CN105025296B - A kind of advanced mathematical encoder and its implementation - Google Patents
A kind of advanced mathematical encoder and its implementation Download PDFInfo
- Publication number
- CN105025296B CN105025296B CN201410183605.9A CN201410183605A CN105025296B CN 105025296 B CN105025296 B CN 105025296B CN 201410183605 A CN201410183605 A CN 201410183605A CN 105025296 B CN105025296 B CN 105025296B
- Authority
- CN
- China
- Prior art keywords
- probabilistic model
- coding
- module
- context
- binaryzation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention discloses a kind of advanced mathematical encoder and its implementation, belong to digital video decoding technical field.The overall structure of the encoder uses parallel efficient design, wherein three layer state machine controlling mechanisms are using three layer state machines control cataloged procedure according to code stream hierarchical structure；It is in encoding current macroblock that technology, which is uniformly processed, in reference data, is simpler intermediate form by portion syntax elements processing, deposit Line Buffer, is used as next code macro block with reference to direct using and without handle again, to reduce logic circuit expense；Algorithm structure parallel optimization is according to special algorithm, is handled using parallel optimization, shortens processing time；Probabilistic model access management mechanism is the characteristics of being directed to high efficient coding of the present invention and the access control method that designs, the two-level memory mode combined using RAM and Local Buffer, shortens the access time to probabilistic model management；Arithmetic coding part is handled by the way of flowing water, reduces feedback time, improves the processing speed of data.
Description
Technical field
The present invention relates to digital video decoding technical field, more particularly to a kind of advanced mathematical coding based on AVS
Device and its implementation.
Background technology
Since nineteen nineties, digital video compaction technique is widely used in communication, personal computer, broadcast
The fields such as TV, consumer electronics, it can be rated as the core technology of digital media industry.The information source that audio frequency and video industry can select at present is compiled
Code standard has four：MPEG-2, MPEG-4, MPEG-4AVC (abbreviation AVC, also referred to as JVT, H.264), AVS, wherein AVS standards are
《Information technology advanced audio/video encodes》The abbreviation of series standard, it is the second generation message sink coding that China has independent intellectual property right
Standard.AVS is a set of full standard system comprising including system, video, audio, media copy management, is digital audio/video
Industry provides more fully solution.AVS video standards are mainly directed towards single-definition/high definition television broadcasting, net
The related application such as network TV and digital storage media, the computation complexity of cataloged procedure are larger.
Entropy code is an important step essential in video coding system, mainly by Entropy principle do not lose appoint
For representing that the symbol of element of video sequence changes into compressed bit stream in the case of what information, using remove comentropy redundancy come
Reach the purpose of compression.Entropy code mainly includes huffman coding, adaptive variable length coding and arithmetic coding etc..Wherein, MPEG-
H.264,2, using huffman coding (Huffman Code), all use with AVS and are based on context-adaptive variable-length encoding (CAVLC)
With based on context adaptive binary arithmetic coding (CABAC).Advanced mathematical coding (AEC) in AVS is will be adaptive
Binary arithmetic coding combines obtained method with a superior context model of design, and its design is based on two-value
3 steps such as change, context modeling, binary arithmetic coding.AEC is mainly to improve compression ratio in terms of two：(1)AEC
Non-static statistical model is provided to each syntactic element, encoder can be adaptively adjusted language according to encoded content and context
The probabilistic model of method element, adapt to different video contents, it is possible to reduce the redundancy of syntactic element, so as to improve compression effect
Rate.(2) for AEC using band as code period, the bin that the syntactic element binarization of a whole piece band is obtained carries out interval iteration
Division, so as to obtain a subinterval, then appoint in this subinterval and take a value to represent the band syntactic element, so,
From average, AEC is the numeral that symbol distributes a non-integer length, overcomes CAVLC codings and is necessary for single symbol
The shortcomings that code word of number distribution integer code length, it is easier to approach symbol entropy, obtain higher code efficiency.But AEC calculating
Complexity is substantially uprised, and challenge is proposed for hardware design.
