CN107040778A - Loop circuit filtering method and loop filter - Google Patents
Loop circuit filtering method and loop filter Download PDFInfo
- Publication number
- CN107040778A CN107040778A CN201610162343.7A CN201610162343A CN107040778A CN 107040778 A CN107040778 A CN 107040778A CN 201610162343 A CN201610162343 A CN 201610162343A CN 107040778 A CN107040778 A CN 107040778A
- Authority
- CN
- China
- Prior art keywords
- sample value
- image unit
- pixel
- loop filtering
- sao
- 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.)
- Withdrawn
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
- H04N19/117—Filters, e.g. for pre-processing or post-processing
-
- 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
- H04N19/174—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 slice, e.g. a line of blocks or a group of blocks
-
- 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
- H04N19/42—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation
-
- 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
- H04N19/70—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
-
- 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
- H04N19/80—Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation
- H04N19/82—Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation involving filtering within a prediction loop
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The present invention provides a kind of device or method of loop filtering processing.According to one or more sample value migration parameters, sample value migration processing is carried out to the pixel treated in current image unit through block elimination filtering.All or part of pixel in the sample value migration bound of parameter of current image unit shares same sample value migration parameter.The vertical sample value migration bound of parameter of current image unit by current image unit vertical edges displacement x s rows, and the horizontal sample value migration bound of parameter of current image unit shifts up ys rows by the horizontal sides of current image unit to the left.In order to reduce line buffer demand, xs is more than m, and m is pixel quantity of the block elimination filtering in each side amendment of horizontal end, and ys is more than or equal to 0.Loop circuit filtering method and loop filter proposed by the invention, can lift hardware efficiency.
Description
Technical field
The invention relates to video coding system, in particular to for combining sample value migration
(Sample Adaptive Offset, abridge SAO) and sample filtering compensation (Adaptive Loop Filter, contracting
Write ALF) video coding system of virtual boundary (virtual boundary), the present invention mends on reduction sample value skew
Repay the line buffer (line buffers) of (SAO) and sample filtering compensation (ALF).
Background technology
Estimation (motion estimation) is a kind of effective interframe encode (Inter-frame coding) technology,
To utilize the time redundancy information (temporal redundancy) in video sequence.Motion compensation (motion
Compensation interframe encode (Inter-frame coding)) has been widely used to encode in various International videos and marked
It is accurate.The estimation applied in various coding standards is often the motion letter used based on block (block)
Cease (motion information) (e.g., coding mode (coding mode) and motion vector (motion vector))
The each macro block (macroblock) of correspondence or similar block structure.In addition, intraframe coding (Intra-coding) also by
(adaptively applied) adaptively is applied, wherein the processing of picture is without reference to other pictures.Inter prediction
Or intra prediction (Inter-predicted or Intra-predicted residues) is generally further changed
(transformation) (quantization) and entropy code (entropy coding), are quantified, to produce regarding for compression
Frequency bit stream.In an encoding process (particularly in quantization program), code impairment (coding artifacts) occurs.
In order to reduce code impairment, newer coded system need to carry out extra processing on the video of reconstruction, to carry
Rise picture quality.Extra processing is standing to count into operation (in-loop operation) in loop, to cause encoder
Same reference picture, lifting system efficiency can be exported with decoder.
Figure 1A describe including between the adaptive frame that is handled in loop/schematic diagram of intraframe video coding system.
For inter prediction (Inter prediction), estimation (Motion Estimation, abridge ME)/fortune
Dynamic compensation (Motion Compensation, abridge MC) 112 is used to be based on other single pictures or multiple pictures
Video data there is provided prediction data.The selection infra-frame prediction of switch 114 (Intra Prediction) 110 or frame
Between prediction data, and selected prediction data forms predicated error (prediction supplied to adder 116
Error), also known as residual error (residues).Predicated error again by converter (Transformation, abridge T) 118 with
And subsequent quantizer (Quantization, abridge Q) 120 processing.Residual error after conversion and quantization is again by entropy
Encoder (Entropy Encoder) 122 is encoded, with the video bit stream for the video data for forming correspondence compression.
The related bit stream of conversion coefficient (transform coefficients) is further filled with auxiliary information (side
Information) ((motion), pattern (mode) or other information related to imagery zone e.g., are acted).
Auxiliary information can also be used for entropy code, to reduce bandwidth demand.Therefore, the related data of auxiliary information can be such as
Entropy coder 122 is supplied to shown in Figure 1A.During using inter-frame forecast mode, single reference picture or multiple ginsengs
Examining picture must also rebuild in encoder-side.Therefore, the residual error after changing and quantify transfers to inverse DCT
(Inverse Quantization, abridge IQ) 124 and decommutator (Inverse Transformation, abbreviation
IT) 126 processing, makes residual error restore (recover).Residual error after recovery can reconstructor (Reconstruction,
Abridge REC) add-back prediction data 136 at 128, to rebuild video data (reconstruct video data).Weight
The video data built can be stored in reference picture buffer (Refernce Picture Biffer) 134 and for predicting it
His frame.
As shown in Figure 1A, the video data of reception in coded system through a series of processing.From reconstructor
128 reconstruction video data may have a variety of damages (impairment) because of a series of processing.Therefore, rebuild
Video data before deposit reference picture buffer 134, be more subjected to processing in a variety of loops, lifting
Video quality.In the hair of efficient video coding (High Efficiency Video Coding, abridge HEVC) standard
Zhan Zhong, has developed de-blocking filter (Deblocking Filter, abridge DF) 130, sample value migration (Sample
Adaptive Offset, abridge SAO) 131 and sample filtering compensation (Adaptive Loop Filter, abbreviation
ALF) 132 picture quality is lifted.Wave filter in loop information (in-loop filter information) may be needed simultaneously
Enter in bit stream so that decoder can suitably restore information needed.Therefore, sample value migration (SAO) is come from
And the wave filter in loop information of sample filtering compensation (ALF) is supplied to entropy coder 122, to be contained in
In bit stream.In Figure 1A, de-blocking filter 130 is applied to reconstruction video (reconstructed video) first,
Sample value migration (SAO) 131 is applied to the video after block elimination filtering (DF) processing, sample filtering compensation again
(ALF) 132 the video after sample value migration (SAO) processing is applied to again.However, de-blocking filter (DF),
The order of sample value migration (SAO) and sample filtering compensation (ALF) is adjustable.System can shown in Figure 1A
Correspondence efficient video coding (HEVC) system (except sample filtering compensation (ALF)) or video encoding standard
The AVS2 video and audio coding standard of Chinese team development (this by).In efficient video coding (HEVC) development
Sample filtering compensation (ALF) had been assessed, but sample filtering compensation (ALF) is in current efficient video coding
(HEVC) not yet it is used.
Figure 1B is system block diagrams, its correspondence include de-blocking filter (DF), sample value migration (SAO) with
And the Video Decoder of sample filtering compensation (ALF).Because encoder may include the sheet for video data reconstruct
Ground decoder (local decoder), thus partial decoding of h device assembly (except entropy decoder 142) used compile
In code device.In addition, decoder end, which is still needed, asks motion compensation 144.Switch 146 selects inter prediction or frame in
Prediction, and selected prediction data is used for reconstructor (REC) 128, is combined with the residual error with recovery.Except right
Compressed video data performs entropy decoding, and entropy decoder 142 also is responsible for the entropy decoding of auxiliary information, for correspondence area
Block provides auxiliary information.For example, frame mode information is believed supplied to infra-frame prediction buffer 111, inter-frame mode
Breath is supplied to sample value supplied to motion compensation 144, adaptive offset information (adaptive offset information)
Migration (SAO) 131, sample filtering compensation information are supplied to sample filtering compensation (ALF) 132 and residual error
Supplied to inverse DCT (IQ) 124.Residual error is through inverse quantization (IQ) 124, inverse transform (IT) 126 and subsequent weight
Program processing is built, video data is reconstructed.Again, the reconstruction video counts that reconstructor (REC) 128 is provided
According to be by as shown in Figure 1B include inverse quantization (IQ) 124, a succession of processing of inverse transform (IT) 126 after,
There are strength offsets (intensity shift).Video data is rebuild further to be mended by deblocking (DF) 130, sample value skew
Repay (SAO) 131 and sample filtering compensation (ALF) 132 is handled.
