CN102006480B - Method for coding and decoding binocular stereoscopic video based on inter-view prediction - Google Patents
Method for coding and decoding binocular stereoscopic video based on inter-view prediction Download PDFInfo
- Publication number
- CN102006480B CN102006480B CN 201010571683 CN201010571683A CN102006480B CN 102006480 B CN102006480 B CN 102006480B CN 201010571683 CN201010571683 CN 201010571683 CN 201010571683 A CN201010571683 A CN 201010571683A CN 102006480 B CN102006480 B CN 102006480B
- Authority
- CN
- China
- Prior art keywords
- vector
- prediction
- coding
- frame
- video
- 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
Links
Images
Abstract
The invention provides a method for coding and decoding binocular stereoscopic video based on inter-view prediction. The coding method comprises the following steps: coding a first way video by using a one-way video coding method; performing motion compensation and parallax compensation on a predictive block of each frame in a second way video, determining predictive vectors and reference frame indexes in accordance with the motion compensation and the parallax compensation, and coding predicted residuals; predicting the predictive vectors to obtain differential predictive vectors; and coding the differential predictive vectors and the reference frame indexes to finally form a second way video code stream. The decoding method is the inverse process of the coding method. The coding efficiency of the binocular stereoscopic video can be improved by using the coding method of the invention.
Description
Technical field
The present invention relates to computer, technology of video compressing encoding field, particularly a kind of coding and coding/decoding method method based on the binocular tri-dimensional video of prediction between looking.
Background technology
Along with the develop rapidly of electronics and computing technique, stereo technology has reached its maturity and has been widely applied among the different fields.Stereoscopic TV, claim again three-dimensional television, because it can provide the more visual information and the viewing effect that has more feeling of immersion of horn of plenty, just causing increasing concern, it will bring into play huge effect in the industries such as the digital home entertainment of Future in China, cultural intention, educational training.Therefore, the traditional two dimensional surface of three-dimensional television replacement shows it also is trend of the times.
Three-dimensional video-frequency is the binocular parallax principle of utilizing eye, binocular receives the left and right sides image from the specific camera point of Same Scene independently of one another, left eye is seen the image of LOOK LEFT, right eye is seen the image of LOOK RIGHT, form binocular parallax, large brain-capacity obtains the depth information of image, makes the image of enjoying that strong depth perception, realism be arranged, and spectators can enjoy very strong stereoscopic visual effect.
Binocular tri-dimensional video comprises a pair of video sequence, respectively the scene of corresponding left eye and right eye viewpoint.Can than traditional video, the data of twice need to be stored and transmit to binocular tri-dimensional video, effectively be collapsed into the key that realize the 3D technology.Because what left view and right view showed is the image of Same Scene different points of view, therefore have very strong correlation, can utilize these correlations effectively to compress.
Summary of the invention
Purpose of the present invention is intended to solve at least one of above-mentioned technological deficiency.
For this reason, one object of the present invention is to propose a kind of coding method based on the binocular tri-dimensional video of prediction between looking, and a kind of coding/decoding method based on look between the binocular tri-dimensional video of prediction corresponding with above-mentioned coding method.Wherein, the code efficiency that this coding method can the Effective Raise binocular tri-dimensional video.
For achieving the above object, one aspect of the present invention has proposed a kind of coding method based on the binocular tri-dimensional video of prediction between looking, may further comprise the steps: according to the single channel method for video coding first via video of binocular tri-dimensional video is encoded, to obtain the code stream of described first via video; Whether the present frame of judging the second road video of described binocular tri-dimensional video is bidirectional frame; If the present frame of described the second road video is not bidirectional frame, then according to upper adjacent reconstructed frame of time the prediction piece of the present frame of described the second road video is carried out motion compensation, and carry out parallax compensation according to looking the prediction piece of a reconstructed frame to the present frame of described the second road video; Determine the first predictive vector and first reference frame index of the prediction piece of described present frame according to described motion compensation and described parallax compensation; Residual error to the reference frame of the present frame of described the second road video and respective reference frame index point is encoded, to obtain the residual error code stream; Obtain the difference prediction vector according to described the first predictive vector; With according to described difference prediction vector and described the first reference frame index the present frame of described the second road video is encoded, and form the code stream of described the second road video present frame with described residual error code stream; If the present frame of described the second road video is bidirectional frame, judge then whether the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame obtains according to Direct Model; If the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame obtain according to Direct Model, then according to described the second predictive vector and the second reference frame index the present frame of described the second road video is encoded, to form the code stream of described the second road video respective frame; If with the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame be not to obtain according to Direct Model, then according to the single channel method for video coding described bidirectional frame is encoded, to obtain the code stream of present frame.
