CN106534851B - Spatial domain scan method and device priority-based in video compress - Google Patents
Spatial domain scan method and device priority-based in video compress Download PDFInfo
- Publication number
- CN106534851B CN106534851B CN201611138887.6A CN201611138887A CN106534851B CN 106534851 B CN106534851 B CN 106534851B CN 201611138887 A CN201611138887 A CN 201611138887A CN 106534851 B CN106534851 B CN 106534851B
- Authority
- CN
- China
- Prior art keywords
- variance
- coefficient
- function
- variance function
- convert
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
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
- H04N19/10—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
- H04N19/12—Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
- H04N19/122—Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
-
- 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
- H04N19/13—Adaptive entropy coding, e.g. adaptive variable length coding [AVLC] or context adaptive binary arithmetic coding [CABAC]
-
- 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
- H04N19/59—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
-
- 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
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/61—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
-
- 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
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/625—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding using discrete cosine transform [DCT]
Abstract
The invention discloses spatial domain scan method and devices priority-based in a kind of video compress.Sweep object of the invention is equal residual coding transformation/transformation bypass coefficient.This method is ranked up scanning by the size of variance function.Variance evaluation method of the invention can make coefficient scanning meet: significantly coefficient is forward, coefficient is rearward by a small margin, in a large amount of 0 coefficient sets and the medium characteristic of tail portion collection, entropy coder can efficiently encode data using the good statistical property of scanning result of the present invention, to improve coding efficiency.
Description
Technical field
The present invention relates to video research area, in particular to spatial domain scanning side priority-based in a kind of video compress
Method and device.
Background technique
Video compression system, it is intended to by being analyzed raw video signal and being handled, with being compared to original signal
Less amount of bit retains the very important visual information in vision signal.One video compression system is divided into coding side and decoding end.
Raw video signal is converted bit stream by the coding side of video compression system.The decoding end of video compression system receives bit
Stream, also original video.If the video being reduced is remained exactly the same with original video, referred to as lossless video encoding.If by going back
There is distortion in former video, then be known as damaging Video coding compared with original signal.
Common video compression system coding side generally comprises following components: 1, piecemeal；2, prediction；3, it converts and measures
Change；4, scanning；5, entropy coding.Decoding end is corresponding to carry out inverse operation.As shown in Figure 1.
Forefathers' research is pointed out: when sweep object is DCT or ADST coefficient, scanning is close to optimal in a zigzag.It is led
Want the reason is as follows that: the export of DCT or ADST coefficient is based on " first-order autoregression Markov model ".First-order autoregression Ma Er
After the covariance matrix of section's husband's model carries out eigen decomposition, characteristic value represents parameter variance, that is to say, that whether coefficient is general
It is easy to take amplitude in rate.Being ranked up to characteristic value can export close to the path of scanning in a zigzag.According to parameter variance into
Row descending sort scanning, can coefficient be met: significantly coefficient be forward, by a small margin coefficient rearward, in a large amount of 0 coefficient sets with
And the medium characteristic of tail portion collection.This appears to be extraordinary statistical property in entropy coder.
However when being transformed to equal transformation/transformation bypass, the size of transformation coefficient is difficult to speculate in advance.Therefore, it scans
As a result usually more in disorder, the statistical law of scan fraction dies down, this is unfavorable for the efficient process of entropy coder.
Summary of the invention
1, the purpose of the present invention.
The present invention is difficult to push away in advance for the equal transformation in real transform/bypass transformation, transformation coefficient scanning sequency
The problem of survey, and propose a kind of scan method and device, better scan path is designed to improve coding efficiency.
2, the technical solution adopted in the present invention.
In the present invention, it proposes that a kind of method estimates the variance of equal transformation coefficient, and proposes according to variance size descending
Scan equal transformation coefficient.So that coefficient meet: significantly coefficient is forward, by a small margin coefficient rearward, in a large amount of 0 coefficient sets and
The medium characteristic of tail portion collection.
Spatial domain scan method priority-based in a kind of video compress proposed by the present invention: sweep object is equal residual
Poor transcoding, coding transform/transformation by-pass factor estimates coefficient relative size by variance function；It is ranked up and is swept according to the size of variance
It retouches.
Based on the preferred embodiment of the present invention, the size according to variance is ranked up as descending sort.
Based on the preferred embodiment of the present invention, the equal residual coding of the scanning converts/and to convert by-pass factor specific
Method is as follows:
Calculate directional prediction residual variance function or inter motion compensation prediction residual variance function in frame；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned.
Based on the preferred embodiment of the present invention, directionality variance function in the frame are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0.L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is edge
Coordinate where prediction pixel.
Based on the preferred embodiment of the present invention, the inter motion compensation variance function are as follows:
Wherein, K is constant,The reference fritter for being gradient squared magnitude function after the motion compensation of reference frame is identical
Value at coordinate.
The invention proposes a kind of using any video-frequency compression method of claim 1-3, in scanning step, scanning
Object be equal residual coding convert/convert by-pass factor, coefficient is estimated by variance function；It is arranged according to the size of variance
Sequence scanning.
Based on the preferred embodiment of the present invention, the size according to variance is ranked up as descending sort.
Based on the preferred embodiment of the present invention, the equal residual coding of the scanning converts/and to convert by-pass factor specific
Method is as follows:
Calculate directionality variance function or inter motion compensation variance function in frame；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned.
Based on the preferred embodiment of the present invention, directionality variance function in the frame are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is edge
Coordinate where prediction pixel.
Based on the preferred embodiment of the present invention, the inter motion compensation variance function are as follows:
Wherein, K is constant,The reference fritter for being gradient squared magnitude function after the motion compensation of reference frame is identical
Value at coordinate.
The invention proposes spatial domain scanning means priority-based in a kind of video compress, comprising:
For sweep object be equal residual coding convert/convert by-pass factor, pass through variance function and estimate that coefficient is estimated
Calculate device；
The sequence scanning means of scanning is ranked up for the size according to variance.
Based on the preferred embodiment of the present invention, the sequence scanning means is ranked up according to the size of variance as drop
Sequence sequence.
Based on the preferred embodiment of the present invention, the equal residual coding converts/converts by-pass factor and specifically includes:
Calculate the estimation device of directionality variance function or inter motion compensation variance function in frame；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned
Sequence scanning means.
Based on the preferred embodiment of the present invention, directionality variance function in frame in the estimation device are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is edge
Coordinate where prediction pixel.
Based on the preferred embodiment of the present invention, the inter motion compensation variance function are as follows:
Wherein, K is constant,The reference fritter for being gradient squared magnitude function after the motion compensation of reference frame is identical
Value at coordinate.
It is proposed by the present invention a kind of using any video compress device of claim 11-13, include the following steps:
In scanning step, sweep object be equal residual coding convert/convert by-pass factor, coefficient is estimated by variance function；According to
The size of variance is ranked up scanning.
Based on the preferred embodiment of the present invention, the size according to variance is ranked up as descending sort.
Based on the preferred embodiment of the present invention, the equal residual coding of the scanning converts/and to convert by-pass factor specific
Method is as follows:
Calculate directionality variance function or inter motion compensation variance function in frame；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned.
Based on the preferred embodiment of the present invention, directionality variance function in the frame are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is edge
Coordinate where prediction pixel.
Based on the preferred embodiment of the present invention, the inter motion compensation variance function are as follows:
Wherein, K is constant,The reference fritter for being gradient squared magnitude function after the motion compensation of reference frame is identical
Value at coordinate.
3, beneficial effects of the present invention.
Variance evaluation method proposed by the present invention may be directly applied to equal transformation/bypass transformation in real transform,
Direct coding is carried out to Pixel Information using variance function, realize that coefficient meets: significantly coefficient is forward, and coefficient leans on by a small margin
Afterwards, in a large amount of 0 coefficient sets and the medium characteristic of tail portion collection, entropy coder can be efficiently good to statistical property of the invention
Data are encoded, to improve coding efficiency.
Detailed description of the invention
Fig. 1 is Video coding and decoding system architecture schematic diagram.
Fig. 2 is the original piecemeal schematic diagram of video frame.