The coded program of efficient video coding (HEVC) be according to maximum coding unit (Logic Coding Unit,
Abridge LCU, also known as code tree unit (Coding Tree Unit, abridge CTU)) implement.Maximum coding unit
(adaptively partitioned) multiple coding units are adaptively divided into using four points of trees (quadtree).Efficiently
In Video coding (HEVC), block elimination filtering (DF) acts on 8x8 block scopes.To each 8x8 blocks, across
The horizontal filtering of vertical block border is implemented first, then implements the vertical filtering across horizontal block border.Figure
2A illustrates block elimination filtering (DF) processing of efficient video coding (HEVC) brightness (luma) component.Block side
The both sides of boundary 210 need to respectively consider 4 boundary pixels.Border can correspond to vertical boundary or horizontal boundary.Border picture
Element is marked as q0, q1, q2 and q3 and p0, p1, p2 and p3.Q0 and two pixels of p0 are close to the side
Boundary.In the processing procedure on luma blocks border, four pixels of each side are derived for filtering parameter, and each side
Up to 3 pixels (that is, p0, p1, p2 or q0, q1, q2) filtered can be corrected.On across vertical block
The horizontal filtering on border, unfiltered reconstruction pixel is derived for filter parameter, and also serves as the original of filtering
Beginning pixel.On the vertical filtering across horizontal block border, the pixel of the treated centre of block elimination filtering (DF) is (i.e.,
Pixel after horizontal filtering) it is used for filter parameter derivation, and also serve as the original pixels of filtering.Efficient video
In block elimination filtering (DF) processing for encoding (HEVC) colourity (chroma) component, each side of block border is used to two
Individual boundary pixel, and an each side only pixel (that is, p0 or q0) is corrected.
Fig. 2 B describe the block elimination filtering (DF) of video encoding standard AVS2 luma components (luma component)
Processing, is directed to each side three of block border 220 boundary pixel.Boundary pixel is marked as q0, q1, q2
And p0, p1 and p2, two pixel q0 and p0 is close to the border.As for the deblocking filter on colourity block border
Ripple (DF) processing, two pixels of each side are derived for filter parameter.Correspondence video encoding standard AVS2,
Block elimination filtering (DF) processing can correct all boundary pixels being related to.In other words, the three of each side of block border
Individual luminance pixel and two chroma pixels can be adjusted.
The sample value migration made according to efficient video coding (HEVC) and video encoding standard AVS2
(SAO) type as shown in figure 3, have 0 degree of four kinds of sample value migration (SAO) types correspondence, 90 degree, 135
Degree and 45 degree of four kinds of sensings.Sample value migration (SAO) is made to filter (per-pixel in loop to each pixel
in-loop filtering).Sample value migration (SAO) parameter corresponds to each maximum coding unit (LCU) or coding
Tree unit (CTU) updates.The sensing type of correspondence sample value migration (SAO), implements pixel classifications first,
Such as according to the assortment of form 1, pixel is classified as multiple groups of (groups, also known as species (categories)
Or grade (classes)).After classification, each pixel rebuild after simultaneously block elimination filtering (DF) is based on selected sense(-)class
Type and classification results are compensated by a shift value.
Form 1
| Species | Situation |
| 1 | C<Two adjacent pixels |
| 2 | C<Another adjacent pixels of one Xiang Linxiangsu ＆＆C== |
| 3 | C>One neighbour's another adjacent pixels of Xiang Su ＆＆C== |
| 4 | C>Two adjacent pixels |
| 0 | It is not any of the above-described situation |
As shown in Table 1, the implementation of sample value migration (SAO) assortment can be by center pixel (C) and two
Individual adjacent pixel splitting ratio compared with.Assortment checks whether center pixel is more than, less than or equal to corresponding
Adjacent pixel.Three kinds of comparative results can use two bit (2-bit) data to represent.
Sample value migration (SAO) parameter (e.g., pixel-shift numerical value (pixel offset) and sample value migration
Type (SAO type)) each code tree unit (CTU) can be corresponded to and adaptively determined.For efficient video coding
(HEVC) for, sample value migration (SAO) bound of parameter (parameter boundary) and code tree unit
(CTU) it is identical.The shared same sample value of sample value migration (SAO) processing of all pixels in bound of parameter
Migration (SAO) type and shift value.Because sample value migration (SAO) is applied to block elimination filtering
(DF) pixel after, the sample value migration processing (SAO process) of present encoding tree unit (CTU) must be waited
It can just be carried out after the completion of to block elimination filtering (DF) processing of present encoding tree unit (CTU).However, until coding
By tree unit (CTU) border before the reconstruction video data of code tree unit (CTU) border opposite side is ready for,
Pixel around code tree unit (CTU) border can not make block elimination filtering (DF) processing.Based on such data-dependent
Property, video encoding standard AVS2 uses sample value migration (SAO) bound of parameter offset.Fig. 4 is described
According to the example of the skew of the bound of parameter of video encoding standard AVS2 sample value migration (SAO).For
For efficient video coding (HEVC), the embodiment 410 of corresponding sample value migration (SAO) bound of parameter
Correspondence code tree unit (CTU) border.For video encoding standard AVS2, corresponding sample value skew is mended
Repay (SAO) bound of parameter 420 and displacement x S and shift up yS to the left corresponding to code tree unit (CTU) border.
Furthermore, in video encoding standard AVS2, xS=4 and yS=4.
Sample filtering compensation (ALF) 132 is video coding tool, for lifting picture quality.Sample filtering is mended
Repay (ALF) and did assessment in efficient video coding (HEVC) developing stage.However, sample filtering compensation (ALF)
And be not used in current efficient video coding (HEVC) standard, but it has been incorporated into video encoding standard AVS2.