In one embodiment of the invention, describedly obtain the first predictive vector and the reference frame index of corresponding prediction piece according to motion compensation and described parallax compensation, further comprise: the motion vector that obtains the prediction piece of described present frame according to motion compensation; Obtain the difference vector of the prediction piece of described present frame according to parallax compensation; In described difference vector and described motion vector, select a vector as described the first predictive vector according to the Optimum Matching criterion, and obtain reference frame index corresponding to described predictive vector.
In one embodiment of the invention, describedly obtain the difference prediction vector according to the first predictive vector, further comprise: if described the first predictive vector is described motion vector, described motion vector is the first reference value; If described the first predictive vector is described difference vector, described difference vector is the second reference value; According to described the first reference value and described the second reference value described predictive vector is carried out predictive coding, to obtain described difference prediction vector.
In one embodiment of the invention, according to described the first reference value and described the second reference value described predictive vector is carried out predictive coding, to obtain described difference prediction vector, further comprise: the reference frame of the correspondence of the described predictive vector of judgement current block belongs to described first via video or the second road video; If described reference frame belongs to described first via video, judge whether the predictive vector of the prediction piece adjacent with described current block belongs to the second road video; If the predictive vector of the prediction piece that described current block is adjacent does not belong to the second road video, predictive vector to the adjacent prediction piece of described current block is encoded, otherwise the predictive vector of the adjacent prediction piece of described current block is not encoded, and the adjacent prediction piece of the described current block of mark is unavailable.
In one embodiment of the invention, according to described the first reference value and described the second reference value described predictive vector is carried out predictive coding, to obtain described difference prediction vector, also comprise: if described reference frame belongs to described the second road video, judge whether the predictive vector of the prediction piece adjacent with described current block belongs to first via video; If the predictive vector of the prediction piece that described current block is adjacent does not belong to first via video, predictive vector to the adjacent prediction piece of described current block is encoded, otherwise the predictive vector of the adjacent prediction piece of described current block is not encoded, and the adjacent prediction piece of the described current block of mark is unavailable.
In one embodiment of the invention, if the prediction piece of described bidirectional frame obtains the second predictive vector of the prediction piece of described bidirectional frame according to Direct Model, further comprise: if the second predictive vector of the prediction piece of the backward reference frame corresponding blocks of described bidirectional frame, belong to respectively the first via and the second road video with the second predictive vector of the prediction piece of described bidirectional frame, perhaps belong to respectively second and first via video with the second predictive vector of the prediction piece of described bidirectional frame, obtain the second predictive vector according to the space vector Forecasting Methodology; Otherwise obtain the second predictive vector according to time domain vector prediction method.
Another aspect of the present invention has proposed a kind of coding/decoding method based on the binocular tri-dimensional video of prediction between looking, may further comprise the steps: according to the single channel video encoding/decoding method first via video code flow of binocular tri-dimensional video coding is decoded, to recover described first via video; Whether the present frame coding of judging the second road video code flow of described binocular tri-dimensional video coding is the bidirectional frame coding; If the present frame of described the second road video code flow coding is not the bidirectional frame coding, then the present frame coding of described the second road video code flow is decoded to obtain difference prediction vector and first reference frame index of described present frame; Present frame coding to described the second road video code flow is decoded with the residual coding of the reference frame coding of described the first reference frame index sensing, to obtain corresponding residual error; Obtain the first predictive vector according to described difference prediction vector; Vow motion vector or the difference vector that obtains described present frame coding according to described the first prediction; According to described motion vector or difference vector, and described residual error is finished the decoding of described present frame coding; If the present frame of described the second road video code flow is encoded to the bidirectional frame coding, judge then whether the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame coding obtains according to Direct Model; If the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame coding obtain according to Direct Model, then according to described the second predictive vector and the second reference frame index the present frame of described the second road video code flow is encoded into decoding, to obtain the corresponding bidirectional frame of described the second road video; If the second predictive vector of the prediction piece of described bidirectional frame and the second reference frame index are not to obtain according to Direct Model, then according to the single channel video encoding/decoding method described bidirectional frame coding is decoded, to recover bidirectional frame.
In one embodiment of the invention, describedly obtain the first predictive vector according to the difference prediction vector, further comprise: if the prediction piece of present frame uses motion compensation to predict, then the first predictive vector of described prediction piece is described difference prediction vector, and the motion vector sum of the prediction piece of the use motion compensation adjacent with described prediction piece; If the prediction piece of present frame uses parallax compensation to predict that then the first predictive vector of described prediction piece is described difference prediction vector, and the difference vector sum of the prediction piece of the use parallax compensation adjacent with described prediction piece.
In one embodiment of the invention, described the second predictive vector that obtains the prediction piece of described bidirectional frame coding according to Direct Model, further comprise: if the reference block of prediction piece corresponding to the backward reference frame of described bidirectional frame coding points to the reference frame of described first via video code flow, obtain the second predictive vector of described Direct Model piece by the method for space vector prediction, if the reference block of the prediction piece that the backward reference frame of described bidirectional frame coding is corresponding points to the reference frame of described the second road video code flow, obtain the second predictive vector of described Direct Model piece by the method for time domain vector prediction.
