CN107786874A - Directional prediction method and apparatus in two-way frame - Google Patents
Directional prediction method and apparatus in two-way frame Download PDFInfo
- Publication number
- CN107786874A CN107786874A CN201610712740.7A CN201610712740A CN107786874A CN 107786874 A CN107786874 A CN 107786874A CN 201610712740 A CN201610712740 A CN 201610712740A CN 107786874 A CN107786874 A CN 107786874A
- Authority
- CN
- China
- Prior art keywords
- pixel
- predicted
- pixelpred2
- pixelpred1
- coordinate
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/103—Selection of coding mode or of prediction mode
- H04N19/109—Selection of coding mode or of prediction mode among a plurality of temporal predictive coding modes
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/587—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal sub-sampling or interpolation, e.g. decimation or subsequent interpolation of pictures in a video sequence
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention provides directional prediction method and apparatus in two-way frame.The inventive method make use of the texture gradually changeable of encoding block, solve infra-frame prediction border discontinuity and infra-frame prediction precision with distance between prediction pixel and reference pixel increase and the defects of reduce.The inventive method considers the roll-off characteristic of encoding block residual error, and prediction is used between weight coefficient and prediction pixel and reference pixel apart from exponent function relation, improves infra-frame prediction performance.The present invention is while directional prediction method in bidirectional frame is proposed, it is also proposed that corresponding device.
Description
Technical field
The present invention relates to the encoding and decoding of video, and relate more specifically to take in two-way frame directional prediction mechanism to carry
Rise the video coding-decoding method and device of intraframe coding performance.
Background technology
Compared with H.264/AVC, HEVC equally employs the pixel prediction on spatial domain, by for increasing predictive mode
Number so that intra prediction direction is finer, the texture of more preferable simulation different directions, so as to further reduce prediction residual energy
Amount, finally lifts coding efficiency.
HEVC be luminance component define 35 kinds of intra prediction modes, including 33 kinds of unidirectional directional prediction patterns with
2 kinds of non-directional predictive mode DC and Planar, as shown in Figure 9.Predictive mode 2~18 belongs to horizontal race's predictive mode, prediction
The vertical race's predictive mode in the waters of pattern 19~34.A kind of predictive mode in predictive mode 2~34 corresponds to a kind of prediction direction.
In view of chrominance texture relatively flat, HEVC only defines 5 candidate patterns for chromatic component, is DC, Planar, water respectively
Flat, vertical and corresponding luma prediction modes.When being predicted, the lower left side pixel of present encoding block, directly to the left picture
Element, top left corner pixel, positive upside pixel and upper right side pixel are incorporated by reference pixel, and texture extrapolation is carried out along specific direction
(directional mode) or linear filtering (non-directional pattern) are to generate predicted value.In order to handle the flat encoding block of texture,
HEVC provides two kinds of non-directional patterns DC and Planar.DC patterns use the zero-order method of relative coarseness, suitable for flat
Encoding block, i.e., all pixels value all identical situations in block.DC patterns are with the DC component of reference pixel come predictive coding block
The interior all pixels in addition to the first row and first row;For the first row and the pixel of first row, respectively with adjacent upside and a left side
The reference pixel of side carries out linear weighted function with DC component, generates final predicted value.The first row and under the above-mentioned pattern to DC
The treatment technology of one row pixel is edge smoothing technology, and it can solve discontinuous problem of the predicted value in boundary.Except answering
Outside DC patterns, edge smoothing technology is also applied in vertically and horizontally predictive mode.Planar patterns employ two dimension
Linear model, to simulate gradual change texture, the encoding block of linear gradient is presented suitable for pixel value.
For directional prediction pattern in frame, the lower left side pixel of present encoding block, directly to the left pixel, top left corner pixel,
Positive upside pixel and upper right side pixel are incorporated by reference pixel, and carrying out texture along specific direction extrapolates to obtain predicted value.
HEVC defines the reference pixel of two dimension, by the way that the reference pixel of one of dimension is projected into another along prediction direction
The purpose of reference pixel dimensionality reduction is can reach in dimension.For the horizontal race direction that coding mode is 2 to 17, left side will be utilized
Reference pixel, now upside pixel is by the row where left side reference pixel is projected to according to prediction direction;It is for coding mode
18 to 34 vertical race direction, the reference pixel of upside will be used, now left side reference pixel will project to according to prediction direction
Row where the reference pixel of top.The generating process of prediction pixel is exactly the position by each pixel to be predicted according to prediction
Row or column where direction projection to reference pixel;If just projecting in whole pixel, the value of direct copying reference pixel
;If projecting on sub-pix, the linear interpolation of 1/32 precision is carried out using two neighboring whole pixel, you can obtain sub- picture
The value of element, using the value of the sub-pix as predicted value.Employed after the video encoding standard of new generation after HEVC in 67 kinds of frames
Predictive mode.HEVC supports 4x4,8x8,16x16,32x32,64x64 coded block size in infra-frame prediction;AVS is in frame
During prediction, 4x4,4x8,8x8,16x16,8x16,32x32,16x32,64x64,32x64 coded block size are supported.
For HEVC predictive mode 2 and predictive mode 34, predicted value by prediction direction with left-hand column and top row phase
The reconstruction pixel value weighted average of friendship obtains.The weight coefficient that prediction direction intersects gained reconstruction pixel value with left-hand column is 1/2,
The weight coefficient for the reconstruction pixel value that prediction direction intersects with top row is 1/2.
In order to further lift HEVC infra-frame prediction performance, the number handled predictive mode increases.It is each to rebuild
The weight coefficient of pixel with the Euclidean distance exponent function relation that is predicted between pixel position and side reference pixel value, and
Predictive mode is unrelated.
Directional prediction technology has the following disadvantages in existing frame:1. for as with directionality roll-off characteristic
The pre- error of measurement of the complex textures such as texture;2. prediction pixel increases with reference pixel distance, precision of prediction reduces;3. directional prediction is deposited
In border discontinuity.
The content of the invention
Therefore, the present invention proposes a kind of intra-frame prediction method, it uses being predicted for one of following methods generation current block
The predicted value of pixel:
A. in the current block at least 2 be predicted pixel PixelPred1 and PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the left of current block
The reference pixel rebuild in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the upside of current block
The reference pixel rebuild in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [xr] [- 1]-R [- 1] [yr] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [- 1] [yr]+weight1*delta,
The predicted value for being predicted pixel PixelPred2 be R [- 1] [yr]+weight2*delta, wherein weight1 by
PixelPred1 and PixelU vertical range determines that weight2 is determined by PixelPred2 and PixelU vertical range, when
When PixelPred1 and PixelU vertical range is more than PixelPred2 and PixelU vertical range, weight coefficient
Weight1 is less than weight coefficient weight2;
B. in the current block at least 2 be predicted pixel PixelPred1 and PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the left of current block
The reference pixel rebuild in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the upside of current block
The reference pixel rebuild in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [- 1] [yr]-R [xr] [- 1] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [xr] [- 1]+weight3*delta,
The predicted value for being predicted pixel PixelPred2 be R [xr] [- 1]+weight4*delta, wherein weight3 by
PixelPred1 and PixelL horizontal range determines that weight4 is determined by PixelPred2 and PixelL horizontal range, when
When PixelPred1 and PixelL horizontal range is more than PixelPred2 and PixelL horizontal range, weight coefficient
Weight3 is less than weight coefficient weight4.
Further, described method also resides in one below:
A. described weight coefficient weight1 is by being predicted pixel PixelPred1 and its upside reference pixel PixelU
Vertical range obtained as power, i.e. weight1=Hconstantp, wherein p is PixelPred1 and its upside reference image
Plain PixelU vertical range, Hconstant are the index truth of a matter;Described weight coefficient weight2 is by being predicted pixel
What PixelPred2 and its upside reference pixel PixelU vertical range obtained as power, i.e. weight2=Hconstantq,
Wherein q is PixelPred2 and its upside reference pixel PixelU vertical range, and Hconstant is the index truth of a matter;
B. described weight coefficient weight3 is by being predicted pixel PixelPred1 and its left side reference pixel PixelL
Horizontal range obtained as power, i.e. weight3=Vconstantp, wherein p is PixelPred1 and its left side reference picture
Plain PixelL horizontal range, Vconstant are the index truth of a matter;Described weight coefficient weight4 is by being predicted pixel
What PixelPred2 and its left side reference pixel PixelL horizontal range obtained as power, i.e. weight4=Vconstantq,
Wherein q is PixelPred2 and its left side reference pixel PixelL horizontal range, and Vconstant is the index truth of a matter.
Further, described method also resides in one below:
A. the prediction direction for being predicted pixel PixelPred1 uses and horizontal direction angle are less than or equal to 45 degree, institute
State weight coefficient weight1 index truth of a matter Hconstant and be predicted the prediction direction and water of pixel PixelPred1 uses
Flat angular separation is into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and horizontal direction angle are less than
Equal to 45 degree, the index truth of a matter Hconstant of the weight coefficient weight2 and be predicted pixel PixelPred2 use it is pre-
Direction and horizontal direction angle are surveyed into sine relation;
B. the prediction direction for being predicted pixel PixelPred1 uses and vertical direction angle are described to add less than 45 degree
Weight coefficient weight3 index truth of a matter Hconstant and the prediction direction and Vertical Square for being predicted pixel PixelPred1 uses
To angle into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and vertical direction angle are less than 45
Degree, the index truth of a matter Hconstant of the weight coefficient weight4 and the prediction direction for being predicted pixel PixelPred2 uses
With vertical direction angle into sine relation.
Second purpose of the invention is to provide a kind of infra-frame prediction device, and it uses one of following device generation current block
It is predicted the predicted value of pixel:
A. first kind intraprediction unit, in the current block at least 2 be predicted pixel PixelPred1 and
PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the left of current block
The reference pixel rebuild in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the upside of current block
The reference pixel rebuild in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [xr] [- 1]-R [- 1] [yr] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [- 1] [yr]+weight1*delta,
The predicted value for being predicted pixel PixelPred2 be R [- 1] [yr]+weight2*delta, wherein weight1 by
PixelPred1 and PixelU vertical range determines that weight2 is determined by PixelPred2 and PixelU vertical range, when
When PixelPred1 and PixelU vertical range is more than PixelPred2 and PixelU vertical range, weight coefficient
Weight1 is less than weight coefficient weight2;
B. the second class intraprediction unit, in the current block at least 2 be predicted pixel PixelPred1 and
PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the left of current block
The reference pixel rebuild in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 on the upside of current block
The reference pixel rebuild in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [- 1] [yr]-R [xr] [- 1] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [xr] [- 1]+weight3*delta,
The predicted value for being predicted pixel PixelPred2 be R [xr] [- 1]+weight4*delta, wherein weight3 by
PixelPred1 and PixelL horizontal range determines that weight4 is determined by PixelPred2 and PixelL horizontal range, when
When PixelPred1 and PixelL horizontal range is more than PixelPred2 and PixelL horizontal range, weight coefficient
Weight3 is less than weight coefficient weight4.
Further, described device, also resides in one below:
A. described weight coefficient weight1 is by being predicted pixel PixelPred1 and its upside reference pixel PixelU
Vertical range obtained as power, i.e. weight1=Hconstantp, wherein p is PixelPred1 and its upside reference image
Plain PixelU vertical range, Hconstant are the index truth of a matter;Described weight coefficient weight2 is by being predicted pixel
What PixelPred2 and its upside reference pixel PixelU vertical range obtained as power, i.e. weight2=Hconstantq,
Wherein q is PixelPred2 and its upside reference pixel PixelU vertical range, and Hconstant is the index truth of a matter;
B. described weight coefficient weight3 is by being predicted pixel PixelPred1 and its left side reference pixel PixelL
Horizontal range obtained as power, i.e. weight3=Vconstantp, wherein p is PixelPred1 and its left side reference picture
Plain PixelL horizontal range, Vconstant are the index truth of a matter;Described weight coefficient weight4 is by being predicted pixel
What PixelPred2 and its left side reference pixel PixelL horizontal range obtained as power, i.e. weight4=Vconstantq,
Wherein q is PixelPred2 and its left side reference pixel PixelL horizontal range, and Vconstant is the index truth of a matter.
Further, described device, also resides in one below:
A. the prediction direction for being predicted pixel PixelPred1 uses and horizontal direction angle are less than or equal to 45 degree, institute
State weight coefficient weight1 index truth of a matter Hconstant and be predicted the prediction direction and water of pixel PixelPred1 uses
Flat angular separation is into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and horizontal direction angle are less than
Equal to 45 degree, the index truth of a matter Hconstant of the weight coefficient weight2 and be predicted pixel PixelPred2 use it is pre-
Direction and horizontal direction angle are surveyed into sine relation;
B. the prediction direction for being predicted pixel PixelPred1 uses and vertical direction angle are described to add less than 45 degree
Weight coefficient weight3 index truth of a matter Hconstant and the prediction direction and Vertical Square for being predicted pixel PixelPred1 uses
To angle into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and vertical direction angle are less than 45
Degree, the index truth of a matter Hconstant of the weight coefficient weight4 and the prediction direction for being predicted pixel PixelPred2 uses
With vertical direction angle into sine relation.
The present invention takes in two-way frame directional prediction mechanism to lift intraframe coding performance, compensate for for side
The defects of border discontinuity be present in the pre- error of measurement of the complex textures such as the texture of tropism roll-off characteristic and directional prediction.
Brief description of the drawings
Fig. 1 is one embodiment of the invention first kind bi-directional intra prediction device.
Fig. 2 is one embodiment of the invention the second class bi-directional intra prediction device.
Fig. 3 is one embodiment of the invention first kind bi-directional intra prediction device.
Fig. 4 is one embodiment of the invention the second class bi-directional intra prediction device.
Fig. 5 is one embodiment of the invention first kind bi-directional intra prediction device.
Fig. 6 is one embodiment of the invention the second class bi-directional intra prediction device.
Fig. 7 is the generating process of the prediction pixel of one embodiment of the invention first kind bi-predictive mode.
Fig. 8 is the generating process of the prediction pixel of one embodiment of the invention the second class bi-predictive mode.
Fig. 9 is HEVC of the embodiment of the present invention 35 kinds of intra prediction modes.
Figure 10 shows the different base coefficient of part first kind bi-predictive mode.
Figure 11 shows the different base coefficient of part the second class bi-predictive mode.
Embodiment
For a further understanding of the present invention, the preferred embodiment of the invention is described with reference to embodiment, still
It should be appreciated that these descriptions are simply further explanation the features and advantages of the present invention, rather than to the claims in the present invention
Limitation.
The embodiment of the present invention one:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 7 shows the generating process of the prediction pixel of first kind bi-predictive mode.Carried out specifically in conjunction with Fig. 7
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr]) × (1/y)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are that 1/y is weight coefficient weight with reference to the row coordinate for rebuilding pixel.Weight coefficient can by directly calculating or
Table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 8 shows the generating process of the prediction pixel of the second class bi-predictive mode.Carried out specifically in conjunction with Fig. 8
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1]) × (1/x)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are that 1/x is weight coefficient weight with reference to the row coordinate for rebuilding pixel.Weight coefficient can by directly calculating or
Table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention two:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 7 shows the generating process of the prediction pixel of first kind bi-predictive mode.Carried out specifically in conjunction with Fig. 7
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr])>>(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (2y+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 8 shows the generating process of the prediction pixel of the second class bi-predictive mode.Carried out specifically in conjunction with Fig. 8
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1])>>(x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (2x+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention three:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 7 shows the generating process of the prediction pixel of first kind bi-predictive mode.Carried out specifically in conjunction with Fig. 7
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr])>>(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (2y+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 8 shows the generating process of the prediction pixel of the second class bi-predictive mode.Carried out specifically in conjunction with Fig. 8
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1])/(3x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (3x+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention four:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 7 shows the generating process of the prediction pixel of first kind bi-predictive mode.Carried out specifically in conjunction with Fig. 7
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr]) × A(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(y+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, its value and prediction direction and horizontal direction angle G_H are into sine relation, i.e.,
A=w_h*sin (G_H)
Wherein w_h is weight coefficient.When w_h is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in table 1.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 8 shows the generating process of the prediction pixel of the second class bi-predictive mode.Carried out specifically in conjunction with Fig. 8
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1]) × A(x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(x+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, angle G_V between its value and prediction direction and vertical direction is into sine relation, i.e.,
A=w_v*sin (G_V)
Wherein w_v is weight coefficient.When w_v is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in table 2.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention five:
The present embodiment inventive technique can be applied to the encoding block or decoding block of all sizes, carry out directionality in two-way frame
Prediction.
Fig. 7 shows the generating process of the prediction pixel of first kind bi-predictive mode.Carried out specifically in conjunction with Fig. 7
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr]) × A(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(y+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, its value and prediction direction and horizontal direction angle G_H are into sine relation, i.e.,
A=w_h*sin (G_H)
Wherein w_h is weight coefficient.When w_h is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in table 1.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 8 shows the generating process of the prediction pixel of the second class bi-predictive mode.Carried out specifically in conjunction with Fig. 8
It is bright.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1]) × A(x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(x+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, angle G_V between its value and prediction direction and vertical direction is into sine relation, i.e.,
A=w_v*sin (G_V)
Wherein w_v is weight coefficient.When w_v is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in table 2.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention six:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 1 shows first kind bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 1 shown devices
It is bright.
In the step s 100, determine to rebuild in pixel on the left of current block according to intra prediction direction and be predicted pixel with described
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step s 102, rebuild on the upside of the current block determined according to intra prediction direction in pixel and be predicted picture with described
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step S104, delta=R [xr] [- 1]-R [- 1] [yr] is calculated.
In step s 106, weight coefficient is determined according to the row coordinate or row coordinate that are predicted pixel.
In step S108, it is R [- 1] [yr]+weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 7.Fig. 7 shows the prediction picture of first kind bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr]) × (1/y)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are that 1/y is weight coefficient weight with reference to the row coordinate for rebuilding pixel.Weight coefficient can by directly calculating or
Table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 2 shows the second class bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 2 shown devices
It is bright.
In step s 200, determine to rebuild in pixel on the left of current block according to intra prediction direction and be predicted pixel with described
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S202, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step S204, delta=R [- 1] [yr]-R [xr] [- 1] is calculated.
In step S206, weight coefficient is determined according to the row coordinate or row coordinate that are predicted pixel.
In step S208, it is R [xr] [- 1] weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 8.Fig. 8 shows the prediction picture of the second class bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1]) × (1/x)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are that 1/x is weight coefficient weight with reference to the row coordinate for rebuilding pixel.Weight coefficient can by directly calculating or
Table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention seven:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 3 shows first kind bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 3 shown devices
It is bright.
In step S300, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step s 302, rebuild on the upside of the current block determined according to intra prediction direction in pixel and be predicted picture with described
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step s 304, delta=R [xr] [- 1]-R [- 1] [yr] is calculated.
In step S306, weighting is determined according to the row coordinate or row coordinate that are predicted pixel and the index truth of a matter specified
Coefficient.
In step S308, it is R [- 1] [yr]+weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 7.Fig. 7 shows the prediction picture of first kind bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr])>>(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (2y+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 4 shows the second class bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 4 shown devices
It is bright.
In step S400, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S402, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step s 404, delta=R [- 1] [yr]-R [xr] [- 1] is calculated.
In step S406, weighting is determined according to the row coordinate or row coordinate that are predicted pixel and the index truth of a matter specified
Coefficient.
In step S408, it is R [xr] [- 1] weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 8.Fig. 8 shows the prediction picture of the second class bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1])>>(x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (2x+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention eight:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 3 shows first kind bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 3 shown devices
It is bright.
In step S300, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step s 302, rebuild on the upside of the current block determined according to intra prediction direction in pixel and be predicted picture with described
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step s 304, delta=R [xr] [- 1]-R [- 1] [yr] is calculated.
In step S306, weighting is determined according to the row coordinate or row coordinate that are predicted pixel and the index truth of a matter specified
Coefficient.
In step S308, it is R [- 1] [yr]+weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 7.Fig. 7 shows the prediction picture of first kind bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr])>>(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (2y+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 4 shows the second class bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 4 shown devices
It is bright.
In step S400, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S402, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step s 404, delta=R [- 1] [yr]-R [xr] [- 1] is calculated.
In step S406, weighting is determined according to the row coordinate or row coordinate that are predicted pixel and the index truth of a matter specified
Coefficient.
In step S408, it is R [xr] [- 1] weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 8.Fig. 8 shows the prediction picture of the second class bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1])/(3x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are to refer to the row coordinate for rebuilding pixel, 1/ (3x+1) it is weight coefficient weight.Weight coefficient can be by directly counting
Calculate or table look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention nine:
The present embodiment inventive technique can be applied to the different size of encoding block such as 4x4,4x8,8x8,8x16 and 16x16 or solution
Code block, carry out directional prediction in two-way frame.
Fig. 5 shows first kind bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 5 shown devices
It is bright.
In step S500, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S502, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step S504, delta=R [xr] [- 1]-R [- 1] [yr] is calculated.
In step S506, weight coefficient is determined according to the row coordinate or row coordinate and prediction direction that are predicted pixel.
In step S508, it is R [- 1] [yr]+weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 7.Fig. 7 shows the prediction picture of first kind bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr]) × A(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(y+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, its value and prediction direction and horizontal direction angle G_H are into sine relation, i.e.,
A=w_h*sin (G_H)
Wherein w_h is weight coefficient.When w_h is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in table 1.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 6 shows the second class bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 6 shown devices
It is bright.
In step S600, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S602, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step s 604, delta=R [- 1] [yr]-R [xr] [- 1] is calculated.
In step S606, weight coefficient is determined according to the row coordinate or row coordinate and prediction direction that are predicted pixel.
In step S608, it is R [xr] [- 1] weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 8.Fig. 8 shows the prediction picture of the second class bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1]) × A(x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(x+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, angle G_V between its value and prediction direction and vertical direction is into sine relation, i.e.,
A=w_v*sin (G_V)
Wherein w_v is weight coefficient.When w_v is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in table 2.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The embodiment of the present invention ten:
The present embodiment inventive technique can be applied to the encoding block or decoding block of all sizes, carry out directionality in two-way frame
Prediction.
Fig. 5 shows the first bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 5 shown devices
It is bright.
In step S500, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S502, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step S504, delta=R [xr] [- 1]-R [- 1] [yr] is calculated.
In step S506, weight coefficient is determined according to the row coordinate or row coordinate and prediction direction that are predicted pixel.
In step S508, it is R [- 1] [yr]+weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 7.Fig. 7 shows the prediction picture of first kind bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [- 1] [yr]+(R [xr] [- 1]-R [- 1] [yr]) × A(y+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(y+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, its value and prediction direction and horizontal direction angle G_H are into sine relation, i.e.,
A=w_h*sin (G_H)
Wherein w_h is weight coefficient.When w_h is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in Figure 10.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 7 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelU vertical range with PixelU vertical range, then the weight coefficient that PixelPred1 is used
Weight1 is less than the weight coefficient weight2 that PixelPred2 is used.
Fig. 6 shows the second class bi-directional intra prediction device.Each step is carried out specifically in conjunction with Fig. 6 shown devices
It is bright.
In step S600, determined to be predicted pixel with described in reconstruction pixel on the left of current block according to intra prediction direction
The value R [- 1] [yr] of reference pixel PixelL in intra prediction direction.
In step S602, picture is predicted with described in the current block upside reconstruction pixel determined according to intra prediction direction
Reference pixel PixelU of the element in intra prediction direction value R [xr] [- 1].
In step s 604, delta=R [- 1] [yr]-R [xr] [- 1] is calculated.
In step S606, weight coefficient is determined according to the row coordinate or row coordinate and prediction direction that are predicted pixel.
In step S608, it is R [xr] [- 1] weight*delta that the predicted value for being predicted pixel, which is calculated,.
Weight coefficient is specifically described with reference to Fig. 8.Fig. 8 shows the prediction picture of the second class bi-predictive mode
The generating process of element.
The coordinate for being predicted pixel in the present encoding block most upper left corner is (0,0).The row coordinate absolute value of current pixel is determined
Justice is current pixel and the horizontal range being predicted between pixel that coordinate is (0,0), and the row coordinate absolute value of current pixel is determined
The vertical range being predicted between pixel that justice is current pixel and coordinate is (0,0).If current pixel be located at coordinate for (0,
0) right side of prediction pixel, then the row coordinate of current pixel is just;If current pixel is located at the prediction picture that coordinate is (0,0)
The left side of element, then the row coordinate of current pixel is negative.If current pixel is located at the downside for the prediction pixel that coordinate is (0,0),
Then the row coordinate of current pixel is just;If current pixel is located at the upside for the prediction pixel that coordinate is (0,0), then current pixel
Row coordinate be negative.Coordinate is worked as the predicted value P [x] [y] for being predicted pixel of (x, y) by what is determined according to intra prediction direction
Preceding piece of left side rebuilds the value R [- 1] in pixel with the reference pixel PixelL in intra prediction direction that is predicted pixel
Rebuild on the upside of [yr] and the current block determined according to intra prediction direction in pixel with the pixel that is predicted in infra-frame prediction side
Upward reference pixel PixelU value R [xr] [- 1] is determined.Coordinate is the predicted value P [x] [y] for being predicted pixel of (x, y)
Calculation expression it is as follows:
P [x] [y]=R [xr] [- 1]+(R [- 1] [yr]-R [xr] [- 1]) × A(x+1)
Wherein y is the row coordinate for being predicted pixel, and x is the row coordinate for being predicted pixel, and xr is with reference to the row for rebuilding pixel
Coordinate, yr are with reference to the row coordinate for rebuilding pixel, A(x+1)It is weight coefficient weight.A is the different base of current predictive pixel
Coefficient, angle G_V between its value and prediction direction and vertical direction is into sine relation, i.e.,
A=w_v*sin (G_V)
Wherein w_v is weight coefficient.When w_v is 0.707, in different base coefficient A and HEVC one-direction prediction modes it
Between corresponding relation it is as shown in figure 11.Weight coefficient can be by directly calculating or tabling look-up or different accuracy approximation obtains.
Along the intra prediction direction, i.e., Fig. 8 be predicted pixel PixelPred1 and PixelPred2 is
PixelPred1 and PixelPred2 uses identical reference pixel PixelL and reference pixel PixelU, and PixelPred1
It is more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, then the weight coefficient that PixelPred1 is used
Weight3 is less than the weight coefficient weight4 that PixelPred2 is used.
In the case of pixel is predicted at 3 or more than 3 using the first kind or the second class intra-frame prediction method, if deposited
It is identical in above-mentioned 2 prediction directions for being predicted pixel being predicted in pixel, then described 2 weightings for being predicted pixel use
Coefficient will meet the relation between PixelPred1 and PixelPred2 weight coefficient in Fig. 7 or Fig. 8;For prediction direction not
Same is predicted pixel, and this, which is predicted pixel, can be respectively adopted the first kind or the second class intra-frame prediction method.
The explanation of above example is only intended to help the method and its core concept for understanding the present invention.It should be pointed out that pair
For those skilled in the art, under the premise without departing from the principles of the invention, the present invention can also be carried out
Some improvement and modification, these are improved and modification is also fallen into the protection domain of the claims in the present invention.
Claims (6)
1. a kind of intra-frame prediction method, it is characterised in that the prediction for being predicted pixel of current block is generated using one of following methods
Value:
A. in the current block at least 2 be predicted pixel PixelPred1 and PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the left of current block
Reference pixel in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the upside of current block
Reference pixel in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [xr] [- 1]-R [- 1] [yr] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [- 1] [yr]+weight1*delta, it is described
The predicted value for being predicted pixel PixelPred2 is R [- 1] [yr]+weight2*delta, and wherein weight1 is by PixelPred1
Determine that weight2 is determined by PixelPred2 and PixelU vertical range, works as PixelPred1 with PixelU vertical range
When being more than PixelPred2 and PixelU vertical range with PixelU vertical range, weight coefficient weight1 is less than weighting
Coefficient weight2;
B. in the current block at least 2 be predicted pixel PixelPred1 and PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the left of current block
Reference pixel in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the upside of current block
Reference pixel in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [- 1] [yr]-R [xr] [- 1] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [xr] [- 1]+weight3*delta, it is described
The predicted value for being predicted pixel PixelPred2 is R [xr] [- 1]+weight4*delta, and wherein weight3 is by PixelPred1
Determine that weight4 is determined by PixelPred2 and PixelL horizontal range, works as PixelPred1 with PixelL horizontal range
When being more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, weight coefficient weight3 is less than weighting
Coefficient weight4.
2. according to the method for claim 1, it is further characterized in that one below:
A. described weight coefficient weight1 is by being predicted hanging down for pixel PixelPred1 and its upside reference pixel PixelU
What straight distance obtained as power, i.e. weight1=Hconstantp, wherein p is PixelPred1 and its upside reference pixel
PixelU vertical range, Hconstant are the index truth of a matter;Described weight coefficient weight2 is by being predicted pixel
What PixelPred2 and its upside reference pixel PixelU vertical range obtained as power, i.e. weight2=Hconstantq,
Wherein q is PixelPred2 and its upside reference pixel PixelU vertical range, and Hconstant is the index truth of a matter;
B. described weight coefficient weight3 is the water by being predicted pixel PixelPred1 and its left side reference pixel PixelL
What flat distance obtained as power, i.e. weight3=Vconstantp, wherein p is PixelPred1 and its left side reference pixel
PixelL horizontal range, Vconstant are the index truth of a matter;Described weight coefficient weight4 is by being predicted pixel
What PixelPred2 and its left side reference pixel PixelL horizontal range obtained as power, i.e. weight4=Vconstantq,
Wherein q is PixelPred2 and its left side reference pixel PixelL horizontal range, and Vconstant is the index truth of a matter.
3. according to the method for claim 2, it is further characterized in that one below:
A. the prediction direction for being predicted pixel PixelPred1 uses and horizontal direction angle are described to add less than or equal to 45 degree
Weight coefficient weight1 index truth of a matter Hconstant and prediction direction and the level side for being predicted pixel PixelPred1 uses
To angle into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and horizontal direction angle are less than or equal to
45 degree, the index truth of a matter Hconstant of the weight coefficient weight2 and be predicted pixel PixelPred2 use prediction side
To with horizontal direction angle into sine relation;
B. the prediction direction for being predicted pixel PixelPred1 uses and vertical direction angle are less than 45 degree, the weighting system
Number weight3 index truth of a matter Hconstant and the prediction direction and vertical direction folder for being predicted pixel PixelPred1 uses
Angle is into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and vertical direction angle are less than 45 degree, institute
State weight coefficient weight4 index truth of a matter Hconstant and be predicted the prediction direction of pixel PixelPred2 uses with hanging down
Straight angular separation is into sine relation.
4. a kind of infra-frame prediction device, it is characterised in that the prediction for being predicted pixel of current block is generated using one of following device
Value:
A. first kind intraprediction unit, in the current block at least 2 be predicted pixel PixelPred1 and
PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the left of current block
Reference pixel in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the upside of current block
Reference pixel in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [xr] [- 1]-R [- 1] [yr] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [- 1] [yr]+weight1*delta, it is described
The predicted value for being predicted pixel PixelPred2 is R [- 1] [yr]+weight2*delta, and wherein weight1 is by PixelPred1
Determine that weight2 is determined by PixelPred2 and PixelU vertical range, works as PixelPred1 with PixelU vertical range
When being more than PixelPred2 and PixelU vertical range with PixelU vertical range, weight coefficient weight1 is less than weighting
Coefficient weight2;
B. the second class intraprediction unit, in the current block at least 2 be predicted pixel PixelPred1 and
PixelPred2,
I. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the left of current block
Reference pixel in pixel is PixelL, and the value of the left side reference pixel is designated as R [- 1] [yr];
Ii. along intra prediction direction determine described in be predicted pixel PixelPred1 and PixelPred2 and rebuild on the upside of current block
Reference pixel in pixel is PixelU, and the value of the upside reference pixel is R [xr] [- 1];
Iii. difference delta=R [- 1] [yr]-R [xr] [- 1] of left side reference pixel and upside reference pixel is calculated;
Iv. be calculated the predicted value for being predicted pixel PixelPred1 be R [xr] [- 1]+weight3*delta, it is described
The predicted value for being predicted pixel PixelPred2 is R [xr] [- 1]+weight4*delta, and wherein weight3 is by PixelPred1
Determine that weight4 is determined by PixelPred2 and PixelL horizontal range, works as PixelPred1 with PixelL horizontal range
When being more than PixelPred2 and PixelL horizontal range with PixelL horizontal range, weight coefficient weight3 is less than weighting
Coefficient weight4.
5. device according to claim 4, is further characterized in that one below:
A. described weight coefficient weight1 is by being predicted hanging down for pixel PixelPred1 and its upside reference pixel PixelU
What straight distance obtained as power, i.e. weight1=Hconstantp, wherein p is PixelPred1 and its upside reference pixel
PixelU vertical range, Hconstant are the index truth of a matter;Described weight coefficient weight2 is by being predicted pixel
What PixelPred2 and its upside reference pixel PixelU vertical range obtained as power, i.e. weight2=Hconstantq,
Wherein q is PixelPred2 and its upside reference pixel PixelU vertical range, and Hconstant is the index truth of a matter;
B. described weight coefficient weight3 is the water by being predicted pixel PixelPred1 and its left side reference pixel PixelL
What flat distance obtained as power, i.e. weight3=Vconstantp, wherein p is PixelPred1 and its left side reference pixel
PixelL horizontal range, Vconstant are the index truth of a matter;Described weight coefficient weight4 is by being predicted pixel
What PixelPred2 and its left side reference pixel PixelL horizontal range obtained as power, i.e. weight4=Vconstantq,
Wherein q is PixelPred2 and its left side reference pixel PixelL horizontal range, and Vconstant is the index truth of a matter.
6. device according to claim 5, is further characterized in that one below:
A. the prediction direction for being predicted pixel PixelPred1 uses and horizontal direction angle are described to add less than or equal to 45 degree
Weight coefficient weight1 index truth of a matter Hconstant and prediction direction and the level side for being predicted pixel PixelPred1 uses
To angle into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and horizontal direction angle are less than or equal to
45 degree, the index truth of a matter Hconstant of the weight coefficient weight2 and be predicted pixel PixelPred2 use prediction side
To with horizontal direction angle into sine relation;
B. the prediction direction for being predicted pixel PixelPred1 uses and vertical direction angle are less than 45 degree, the weighting system
Number weight3 index truth of a matter Hconstant and the prediction direction and vertical direction folder for being predicted pixel PixelPred1 uses
Angle is into sine relation;The prediction direction for being predicted pixel PixelPred2 uses and vertical direction angle are less than 45 degree, institute
State weight coefficient weight4 index truth of a matter Hconstant and be predicted the prediction direction of pixel PixelPred2 uses with hanging down
Straight angular separation is into sine relation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610712740.7A CN107786874A (en) | 2016-08-24 | 2016-08-24 | Directional prediction method and apparatus in two-way frame |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610712740.7A CN107786874A (en) | 2016-08-24 | 2016-08-24 | Directional prediction method and apparatus in two-way frame |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107786874A true CN107786874A (en) | 2018-03-09 |
Family
ID=61388744
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610712740.7A Pending CN107786874A (en) | 2016-08-24 | 2016-08-24 | Directional prediction method and apparatus in two-way frame |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107786874A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110072112A (en) * | 2019-03-12 | 2019-07-30 | 浙江大华技术股份有限公司 | Intra-frame prediction method, encoder and storage device |
CN110971911A (en) * | 2018-09-30 | 2020-04-07 | 北京三星通信技术研究有限公司 | Method and apparatus for intra prediction in video encoding and decoding |
CN112437313A (en) * | 2019-08-26 | 2021-03-02 | 浙江大学 | Intra-frame prediction pixel generation method and device |
RU2771669C1 (en) * | 2018-07-06 | 2022-05-11 | Мицубиси Электрик Корпорейшн | Video encoder, video decoder, method for video encoding, method for video decoding |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102484710A (en) * | 2009-07-29 | 2012-05-30 | 高通股份有限公司 | Method and device for pixel interpolation |
CN102695061A (en) * | 2011-03-20 | 2012-09-26 | 华为技术有限公司 | Method and apparatus for determining weight factors, and method and apparatus for predicting intra-frame weighting |
US20120300835A1 (en) * | 2007-04-17 | 2012-11-29 | Qualcomm Incorporated | Pixel-by-pixel weighting for intra-frame coding |
CN103597831A (en) * | 2011-06-20 | 2014-02-19 | 联发科技(新加坡)私人有限公司 | Method and apparatus of directional intra prediction |
CN103765901A (en) * | 2011-06-28 | 2014-04-30 | 三星电子株式会社 | Method and apparatus for image encoding and decoding using intra prediction |
CN103957421A (en) * | 2014-04-14 | 2014-07-30 | 上海大学 | HEVC coding size rapid determining method based on texture complexity |
CN104967852A (en) * | 2010-07-16 | 2015-10-07 | 三星电子株式会社 | Method and apparatus for encoding and decoding image through intra prediction |
CN105874797A (en) * | 2014-01-09 | 2016-08-17 | 高通股份有限公司 | Intra prediction from a predictive block |
-
2016
- 2016-08-24 CN CN201610712740.7A patent/CN107786874A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120300835A1 (en) * | 2007-04-17 | 2012-11-29 | Qualcomm Incorporated | Pixel-by-pixel weighting for intra-frame coding |
CN105430403A (en) * | 2007-04-17 | 2016-03-23 | 高通股份有限公司 | Pixel-by-pixel weighting for intra-frame coding |
CN102484710A (en) * | 2009-07-29 | 2012-05-30 | 高通股份有限公司 | Method and device for pixel interpolation |
CN104967852A (en) * | 2010-07-16 | 2015-10-07 | 三星电子株式会社 | Method and apparatus for encoding and decoding image through intra prediction |
CN102695061A (en) * | 2011-03-20 | 2012-09-26 | 华为技术有限公司 | Method and apparatus for determining weight factors, and method and apparatus for predicting intra-frame weighting |
CN103597831A (en) * | 2011-06-20 | 2014-02-19 | 联发科技(新加坡)私人有限公司 | Method and apparatus of directional intra prediction |
CN103765901A (en) * | 2011-06-28 | 2014-04-30 | 三星电子株式会社 | Method and apparatus for image encoding and decoding using intra prediction |
CN105874797A (en) * | 2014-01-09 | 2016-08-17 | 高通股份有限公司 | Intra prediction from a predictive block |
CN103957421A (en) * | 2014-04-14 | 2014-07-30 | 上海大学 | HEVC coding size rapid determining method based on texture complexity |
Non-Patent Citations (1)
Title |
---|
JIANLE CHEN等: ""Algorithm Description of Joint Exploration Test Model 1"", 《JOINT VIDEO EXPLORATION TEAM (JVET) OF ITU-T SG 16 WP 3 AND ISO/IEC JTC 1/SC 29/WG 11 1ND MEETING: GENEVA, CH, 19–21 OCTOBER 2015,JVET-A1001》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
RU2771669C1 (en) * | 2018-07-06 | 2022-05-11 | Мицубиси Электрик Корпорейшн | Video encoder, video decoder, method for video encoding, method for video decoding |
CN110971911A (en) * | 2018-09-30 | 2020-04-07 | 北京三星通信技术研究有限公司 | Method and apparatus for intra prediction in video encoding and decoding |
CN110971911B (en) * | 2018-09-30 | 2023-09-19 | 北京三星通信技术研究有限公司 | Method and apparatus for intra prediction in video coding and decoding |
CN110072112A (en) * | 2019-03-12 | 2019-07-30 | 浙江大华技术股份有限公司 | Intra-frame prediction method, encoder and storage device |
CN112437313A (en) * | 2019-08-26 | 2021-03-02 | 浙江大学 | Intra-frame prediction pixel generation method and device |
CN112437313B (en) * | 2019-08-26 | 2022-08-12 | 浙江大学 | Intra-frame prediction pixel generation method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104780366B (en) | Determine the intra prediction mode of image coding unit and image decoding unit | |
CN102648631B (en) | For the method and apparatus of coding/decoding high-definition picture | |
CN104871537B (en) | The method of infra-frame prediction between color | |
CN105959706B (en) | Image encoding device and method, and image decoding device and method | |
CN109792516A (en) | Method and device in image and coding and decoding video for coloration encoding and decoding in frame | |
CN103119945B (en) | The method and apparatus that image is coded and decoded by intra prediction | |
Zhang et al. | Gradient-based fast decision for intra prediction in HEVC | |
CA2634050C (en) | Video encoding method and decoding method, apparatuses therefor, programs therefor, and storage media for storing the programs | |
CN105208387B (en) | A kind of HEVC Adaptive Mode Selection Method for Intra-Prediction | |
MXPA05009250A (en) | Fast mode decision algorithm for intra prediction for advanced video coding. | |
JP6784836B2 (en) | Equipment and methods for video coding | |
CN103686165B (en) | Decoding method and Video Codec in depth image frame | |
IL229715A (en) | Method and apparatus of directional intra prediction | |
WO2008042127A2 (en) | Geometric intra prediction | |
CN107864380B (en) | 3D-HEVC fast intra-frame prediction decision method based on DCT | |
CN109845263A (en) | Apparatus for video encoding and method | |
CN105491390B (en) | Intra-frame prediction method in hybrid video coding standard | |
CN107786874A (en) | Directional prediction method and apparatus in two-way frame | |
CN108366256A (en) | A kind of HEVC intra prediction modes quickly select system and method | |
CN101820546A (en) | Intra-frame prediction method | |
CN107027025A (en) | A kind of light field image compression method based on grand block of pixels adaptive prediction | |
CN104853192A (en) | Prediction mode selection method and device | |
CN104333761B (en) | A kind of HEVC elementary cells level bit-rate distribution method | |
CN104284186A (en) | Fast algorithm suitable for HEVC standard intra-frame prediction mode judgment process | |
Xu et al. | Rate-distortion optimized scan for point cloud color compression |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180309 |