17 taps (17-tap) balanced sample for video encoding standard AVS2 particularly as shown in Figure 5, which is filtered, to be mended
Repay (ALF).Under the filtering operation of 17 tap balanced sample filtering compensations (ALF) meaning current pixel may need
Fang Sanhang (lines) data.If those rows (are particularly at subsequent from another code tree unit (CTU)
The code tree unit (CTU) of code tree cell row (CTU row)), sample filtering compensation (ALF) processing is necessary
Delay is until obtain follow-up related data.Features above means that line buffer (line buffer) need to be configured, with
The related data of instant memory present encoding tree unit (CTU), is used for subsequent treatment.In order to deal with data according to
The problem of relying, video encoding standard AVS2 uses sample filtering compensation (ALF) virtual boundary (virtual
Boundary), handled with limiting sample filtering compensation (ALF) not across the virtual boundary.Fig. 6 describes basis
The example of sample filtering compensation (ALF) virtual boundary of video encoding standard AVS2 luma component, wherein retouching
State on selecting pixel that (e.g., pixel a, b, c and sample filtering compensation (ALF) d) are handled.Row (line) 610
For code tree unit (CTU) border between code tree unit (CTU) X and code tree unit (CTU) Y.OK
620 be sample filtering compensation (ALF) virtual boundary (i.e. yC-4) of brightness, and it is according to video encoding standard AVS2
Specification and above code tree unit (CTU) border (i.e. yC) row of distance four (i.e. positioned at yC-4) part.Close
In chrominance components, (referred to according to video encoding standard AVS2：Information Technology–Advanced
Media Coding Part2:Video Final Committee Draft,Audio and Video Coding
Standard Workgroup of China,Feb.7,2015,Document:N2120.D3), sample filtering compensation
(ALF) virtual boundary is located above code tree unit (CTU) border in place of the row of distance 3 (i.e. positioned at yC-3).Picture
Plain a, b and c sample filtering compensation (ALF) processing are implemented in code tree unit (CTU) X processing stages.This
Outside, the information above virtual boundary is used only in pixel a, b and c sample filtering compensation (ALF) processing.Extremely
Pixel d below virtual boundary, the processing of its sample filtering compensation (ALF) is in code tree unit (CTU) Y
Processing stage implements, and using only the information below virtual boundary.The use of virtual boundary will suppress data according to
Lai Xing, reduces the capacity requirement of row buffering.
Such as above-mentioned content, de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF)
Program is related to adjacent data.In efficient video coding (HEVC) and video encoding standard (AVS2), coding
Tree unit (CTU) has been used as the unit of coded treatment.De-blocking filter (DF), sample value migration (SAO) with
And sample filtering compensation (ALF), when handling across code tree unit (CTU) border and implementing, data dependencies must quilt
Management with caution, to reduce the demand of line buffer.Because de-blocking filter (DF), sample value migration (SAO)
And sample filtering compensation (ALF) processing is implemented in each code tree unit (CTU), corresponding hardware practice in succession
It must be designed to Pipelining (pipeline) mode.Fig. 7 describe correspondence video encoding standard AVS2 decoders,
The data dependence of de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF) processing
Property.Using the processing sequence 700 of code tree unit (CTU) as shown in fig. 7, and code tree unit (CTU) X
And code tree unit (CTU) border between code tree unit (CTU) Y is marked as 705.As shown in fig. 7,
Reconstruction video from reconstruction square 710 is by de-blocking filter (DF) 720, sample value migration (SAO) 730
And sample filtering compensation (ALF) 740 is handled.The output of sample filtering compensation (ALF) 740, which is stored in, have been decoded
Frame buffer.
De-blocking filter (DF) 720, sample value migration (SAO) 730 and sample filtering compensation (ALF) 740
Handle corresponding processing state difference label 725,735 and 745.The code displaying tree unit of figure 725
(CTU) block elimination filtering (DF) processing state of the X block elimination filtering (DF) at the end of processing stage.The top of row 722
Luminance pixel and the chroma pixel of the top of row 724 completed block elimination filtering (DF).Because block border is another
Pixel (i.e. code tree unit (CTU) border is below 705) the still non-availability of side, the brightness picture of the lower section of row 722
Element and the chroma pixel of the lower section of row 724 can not handle rank in code tree unit (CTU) X block elimination filtering (DF)
It is processed in section.Figure 735 code displaying tree unit (CTU) X sample value migration (SAO) processing stage
At the end of sample value migration (SAO) processing state.The luminance pixel of the top of row 732 and the top of row 734
Chroma pixel completed sample value migration (SAO), wherein row 732 aligns with row 734.Figure 745 shows
Show sample filtering compensation of code tree unit (CTU) X sample filtering compensation (ALF) at the end of processing stage
(ALF) state is handled.Similarly, because the sample value skew for being directed to code tree unit (CTU) Y marks is mended
(SAO) parameter is repaid, it is not yet handled by length-changeable decoding (variable length decode, VLD) and completed,
The luminance pixel of the lower section of row 732 and the chroma pixel of the lower section of row 734 there is no method to make code tree unit (CTU) X
Sample value migration (SAO).Brightness above row 742 (sample filtering compensation (ALF) virtual boundary of brightness)
Pixel is based on video encoding standard AVS2 drafts and handled through sample filtering compensation (ALF).(the sample of colourity of row 744
This filtering compensation (ALF) virtual boundary) above chroma pixel can be handled through sample filtering compensation (ALF).However,
The sample filtering compensation (ALF) of chrominance components can not be in code tree unit (CTU) X processing stages to colourity row
A to D is performed.For example, sample filtering compensation (ALF) processing of pixel 746 will use pixel 748.Because
Chroma pixel 748 is located at the lower section of colourity sample value migration (SAO) border 734, and chroma pixel 748 is in coding
Tree unit (CTU) X processing stages not yet make sample value migration (SAO) processing.Therefore, although positioned at colourity
Sample filtering compensation (ALF) virtual boundary above, chroma pixel 746 can not still make sample filtering compensation (ALF)
Processing.Therefore, six rows of sample value migration (SAO) processing had been made in the top of pixel 748 (at once above D)
It must store to buffer, to make row A to D sample in code tree unit (CTU) Y processing stages later
Used during filtering compensation (ALF) processing.Wherein, above row A three rows at code tree unit (CTU) X
Reason rank made sample filtering compensation (ALF), it also requires being handled using the sample filtering compensation (ALF) for the A that is expert at.
On hardware is realized, the chroma samples of the row of picture width six must be stored in line buffer, typically using embedding
Enter the realization of formula memory, and so application may need high chip cost.Therefore, it is desirable to develop a kind of method
With device, (such as de-blocking filter (DF), sample value migration (SAO), sample are handled for reducing loop filtering
This filtering compensation (ALF) processing, other loop filterings processing or its combination) need line buffer quantity.This
Outside, for different sample value migration (SAO) bound of parameter, system is by different sample value migrations
(SAO) switch between parameter.System complexity and power consumption can thus be increased.Therefore, it is desirable to develop
Loop filtering processing (such as de-blocking filter (DF), sample value migration with appropriate System Parameter Design
(SAO), sample filtering compensation (ALF) handles other loop filtering programs or its combination), needed with reducing row buffering
Ask, system complexity and system power consumption or the combination benefit for obtaining any of the above improvement.In another viewpoint
In, this area is expected to develop a kind of method and apparatus, makees to include in efficient and cost-effective mode
Blocking filter (DF), sample value migration (SAO), sample filtering compensation (ALF), other loop filtering programs
Or the loop filtering processing of its combination, for combining the video coding system application that such loop filtering is handled.
The content of the invention
The present invention discloses the method and device of a kind of loop filtering processing for rebuild video data.In order to
Reduce the computation complexity of sample value migration (SAO) parameter switching and reduce line buffer demand, the present invention
Embodiment according to correspondence target horizontal direction and Vertical Square are made to sample value migration (SAO) bound of parameter
To displacement.According to the embodiment of the present invention, block elimination filtering (DF) processing is first implemented in reconstruction pixel.Go
Repair each side of the corresponding horizontal end in image unit border of block filtering (DF) processing between two image units
Just up to m pixel.According to one or more sample value migration (SAO) parameters, to current image unit
The middle pixel treated through block elimination filtering (DF) carries out sample value migration (SAO) processing.Current image unit
All or part of pixel in sample value migration (SAO) bound of parameter shares same said one or multiple
Sample value migration (SAO) parameter.Vertical sample value migration parameter (SAO) border of current image unit by
The vertical edges of current image unit displacement x s rows, and the horizontal sample value migration of current image unit to the left
(SAO) bound of parameter shifts up ys rows by the horizontal sides of current image unit.According to one or more spaces
Loop filter parameters limit sample value skew above border to hollow loop filter of current image unit
The treated pixel of compensation (SAO) makees space loop filtering process.The space loop filtering of current image unit
Device limits border and shifts up yv rows from the base of current image unit.In order to reduce line buffer demand, m,
Xs, ys and yv are positive integer, and xs is more than m, and ys is more than or equal to 0, and ys is less than yv, and yv roots
Set according to m.
