then the ladder The pixel of angle value both sides belongs to different divisions.
Referring to Fig. 3, so that the image block after division is two pieces as an example, Fig. 3 lower left corners give a specific division result Example.The division is the division on reference block.
23：The division of reference block is applied on former video block.
Referring to Fig. 3 the 6th step, will be used to divide former video block with reference block identical dividing mode, such as Fig. 3 upper right corner It is shown.
MV0 dividing mode is described above, it is similar, MV1 and MV2 dividing mode can be obtained.Wherein, fortune can be used Dynamic information is one-to-one with dividing mode.
It is understood that the present embodiment, by taking three movable informations as an example, the method for the embodiment of the present invention is not limited to this Three movable informations, the movable information can be multiple, and the plurality of movable information can form a list, referred to as movable information List, and each movable information can be corresponded in the movable information list and uniquely indexed, and therefore, each movable information can To be indexed with a movable information.
24：Former video block is encoded using every kind of dividing mode, obtains evaluation of estimate corresponding to every kind of dividing mode, will The optimal dividing mode of evaluation of estimate is defined as the dividing mode of former video block.
Wherein, evaluation of estimate can refer to rate distortion costs value, and now, evaluation of estimate is optimal to refer to rate distortion costs value minimum. It is understood that the evaluation of estimate is not limited to rate distortion costs value, such as every kind of division side can also be determined according to residual values Residual values corresponding to formula, the minimum dividing mode of residual values is defined as final dividing mode.The embodiment of the present invention is with rate distortion Exemplified by cost value.
As it appears from the above, every kind of movable information will correspond to a kind of dividing mode, one can be obtained again according to every kind of dividing mode Individual rate distortion costs value, after rate distortion costs value corresponding to all movable informations difference is calculated, it is possible to determine final Dividing mode.For example, using the rate distortion costs value that dividing mode corresponding to MV0 obtains as cost0, drawn using corresponding to MV1 The rate distortion costs value that point mode obtains is costl, use the rate distortion costs value that dividing mode corresponding to MV2 obtains for Cost2, it is assumed that cost0 is minimum, then dividing mode corresponding to MV0 can be defined as into final dividing mode.
The present embodiment uses motion compensating module in adaptive division, can because motion compensating module is general module To reduce motion complexity and improve applicability.
Optionally, referring to Fig. 4, former video block is encoded using every kind of dividing mode, obtains every kind of dividing mode pair The rate distortion costs value answered, can include：
41：Corresponding every kind of dividing mode, the video block after being divided, and determine the motion of the video block after each division Information；
Optionally, estimation can be carried out to the video block after each division, obtains corresponding movable information.For example, Assuming that obtain two video blocks, respectively the first video block and the second video block after being divided to former video block, then, can be to the One video block carries out estimation, obtains movable information corresponding to the first video block, and, motion is carried out to the second video block and estimated Meter, obtains movable information corresponding to the second video block.Or
Optionally, can also be using the available motion information of use as regarding after the division adjacent with the available motion information The movable information of frequency block.For example, the movable information used is MV0, it is assumed that the video block adjacent with the MV0 regards for first after division Frequency block (video block in the left side after such as Fig. 3 division), then the movable information of the first video block can be defined as MV0, similar , as Fig. 3 division after the movable information of video block on right side can be defined as MV1.
42：According to the movable information of the video block after each division, the video block after each division is moved Compensation, obtain the prediction block of the video block after the prediction block of the video block after each division, and each division with The residual block of video block after each division；
For example, the video block after division is the first video block and the second video block, movable information corresponding to two video blocks For the first movable information and the second movable information, then motion benefit can be carried out to the first image block using the first movable information Repay, obtain the first prediction block, the first prediction block is the prediction block of the first video block, and obtains the first prediction block and the first video block Residual block, the residual block is properly termed as the first residual block；Similar, the second video block can be entered using the second movable information Row motion compensation, the second prediction block is obtained, the second prediction block is the prediction block of the second video block, and obtains the second prediction block and the The residual block of two video blocks, the residual block are properly termed as the second residual block.
It is understood that the particular content of prediction block and residual block is obtained in the embodiment of the present invention during motion compensation, And estimation obtains the particular content of movable information and may refer to the specific of motion compensation in the prior art and estimation Content, it will not be repeated here.
43：According to residual block corresponding to the video block after each division, carry out orthogonal transformation, quantization, re-quantization and it is inverse just After alternation changes, the residual block of the reconstruct of the former video block is obtained.
Optionally, orthogonal transformation can be regular shape conversion, for example, rectangular transform, when using regular shape conversion, Split first can be carried out to the residual block obtained after each division, regular residual block corresponding to former video block be obtained, such as by first Residual block and the second residual block carry out split, carry out orthogonal transformation etc. to the residual block after split again afterwards.Or
Orthogonal transformation can also be non-regular shape conversion, at this point it is possible to respectively to the first residual block and the second residual block Carry out orthogonal transformation etc..