The content of the invention
It is an object of the invention to considered to provide a kind of height in terms of chip area and processing speed two
Level arithmetic coding structure.
In order to solve problem above, the invention discloses a kind of advanced mathematical encoder, mainly by binaryzation and context
Index computing module, probabilistic model access management module and arithmetic coding module composition.Advanced mathematical coding is by syntactic element
Value carries out corresponding binaryzation, the corresponding probabilistic model of the binary value each obtained (bin), by obtained probabilistic model and
Binary value carries out arithmetic coding, obtains final code stream (bit).According to the difference of functions of modules, advanced mathematical encoder is adopted
With parallel efficient design mode.Because the binarization method of different syntactic elements and the method for probabilistic model index selection are different,
Its determination methods is frame field type, the macro block (mb) type of current coding macro block and the syntax element type of present encoding according to band
Etc. determining.So syntactic element only carried out into binaryzation in design, and based on context select each bin probability mould
Type indexes.Condition need not be seen again, can directly be encoded after obtaining binary value and corresponding probabilistic model index.So will
The selection of probabilistic model and arithmetic coding uniformly carry out outlet processing.
Further, advanced mathematical encoder adoption status control mode encodes to syntactic element, and this module needs to know
Road some specific information judge the syntactic element to be encoded.By taking AVS standards as an example, advanced mathematical coding is using band to be single
Position, band are made up of several macro blocks, and each macro block contains different syntactic elements, in advanced mathematical coding,
The coded sequence of syntactic element is fixed, but according to frame field type and the difference of macro block (mb) type, the language included in macro block
Method element is different.So control coded sequence using three layer state machines in advanced mathematical coding control module.First
Layer is that band redirects, and the second layer is redirecting for macro block, and third layer is redirecting for syntactic element coded sequence.
Context coding feature based on arithmetic coding, it is by data when using line buffer storing one row macro blocks
It is restored again into after processing in line buffer., can be from line so when encoding current macroblock needs its top macro block
Data are taken out in buffer directly to use, and without carrying out a series of logic judgment analysis again, are deposited again after such united analysis data
The mode of storage reduces logical complexity in the case where memory space is suitable, reduces logic circuit area.
Parallel data processing is used inside binaryzation and context index computing module, reduces operation time, guarantee processing
Speed.This module will be averaging a cycle (cycle) and export a bin and corresponding probabilistic model index ctxIdx,
Parallel data processing is made according to the characteristic of syntactic element binaryzation when designing module hardware structure, can be simultaneously within the identical time
Row handles more data, ensures design module performance.
Probabilistic model management module, first have to be encoded to choose probability according to probabilistic model index.To probability updating
The probability of renewal is stored back into the RAM of storage probabilistic model afterwards.In order to reduce the feedback processing time, using RAM and local
Caching secondary storage mode simultaneously increases when a probabilistic model control module takes probabilistic model to manage, and when deposits probability mould
Type.Its major function is that the probabilistic model as used in current probability model with last bin is identical, then does not have to take new probability, directly
Connect and be updated.When running into different probability model, take new probabilistic model, while probability updating is carried out by last renewal after
Probabilistic model be stored back in RAM, using parallel processing and control mode come reach reduce processing time purpose.
In advanced mathematical coding, arithmetic coding module (BAC) is the core of algorithm, due to the height in AVS standards
Level arithmetic coding is not look-up table, but to carry out update probability by calculating, and computation complexity is relative to be uprised.So arithmetic is compiled
Code module uses the design method of streamline.It is improved overall processing speed.
Brief description of the drawings
When considered in conjunction with the accompanying drawings, by referring to following detailed description, can more completely more fully understand the present invention with
And easily learn many of which with the advantages of, but accompanying drawing described herein be used for a further understanding of the present invention is provided,
The part of the present invention is formed, schematic description and description of the invention is used to explain the present invention, do not formed to this hair
Bright improper restriction, wherein：
Fig. 1, advanced mathematical encoder basic structure schematic diagram；
Fig. 2, advanced mathematical encoder concrete structure schematic diagram；
Fig. 3, the storage schematic diagram of macro block adjacent block；
Storage mode schematic diagram in Fig. 4, line buffer；
Fig. 5, advanced mathematical coding module first layer state machine diagram；
Fig. 6, advanced mathematical coding module second layer state machine diagram；
Fig. 7, advanced mathematical coding module third layer view；
Fig. 8, binaryzation and context index computing module schematic diagram；
Fig. 9, probabilistic model access management module schematic diagram；
Figure 10, arithmetic coding module schematic diagram.
Embodiment
Illustrated referring to Fig. 1-10 pairs of embodiments of the invention.
It is right with reference to the accompanying drawings and detailed description to enable above-mentioned purpose, feature and advantage more obvious understandable
The present invention is described in further detail.
As shown in figure 1, a kind of realization of advanced mathematical encoder based on AVS proposed by the present invention, main to include three layers
Technology, algorithm structure parallel optimization and the efficient flowing structure five of arithmetic coding is uniformly processed in state machine controlling mechanism, reference data
Kind technology.Just the implementation of the present invention makees example elaboration separately below.As shown in Figure 1 for binaryzation and context modeling
The coding-control of input employs three layer state machine controlling mechanisms, and binaryzation and context modeling are employed at reference data unification
Reason technology and algorithm structure parallel optimization, specifically include analysis syntactic element, binaryzation and context index computing module and
Line buffer participation.Probabilistic model access management employs probabilistic model access management mechanism, specifically includes analysis up and down
Literary index module, context RAM1 and context RAM2 and local cache (Local Buffer) participation.Arithmetic coding is adopted
With efficient flowing structure.
The operation principle of encoder shown in Fig. 1 is：The syntax element value of input is subjected to binaryzation and produces binary value, and
Based on context the probabilistic model corresponding to binary value is selected to index, according to the corresponding probability mould of probabilistic model index selection
Type, binary value probabilistic model corresponding with its is subjected to arithmetic coding, and the code stream after exports coding；Complete the arithmetic
Also include being updated probabilistic model corresponding to the binary value after coding.
A kind of advanced mathematical encoder as shown in Figure 2, mainly by binaryzation and context index computing module, probabilistic model
Access management module and arithmetic coding module composition, the binaryzation and context index computing module are used for syntax element value
Carry out binaryzation and produce binary value, and based on context select the probabilistic model corresponding to binary value to index；It is described general
Rate model access management module, for according to the corresponding probabilistic model of probabilistic model index selection, for the binary value
Corresponding probabilistic model is updated；The arithmetic coding module is used to be calculated binary value probabilistic model corresponding with its
Art encodes, and the code stream after exports coding；The encoder is used as data buffering using binary value and probabilistic model index FIFO
Device, binaryzation and context index computing module, probabilistic model access management module and arithmetic coding module concurrent operation.It is described
Using binaryzation with based on context calculating probabilistic model index parallel processing inside binaryzation and context index computing module,
In the inside of binaryzation, the binaryzation of macroblock coding template can directly obtain, the binaryzation of motion vector components difference (mvd)
Need to be by being calculated, it is possible to by cbp and mvd binaryzation with organizing top block and left side block parallel data processing,
Hardware resource area reduces operation time in the case of being guaranteed, ensure processing speed.