Each image unit can correspond to code tree unit (CTU).Space loop filter processing can correspond to sample filtering
Compensate (ALF) processing.
If above-mentioned reconstruction video data report includes luma component and chrominance components, above-mentioned block elimination filtering processing, sample
It is worth migration processing and space loop filtering process point luma component and chrominance components is carried out, each makes
Above-mentioned m values are marked as M and N, and the above-mentioned xs each used is each used marked as xS and xSC
Above-mentioned ys marked as yS and ySC.And, the above-mentioned yv each used is marked as yV and yVC.One
Plant in embodiment, yS and ySC is equal to 0.YV can be more than M and yVC can be more than N.For example, yV
It is equal to (N+1) equal to (M+1) and yVC.In one embodiment, M is equal to 3 and N and is equal to 2.
In another embodiment, yS is equal to ySC, yV equal to yVC and yVC is more than MAX (M, N).
For example, yV and yVC is equal to MAX (M, N)+1 and yS and ySC can be 0 to MAX (M, N) integer.
In one embodiment, M is equal to 3 and N and is equal to 2.
In another embodiment, the current line handled in the currently processed stage of current image unit work as
Symbol data obtained by the comparison of the adjacent pixel of the adjacent lines handled in preceding pixel and subsequent processing stage is stored up
Leave and, above-mentioned symbol data correspondence " be more than ", " be less than " or " be equal to ".Above-mentioned symbol data can be used 2
Position (bit) storage.
Loop circuit filtering method and loop filter proposed by the invention, can lift hardware efficiency.
Brief description of the drawings
Figure 1A describes between adaptive frame/intraframe video coding system, wherein with reference to loop filter, it is described
Loop filter includes de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF)
Video data to processing reconstructed；
Figure 1B describes the system block diagrams of correspondence Video Decoder, and Video Decoder includes de-blocking filter
(DF), sample value migration (SAO) and sample filtering compensation (ALF)；
Fig. 2A is described makees block elimination filtering (DF) processing institute to luma component according to efficient video coding (HEVC)
The pixel for the block border both sides being related to；
Fig. 2 B are described according to involved by video encoding standard AVS2 makees block elimination filtering (DF) processing to luma component
And block border both sides pixel；
Fig. 3 describes the pixel classifications made based on 3x3 windows, there is four kinds of sensing forms, correspondingly 0 degree,
90 degree, 135 degree and 45 degree；
Fig. 4 describes sample value migration (SAO) boundary displacement according to video encoding standard AVS2；
Fig. 5 describes 17 taps (17-tap) balanced sample filtering compensation for video encoding standard AVS2
(ALF)；
Fig. 6 describes virtual according to the sample filtering compensation (ALF) of video encoding standard AVS2 luma component
Border；
Fig. 7 describes decoder and de-blocking filter (DF), the sample value for video encoding standard AVS2
Migration (SAO) and the related data dependency of sample filtering compensation (ALF) processing；
Fig. 8 describes de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF)
The processing state of processing, wherein the related boundary parameter of sign loop filter；
Fig. 9 is described has different horizontal sample value migration (SAO) parameters according to embodiment of the present invention
Border and sample value migration (SAO) processing on horizontal sample value migration (SAO) processing border；
Figure 10 is described to be tied according to present encoding tree unit (CTU) processing stage of first embodiment of the invention
The processing of de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF) processing during beam
State；
Figure 11 is described to be tied according to present encoding tree unit (CTU) processing stage of second embodiment of the invention
The processing of de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF) processing during beam
State；
Figure 12 describes the flow chart of the coded system including embodiment of the present invention, the system alignment loop
Wave filter relevant border, to reduce line buffer demand.
Embodiment
The following describe the better embodiment of the invention enumerated.Purpose described below is to introduce the present invention
Basic conception, is not intended to limit present invention.Protection scope of the present invention should be determined according to claim
Scope be defined.
Data dependency between the different loop processed stages is discussed for convenience, the application introduces boundary parameter
Related loop filter.Fig. 8 reproduces Fig. 7 de-blocking filter (Deblocking Filter, DF), sample value
Migration (Sample Adaptive Offset, SAO) and sample filtering compensation (Adaptive Loop Filter,
ALF) the relevant treatment state of processing, and by the related boundary parameter sign of loop filter wherein.Deblocking is filtered
The processing state difference label 825 of ripple device (DF), sample value migration (SAO) and sample filtering compensation (ALF),
835 and 845.Code displaying tree unit (CTU) X of figure 825 is in block elimination filtering (DF) processing stage end
Block elimination filtering (DF) treatment situation.Luminance pixel above row 822 (i.e. luma deblocking filter border) and
Chroma pixel above row 824 (i.e. chroma deblocking filters border) is handled through block elimination filtering (DF).By deblocking
The luminance pixel maximum quantity of each side of block border updated in filtering (DF) processing is M, and chroma pixel
Maximum quantity is N.In Fig. 8, code tree unit (CTU) boundary line 805 is also block border.
Code displaying tree unit (CTU) X of figure 835 is in the sample of sample value migration (SAO) processing stage end
It is worth migration (SAO) treatment situation.It is bright above row 832 (i.e. sample value migration (SAO) border of brightness)
The chroma pixel spent above pixel and row 834 (i.e. sample value migration (SAO) border of colourity) is inclined through sample value
Move compensation (SAO) processing.Sample value migration (SAO) the parameter side proposed based on video encoding standard AVS2
Boundary's shift technique, row 832 aligns with row 834.In order to avoid in the processing stage of each code tree unit (CTU)
Middle generation sample value migration (SAO) parameter switching, sample value migration (SAO) bound of parameter is in luma component
Displacement (xS, yS), and at chrominance components displacement (xSC, ySC).In other words, correspondence upper left side point for (xC,
YC code tree unit (CTU)), the top margin position of sample value migration (SAO) parameter of corresponding brightness component
The top margin for moving to (yC-yS) and sample value migration (SAO) parameter of correspondence chrominance components is moved to
(yC-ySC) (show as depicted in figure 8).Similarly, sample value migration (SAO) bound of parameter displacement is also embodied in
X directions.Originally, under the specification of efficient video coding (HEVC), sample value migration (SAO) parameter is according to volume
Code tree unit (CTU) is determined.In the prior art, in order to reduce the fortune of sample value migration (SAO) parameter switching
Complexity is calculated, sample value migration (SAO) parameter is based on after block elimination filtering (DF) processing and prepares to make sample value
Those pixels of migration (SAO) processing are determined.It is, the sample value migration (SAO) of prior art
Bound of parameter is equal to the picture for getting out block elimination filtering (DF) processing for making sample value migration (SAO) processing
Plain border.Get out make the pixel boundary of block elimination filtering (DF) processing of sample value migration (SAO) processing again
Claim sample value migration (SAO) processing border (processing boundary).
Code displaying tree unit (CTU) X of figure 845 is in the sample of sample filtering compensation (ALF) processing stage end
The treatment situation of this filtering compensation (ALF).On row 842 (i.e. sample filtering compensation (ALF) virtual boundary of brightness)
The luminance pixel of side is handled through sample filtering compensation (ALF).(the i.e. sample filtering compensation (ALF) of colourity of row 844
Virtual boundary) above chroma pixel can be handled through sample filtering compensation (ALF).However, colourity row D can not
Implement sample filtering compensation (ALF) processing of chrominance components in code tree unit (CTU) X processing stages.Correspondence
Sample filtering compensation (ALF) virtual boundary of luma component is (yC-yV), the sample filtering of correspondence chrominance components
It is (yC-yVC) to compensate (ALF) virtual boundary.Wherein, yV and yVC distinguish corresponding brightness and chrominance components
Border vertical displacement.For video encoding standard AVS2 drafts, by the brightness being updated and chrominance components
Boundary pixel quantity (i.e. M and N) be respectively 3 and 2.Brightness and the sample value migration of chrominance components
(SAO) bound of parameter vertical displacement is all set to 4.On the other hand, the sample filtering of brightness and chrominance components is mended
The vertical displacement (i.e. yV and yVC) for repaying (ALF) virtual boundary is set to 4 and 3.
In order to reduce line buffer size requirements simultaneously and reduce the sample value of code tree unit (CTU) processing stage
The computational complexity of migration (SAO) parameter switching, the application discloses a kind of sample value migration (SAO) ginseng
Number borders setting skills, according to each target, both horizontally and vertically displacement sample value migration (SAO)
Bound of parameter, and have different sample value migration (SAO) bound of parameter and sample value migration (SAO) place
Manage border.Such as foregoing teachings, in the prior art, sample value migration (SAO) bound of parameter and sample value skew
Compensate (SAO) processing border identical.According to the present invention, vertical sample value migration (SAO) bound of parameter is maintained
Border is handled equal to sample value migration (SAO), but horizontal sample value migration (SAO) bound of parameter can be different
Border is handled in sample value migration (SAO).Particularly, sample value migration (SAO) processing border is basis
The position of the treated pixel data of block elimination filtering is determined.Fig. 9 is described according to embodiment of the present invention
Unequal horizontal sample value migration (SAO) parameter and processing border.In one embodiment, level
Sample value migration (SAO) bound of parameter 912 equivalent code tree unit (CTU) horizontal boundary 910.The opposing party
Face, horizontal sample value migration (SAO) processing border 920, which is located at, removes the lastrow of block boundary 930.As for level
The pixel 940 of the top of sample value migration (SAO) processing border 920, the processing of its sample value migration (SAO)
It is the sample value of (i.e. its top) in sample value migration (SAO) bound of parameter based on code tree unit (CTU) X
Migration (SAO) parameter is carried out.However, according to present context, the pixel 950 on row D is also base
Sample value migration (SAO) processing is carried out in code tree unit (CTU) X sample value migration (SAO) parameter,
Because pixel 950 is located within horizontal sample value migration (SAO) bound of parameter 912 (i.e. its top).Tradition
In sample value migration (SAO) processing, the pixel 950 on row D is the sample according to code tree unit (CTU) Y
It is worth migration (SAO) parameter and carries out sample value migration (SAO).
In described above, an image is divided into multiple code tree units, and each code tree dividing elements are
One or more coding units (CUs).De-blocking filter (DF), sample value migration (SAO) and sample filter
Ripple compensation (ALF) processing is implemented in block border, is reduced in block border or neighbouring pseudomorphism (artifacts).
For the coded system that wherein code tree unit (CTUs) is dealt with horizontal raster scan order, code tree
Unit (CTU, also known as block scope) is the de-blocking filter (DF) on border, sample value migration (SAO) and
Sample filtering compensation (ALF) processing will need line buffer to store across code tree cell row (CTU row) border
Information.However, image, which can be also divided, makees other image units (such as macro block (macroblock) or ceramic tile shape (tile))
To make coded treatment.(image unit boundary) equally faces code tree unit (CTU) on image unit border
The line buffer subject under discussion that border faces.
Above-mentioned sample filtering compensation (ALF) is only an example, and the present invention can be applicable to any space loop filter
Ripple device (spatial loop filter).For example, the two-dimensional finite impulse with one group of space loop filter parameters rings
(finite impulse response, abridge FIR) wave filter is answered to can be used for substitution sample filtering compensation (ALF).For
The line buffer demand of reduction space loop filtering process, the space loop filtering boundary that limits can be used for limiting
Loop filtering process in space processed, which is used only, to be limited in the loop filtering boundary of space through sample value migration (SAO) place
The data managed.For example, the space loop filter border limited can be located on code tree unit (CTU) border
Side is at y rows.Space loop filter is implemented on above the space loop filter border limited through sample value
The treated pixel of migration (SAO), and will can only use warp above the space loop filter border limited
The treated pixel of sample value migration (SAO) as space loop filter input.
Fig. 9 describes one embodiment of the present invention, the horizontal nothing of sample value migration (SAO) bound of parameter 912
Must be identical with code tree unit (CTU) horizontal boundary 910.According to the embodiment of the present invention, horizontal sample value is inclined
Moving compensation (SAO) bound of parameter 912 can design in code tree unit (CTU) horizontal boundary 910 to horizontal sample
It is worth any row in migration (SAO) the processing next line of border 920.For code tree unit (CTU) X reclaimed waters
The pixel of the flat lower section of sample value migration (SAO) processing border 920, because the sample value migration of area's pixel
(SAO) operation is sample value migration (SAO) parameter information being based in code tree unit (CPU) X, therefore sample value
Sample value migration (SAO) parameter information needed for migration (SAO) operation needs buffer storage.However,
In efficient video coding (HEVC) and video encoding standard AVS2, sample value migration (SAO) parameter can be compiled
Code is merging (merge-up) upwards or merges (merge-left) grammer (syntax) to the left, it is also desirable to which buffer storage is whole
Sample value migration (SAO) parameter of individual code tree cell row (CTU row), therefore the sample value migration of storage
(SAO) parameter can be shared.
If the video data correspondence color video data in processing, this case technology can be applicable to luma component and color
Spend component.In the first embodiment, the vertical displacement yS of horizontal sample value migration (SAO) bound of parameter
Vertical displacement yV and yVC with ySC and sample filtering compensation (ALF) virtual boundary is with formula (1) and (2)
Determine：
0≤yS<YV=M+1, and (1)
0≤ySC<YVC=N+1 (2)
The main function of line buffer demand arises from a code tree cell row to next code tree cell row
The storage requirements of border loop filtering process.Because an image may be very wide, corresponding line buffer size
May be very big.Therefore, it is an object of the present invention to reduce to encode between two code tree cell rows (CTU row)
Set the line buffer demand of the loop filtering processing on unit (CTU) border.If the boundary displacement of correspondence vertical boundary
Have an impact to line buffer demand, it is also very small.Brightness and sample value migration (SAO) parameter of chrominance components
Border horizontal displacement xS and xSC maintain (that is, xS=M+1 and xSC=N+1) identical with traditional approach.
When system is with vertical scanning sequential processes picture, code tree unit column (CTU column) processing mode is with coding
Set unit row (CTU row).
Figure 10 is described according to the embodiment of the present invention, wherein M=3, N=2, yS=3, ySC=2, yV=4
And yVC=3.De-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF) processing
Processing state respectively in Figure 10 labels 1025,1035 and 1045.Shown in loop processed border and Fig. 8
Content part is identical.In Figure 10, represent that those loop processed borders are identical with Fig. 8 Ref. No. persons.Figure 10
Indicate that the sample value migration (SAO) for brightness and chrominance components handles border, the sample value skew of luma component
The vertical displacement (yS ') for compensating (SAO) processing is 4, and the sample value migrations (SAO) of chrominance components processing is hung down
Straight displacement (ySC ') is 3.As shown in Figure 10, sample value migration (SAO) parameter of brightness and chrominance components
Border 1032 and 1034 is in the lower section of corresponding sample value migration (SAO) processing border 1036 and 1038
A line.
