CN105939479B - A kind of parallel encoding method and apparatus - Google Patents
A kind of parallel encoding method and apparatus Download PDFInfo
- Publication number
- CN105939479B CN105939479B CN201610425871.7A CN201610425871A CN105939479B CN 105939479 B CN105939479 B CN 105939479B CN 201610425871 A CN201610425871 A CN 201610425871A CN 105939479 B CN105939479 B CN 105939479B
- Authority
- CN
- China
- Prior art keywords
- unit
- coding
- group
- adjacent encoder
- present encoding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/42—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
- H04N19/436—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 using parallelised computational arrangements
-
- 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
- H04N19/119—Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
Abstract
The embodiment of the invention provides a kind of parallel encoding method and apparatus, are related to technical field of video coding.The described method includes: being evenly dividing the coding unit of image by column for M code set, the M code set corresponds to M thread；Start each thread and the parallel encoding predicted based on adjacent column group is carried out between corresponding code set.The present invention solves in first technology such as the corresponding primary initialization coding of the every row coding unit of waveform parallel encoding and primary ending coding；And without motion prediction between different zones in tile parallel encoding, and the first row coding unit in the region started for non-the first row, it is not available negative motion vector to be predicted, the lower problem of caused code efficiency, achieves the beneficial effect for improving code efficiency.
Description
Technical field
The present invention relates to technical field of video coding, more particularly to a kind of parallel encoding method and apparatus.
Background technique
The appearance of multi-core CPU (Central Processing Unit, central processing unit) brings huge for people's lives
Big change.In order to make full use of the advantage of multi-core CPU, starts multiple threads and parallel encoding is carried out to video image.
In first technology, there are mainly two types of parallel encodings: waveform parallel encoding and tile parallel encoding.Such as Fig. 1, waveform is simultaneously
Row coding starts multiple threads and carries out parallel encoding to four row coding units, and every row coding unit is suitable according to from left to right
Sequence coding, every row coding unit correspond to independent entropy coding.Wherein, for second coding unit 42 of fourth line, it is necessary to
Start to encode again after the completion of three row third coding units 33 coding.Tiles code divides video into multiple do not overlap first
Rectangular area, then start multiple threads to each region carry out parallel encoding.Wherein, the corresponding independent entropy in each region
It encodes, without motion prediction between each region.
As can be seen that the corresponding primary initialization coding of every row coding unit and primary ending are compiled in waveform parallel encoding
Code, when video image is bigger, will lead to the waste of code efficiency.Fortune has been blocked in tile parallel encoding between different zones
Dynamic prediction, will lead to reduction code efficiency；In addition, when video image horizontal parallel, in the region that starts for non-the first row
The first row coding unit is not available negative motion vector and is predicted, to also result in reduction code efficiency.
Summary of the invention
In view of the above problems, it proposes on the present invention overcomes the above problem or at least be partially solved in order to provide one kind
State the parallel encoding method and apparatus of problem.
According to one aspect of the present invention, a kind of parallel encoding method is provided, comprising:
The coding unit of image is evenly dividing by column as M code set, the M code set corresponds to M thread；
Start each thread and the parallel encoding predicted based on adjacent column group is carried out between corresponding code set.
Preferably, each thread of starting carries out the parallel encoding predicted based on adjacent column group to corresponding code set
Step, comprising:
For each coding unit of present encoding group, according to the coding unit in present encoding group, present encoding group
The encoded content of adjacent encoder unit in adjacent encoder group starts each thread by row and carries out predictive coding from left to right.
Preferably, each coding unit for present encoding group, according to the coding unit present encoding group, when
Adjacent encoder unit in the adjacent encoder group of preceding code set, starting thread are carried out the step of predictive coding by row from left to right,
Include:
Judge the position of current coded unit locating column in present encoding group；
If current coded unit is obtained from the left adjacent encoder group of present encoding group in the first row of present encoding group
The encoded content of the left adjacent encoder unit of current coded unit, upper left adjacent encoder unit, obtains from present encoding group and works as
The encoded content of the upper adjacent encoder unit of preceding coding unit, upper right adjacent encoder unit carries out predictive coding；
If current coded unit is obtained from the right adjacent encoder group of present encoding group in last column of present encoding group
The encoded content for taking the upper right adjacent encoder unit of current coded unit obtains a left side for current coded unit from present encoding group
Adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit encoded content, carry out predictive coding；
If current coded unit in the inner array of present encoding group, obtains current coded unit from present encoding group
Left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right adjacent encoder unit encoded content, into
Row predictive coding.
Preferably, further includes:
The code stream length of each code set is encoded；
By the code stream length after coding according to code set from left to right be sequentially written in total code stream；
By the corresponding code stream of each code set according to code set from left to right be sequentially written in total code stream.
Preferably, each coding unit for present encoding group, according to the coding unit present encoding group, when
Adjacent encoder unit in the adjacent encoder group of preceding code set, starting thread are carried out the step of predictive coding by row from left to right,
Include:
Judge the position of current coded unit locating column in present encoding group；
If current coded unit is obtained from the left adjacent encoder group of present encoding group in the first row of present encoding group
The encoded content for taking the lower-left adjacent encoder unit of current coded unit, left adjacent encoder unit, upper left coding unit, from current
In code set, the upper adjacent encoder unit of current coded unit, the encoded content of upper right adjacent encoder unit are obtained, is predicted
Coding；
If current coded unit obtains present encoding list from present encoding group in last column of present encoding group
Left adjacent encoder unit, upper left adjacent encoder unit, the upper adjacent encoder unit of member carry out predictive coding；
If current coded unit is in the inner array of present encoding group, from present encoding group, current coded unit is obtained
Left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right adjacent encoder unit carry out prediction volume
Code.
Preferably, further includes:
By the corresponding code stream of each code set according to code set from left to right be sequentially written in total code stream.
Another aspect according to the present invention provides a kind of parallel encoding device characterized by comprising
Division module, for being evenly dividing the coding unit of image by column for M code set, the M code set pair
Answer M thread；
Coding module carries out the parallel volume predicted based on adjacent column group to corresponding code set for starting each thread
Code.