Residual block quantifies through orthogonal transformation, one-dimensional quantization transform coefficient is obtained after scanning；Phase is made to quantization transform coefficient The inverse scan answered, re-quantization, inverse orthogonal transformation obtain the residual block of the reconstruct of former video block.
44：The prediction block of former video block is obtained according to the prediction block of the video block after each division, according to the original The residual block of the reconstruct of the prediction block of video block and the former video block, the former video block reconstructed.
Wherein it is possible to the prediction block of source video block is obtained after the first prediction block and the second prediction block are carried out into split, afterwards, The prediction block of source video block is added with the residual block of the reconstruct of the former video block again, the former video block reconstructed.
In addition, ginseng can be used as after former video block/frame of reconstruct post-treated (including but not limited to deblocking effect etc.) Frame deposit frame memory is examined, for being used during encoded image frame estimation below.
45：Compare the former video block of the reconstruct and the former video block, obtain coding distortion, and by described adaptive Coding mode corresponding to division is encoded to obtain coded-bit to former video block, according to coding distortion and coded-bit, is obtained Rate distortion costs value.
Above-mentioned coding distortion can be obtained by comparing the former video block of reconstruct with former video block.
Above-mentioned coded-bit can include：It is the bit adaptively divided to show dividing mode, shows former video block The bit of available motion information corresponding to dividing mode.
Available motion information can form available motion information list, and the movable information included in above-mentioned coded-bit can be with It is the index of available motion information, or directly the movable information of use is encoded.The motion letter of video block after division Breath can be the information that direct coding obtains, or can also be the difference of available motion information corresponding with final dividing mode Information.
I.e., optionally, the coded-bit includes：
Bit corresponding to dividing mode, the dividing mode show to be divided into adaptive division；And
Bit corresponding to first available motion information, bit corresponding to the first available motion information can for described first With index of the movable information in the available motion information list, or, bit corresponding to the first available motion information For the bit after being encoded to the first available motion information.
For example, the movable information of the final division (dividing mode for the former video block determined) of generation is MV0, it is assumed that Indexes of the MV0 in movable information list is 0, then movable information corresponding to the dividing mode included in the coded-bit is 0； Needed it is understood that be one group of available motion information corresponding to the dividing mode, during coding to one group of available motion letter Breath is encoded.
Further, if carrying out estimation to each video block after division, coded-bit can also include each Bit corresponding to the movable information of video block after division, bit corresponding to the movable information include but is not limited to：Direct coding Obtained bit, or the difference value of the movable information movable information corresponding with each image block after division for generating division Coded-bit.By taking difference value as an example, for example, the video block after division is the first video block and the second video block, the first video Movable information corresponding to block is the first movable information, and movable information corresponding to the second video block is the second movable information, for giving birth to Movable information collection into division is combined into { MV0, MV1, MV2 }, it is assumed that the first movable information connects the most with MV0 in movable information set Closely, and the second movable information and MV1 in movable information set are closest, then the difference also including MV0 and the first movable information Score value and MV1 and the second movable information difference value.It is described closest to refer to use some in movable information set When movable information encodes to the first or second movable information, the ratio of the difference value needs in corresponding index and x and y directions For example, it is assumed that the first movable information is NMV, corresponding above-mentioned three movable informations MV0, MV1 and MV3 can be obtained the special minimum of number Arrive：(1) index value 0, NMV (x)-MV0 (x), NMV (y)-MV0 (y)；(2) index value 1, NMV (x)-MV1 (x), NMV (y)-MV1 (y)；(3) index value 2, NMV (x)-MV2 (x), NMV (y)-MV2 (y)；Above-mentioned three kinds of information is encoded respectively, obtains three Bit after kind coding, it is assumed that minimum to the encoding bit overhead of (1), then MV0 is NMV immediate movable information.It is above-mentioned (x) and (y) respectively represent corresponding sports information x and y directions value, as NMV (x) represent NMV in the value in x directions, NMV (y) values of the NMV in y directions is represented.In addition, coded-bit can also include：The structural information of orthogonal transformation, complete the zero of minor structure Coefficient indicates, and orthogonal transform coefficient quantified etc..
After obtaining coding distortion and coded-bit, it is possible to obtain rate distortion costs value.Wherein, rate distortion costs value is not Only include the bit number that distortion metrics also include coded-bit corresponding to the distortion, for example, being obtained using dividing mode corresponding to MV0 To coding distortion have identical distortion metrics with the coding distortion obtained using the corresponding dividing modes of MV1, but corresponding to MV0 The bit number for the coded-bit that dividing mode obtains is less than the bit number for the coded-bit that dividing mode corresponding to MV1 obtains, that It can determine that rate distortion costs value corresponding to dividing mode corresponding to MV0 is less than rate corresponding to dividing mode corresponding to MV1 and lost True cost value.
The present embodiment can reduce overhead bit by the index in coding using movable information.When there is new motion When information coding method can further improve the code efficiency of movable information, thought proposed by the present invention stands good.
Adaptive division to former video block can be completed by foregoing description, for example, with reference to Fig. 5, to be existing on the left of Fig. 5 HEVC divisions an example, be the division completed using the embodiment of the present invention on the right side of Fig. 5, be all bent in two of which example Scope where on the right side of line is moving object, and left side is static background.From fig. 5, it can be seen that although existing HEVC divisions are right The division that former video block has carried out profound level can not accurately still describe moving object, thus reduce forecasting accuracy, influence to press Contracting performance, and by the way of the embodiment of the present invention, effectively moving object can be marked off to come, improves forecasting accuracy, And then improve compression performance.