Code And Decode method based on the binocular tri-dimensional video of prediction between looking of the present invention, take full advantage of the correlation of time domain and the correlation between viewpoint, by the prediction piece is carried out motion compensation or parallax compensation, can effectively improve the code efficiency of binocular tri-dimensional video.In addition, the present invention asks for algorithm to the second predictive vector of the piece of employing Direct Model coding in the bidirectional frame and improves, and has more improved the accuracy that the second predictive vector of Direct Model piece is asked in the bidirectional frame.Simultaneously, the present invention improves the predictive coding algorithm of predictive vector, has improved the accuracy of prediction of predictive vector and the efficient of predictive vector differential coding.
The aspect that the present invention adds and advantage in the following description part provide, and part will become obviously from the following description, or recognize by practice of the present invention.
Description of drawings
Above-mentioned and/or the additional aspect of the present invention and advantage are from obviously and easily understanding becoming the description of embodiment below in conjunction with accompanying drawing, wherein:
Fig. 1 is the flow chart based on the method for coding binocular stereo video of prediction between looking of the embodiment of the invention;
Fig. 2 carries out the schematic diagram that predictive coding obtains the difference prediction vector to predictive vector in the coding method of the present invention;
The direct mode motion vector derivation schematic diagram of B frame in Fig. 3 a-3b coding method of the present invention and the coding/decoding method;
Fig. 4 is IBBPBBP... predict schematic diagram among the embodiment of coding method of the present invention; With
Fig. 5 is the flow chart based on the binocular tri-dimensional video coding/decoding method of prediction between looking of the embodiment of the invention.
Embodiment
The below describes whole embodiment of the present invention in detail, and the example of described embodiment is shown in the drawings, and wherein identical or similar label represents identical or similar element or the element with identical or similar functions from start to finish.Be exemplary below by the embodiment that is described with reference to the drawings, only be used for explaining the present invention, and can not be interpreted as limitation of the present invention.
The embodiment of the invention is a kind of coding method based on the binocular tri-dimensional video of prediction between looking that propose for having now not high to the binocular tri-dimensional video code efficiency.The coding method based on the binocular tri-dimensional video of prediction between looking by the embodiment of the invention proposes can improve the code efficiency to binocular tri-dimensional video effectively.Below with reference to accompanying drawing detailed introduction is done by coding method and corresponding decoding side that the embodiment of the invention proposes.
As shown in Figure 1, show the flow chart based on the method for coding binocular stereo video of prediction between looking of the embodiment of the invention.The coding method that proposes for binocular tri-dimensional video owing to the embodiment of the invention, so the wherein coding method of one tunnel employing single channel video to binocular tri-dimensional video is encoded, in specific embodiments of the invention, wherein one road video that the coding method of adopting the single channel video is encoded is defined as first via video, and another road video is defined as the second road video, certainly, so-called first via video and the second road video can exchange.And, according to the correlation of binocular tri-dimensional video two-path video, the second road video of the embodiment of the invention is encoded, thereby reduced coding to redundant video, improve code efficiency.Specifically may further comprise the steps:
Step S101 judges whether current prediction piece belongs to the second road video, if do not belong to the second road video, and execution in step S102 then, otherwise execution in step S103;
Step S102 utilizes the method for single channel Video coding first via video is predicted and to be encoded;
Step S103 judges whether bidirectional frame (B frame) prediction of present frame of the second road video, if present frame is bidirectional frame, and execution in step S109 then, otherwise execution in step S104;
Step S104, because the reference frame of the prediction piece of current bidirectional frame may be from the reconstructed frame of current view point or the reconstructed frame of first via viewpoint, so simultaneously this piece is carried out motion compensation and parallax compensation, and according to the Optimum Matching criterion, (such as Averaged Square Error of Multivariate, the average absolute value error) from the motion vector of current prediction piece and difference vector, selects according to the Optimum Matching criterion, select the vector of one of them as the predictive vector of current prediction piece, in specific embodiments of the invention, this predictive vector is defined as the first predictive vector, with distinguish to adopt that distinct methods gets predictive vector, below for the description of this predictive vector, all describe as an example of the first predictive vector example.But those skilled in the art will appreciate that the definition for the first predictive vector, also can be defined as other titles, but the implication of its expression is identical with predictive vector;
Step S105 judges whether the first predictive vector of current prediction piece points to the reference frame of the second road video, if point to the reference frame of the second road video, and execution in step S106 then, otherwise execution in step S107;
Step S106 points to the adjacent block of first via video with unavailable in the prediction of the first predictive vector of the prediction piece of present frame.