Fig. 3 is original block data schematic diagram.
Fig. 4 is vertical prediction piecemeal schematic diagram.
Fig. 5 is the predicted value schematic diagram obtained after Fig. 2 data boundary vertically replicates.
Fig. 6 be Fig. 2 in the predicted value error schematic diagram in Fig. 4.
Fig. 7 be Fig. 2 in the predicted value error information schematic diagram in Fig. 4.
Fig. 8 is zigzag coefficient scanning schematic diagram.
Fig. 9 is directional prediction schematic diagram in prediction inexactness model frame.
Figure 10 is standard deviation schematic diagram of the present invention.
Figure 11 is example criteria difference function values schematic diagram of the present invention.
Figure 12 is sample scan sequential schematic of the present invention.
Figure 13 is zigzag scanning result schematic diagram.
Figure 14 is the scanning result schematic diagram of this patent of the present invention.
Specific embodiment
Regular statistical data how is inputted, entropy coder is enable efficiently to compile to the good data of statistical property
Code, thus improve coding efficiency be the invention solves root problem.It is when scanning equal residual coding transformation/transformation bypass
When number, the size of transformation coefficient is difficult to speculate in advance.Therefore, scanning result is usually more in disorder, the statistical law of scan fraction
It dies down, this is unfavorable for the efficient process of entropy coder.The scanning problem of equal transformation coefficient, is a lack of one after all
A good method estimates the variance of coefficient.This is also that the present invention endeavours to solve the problems, such as.In the present invention, a kind of method is proposed
It estimates the variance of equal transformation coefficient, and proposes to scan equal transformation coefficient according to variance size descending.
Common video compression system coding atmosphere piecemeal, prediction, transformation and quantization, scanning, entropy coding.
Blocking step: a video includes many frames, and a video coding system respectively encodes each frame.It is compiling
When one frame of code, piecemeal is carried out to a video first.Then each fritter is separately encoded.Common partition strategy packet
Contain: fixed size piecemeal: such as JPEG system is using fixed 8x8 blocks of pixels.H.264 the system variable-size based on macro block point
Block strategy.Partition strategy of the HEVC system based on quaternary tree.And partition strategy of the MPEG4 system based on shape.
Piecemeal, which is conducive to video coding system, carries out adaptively, to reach ideal encoding efficiency video content.
Prediction steps: by after piecemeal, each fritter encodes video frame one by one.Exist in the Pixel Information of each fritter
Redundancy, this redundancy are very common in video.For example, there are similitudes for a fritter and surrounding fritter, a fritter is with before
There are similitudes for fritter in one frame video.In order to tentatively reduce data redundancy, the content of this fritter is carried out first pre-
It surveys, and prediction error is being encoded in next step.By well-designed prediction technique, can greatly reduce in each fritter
The data volume for including.Common prediction technique includes: directional intra prediction, DC component intra prediction, motion compensation interframe
Prediction etc..
In directional intra prediction, the border pixel values for the fritter being encoded are copied to along some direction to be worked as
Preceding fritter, the prediction data as the fritter.Fig. 2-7 shows the example of 4x4 pixel fritter height pattern prediction.?
In this example, there are bulk informations for initial data (Fig. 2 and Fig. 3).Due to hanging down in vertical prediction (Fig. 4 and Fig. 5) and fritter
Straight edge matches, so vertical prediction data and initial data are very close.Prediction error (Fig. 6 and Fig. 7) is compared to original
Beginning data are very small.In this example, the data volume encoded required for vertical prediction substantially reduces.
Shift step: previous step predicts that error, also known as prediction residual are further encoded.Residual coding is usually become
It changes, is converted to transformation coefficient.The purpose of transformation is to find one group of efficient residual pixel expression way, mathematically usually requires that change
It is orthonormal for changing.After transformation coefficient is quantified as integer, next step scanning is carried out.