After the dividing mode of former video block is determined, it can complete to encode according to the dividing mode.For example, with reference to Fig. 6 And corresponding Fig. 7 structural representation, a kind of schematic flow sheet of the coding method proposed for the embodiment of the present invention, including：
61：Former video block is divided using the dividing mode of former video block, the video block after being divided；
It is determined that the particular content of the dividing mode of former video block may refer to the embodiment shown in above-mentioned Fig. 2.
The step can be as shown in Figure 7 division module complete.
62：Motion compensation is carried out to the video block after each division, obtains prediction block corresponding to each video block, and often Prediction block corresponding to individual video block and the residual block of the video block；
The step can be as shown in Figure 7 motion compensating module complete, residual block can be obtained, in addition, prediction when also Header information can be obtained, header information includes the segmentation size of the video block after predictive mode, each division, movable information Deng.
63：Residual coding is obtained after carrying out orthogonal transformation, quantization and coefficient scanning to residual block；
Residual coding includes：The structural information of orthogonal transformation, the full zero coefficient instruction of minor structure, and the orthogonal transformation quantified Coefficient etc..
Orthogonal transformation module, quantization modules and the coefficient scanning module that the step can distinguish as shown in Figure 7 are completed.
64：Entropy code is carried out to residual coding and header information, the bit stream after being encoded, exported afterwards to decoding Device.
The step can be as shown in Figure 7 entropy code module complete.
In addition, reconstructed video block can be stored in frame storaging piece, for the coded reference of subsequent video frame.Such as Shown in Fig. 7, inverse quantization module, inverse orthogonal transformation module, frame storaging piece can also be included.
The behavior of coding side is described above, the corresponding behavior for decoding side can be with following examples.
Referring to Fig. 8 and corresponding structural representation Fig. 9, the flow that decoder performs can include：
81：Decoder carries out entropy decoding to the bit stream of reception.
The step can be as shown in Figure 9 entropy decoder module complete, obtain header information and 1 and maintain number.
82：Decoder divides to reference block.
The step can specifically include：
Available motion information is obtained from the bit stream (header information described above) of reception.Assuming that what is included is available fortune During dynamic information index, it can pre-established and encoder identical available motion information according to the available motion information index Searched in list, determine the available motion information；
Using the available motion information, motion compensation is carried out to the reference block, obtains residual block；
Statistical analysis is carried out to the residual block, obtains dividing mode corresponding to the available motion information；
The reference block is divided using dividing mode corresponding to the available motion information, regarding after being divided Frequency block.
The step can be as shown in Figure 9 division module complete.The acquisition for the reference block that the division module uses can be as Described in above-described embodiment, pre-defined video block is such as defined as reference block, or one in the reference frame of storage is regarded Frequency block is defined as reference block.
83：Decoder is predicted according to header information to the video block after division, obtains prediction block.
The step can be as shown in Figure 9 motion compensating module complete, obtain prediction block.
84：Decoder carries out coefficient scanning, re-quantization, inverse orthogonal transformation, the residual block reconstructed to one-dimensional coefficient.
The step can be as shown in Figure 9 coefficient scanning module, inverse quantization module, inverse orthogonal transformation module complete, obtain The residual block of reconstruct.
85：The block reconstructed according to the residual block of prediction block and reconstruct.
The step can be completed by Fig. 9 reconstructed module.The former video block that the block of the reconstruct namely recovers.It can manage Solution, due to the reasons such as coding loss be present, the former video block that the former video block and the encoder that recover are encoded is not It is identical.
Afterwards, after all video blocks of current decoded frame all complete 81~85 above-mentioned processing, can be reconstructed Decoding frame, then the decoding frame of reconstruct is post-processed and is put into reference to frame buffer, removed the image lesion of coding-belt, carry Rise subjective picture quality.
It should be noted that in above-described embodiment, by taking hierarchical coding as an example, it is to be understood that the embodiment of the present invention is simultaneously Hierarchical coding is not limited to, can also be applied in single layer coding.Unlike hierarchical coding, in single layer coding, reference block leads to Cross the determination of division vector.
Above-mentioned predicting unit can use single directional prediction or bi-directional predicted, for bi-directional predicted, the generation of prediction pixel Formed by two prediction signal weightings.Further, the prediction signal that predicting unit uses can be at least three, by least three Individual prediction signal weights to obtain prediction pixel.
In one embodiment of the invention, in video coding technique, redundancy is removed using the spatial coherence of pixel Information is an effective technology.This redundancy is embodied on movable information, is exactly the strong correlation between motion vector Property.Specifically, due to this redundancy, the motion vector value between adjacent different motion block is sufficiently close to.
Using the strong correlation between neighboring block motion vector, by the present invention in that with the motion vector of contiguous block to current Coding/decoding unit carries out motion compensation and the statistical property to caused motion compensated residual is analyzed.Due to current volume/solution In code unit, different moving object has different kinetic characteristics, and making motion compensation with same motion vector will produce The different residual error of statistical property.Using the statistical property of residual error effective division can be made to different moving object.