In specific embodiments of the invention, as shown in Figure 2, show in the coding method of the embodiment of the invention predictive vector is carried out predictive coding, thereby obtain the schematic diagram of difference prediction vector, below in conjunction with specific embodiment step S106 is further explained, if current block E points to the second road video, and adjacent block A points to first via video, so current block E and the first predictive vector of adjacent block A belong to different videos, therefore can't utilize the first predictive vector of adjacent block A to carry out predictive coding, so tag block A is unavailable, ignore the coding to the first predictive vector of adjacent block A;
Step S107 points to the adjacent block of the second road video with unavailable in the prediction of the first predictive vector of the prediction piece of present frame.
In specific embodiments of the invention, equally with reference to figure 2, if current block E points to first via viewpoint, and adjacent block A points to the second tunnel viewpoint, so current block E and the first predictive vector of adjacent block A belong to different videos, therefore can't utilize the first predictive vector of adjacent block A to carry out predictive coding, so tag block A is unavailable, ignore equally the coding to the predictive vector of adjacent block A;
Step S108 to there not being disabled the first predictive vector of mark prediction piece to carry out predictive coding, obtains the difference prediction vector according to step S106 or S107.
Particularly, if described the first predictive vector is described motion vector, described motion vector is the first reference value; If described the first predictive vector is described difference vector, described difference vector is the second reference value; According to described the first reference value and described the second reference value described the first predictive vector is carried out predictive coding, to obtain described difference prediction vector, if namely current prediction piece uses motion compensation to predict, the motion vector of motion compensation block that then will be adjacent with current prediction piece is as predicted value, if current prediction piece uses parallax compensation to predict that the difference vector of parallax compensation piece that then will be adjacent with current prediction piece is as predicted value.Thereby, according to predicted value the first predictive vector of predicting piece is carried out predictive coding, obtain the difference prediction vector.According to difference prediction vector and the first reference frame index, each frame of the second road video is encoded at last, finally obtain the second road video code flow;
Step S109, because described present frame is bidirectional frame, whether adopt Direct Model to ask for the second predictive vector so judge current block, from Fig. 1, can find out, in this specific embodiment, the motion vector of this second predictive vector for adopting Direct Model to ask for is if adopt Direct Model to ask for the second predictive vector, execution in step S110 then, otherwise execution in step S113;
Step S110, if the second predictive vector of the prediction piece of the backward reference frame corresponding blocks of described bidirectional frame, belong to respectively the first via and the second road video with the second predictive vector of the prediction piece of described bidirectional frame, perhaps belong to respectively second and first via video with the second predictive vector of the prediction piece of described bidirectional frame, obtain the second predictive vector according to the space vector Forecasting Methodology; Otherwise obtain the second predictive vector according to time domain vector prediction method, judge namely whether the second predictive vector of the backward corresponding blocks of current prediction piece points to the reference frame of first via video, if the second predictive vector of the backward corresponding blocks of current prediction piece points to first via video reference frame, execution in step S111 then, otherwise execution in step S112;
Step S111 adopts the spatial domain prediction pattern that the second predictive vector is predicted.
In specific embodiments of the invention, adopt the prediction schematic diagram of spatial domain prediction pattern shown in Fig. 3 a, be the Direct Model second predictive vector derivation principle figure of bidirectional frame in coding method of the present invention and the coding/decoding method;
Step S112 adopts the time domain prediction pattern that the second predictive vector is predicted.
In specific embodiments of the invention, adopt the prediction schematic diagram of time domain prediction pattern shown in Fig. 3 b, be the Direct Model second predictive vector derivation principle figure of bidirectional frame in coding method of the present invention and the coding/decoding method;
Step S113 utilizes the method for single channel Video coding remaining bidirectional frame is predicted and to be encoded.
In specific embodiments of the invention, coding side carries out motion compensation and parallax compensation to the second road video in the binocular tri-dimensional video, prediction between coding side is looked macroblock to be encoded as the first predictive vector according to one of Optimum Matching criterion selection difference vector and motion vector.Coding side transfers to decoding end with the first predictive vector first and reference frame index, so that decoding end can be carried out correct decoding to the second road code stream.In a preferred embodiment of the invention, as shown in Figure 4, be IBBPBBP... predict schematic diagram.The instantiation that adopts the binocular video coding that coding method of the present invention obtains is shown.But those of ordinary skill in the art instructs, and the IBBPBBP... coding structure is just in order to illustrate the effect of the embodiment of the invention, and wherein, coding structure provided by the invention is not limited to this.These also should belong to protection scope of the present invention.
In the embodiment of the invention on the other hand, a kind of coding/decoding method based on the binocular tri-dimensional video of prediction between looking has been proposed.Be described in detail below in conjunction with the coding/decoding method of accompanying drawing to the embodiment of the invention.
As shown in Figure 5, show the particular flow sheet based on the coding/decoding method of the binocular tri-dimensional video of prediction between looking of the embodiment of the invention.