From the overall structure of Fig. 2 arithmetic encoders, arithmetic encoder is using coding control module come to whole
Cataloged procedure is controlled.Arithmetic encoder sends request as dynamic RAM first, and dynamic RAM can be by one
The individual macro block all data to be encoded are sent in syntactic element data buffer, and analytic language is controlled by coding control module
Method element module reads data and analyzed from buffer, be stored into after data are uniformly processed line buffer to
Next code reads top block and used.Subsequent binaryzation can read the information of current syntax element with context index computing module
With the syntax element information of top block, binaryzation is carried out, and the general of each binary value is calculated according to top block and left side block
Rate model is indexed, and binary value and probabilistic model index are spliced, and is exported to binary value and probabilistic model index FIFO
Buffer.Analysis context index module reads each binary value of data post analysis and corresponding probabilistic model from buffer
Index, from the RAM (context RAM1 and context RAM2) of two storage probabilistic models after the corresponding probabilistic model of extraction, output
Encoded to binary arithmetic coding module, the code stream to be spued after coding is stored back into dynamic memory, obtained to the end
Encode file.In Fig. 2, need to learn that specific syntax element type, frame, field type and macro block (mb) type etc. is come to language in part A
Method element carries out binaryzation and context index calculates.But need not just know the occurrence of these variables again in B and C portion,
Coding method unitizes, as long as learning each bin (binary value after binaryzation) and corresponding probabilistic model index address, just
It can be encoded.Function division is carried out according to the characteristics of above arithmetic coding, using parallel efficient design structure.Wherein two enter
Value processed plays cushioning effect with probabilistic model index FIFO so that two different modules of processing speed can preferably coordinate.A
As long as part sees that FIFO is discontented with can and continues with syntactic element, by the bin after binaryzation and corresponding probabilistic model
In index press-in FIFO, as long as and coding detection FIFO below is not empty, it is possible to it is continuous to read and coded data.So design
Internal structure can be made relatively independent, and judge whether to need work by best detection signal.
Because encoding current macroblock is come select probability model, so needing according to its top macro block and left side macro block information
Top macro block and the access of left side macro block are come.So as shown in figure 3, left side macro block is stored using buffer, top macro block uses
One line buffer is stored.In data storage, we first store data again after treatment, are so sequeled after
During code, data can be taken out from line buffer and directly used, without carrying out a series of logic judgment analysis again, such as Fig. 4 institutes
Show, a macro block is divided into four fritters by we, and according to the difference of macro block (mb) type, four fritters are stored with corresponding grammer member
The data storage of element, such as 8X16 macro block (mb) type, No. 0 block and No. 2 blocks is identical, and the data storage of No. 1 block and No. 3 blocks is identical.
The information of No. 2 blocks and No. 3 blocks need to be only stored when storing top macro block, and left side macro block stores No. 1 block and No. 3 blocks.This
The mode stored again after sample united analysis data reduces logical complexity in the case where memory space is suitable, reduces circuit area
Expense.
According to arithmetic coding with band (slice) for unit, slice includes macro block, and syntactic element is included in macro block
Characteristic, the coding of advanced mathematical coding module is controlled with three-level state machine.Jumped as shown in figure 5, first layer is slice states
Turn, into AEC MB states after start encode a macro block, in this condition, separate shape of the second layer for a macroblock coding
State machine.Fig. 6 is that the second layer redirects for macroblock status, under MB One states, separates third layer state machine.Fig. 7 third layer is then
The state transition of syntactic element, mainly the syntactic element in a macro block is redirected according to condition progress order.
Binaryzation and probabilistic model index computing module are as shown in figure 8, the module is indexed using binaryzation and probabilistic model
Control module is controlled to whole process, reads linebuffer first, obtains the data of top block.Then by organizing number
Pass through binaryzation advanced processing while according to module the syntax element value of current macro and top macro block to be uniformly processed
Module carries out binaryzation in advance to some special syntactic elements, and the data output after processing indexes to binaryzation and probabilistic model
Module.Number of the control module according to state transition actuation techniques module according to current syntactic element Data-Statistics binary value,
Give the data counted to binaryzation and context index output module, the context that the module provides according to tissue data is believed
The statistics number of breath and counter, corresponding binary value and probabilistic model index output is carried out to syntactic element.Due to the mould
Block will be averaging a cycle (cycle) and export a bin and corresponding probabilistic model index (ctxIdx), in design module
Parallel data processing is made according to the characteristic of syntactic element binaryzation during hardware configuration, reduces data processing time.Used as
The syntactic element of unitary code binaryzation method, is first counted with counting module, then exports corresponding bin.But mvd (motion arrows
Measure difference) binarization method it is more special, if synElVal (syntactic element) value is more than or equal to 3 and synElVal
Value be odd number, first four of binary symbols string are " 1110 ", and subsequent bit is 0 rank index brother's human relations corresponding to (synElVal-3)/2
Cloth code；If value of the synElVal value more than 3 and synElVal is even number, first four of binary symbols string are " 1111 ",
Subsequent bit is 0 rank Exp- Golomb corresponding to (synElVal-3)/2.So Columbus's binarizing portion is located can in advance
Reason, and tissue data parallel perform.Syntactic element CBP (macroblock coding template) maximum is 63, and its binarization method is fixed,
There is no rule, so above parallel processing can also be placed on.Processing time can be reduced by being achieved in that, ensure efficient process
Energy.
, it is necessary to be upgraded in time to probability after extraction probabilistic model carries out arithmetic coding probability updating, whole place result in
Managing speed reduces.In order to solve this problem, as shown in figure 9, the present invention uses two RAM and local caches secondary storage knot
Structure, and add the probabilistic model access management module control access time.Wherein context RAM1 and context RAM2 is two pieces
RAM, it is respectively intended to store 195 probabilistic models.Probabilistic model access management module is control module, in order to avoid by probability with
The probabilistic model that reselection repeats after newly going back, after receiving probability index, the module can judge, if to use same
Individual probabilistic model is then directly with the probability after previous renewal, without being read from RAM, if probabilistic model used is not same
It is individual, then corresponding probabilistic model is chosen in RAM.
Arithmetic coding is algorithm nucleus module, and as shown in Figure 10, the arithmetic coding module uses the design of Pyatyi flowing water
Method, first order flowing water are used for probabilistic model being updated；Module in the flowing water of the second level is in order to avoid probability updating is returned
The probabilistic model for going rear reselection to repeat, after receiving probabilistic model index, it can judge, if to use same probability
Model is then directly with the probabilistic model after previous renewal, without being read from context RAM, if probabilistic model used is not
It is same, then corresponding probabilistic model is chosen in context RAM；Third level flowing water is mainly updated to region, first to area
Domain calculates, and gives the data after calculating to area update module and region is safeguarded, the data after maintenance can feed back to region
Computing module exports；Fourth stage flowing water is updated to region starting point；After level V flowing water mainly completes coding reformation normalization
Discharge code stream.Probability updating carries out addition and subtraction using non-multiplication but after changing into log-domain to calculate new probability, though it is simple
Change, but algorithm is still very complicated.But it have been found that many computings are not fed back during update probability, it is single renewal budget
, so the present invention uses flowing water method, by the computing unification of complexity.After binaryzation above, after binaryzation
In bin and probabilistic model index deposit FIFO, the processing speed of this process can be quickly.And arithmetic coding is set using flowing water below
Meter, the bin in FIFO can be constantly encoded, is so improved in bulk velocity.
Finally it should be noted that：Obviously, above-described embodiment is only intended to clearly illustrate example of the present invention, and simultaneously
The non-restriction to embodiment.For those of ordinary skill in the field, can also do on the basis of the above description
Go out other various forms of changes or variation.There is no necessity and possibility to exhaust all the enbodiments.And thus drawn
Among the obvious changes or variations that Shen goes out is still in protection scope of the present invention.
Claims (7)
- A kind of 1. advanced mathematical encoder, it is characterised in that：Mainly by binaryzation and context index computing module, probabilistic model Access management module and arithmetic coding module composition, the binaryzation and context index computing module use three layer state machine controls Syntax element value is carried out binaryzation and produces binary value by making mechanism, while technical finesse part is uniformly processed using reference data Syntactic element, and based on context select the probabilistic model corresponding to binary value to index；The probabilistic model access management Module, for according to the corresponding probabilistic model of probabilistic model index selection, for probabilistic model corresponding to the binary value It is updated；The arithmetic coding module is used to binary value probabilistic model corresponding with its carrying out arithmetic coding, and exports Code stream after coding；The encoder is controlled using coding control module to whole cataloged procedure, and with binary value and probabilistic model Buffer is indexed as data buffer so that binaryzation and context index computing module, probabilistic model access management module Can concurrent operation with arithmetic coding module.