Preferably, the coding module, comprising:
First encoding submodule, for each coding unit for present encoding group, according to the coding unit current
The encoded content of code set, adjacent encoder unit in the adjacent encoder group of present encoding group starts each thread by row from left-hand
Right carry out predictive coding.
Preferably, first encoding submodule, comprising:
First judging unit, for judging the position of current coded unit locating column in present encoding group；
First coding unit, if for current coded unit present encoding group first row, from present encoding group
The left adjacent encoder unit of current coded unit, the encoded content of upper left adjacent encoder unit are obtained in left adjacent encoder group, from
The upper adjacent encoder unit of current coded unit, the encoded content of upper right adjacent encoder unit are obtained in present encoding group, are carried out
Predictive coding；
Second coding unit, if for current coded unit present encoding group last column, from present encoding group
Right adjacent encoder group in obtain current coded unit upper right adjacent encoder unit encoded content, obtained from present encoding group
The encoded content of the left adjacent encoder unit of current coded unit, upper left adjacent encoder unit, upper adjacent encoder unit is taken, is carried out
Predictive coding；
Third coding unit, if for current coded unit present encoding group inner array, from present encoding group
Obtain left adjacent encoder unit, upper left adjacent encoder unit, the upper adjacent encoder unit, upper right adjacent encoder of current coded unit
The encoded content of unit carries out predictive coding.
Preferably, further includes:
First code stream VLC module is encoded for the code stream length to each code set；
First code stream length writing module, the sequential write for the code stream length after encoding according to code set from left to right
Enter total code stream；
First code stream writing module, for by the corresponding code stream of each code set according to code set being sequentially written in from left to right
Total code stream.
Preferably, first encoding submodule, comprising:
Second judgment unit, for judging the position of current coded unit locating column in present encoding group；
4th coding unit, if for current coded unit present encoding group first row, from present encoding group
In left adjacent encoder group, lower-left adjacent encoder unit, the left adjacent encoder unit, upper left coding unit of current coded unit are obtained
Encoded content obtain the upper adjacent encoder unit of current coded unit, upper right adjacent encoder unit from present encoding group
Encoded content carries out predictive coding；
5th coding unit, if for current coded unit present encoding group last column, from present encoding group
In, the left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit for obtaining current coded unit carry out prediction volume
Code；
6th coding unit, if for current coded unit present encoding group inner array, from present encoding group,
Obtain left adjacent encoder unit, upper left adjacent encoder unit, the upper adjacent encoder unit, upper right adjacent encoder of current coded unit
Unit carries out predictive coding.
Preferably, further includes:
Second code stream writing module, for by the corresponding code stream of each code set according to code set being sequentially written in from left to right
Total code stream.
The embodiment of the present invention has the advantages that
A kind of parallel encoding method and apparatus according to the present invention, can by the coding unit of image by column be evenly dividing for
M code set, the M code set correspond to M thread；Start adjacent column each thread carries out based on group corresponding code set
The parallel encoding of prediction.Thus the corresponding primary initialization coding of every row coding unit and primary ending in first technology are solved
Coding, without motion prediction between different zones, when video image horizontal parallel, in the region that starts for non-the first row
The first row coding unit is not available negative motion vector and is predicted, the lower problem of caused code efficiency is achieved and mentioned
The beneficial effect of high coding efficiency.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And it can be implemented in accordance with the contents of the specification, and in order to allow above and other objects of the present invention, feature and advantage can
It is clearer and more comprehensible, the followings are specific embodiments of the present invention.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows the waveform parallel encoding schematic diagram of first technology；
Fig. 2 shows a kind of step flow charts of parallel encoding embodiment of the method one according to the present invention；
Fig. 2A shows the parallel encoding schematic diagram of the embodiment of the present invention；
Fig. 3 shows a kind of step flow chart of parallel encoding embodiment of the method two according to the present invention；
Fig. 4 shows a kind of step flow chart of parallel encoding embodiment of the method three according to the present invention；
Fig. 5 shows a kind of structural block diagram of parallel encoding Installation practice four according to the present invention；
Fig. 6 shows a kind of structural block diagram of parallel encoding Installation practice five according to the present invention；
Fig. 7 shows a kind of structural block diagram of parallel encoding Installation practice six according to the present invention.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
Embodiment one
Referring to Fig. 2, a kind of step flow chart of parallel encoding embodiment of the method one according to the present invention is shown, specifically may be used
To include the following steps:
The coding unit of image is evenly dividing by column as M code set by step 101, and the M code set corresponds to M
Thread；
Wherein, image is made of pixel, and image can be indicated with pixel number.Under normal conditions, image is used
Width and the height pixel number that is included indicate image.For example, 1024*528 indicates the wide and high pixel number of image
Respectively 1024 and 528.
In practical applications, usually image is encoded as unit of coding unit.Wherein, coding unit can be more
The rectangular area of a pixel composition, for example, the rectangular area of 8*8 pixel composition can be used as a coding unit, then,
The image that can be 128*256 by size is divided into the coding unit that 16*32 size is 8*8.Certainly, different coding standard
There is different definition to coding unit, can also be set according to concrete application.
The embodiment of the present invention is further divided into code set on the basis of coding unit, by image.Such as Fig. 2A, image has
The coding unit of image is divided into 3 code sets according to column by 9 column coding units, wherein each code set includes 3 column codings
Unit.Since each column coding unit includes equal number of coding unit, so each code set includes equal number of coding
Unit.
In practical applications, the coding unit columns of image is not usually the integral multiple of Thread Count.In these cases, it is
Make multithreading to the overall coding efficiency highest of each code set, the coding unit number that each code set includes is equal as far as possible or close
Patibhaga-nimitta etc..It is equal with number of threads in code set number when the columns of the coding unit of image is not the integral multiple of Thread Count
Under conditions of, coding unit can be divided equally as far as possible in each code set.For example, if image has 11 column coding units, Thread Count
When being 3, then using 24 column coding units as 2 code sets, 13 column coding unit is as 1 code set, 3 codings
Corresponding 3 threads of group.
It is appreciated that the corresponding relationship between thread and code set can be set according to concrete application.Implement in the present invention
It is according to sequence from left to right that each code set is corresponding with each thread in example.
Step 102, each thread of starting carry out the parallel encoding predicted based on adjacent column group to corresponding code set.
In multi-core CPU, it can star multiple threads and parallel processing carried out to data.In embodiments of the present invention, start
Multiple threads carry out parallel encoding to ready-portioned each code set.
In practical applications, in order to save the bandwidth resources for transmitting image, image is compressed before transmitting image
Coding.Predictive coding is the most common compressed encoding, using the redundancy between pixel in image, according to the value of neighbor pixel
It predicts the value of current pixel point, and the difference for the actual value for predicting obtained value and current pixel point is encoded.
The embodiment of the present invention can permit using the adjacent encoder unit in adjacent encoder group, to the coding unit of code set
Carry out predictive coding.Such as Fig. 2A, the coding unit in the column L3 of the first code set can be according in the column L4 of the second code set
Coding unit carries out predictive coding, and the coding unit in the column L6 of the second code set can be according in the column L7 of third code set
Coding unit carries out predictive coding, and the coding unit in the column L4 of the second code set can also be according in the column L3 of the first code set
Coding unit carry out predictive coding, the coding unit in the column L7 of third code set can also be according to the column L6 of the second code set
In coding unit carry out predictive coding.
In order to improve code efficiency, each thread carries out parallel encoding to code set.Before the coding for starting each code set, need
Encoder is initialized, be one initial value of encoder settings.In practical applications, starting to encode present encoding group
The first row coding unit before, can be using the arithmetic coding content of the adjacent encoder unit in left adjacent sets as present encoding
The initial value of device.Encoded content of the initial value as first coding unit of the first row coding unit of present encoding group, from
Second coding unit of the first row coding unit starts to be encoded.Described in being obtained from the corresponding thread of left adjacent sets
Arithmetic coding content needs to carry out inter-thread communication.For each code speech, corresponding existing inter-thread communication side can be used
Method, the embodiment of the present invention to the specific method of inter-thread communication with no restrictions.
It is appreciated that encoder is without initialization, directly since first coding unit for first code set
It is encoded.
Preferably, step 102 may include sub-step 1021:
Sub-step 1021, each coding unit for present encoding group, according to the coding unit present encoding group, when
The encoded content of adjacent encoder unit in the adjacent encoder group of preceding code set starts each thread and is predicted from left to right by row
Coding.
As Fig. 2A starts thread by row and carry out predictive coding from left to right, respectively for each coding unit of the first code set
The corresponding coded sequence of coding unit are as follows: coding unit 11, coding unit 12, coding unit 13, coding unit 21, coding unit
22, coding unit 23, coding unit 31, coding unit 32, coding unit 33, coding unit 41, coding unit 42, coding unit
43, coding unit 51, coding unit 52, coding unit 53, coding unit 61, coding unit 62, coding unit 63, coding unit
71, coding unit 72, coding unit 73.
In practical applications, for a coding unit, predictive coding is carried out according to the adjacent encoder unit of the coding unit.
Wherein, the adjacent encoder unit of the coding unit can be the coding unit in present encoding group, or present encoding group
Adjacent encoder group in coding unit.The coding list of coding unit and rightmost side column for the leftmost column in code set
Member can use the coding unit of the rightmost side column in left adjacent encoder group respectively, the leftmost column in right adjacent encoder group
Coding unit carries out predictive coding.
It is appreciated that needing the reference encoder cell encoding relied in the coding unit to the coding of any coding unit
Start after terminating.
The coding unit divided by column of image can be M code set by the present embodiment, and start M thread to each coding
Group carries out the coding predicted based on adjacent column between group, it is possible to reduce the initialization times and coding ending of encoder, Ke Yi
Predictive coding is carried out between the adjacent column of adjacent encoder group, so that code efficiency is improved, further, since the embodiment of the present invention is not
Divided by row code set, so the first row coding unit in the region started there is no non-the first row, is not available negative movement
The problem of vector is predicted, to improve code efficiency.
Embodiment two
Referring to Fig. 3, a kind of step flow chart of parallel encoding embodiment of the method two according to the present invention is shown, specifically may be used
To include the following steps:
The coding unit of image is evenly dividing by column as M code set by step 201, and the M code set corresponds to M
Thread；
Step 202, the position for judging current coded unit locating column in present encoding group；
In the embodiment of the present invention, according to the adjacent encoder unit of current coded unit, current coded unit is predicted
Coding.When present encoding group includes multiple coding unit column, the adjacent encoder unit of different lines coding unit belongs to different volumes
Code character.
Wherein, the adjacent encoder unit of first row coding unit includes the left adjacent encoder of the affiliated code set of the coding unit
The coding unit in coding unit and the affiliated code set of the coding unit in group.The adjacent encoder of last column coding unit
Unit includes the right adjacent encoder group of the coding unit and the affiliated code set of the coding unit in the affiliated code set of the coding unit
In coding unit.The adjacent encoder unit of middle column coding unit is the coding list in the affiliated code set of the coding unit
Member.
If step 203, current coded unit are in the first row of present encoding group, from the left adjacent encoder of present encoding group
The left adjacent encoder unit of current coded unit, the encoded content of upper left adjacent encoder unit are obtained in group, from present encoding group
The encoded content of the middle upper adjacent encoder unit for obtaining current coded unit, upper right adjacent encoder unit, carries out predictive coding；
In the embodiment of the present invention, before being encoded to present encoding group, according in the first row of left adjacent encoder group
Rightmost side coding unit initializes entropy coder.Specifically, being compiled the arithmetic coding content of the rightmost side coding unit as entropy
The initial value of code device.Such as Fig. 2A, after the rightmost side coding unit 13 in the first row of the first code set completes coding, starting
The coding of second code set, and using the arithmetic coding content of coding unit 13 as the initial of the entropy coder of the second code set
Value.
Such as Fig. 2A, predictive coding is carried out to the coding unit 24 in the first row L4 in the second code set, needs to rely on phase
Adjacent coding unit, comprising: left adjacent encoder unit 23, upper left adjacent encoder unit 13, upper adjacent encoder unit 14, upper right are adjacent
Coding unit 15.Wherein, left adjacent encoder unit 23, upper left adjacent encoder unit 13 belong to the first code set.Upper adjacent encoder
Unit 14, upper right adjacent encoder unit 15 belong to the second code set.
It according to above-mentioned dependence, needs after coding unit 23 completes coding, starts the coding to coding unit 24.
Due to being encoded from left to right by row in same code set, so when coding unit 23 is completed to encode, coding unit 13
It has completed to encode, coding is completed in coding unit 14, coding unit 15 when starting the coding of coding unit 24.
As can be seen that the predictive coding to the second code set, depends on left the first code set of adjacent sets.
It is appreciated that when present encoding group is the leftmost side code set of image, at this point, first for present encoding group
Column coding unit carries out predictive coding according to upper adjacent encoder unit, upper right adjacent encoder unit.
If step 204, current coded unit are in last column of present encoding group, from the adjacent volume in the right side of present encoding group
The encoded content that the upper right adjacent encoder unit of current coded unit is obtained in code character, obtains present encoding from present encoding group
The left adjacent encoder unit of unit, upper left adjacent encoder unit, upper adjacent encoder unit encoded content, carry out predictive coding；
Such as Fig. 2A, predictive coding is carried out to the coding unit 26 in last column L6 in the second code set, needs to rely on
Adjacent encoder unit, comprising: left adjacent encoder unit 25, upper left adjacent encoder unit 15, upper adjacent encoder unit 16, upper right phase
Adjacent coding unit 17.Wherein, left adjacent encoder unit 25, upper left adjacent encoder unit 15, upper adjacent encoder unit 16 belong to
Two code sets, upper right adjacent encoder unit 17 belong to third code set.
It according to above-mentioned dependence, needs after coding unit 17 encodes completion, starts the coding of coding unit 26.By
In in same code set, encoded from left to right by row, thus when starting the coding of coding unit 26 coding unit 15,
Coding is completed in coding unit 16, coding unit 25.
As can be seen that the predictive coding to the second code set, depends on right adjacent sets third code set.
It is appreciated that when present encoding group is the rightmost side code set of image, at this point, for the last of present encoding group
One column coding unit carries out predictive coding according to left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit.
If step 205, current coded unit in the inner array of present encoding group, obtain current compile from present encoding group
The code left adjacent encoder unit of unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right adjacent encoder unit coding
Content carries out predictive coding；
Such as Fig. 2A, predictive coding is carried out to the coding unit 25 in the middle column L5 in the second code set, needs to rely on phase
Adjacent coding unit, comprising: left adjacent encoder unit 24, upper left adjacent encoder unit 14, upper adjacent encoder unit 15, upper right are adjacent
Coding unit 16.Wherein, left adjacent encoder unit 24, upper left adjacent encoder unit 14, upper adjacent encoder unit 15, upper right are adjacent
Coding unit 16 belongs to the second code set.
According to above-mentioned dependence, need to start the volume to coding unit 25 after the coding of coding unit 24 is completed
Code.Due to being encoded from left to right by row in same code set, so coding is single when starting the coding of coding unit 25
Coding is completed in member 14, coding unit 15, coding unit 24.
As can be seen that the predictive coding to the second code set, dependent on other coding units in the second code set.
It is appreciated that for the coding unit for being in image border, since there is no part adjacent encoder unit, so root
According to existing part adjacent encoder unit, predictive coding is carried out.
In conclusion the predictive coding to present encoding group, dependent on adjacent with present encoding group in left adjacent encoder group
Coding unit, coding list adjacent with present encoding group in the coding unit in present encoding group and right adjacent encoder group
Member.
Step 206 encodes the code stream length of each code set；
After encoding to each code set, code stream is formed.The prior art, the present invention can be used to the coding of code stream length
Embodiment to the encryption algorithm of code stream length with no restrictions.
Step 207, by the code stream length after coding according to code set from left to right be sequentially written in total code stream；
Wherein, it before total code stream being written in code stream length, needs to initialize total code stream, sets the initial value of total code stream to
It is empty.
Total code stream is written into code stream length in sequence, so that positioning the start bit of each code set when parallel decoding
It sets.
Step 208, by the corresponding code stream of each code set according to code set from left to right be sequentially written in total code stream.
Wherein, the sequence consensus of each code set corresponding code stream sequence and code stream length.As total code stream in Fig. 2A can be with
For (the code stream length of the first code set, the code stream length of the second code set, the code stream length of third code set, the first code set
Code stream, the code stream of the second code set, the code stream of third code set).
The coding unit divided by column of image can be M code set by the present embodiment, and start M thread to each coding
Group carries out the coding predicted based on adjacent column between group, it is possible to reduce the initialization times and coding ending of encoder, Ke Yi
Predictive coding is carried out between the adjacent column of adjacent encoder group, so that code efficiency is improved, further, since the embodiment of the present invention is not
Divided by row code set, so the first row coding unit in the region started there is no non-the first row, is not available negative movement
The problem of vector is predicted, to improve code efficiency.
And it can be adjacent with present encoding group according in the left adjacent encoder group and right adjacent encoder group of present encoding group
Coding unit carries out predictive coding, to improve code efficiency.
Embodiment three
Referring to Fig. 4, a kind of step flow chart of parallel encoding embodiment of the method three according to the present invention is shown, specifically may be used
To include the following steps:
The coding unit of image is evenly dividing by column as M code set by step 301, and the M code set corresponds to M
Thread；
Step 302, the position for judging current coded unit locating column in present encoding group；
If step 303, current coded unit are in the first row of present encoding group, from the left adjacent encoder of present encoding group
In group, in the coding that obtains the lower-left adjacent encoder unit of current coded unit, left adjacent encoder unit, upper left coding unit
Hold, from present encoding group, obtains the upper adjacent encoder unit of current coded unit, in the coding of upper right adjacent encoder unit
Hold, carries out predictive coding；
Compared with embodiment two, the present embodiment uses prediction mode between another different group.
In the embodiment of the present invention, before being encoded to present encoding group, according in the second row of left adjacent encoder group
Rightmost side coding unit initializes entropy coder.Specifically, being compiled the arithmetic coding content of the rightmost side coding unit as entropy
The initial value of code device.In Fig. 2A, after the rightmost side coding unit 23 in the second row of the first code set completes coding, open
The coding of dynamic second code set, and using the arithmetic coding content of coding unit 23 as the initial of the entropy coder of the second code set
Value.
Such as Fig. 2A, predictive coding is carried out to the coding unit 24 in the first row L4 in the second code set, needs to rely on phase
Adjacent coding unit, comprising: lower-left adjacent encoder unit 33, left adjacent encoder unit 23, upper left coding unit 13, upper adjacent encoder
Unit 14, upper right adjacent encoder unit 15.Wherein, lower-left adjacent encoder unit 33, left adjacent encoder unit 23, upper left coding are single
Member 13 belongs to the first code set.Upper adjacent encoder unit 14, upper right adjacent encoder unit 15 belong to the second code set.
It according to above-mentioned dependence, needs after coding unit 33 completes coding, starts the coding to coding unit 24.
Due to being encoded from left to right by row in same code set, so when coding unit 33 is completed to encode, coding unit
13, it 23 has completed to encode, coding is completed in coding unit 14, coding unit 15 when starting the coding of coding unit 24.
As can be seen that the predictive coding to the second code set, depends on left the first code set of adjacent sets.
It is appreciated that when present encoding group is the leftmost side code set of image, at this point, first for present encoding group
Column coding unit carries out predictive coding according to upper adjacent encoder unit, upper right adjacent encoder unit.
If step 304, current coded unit are in last column of present encoding group, from present encoding group, acquisition is worked as
Left adjacent encoder unit, upper left adjacent encoder unit, the upper adjacent encoder unit of preceding coding unit carry out predictive coding；
Such as Fig. 2A, predictive coding is carried out to the coding unit 26 in last column L6 in the second code set, needs to rely on
Adjacent encoder unit, comprising: left adjacent encoder unit 25, upper left adjacent encoder unit 15, upper adjacent encoder unit 16.Wherein,
Left adjacent encoder unit 25, upper left adjacent encoder unit 15, upper adjacent encoder unit 16 belong to the second code set.
It according to above-mentioned dependence, needs after coding unit 25 encodes completion, starts the coding of coding unit 26.By
In in same code set, encoded from left to right by row, thus when starting the coding of coding unit 26 coding unit 15,
Coding is completed in coding unit 16, coding unit 25.
As can be seen that the predictive coding to the second code set, independent of right adjacent sets third code set.
If step 305, current coded unit are in the inner array of present encoding group, from present encoding group, obtain current
Left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, the upper right adjacent encoder unit of coding unit carry out
Predictive coding；
In conclusion the predictive coding to present encoding group, dependent on adjacent with present encoding group in left adjacent encoder group
Coding unit and present encoding group in coding unit.
Step 306, by the corresponding code stream of each code set according to code set from left to right be sequentially written in total code stream.
The coding unit divided by column of image can be M code set by the present embodiment, and start M thread to each coding
Group carries out the coding predicted based on adjacent column between group, it is possible to reduce the initialization times and coding ending of encoder, Ke Yi
Predictive coding is carried out between the adjacent column of adjacent encoder group, so that code efficiency is improved, further, since the embodiment of the present invention is not
Divided by row code set, so the first row coding unit in the region started there is no non-the first row, is not available negative movement
The problem of vector is predicted, to improve code efficiency.
And can be according in the left adjacent encoder group of present encoding group, the coding unit adjacent with present encoding group carries out pre-
Coding is surveyed, to improve code efficiency.
Example IV
Referring to Fig. 5, a kind of structural block diagram of parallel encoding Installation practice four according to the present invention is shown, it specifically can be with
Including following module:
Division module 401, for being evenly dividing the coding unit of image by column for M code set, the M code set
Corresponding M thread；
Coding module 402, for start each thread between corresponding code set carry out based on group adjacent column predict it is parallel
Coding.
Preferably, coding module 402 includes:
First encoding submodule, for each coding unit for present encoding group, according to the coding unit current
The encoded content of code set, adjacent encoder unit in the adjacent encoder group of present encoding group starts each thread by row from left-hand
Right carry out predictive coding.
The coding unit divided by column of image can be M code set by the present embodiment, and start M thread to each coding
Group carries out the coding predicted based on adjacent column between group, it is possible to reduce the initialization times and coding ending of encoder, Ke Yi
Predictive coding is carried out between the adjacent column of adjacent encoder group, so that code efficiency is improved, further, since the embodiment of the present invention is not
Divided by row code set, so the first row coding unit in the region started there is no non-the first row, is not available negative movement
The problem of vector is predicted, to improve code efficiency.
Embodiment five
Referring to Fig. 6, a kind of structural block diagram of parallel encoding Installation practice five according to the present invention is shown, it specifically can be with
Including following module:
Division module 501, for being evenly dividing the coding unit of image by column for M code set, the M code set
Corresponding M thread；
Coding module 502, for start each thread between corresponding code set carry out based on group adjacent column predict it is parallel
Coding；May include:
First judging unit 5021, for judging the position of current coded unit locating column in present encoding group；
First coding unit 5022, if for current coded unit present encoding group first row, from present encoding
In the coding for obtaining the left adjacent encoder unit of current coded unit, upper left adjacent encoder unit in the left adjacent encoder group of group
Hold, the upper adjacent encoder unit of current coded unit, the encoded content of upper right adjacent encoder unit obtained from present encoding group,
Carry out predictive coding；
Second coding unit 5023, if being compiled in last column of present encoding group from current for current coded unit
The encoded content that the upper right adjacent encoder unit of current coded unit is obtained in the right adjacent encoder group of code character, from present encoding group
The middle encoded content for obtaining the left adjacent encoder unit of current coded unit, upper left adjacent encoder unit, upper adjacent encoder unit,
Carry out predictive coding；
Third coding unit 5024, if for current coded unit present encoding group inner array, from present encoding
The left adjacent encoder unit of acquisition current coded unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right are adjacent in group
The encoded content of coding unit carries out predictive coding.
First code stream VLC module 503 is encoded for the code stream length to each code set；
First code stream length writing module 504, for the code stream length after encoding according to code set from left to right suitable
Total code stream is written in sequence；
First code stream writing module 505, for the sequence by the corresponding code stream of each code set according to code set from left to right
Total code stream is written.
The coding unit divided by column of image can be M code set by the present embodiment, and start M thread to each coding
Group carries out the coding predicted based on adjacent column between group, it is possible to reduce the initialization times and coding ending of encoder, Ke Yi
Predictive coding is carried out between the adjacent column of adjacent encoder group, so that code efficiency is improved, further, since the embodiment of the present invention is not
Divided by row code set, so the first row coding unit in the region started there is no non-the first row, is not available negative movement
The problem of vector is predicted, to improve code efficiency.
And it can be adjacent with present encoding group according in the left adjacent encoder group and right adjacent encoder group of present encoding group
Coding unit carries out predictive coding, to improve code efficiency.
Embodiment six
Referring to Fig. 7, a kind of structural block diagram of parallel encoding Installation practice six according to the present invention is shown, it specifically can be with
Including following module:
Division module 601, for being evenly dividing the coding unit of image by column for M code set, the M code set
Corresponding M thread；
Coding module 602, for start each thread between corresponding code set carry out based on group adjacent column predict it is parallel
Coding；May include:
Second judgment unit 6021, for judging the position of current coded unit locating column in present encoding group；
4th coding unit 6022, if for current coded unit present encoding group first row, from present encoding
In the left adjacent encoder group of group, lower-left adjacent encoder unit, left adjacent encoder unit, the upper left coding of current coded unit are obtained
The encoded content of unit obtains upper adjacent encoder unit, the upper right adjacent encoder list of current coded unit from present encoding group
The encoded content of member carries out predictive coding；
5th coding unit 6023, if being compiled in last column of present encoding group from current for current coded unit
In code character, the left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit for obtaining current coded unit are carried out in advance
Survey coding；
6th coding unit 6024, if for current coded unit present encoding group inner array, from present encoding
In group, the left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right phase of current coded unit are obtained
Adjacent coding unit carries out predictive coding.
Second code stream writing module 603, for the sequence by the corresponding code stream of each code set according to code set from left to right
Total code stream is written.
The coding unit divided by column of image can be M code set by the present embodiment, and start M thread to each coding
Group carries out the coding predicted based on adjacent column between group, it is possible to reduce the initialization times and coding ending of encoder, Ke Yi
Predictive coding is carried out between the adjacent column of adjacent encoder group, so that code efficiency is improved, further, since the embodiment of the present invention is not
Divided by row code set, so the first row coding unit in the region started there is no non-the first row, is not available negative movement
The problem of vector is predicted, to improve code efficiency.
And can be according in the left adjacent encoder group of present encoding group, the coding unit adjacent with present encoding group carries out pre-
Coding is surveyed, to improve code efficiency.
For embodiment of the method, for simple description, therefore, it is stated as a series of action combinations, but this field
Technical staff should be aware of, and embodiment of that present invention are not limited by the describe sequence of actions, because implementing according to the present invention
Example, some steps may be performed in other sequences or simultaneously.Secondly, those skilled in the art should also know that, specification
Described in embodiment belong to preferred embodiment, the actions involved are not necessarily necessary for embodiments of the present invention.
For device embodiment, since it is basically similar to the method embodiment, related so being described relatively simple
Place illustrates referring to the part of embodiment of the method.
Algorithm and display are not inherently related to any particular computer, virtual system, or other device provided herein.
Various general-purpose systems can also be used together with teachings based herein.As described above, it constructs required by this kind of system
Structure be obvious.In addition, the present invention is also not directed to any particular programming language.It should be understood that can use various
Programming language realizes summary of the invention described herein, and the description done above to language-specific is to disclose this hair
Bright preferred forms.
In the instructions provided here, numerous specific details are set forth.It is to be appreciated, however, that implementation of the invention
Example can be practiced without these specific details.In some instances, well known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this specification.
Similarly, it should be understood that in order to simplify the disclosure and help to understand one or more of the various inventive aspects,
Above in the description of exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the disclosed method should not be interpreted as reflecting the following intention: i.e. required to protect
Shield the present invention claims features more more than feature expressly recited in each claim.More precisely, as following
Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following specific embodiment are expressly incorporated in the specific embodiment, wherein each claim itself
All as a separate embodiment of the present invention.
Those skilled in the art will understand that can be carried out adaptively to the module in the equipment in embodiment
Change and they are arranged in one or more devices different from this embodiment.It can be the module or list in embodiment
Member or component are combined into a module or unit or component, and furthermore they can be divided into multiple submodule or subelement or
Sub-component.Other than such feature and/or at least some of process or unit exclude each other, it can use any
Combination is to all features disclosed in this specification (including adjoint claim, abstract and attached drawing) and so disclosed
All process or units of what method or apparatus are combined.Unless expressly stated otherwise, this specification is (including adjoint power
Benefit requires, abstract and attached drawing) disclosed in each feature can be by providing identical, equivalent, or similar purpose alternative features come generation
It replaces.
In addition, it will be appreciated by those of skill in the art that although some embodiments described herein include other embodiments
In included certain features rather than other feature, but the combination of the feature of different embodiments mean it is of the invention
Within the scope of and form different embodiments.For example, in the following claims, embodiment claimed is appointed
Meaning one of can in any combination mode come using.
Various component embodiments of the invention can be implemented in hardware, or to run on one or more processors
Software module realize, or be implemented in a combination thereof.It will be understood by those of skill in the art that can be used in practice
Microprocessor or digital signal processor (DSP) come realize some in parallel encoding equipment according to an embodiment of the present invention or
The some or all functions of person's whole component.The present invention is also implemented as one for executing method as described herein
Point or whole device or device programs (for example, computer program and computer program product).Such this hair of realization
Bright program can store on a computer-readable medium, or may be in the form of one or more signals.It is such
Signal can be downloaded from an internet website to obtain, and is perhaps provided on the carrier signal or is provided in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and ability
Field technique personnel can be designed alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between parentheses should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" located in front of the element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of several different elements and being come by means of properly programmed computer real
It is existing.In the unit claims listing several devices, several in these devices can be through the same hardware branch
To embody.The use of word first, second, and third does not indicate any sequence.These words can be explained and be run after fame
Claim.
Claims (6)
1. a kind of parallel encoding method characterized by comprising
The coding unit of image is evenly dividing by column as M code set, the M code set corresponds to M thread；
Start each thread and the parallel encoding predicted based on adjacent column group is carried out between corresponding code set；
Wherein, the step of each thread of starting carries out the parallel encoding predicted based on adjacent column group to corresponding code set,
Include:
For each coding unit of present encoding group, according to the coding unit present encoding group, present encoding group it is adjacent
The encoded content of adjacent encoder unit in code set starts each thread by row and carries out predictive coding from left to right；
Wherein, each coding unit for present encoding group, according to the coding unit in present encoding group, present encoding
The encoded content of adjacent encoder unit in the adjacent encoder group of group starts each thread by row and carries out predictive coding from left to right
Step, comprising:
Judge the position of current coded unit locating column in present encoding group；
If current coded unit in the first row of present encoding group, obtains current from the left adjacent encoder group of present encoding group
The encoded content of the left adjacent encoder unit of coding unit, upper left adjacent encoder unit obtains current compile from present encoding group
The encoded content of the code upper adjacent encoder unit of unit, upper right adjacent encoder unit, carries out predictive coding；
If current coded unit obtains from the right adjacent encoder group of present encoding group and works as in last column of present encoding group
The encoded content of the upper right adjacent encoder unit of preceding coding unit, the left side that current coded unit is obtained from present encoding group are adjacent
Coding unit, upper left adjacent encoder unit, upper adjacent encoder unit encoded content, carry out predictive coding；
If current coded unit in the inner array of present encoding group, obtains the left phase of current coded unit from present encoding group
Adjacent coding unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right adjacent encoder unit encoded content, carry out pre-
Survey coding.
2. the method according to claim 1, wherein further include:
The code stream length of each code set is encoded；
By the code stream length after coding according to code set from left to right be sequentially written in total code stream；
By the corresponding code stream of each code set according to code set from left to right be sequentially written in total code stream.
3. a kind of parallel encoding method characterized by comprising
The coding unit of image is evenly dividing by column as M code set, the M code set corresponds to M thread；
Start each thread and the parallel encoding predicted based on adjacent column group is carried out between corresponding code set；
Wherein, the step of each thread of starting carries out the parallel encoding predicted based on adjacent column group to corresponding code set,
Include:
For each coding unit of present encoding group, according to the coding unit present encoding group, present encoding group it is adjacent
The encoded content of adjacent encoder unit in code set starts each thread by row and carries out predictive coding from left to right；
Wherein, each coding unit for present encoding group, according to the coding unit in present encoding group, present encoding
The encoded content of adjacent encoder unit in the adjacent encoder group of group starts each thread by row and carries out predictive coding from left to right
Step, comprising:
Judge the position of current coded unit locating column in present encoding group；
If current coded unit is in the first row of present encoding group, from the left adjacent encoder group of present encoding group, acquisition is worked as
The lower-left adjacent encoder unit of preceding coding unit, left adjacent encoder unit, upper left adjacent encoder unit encoded content, from current
In code set, the upper adjacent encoder unit of current coded unit, the encoded content of upper right adjacent encoder unit are obtained, is predicted
Coding；
If current coded unit from present encoding group, obtains current coded unit in last column of present encoding group
Left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit carry out predictive coding；
If current coded unit is in the inner array of present encoding group, from present encoding group, a left side for current coded unit is obtained
Adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, upper right adjacent encoder unit carry out predictive coding；
By the corresponding code stream of each code set according to code set from left to right be sequentially written in total code stream.
4. a kind of parallel encoding device characterized by comprising
Division module, for being evenly dividing the coding unit of image by column for M code set, the M code set corresponds to M
Thread；
Coding module carries out the parallel encoding predicted based on adjacent column group to corresponding code set for starting each thread；
Wherein, the coding module, comprising:
First encoding submodule, for each coding unit for present encoding group, according to the coding unit in present encoding
The encoded content of group, adjacent encoder unit in the adjacent encoder group of present encoding group, start each thread by row from left to right into
Row predictive coding；
Wherein, first encoding submodule, comprising:
First judging unit, for judging that the current coded unit position of locating column or second judgement in present encoding group are single
Member, for judging the position of current coded unit locating column in present encoding group；
First coding unit, if for current coded unit present encoding group first row, from the left phase of present encoding group
The encoded content that the left adjacent encoder unit of current coded unit, upper left adjacent encoder unit are obtained in adjacent code set, from current
The upper adjacent encoder unit of current coded unit, the encoded content of upper right adjacent encoder unit are obtained in code set, are predicted
Coding；
Second coding unit, if for current coded unit present encoding group last column, from the right side of present encoding group
The encoded content that the upper right adjacent encoder unit of current coded unit is obtained in adjacent encoder group, obtains from present encoding group and works as
The left adjacent encoder unit of preceding coding unit, upper left adjacent encoder unit, upper adjacent encoder unit encoded content, predicted
Coding；
Third coding unit, if being obtained from present encoding group for current coded unit in the inner array of present encoding group
Left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, the upper right adjacent encoder unit of current coded unit
Encoded content, carry out predictive coding.
5. device according to claim 4, which is characterized in that further include:
First code stream VLC module is encoded for the code stream length to each code set；
First code stream length writing module, for the code stream length after coding is total according to code set being sequentially written in from left to right
Code stream；
First code stream writing module, for by the corresponding code stream of each code set according to code set from left to right be sequentially written in total code
Stream.
6. a kind of parallel encoding device characterized by comprising
Division module, for being evenly dividing the coding unit of image by column for M code set, the M code set corresponds to M
Thread；
Coding module carries out the parallel encoding predicted based on adjacent column group to corresponding code set for starting each thread；
Wherein, the coding module, comprising:
First encoding submodule, for each coding unit for present encoding group, according to the coding unit in present encoding
The encoded content of group, adjacent encoder unit in the adjacent encoder group of present encoding group, start each thread by row from left to right into
Row predictive coding；
Wherein, first encoding submodule, comprising:
Second judgment unit, for judging the position of current coded unit locating column in present encoding group；
4th coding unit, if for current coded unit present encoding group first row, from the left phase of present encoding group
In adjacent code set, the lower-left adjacent encoder unit, left adjacent encoder unit, upper left adjacent encoder unit of current coded unit are obtained
Encoded content obtain the upper adjacent encoder unit of current coded unit, upper right adjacent encoder unit from present encoding group
Encoded content carries out predictive coding；
5th coding unit, if from present encoding group, being obtained for current coded unit in last column of present encoding group
The left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit of current coded unit is taken to carry out predictive coding；
6th coding unit, if the inner array for current coded unit in present encoding group is obtained from present encoding group
Left adjacent encoder unit, upper left adjacent encoder unit, upper adjacent encoder unit, the upper right adjacent encoder unit of current coded unit
Carry out predictive coding；
Second code stream writing module, for by the corresponding code stream of each code set according to code set from left to right be sequentially written in total code
Stream.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610425871.7A CN105939479B (en) | 2016-06-15 | 2016-06-15 | A kind of parallel encoding method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610425871.7A CN105939479B (en) | 2016-06-15 | 2016-06-15 | A kind of parallel encoding method and apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105939479A CN105939479A (en) | 2016-09-14 |
| CN105939479B true CN105939479B (en) | 2019-06-21 |
Family
ID=56872615
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610425871.7A Active CN105939479B (en) | 2016-06-15 | 2016-06-15 | A kind of parallel encoding method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105939479B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101889449A (en) * | 2007-06-28 | 2010-11-17 | 三菱电机株式会社 | Image encoder and image decoder |
| CN102006469A (en) * | 2010-08-24 | 2011-04-06 | 吉林大学 | Three-dimensional element image based multi-level mixed predictive coding structure parallel implementation method |
| CN102223534A (en) * | 2011-06-03 | 2011-10-19 | 东南大学 | All-parallel bit plane coding method for image compression |
| CN103797793A (en) * | 2011-07-15 | 2014-05-14 | 弗兰霍菲尔运输应用研究公司 | Sample array coding for low-delay |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130003829A1 (en) * | 2011-07-01 | 2013-01-03 | Kiran Misra | System for initializing an arithmetic coder |
-
2016
- 2016-06-15 CN CN201610425871.7A patent/CN105939479B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101889449A (en) * | 2007-06-28 | 2010-11-17 | 三菱电机株式会社 | Image encoder and image decoder |
| CN102006469A (en) * | 2010-08-24 | 2011-04-06 | 吉林大学 | Three-dimensional element image based multi-level mixed predictive coding structure parallel implementation method |
| CN102223534A (en) * | 2011-06-03 | 2011-10-19 | 东南大学 | All-parallel bit plane coding method for image compression |
| CN103797793A (en) * | 2011-07-15 | 2014-05-14 | 弗兰霍菲尔运输应用研究公司 | Sample array coding for low-delay |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105939479A (en) | 2016-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3343355B1 (en) | Dimension shuffling using matrix processors | |
| US10198401B2 (en) | Max pooling in a matrix processing architecture | |
| EP3346391A1 (en) | Distributed convolution for neural networks | |
| KR101797845B1 (en) | Parallel video processing apparatus using multicore system and method thereof | |
| CN103020890B (en) | Based on the visual processing apparatus of multi-level parallel processing | |
| CN101068364B (en) | Video encoder and graph processing unit | |
| CN104933747B (en) | Vector animation is converted to the method and device of bitmaps animation | |
| WO2009033152A3 (en) | Real-time video coding/decoding | |
| CN106034236B (en) | A kind of selection method, device and the encoder of HEVC coding optimal reference frame | |
| CN106921863A (en) | Use the method for multiple decoder core decoding video bit streams, device and processor | |
| US9667985B1 (en) | Sweep dependency based graphics processing unit block scheduling | |
| CN109544439B (en) | Decoding method based on multi-core processor, terminal equipment and storage medium | |
| CN104375804A (en) | Multiple threads execution processor and operating method thereof | |
| US10706365B2 (en) | Local optimization of quantum circuits | |
| CN102857758B (en) | Reusable pixel processing method and reusable video processing chip | |
| CN105939479B (en) | A kind of parallel encoding method and apparatus | |
| JP2016123089A (en) | Pad encoding and decoding | |
| CN103686191A (en) | Method of processing multi-view image and apparatus for executing same | |
| CN105874800A (en) | Syntax parsing apparatus with multiple syntax parsing circuits for processing multiple image regions within same frame or processing multiple frames and related syntax parsing method | |
| EP3552396A1 (en) | Lightfield compression using disparity predicted replacement | |
| WO2020095155A1 (en) | Large model support in deep learning | |
| TW201534107A (en) | Non-sequential run length encoding system and method | |
| CN106231307B (en) | A kind of compression of images intra-coding prediction method and its hardware realization | |
| CN115082371B (en) | Image fusion method and device, mobile terminal equipment and readable storage medium | |
| CN103533331B (en) | Encoding and decoding method and device for image |
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 |