Particularly, the decode procedure of binocular tri-dimensional video be may further comprise the steps:
Step S201 judges whether the prediction piece of present frame belongs to the second road video.If belong to the second road video, execution in step S203 then, otherwise execution in step S202;
Step S202, the present frame coding is predicted and decoded according to the method for single channel video decode, obtain the reconstructed frame of present frame coding;
Step S203, judge whether bidirectional frame (B frame) prediction of present frame, if present frame is the bidirectional frame prediction, execution in step S209 then, otherwise execution in step S204;
Step S204, from the second road video code flow, obtain the difference prediction vector of current block, and judge that according to the respective reference frame index of difference prediction vector this reference frame points to the reference frame in the former moment of the second road video, or the parallax reference frame of first via video current time;
Step S205, judge whether the difference prediction vector of current block points to the reference frame of the second road video.If the reference frame of difference prediction vectors directed the second road video of current block, execution in step S206 then, otherwise execution in step S207;
Step S206, in the recovery of predictive vector, point to the adjacent block of first via video with unavailable.In specific embodiments of the invention, this predictive vector is corresponding to the first predictive vector of the coding method of the present invention's proposition.
In a specific embodiment of the present invention, as shown in Figure 2, show in the coding method of the present invention predictive vector is carried out the schematic diagram that predictive coding obtains the difference prediction vector.By specific embodiment step S206 is done detailed explanation, if current block E points to the second road video, and adjacent block A points to first via video, because the first predictive vector of current block E and adjacent block A belongs to different videos, therefore can't utilize the first predictive vector of adjacent block A to carry out predictive coding, so tag block A is unavailable;
Step S207, in the recovery of predictive vector, point to the adjacent block of the second tunnel viewpoint with unavailable.
In a specific embodiment of the present invention, as shown in Figure 2, show in the coding method of the present invention predictive vector is carried out the schematic diagram that predictive coding obtains the difference prediction vector.If by specific embodiment step S207 current block E is pointed to first via viewpoint, and adjacent block A points to the second tunnel viewpoint, because their the first predictive vector belongs to different viewpoints, therefore can't utilize the first predictive vector of adjacent block A to carry out predictive coding, so tag block A is unavailable;
The difference prediction vector that obtains in the predicted value of step S208, the first predictive vector of obtaining according to step S206 or S207 and the second road video code flow recovers original difference vector or motion vector.
Particularly, if current prediction piece uses motion compensation to predict, the first predictive vector of then predicting piece is the vector predictor sum that the difference prediction vector reaches the piece of the use motion compensation adjacent with current prediction piece, if current prediction piece uses parallax compensation to predict, the first predictive vector of then predicting piece is the vector predictor sum that the difference prediction vector reaches the piece of the use parallax compensation adjacent with current prediction piece, then current prediction piece is carried out parallax compensation or motion compensation.And according to the compensation result to the prediction piece, the second road video code flow is reconstructed, finish the decoding to the second road video code flow;
Step S209, judge whether current block adopts Direct Model to ask for the second predictive vector, ask for the second predictive vector if adopt Direct Model, execution in step S210 then, otherwise execution in step S210, in specific embodiments of the invention, can find out that from Fig. 5 this second predictive vector is the motion vector during the employing Direct Model is asked for, at this, motion vector is defined as the second predictive vector;
Step S210, judge whether the second predictive vector of the backward corresponding blocks of current block points to the reference frame of first via video, if it is execution in step S211, otherwise execution in step S212;
Step S211, employing spatial domain prediction pattern are predicted the second predictive vector.
In specific embodiments of the invention, adopt the prediction schematic diagram of spatial domain prediction pattern shown in Fig. 3 a, be the Direct Model second predictive vector derivation principle figure of bidirectional frame in coding method of the present invention and the coding/decoding method;
Step S212, employing time domain prediction pattern are predicted the second predictive vector.
In specific embodiments of the invention, adopt the prediction schematic diagram of time domain prediction pattern shown in Fig. 3 b, be the Direct Model second predictive vector derivation principle figure of bidirectional frame in coding method of the present invention and the coding/decoding method;
Step S213, utilize the single channel video encoding/decoding method to carry out bi-directional predicted and decoding to residue bidirectional frame coding, obtain remaining the reconstructed frame of bidirectional frame coding, thereby finish whole decode procedure.
Need to prove, in an embodiment of the present invention, at first, the coding of binocular tri-dimensional video not only can be according to the IBBPBBP... coding structure of instantiation of the present invention, also can be according to structured codings such as IPPP..., secondly, parallax compensation of the present invention adopts the image of another video of synchronization to carry out parallax compensation as the reference image, but be not limited to this, the reference frame of predicting between video also can be a plurality of, but those skilled in the art instructs, and these all can be changed according to the thought of put forward the methods of the present invention.Although but these carry out multiple different the modification to the present invention, equally all should belong to protection scope of the present invention.
The coding method based on the binocular tri-dimensional video of prediction between looking by embodiment of the invention proposition, take full advantage of the correlation of time domain and the correlation between viewpoint, by the prediction piece is carried out motion compensation or parallax compensation, can effectively improve the code efficiency of binocular tri-dimensional video.In addition, the present invention asks for algorithm to the second predictive vector of the piece of employing Direct Model coding in the bidirectional frame and improves, and has more improved the accuracy that the second predictive vector of Direct Model piece is asked in the bidirectional frame.Simultaneously, the present invention improves the predictive coding algorithm of predictive vector, has improved the accuracy of prediction of predictive vector and the efficient of predictive vector differential coding.And a kind of coding/decoding method based on the binocular tri-dimensional video of prediction between looking that the present invention proposes, can think the inverse process based on the coding method of the binocular tri-dimensional video of prediction between looking that the present invention proposes, so this coding/decoding method has improved decoding efficiency equally to the decoding of video.
Although illustrated and described embodiments of the invention, for the ordinary skill in the art, be appreciated that without departing from the principles and spirit of the present invention and can carry out multiple variation, modification, replacement and modification to these embodiment that scope of the present invention is by claims and be equal to and limit.
Claims (9)
1. the coding method based on the binocular tri-dimensional video of prediction between looking is characterized in that, may further comprise the steps:
According to the single channel method for video coding first via video of binocular tri-dimensional video is encoded, to obtain the code stream of described first via video;
Whether the present frame of judging the second road video of described binocular tri-dimensional video is bidirectional frame;
If the present frame of described the second road video is not bidirectional frame, then according to upper adjacent reconstructed frame of time the prediction piece of the present frame of described the second road video is carried out motion compensation, and carry out parallax compensation according to looking the prediction piece of a reconstructed frame to the present frame of described the second road video;
Determine the first predictive vector and first reference frame index of the prediction piece of described present frame according to described motion compensation and described parallax compensation;
Residual error to the reference frame of the present frame of described the second road video and respective reference frame index point is encoded, to obtain the residual error code stream;
Obtain the difference prediction vector according to described the first predictive vector; With
According to described difference prediction vector and described the first reference frame index the present frame of described the second road video is encoded, and form the code stream of described the second road video present frame with described residual error code stream;
If the present frame of described the second road video is bidirectional frame, judge then whether the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame obtains according to Direct Model;
If the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame obtain according to Direct Model, then according to described the second predictive vector and the second reference frame index the present frame of described the second road video is encoded, to form the code stream of described the second road video respective frame; With
If the second predictive vector of the prediction piece of described bidirectional frame and the second reference frame index are not to obtain according to Direct Model, then according to the single channel method for video coding described bidirectional frame is encoded, to obtain the code stream of present frame.
2. the coding method based on the binocular tri-dimensional video of prediction between looking as claimed in claim 1, it is characterized in that, describedly determine further to comprise the first predictive vector and first reference frame index of the prediction piece of described present frame according to motion compensation and described parallax compensation:
Obtain the motion vector of the prediction piece of described present frame according to motion compensation;
Obtain the difference vector of the prediction piece of described present frame according to parallax compensation;
In described difference vector and described motion vector, select a vector as described the first predictive vector according to the Optimum Matching criterion, and obtain described the first reference frame index.
3. the coding method of binocular tri-dimensional video based on prediction between looking as claimed in claim 2 is characterized in that, describedly obtains the difference prediction vector according to the first predictive vector, further comprises:
If described the first predictive vector is described motion vector, described motion vector is the first reference value;
If described the first predictive vector is described difference vector, described difference vector is the second reference value;
According to described the first reference value and described the second reference value described the first predictive vector is carried out predictive coding, to obtain described difference prediction vector.
4. the coding method based on the binocular tri-dimensional video of prediction between looking as claimed in claim 3, it is characterized in that, according to described the first reference value and described the second reference value described the first predictive vector is carried out predictive coding, to obtain described difference prediction vector, further comprises:
The reference frame of the correspondence of described first predictive vector of judgement current block belongs to described first via video or the second road video;
If described reference frame belongs to described first via video, judge whether the first predictive vector of the prediction piece adjacent with described current block belongs to the second road video;
If described first predictive vector of the prediction piece that described current block is adjacent does not belong to the second road video, the first predictive vector to the adjacent prediction piece of described current block is encoded, otherwise the first predictive vector of the adjacent prediction piece of described current block is not encoded, and the adjacent prediction piece of the described current block of mark is unavailable.
5. the coding method based on the binocular tri-dimensional video of prediction between looking as claimed in claim 4, it is characterized in that, according to described the first reference value and described the second reference value described the first predictive vector is carried out predictive coding, to obtain described difference prediction vector, also comprises:
If described reference frame belongs to described the second road video, judge whether the predictive vector of the prediction piece adjacent with described current block belongs to first via video;
If the predictive vector of the prediction piece that described current block is adjacent does not belong to first via video, predictive vector to the adjacent prediction piece of described current block is encoded, otherwise the predictive vector of the adjacent prediction piece of described current block is not encoded, and the adjacent prediction piece of the described current block of mark is unavailable.
6. the coding method of binocular tri-dimensional video based on prediction between looking as claimed in claim 1 is characterized in that, if the prediction piece of described bidirectional frame obtains the second predictive vector of the prediction piece of described bidirectional frame according to Direct Model, further comprises:
If the second predictive vector of the prediction piece of the backward reference frame corresponding blocks of described bidirectional frame, belong to respectively the first via and the second road video with the second predictive vector of the prediction piece of described bidirectional frame, perhaps belong to respectively second and first via video with the second predictive vector of the prediction piece of described bidirectional frame, obtain the second predictive vector according to the space vector Forecasting Methodology; Otherwise obtain the second predictive vector according to time domain vector prediction method.
7. the coding/decoding method based on the binocular tri-dimensional video of prediction between looking is characterized in that, may further comprise the steps:
According to the single channel video encoding/decoding method first via video code flow of binocular tri-dimensional video coding is decoded, to recover described first via video;
Whether the present frame coding of judging the second road video code flow of described binocular tri-dimensional video coding is the bidirectional frame coding;
If the present frame of described the second road video code flow coding is not the bidirectional frame coding, then the present frame coding of described the second road video code flow is decoded to obtain difference prediction vector and first reference frame index of described present frame;
Present frame coding to described the second road video code flow is decoded with the residual coding of the reference frame coding of described the first reference frame index sensing, to obtain corresponding residual error;
Obtain the first predictive vector according to described difference prediction vector;
Obtain motion vector or the difference vector of described present frame coding according to described the first predictive vector;
If the present frame of described the second road video code flow is encoded to the bidirectional frame coding, judge then whether the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame coding obtains according to Direct Model;
If the second predictive vector and second reference frame index of the prediction piece of described bidirectional frame coding obtain according to Direct Model, then according to described the second predictive vector and the second reference frame index the present frame of described the second road video code flow is encoded into decoding, to obtain the corresponding bidirectional frame of described the second road video;
If the second predictive vector of the prediction piece of described bidirectional frame and the second reference frame index are not to obtain according to Direct Model, then according to the single channel video encoding/decoding method described bidirectional frame coding is decoded, to recover bidirectional frame.
8. the coding/decoding method of binocular tri-dimensional video based on prediction between looking as claimed in claim 7 is characterized in that, describedly obtains the first predictive vector according to the difference prediction vector, further comprises:
If the prediction piece of present frame uses motion compensation to predict that then the first predictive vector of described prediction piece is described difference prediction vector, and the motion vector sum of the prediction piece of the use motion compensation adjacent with described prediction piece;
If the prediction piece of present frame uses parallax compensation to predict that then the first predictive vector of described prediction piece is described difference prediction vector, and the difference vector sum of the prediction piece of the use parallax compensation adjacent with described prediction piece.
9. the coding/decoding method of binocular tri-dimensional video based on prediction between looking as claimed in claim 8 is characterized in that, obtains the second predictive vector of the prediction piece of described bidirectional frame coding according to Direct Model, further comprises:
If the reference block of the prediction piece that the backward reference frame of described bidirectional frame coding is corresponding points to the reference frame of described first via video code flow, obtain the second predictive vector of described Direct Model piece by the method for space vector prediction, if the reference block of the prediction piece that the backward reference frame of described bidirectional frame coding is corresponding points to the reference frame of described the second road video code flow, obtain the second predictive vector of described Direct Model piece by the method for time domain vector prediction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010571683 CN102006480B (en) | 2010-11-29 | 2010-11-29 | Method for coding and decoding binocular stereoscopic video based on inter-view prediction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010571683 CN102006480B (en) | 2010-11-29 | 2010-11-29 | Method for coding and decoding binocular stereoscopic video based on inter-view prediction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102006480A CN102006480A (en) | 2011-04-06 |
| CN102006480B true CN102006480B (en) | 2013-01-30 |
Family
ID=43813496
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010571683 Active CN102006480B (en) | 2010-11-29 | 2010-11-29 | Method for coding and decoding binocular stereoscopic video based on inter-view prediction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102006480B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120262545A1 (en) * | 2011-04-18 | 2012-10-18 | Paul Kerbiriou | Method for coding and decoding a 3d video signal and corresponding devices |
| KR20140057683A (en) * | 2011-08-29 | 2014-05-13 | 주식회사 아이벡스피티홀딩스 | Method of encoding moving pictures in merge mode |
| CN102413342A (en) * | 2011-09-05 | 2012-04-11 | 清华大学 | Geometrical correction method and system of binocular stereoscopic video |
| JP5514372B2 (en) * | 2011-10-05 | 2014-06-04 | パナソニック株式会社 | Encoding method and encoding apparatus |
| CN102438141B (en) * | 2011-10-25 | 2013-03-27 | 中国科学技术大学 | Processing method of code stream of stereo video and apparatus thereof |
| EP3247116A1 (en) * | 2012-07-27 | 2017-11-22 | HFI Innovation Inc. | Method of constrain disparity vector derivation in 3d video coding |
| US10136143B2 (en) | 2012-12-07 | 2018-11-20 | Qualcomm Incorporated | Advanced residual prediction in scalable and multi-view video coding |
| CN104902256B (en) * | 2015-05-21 | 2018-01-09 | 南京大学 | A kind of binocular stereo image decoding method based on motion compensation |
| US10560678B2 (en) | 2016-11-09 | 2020-02-11 | Mediatek Inc. | Method and apparatus having video encoding function with syntax element signaling of rotation information of content-oriented rotation applied to 360-degree image content or 360-degree video content represented in projection format and associated method and apparatus having video decoding function |
| US20180262774A1 (en) * | 2017-03-09 | 2018-09-13 | Mediatek Inc. | Video processing apparatus using one or both of reference frame re-rotation and content-oriented rotation selection and associated video processing method |
| CN107071385B (en) * | 2017-04-18 | 2019-01-25 | 杭州派尼澳电子科技有限公司 | A kind of method for encoding stereo video introducing parallax compensation based on H265 |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007024072A1 (en) * | 2005-08-22 | 2007-03-01 | Samsung Electronics Co., Ltd. | Method and apparatus for encoding multiview video |
| CN101222640A (en) * | 2007-01-09 | 2008-07-16 | 华为技术有限公司 | Method an device for confirming reference frame |
| CN101547350A (en) * | 2008-03-28 | 2009-09-30 | 陈意辉 | Efficient plane and stereo digital video coding and decoding method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1225127C (en) * | 2003-09-12 | 2005-10-26 | 中国科学院计算技术研究所 | A coding/decoding end bothway prediction method for video coding |
| US8982951B2 (en) * | 2008-03-10 | 2015-03-17 | Mediatek Inc. | Adaptive motion estimation coding |
-
2010
- 2010-11-29 CN CN 201010571683 patent/CN102006480B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007024072A1 (en) * | 2005-08-22 | 2007-03-01 | Samsung Electronics Co., Ltd. | Method and apparatus for encoding multiview video |
| CN101222640A (en) * | 2007-01-09 | 2008-07-16 | 华为技术有限公司 | Method an device for confirming reference frame |
| CN101547350A (en) * | 2008-03-28 | 2009-09-30 | 陈意辉 | Efficient plane and stereo digital video coding and decoding method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102006480A (en) | 2011-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102006480B (en) | Method for coding and decoding binocular stereoscopic video based on inter-view prediction | |
| CN104412597B (en) | The method and device that unified difference vector for 3D Video codings is derived | |
| EP2721823B1 (en) | Method and apparatus of texture image compression in 3d video coding | |
| CN106471807B (en) | The coding method of three-dimensional or multi-angle video including visual angle synthesis prediction | |
| US20140002599A1 (en) | Competition-based multiview video encoding/decoding device and method thereof | |
| KR101747434B1 (en) | Apparatus and method for encoding and decoding motion information and disparity information | |
| EP2898688B1 (en) | Method and apparatus for deriving virtual depth values in 3d video coding | |
| CN104995916B (en) | Video data decoding method and video data decoding device | |
| KR100738867B1 (en) | Method for Coding and Inter-view Balanced Disparity Estimation in Multiview Animation Coding/Decoding System | |
| CN102801995B (en) | A kind of multi-view video motion based on template matching and disparity vector prediction method | |
| KR101893559B1 (en) | Apparatus and method for encoding and decoding multi-view video | |
| CN101222627A (en) | Multi-viewpoint video coding and decoding system, method and device for estimating vector | |
| CN103051894B (en) | A kind of based on fractal and H.264 binocular tri-dimensional video compression & decompression method | |
| KR101598855B1 (en) | Apparatus and Method for 3D video coding | |
| CN104798375A (en) | Method and apparatus of constrained disparity vector derivation in 3d video coding | |
| CN104935940B (en) | The signal transfer method of block segmentation based on depth | |
| KR20090122633A (en) | Method and its apparatus for fast mode decision in multi-view video coding | |
| CN106210741B (en) | A kind of deep video encryption algorithm based on correlation between viewpoint | |
| WO2006110007A1 (en) | Method for coding in multiview video coding/decoding system | |
| CN103220532A (en) | Joint prediction encoding method and joint predication encoding system for stereoscopic video | |
| KR20070076356A (en) | Method and apparatus for coding and decoding of video sequence | |
| CN103765895A (en) | Encoding/decoding method and apparatus using a skip mode | |
| CN104782126A (en) | Apparatus and method for coding/decoding multi-view image | |
| KR101078525B1 (en) | Method for coding of multi-view video | |
| CN102263952A (en) | Quick fractal compression and decompression method for binocular stereo video based on object |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |