CN103888782B - Parallel task partitioning method for HEVC decoder - Google Patents
Parallel task partitioning method for HEVC decoder Download PDFInfo
- Publication number
- CN103888782B CN103888782B CN201410077419.7A CN201410077419A CN103888782B CN 103888782 B CN103888782 B CN 103888782B CN 201410077419 A CN201410077419 A CN 201410077419A CN 103888782 B CN103888782 B CN 103888782B
- Authority
- CN
- China
- Prior art keywords
- decoding
- unit
- hevc decoder
- parallel task
- code tree
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 84
- 238000000638 solvent extraction Methods 0.000 title abstract 3
- 230000008569 process Effects 0.000 claims abstract description 43
- 238000012545 processing Methods 0.000 claims abstract description 18
- 230000001360 synchronised effect Effects 0.000 claims abstract description 9
- 238000013139 quantization Methods 0.000 claims abstract description 8
- 230000009466 transformation Effects 0.000 claims abstract description 8
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 230000003139 buffering effect Effects 0.000 claims description 6
- 230000008030 elimination Effects 0.000 claims description 6
- 238000003379 elimination reaction Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 5
- 238000013404 process transfer Methods 0.000 claims description 2
- 239000013589 supplement Substances 0.000 claims 1
- 238000012546 transfer Methods 0.000 claims 1
- 238000013461 design Methods 0.000 description 7
- 238000012360 testing method Methods 0.000 description 6
- 238000001514 detection method Methods 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 241001209435 Actus Species 0.000 description 1
- 241000796385 Choreutis chi Species 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000011056 performance test Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Abstract
The invention discloses a parallel task partitioning method for an HEVC decoder. According to the method, processing of a network abstraction layer unit, decoding of a parameter set and header information and management work of a decoding image buffer area are executed by one control unit, entropy coding, inverse quantization, inverse transformation, intra-frame prediction, inter-frame prediction and deblocking filter required by a decoding and coding tree unit and pixel self-adaptive compensation are executed by a plurality of work units, and a steamline type work mode is adopted in the control unit and the work units. The parallel task partitioning method for the HEVC decoder can ensure load balance better and eliminate dependence on a lock structure in the synchronous process, so that the computing resource utilization rate is improved, and synchronous overhead is saved.
Description
Technical field
The present invention relates to a kind of decoder implementation method in video decoding techniques field, specifically one kind are directed to HEVC standard
Decoder parallel task division methods.
Background technology
In order to better adapt to ultra high-definition video traffic, International Telecommunication Union and International Organization for Standardization combine formulated new
Generation video compression standard HEVC.For HEVC decoder, the huge data processing needs that ultra high-definition video brings are
Exceed the disposal ability of existing monokaryon CPU, this proposes stern challenge to decoding real-time.
At present, the best mode solving the above problems is parallel processing technique.Paralell design needs are carried out to decoder
Solve the problems, such as two classes, that is, task divides and synchronous.The first kind of way that task divides is that function divides, will be in HEVC decoder
Including each functions such as entropy decoding, inverse quantization, inverse transformation, infra-frame prediction, inter prediction, block elimination filtering and pixel adaptive equalizations
Module assignment is on different processor units；The second way that task divides is that data divides, will picture frame to be decoded
Data spatially carries out region division according to certain mode, then the decoding task of zones of different is distributed to one group of process
Device completes parallel.Decoder synchronization work purpose be constraint one group of concurrent working TU task unit between execution sequence it is ensured that
The correctness of decoding process.
So far, the traveling wave flowing water of the most efficient HEVC decoding architecture eclipsed form of industrial quarters and academia announcement
Method for parallel processing, is in periodical Journal of Signal Processing Systems, Volume by authors such as C.Chi
Entitled Parallel HEVC Decoding on Multi-and Many-core in 71, Issue 3, pp 247-260
Deliver in the paper of Architectures.The method can be summarized as follows：For the decoding process being unable to parallel processing,
Distribution scanning and two processing units of output.Scan process unit is mainly responsible for detection, is extracted and process network abstract layer list
Unit, decoding parametric collection categorical data, obtain band header and data to be decoded is drawn for unit according to code tree cell row
Being divided into several can be that current decoded frame distributes memory area simultaneously in time in decoding image buffering area with the unit of parallel processing
Release buffer area.The image that on the one hand output processing unit needs to be responsible for complete decoding is put into slow according to DISPLAY ORDER sequence
Rush in area, on the other hand need release in time to have shown that the memory space shared by image of output.
For can be with the process of parallel processing, the method can be corresponding simultaneously according to the sum distribution of computer available resources
Row evaluation work unit.Above-mentioned working cell is responsible for the row in units of code tree cell row, specifying according to traveling wave flowing water method
Between data dependence relation carry out parallel decoding.It should be noted that the traveling wave flowing water method for parallel processing of eclipsed form is in compatibility
On the premise of HEVC standard code stream, traveling wave flowing water method is expanded in time domain, and obtain bigger can be parallel
Property.The improved principle of time domain is：If can not continuing to provide for working cell in current decoded frame and needing the code tree of decoding
During cell row, decoder just can start the decoding process to next two field picture.It is single that this mechanism can efficiently reduce calculating
Unit's idle time because of the task of wait, it is favorably improved overall decoding performance.
Although the traveling wave flowing water method for parallel processing of eclipsed form obtains preferable real-time decoding speed, it in performance still
There is the leeway that can be lifted.By analysis it has been found that for scan process unit and exporting processing unit corresponding function distribution two
Individual parallel computation unit has following both sides design drawback：
First drawback is that resource utilization is low.Generally, the task of scanning work unit mainly includes solving harbour letter
Memory area in breath and application, release display image relief area, and the task of output services unit is mainly included to display figure
As the data in buffer area is ranked up and discharges.
The computation complexity of our analytic solution terminal information first.The document delivered in F.Bossen et al.
HEVCcomplexity and implementation analysis.IEEE Transaction on Circuits and
Systems for Video Technology 2012:The document that 22 (12), pp1685-1696. and J.Vanne et al. deliver
Comparative Rate-Distortion-Complexity Analysis of HEVC and AVC Video
Codecs.IEEE Transaction on Circuits and Systems for Video Technology2012.22
(12) in .1885-1898, comprehensive test has been carried out to the complexity of HEVC decoder.By the meter to each functional module of decoder
Calculate complexity analyzing result and understand, entropy decoding part accounts for the 5% to 6% of decoder overall complexity, and solves terminal information and made
Columbus's coding/decoding method complexity is relatively low and data volume is less for global entropy solution code process, during this partial decoding of h
Between only account for the very little part of entropy decoding, account for and overall decode less than the 1% of the time.Even if remaining entropy decoding and prediction process of reconstruction
Workload makes the task amount being assigned to each computing unit decrease, for common 8 cores because of paralell design
Or for the processor of 4 cores, each processor at least will undertake 90.4%/8=11.3% or 90.4%/4=22.6%
Computation complexity.And this complexity is apparently higher than the complexity of solution terminal information.
Next the computation complexity of further analysis management relief area.Management work master for display image relief area
The following aspects to be included：1) it is current decoding video frame application memory space in the buffer；2) due to decoding order and
It is possible to difference, the frame of video of decoding and rebuilding needs before output in time sequencing enterprising rearrangement sequence between DISPLAY ORDER；
3) release has shown that output and is not used again as the image of reference frame.In these work, decoding frame being carried out reorder can
To complete in linear session (such as chain structure) by efficient data structure；And in display image relief area application or
The time consuming required for release memory space is also less.Divided according to the test result that F.Bossen and J.Vanne delivers document
Analysis, its method complexity accounts for overall less than 1%.
To sum up, although the header that scanning work unit and output services unit can process two decoding frames eases up simultaneously
Rush area's information, because the complexity of itself method is significantly lower than the work complexity of decoded slices cutting unit, above-mentioned two
Computing unit occurs the longer idle waiting time in running, and this can cause the waste of computing resource.
Second drawback is to produce synchronization overhead.For the angle of parallel computation, decoding image display buffer is
One group of shared data, and in scanning work unit and output services unit, control the application of this relief area and the journey of release function
Sequence broadly falls into critical section.Execution sequence between critical section program must be controlled effectively, and otherwise will malfunction.One letter
Single example is, in the case of not having synchronization mechanism, when scanning work unit no longer makes because detecting in the buffer
Frame of video and execute memory space release operation when, output services unit is likely to also want because identical phenomenon is detected
Execute same operation.Computing resource not only can be wasted twice to identical operation execution and be also very likely to cause mistake, therefore
Need to introduce synchronization mechanism for this two critical sections.And the cost introducing synchronization mechanism is exactly to produce and synchronous related expense,
This includes waiting another working cell to complete what the exclusive formula to shared data unit accessed and caused because of a working cell
Resources idle, and the synchronization mechanism that provided using computer software and hardware resource and the expense that causes.
In sum it would be desirable to a kind of can improve computing resource utilization rate further and save the HEVC of synchronization overhead
Decoder parallel task division methods.
Content of the invention
The present invention is directed to the deficiency of existing HEVC decoder design method, and a kind of HEVC decoder parallel task of design divides
Method, thus realize the purpose improving computing resource utilization rate and saving synchronization overhead.
The present invention is achieved by the following technical solutions, HEVC decoder parallel task division side of the present invention
Method is it is stipulated that entropy decoding, inverse quantization, inverse transformation, infra-frame prediction, inter prediction, block elimination filtering and pixel adaptive equalization work(
The collaborative work mode of energy module, that is,：By to the process of network abstraction layer unit, the decoding to parameter set and header, and
The management work of decoding image buffering area is transferred to a control unit execution, the entropy decoding that decoding code tree unit is needed,
Inverse quantization, inverse transformation, infra-frame prediction, inter prediction, block elimination filtering and pixel adaptive equalization process transfer to several work
Unit executed in parallel, meets the working method of pipeline system between above-mentioned control unit and working cell.
In above-mentioned HEVC decoder parallel task division methods, the process work to network abstraction layer unit for the control unit
Including extract network abstraction layer unit and remove competition position；The parameter set information of control unit decoding includes video parameter collection, sequence
Row parameter set, image parameters collection and supplemental enhancement information；The header of control unit decoding includes network abstraction layer unit head
Information, access unit define symbol information, sequence ends information, bit stream termination message, filling data message, band segmentation head letter
Breath, band segmentation suffix information；Control unit is included to the management work of decoding image buffering area for image distribution to be decoded
Memory space, to the image in relief area according to display output order sort and release no longer be subsequent decoding process needed for
The space shared by image wanted.
In above-mentioned HEVC decoder parallel task division methods, a code tree cell row is responsible in each working cell
The updating context of decoding effort, wherein entropy decoding is according to the regulation to traveling wave flowing water parallel processing for the HEVC standard.Working cell
Between using no locking synchronous collaborative work mode, for the job note processing two spatially adjacent code tree cell row
Unit, in decoding progress, next line to be ensured at least lags behind two code tree units of lastrow.
Compared with prior art, the present invention has following beneficial effect：
In above-mentioned HEVC decoder parallel task division methods, between control unit and working cell, meet pipeline system
Working method, that is, working cell process current frame data when, control unit can process next frame data.Control unit and
Using no locking synchronous collaborative work mode between working cell.The present invention can better ensure that load balance and remove same
Step process is for the dependence of lock construction, thus improve computing resource utilization rate and having saved synchronization overhead.
Brief description
The detailed description with reference to the following drawings, non-limiting example made by reading, the further feature of the present invention,
Objects and advantages will become more apparent upon：
Fig. 1 be present invention provide that HEVC decoder parallel task dividing system frame diagram.
Fig. 2 is the workflow diagram of one embodiment of the invention control unit.
Fig. 3 is the workflow diagram of one embodiment of the invention working cell.
Fig. 4 is the no lock synchronizing process between one embodiment of the invention control unit and working cell.
Fig. 5 is the synchronous mistake of the spatially adjacent two code tree cell row up and down of one embodiment of the invention parallel decoding
Journey.
Specific embodiment
With reference to specific embodiment, the present invention is described in detail.Following examples will be helpful to the technology of this area
Personnel further understand the present invention, but the invention is not limited in any way.It should be pointed out that the ordinary skill to this area
For personnel, without departing from the inventive concept of the premise, some deformation can also be made and improve.These broadly fall into the present invention
Protection domain.
Shown in Fig. 1 be present invention provide that HEVC decoder task division methods.Wherein, control unit is responsible for from coding
Extract in bit stream and process network abstraction layer unit, parameter set information therein and header are parsed, then according to
The subcode stream of code tree row partition is stored in task pool, uses for follow-up decoding process.Additionally, control unit is also negative
Duty management decoding image buffering area, for the image distribution memory space that will decode, and discharges and is no longer made by subsequent decoding process
Memory area.In addition to a control unit, the system shown in Fig. 1 also can be distributed some according to specific hardware resource
Individual working cell.Mutually coordinated between each working cell, the concurrently subcode stream in decoding task pond.
The workflow for one embodiment of the invention control unit shown in Fig. 2.First determine whether that head is believed in original state
Whether have been obtained for the header less than a two field picture in breath relief area, if condition is unsatisfactory for, again detect this
Part；If condition meets, detect network abstraction layer unit further in the bitstream.Without network abstract layer is detected
Unit, then repeat the detection work to network abstraction layer unit, if network abstraction layer unit is detected, extracts this unit simultaneously
Remove competition position.The network abstraction layer unit later for removing competition position, can execute difference according to its Type Control unit
Parsing operation, if network abstraction layer unit is Parameter Set Type, preserve corresponding parse data after, control unit is returned again
Original state to flow process；If network abstraction layer unit splits type for band, control unit starts to parse band segmentation head
Information, code stream is divided in units of the code tree cell row, subcode stream for traveling wave flowing water parallel decoding, then in decoding
It is the image distribution memory area that will decode in image buffer, and discharge the image institute no longer being used by subsequent decoding process
The buffer area taking, eventually passes back to the starting stage of flow process.
The workflow for one embodiment of the invention working cell shown in Fig. 3.In the Detection task pond first of working cell
Whether there is the subcode stream corresponding with code tree cell row being available for decoding, without then duplicate detection process；If there are
Then determine whether whether first code tree unit of current line meets the data dependency condition needed for parallel decoding.Here
Data dependency condition refers to, after current decoding picture frame is divided in units of code tree unit, if present encoding tree
Unit is in line n m row, if the non-existent words of code tree unit of the (n-1)th row m+1 row, present encoding tree unit
Data dependency condition be defaulted as meet, if (n-1)th row m+1 row code tree unit exist if, then investigate this list
Whether the decoding process of unit completes, if completed, the data dependency condition of present encoding tree unit meets, if do not completed
Then show that data dependency condition is unsatisfactory for.Present encoding tree unit data dependency condition meet after, start to work as
Front code tree unit carries out entropy decoding, inverse quantization, inverse transformation, prediction reconstruction and loop filtering process.Complete in said process
Afterwards, investigate position in present encoding tree cell row for the present encoding tree unit, if present encoding tree unit is in current line
Low order end, then workflow return to the original state of flow chart shown in Fig. 3；If present encoding tree unit is not right in one's own profession
End, then be set to investigation object the code tree unit on the present encoding tree unit right side, and start to investigate whether this unit meets
Data dependency condition.
Shown in Fig. 4 for the no lock synchronizing process between one embodiment of the invention control unit and working cell.Above-mentioned two
Synchronizing process between class unit is controlled by a flag bit, and this flag bit is used for judging whether current frame pixel decoding completes,
And it is initialized to true.All this flag bit can be investigated after control unit is parsed a band header, if this mark
Position is SM set mode, then can start the parsing work of next band header and this flag bit resets；Otherwise, if should
Flag bit is reset state, then this flag bit of duplicate detection, until this flag bit is done.Last code tree of present frame is single
After unit's decoding, working cell set again.
The up and down two code tree cell row spatially adjacent for one embodiment of the invention parallel decoding shown in Fig. 5
Synchronizing process.Above-mentioned synchronizing process is to be controlled by each corresponding flag bit of code tree unit.When a code tree unit
Represent when flag bit is for SM set mode that the decoding process of this code tree unit has completed；If this flag bit is reset state
When, represent that the decoding process of this code tree unit not yet completes.The picture frame of current decoding is entered in units of code tree unit
Row divides, if code tree unit A is located at line n m and arranges, and code tree unit B is located at the (n-1)th row m+1 row.Judging code tree
When whether unit A can start decoding process, need to investigate the flag bit of code tree unit B, if this flag bit is set shape
State, then can start code tree unit A include entropy decoding, inverse quantization, inverse transformation, infra-frame prediction, inter prediction, block elimination filtering with
And the decoding process of pixel adaptive equalization etc., then the flag bit of code tree unit B is resetted, and to code tree unit A's
Flag bit set；If the flag bit of code tree unit B is reset state, the decoding effort of code tree unit A is needed by
Postpone, until the flag bit of code tree unit B is set because its decoding process completes.
Under test code streams encoding condition shown in experiment porch shown in table 1, table 2, to according to method institute in the present invention
The performance of the decoder of design is tested, and cycle testss are 3840x2160 for resolution, and bit-depth is 8, and aberration is adopted
Original mold formula is 4:2:0 ultra high-definition sequence C actus, Coastguard and foreman.The performance test results as shown in table 3, should
Test result proves that the method for present invention design has good real-time, which reflects the height to computing resource in decoding process
Effect uses and the saving to synchronization overhead.
Table 1 experiment porch information
Table 2 test code streams encoding condition
Table 3 ultra high-definition sequential decoding performance
Above the specific embodiment of the present invention is described.It is to be appreciated that the invention is not limited in above-mentioned
Particular implementation, those skilled in the art can make various modifications or modification within the scope of the claims, this not shadow
Ring the flesh and blood of the present invention.
Claims (6)
1. a kind of HEVC decoder parallel task division methods, it is pre- that the method defines entropy decoding, inverse quantization, inverse transformation, frame in
Survey, inter prediction, block elimination filtering and pixel adaptive equalization functional module collaborative work mode it is characterised in that：Described
Method is by the process of network abstraction layer unit, the decoding to parameter set and header, and the pipe to decoding image buffering area
Science and engineering transfers to a control unit execution, and the entropy decoding that decoding code tree unit is needed, inverse quantization, inverse transformation, frame in are pre-
Survey, inter prediction, block elimination filtering and pixel adaptive equalization process transfer to several working cell executed in parallel, above-mentioned control
The working method of pipeline system is met between unit and working cell.
2. a kind of HEVC decoder parallel task division methods according to claim 1 are it is characterised in that described control list
Unit includes to the process work of network abstraction layer unit extracting network abstraction layer unit and removing competition position；
The parameter set information of described control unit decoding includes video parameter collection, sequence parameter set, image parameters collection and supplement
Enhancement information；
The header of described control unit decoding includes network abstraction layer unit header, access unit defines symbol information, sequence
Termination message, bit stream termination message, filling data message, band segmentation header, band segmentation suffix information；
Described control unit includes to the management work of decoding image buffering area distributing memory space for image to be decoded, to slow
Rush the image in area to sort according to the order of display output and discharge shared by the image no longer required for subsequent decoding process
Space.
3. a kind of HEVC decoder parallel task division methods according to claim 1 and 2 are it is characterised in that each institute
State the decoding effort that a code tree cell row is responsible in working cell, wherein the updating context of entropy decoding is according to HEVC standard pair
The regulation of middle traveling wave flowing water parallel processing.
4. a kind of HEVC decoder parallel task division methods according to claim 3 are it is characterised in that described job note
Using no locking synchronous collaborative work mode between unit, and for processing the work of two spatially adjacent code tree cell row
Unit, in decoding progress, next line to be ensured at least lags behind two code tree units of lastrow.
5. a kind of HEVC decoder parallel task division methods according to claim 1 and 2 are it is characterised in that described control
The working method of pipeline system is met, that is, working cell, when processing current frame data, controls between unit processed and working cell
Unit can process next frame data.