- 2. advanced mathematical encoder according to claim 1, it is characterised in that：The binaryzation and context index calculate Inside modules using binaryzation with based on context calculating probabilistic model index parallel processing, specifically in the inside of binaryzation, By the binaryzation of macroblock coding module and motion vector components difference and tissue top block and left side block parallel data processing.
- 3. advanced mathematical encoder according to claim 1, it is characterised in that：Upper phase is stored using L i neBuffer Adjacent block, technology is uniformly processed using reference data, L i ne Buffer, specific method are restored again into after unified progress data processing It is that a macro block is divided into four fritters 0,1,2,3, the information of four fritters determines according to the type of macro block, adjacent in storage Only need to store No. 2 blocks and No. 3 blocks during macro block, No. 1 block and No. 3 blocks need to be only stored when storing left adjacent macroblocks.
- 4. advanced mathematical encoder according to claim 1, it is characterised in that：The probabilistic model access management module is adopted With two-level memory, including context RAM and local caches, wherein local caches be used to storing nearest meeting uses up and down Literary probabilistic model, context RAM is two block RAMs, for depositing 195 context probability modeling information respectively, in context RAM Controlled whether between local caches using probabilistic model access control module according to the probabilistic model of adjacent binary value New probabilistic model is read from RAM, whether the probabilistic model after renewal will be stored back to RAM.
- 5. advanced mathematical encoder according to claim 1, it is characterised in that：The arithmetic coding module uses Pyatyi stream The design method of water, first order flowing water are used for probabilistic model being updated；Module in the flowing water of the second level is in order to avoid by generally The probabilistic model that reselection repeats after rate renewal is gone back, after receiving probabilistic model index, it can judge, if to use same One probabilistic model is then directly with the probabilistic model after previous renewal, without being read from context RAM, if probability used Model is not same, then corresponding probabilistic model is chosen in context RAM；Third level flowing water is mainly updated to region； Fourth stage flowing water is updated to region starting point；Level V flowing water mainly completes the code stream that spued after coding reformation normalizes.
- A kind of 6. implementation method of advanced mathematical encoder, it is characterised in that：Whole cataloged procedure is controlled using three layer state machines The syntax element value of input is carried out binaryzation and produces binary value by mechanism, while technical finesse is uniformly processed using reference data Portion syntax elements, and based on context select the probabilistic model corresponding to binary value to index, indexed according to probabilistic model Corresponding probabilistic model is selected, binary value probabilistic model corresponding with its is subjected to arithmetic coding, and the code after exports coding Stream；Also include being updated probabilistic model corresponding to the binary value after the arithmetic coding is completed.
- 7. according to the method for claim 6, it is characterised in that：The syntax element value by input carries out binaryzation generation Binary value, using three layer state machines control, the coding of first layer state machine major control band, including band initialization and Band end-of-encode needs to flush；The coding of second layer state machine major control macro block, including whether macro block is last Macro block, if last macro block wants direct coding 1, otherwise encode 0；Language in one macro block of third layer state machine major control The order of method element, the type of syntax elements encoded and order are determined according to frame/field type, macro block (mb) type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410183605.9A CN105025296B (en) | 2014-04-30 | 2014-04-30 | A kind of advanced mathematical encoder and its implementation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410183605.9A CN105025296B (en) | 2014-04-30 | 2014-04-30 | A kind of advanced mathematical encoder and its implementation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105025296A CN105025296A (en) | 2015-11-04 |
| CN105025296B true CN105025296B (en) | 2018-02-23 |
Family
ID=54414957
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410183605.9A Active CN105025296B (en) | 2014-04-30 | 2014-04-30 | A kind of advanced mathematical encoder and its implementation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105025296B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107071494B (en) * | 2017-05-09 | 2019-10-11 | 珠海市杰理科技股份有限公司 | The generation method and system of the binary syntax element of video image frame |
| CN107659814A (en) * | 2017-09-21 | 2018-02-02 | 深圳市德赛微电子技术有限公司 | Entropy decoding structure in a kind of bimodulus decoder of AVS and MPEG 2 |
| CN111683253A (en) * | 2020-06-12 | 2020-09-18 | 浪潮(北京)电子信息产业有限公司 | Parameter set decoding method and device |
| CN111787325B (en) * | 2020-07-03 | 2022-03-08 | 北京博雅慧视智能技术研究院有限公司 | Entropy encoder and encoding method thereof |
| WO2023197104A1 (en) * | 2022-04-11 | 2023-10-19 | Oppo广东移动通信有限公司 | Coding method, decoding method, encoders, decoder and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102231830A (en) * | 2010-11-23 | 2011-11-02 | 浙江大学 | Arithmetic unit used for context arithmetic encoding and decoding |
| CN103227924A (en) * | 2013-05-07 | 2013-07-31 | 北京大学 | Arithmetic coder and coding method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI487295B (en) * | 2011-05-17 | 2015-06-01 | Univ Nat Cheng Kung | Method for high throughput parallel avc/h.264 context-based adaptive binary arithmetic decoder |
-
2014
- 2014-04-30 CN CN201410183605.9A patent/CN105025296B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102231830A (en) * | 2010-11-23 | 2011-11-02 | 浙江大学 | Arithmetic unit used for context arithmetic encoding and decoding |
| CN103227924A (en) * | 2013-05-07 | 2013-07-31 | 北京大学 | Arithmetic coder and coding method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105025296A (en) | 2015-11-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105025296B (en) | A kind of advanced mathematical encoder and its implementation | |
| CN103733622B (en) | Context initialization in entropy code | |
| TWI514779B (en) | Entropy coding | |
| Zhang et al. | Fast CU partition and intra mode decision method for H. 266/VVC | |
| JP6130839B2 (en) | Entropy coding | |
| Pastuszak et al. | Algorithm and architecture design of the H. 265/HEVC intra encoder | |
| CN105393537B (en) | Hash table building and availability inspection for the Block- matching based on hash | |
| CN104811707B (en) | The equipment decoded to video | |
| KR101941955B1 (en) | Recursive block partitioning | |
| CN103416057B (en) | Image coding apparatus and method and image decoding apparatus and method | |
| CN105357540A (en) | Method and apparatus for decoding video | |
| CN103227924B (en) | A kind of arithmetic encoder and coded method | |
| CN109076212A (en) | The method and code device of code rate control | |
| CN104094600A (en) | Method and apparatus for hierarchical data unit-based video encoding and decoding comprising quantization parameter prediction | |
| CN104365099A (en) | Parameter update method for entropy coding and decoding of conversion coefficient level, and entropy coding device and entropy decoding device of conversion coefficient level using same | |
| KR102165070B1 (en) | Coding of video syntax elements using context tree | |
| CN106576168A (en) | Image compression/decompression device | |
| CN108028933A (en) | Video coding and decoding method and device | |
| CN102238383B (en) | For the multibus system architecture of Video Codec | |
| Srinivasarao et al. | High‐speed low‐power very‐large‐scale integration architecture for dual‐standard deblocking filter | |
| CN110191339A (en) | Rate estimation core cell, rate estimation device and code rate estimation method | |
| WO2024037091A1 (en) | Point cloud processing method and apparatus, computer device, and storage medium | |
| Menasri et al. | Field programmable gate array implementation of variable‐bins high efficiency video coding CABAC decoder with path delay optimisation | |
| CN108353174A (en) | Parallel arithmetic decoding technique | |
| KR101108674B1 (en) | Performance inhancement method of entophy coding and video coding method using the performance inhancement method and appartus thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |