CN103763557A - Do-NBDV acquiring method and video decoding device - Google Patents
Do-NBDV acquiring method and video decoding device Download PDFInfo
- Publication number
- CN103763557A CN103763557A CN201410003938.9A CN201410003938A CN103763557A CN 103763557 A CN103763557 A CN 103763557A CN 201410003938 A CN201410003938 A CN 201410003938A CN 103763557 A CN103763557 A CN 103763557A
- Authority
- CN
- China
- Prior art keywords
- value
- nbdv
- view
- sequence number
- pixel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention provides a Do-NBDV acquiring method and a video decoding device, and relates to the field of video processing. The computation complexity of a Do-NBDV can be lowered, and the coding efficiency can be improved. The method comprises the steps of determining a reference view, acquiring an NBDV value and the serial number value of the reference view, acquiring the depth map information corresponding to the reference view according to the serial number value, acquiring the CU address and the PU division mode of the reference view according to the serial number value, acquiring the position information of a first PU block through calculation according to the NBDV value, the CU address and the PU division mode, determining M pixels in the first PU block, acquiring the depth values of the M pixels according to the depth map information, using the mapping value of the pixel with the largest depth value in the M pixels as the horizontal component of a first prediction motion vector, and calculating and obtaining the horizontal component of the first Do-NBDV according to the horizontal component of the first prediction motion vector, wherein the M is equal to two or three.
Description
Technical field
The present invention relates to field of video processing, relate in particular to a kind of Do-NBDV (Depth-oriented Neighboring Block Disparity Vector, the degree of depth is based on adjacent block difference vector) acquisition methods and video decoder.
Background technology
Because the image in synchronization different points of view may have similar picture material, prediction between carry out in Video coding, also needing to look, further to eliminate the information redundancy between different points of view image.Difference vector has represented the vector correlation of corresponding similar macro blocks between synchronization different points of view image, depending on the accuracy of a vector, directly affects coding and decoding video efficiency.
Difference vector can obtain by Do-NBDV, wherein traditional Do-NBDV computational process redundancy.Existing Do-NBDV calculation process also can affect computational accuracy in addition, and then affects code efficiency.
Summary of the invention
Embodiments of the invention provide a kind of Do-NBDV acquisition methods and video decoder, can reduce the computation complexity of Do-NBDV, improve code efficiency.
For achieving the above object, embodiments of the invention adopt following technical scheme:
First aspect, discloses a kind of Do-NBDV acquisition methods, and described method comprises:
Determine reference-view;
Obtain the sequence number value based on adjacent block difference vector NBDV value and described reference-view, according to the sequence number value of described reference-view, obtain depth map information corresponding to described reference-view;
According to the sequence number value of described reference-view, obtain codec unit CU address and the predicting unit PU partitioning scheme of described reference-view, described CU address is coding CU address or decoding CU address;
According to described NBDV value, described CU address and described PU partitioning scheme, calculate the positional information that obtains a PU piece;
In a described PU piece, determine M pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel;
Horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate and obtain the horizontal component of first degree of depth based on adjacent block difference vector Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV; Described M equals 2 or 3.
In conjunction with first aspect, in the possible implementation of the first, before described definite reference-view, described method also comprises:
The basic view of video is encoded and carried out difference vector and calculate;
Or, the basic view of described video is decoded and is carried out difference vector and calculate.
In conjunction with first aspect, in the possible implementation of the second, after calculating the horizontal component that obtains a Do-NBDV, described method also comprises:
According to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece;
In described the 2nd PU piece, determine N pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described N pixel;
Horizontal component using the mapping value of the pixel of depth value maximum in a described N pixel as the second motion vectors, according to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV, the vertical component that the vertical component of described the 2nd Do-NBDV is described NBDV; Described N equals 2 or 3.
In conjunction with first aspect, in the third possible implementation, described in a described PU piece determine M pixel, horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV, comprising:
In a described PU piece, determine lower-left and 2 of bottom rights pixel, and obtain the depth value of 2 pixels in described lower-left and bottom right;
Horizontal component using the mapping value of pixel larger depth value as the first motion vectors;
According to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV.
Or, in a described PU piece, determine upper right, lower-left and 3 of bottom rights pixel, and obtain the depth value of described upper right, 3 pixels in lower-left and bottom right;
Horizontal component using the mapping value of the pixel of depth value maximum as the first motion vectors;
According to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV.
In conjunction with first aspect, in the 4th kind of possible implementation, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Obtain mobile vector MV value and the sequence number value of the time domain adjacent block of the coding unit CU piece of described reference-view;
Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described time domain adjacent block as described reference-view.
In conjunction with first aspect, in the 5th kind of possible implementation, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Obtain MV value and the sequence number value of the spatial domain adjacent block of the CU piece of described reference-view;
Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described spatial domain adjacent block as described reference-view.
In conjunction with first aspect, in the 6th kind of possible implementation, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Detect described reference-view and whether have motion prediction compensation block;
If exist, obtain MV value and the sequence number value of described motion prediction compensation block;
Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of described motion prediction compensation block as described reference-view.
In conjunction with first aspect, in the 7th kind of possible implementation, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Described NBDV is preset as to 0 vector;
Obtain the sequence number value of looking a reference frame of the CU piece of described reference-view, and using the described sequence number value of a reference frame of looking as the sequence number value of described reference-view.
Second aspect, a kind of video processor, comprising:
Determining unit, for determining reference-view;
Acquiring unit, for obtaining the sequence number value based on adjacent block difference vector NBDV value and described reference-view, and obtains depth map information corresponding to described reference-view according to the sequence number value of described reference-view;
Described acquiring unit also for, according to the sequence number value of described reference-view, obtain codec unit CU address and the predicting unit PU partitioning scheme of described reference-view, described CU address is coding CU address or decoding CU address;
Computing unit, calculates for the described NBDV value, described CU address and the described PU partitioning scheme that obtain according to described acquiring unit the positional information that obtains a PU piece;
Described determining unit also for, at described computing unit, calculate in the described PU piece obtaining and determine M pixel, described M equals 2 or 3;
Described acquiring unit also for, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel;
Comparing unit, the depth value of described M the pixel of determining for more described determining unit, the pixel of acquisition depth value maximum;
Described acquiring unit also for, obtain the mapping value of the pixel of described depth value maximum;
Described computing unit also for, horizontal component using the mapping value of the pixel of described depth value maximum as the first motion vectors, according to the horizontal component of described the first motion vectors calculate obtain a Do-NBDV horizontal component, the vertical component that the vertical component of a described Do-NBDV is described NBDV.。
In conjunction with second aspect, in the possible implementation of the first, also comprise coding unit and decoding unit,
Described coding unit, before determining reference-view in described determining unit, encodes and carries out difference vector and calculate the basic view of video;
Described decoding unit, before determining reference-view in described determining unit, decodes and carries out difference vector and calculate the basic view of described video.
In conjunction with second aspect, in the possible implementation of the second, described computing unit also for, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece;
Described determining unit also for, at described computing unit, calculate in described the 2nd PU piece obtaining and determine N pixel, described N equals 2 or 3;
Described acquiring unit also for, according to depth map information corresponding to described reference-view, obtain the depth value of a described N pixel;
Comparing unit, the depth value of described N the pixel of determining for more described determining unit, the pixel of acquisition depth value maximum;
Described acquiring unit also for, obtain the mapping value of the pixel of depth value maximum in a described N pixel;
Described computing unit also for, horizontal component using the mapping value of the pixel of described depth value maximum as the second motion vectors, according to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV, the vertical component that the vertical component of described the 2nd Do-NBDV is described NBDV.
In conjunction with second aspect, in the third possible implementation, described determining unit specifically for, in a described PU piece, determine lower-left and 2 of bottom rights pixel;
Described determining unit specifically for, in a described PU piece determine upper right, lower-left and 3 of bottom rights pixel.
In conjunction with second aspect, in the 4th kind of possible implementation, described acquiring unit specifically for, obtain mobile vector MV value and the sequence number value of the time domain adjacent block of the coding unit CU piece of described reference-view; Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described time domain adjacent block as described reference-view.
In conjunction with second aspect, in the 5th kind of possible implementation, described acquiring unit specifically for, obtain MV value and the sequence number value of the spatial domain adjacent block of the CU piece of described reference-view; Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described spatial domain adjacent block as described reference-view.
In conjunction with second aspect, in the 6th kind of possible implementation, also comprise detecting unit,
Described detecting unit is used for, and detects described reference-view and whether has motion prediction compensation block;
Described acquiring unit specifically for, after described detecting unit detects that described reference-view exists motion prediction compensation block, obtain MV value and the sequence number value of described motion prediction compensation block; Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of described motion prediction compensation block as described reference-view.
In conjunction with second aspect, in the 7th kind of possible implementation, also comprise setting unit,
Described setting unit is used for, and described NBDV is preset as to 0 vector;
Described acquiring unit specifically for, obtain the sequence number value of looking a reference frame of the CU piece of described reference-view, and using the described sequence number value of a reference frame of looking as the sequence number value of described reference-view.
Do-NBDV acquisition methods and video decoder that the embodiment of the present invention provides, in calculating Do-NBDV process, utilize less pixel information to calculate and obtain Do-NBDV value, and recalculate and obtain Do-NBDV value more accurately the Do-NBDV value of acquisition as input, with in having technology, calculate Do-NBDV process redundancy and Do-NBDV value precision not high compared with, method provided by the invention and device, reduce computation complexity, improved the precision of Do-NBDV value, improved code efficiency.
Accompanying drawing explanation
The schematic flow sheet of the Do-NBDV acquisition methods that Fig. 1 provides for the embodiment of the present invention 1;
The schematic flow sheet of the Do-NBDV acquisition methods that Fig. 2 provides for the embodiment of the present invention 2;
The schematic flow sheet of the another kind of Do-NBDV acquisition methods that Fig. 3 provides for the embodiment of the present invention 2;
Employing 2 pixels that Fig. 4 provides for the embodiment of the present invention 2 calculate the schematic flow sheet of Do-NBDV;
The structured flowchart of the video decoder that Fig. 5 provides for the embodiment of the present invention 3;
The structured flowchart of the video processor that Fig. 6 provides for the embodiment of the present invention 4.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.
Current block-based coding and decoding video combination frame generally includes prediction module, conversion module, quantization modules and coding/decoding module; Wherein prediction module obtains the prediction block message of the image block of video sequence coding image by coding, and then obtain image block residual error, predictive compensation module obtains the prediction block message of current decoded image blocks by decoding, then obtains current decoded image blocks according to described image block residual error.Conventionally, prediction module comprises infra-frame prediction and inter prediction, wherein, inter prediction technology utilize present image contiguous encoding or decoded picture Pixel Information has been removed current image block redundant information to obtain residual error.Because the image in synchronization different points of view may have similar picture material, when inter prediction adopt depending between prediction can further eliminate the information redundancy between different points of view image.Difference vector can represent the vector correlation of corresponding similar macro blocks between synchronization different points of view image.Looking in a predictive coding process, difference vector is as determine the corresponding macro block position between different points of view depending on a motion vector.And similarity degree between corresponding macro block has determined the redundancy degree of coding, therefore depending on accurately whether directly having affected of a vector, look a predictive coding performance, reasonably difference vector computational methods can effectively improve coding and decoding video efficiency.
Embodiment 1:
The invention provides a kind of Do-NBDV acquisition methods, as shown in Figure 1, said method comprising the steps of:
101, determine reference-view.
Before this, need to encode to the basic view of video, or video basic point is decoded.A viewpoint in a two field picture being completed after coding (or decoding), when second viewpoint and viewpoint are afterwards encoded to (or decoding), just need to calculate difference vector.Difference vector can calculate acquisition by Do-NBDV, while therefore carrying out Video coding (or decoding), can first calculate and obtain Do-NBDV value, according to described Do-NBDV value, is calculating and is obtaining difference vector.
Here, more than one of the viewpoint that has same image content with described basic view, when determining reference-view, judge the CU(Coding Unit of selected viewpoint, coding unit) whether the image sequence of sequence number value corresponding to piece and basic view number consistent, if consistent, determine that this viewpoint is reference-view; Or, judge that whether the sequence number value of reference frame viewpoint of the CU piece of selected viewpoint is consistent with the sequence number value of basic view, if consistent, determine that this viewpoint is reference-view.
102, obtain the sequence number value based on adjacent block difference vector NBDV value and described reference-view, according to the sequence number value of described reference-view, obtain depth map information corresponding to described reference-view.
Need to calculate NBDV value calculating during described Do-NBDV, and NBDV can be used in its correspondence depending between predict that encoding and decoding, virtual view are predicted and residual prediction on.At present, 3D video has adopted the mode of many views plus depth to carry out encoding and decoding, carrying out in multi-view coded process, give each view allocation an id value to distinguish each view.The sequence number value of reference-view of the present invention is exactly above-mentioned id value.In addition, depth map information has comprised all information entrained in certain depth map, mainly contains: information of the reference frame of the depth value of each pixel, motion vector, this depth map etc. in the size of figure, this depth map.
The sequence number value of NBDV value described here and described reference-view can be to calculate and obtain according to the time domain adjacent block information of the CU piece of described reference-view, can be to calculate and obtain according to the spatial domain adjacent block information of the CU piece of described reference-view, also can be to calculate and obtain according to the motion prediction compensation block information of described reference-view, or described NBDV value is preset as to 0 vector, the sequence number value with reference to the sequence number value of looking a reference frame of the CU piece of view as described reference-view.It should be noted that, each view can be to there being multiple reference frames, in the embodiment of the present invention, described reference frame is one that in multiple reference frames of reference-view, determines in advance, therefore can obtain by obtaining the sequence number value of this definite reference frame the sequence number value of described reference-view.
Like this, just can obtain depth map information corresponding to described reference-view according to the sequence number value of described reference-view, to calculate, obtain described Do-NBDV.
103, according to the sequence number value of described reference-view, obtain CU address and the predicting unit PU partitioning scheme of described reference-view, described CU address is coding CU address or decoding CU address.
Wherein.The CU address of need to encoding when encoding, the CU address of need to decoding when decoding.Before this, will record the corresponding relation between the PU partitioning scheme of the sequence number value of reference-view and the CU block address of described reference-view, employing, therefore, according to the sequence number value of described reference-view, just can obtain coding unit CU address and the predicting unit PU partitioning scheme of described reference-view.
104, according to described NBDV value, described CU address and described PU partitioning scheme, calculate the positional information that obtains a PU piece.
According to current CU address, can obtain corresponding CU piece, and then just can obtain the coordinate figure of the pixel in this CU piece upper left corner, then in conjunction with described NBDV, shine upon and can obtain PU positional information.
105, in a described PU piece, determine M pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel.
106, the horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, calculates the horizontal component that obtains a Do-NBDV according to the horizontal component of described the first motion vectors, described M equals 2 or 3.
The mapping value of pixel is to start initialized time according to the parallax parameter value of each pixel that camera position writes at coding and decoding video.Therefore, just can corresponding its correspondence of acquisition according to certain pixel mapping value.In addition, method correction provided by the invention the horizontal component of Do-NBDV, the vertical component of described Do-NBDV will not change, still for before the vertical component of the NBDV that obtains.
Here, in a described PU piece, determine lower-left and 2 of bottom rights pixel, and obtain the depth value of 2 pixels in described lower-left and bottom right; Horizontal component using the mapping value of pixel larger depth value as the first motion vectors; According to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV.It should be noted that, at PU piece one jiao, can obtain a pixel, therefore in lower-left and bottom right can correspondingly obtain two pixels.
Or, in a described PU piece, determine upper right, lower-left and 3 of bottom rights pixel, and obtain the depth value of described upper right, 3 pixels in lower-left and bottom right; Horizontal component using the mapping value of the pixel of depth value maximum as the first motion vectors; According to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV.
In addition, after step 104, can also calculate the positional information that obtains the 2nd PU piece according to a described Do-NBDV, described CU address and described PU partitioning scheme.
In described the 2nd PU piece, determine N pixel, described N equals 2 or 3, horizontal component using the mapping value of the pixel of depth value maximum in a described N pixel as the second motion vectors, calculates the horizontal component that obtains the 2nd Do-NBDV according to the horizontal component of described the second motion vectors.The vertical component of the 2nd Do-NBDV is still the vertical component of described NBDV.The Do-NBDV value precision obtaining is like this higher, further adopts described Do-NBDV to encode and can improve code efficiency.
The Do-NBDV acquisition methods that the embodiment of the present invention provides, in calculating Do-NBDV process, utilize less pixel information to calculate and obtain Do-NBDV value, and recalculate and obtain Do-NBDV value more accurately the Do-NBDV value of acquisition as input, with in having technology, calculate Do-NBDV process redundancy and Do-NBDV value precision not high compared with, method provided by the invention and device, reduce computation complexity, improved the precision of Do-NBDV value, improved code efficiency.
Embodiment 2:
The invention provides a kind of Do-NBDV acquisition methods, as shown in Figure 2, said method comprising the steps of:
201, the basic viewpoint of video is encoded, carry out difference vector and calculate and determine reference-view.
Under normal circumstances, a viewpoint in a two field picture is completed after coding, when second viewpoint and viewpoint are afterwards encoded, just need to calculate difference vector.This is because difference vector refers to the vector correlation of corresponding similar macro blocks between the image of synchronization different points of view.Difference vector can calculate acquisition by Do-NBDV, while therefore carrying out Video coding, can first calculate and obtain Do-NBDV value, according to described Do-NBDV value, is calculating and is obtaining difference vector.
Here, more than one of the view that has same image content with described basic view, when determining reference-view, judges whether the image sequence of sequence number value that the CU piece of selected view is corresponding and basic view is number consistent, if consistent, definite this viewpoint is reference-view; Or, judge that whether the sequence number value of reference frame view of the CU piece of selected viewpoint is consistent with the sequence number value of basic view, if consistent, determine that this view is reference-view.
Certainly, difference vector can calculate acquisition by NBDV, calculates situation about obtaining in this consideration by Do-NBDV.
202, obtain the sequence number value of NBDV value, described reference-view, according to the sequence number value of described reference-view, obtain depth map information corresponding to described reference-view.
Generally, can obtain the mobile vector MV value of time domain adjacent block of the coding unit CU piece of described reference-view and the sequence number value of reference-view.Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of the reference-view of described time domain adjacent block as described reference-view.
Or, obtain the MV value of spatial domain adjacent block of the CU piece of described reference-view and the sequence number value of reference-view.Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of the reference-view of described spatial domain adjacent block as described reference-view.
If time domain adjacent block and the spatial domain adjacent block of the CU piece of described reference-view are all unavailable, detect described reference-view and whether have motion prediction compensation block.If described reference-view exists motion prediction compensation block, obtain the MV value of described motion prediction compensation block and the sequence number value of reference-view.Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of the reference-view of described motion prediction compensation block as described reference-view.
If cannot obtain the MV value of described motion prediction compensation block, described NBDV is preset as to 0 vector.Obtain the sequence number value of the reference-view of looking a reference frame of the CU piece of described reference-view, and the sequence number value using the sequence number value of described reference-view of looking a reference frame as described reference-view.
203, according to the sequence number value of described reference-view, obtain CU address and the predicting unit PU partitioning scheme of described reference-view.
Here, described CU address is coding CU address.The CU address of need to encoding when encoding also needs the CU address of decoding when decoding.
204, according to described NBDV value, described CU address and described PU partitioning scheme, calculate the positional information that obtains the first predicting unit PU piece.
Here, adopt preset algorithm, according to described NBDV value, described CU address and described PU partitioning scheme, calculate the positional information that obtains the first predicting unit PU piece.Preset algorithm is not limited.Obtaining after the positional information of PU piece, just can on PU piece corresponding to described positional information, choose pixel, and then calculate acquisition Do-NBDV value.
205, in a described PU piece, determine M pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel, horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, calculates the horizontal component that obtains a Do-NBDV according to the horizontal component of described the first motion vectors.
Wherein, described M equals 2 or 3.If described M equals 2, in a described PU piece, determine lower-left and 2 of bottom rights pixel, and obtain the depth value of 2 pixels in described lower-left and bottom right.By relatively obtaining the larger pixel of depth value in these 2 pixels, and then the mapping value of this pixel of calculating acquisition, the mapping value of this pixel, as the horizontal component of the first motion vectors, is calculated the horizontal component that obtains a Do-NBDV according to the horizontal component of described the first motion vectors.Example, in video coding process, viewpoint 1 (viewpoint 0 being encoded) is encoded, suppose that in difference vector computational process, can carry out Do-NBDV calculates.As shown in Figure 4, if the NBDV obtaining is (176,112).According to NBDV, calculate corresponding PU piece, and determine lower-left and 2 of bottom rights pixel on described PU piece.Obtain the depth value of these 2 pixels and get larger depth value, according to described larger depth value, correspond to the pixel of bottom right.Obtain bottom right pixel and put corresponding mapping value 183, and as the horizontal component of 183 motion vectors, obtain the horizontal component of Do-NBDV after clip operation, the vertical component of Do-NBDV is still the vertical component of described NBDV in addition.
If described M equals 3, in a described PU piece, determine upper right, lower-left and 3 of bottom rights pixel, and obtain the depth value of described upper right, 3 pixels in lower-left and bottom right.And calculate the depth value of these three pixels, and then obtain the pixel of depth value maximum.Horizontal component using the mapping value of the pixel of described depth value maximum as the first motion vectors, calculates the horizontal component that obtains a Do-NBDV according to the horizontal component of described the first motion vectors.
206, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece.
Here, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains a more accurate PU piece, just can continue to calculate the more accurate Do-NBDV of acquisition.
In video coding process, viewpoint 1 (viewpoint 0 being encoded) is encoded, establish in difference vector computational process and can carry out Do-NBDV calculating.Example, if the NBDV obtaining is (176,112).According to obtaining Do-NBDV after the first calculating of the corresponding depth map information of the sequence number value of NBDV, reference-view and reference-view, be (184,112), by (184,112) as input NBDV value, once calculate again, and (186,112) that precision is higher final result as Do-NBDV will be obtained.
207, in described the 2nd PU piece, determine N pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described N pixel, horizontal component using the mapping value of the pixel of depth value maximum in a described N pixel as the second motion vectors, calculates the horizontal component that obtains the 2nd Do-NBDV according to the horizontal component of described the second motion vectors.
Here, described N equals 2 or 3.The value of described N is not subject to the impact of M.When described M gets 2, described N can get 2 or 3, and described M gets 3 o'clock, and described N can get 2 or 3.
When described N equals 2, in described the 2nd PU piece, determine lower-left and 2 of bottom rights pixel, and obtain the depth value of 2 pixels in described lower-left and bottom right.Horizontal component using the mapping value of pixel larger depth value as the second motion vectors.According to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV.