The displacement of sample value migration (SAO) bound of parameter horizontal direction is equal to sample value migration (SAO) processing
Border.For example, sample value migration (SAO) the bound of parameter horizontal displacement xS of brightness and chrominance components with
And xSC is identical with the displacement that sample value migration (SAO) handles border horizontal direction.
In second of embodiment, sample value migration (SAO) bound of parameter vertical displacement yS and ySC
And sample filtering compensation (ALF) virtual boundary vertical displacement yV and yVC are determined according to formula (3)：
0≤yS=ySC<YV=yVC=MAX (M, N)+1 (3)
In other words, sample value migration (SAO) bound of parameter of brightness and chrominance components is identical, makes internal memory
Access behavior is consistent.Sample filtering compensation (ALF) virtual boundary of brightness and chrominance components is also identical.In addition,
Sample filtering compensation (ALF) virtual boundary is located above sample value migration (SAO) bound of parameter at least a line.
Figure 11 is described according to the embodiment of the present invention, wherein M=3, N=2, yS=ySC=2, yV=VC=4.
The processing state of de-blocking filter (DF), sample value migration (SAO) and sample filtering compensation (ALF) processing
Respectively in Figure 11 labels 1125,1135 and 1145.Loop processed border and the phase of content part shown in Fig. 8
Together.In Figure 11, identical loop processed border is represented with Fig. 8 identical Ref. No. persons.Figure 11 is also retouched
State and handle border for the sample value migration (SAO) of brightness and chrominance components, luma component and chrominance components
Sample value migration (SAO) processing border vertical displacement (i.e. marked as yS ' and ySC ') it is 4.As shown in figure 11,
Brightness and sample value migration (SAO) bound of parameter 1132 of chrominance components and 1134 are in corresponding sample
It is worth in place of migration (SAO) the processing row of 1136 and 1138 lower section distance of border two.Further indicated in figure
Brightness and sample filtering compensation (ALF) virtual boundary 1142 and 1044 of chrominance components.
Sample value migration (SAO) bound of parameter of horizontal direction is equal to the sample value migration of horizontal direction
(SAO) border is handled.For example, sample value migration (SAO) bound of parameter level of brightness and chrominance components
Displacement x S and xSC may be set to xS=xSC=MAX (M, N)+1.
In sample value migration (SAO) processing, the pixel count in present encoding tree unit (CTU) processing border
According to available for sample value migration (SAO) processing later.For example, Fig. 9 rows D is in code tree unit
(CTU) Y pipeline stages (pipeline stage) processing.Row D sample value migration (SAO) processing will need row
C data；The row C data is sample value migration (SAO) the processing rank in code tree unit (CTU) X
Handled in section.Such as foregoing teachings, sample value migration (SAO) classification is handled center pixel data and two
Other adjacent pixel is compared.Comparative result judges whether the center pixel is more than, less than or equal to selection
Adjacent pixel.Accordingly, sample value migration (SAO) operation module can be pre-processed between row C and row D
Compare behavior, and store its result (that is, ">”、”<" or "=").Because unjustified sample value migration (SAO)
Bound of parameter and processing border can cause row D to go together C using same sample value migration (SAO) ginseng
Number makees sample value migration (SAO) processing, pipeline stages processing pair of the comparative result in code tree unit (CTU) X
Row C makees to obtain during sample value migration (SAO) processing.
Comparative result between pixel column C and adjacent lines of pixels D can one by one pixel by two bit data representations,
To indicate one of three kinds of comparative results.Compare the complete pixel storage of generally 8 or more positions, described two
The data volume of bit symbol data is relatively low.Accordingly, line buffer cost can be downgraded greatly.
Loop filtering processing boundary design disclosed above can be used for solving to include de-blocking filter (DF), sample value
Migration (SAO) and the video system (such as video encoding standard AVS2) of sample filtering compensation (ALF) processing
Data dependence situation trigger substantial amounts of line buffer demand.The present invention can also be implemented in any including deblocking
The high-order Video coding of wave filter (DF), sample value migration (SAO) and sample filtering compensation (ALF) processing
System.
Form 2 compares the line buffer demand of video encoding standard AVS2 and the embodiment of the present invention.It is such as foregoing
Content, all of above embodiment respectively needs three rows storage data to brightness and chrominance components, as deblocking filter
Ripple (DF).To sample value migration (SAO) processing, all systems are required for brightness and chrominance components stored row
C and row D data.However, the storage of row C pixel datas is substituted, can be by between stored row C and row D
Comparative result reduce storage requirements.Such as foregoing teachings, each comparative result only needs to be stored with dibit.
According to conventional video coding standard AVS2, storage is made colourity group by sample value migration (SAO) result of six rows
Sample filtering compensation (ALF) processing divided.System including any embodiment of the invention can solve slow with six rows
Rush the demand of device storage chrominance components sample filtering compensation (ALF) processing.Correspondence conventional video coding standard
AVS2, first embodiment and second embodiment, de-blocking filter (DF), sample value migration (SAO) with
And the line buffer sum required for sample filtering compensation (ALF) is respectively 16,7.5 and 8.5.By storing sample
The symbol for the comparative result that value migration (SAO) processing is related to, two embodiments all reach that extra storage is empty
Between simplify effect.In other words, first and second embodiment can reduce 8.5 and 7.5 line buffers respectively.
In the example of table 2, conventional video coding standard AVS2 sample filtering compensation (ALF) colourity virtual boundary is
YC-3, sample value migration (SAO) displacement amount of pixels (being expressed by parameter SAO_SHIFT_PIX_NUM) is 4；
The content of first embodiment corresponding diagram 9；And the content of second embodiment corresponding diagram 10.
Form 2
Figure 12 describes the flow chart of the video coding system including embodiment of the present invention, the video system
System's alignment loop filter relevant border, to reduce line buffer demand.Step 1210, video system is received
The reconstruction video data of one image unit of correspondence.The video data of reconstruction may be derived from memory (for example, calculating
Machine memory, buffer (random access memory (RAM) or DRAM (DRAM) or other matchmakers
Body), or it is derived from processor.Block elimination filtering (DF) processing is then performed in reconstruction pixel in step 1220.Deblocking
Image unit border each side of corresponding horizontal boundary between filtering (DF) processing two image units of amendment is up to
M pixel.Step 1230, based on one or more sample value migration (SAO) parameters, sample value skew is mended
Repay (SAO) processing and be implemented in the treated pixel of block elimination filtering (DF) in current image unit.Current image unit
Sample value migration (SAO) bound of parameter in all or part of pixel share same sample value migration
(SAO) parameter.Vertical sample value migration (SAO) bound of parameter of current image unit is from current image unit
Vertical boundary displacement x s rows, and horizontal sample value migration (SAO) the parameter side of current image unit to the left
Boundary shifts up ys rows from the horizontal boundary of current image unit.Step 1240, according to one or more spaces
Loop filter parameters, it is inclined through sample value above border to limiting for the space loop filter of current image unit
Move compensation (SAO) treated pixel and implement space loop filtering process, wherein from the base of current image unit
Shift up yV rows and obtain the space loop filter of the current image unit and limit border.In order to reduce
Line buffer demand and/or reduce loop filtering process switching, loop filter relevant border according to positive integer m,
Xs, ys and yv are set, and xs is more than m, and ys is more than or equal to 0 and ys and is less than yv, and yv is according to m
Setting is (as shown in step 1250).
Flow chart illustrated above is intended to illustrate the example that loop filtering is handled according to the present invention.This area
Technical staff can change each step, rearrange sequence of steps, segmentation step, or combine step to implement
Spirit Essence of the present invention without departing from the present invention.In the disclosure, specific syntax and semantics have been used to
Bright embodiment.Those skilled in the art can not depart from the spirit of the present invention by equivalence replacement syntax and semantics
Premise implements the present invention.
It is described above to enable those skilled in the art real according to the interior text and condition of disclosed application-specific
The existing present invention.This case technical staff, which works as, can change above example content in many ways, and this specification is fixed
The rule possible application of justice is in other embodiments.Therefore, this case, which is not intended to be defined in the above, shows
With the specific embodiment of description, and described principle should be met and novel feature authorizes widest scope.With
In upper detailed description, the description of a variety of details is used to help thoroughly understand this case.Those skilled in the art
It should be understood that the present invention can be put into practice.
The embodiment of invention as described above can by various hardware, software code, or both combination realize.
For example, an embodiments of the invention, which can be one or more electronic circuits, is integrated into a video pressure
Shrinking chip or program code are integrated into video compression software to perform processing routine described herein.The present invention
An embodiment can also be program code, by digital signal processor (DSP) perform, to perform sheet
The processing routine of text description.It is fast that this case may also be related to a number of function side, by calculator processor,
Digital signal processor, microprocessor or field effect FPGA array (FPGA) are performed.These processing
Device can be configured to carry out particular task according to this case, determine wherein performing machine-readable software code or performing
The adopted firmware code by presently embodied specific method.Software code or firmware code can be by different volumes
Cheng Yuyan and the exploitation of different forms or pattern.The software code is compiled also for different target platforms.
However, perform the different coded formats of task according to this case technology, style and software code language and other
Code device would not depart from the spirit and scope of the present invention.
In the case where not departing from spirit or essential characteristics of the present invention, the present invention can be implemented in other specific forms.
Description example is considered as only illustrating in all respects and not restricted.Therefore, model of the invention
Enclose and indicated by claims, rather than it is previously mentioned.In all methods and scope equivalent in claim
Change belongs to the covering scope of the present invention.
Claims (26)
1. a kind of method of loop filtering processing, it is characterised in that methods described is entered for video coding system
Row rebuilds video data, and above-mentioned reconstruction video data is divided into multiple image units, and methods described includes：
Receive the reconstruction video data of image unit；
To rebuild pixel implement block elimination filtering processing, wherein, block elimination filtering processing two image units it
Between the corresponding horizontal end in image unit border each side correct up to m pixel；
According to one or more sample value migration parameters, to being handled in current image unit through block elimination filtering
The pixel crossed carries out the sample value migration parameter side of sample value migration processing, wherein current image unit
All or part of pixel in boundary shares same said one or multiple sample value migration parameters, wherein
The vertical sample value migration bound of parameter of current image unit by current image unit vertical boundary to the left
Displacement x s rows, and current image unit horizontal sample value migration bound of parameter by current image unit
Horizontal boundary shifts up ys rows；And
According to one or more space loop filter parameters to hollow loop filter of current image unit
Limit the pixel that sample value migration has been treated above border and make space loop filtering process, wherein when
The space loop filter of preceding image unit limits border and shifts up yv from the base of current image unit
OK；
Wherein, m, xs, ys and yv are positive integer, and xs is more than m, and ys is more than or equal to 0, and ys
Less than yv, and yv is set according to m.
2. the method for loop filtering processing according to claim 1, it is characterised in that each shadow
As unit one code tree unit of correspondence.
3. the method for loop filtering processing according to claim 1, it is characterised in that above-mentioned space
Loop filtering processing correspondence adaptive loop filter processing.
4. the method for loop filtering processing according to claim 1, it is characterised in that above-mentioned reconstruction
Video data is included at luma component and chrominance components, above-mentioned block elimination filtering processing, sample value migration
Reason and space loop filtering process point luma component and chrominance components are carried out, the above-mentioned m each used
Value is marked as M and N, and the above-mentioned xs each used is marked as xS and xSC, the above-mentioned ys each used
Marked as yS and ySC, and the above-mentioned yv each used is marked as yV and yVC.
5. the method method of loop filtering according to claim 4 processing, it is characterised in that yS with
YSC is equal to 0.
6. the method for loop filtering processing according to claim 4, it is characterised in that yV is more than
M and yVC are more than N.
7. the method for loop filtering processing according to claim 6, it is characterised in that yV is equal to
(M+1) and yVC be equal to (N+1).
8. the method for loop filtering processing according to claim 7, it is characterised in that M is equal to 3
And N is equal to 2.
9. the method for loop filtering processing according to claim 4, it is characterised in that yS is equal to
YSC, yV are equal to yVC, and yVC is more than MAX (M, N).
10. the method for loop filtering processing according to claim 9, it is characterised in that yV and yVC
Equal to MAX (M, N)+1.
11. the method for loop filtering processing according to claim 10, it is characterised in that M is equal to
3 and N is equal to 2.
12. the method for loop filtering processing according to claim 1, it is characterised in that further
By the current pixel of the current line handled in the currently processed stage of current image unit and subsequent processing rank
Symbol data obtained by the comparison of the adjacent pixel of the adjacent lines handled in section is stored, above-mentioned symbolic number
According to corresponding to " being more than ", " being less than " or " being equal to ".
13. the method for loop filtering processing according to claim 12, it is characterised in that above-mentioned symbol
Number is stored with two bits.
14. a kind of device of loop filtering processing, it is characterised in that described device is entered for video coding system
Row rebuilds video data, and above-mentioned reconstruction video data is divided into multiple code tree units, and described device bag
Include：
One or more electronic circuits, are coupled to line buffer, and be configured as：
Receive the reconstruction video data of image unit；
To rebuild pixel implement block elimination filtering processing, wherein, block elimination filtering processing two image units it
Between the corresponding horizontal end in image unit border each side correct up to m pixel；
According to one or more sample value migration parameters, to being handled in current image unit through block elimination filtering
The pixel crossed carries out the sample value migration parameter side of sample value migration processing, wherein current image unit
All or part of pixel in boundary shares same said one or multiple sample value migration parameters, wherein
The vertical sample value migration bound of parameter of current image unit by current image unit vertical boundary to the left
Displacement x s rows, and current image unit horizontal sample value migration bound of parameter by current image unit
Horizontal boundary shifts up ys rows；And
According to one or more space loop filter parameters to hollow loop filter of current image unit
Limit the pixel that sample value migration has been treated above border and make space loop filtering process, wherein when
The space loop filter of preceding image unit limits border and shifts up yv from the base of current image unit
OK；
Wherein, m, xs, ys and yv are positive integer, and xs is more than m, and ys is more than or equal to 0 and small
In yv, and yv is set according to m.
15. the device of loop filtering processing according to claim 14, it is characterised in that each
Image unit one code tree unit of correspondence.
16. the device of loop filtering processing according to claim 14, it is characterised in that above-mentioned sky
Between loop filtering processing correspondence adaptive loop filter processing.
17. the device of loop filtering processing according to claim 14, it is characterised in that above-mentioned heavy
Building video data report includes luma component and chrominance components, and above-mentioned block elimination filtering processing, sample value skew are mended
Repay processing and space loop filtering process point luma component and chrominance components are carried out, what is each used is upper
M values are stated marked as M and N, the above-mentioned xs each used is marked as xS and xSC, and what is each used is upper
Ys is stated marked as yS and ySC, and the above-mentioned yv each used is marked as yV and yVC.
18. the device of loop filtering according to claim 17 processing, it is characterised in that yS with
YSC is equal to 0.
19. the device of loop filtering processing according to claim 17, it is characterised in that yV is big
It is more than N in M and yVC.
20. the device of loop filtering processing according to claim 19, it is characterised in that yV etc.
It is equal to (N+1) in (M+1) and yVC.
21. the device of loop filtering processing according to claim 20, it is characterised in that M is equal to
3 and N is equal to 2.
22. the device of loop filtering processing according to claim 17, it is characterised in that yS is equal to
YSC, yV are equal to yVC and yVC is more than MAX (M, N).
23. the device of loop filtering according to claim 22 processing, it is characterised in that yV with
YVC is equal to MAX (M, N)+1.
24. the device of loop filtering processing according to claim 23, it is characterised in that M is equal to
3 and N is equal to 2.
25. the device of loop filtering processing according to claim 14, it is characterised in that current shadow
As unit the currently processed stage in the current pixel of current line that handles with being handled in subsequent processing stage
Adjacent lines adjacent pixel comparison obtained by symbol data be stored, above-mentioned symbol data correspondence "
More than ", " be less than " or " be equal to ".
26. the device of loop filtering processing according to claim 25, it is characterised in that above-mentioned symbol
Number is stored with two bits.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US15/015,552 US20160241881A1 (en) | 2015-02-13 | 2016-02-04 | Method and Apparatus of Loop Filters for Efficient Hardware Implementation |
| US15/015,552 | 2016-02-04 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107040778A true CN107040778A (en) | 2017-08-11 |
Family
ID=59532170
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610162343.7A Withdrawn CN107040778A (en) | 2016-02-04 | 2016-03-22 | Loop circuit filtering method and loop filter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107040778A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020043192A1 (en) * | 2018-08-31 | 2020-03-05 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries in video coding |
| US11089335B2 (en) | 2019-01-14 | 2021-08-10 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries |
| US11094088B2 (en) | 2018-08-31 | 2021-08-17 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries in video coding |
| CN113489984A (en) * | 2021-05-25 | 2021-10-08 | 杭州博雅鸿图视频技术有限公司 | Sample adaptive compensation method and device of AVS3, electronic equipment and storage medium |
| US11765349B2 (en) | 2018-08-31 | 2023-09-19 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103238318A (en) * | 2010-11-02 | 2013-08-07 | 联发科技股份有限公司 | Method and apparatus of slice boundary filtering for high efficiency video coding |
| US20130322523A1 (en) * | 2011-05-10 | 2013-12-05 | Mediatek Inc. | Method and apparatus for reduction of in-loop filter buffer |
| CN103843350A (en) * | 2011-10-14 | 2014-06-04 | 联发科技股份有限公司 | Method and apparatus for loop filtering |
| CN103891277A (en) * | 2011-10-14 | 2014-06-25 | 联发科技股份有限公司 | Method and apparatus for loop filtering |
| CN104067245A (en) * | 2012-08-03 | 2014-09-24 | 联发科技股份有限公司 | Video processing system with shared/configurable in-loop filter data buffer architecture and related video processing method thereof |
| CN104584561A (en) * | 2012-08-06 | 2015-04-29 | 联发科技股份有限公司 | Method and apparatus for sample adaptive offset in a video decoder |
| CN104823446A (en) * | 2012-12-06 | 2015-08-05 | 索尼公司 | Image processing device, image processing method, and program |
-
2016
- 2016-03-22 CN CN201610162343.7A patent/CN107040778A/en not_active Withdrawn
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103238318A (en) * | 2010-11-02 | 2013-08-07 | 联发科技股份有限公司 | Method and apparatus of slice boundary filtering for high efficiency video coding |
| US20130322523A1 (en) * | 2011-05-10 | 2013-12-05 | Mediatek Inc. | Method and apparatus for reduction of in-loop filter buffer |
| CN103843350A (en) * | 2011-10-14 | 2014-06-04 | 联发科技股份有限公司 | Method and apparatus for loop filtering |
| CN103891277A (en) * | 2011-10-14 | 2014-06-25 | 联发科技股份有限公司 | Method and apparatus for loop filtering |
| CN104067245A (en) * | 2012-08-03 | 2014-09-24 | 联发科技股份有限公司 | Video processing system with shared/configurable in-loop filter data buffer architecture and related video processing method thereof |
| CN104584561A (en) * | 2012-08-06 | 2015-04-29 | 联发科技股份有限公司 | Method and apparatus for sample adaptive offset in a video decoder |
| CN104823446A (en) * | 2012-12-06 | 2015-08-05 | 索尼公司 | Image processing device, image processing method, and program |
Non-Patent Citations (1)
| Title |
|---|
| ESENLIK S ET AL: "《Grid displacements for in-loop filtering》", 《JOINT COLLABORATIVE TEAM ON VIDEO CODING (JCT-VC)》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020043192A1 (en) * | 2018-08-31 | 2020-03-05 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries in video coding |
| US11094088B2 (en) | 2018-08-31 | 2021-08-17 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries in video coding |
| US11765349B2 (en) | 2018-08-31 | 2023-09-19 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries |
| US11089335B2 (en) | 2019-01-14 | 2021-08-10 | Mediatek Inc. | Method and apparatus of in-loop filtering for virtual boundaries |
| CN113489984A (en) * | 2021-05-25 | 2021-10-08 | 杭州博雅鸿图视频技术有限公司 | Sample adaptive compensation method and device of AVS3, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105898335B (en) | Promote the loop circuit filtering method and loop filter of hardware efficiency | |
| EP2708027B1 (en) | Method and apparatus for reduction of in-loop filter buffer | |
| EP3057320A1 (en) | Method and apparatus of loop filters for efficient hardware implementation | |
| TWI738092B (en) | Method and apparatus for reduction of in-loop filter buffer | |
| CN103947208B (en) | Reduce the method and device of deblocking filter | |
| CN103891277B (en) | Loop filter method and its device | |
| CN104185989B (en) | For the improvement infra-frame prediction in lossless coding in the binaryzation scheme and HEVC of intra prediction | |
| US20060133504A1 (en) | Deblocking filters for performing horizontal and vertical filtering of video data simultaneously and methods of operating the same | |
| KR102359415B1 (en) | Interpolation filter for inter prediction apparatus and method for video coding | |
| CN107040778A (en) | Loop circuit filtering method and loop filter | |
| TWI737137B (en) | Method and apparatus for non-linear adaptive loop filtering in video coding | |
| CN107172424A (en) | Loop circuit filtering method and its equipment | |
| CN103051892B (en) | Embedded loop filter method and embedded loop filter | |
| CN105306957B (en) | Adaptive loop filter method and apparatus | |
| CN104521234B (en) | Merge the method for processing video frequency and device for going block processes and sampling adaptive migration processing | |
| WO2022116113A1 (en) | Intra-frame prediction method and device, decoder, and encoder | |
| CN103491372A (en) | Filtering method of blocking-removing filter suitable for HEVC standard | |
| KR20050121627A (en) | Filtering method of audio-visual codec and filtering apparatus thereof | |
| JP7005480B2 (en) | Image decoder, image coding device, program and image processing system | |
| CN110720216B (en) | Apparatus and method for video encoding | |
| TW202034699A (en) | Image coding device, image decoding device, image coding method, image decoding method, and program |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20170811 |