6. a kind of HEVC decoder parallel task division methods according to claim 1 are it is characterised in that described control list
Using no locking synchronous collaborative work mode between unit and working cell.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410077419.7A CN103888782B (en) | 2014-03-04 | 2014-03-04 | Parallel task partitioning method for HEVC decoder |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410077419.7A CN103888782B (en) | 2014-03-04 | 2014-03-04 | Parallel task partitioning method for HEVC decoder |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103888782A CN103888782A (en) | 2014-06-25 |
| CN103888782B true CN103888782B (en) | 2017-02-08 |
Family
ID=50957470
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410077419.7A Expired - Fee Related CN103888782B (en) | 2014-03-04 | 2014-03-04 | Parallel task partitioning method for HEVC decoder |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103888782B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9477533B2 (en) * | 2014-12-26 | 2016-10-25 | Intel Corporation | Progress meters in parallel computing |
| CN104954808B (en) * | 2015-06-17 | 2018-03-23 | 福州瑞芯微电子股份有限公司 | One kind realizes rotation operating method and system based on code tree unit CTU buffer |
| CN105847840B (en) * | 2015-11-18 | 2018-12-07 | 西安邮电大学 | A kind of parallel organization building method for efficient video coding inverse transformation operation |
| CN106658012B (en) * | 2017-01-06 | 2020-06-19 | 华南理工大学 | Parallel pipeline task division method for VP9 decoder |
| CN111787330B (en) * | 2020-06-16 | 2022-03-08 | 眸芯科技（上海）有限公司 | Coding method supporting decoding compression frame buffer self-adaptive distribution and application |
| CN114125464B (en) * | 2020-08-27 | 2024-02-06 | 扬智科技股份有限公司 | Video decoding method and video decoding device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103024389A (en) * | 2012-12-24 | 2013-04-03 | 芯原微电子(北京)有限公司 | HEVC (high efficiency video coding) decoding device and method |
| CN103379318A (en) * | 2012-04-12 | 2013-10-30 | 中兴通讯股份有限公司 | Video bitstream coding method and apparatus |
| CN103414895A (en) * | 2013-07-29 | 2013-11-27 | 复旦大学 | Encoder intra-frame prediction device and method applicable to HEVC standards |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120257702A1 (en) * | 2011-04-11 | 2012-10-11 | Matthias Narroschke | Order of deblocking |
-
2014
- 2014-03-04 CN CN201410077419.7A patent/CN103888782B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103379318A (en) * | 2012-04-12 | 2013-10-30 | 中兴通讯股份有限公司 | Video bitstream coding method and apparatus |
| CN103024389A (en) * | 2012-12-24 | 2013-04-03 | 芯原微电子(北京)有限公司 | HEVC (high efficiency video coding) decoding device and method |
| CN103414895A (en) * | 2013-07-29 | 2013-11-27 | 复旦大学 | Encoder intra-frame prediction device and method applicable to HEVC standards |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103888782A (en) | 2014-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103888782B (en) | Parallel task partitioning method for HEVC decoder | |
| CN102098503B (en) | Method and device for decoding image in parallel by multi-core processor | |
| CN105992008B (en) | A kind of multi-level multi-task parallel coding/decoding method in multi-core processor platform | |
| CN105245897A (en) | Efficient software for transcoding to hevc on multi-core processors | |
| CN105163127B (en) | video analysis method and device | |
| CN102625108B (en) | Multi-core-processor-based H.264 decoding method | |
| US8660191B2 (en) | Software video decoder display buffer underflow prediction and recovery | |
| CN103974081B (en) | HEVC coding method based on multi-core processor Tilera | |
| CN101860752B (en) | Video code stream parallelization method for embedded multi-core system | |
| CN105554591A (en) | Video analysis method and device | |
| CN103297767A (en) | JPEG image decoding method and decoder suitable for multi-core embedded type platform | |
| CN101616323A (en) | The system and method that encoded video data stream is decoded | |
| CN105981383A (en) | Method and apparatus for processing video | |
| CN101222646A (en) | Infra-frame prediction device and method suitable for AVS encoding | |
| CN106851298B (en) | High-efficiency video coding method and device | |
| CN102572430A (en) | Method for implementing H.264 deblocking filter algorithm based on reconfigurable technique | |
| CN108540797A (en) | HEVC based on multi-core platform combines WPP coding methods within the frame/frames | |
| CN104980764A (en) | Parallel coding/decoding method, device and system based on complexity balance | |
| CN101383971A (en) | Intra-frame prediction processing method based on image encoding and decoding | |
| CN101616327A (en) | The method that multi-DSP core framework and fast parallel vision signal are handled | |
| CN106228506A (en) | A kind of method of multiple image parallel processing based on GPU | |
| CN101841722B (en) | Detection method of detection device of filtering boundary strength | |
| Gudumasu et al. | Software-based versatile video coding decoder parallelization | |
| CN101256668A (en) | Method for implementing video filtering to working balanced multiple nucleus | |
| US8111749B2 (en) | Image encoding or decoding method and device, with parallelization of processing over several processors and coprocessors, corresponding computer-readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170208 |