On the other hand, because motion estimation/compensation module is the general module in current Video Codec, its workability It and can realize and be widely studied and optimized, computational complexity can meet the requirement of Video Codec completely.
In the following description, we will provide a specific embodiment based on the present invention.For convenience of understanding, this implementation Example is the structure based on hierarchical coding., might not be by but just itself, this is a general coding techniques to the present invention It is confined to the structure of hierarchical coding.
In layered encoding structure, a vision signal is divided into Primary layer and enhancement layer.It is the wherein time of Primary layer, empty Between or mass resolution be less than enhancement layer.
In one embodiment of the invention, in coder side,
In layered video coding, it is assumed that Primary layer has the resolution ratio as enhancement layer.In space layered coding, Above-mentioned hypothesis seems invalid, but now it is assumed above that Primary layer be defined as by up-sampling after Primary layer, still have There is the spatial resolution as enhancement layer.
The methods of present invention is by the statistical property threshold value that adjusts motion compensated residual, can will be current in enhancement layer Coding unit is divided into multiple divisions.Without loss of generality, description below is assumed current coded unit being divided into two divisions, The situation of multiple divisions can be by that analogy.Specific division number can be made an appointment by encoding and decoding end, can also be by encoder Write code stream and decoder is obtained by parsing code stream.
The operation of encoder may be summarized to be following steps：
The first step：For the current coded unit (can be regular shape or irregular shape) in enhancement layer, it is determined that can Use movable information
To the current coded unit in enhancement layer, find corresponding Primary layer coding unit and obtain Primary layer coding unit Correlative coding information, including but not limited to movable information etc..
Determine the movable information in available motion information list.Movable information in movable information list is essentially from three Aspect：The adjacent movable information in the space of current coded unit in enhancement layer；Current coded unit is temporally adjacent in enhancement layer Movable information；The movable information of Primary layer coding unit corresponding to current coded unit in enhancement layer.Available motion packet In terms of including but being not limited to three above, such as can also add can for the room and time movable information of corresponding basic layer unit With movable information list.Movable information quantity in movable information list can be the number made an appointment or be taken from code stream , movable information therein can have identical or must be different.
Second step：For a movable information in movable information list, generation division
Motion compensation is carried out to whole current EL coding unit using this movable information, obtains corresponding residual error.
To the caused residual error progress statistical analysis from previous step, such as the intermediate value of residual error can be calculated.
By the use of caused statistical property (such as intermediate value is used as the threshold value of division) from previous step, to from step 1 Obtained Residual Generation division (is divided into division 0 less than what the pixel of intermediate value formed by corresponding residual error as defined, is otherwise division 1)。
3rd step：For the division generated in second step
Estimation can further be made：Estimation can reduce computational complexity using various fast algorithms.
Rate distortion function estimation in estimation needs the coding method of specific motion vector, can use fully profit With the encryption algorithm of movable information correlation.For example, the rope in the motion vector list to being generated in the first step can be passed through Draw the mode of coding.The bit number that specific motion vector coding method needs will be included in the calculating of rate distortion function.
The minimum motion vector of rate distortion costs value is final motion vector.Using this motion vector to present encoding Unit is taken exercises compensation, generates motion compensated residual.
4th step：For the motion compensated residual generated in the 3rd step, spatial redundancy is removed using orthogonal transformation.Due to working as The shape of preceding coding unit is not necessarily regular shape, and orthogonal transformation also not necessarily uses regular shape (current Video coding The orthogonal transformation of regular shape is used in standard).
5th step：It is residual to being rebuild after the orthogonal transform coefficient progress inverse quantization of the 4th step and reverse orthogonal transformation Difference (if the obtained all orthogonal transform coefficients of the 4th step are 0, defines reconstructive residual error now as 0), plus from the 3rd step Obtained motion-compensated values, and the coding and rebuilding of current coded unit can be obtained after clipping operation.By with before coding Original current coded unit pixel value, which is compared, can relatively obtain coding distortion.The bits of encoded of current coded unit is estimated Calculate, including：For generating the index in the motion vector list of the motion vector of division in the first step；If further transport Dynamic estimation, the index of reference motion vector, and the difference value of final motion vector and reference motion vector；Orthogonal transformation Structural information, the full zero coefficient instruction of minor structure, and the orthogonal transform coefficient quantified.Comprehensive coding distortion and bit expend, can Obtain the rate distortion costs value of this movable information division.
6th step：If all movable informations and its division have all been estimated in second step movable information list, select Select final division of the minimum division of rate distortion costs as current coded unit.Otherwise, repeat step two to five.
7th step：Final division to current coded unit, module is entropy encoded by the coded-bit of current coded unit Output bit flow is write, including：Rope in the motion vector list that motion vector for generating division generates in the first step Draw；If further making estimation, include the index of reference motion vector, and final motion vector with being sweared with reference to motion The difference value of amount；The structural information of orthogonal transformation, the full zero coefficient instruction of minor structure, and orthogonal transform coefficient quantified etc..
8th step：All coding units of current encoded frame all after the completion of, can be to locating after the progress of the coded frame of reconstruction Reason operation, the image lesion of coding-belt is removed, lifts subjective picture quality.
Algorithm above describes algorithm of the invention by the example encoded in a hierarchical coding to enhancement layer frame Flow.As it was previously stated, the present invention is a universal coding algorithm, it can both apply in the scene of hierarchical coding, can also Apply in the scene of single-layer video coding.When single-layer video encodes, it is necessary to which one is similar in layered video coding and corresponds to Up-sampling Primary layer coding unit reference encoder unit, present encoding is estimated as basis with this reference encoder unit The division of unit.Using the temporal correlation of vision signal, this reference encoder unit can be sought in reconstructed reference frame Look for.The method for being similar to motion vector encoder can be used to the coding of this reference encoder unit, that is, encode this reference encoder list The list of reference picture, location index in lists where first, and this reference encoder unit is relative to current coded unit Displacement.In order to be different from motion vector, this information can be defined as shape vector or division vector.
Algorithm above is applied to unidirectional and bi-directional predicted situation.For bi-directional predicted, the process of motion compensation is than list It is more complex to the process of prediction：The generation of prediction pixel is formed by two prediction signal weightings；And motion estimation process Also it is more complicated than single directional prediction more, therefore it is required that computational complexity it is also higher than single directional prediction.
Further genralrlization, present invention could apply to have the situation of multiple prediction signals：Under this application scenarios, prediction Pixel is to be weighted to generate by multiple prediction signals.
Remark additionally in the example that the present invention is applied to hierarchical coding, above-mentioned encryption algorithm also has generalization. In layered coding technique, there can be two class technologies.One is the coding unit of the technology based on coding unit, i.e. enhancement layer is to base This layer of information reuse is the coding unit level based on Primary layer；Another is the skill of the reference frame level based on Primary layer generation Art, that is, the Primary layer rebuild inserts the reference frame buffer of enhancement layer after processing, and the coding unit of enhancement layer is to Primary layer Information reuse is based on reference to the Primary layer reference frame in frame buffer.The example that the present invention is applied to hierarchical coding can be with general More than in two class technologies, it is only difference is in the acquisition of base layer motion vector information：In the skill based on coding unit In art, the acquisition of base layer motion vector information is more direct, and movable information has direct corresponding relation；Based on reference to frame level Technology in, base layer motion vector information needs one and hinted obliquely at, after being converted into enhancement layer with reference to the movable information in frame buffer, Just can also easily it with reference to.
Finally, the embodiment of the present invention above is described as universal description, and is not specific to a certain special implementation.Than Specific limit is not made in specific implementation such as to entropy code：It can be variable-length encoding；It can also be the coding that counts.For counting Both can be binary arithmetic coding or the general coding that counts for coding.
In one embodiment of the invention, side is being decoded,
After receiving coded bit stream, decoder parsing bit stream is simultaneously solved accordingly according to the syntactic element being resolved to Code reconstruction operation.Specific step and flow is as follows.
The first step：For the current coded unit (can be regular shape or irregular shape) in enhancement layer, it is determined that can Use movable information
To the current coded unit in enhancement layer, find corresponding Primary layer coding unit and obtain Primary layer coding unit Correlative coding information, including but not limited to movable information etc..
Determine the movable information in available motion information list.Movable information in movable information list is essentially from three Aspect：The adjacent movable information in the space of current coded unit in enhancement layer；Current coded unit is temporally adjacent in enhancement layer Movable information；The movable information of Primary layer coding unit corresponding to current coded unit in enhancement layer.Available motion packet In terms of including but being not limited to three above, such as can also add can for the room and time movable information of corresponding basic layer unit With movable information list.Movable information quantity in movable information list can be the number made an appointment or be taken from code stream , movable information therein can have identical or must be different.
Second step：The movable information index in code stream is parsed, and corresponding movable information is obtained from movable information list, Generated and divided using this movable information
Motion compensation is carried out to whole current EL coding unit using this movable information, obtains corresponding residual error.
To the caused residual error progress statistical analysis from previous step, such as the intermediate value of residual error can be calculated.
By the use of caused statistical property (such as intermediate value is used as threshold value) from previous step, to what is obtained from step 1 Residual Generation division (is divided into division 0, otherwise for division 1) by corresponding residual error as defined less than what the pixel of intermediate value formed.
3rd step：For the division generated in second step, the precise motion that code stream determines each division is further parsed Information；And obtain the motion compensation that current coded unit respectively divides using obtained movable information
4th step：Parsing code stream obtains the residual error of current coded unit, includes the structure of orthogonal transformation, and complete the zero of minor structure Coefficient instruction, orthogonal transform coefficient of quantization etc..
5th step：To obtaining reconstructive residual error after the orthogonal transform coefficient progress inverse quantization of the 4th step and reverse orthogonal transformation If (all orthogonal transform coefficients that the 4th step obtains are 0, and the reconstructive residual error defined now is 0), to be obtained plus from the 3rd step The motion-compensated values arrived, and the reconstruction of current coded unit can be obtained after clipping operation.
6th step：All coding units of current encoded frame all after the completion of, can be to locating after the progress of the coded frame of reconstruction Reason operation, the image lesion of coding-belt is removed, lifts subjective picture quality.
Above-mentioned that method flow is described, accordingly, the embodiment of the present invention is additionally provided for performing the above method Device.
Referring to Figure 10, the embodiments of the invention provide a kind of division device 10 of image block, the device can be located at coding In device, the device 10 includes acquisition module 101, motion compensating module 102, motion division module 103 and coding module 104；Obtain Modulus block 101 is used to obtain the reference block of object block and the movable information of reference block, sampled point and object block in reference block In sampled point correspond；The movable information that motion compensating module 102 is used to obtain using acquisition module 101 is to acquisition module 101 reference blocks obtained carry out motion compensation, obtain residual block；Division module 103 is moved to be used in motion compensating module 102 The value of each sampled point is adopted compared with the division threshold value that encoder determines, and according to each in obtained residual block The comparative result of sampling point carries out motion division to object block；Coding module 104, for what is obtained according to motion division module 103 The result of motion division encodes to object block.
Figure 11 is participated in, the embodiments of the invention provide a kind of division device 11 of image block, the device can be located at coding In device, the device 10 includes acquisition module 111, motion compensating module 112, motion division module 113 and coding module 114；Obtain Modulus block 111 is used to obtain the reference block of object block and the movable information of reference block, sampled point and object block in reference block In sampled point correspond；The movable information that motion compensating module 112 is used to obtain using acquisition module 111 is to acquisition module 111 reference blocks obtained carry out motion compensation, obtain residual block；Division module 113 is moved to be used in motion compensating module 112 The value of each sampled point is adopted compared with the division threshold value that decoder determines, and according to each in obtained residual block The comparative result of sampling point carries out motion division to object block；Decoder module 114, for what is obtained according to motion division module 113 The result of motion division decodes to object block.
In hardware realization, above-mentioned acquisition module, motion division module motion compensating module, coding module and decoding mould Block can be completed specifically by processor, and processor can be CPU (CPU), microprocessor, single-chip microcomputer etc..Above-mentioned Acquisition module can be specially receiver or transceiver.
Further, the embodiment of the present invention can also provide a kind of encoder, including the above-mentioned division being located in encoder Device, motion compensating module, orthogonal transform unit, quantifying unit, coefficient scanning unit and entropy code unit can also be included.Tool Hold in vivo and may refer to Fig. 7.
Further, the embodiment of the present invention can also provide a kind of decoder, including the above-mentioned division being located in decoder Device, entropy decoding unit, coefficient scanning unit, inverse quantization unit, inverse orthogonal transformation unit, predicting unit, again can also be included Structure unit etc..Particular content may refer to Fig. 9.
The present embodiment carries out motion compensation by using available motion information to reference block, and motion compensation is obtained residual Poor block carries out statistical analysis, and reference block is divided according to the result of statistical analysis, can be completed to the adaptive of reference block Division, more accurately movable information can be obtained by adaptively dividing, and then can carry out accurate motion compensated prediction, carried High compression performance.
It is apparent to those skilled in the art that for convenience and simplicity of description, only with above-mentioned each function The division progress of module, can be as needed and by above-mentioned function distribution by different function moulds for example, in practical application Block is completed, i.e., the internal structure of device is divided into different functional modules, to complete all or part of work(described above Energy.The specific work process of the system, apparatus, and unit of foregoing description, it may be referred to corresponding in preceding method embodiment Journey, it will not be repeated here.
In several embodiments provided herein, it should be understood that disclosed system, apparatus and method can be with Realize by another way.For example, device embodiment described above is only schematical, for example, the module or The division of unit, only a kind of division of logic function, can there are other dividing mode, such as multiple units when actually realizing Or component can combine or be desirably integrated into another system, or some features can be ignored, or not perform.It is another, institute Display or the mutual coupling discussed or direct-coupling or communication connection can be by some interfaces, device or unit INDIRECT COUPLING or communication connection, can be electrical, mechanical or other forms.
The unit illustrated as separating component can be or may not be physically separate, show as unit The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple On NE.Some or all of unit therein can be selected to realize the mesh of this embodiment scheme according to the actual needs 's.
In addition, each functional unit in each embodiment of the application can be integrated in a processing unit, can also That unit is individually physically present, can also two or more units it is integrated in a unit.Above-mentioned integrated list Member can both be realized in the form of hardware, can also be realized in the form of SFU software functional unit.
If the integrated unit is realized in the form of SFU software functional unit and is used as independent production marketing or use When, it can be stored in a computer read/write memory medium.Based on such understanding, the technical scheme of the application is substantially The part to be contributed in other words to prior art or all or part of the technical scheme can be in the form of software products Embody, the computer software product is stored in a storage medium, including some instructions are causing a computer It is each that equipment (can be personal computer, server, or network equipment etc.) or processor (processor) perform the application The all or part of step of embodiment methods described.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD Etc. it is various can be with the medium of store program codes.
Described above, above example is only to illustrate the technical scheme of the application, rather than its limitations；Although with reference to before Embodiment is stated the application is described in detail, it will be understood by those within the art that：It still can be to preceding State the technical scheme described in each embodiment to modify, or equivalent substitution is carried out to which part technical characteristic；And these Modification is replaced, and the essence of appropriate technical solution is departed from the spirit and scope of each embodiment technical scheme of the application.

Claims (23)

1. a kind of image block coding method, it is characterised in that methods described includes：
Encoder determines the reference block of object block and the movable information of reference block, the sampled point in the reference block and the mesh The sampled point marked in block corresponds；
The encoder carries out motion compensation using the movable information of the reference block to the reference block, obtains residual block；
The encoder carries out the division threshold value that the value of each sampled point in the residual block and the encoder determine Compare, and motion division, the division threshold value are carried out to the object block according to the comparative result of each sampled point For the intermediate value of whole residual values in the residual block；
The encoder encodes according to the result of the motion division to the object block.
2. according to the method for claim 1, it is characterised in that the encoder is by each sampled point in the residual block Value compared with the division threshold value that the encoder determines, and according to the comparative result of each sampled point to institute State object block carry out move division include：
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the second motion division of the object block.
3. according to the method for claim 1, it is characterised in that the encoder is by each sampled point in the residual block Value compared with the division threshold value that the encoder determines, and according to comparative result to the object block carry out motion draw Dividing includes：
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the second motion division of the object block.
4. according to the method described in claim 1-3 any one, it is characterised in that the encoder determines the motion of reference block Information, including：
When the encoder encodes according to non-layered coded system to the object block, at least one in following item is obtained Item is as the reference block movable information：
The movable information of the video block adjacent with the target block space；
The movable information of temporally adjacent video block with the object block.
5. according to the method described in claim 1-3 any one, it is characterised in that the reference block for determining object block, bag Include：
When the encoder encodes according to hierarchical coding mode to the object block, reconstruction image is layered according to low quality Described in the correspondence image block of object block determine the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than the target The reconstruction image of layering where block.
6. according to the method described in claim 1-3 any one, it is characterised in that the movable information for determining reference block, Including：
When the encoder encodes according to hierarchical coding mode to the object block, at least one in following item is obtained As the movable information of the reference block：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
7. according to the method described in claim any one of 1-3, it is characterised in that the reference block for determining object block, including：
When the encoder encodes according to non-layered coded system to the object block, by the reference frame of storage Three image blocks are defined as reference block, and the 3rd image block refers under the distortion metrics of setting, in the reference frame of the storage There is the image block minimum with the distortion metrics of the object block.
8. a kind of image block coding/decoding method, it is characterised in that methods described includes：
Decoder determines the reference block of object block and the movable information of reference block, the sampled point in the reference block and the mesh The sampled point marked in block corresponds；
The decoder carries out motion compensation using the movable information of the reference block to the reference block, obtains residual block；
The decoder carries out the division threshold value that the value of each sampled point in the residual block and the decoder determine Compare, and motion division, the division threshold value are carried out to the object block according to the comparative result of each sampled point For the intermediate value of whole residual values in the residual block；
The decoder decodes according to the result of the motion division to the object block.
9. according to the method for claim 8, it is characterised in that the decoder is by each sampled point in the residual block Value compared with the division threshold value that the decoder determines, and according to the comparative result of each sampled point to institute State object block carry out move division include：
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the second motion division of the object block.
10. according to the method for claim 8, it is characterised in that the decoder samples each in the residual block The value of point is compared with the division threshold value that the decoder determines, and according to the comparative result pair of each sampled point The object block, which carries out motion division, to be included：
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the second motion division of the object block.
11. according to the method described in claim 8-10 any one, it is characterised in that the reference block for determining object block, Including：
When encoder encodes according to hierarchical coding mode to the object block, the decoder is layered weight according to low quality The correspondence image block for building object block described in image determines the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than the target The reconstruction image of layering where block.
12. according to the method described in claim 8-10 any one, it is characterised in that the motion letter for determining reference block Breath, including：
When encoder encodes according to hierarchical coding mode to the object block, the decoder is obtained in following item extremely One item missing is as the reference block movable information：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
13. according to the method described in claim any one of 8-10, it is characterised in that the reference block for determining object block, bag Include：
When encoder encodes according to non-layered coded system to the object block, the decoder is by the reference frame of storage In the 3rd image block be defined as reference block, the 3rd image block refers under the distortion metrics of setting, the ginseng of the storage Examining in frame has the image block minimum with the distortion metrics of the object block.
14. the division device of a kind of image block, it is characterised in that described device is located in encoder, and described device includes：
Acquisition module, for obtaining the reference block of object block and the movable information of reference block, the sampled point in the reference block Corresponded with the sampled point in the object block；
Motion compensating module, the institute that the movable information for being obtained using the acquisition module is obtained to the acquisition module State reference block and carry out motion compensation, obtain residual block；
Division module is moved, for the division thresholding for determining the value of each sampled point in the residual block and the encoder Value is compared, and motion division, the division are carried out to the object block according to the comparative result of each sampled point Threshold value is the intermediate value of whole residual values in the residual block；
Coding module, for being encoded according to the result of the motion division to the object block.
15. device according to claim 14, it is characterised in that the acquisition module is specifically used for：
Obtain the reference block of the object block；
When the encoder encodes according to hierarchical coding mode to the object block, at least one in following item is obtained As the movable information of the reference block：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
16. device according to claim 14, it is characterised in that the acquisition module is specifically used for：
When the encoder encodes according to hierarchical coding mode to the object block, reconstruction image is layered according to low quality Described in the correspondence image block of object block determine the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than the target The reconstruction image of layering where block；
Obtain the movable information of the reference block.
17. according to the device described in claim 14-16 any one, it is characterised in that the motion division module is specifically used In：
By the value of each sampled point in the residual block compared with the division threshold value that the encoder determines；
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the second motion division of the object block.
18. according to the device described in claim 14-16 any one, it is characterised in that the motion division module is specifically used In：
By the value of each sampled point in the residual block compared with the division threshold value that the encoder determines；
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the second motion division of the object block.
19. the division device of a kind of image block, it is characterised in that described device is located in decoder, and described device includes：
Acquisition module, for obtaining the reference block of object block and the movable information of reference block, the sampled point in the reference block Corresponded with the sampled point in the object block；
Motion compensating module, the institute that the movable information for being obtained using the acquisition module is obtained to the acquisition module State reference block and carry out motion compensation, obtain residual block；
Division module is moved, for the division thresholding for determining the value of each sampled point in the residual block and the decoder Value is compared, and motion division, the division are carried out to the object block according to the comparative result of each sampled point Threshold value is the intermediate value of whole residual values in the residual block；
Decoder module, for being decoded according to the result of the motion division to the object block.
20. device as claimed in claim 19, it is characterised in that the acquisition module is specifically used for：
Obtain the reference block of the object block；
When encoder encodes according to hierarchical coding mode to the object block, obtain in following item at least one of as The movable information of the reference block：
The movable information of the image block adjacent with the target block space；
The movable information of temporally adjacent image block with the object block；
The movable information of reference block；
With the movable information of the image block adjacent with reference to block space；
The movable information of temporally adjacent image block with the reference block.
21. device according to claim 19, it is characterised in that the acquisition module is specifically used for：
When encoder encodes according to hierarchical coding mode to the object block, institute in reconstruction image is layered according to low quality The correspondence image block for stating object block determines the reference block：
Wherein, the spatial resolution of the low quality layering reconstruction image, at least one index of signal to noise ratio are less than the target The reconstruction image of layering where block；
Obtain the movable information of the reference block.
22. according to the device described in claim 19-21 any one, it is characterised in that the motion division module is specifically used In：
By the value of each sampled point in the residual block compared with the division threshold value that the decoder determines；
If the value of the sampled point in the residual block is more than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is not more than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the second motion division of the object block.
23. according to the device described in claim 19-21 any one, it is characterised in that the motion division module is specifically used In：
By the value of each sampled point in the residual block compared with the division threshold value that the decoder determines；
If the value of the sampled point in the residual block is not less than the division threshold value, it is determined that corresponding in the object block Sampled point belongs to the first motion division of the object block；
If the value of the sampled point in the residual block is less than the division threshold value, it is determined that is adopted corresponding in the object block Sampling point belongs to the second motion division of the object block.
CN201310123316.5A 2013-04-10 2013-04-10 The decoding method and device of a kind of image block Active CN104104955B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310123316.5A CN104104955B (en) 2013-04-10 2013-04-10 The decoding method and device of a kind of image block

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310123316.5A CN104104955B (en) 2013-04-10 2013-04-10 The decoding method and device of a kind of image block

Publications (2)

Publication Number Publication Date
CN104104955A CN104104955A (en) 2014-10-15
CN104104955B true CN104104955B (en) 2017-11-17

Family

ID=51672712

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310123316.5A Active CN104104955B (en) 2013-04-10 2013-04-10 The decoding method and device of a kind of image block

Country Status (1)

Country Link
CN (1) CN104104955B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110572673B (en) * 2019-09-27 2024-04-09 腾讯科技（深圳）有限公司 Video encoding and decoding method and device, storage medium and electronic device

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7098959B2 (en) * 2002-09-10 2006-08-29 Kabushiki Kaisha Toshiba Frame interpolation and apparatus using frame interpolation
CN101873500A (en) * 2009-04-24 2010-10-27 华为技术有限公司 Interframe prediction encoding method, interframe prediction decoding method and equipment
WO2010144408A1 (en) * 2009-06-11 2010-12-16 Motorola Mobility, Inc. Digital image compression by adaptive macroblock resolution coding
US20110170608A1 (en) * 2010-01-08 2011-07-14 Xun Shi Method and device for video transcoding using quad-tree based mode selection
CN102611891A (en) * 2012-02-07 2012-07-25 中国电子科技集团公司第三研究所 Method for directly performing transform coding in transform domain
CN104104966A (en) * 2013-04-08 2014-10-15 华为技术有限公司 Video block dividing method and device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7098959B2 (en) * 2002-09-10 2006-08-29 Kabushiki Kaisha Toshiba Frame interpolation and apparatus using frame interpolation
CN101873500A (en) * 2009-04-24 2010-10-27 华为技术有限公司 Interframe prediction encoding method, interframe prediction decoding method and equipment
WO2010144408A1 (en) * 2009-06-11 2010-12-16 Motorola Mobility, Inc. Digital image compression by adaptive macroblock resolution coding
US20110170608A1 (en) * 2010-01-08 2011-07-14 Xun Shi Method and device for video transcoding using quad-tree based mode selection
CN102611891A (en) * 2012-02-07 2012-07-25 中国电子科技集团公司第三研究所 Method for directly performing transform coding in transform domain
CN104104966A (en) * 2013-04-08 2014-10-15 华为技术有限公司 Video block dividing method and device

Also Published As

Publication number Publication date
CN104104955A (en) 2014-10-15

Similar Documents

Publication Publication Date Title
CN102714736B (en) Method and apparatus for encoding and decoding motion vector based on reduced motion vector predictor candidates
CN103260031B (en) Method and apparatus for encoding/decoding to motion vector
CN102792697B (en) For method and apparatus motion vector being encoded and decoding
CN104735454B (en) Method and apparatus for encoding and decoding image by using large transform unit
CN106101720B (en) Video encoding/decoding method
CN102893604B (en) Data compression for video
CN104104966B (en) The division methods and device of video block
CN106131557A (en) The method and apparatus of decoding high resolution image
CN103826125B (en) Concentration analysis method and device for compression monitor video
Vasa et al. Coddyac: Connectivity driven dynamic mesh compression
CN101742305B (en) Method for estimating motions of scalable video coding and decoding system based on Markov chain model
CN105025298A (en) A method and device of encoding/decoding an image
CN102065293B (en) Image compression method based on space domain predictive coding
CN104104955B (en) The decoding method and device of a kind of image block
Heu et al. SNR and temporal scalable coding of 3-D mesh sequences using singular value decomposition
WO2021095245A1 (en) Image processing method, data processing method, image processing device, and program
CN105025301B (en) Image decoding apparatus
JP2018093504A (en) Method and device for coding/decoding motion vector, based on candidate of reduced prediction motion vector

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20201217

Address after: Room 400, building 5, No.11 Deshengmenwai street, Xicheng District, Beijing 100032 (Desheng Park)

Patentee after: MIGU CULTURE TECHNOLOGY Co.,Ltd.

Address before: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen

Patentee before: HUAWEI TECHNOLOGIES Co.,Ltd.