Or, when described N equals 3, in described the 2nd PU piece, determine upper right, lower-left and 3 of bottom rights pixel, and obtain the depth value of described upper right, 3 pixels in lower-left and bottom right.Horizontal component using the mapping value of the pixel of depth value maximum as the second motion vectors.According to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV.
The present invention also provides a kind of Do-NBDV acquisition methods, as shown in Figure 3, said method comprising the steps of:
301, the basic viewpoint of video is decoded, carry out difference vector and calculate and determine reference-view.
Under normal circumstances, a viewpoint in a two field picture is completed after decoding, when second viewpoint and viewpoint are afterwards decoded, just need to calculate difference vector.This is because difference vector refers to the vector correlation of corresponding similar macro blocks between the image of synchronization different points of view.Difference vector can calculate acquisition by Do-NBDV, while therefore carrying out video decode, can first calculate and obtain Do-NBDV value, according to described Do-NBDV value, is calculating and is obtaining difference vector.
302, obtain the sequence number value of NBDV value, described reference-view, according to the sequence number value of described reference-view, obtain depth map information corresponding to described reference-view.
303, according to the sequence number value of described reference-view, obtain CU address and the predicting unit PU partitioning scheme of described reference-view.
304, according to described NBDV value, described CU address and described PU partitioning scheme, calculate the positional information that obtains the first predicting unit PU piece.
305, in a described PU piece, determine M pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel, horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, calculates the horizontal component that obtains a Do-NBDV according to the horizontal component of described the first motion vectors.
306, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece.
Here, described CU address is decoding CU address.
307, in described the 2nd PU piece, determine N pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described N pixel, horizontal component using the mapping value of the pixel of depth value maximum in a described N pixel as the second motion vectors, calculates the horizontal component that obtains the 2nd Do-NBDV according to the horizontal component of described the second motion vectors.
Do-NBDV acquisition methods and video decoder that the embodiment of the present invention provides, in calculating Do-NBDV process, utilize less pixel information to calculate and obtain Do-NBDV value, and recalculate and obtain Do-NBDV value more accurately the Do-NBDV value of acquisition as input, with in having technology, calculate Do-NBDV process redundancy and Do-NBDV value precision not high compared with, method provided by the invention and device, reduce computation complexity, improved the precision of Do-NBDV value, improved code efficiency.
Embodiment 3:
The invention provides a kind of video decoder, as shown in Figure 5, described video processor comprises: determining unit 501, acquiring unit 502, computing unit 503 and comparing unit 504.
Determining unit 501, for determining reference-view.
Acquiring unit 502, for obtaining the sequence number value based on adjacent block difference vector NBDV value and described reference-view, and obtains depth map information corresponding to described reference-view according to the sequence number value of described reference-view.
Described acquiring unit 502 also for, according to the sequence number value of described reference-view, obtain coding unit CU address and the predicting unit PU partitioning scheme of described reference-view.
Described determining unit 501 also for, in described computing unit 503 calculates the described PU piece obtaining, determine M pixel, described M equals 2 or 3.
Described acquiring unit 502 also for, according to depth map information corresponding to described reference-view, obtain depth value and the mapping value of a described M pixel.
Comparing unit 504, the depth value of described M the pixel of determining for more described determining unit 501, the pixel of acquisition depth value maximum.
Described acquiring unit 502 also for, obtain the mapping value of the pixel of described depth value maximum.
Described computing unit 503 also for, horizontal component using the mapping value of the pixel of described depth value maximum as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV.
Described video processor also comprises coding unit and decoding unit,
Described coding unit, before determining reference-view in described determining unit 501, encodes and carries out difference vector and calculate the basic view of video.
Described decoding unit, before determining reference-view in described determining unit 501, decodes and carries out difference vector and calculate the basic view of video.
Described computing unit 503 also for, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece.
Described determining unit 501 also for, in described computing unit 503 calculates described the 2nd PU piece obtaining, determine N pixel, described N equals 2 or 3.
Described acquiring unit 503 also for, according to depth map information corresponding to described reference-view, obtain depth value and the mapping value of a described N pixel.
Comparing unit 504, the depth value of described N the pixel of determining for more described determining unit 501, the pixel of acquisition depth value maximum.
Described computing unit 503 also for, horizontal component using the mapping value of the pixel of described depth value maximum as the second motion vectors, according to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV.
Described determining unit 501 specifically for, in a described PU piece, determine lower-left and 2 of bottom rights pixel.
Described determining unit 501 specifically for, in a described PU piece determine upper right, lower-left and 3 of bottom rights pixel.
Described acquiring unit 502 specifically for, obtain mobile vector MV value and the sequence number value of the time domain adjacent block of the coding unit CU piece of described reference-view.Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described time domain adjacent block as described reference-view.
Described acquiring unit 502 specifically for, obtain MV value and the sequence number value of the spatial domain adjacent block of the CU piece of described reference-view.Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described spatial domain adjacent block as described reference-view.
Described video processor also comprises detecting unit, and described detecting unit is used for, and detects described reference-view and whether has motion prediction compensation block.
Described acquiring unit 502 specifically for, after described detecting unit detects that described reference-view exists motion prediction compensation block, obtain MV value and the sequence number value of described motion prediction compensation block.Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of described motion prediction compensation block as described reference-view.
Described video processor also comprises setting unit, and described setting unit is used for, and described NBDV is preset as to 0 vector.
Described acquiring unit 502 specifically for, obtain the sequence number value of looking a reference frame of the CU piece of described reference-view, and using the described sequence number value of a reference frame of looking as the sequence number value of described reference-view.
The video decoder that the embodiment of the present invention provides, in calculating Do-NBDV process, utilize less pixel information to calculate and obtain Do-NBDV value, and recalculate and obtain Do-NBDV value more accurately the Do-NBDV value of acquisition as input, with in having technology, calculate Do-NBDV process redundancy and Do-NBDV value precision not high compared with, method provided by the invention and device, reduce computation complexity, improved the precision of Do-NBDV value, improved code efficiency.
Embodiment 4:
The embodiment of the present invention provides a kind of video processor, on hardware is realized, coding unit, decoding unit, detecting unit and the setting unit described in the determining unit described in Fig. 5, acquiring unit, computing unit and comparing unit and above embodiment can be embedded in the processor of described video processor with example, in hardware or form of software.This processor can be CPU (CPU), also can single-chip microcomputer.
As shown in Figure 6, described video processor comprises: memory 601 and processor 602.Wherein, in memory 601, store batch processing code, and processor 602 is for calling the program code of memory 601 storages, for carrying out following operation:
Processor 602, for determining reference-view.
Processor 602, for obtaining the sequence number value based on adjacent block difference vector NBDV value and described reference-view, and obtains depth map information corresponding to described reference-view according to the sequence number value of described reference-view.
Described processor 602 also for, according to the sequence number value of described reference-view, obtain coding unit CU address and the predicting unit PU partitioning scheme of described reference-view.
Processor 602, for calculating the positional information that obtains a PU piece according to described NBDV value, described CU address and described PU partitioning scheme.
Described processor 602 also for, calculate obtain a described PU piece in determine M pixel, described M equals 2 or 3.
Described processor 602 also for, according to depth map information corresponding to described reference-view, obtain depth value and the mapping value of a described M pixel.
Processor 602, for the depth value of a more described M pixel, obtains the pixel of depth value maximum.
Described processor 602 also for, obtain the mapping value of the pixel of described depth value maximum.
Described processor 602 also for, horizontal component using the mapping value of the pixel of described depth value maximum as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV.
Described processor 602, for before determining reference-view, encodes and carries out difference vector and calculate the basic view of video.
Described processor 602, for before determining reference-view, decodes and carries out difference vector and calculate the basic view of video.
Described processor 602 also for, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece.
Described processor 602 also for, calculate obtain described the 2nd PU piece in determine N pixel, described N equals 2 or 3.
Described processor 602 also for, according to depth map information corresponding to described reference-view, obtain depth value and the mapping value of a described N pixel.
Processor 602, for the depth value of a more described N pixel, obtains the pixel of depth value maximum.
Described processor 602 also for, horizontal component using the mapping value of the pixel of described depth value maximum as the second motion vectors, according to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV.
Described processor 602 specifically for, in a described PU piece, determine lower-left and 2 of bottom rights pixel.
Described processor 602 specifically for, in a described PU piece determine upper right, lower-left and 3 of bottom rights pixel.
Described processor 602 specifically for, obtain mobile vector MV value and the sequence number value of the time domain adjacent block of the coding unit CU piece of described reference-view.Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described time domain adjacent block as described reference-view.
Described processor 602 specifically for, obtain MV value and the sequence number value of the spatial domain adjacent block of the CU piece of described reference-view.Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described spatial domain adjacent block as described reference-view.
Processor 602 for, detect described reference-view and whether have motion prediction compensation block.
Described processor 602 specifically for, after detecting that described reference-view exists motion prediction compensation block, obtain MV value and the sequence number value of described motion prediction compensation block.Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of described motion prediction compensation block as described reference-view.
Described processor 602 for, described NBDV is preset as to 0 vector.
Described processor 602 specifically for, obtain the sequence number value of looking a reference frame of the CU piece of described reference-view, and using the described sequence number value of a reference frame of looking as the sequence number value of described reference-view.
The video processor that the embodiment of the present invention provides, in calculating Do-NBDV process, utilize less pixel information to calculate and obtain Do-NBDV value, and recalculate and obtain Do-NBDV value more accurately the Do-NBDV value of acquisition as input, with in having technology, calculate Do-NBDV process redundancy and Do-NBDV value precision not high compared with, method provided by the invention and device, reduce computation complexity, improved the precision of Do-NBDV value, improved code efficiency.
One of ordinary skill in the art will appreciate that: all or part of step that realizes said method embodiment can complete by the relevant hardware of program command, aforesaid program can be stored in computer read/write memory medium, this program, when carrying out, is carried out the step that comprises said method embodiment; And aforesaid storage medium comprises: various media that can be program code stored such as ROM, RAM, magnetic disc or CDs.
The above; be only the specific embodiment of the present invention, but protection scope of the present invention is not limited to this, any be familiar with those skilled in the art the present invention disclose technical scope in; can expect easily changing or replacing, within all should being encompassed in protection scope of the present invention.Therefore, protection scope of the present invention should described be as the criterion with the protection range of claim.
Claims (16)
1. the degree of depth, based on an adjacent block difference vector Do-NBDV acquisition methods, is characterized in that, described method comprises:
Determine reference-view;
Obtain the sequence number value based on adjacent block difference vector NBDV value and described reference-view, according to the sequence number value of described reference-view, obtain depth map information corresponding to described reference-view;
According to the sequence number value of described reference-view, obtain codec unit CU address and the predicting unit PU partitioning scheme of described reference-view, described CU address is coding CU address or decoding CU address;
According to described NBDV value, described CU address and described PU partitioning scheme, calculate the positional information that obtains a PU piece;
In a described PU piece, determine M pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel;
Horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate and obtain the horizontal component of first degree of depth based on adjacent block difference vector Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV; Described M equals 2 or 3.
2. method according to claim 1, is characterized in that, before described definite reference-view, described method also comprises:
The basic view of video is encoded and carried out difference vector and calculate;
Or, the basic view of described video is decoded and is carried out difference vector and calculate.
3. method according to claim 1, is characterized in that, after calculating the horizontal component that obtains a Do-NBDV, described method also comprises:
According to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece;
In described the 2nd PU piece, determine N pixel, according to depth map information corresponding to described reference-view, obtain the depth value of a described N pixel;
Horizontal component using the mapping value of the pixel of depth value maximum in a described N pixel as the second motion vectors, according to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV, the vertical component that the vertical component of described the 2nd Do-NBDV is described NBDV; Described N equals 2 or 3.
4. method according to claim 1, it is characterized in that, described in a described PU piece determine M pixel, horizontal component using the mapping value of the pixel of depth value maximum in a described M pixel as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV, comprising:
In a described PU piece, determine lower-left and 2 of bottom rights pixel, and obtain the depth value of 2 pixels in described lower-left and bottom right;
Horizontal component using the mapping value of pixel larger depth value as the first motion vectors;
According to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV;
Or, in a described PU piece, determine upper right, lower-left and 3 of bottom rights pixel, and obtain the depth value of described upper right, 3 pixels in lower-left and bottom right;
Horizontal component using the mapping value of the pixel of depth value maximum as the first motion vectors;
According to the horizontal component of described the first motion vectors, calculate the horizontal component that obtains a Do-NBDV.
5. method according to claim 1, is characterized in that, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Obtain mobile vector MV value and the sequence number value of the time domain adjacent block of the coding unit CU piece of described reference-view;
Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described time domain adjacent block as described reference-view.
6. method according to claim 1, is characterized in that, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Obtain MV value and the sequence number value of the spatial domain adjacent block of the CU piece of described reference-view;
Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described spatial domain adjacent block as described reference-view.
7. method according to claim 1, is characterized in that, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Detect described reference-view and whether have motion prediction compensation block;
If exist, obtain MV value and the sequence number value of described motion prediction compensation block;
Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of described motion prediction compensation block as described reference-view.
8. method according to claim 1, is characterized in that, described in obtain the sequence number value of NBDV value and described reference-view, comprising:
Described NBDV is preset as to 0 vector;
Obtain the sequence number value of looking a reference frame of the CU piece of described reference-view, and using the described sequence number value of a reference frame of looking as the sequence number value of described reference-view.
9. a video decoder, is characterized in that, comprising:
Determining unit, for determining reference-view;
Acquiring unit, for obtaining the sequence number value based on adjacent block difference vector NBDV value and described reference-view, and obtains depth map information corresponding to described reference-view according to the sequence number value of described reference-view;
Described acquiring unit also for, according to the sequence number value of described reference-view, obtain codec unit CU address and the predicting unit PU partitioning scheme of described reference-view, described CU address is coding CU address or decoding CU address;
Computing unit, calculates for the described NBDV value, described CU address and the described PU partitioning scheme that obtain according to described acquiring unit the positional information that obtains a PU piece;
Described determining unit also for, at described computing unit, calculate in the described PU piece obtaining and determine M pixel, described M equals 2 or 3;
Described acquiring unit also for, according to depth map information corresponding to described reference-view, obtain the depth value of a described M pixel;
Comparing unit, the depth value of described M the pixel of determining for more described determining unit, the pixel of acquisition depth value maximum;
Described acquiring unit also for, obtain the mapping value of the pixel of described depth value maximum;
Described computing unit also for, horizontal component using the mapping value of the pixel of described depth value maximum as the first motion vectors, according to the horizontal component of described the first motion vectors, calculate and obtain the horizontal component of first degree of depth based on adjacent block difference vector Do-NBDV, the vertical component that the vertical component of a described Do-NBDV is described NBDV.
10. video decoder according to claim 9, is characterized in that, also comprises coding unit and decoding unit,
Described coding unit, before determining reference-view in described determining unit, encodes and carries out difference vector and calculate the basic view of video;
Described decoding unit, before determining reference-view in described determining unit, decodes and carries out difference vector and calculate the basic view of described video.
11. video decoders according to claim 9, is characterized in that,
Described computing unit also for, according to a described Do-NBDV, described CU address and described PU partitioning scheme, calculate the positional information that obtains the 2nd PU piece;
Described determining unit also for, at described computing unit, calculate in described the 2nd PU piece obtaining and determine N pixel, described N equals 2 or 3;
Described acquiring unit also for, according to depth map information corresponding to described reference-view, obtain the depth value of a described N pixel;
Comparing unit, the depth value of described N the pixel of determining for more described determining unit, the pixel of acquisition depth value maximum;
Described acquiring unit also for, obtain the mapping value of the pixel of depth value maximum in a described N pixel;
Described computing unit also for, horizontal component using the mapping value of the pixel of described depth value maximum as the second motion vectors, according to the horizontal component of described the second motion vectors, calculate the horizontal component that obtains the 2nd Do-NBDV, the vertical component that the vertical component of described the 2nd Do-NBDV is described NBDV.
12. video decoders according to claim 9, is characterized in that,
Described determining unit specifically for, in a described PU piece, determine lower-left and 2 of bottom rights pixel;
Described determining unit specifically for, in a described PU piece determine upper right, lower-left and 3 of bottom rights pixel.
13. video decoders according to claim 9, is characterized in that,
Described acquiring unit specifically for, obtain mobile vector MV value and the sequence number value of the time domain adjacent block of the coding unit CU piece of described reference-view; Using the MV value of described time domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described time domain adjacent block as described reference-view.
14. video decoders according to claim 9, is characterized in that,
Described acquiring unit specifically for, obtain MV value and the sequence number value of the spatial domain adjacent block of the CU piece of described reference-view; Using the MV value of described spatial domain adjacent block as described NBDV value, the sequence number value using the sequence number value of described spatial domain adjacent block as described reference-view.
15. video decoders according to claim 9, is characterized in that, also comprise detecting unit,
Described detecting unit is used for, and detects described reference-view and whether has motion prediction compensation block;
Described acquiring unit specifically for, after described detecting unit detects that described reference-view exists motion prediction compensation block, obtain MV value and the sequence number value of described motion prediction compensation block; Using the MV value of described motion prediction compensation block as described NBDV, the sequence number value using the sequence number value of described motion prediction compensation block as described reference-view.
16. video decoders according to claim 9, is characterized in that, also comprise setting unit,
Described setting unit is used for, and described NBDV is preset as to 0 vector;
Described acquiring unit specifically for, obtain the sequence number value of looking a reference frame of the CU piece of described reference-view, and using the described sequence number value of a reference frame of looking as the sequence number value of described reference-view.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410003938.9A CN103763557B (en) | 2014-01-03 | 2014-01-03 | A kind of Do NBDV acquisition methods and video decoder |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410003938.9A CN103763557B (en) | 2014-01-03 | 2014-01-03 | A kind of Do NBDV acquisition methods and video decoder |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103763557A true CN103763557A (en) | 2014-04-30 |
| CN103763557B CN103763557B (en) | 2017-06-27 |
Family
ID=50530705
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410003938.9A Expired - Fee Related CN103763557B (en) | 2014-01-03 | 2014-01-03 | A kind of Do NBDV acquisition methods and video decoder |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103763557B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101222627A (en) * | 2007-01-09 | 2008-07-16 | 华为技术有限公司 | Multi-viewpoint video coding and decoding system, method and device for estimating vector |
| WO2012171442A1 (en) * | 2011-06-15 | 2012-12-20 | Mediatek Inc. | Method and apparatus of motion and disparity vector prediction and compensation for 3d video coding |
| WO2013053309A1 (en) * | 2011-10-11 | 2013-04-18 | Mediatek Inc. | Method and apparatus of motion and disparity vector derivation for 3d video coding and hevc |
| WO2013068547A2 (en) * | 2011-11-11 | 2013-05-16 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Efficient multi-view coding using depth-map estimate and update |
| CN103220532A (en) * | 2013-05-02 | 2013-07-24 | 清华大学 | Joint prediction encoding method and joint predication encoding system for stereoscopic video |
-
2014
- 2014-01-03 CN CN201410003938.9A patent/CN103763557B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101222627A (en) * | 2007-01-09 | 2008-07-16 | 华为技术有限公司 | Multi-viewpoint video coding and decoding system, method and device for estimating vector |
| WO2012171442A1 (en) * | 2011-06-15 | 2012-12-20 | Mediatek Inc. | Method and apparatus of motion and disparity vector prediction and compensation for 3d video coding |
| WO2013053309A1 (en) * | 2011-10-11 | 2013-04-18 | Mediatek Inc. | Method and apparatus of motion and disparity vector derivation for 3d video coding and hevc |
| WO2013068547A2 (en) * | 2011-11-11 | 2013-05-16 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Efficient multi-view coding using depth-map estimate and update |
| CN103220532A (en) * | 2013-05-02 | 2013-07-24 | 清华大学 | Joint prediction encoding method and joint predication encoding system for stereoscopic video |
Non-Patent Citations (1)
| Title |
|---|
| JIWOOK JUNG, JAEWON SUNG,SEHOON YEA: "3D-CE1.h related : Adaptive method for Depth Oriented Neighboring Block Disparity Vector", 《JOINT COLLABORATIVE TEAM ON 3D VIDEO CODING EXTENSION DEVELOPMENT OF ITU-T SG 16 WP 3 AND ISO/IEC JTC 1/SC 29/WG 11, 3RD MEETING: GENEVA, CH, 17-23 JAN. 2013, JCT3V-C0112》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103763557B (en) | 2017-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10404993B2 (en) | Picture prediction method and related apparatus | |
| US9736498B2 (en) | Method and apparatus of disparity vector derivation and inter-view motion vector prediction for 3D video coding | |
| US9992494B2 (en) | Method of depth based block partitioning | |
| US20150172714A1 (en) | METHOD AND APPARATUS of INTER-VIEW SUB-PARTITION PREDICTION in 3D VIDEO CODING | |
| CN104412597A (en) | Method and apparatus of unified disparity vector derivation for 3d video coding | |
| JP2015533038A5 (en) | ||
| CN101563927A (en) | Method for decoding a block of a video image | |
| RU2015147403A (en) | METHOD AND DEVICE FOR ENCODING OR DECODING AN IMAGE WITH PREDICTING MOTION INFORMATION BETWEEN LEVELS IN ACCORDANCE WITH THE MOTION INFORMATION COMPRESSION SCHEME | |
| CA2692250A1 (en) | Video encoding and decoding methods using residual prediction, and corresponding apparatuses | |
| CN108989799B (en) | Method and device for selecting reference frame of coding unit and electronic equipment | |
| WO2020181428A1 (en) | Prediction method, encoder, decoder, and computer storage medium | |
| JP2017526268A (en) | Arrangement of adaptive search windows for video coding | |
| CN105474640A (en) | Method and apparatus of camera parameter signaling in 3d video coding | |
| US20150264356A1 (en) | Method of Simplified Depth Based Block Partitioning | |
| CN102377992B (en) | Method and device for obtaining predicted value of motion vector | |
| CN103108183B (en) | Skip mode and Direct mode motion vector predicting method in three-dimension video | |
| CN104333758A (en) | Depth map prediction method, pixel detection method and related devices | |
| CN103079072A (en) | Inter-frame prediction method, encoding equipment and decoding equipment | |
| CN111066322B (en) | Intra-prediction for video coding via perspective information | |
| CN103763557A (en) | Do-NBDV acquiring method and video decoding device | |
| CN106534850B (en) | Image processing apparatus, image interpolation method, and image encoding method | |
| TWI704800B (en) | Boundary filtering for sub-block | |
| CN103997635B (en) | The synthesis viewpoint distortion prediction method of free viewpoint video and coding method | |
| CN103139569A (en) | Multi-viewpoint video coding and decoding method and device and coder and decoder | |
| CN103747265A (en) | NBDV (Disparity Vector from Neighboring Block) acquisition method and video decoding device |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170627 Termination date: 20190103 |