Common converter technique has: discrete cosine transform (DCT) and asymmetric sine transform (ADST).Both become
There are floating number and quantization during changing, conversion process can introduce a small amount of video distortion, be typically used in and damage video compress
Another kind transformation is known as equal residual coding transformation/transformation bypass.Residual pixel is only done in this transformation to be quantified,
Or it is directly exported residual pixel as transformation coefficient.Since floating-point operation is not present in equal transformation, it can guarantee essence
Really retain residual error integer value, can be used for damaging video compress or lossless video compression.
Scanning step: the transformation coefficient of previous step is usually two-dimensional, in order to encode to two-dimensional transform coefficient, first
Need to be converted into one group of one-dimensional array.This process, which is referred to as, to be scanned.During scanning, it is desirable to effectively in view of becoming
The statistical property for changing coefficient assembles the similar coefficient of statistical property, so that next step entropy coding is more effective.
There are many kinds of existing scanning techniques.Including being directed to DCT, the zigzag of ADST coefficient scans (Zig-zag),
It is used in HEVC system and is directed to the adaptive scanning mode of prediction direction [3].
Fig. 8 illustrates the two-dimentional quantization parameter that one piece of 8x8 is scanned with zigzaggery.Blank indicates that coefficient is 0, scanning
Result afterwards is corresponding are as follows: 15,2,0,4,1,0,0,0 ... this one-dimensional data are admitted to carries out entropy coding in next step.
Entropy code step: the one-dimensional coefficient after scanned is sent into entropy coder and is converted to final binary stream.Entropy coding
It is studied by the statistical property to the one-dimensional data after scanning, to reduce statistical redundancy.
Statistical property of the design of entropy coder dependent on coefficient after scanning.For common coded system, utilize
To statistical property include: significantly coefficient is forward, and coefficient is rearward by a small margin；A large amount of 0 coefficient aggregations；Tail portion includes a large amount of 0
Coefficient etc..Usually, the scan fraction for being sent into entropy coder is regular, and the efficiency of entropy coder is higher.
Embodiment 1
Spatial domain scan method priority-based in video compress proposed by the present invention, specifically in accordance with the following steps:
Step 1: it is assumed that equal transformation coefficient coordinate is (m, n), current block coordinate range is 1≤m≤M, 1≤n≤N, block
Edge coordinate is m=0 or n=0.Then parameter variance can be estimated by prediction inexactness model [6].Specifically:
Situation one: when selecting directional prediction in frame, variance function is given by the following formula:
Each variable meaning in formula referring to FIG. 9, L be the coefficient to edge prediction pixel distance, α be prediction direction with
Edge normal angle, n ' are coordinate where edge prediction pixel.It is illustrated as the case where prediction pixel is derived from left edge.When prediction picture
When element is derived from top edge, the position of m and n is replaced.
Situation two: when selecting inter motion compensation, variance function is given by the following formula:
Wherein, K is constant,The reference fritter for being gradient squared magnitude function after the motion compensation of reference frame is identical
Value at coordinate.
Step 2: being directed toIt is ranked up from big to small, foundationSize sequence, to residual pixel/phase
Equal transformation coefficients are scanned.
Embodiment 2
Assuming that needing to be scanned the residual error coefficient of Fig. 6.
1, it estimates first
By prediction inexactness model formation,Function is as shown in Figure 10 and Figure 11: (having used when calculating gradient value flat
Equal difference)
2, byIt determines scanning sequency, it is ranked up, scanning sequency is obtained and is illustrated in fig. 12 shown below:
The residual error data of scanning figure 7 in this order, result after being scanned are as follows:
0.1,0.2,0.1,0.1,0.1,0,0,0 ... (rear full 0)
Under comparing, if scanned according to zigzag, obtained scanning result are as follows:
0, 0, 0, 0, 0.1, 0, 0, 0, 0.1, 0, 0.1, 0.1, 0, 0, 0.2, 0
Compare two groups of results, it is apparent that having following characteristic using the scanning result of the patent scanning technique:
A large amount of 0 coefficients concentrate on tail portion, and substantially value coefficient is forward, and small size value coefficient is rearward.This is for entropy coder right and wrong
It is often advantageous.
Real video signal testing result:
4x4 piecemeal, intra prediction is fixed in video standard test signal ice_qcif by us.By equal transformation series
Number is scanned using two methods.Method 1: it scans in a zigzag.Method 2: the method for this patent.
We study scan method for entirely testing the average effect of 4x4 fritter in signal.For this purpose, we are more each
Locate the amplitude summation of coefficient.One good scan method, the meeting on average is so that big coefficient is forward, and small coefficient is rearward.Cause
This, a good scan method should make scanning result rapid decrement.
Comparing Figure 13 and Figure 14, we should be apparent that.Under average, scan method of the invention to sweep
The statistical law for retouching result is very regular, and meets the needs of entropy coder.
Claims (12)
1. spatial domain scan method priority-based in a kind of video compress, it is characterised in that:
Sweep object be equal residual coding convert/convert by-pass factor, by variance function estimate coefficient relative size or
Order statistics calculate directional prediction variance function or inter motion compensation variance function in frame；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned；
Scanning is ranked up according to the size of variance；
The variance function of directional prediction residual error in the frame are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is side
Coordinate where edge prediction pixel.
2. spatial domain scan method priority-based in video compress according to claim 1, it is characterised in that: described
The size according to variance be ranked up as descending sort.
3. spatial domain scan method priority-based in video compress according to claim 1, it is characterised in that described
Inter motion compensation prediction residual variance function are as follows:
Wherein, K is constant,The reference fritter same coordinate for being gradient squared magnitude function after the motion compensation of reference frame
Locate value.
4. a kind of video-frequency compression method using any spatial domain scan method of claim 1-2, it is characterised in that:
In scanning step, sweep object be equal residual coding convert/convert by-pass factor, pass through variance function and estimate coefficient
Relative size or order statistics calculate directional prediction residual variance function or inter motion compensation residual variance letter in frame
Number；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned；
Scanning is ranked up according to the size of variance；
Directionality variance function in the frame are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is side
Coordinate where edge prediction pixel.
5. video-frequency compression method according to claim 4, it is characterised in that: described to be ranked up according to the size of variance
For descending sort.
6. video-frequency compression method according to claim 4, it is characterised in that the inter motion compensation variance function are as follows:
Wherein, K is constant,The reference fritter same coordinate for being gradient squared magnitude function after the motion compensation of reference frame
Locate value.
7. spatial domain scanning means priority-based in a kind of video compress, it is characterised in that:
For sweep object be equal residual coding convert/convert by-pass factor, pass through variance function estimate coefficient it is relatively large
Small or order statistics estimation devices, comprising:
Calculate frame in the estimation device of directionality variance function or inter motion compensation variance function, for variance function carry out from
Small sequence is arrived greatly, converts/convert the sequence scanning means that by-pass factor is scanned to equal residual coding；
The sequence scanning means of scanning is ranked up for the size according to variance；
Directionality variance function in frame in the estimation device are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is side
Coordinate where edge prediction pixel.
8. spatial domain scanning means priority-based in video compress according to claim 7, it is characterised in that: described
Sequence scanning means be ranked up according to the size of variance as descending sort.
9. spatial domain scanning means priority-based in video compress according to claim 7, it is characterised in that described
Inter motion compensation variance function are as follows:
Wherein, K is constant,The reference fritter same coordinate for being gradient squared magnitude function after the motion compensation of reference frame
Locate value.
10. a kind of video compress device using any spatial domain scanning means of claim 7-8, it is characterised in that:
In scanning step, sweep object be equal residual coding convert/convert by-pass factor, pass through variance function estimate coefficient it is opposite
Size or order statistics, the specific method is as follows:
Calculate directionality variance function or inter motion compensation variance function in frame；
It is sorted from large to small for variance function, converts/convert by-pass factor to equal residual coding and be scanned；According to
The size of variance is ranked up scanning；
Directionality variance function in the frame are as follows:
Wherein equal transformation coefficient coordinate is (m, n), and current block coordinate range is 1≤m≤M, 1≤n≤N, block edge seat
It is designated as m=0 or n=0, L is distance of the coefficient to edge prediction pixel, and α is prediction direction and edge normal angle, and n ' is side
Coordinate where edge prediction pixel.
11. video compress device according to claim 10, it is characterised in that: described to be arranged according to the size of variance
Sequence is descending sort.
12. video compress device according to claim 10, it is characterised in that the inter motion compensation variance function
Are as follows:
Wherein, K is constant,The reference fritter same coordinate for being gradient squared magnitude function after the motion compensation of reference frame
Locate value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611138887.6A CN106534851B (en) | 2016-12-12 | 2016-12-12 | Spatial domain scan method and device priority-based in video compress |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611138887.6A CN106534851B (en) | 2016-12-12 | 2016-12-12 | Spatial domain scan method and device priority-based in video compress |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106534851A CN106534851A (en) | 2017-03-22 |
| CN106534851B true CN106534851B (en) | 2019-07-09 |
Family
ID=58343090
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611138887.6A Active CN106534851B (en) | 2016-12-12 | 2016-12-12 | Spatial domain scan method and device priority-based in video compress |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106534851B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH05199418A (en) * | 1992-01-22 | 1993-08-06 | Oki Electric Ind Co Ltd | Method for scanning picture block data |
| JP2004312693A (en) * | 2003-03-24 | 2004-11-04 | Kitakyushu Foundation For The Advancement Of Industry Science & Technology | Image encoding apparatus and image encoding method |
-
2016
- 2016-12-12 CN CN201611138887.6A patent/CN106534851B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH05199418A (en) * | 1992-01-22 | 1993-08-06 | Oki Electric Ind Co Ltd | Method for scanning picture block data |
| JP2004312693A (en) * | 2003-03-24 | 2004-11-04 | Kitakyushu Foundation For The Advancement Of Industry Science & Technology | Image encoding apparatus and image encoding method |
Non-Patent Citations (1)
| Title |
|---|
| Scene Adaptive Coder;Wen-Hsiung Chen等;《IEEE Transactions on Communications》;19840331;第32卷(第3期);第225-232页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106534851A (en) | 2017-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104602003B (en) | Prediction block produces equipment | |
| CN101325707B (en) | System for encoding and decoding texture self-adaption video | |
| CN107343204B (en) | Utilize the method for decoding apparatus decoding video signal | |
| EP1796395A2 (en) | Method and device for intra prediction coding and decoding of images | |
| CN107277509B (en) | A kind of fast intra-frame predicting method based on screen content | |
| CN111355956B (en) | Deep learning-based rate distortion optimization rapid decision system and method in HEVC intra-frame coding | |
| EP2579598A2 (en) | Method for encoding/decoding high-resolution image and device for performing same | |
| US20070201751A1 (en) | Block-Based Fast Image Compression | |
| CN103081474A (en) | Intra-prediction decoding device | |
| CN1937775A (en) | Method and apparatus for image intraperdiction encoding/decoding | |
| CN103517069A (en) | HEVC intra-frame prediction quick mode selection method based on texture analysis | |
| CN103024383A (en) | Intra-frame lossless compression coding method based on HEVC (high efficiency video coding) frame | |
| CN106878729A (en) | Adaptive loop filter method and apparatus based on subregion basis | |
| CN108028941A (en) | Method and apparatus for coding and decoding digital picture by super-pixel | |
| CN106170093B (en) | Intra-frame prediction performance improving coding method | |
| CN104396247A (en) | Lcu-based adaptive loop filtering for video coding | |
| KR20110112224A (en) | Method and apparatus for encdoing/decoding information regarding encoding mode | |
| CN104199627B (en) | Gradable video encoding system based on multiple dimensioned online dictionary learning | |
| CN101155302B (en) | Video coding and decoding method based on image block data rotating and transforming | |
| CN103647975B (en) | Improved sample adaptive offset filtering method based on histogram analysis | |
| CN105306957A (en) | Adaptive loop filtering method and device | |
| CN105898300A (en) | Improved transformation coefficient sign data hiding method based on recovered transformation coefficient | |
| CN104702959B (en) | A kind of intra-frame prediction method and system of Video coding | |
| CN107682699B (en) | A kind of nearly Lossless Image Compression method | |
| CN106878754B (en) | A kind of 3D video depth image method for choosing frame inner forecast mode |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |