CN104079937A - H.264-to-HEVC fast inter-frame transcoding method and device based on motion vector analysis - Google Patents

H.264-to-HEVC fast inter-frame transcoding method and device based on motion vector analysis Download PDF

Info

Publication number
CN104079937A
CN104079937A CN201410340334.3A CN201410340334A CN104079937A CN 104079937 A CN104079937 A CN 104079937A CN 201410340334 A CN201410340334 A CN 201410340334A CN 104079937 A CN104079937 A CN 104079937A
Authority
CN
China
Prior art keywords
motion vector
coding unit
unit
macro block
block size
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.)
Granted
Application number
CN201410340334.3A
Other languages
Chinese (zh)
Other versions
CN104079937B (en
Inventor
张海霞
孙彬
刘治
尚蕾
金蕾
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong University
Original Assignee
Shandong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shandong University filed Critical Shandong University
Priority to CN201410340334.3A priority Critical patent/CN104079937B/en
Publication of CN104079937A publication Critical patent/CN104079937A/en
Application granted granted Critical
Publication of CN104079937B publication Critical patent/CN104079937B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention relates to an H.264-to-HEVC fast inter-frame transcoding method based on motion vector analysis. The H.264-to-HEVC fast inter-frame transcoding method based on motion vector analysis includes the concrete steps that (1) H.264 code streams are decoded by an H.264 decoder terminal, and reference frame information, macro-block sizes, macro-block positions, motion vectors (MV) of macro-blocks and prediction modes of the macro-blocks in the H.264 code streams are extracted; (2) the coding unit (CU) division mode is determined according to the motion vectors (MV) of the macro-blocks of the H.264 code streams obtained on the H.264 decoder terminal; (3) the division mode of prediction units (PU) is determined according to the partitioning of the coding units (CU); (4) the motion vectors of the prediction units (PU) in the current HEVC are directly predicted through motion vector information extracted from the H.264 code streams. Through the fast inter-frame transcoding method, relevant information in the H.264 code streams can be utilized to the maximum extent, the complexity of the HEVC coding end is well reduced, the HEVC coding time is shortened, and the coding efficiency is greatly improved.

Description

A kind of based on motion vector analysis by H.264 to code-transferring method and transcoding device the fast frame of HEVC
Technical field
The present invention relates to a kind of based on motion vector analysis by H.264 to code-transferring method and transcoding device the fast frame of HEVC, belong to video code conversion field.
Background technology
H.264/AVC current most widely used video encoding and decoding standard is, it is widely used in cable TV and satellite television, the wireless and field such as mobile network, the service of multimedia streaming.But along with the rise of the universal and ultra high-definition form of HD video, original H.264/AVC standard can not meet the demand of coding and decoding video in effect and speed well.Therefore, the ITU-T of International Organization for standardization and ISO/IEC have proposed a kind of high-performance video coding HEVC as video encoding standard of new generation.More H.264/AVC, HEVC has promoted 30% to 50% at coding and decoding video aspect of performance, but HEVC has paid higher calculation cost and more complicated storage organization at coding side simultaneously.And, nowadays H.264 broader applications in video camera, mobile phone, at a distance of etc. in the video capture device of main flow.HEVC cannot replace status H.264 in a short time, that is to say for a comparatively long period of time, and two kinds of video standards will coexist.Therefore, to H.26/AVC4 to the research of HEVC transcoding by contributing to two kinds of fusion and transition between video standard, have important practical significance.
H.264/AVC be the one conversion to video format to the process of HEVC transcoding, the most direct method is that the encoder of decoder H.264 and HEVC is carried out to cascade, and H.264 decoded video is directly inputted in HEVC encoder and encoded.But, owing to having introduced some algorithms of intensive and high complexity in HEVC encoder, therefore, utilize H.264 entrained partial information in code stream to be optimized the algorithm of this part amount of calculation and high complexity, just become a problem of research.
With respect to the macro block MB in coding standard H.264, the difference of HEVC is the code tree unit CTU that the introduced self adaptation quad-tree structure elementary cell as prediction, conversion, quantification, entropy coding.A code tree unit CTU comprises a luminance coding tree piece CTB and two chroma coder tree pieces.Its affiliated brightness and size and the position of chroma coder piece CB formulated in the quaternary tree syntax of code tree unit CTU, common luminance coding tree piece CTB and two chroma coders tree pieces and a coding unit CU of their relevant syntaxes composition.The full-size of coding unit CU is the size of the luminance coding tree piece CTB under it.The size of coding unit CU or the pixel count comprising can be set to 64 × 64,32 × 32,16 × 16 and 8 × 8.In general, coding unit CU size is larger, and compression performance is better.Each coding unit CU is comprising predicting unit PU associated therewith and converter unit TU.For predicting unit PU, can be divided into the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N.Compared with fixing the coding unit of macroblock size H.264, HEVC is more flexible to the division of coded image.But, in improving video coding performance, also increase the complexity of cataloged procedure.
H.264, traditional is by the H.264 encoder cascade of decoder and HEVC to the method for HEVC transcoding, directly the image obtaining of decoding of standard is H.24 carried out to HEVC standard code.This method has traveled through all processes in coding, and complexity is higher, the information comprising in code stream is not H.264 made full use of.
At present comparatively comprehensive for the transcoding research between various video encoding standards in the past, also carry out for the research to HEVC transcoding H.264.The people such as Tamer Shanabieh and Eduardo Peixoto have proposed a kind of by linear discriminant function, between the characteristic information in code stream H.264 and the partitioned mode in HEVC coding standard, set up a mapping.Utilize the thought of pattern classification, extract the characteristic information in code stream H.264, directly determine the dividing mode of the coding unit of image in HEVC encoder.This method has reduced coding side and has determined the amount of calculation of partitioned mode, thereby has promoted code rate.
In addition the spectra calculation method that, the people such as Dong Zhang and Bin Li has proposed based on rate-distortion optimization model is predicted the CU of I picture and the macroblock mode of PU.Reduce little in the situation that at encoding efficiency, larger reduction the computation complexity of coding side.
Due to H.264 with the code stream of HEVC coding in, the prediction mode that reference frame number, motion vector, minimum movement are estimated etc. is identical.Therefore, can be H.264 extracting the information such as reference frame number, macro block position coordinate, macro block (mb) type, motion vector in code stream.And be applied in HEVC coding side, the link such as interframe and infra-frame prediction is optimized, thereby significantly reduces the complexity of whole transcoding process.
Summary of the invention
For deficiency of the prior art, the present invention announced a kind of based on motion vector analysis by H.264 to code-transferring method the fast frame of HEVC;
The present invention has also announced a kind of transcoding device of realizing above-mentioned code-transferring method.
Technical scheme of the present invention is:
Based on motion vector analysis by H.264 to a code-transferring method the fast frame of HEVC, concrete steps comprise:
(1) in decoder terminal H.264, code stream is H.264 decoded, extract reference frame information, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block in described H.264 code stream in decoder terminal;
(2) determine coding unit CU dividing mode according to the motion vector MV of the macro block at the H.264 code stream that H.264 decoder terminal obtains, i.e. the piecemeal of coding unit CU, concrete steps are as follows:
The initial block size of a, setting coding unit CU is 64 × 64; Calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 64 × 64 coding unit CU;
B, according to the motion vector MV that extracts the macro block described in the step a obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 64 × 64 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 64 × 64, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
C, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 64 × 64 is without dividing, and block size 64 × 64 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 64 × 64 divides processing, splits into 4 block sizes and is 32 × 32 coding unit CU, enters steps d;
D, the initial block size of setting coding unit CU are 32 × 32, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 32 × 32 coding unit CU;
E, according to the motion vector MV that extracts the macro block described in the steps d obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 32 × 32 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 32 × 32, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
F, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 32 × 32 is without dividing, and block size 32 × 32 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 32 × 32 divides processing, splits into the coding unit of 4 block sizes 16 × 16, enters step g;
G, the initial block size of setting coding unit CU are 16 × 16, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 16 × 16 coding unit CU;
H, according to the motion vector MV that extracts the described macro block of the step (g) that obtains in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 16 × 16 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 16 × 16, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
I, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 16 × 16 is without dividing, and block size 16 × 16 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 16 × 16 divides processing, splits into the coding unit of 4 block sizes 8 × 8, and block size 8 × 8 is now exactly coding unit CU dividing mode;
Work as D mvbe less than T 1time, the motion vector difference of each macro block comprising in the coding unit CU that block size is 64 × 64 is little, without dividing; D mvbe more than or equal to T 1time, the motion vector difference of each macro block comprising in the coding unit CU that block size is 64 × 64 is larger, divides processing;
(3) determine the dividing mode of predicting unit PU according to the piecemeal of coding unit CU, determine the dividing mode of predicting unit PU according to coding unit CU dividing mode, each block size is the dividing mode that the coding unit CU of 2N × 2N has 7 kinds of predicting unit PU, wherein, described N=4,8,16,32; The dividing mode of 7 kinds of described predicting unit PU is respectively the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N, wherein, the dividing mode of 7 kinds of described predicting unit PU is international standard, U is up, D is down, R is right, L is left, n is the mark of asymmetric mode, 2N × the nU with n, 2N × nD, nL × 2N, nR × 2N are asymmetric mode, and 2N × the 2N with n, 2N × N, N × 2N are not symmetric pattern; Described 2N × nU is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 1:3 from top to bottom; Described 2N × nD is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 3:1 from top to bottom; Described nL × 2N is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 1:3 from left to right; Described nR × 2N is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 3:1 from left to right; Concrete steps are as follows:
If the block size of j coding unit CU is, described coding unit CU is divided into the fritter that 4 block sizes are 4 × 4 at 8 × 8 o'clock; Otherwise, described coding unit CU is divided into 16 fritters that block size is identical;
K, according to the motion vector information obtaining in code stream H.264, extract the motion vector of each fritter in coding unit CU, and calculate the phase place of the motion vector of the fritter described in each ;
By the variance sum D of the phase place of each fritter in coding unit CU under the dividing mode of 7 kinds of predicting unit PU of following formula II calculating ψ, the dividing mode of 7 kinds of described predicting unit PU is the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N;
D ψ＝D ψ1+D ψ2 ⅱ
In formula II, ψ 1, ψ 2represent respectively two segmented areas in predicting unit PU dividing mode; wherein, (i, j) is ψ 1the transverse and longitudinal coordinate of piecemeal; (i ', j ') be ψ 2the transverse and longitudinal coordinate of piecemeal;
Relatively D in the dividing mode situation of 7 kinds of described predicting unit PU ψ, choose and make D ψthe value dividing mode of hour predicting unit PU is defined as the dividing mode of predicting unit PU;
(4) directly dope the motion vector of predicting unit PU in current HEVC by the motion vector information extracting in code stream H.264, concrete steps are:
L, basis at the motion vector of the H.264 code stream that H.264 decoder terminal extracts, find respectively the corresponding best matching blocks in reference frame unit of each fritter comprising in the predicting unit PU that determines dividing mode in step (l); Described best matching blocks refer to described predicting unit PU in the nearest reference frame unit of each fritter of comprising; Calculate respectively the variance distance B of each fritter comprising in reference frame unit and predicting unit PU by following formula III,
In formula III, for the position of t fritter in predicting unit PU; for the position of best matching blocks corresponding to t fritter in PU unit;
Wherein, with described predicting unit PU in the variance of each fritter of comprising be described best matching blocks apart from the reference frame unit of sum minimum;
If m obtains a best matching blocks by step (l), directly calculate skew between predicting unit PU and described best matching blocks and be the motion vector of described predicting unit PU; If obtain two or more best matching blocks by step (l), the summation absolute error criterion SAD of the predicting unit PU of calculative determination dividing mode and described two or more best matching blocks respectively, choose the best matching blocks of sad value minimum as best matching blocks, directly calculate skew between predicting unit PU and described best matching blocks and be the motion vector of described predicting unit PU.
A kind of realize based on motion vector analysis by the transcoding device to code-transferring method the fast frame of HEVC H.264, comprise: H.264 decoder module, motion vector information extraction module, storage module, computing module and HEVC coding module, described H.264 decoder module connects described motion vector information extraction module, described motion vector information extraction module connects described storage module, described storage module connects described computing module, and described computing module connects described HEVC coding module.
Preferred according to the present invention, described H.264 decoder module is decoded to code stream H.264 in decoder terminal H.264; Described motion vector information extraction module is for the reference frame information that extracts described H.264 code stream in decoder terminal H.264 and comprise, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block; Described storage module, for storing reference frame information that described H.264 code stream that motion vector information extraction module extracts comprises, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block; Described computing module is for the variance distance of the motion vector of interior each macro block comprising of calculation code unit CU, be used for the variance sum of the phase place of calculation code unit each fritter of CU, the variance of each fritter comprising for computing reference frame unit distance and predicting unit PU is apart from sum, also for calculating the skew between predicting unit PU and described best matching blocks; Described HEVC coding module, for according to the data result obtaining from motion vector information that H.264 code stream extracts and computing module, is optimized coding to present frame.
Beneficial effect of the present invention is:
1, quick inter prediction code-transferring method provided by the invention maximizes the use the relevant information in code stream H.264, has reduced well the complexity of HEVC coding side.
2, the present invention can reduce the time of HEVC coding, greatly improves code efficiency.
Brief description of the drawings:
Fig. 1 be the present invention a kind of realize based on motion vector analysis by the transcoding device structure principle chart to code-transferring method the fast frame of HEVC H.264.
Embodiment
Below in conjunction with Figure of description and embodiment, the invention will be further described, and following embodiment is only not used in and limits the scope of the invention for the present invention is described.
Embodiment 1
Based on motion vector analysis by H.264 to a code-transferring method the fast frame of HEVC, concrete steps comprise:
(1) in decoder terminal H.264, code stream is H.264 decoded, extract reference frame information, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block in described H.264 code stream in decoder terminal;
(2) determine coding unit CU dividing mode according to the motion vector MV of the macro block at the H.264 code stream that H.264 decoder terminal obtains, i.e. the piecemeal of coding unit CU, concrete steps are as follows:
The initial block size of a, setting coding unit CU is 64 × 64; Calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 64 × 64 coding unit CU;
B, according to the motion vector MV that extracts the macro block described in the step a obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 64 × 64 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 64 × 64, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
C, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 64 × 64 is without dividing, and block size 64 × 64 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 64 × 64 divides processing, splits into 4 block sizes and is 32 × 32 coding unit CU, enters steps d;
D, the initial block size of setting coding unit CU are 32 × 32, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 32 × 32 coding unit CU;
E, according to the motion vector MV that extracts the macro block described in the steps d obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 32 × 32 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 32 × 32, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
F, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 32 × 32 is without dividing, and block size 32 × 32 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 32 × 32 divides processing, splits into the coding unit of 4 block sizes 16 × 16, enters step g;
G, the initial block size of setting coding unit CU are 16 × 16, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 16 × 16 coding unit CU;
H, according to the motion vector MV that extracts the described macro block of the step (g) that obtains in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 16 × 16 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 16 × 16, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
I, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 16 × 16 is without dividing, and block size 16 × 16 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 16 × 16 divides processing, splits into the coding unit of 4 block sizes 8 × 8, and block size 8 × 8 is now exactly coding unit CU dividing mode;
Work as D mvbe less than T 1time, the motion vector difference of each macro block comprising in the coding unit CU that block size is 64 × 64 is little, without dividing; D mvbe more than or equal to T 1time, the motion vector difference of each macro block comprising in the coding unit CU that block size is 64 × 64 is larger, divides processing;
(3) determine the dividing mode of predicting unit PU according to the piecemeal of coding unit CU, determine the dividing mode of predicting unit PU according to coding unit CU dividing mode, each block size is the dividing mode that the coding unit CU of 2N × 2N has 7 kinds of predicting unit PU, wherein, described N=4,8,16,32; The dividing mode of 7 kinds of described predicting unit PU is respectively the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N, wherein, the dividing mode of 7 kinds of described predicting unit PU is international standard, U is up, D is down, R is right, L is left, n is the mark of asymmetric mode, 2N × the nU with n, 2N × nD, nL × 2N, nR × 2N are asymmetric mode, and 2N × the 2N with n, 2N × N, N × 2N are not symmetric pattern; Described 2N × nU is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 1:3 from top to bottom; Described 2N × nD is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 3:1 from top to bottom; Described nL × 2N is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 1:3 from left to right; Described nR × 2N is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 3:1 from left to right; Concrete steps are as follows:
If the block size of j coding unit CU is, described coding unit CU is divided into the fritter that 4 block sizes are 4 × 4 at 8 × 8 o'clock; Otherwise, described coding unit CU is divided into 16 fritters that block size is identical;
K, according to the motion vector information obtaining in code stream H.264, extract the motion vector of each fritter in coding unit CU, and calculate the phase place of the motion vector of the fritter described in each ;
By the variance sum D of the phase place of each fritter in coding unit CU under the dividing mode of 7 kinds of predicting unit PU of following formula II calculating ψ, the dividing mode of 7 kinds of described predicting unit PU is the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N;
D ψ＝D ψ1+D ψ2 ⅱ
In formula II, ψ 1, ψ 2represent respectively two segmented areas in predicting unit PU dividing mode; wherein, (i, j) is ψ 1the transverse and longitudinal coordinate of piecemeal; (i ', j ') be ψ 2the transverse and longitudinal coordinate of piecemeal;
Relatively D in the dividing mode situation of 7 kinds of described predicting unit PU ψ, choose and make D ψthe value dividing mode of hour predicting unit PU is defined as the dividing mode of predicting unit PU;
(4) directly dope the motion vector of predicting unit PU in current HEVC by the motion vector information extracting in code stream H.264, concrete steps are:
L, basis at the motion vector of the H.264 code stream that H.264 decoder terminal extracts, find respectively the corresponding best matching blocks in reference frame unit of each fritter comprising in the predicting unit PU that determines dividing mode in step (l); Described best matching blocks refer to described predicting unit PU in the nearest reference frame unit of each fritter of comprising; Calculate respectively the variance distance B of each fritter comprising in reference frame unit and predicting unit PU by following formula III,
In formula III, for the position of t fritter in predicting unit PU; for the position of best matching blocks corresponding to t fritter in PU unit;
Wherein, with described predicting unit PU in the variance of each fritter of comprising be described best matching blocks apart from the reference frame unit of sum minimum;
M, obtain a best matching blocks by step (l), directly calculate skew between predicting unit PU and described best matching blocks and be the motion vector of described predicting unit PU.
Embodiment 2
Based on motion vector analysis by H.264 to a code-transferring method the fast frame of HEVC, concrete steps comprise:
(1) in decoder terminal H.264, code stream is H.264 decoded, extract reference frame information, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block in described H.264 code stream in decoder terminal;
(2) determine coding unit CU dividing mode according to the motion vector MV of the macro block at the H.264 code stream that H.264 decoder terminal obtains, i.e. the piecemeal of coding unit CU, concrete steps are as follows:
The initial block size of a, setting coding unit CU is 64 × 64; Calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 64 × 64 coding unit CU;
B, according to the motion vector MV that extracts the macro block described in the step a obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 64 × 64 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 64 × 64, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
C, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 64 × 64 is without dividing, and block size 64 × 64 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 64 × 64 divides processing, splits into 4 block sizes and is 32 × 32 coding unit CU, enters steps d;
D, the initial block size of setting coding unit CU are 32 × 32, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 32 × 32 coding unit CU;
E, according to the motion vector MV that extracts the macro block described in the steps d obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 32 × 32 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 32 × 32, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
F, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 32 × 32 is without dividing, and block size 32 × 32 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 32 × 32 divides processing, splits into the coding unit of 4 block sizes 16 × 16, enters step g;
G, the initial block size of setting coding unit CU are 16 × 16, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 16 × 16 coding unit CU;
H, according to the motion vector MV that extracts the described macro block of the step (g) that obtains in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 16 × 16 mv,
D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 16 × 16, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
I, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 16 × 16 is without dividing, and block size 16 × 16 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 16 × 16 divides processing, splits into the coding unit of 4 block sizes 8 × 8, and block size 8 × 8 is now exactly coding unit CU dividing mode;
Work as D mvbe less than T 1time, the motion vector difference of each macro block comprising in the coding unit CU that block size is 64 × 64 is little, without dividing; D mvbe more than or equal to T 1time, the motion vector difference of each macro block comprising in the coding unit CU that block size is 64 × 64 is larger, divides processing;
(3) determine the dividing mode of predicting unit PU according to the piecemeal of coding unit CU, determine the dividing mode of predicting unit PU according to coding unit CU dividing mode, each block size is the dividing mode that the coding unit CU of 2N × 2N has 7 kinds of predicting unit PU, wherein, described N=4,8,16,32; The dividing mode of 7 kinds of described predicting unit PU is respectively the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N, wherein, the dividing mode of 7 kinds of described predicting unit PU is international standard, U is up, D is down, R is right, L is left, n is the mark of asymmetric mode, 2N × the nU with n, 2N × nD, nL × 2N, nR × 2N are asymmetric mode, and 2N × the 2N with n, 2N × N, N × 2N are not symmetric pattern; Described 2N × nU is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 1:3 from top to bottom; Described 2N × nD is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 3:1 from top to bottom; Described nL × 2N is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 1:3 from left to right; Described nR × 2N is asymmetric cuts apart the ratio cut partition predicting unit PU that pattern is 3:1 from left to right; Concrete steps are as follows:
If the block size of j coding unit CU is, described coding unit CU is divided into the fritter that 4 block sizes are 4 × 4 at 8 × 8 o'clock; Otherwise, described coding unit CU is divided into 16 fritters that block size is identical;
K, according to the motion vector information obtaining in code stream H.264, extract the motion vector of each fritter in coding unit CU, and calculate the phase place of the motion vector of the fritter described in each ;
By the variance sum D of the phase place of each fritter in coding unit CU under the dividing mode of 7 kinds of predicting unit PU of following formula II calculating ψ, the dividing mode of 7 kinds of described predicting unit PU is the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N;
D ψ＝D ψ1+D ψ2 ⅱ
In formula II, ψ 1, ψ 2represent respectively two segmented areas in predicting unit PU dividing mode; wherein, (i, j) is ψ 1the transverse and longitudinal coordinate of piecemeal; (i ', j ') be ψ 2the transverse and longitudinal coordinate of piecemeal;
Relatively D in the dividing mode situation of 7 kinds of described predicting unit PU ψ, choose and make D ψthe value dividing mode of hour predicting unit PU is defined as the dividing mode of predicting unit PU;
(4) directly dope the motion vector of predicting unit PU in current HEVC by the motion vector information extracting in code stream H.264, concrete steps are:
L, basis at the motion vector of the H.264 code stream that H.264 decoder terminal extracts, find respectively the corresponding best matching blocks in reference frame unit of each fritter comprising in the predicting unit PU that determines dividing mode in step (l); Described best matching blocks refer to described predicting unit PU in the nearest reference frame unit of each fritter of comprising; Calculate respectively the variance distance B of each fritter comprising in reference frame unit and predicting unit PU by following formula III,
In formula III, for the position of t fritter in predicting unit PU; for the position of best matching blocks corresponding to t fritter in PU unit;
Wherein, with described predicting unit PU in the variance of each fritter of comprising be described best matching blocks apart from the reference frame unit of sum minimum;
M, obtain two or more best matching blocks by step (l), the summation absolute error criterion SAD of the predicting unit PU of calculative determination dividing mode and described two or more best matching blocks respectively, choose the best matching blocks of sad value minimum as best matching blocks, directly calculate skew between predicting unit PU and described best matching blocks and be the motion vector of described predicting unit PU.
Embodiment 3
A kind of realize based on motion vector analysis by the transcoding device to code-transferring method the fast frame of HEVC H.264, comprise: H.264 decoder module, motion vector information extraction module, storage module, computing module and HEVC coding module, described H.264 decoder module connects described motion vector information extraction module, described motion vector information extraction module connects described storage module, described storage module connects described computing module, and described computing module connects described HEVC coding module.
Described H.264 decoder module is decoded to code stream H.264 in decoder terminal H.264; Described motion vector information extraction module is for the reference frame information that extracts described H.264 code stream in decoder terminal H.264 and comprise, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block; Described storage module, for storing reference frame information that described H.264 code stream that motion vector information extraction module extracts comprises, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block; Described computing module is for the variance distance of the motion vector of interior each macro block comprising of calculation code unit CU, be used for the variance sum of the phase place of calculation code unit each fritter of CU, the variance of each fritter comprising for computing reference frame unit distance and predicting unit PU is apart from sum, also for calculating the skew between predicting unit PU and described best matching blocks; Described HEVC coding module, for according to the data result obtaining from motion vector information that H.264 code stream extracts and computing module, is optimized coding to present frame.

Claims (3)

  1. Based on motion vector analysis by H.264 to a code-transferring method the fast frame of HEVC, it is characterized in that, concrete steps comprise:
    (1) in decoder terminal H.264, code stream is H.264 decoded, extract reference frame information, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block in described H.264 code stream in decoder terminal;
    (2) determine coding unit CU dividing mode according to the motion vector MV of the macro block at the H.264 code stream that H.264 decoder terminal obtains, i.e. the piecemeal of coding unit CU, concrete steps are as follows:
    The initial block size of a, setting coding unit CU is 64 × 64; Calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 64 × 64 coding unit CU;
    B, according to the motion vector MV that extracts the macro block described in the step a obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 64 × 64 mv,
    D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
    In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 64 × 64, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
    C, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 64 × 64 is without dividing, and block size 64 × 64 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 64 × 64 divides processing, splits into 4 block sizes and is 32 × 32 coding unit CU, enters steps d;
    D, the initial block size of setting coding unit CU are 32 × 32, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 32 × 32 coding unit CU;
    E, according to the motion vector MV that extracts the macro block described in the steps d obtaining in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 32 × 32 mv,
    D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
    In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 32 × 32, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
    F, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 32 × 32 is without dividing, and block size 32 × 32 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 32 × 32 divides processing, splits into the coding unit of 4 block sizes 16 × 16, enters step g;
    G, the initial block size of setting coding unit CU are 16 × 16, calculate initial block size and be number and the type of the H.264 substandard macro block comprising in 16 × 16 coding unit CU;
    H, according to the motion vector MV that extracts the described macro block of the step (g) that obtains in decoding end H.264, the variance distance B of the motion vector of interior each macro block comprising of coding unit CU that to calculate block size be 16 × 16 mv,
    D mv = ( MV x m - MV x n ) 2 + ( MV y m - MV y n ) 2 - - - i
    In formula I, D mvin order to represent the otherness of motion vector of each macro block comprising in coding unit CU that block size is 16 × 16, wherein with represent respectively H.264 and be positioned at (x in standard m, y m) and (x n, y n) abscissa and the ordinate of the m macro block of locating and the motion vector MV of n macro block;
    I, a threshold value T of setting 1if, D mvbe less than T 1time, the coding unit CU that block size is 16 × 16 is without dividing, and block size 16 × 16 is now exactly coding unit CU dividing mode; If D mvbe more than or equal to T 1time, the coding unit CU that block size is 16 × 16 divides processing, splits into the coding unit of 4 block sizes 8 × 8, and block size 8 × 8 is now exactly coding unit CU dividing mode;
    (3) determine the dividing mode of predicting unit PU according to the piecemeal of coding unit CU, determine the dividing mode of predicting unit PU according to coding unit CU dividing mode, each block size is the dividing mode that the coding unit CU of 2N × 2N has 7 kinds of predicting unit PU, wherein, described N=4,8,16,32; The dividing mode of 7 kinds of described predicting unit PU is respectively the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N, and concrete steps are as follows:
    If the block size of j coding unit CU is, described coding unit CU is divided into the fritter that 4 block sizes are 4 × 4 at 8 × 8 o'clock; Otherwise, described coding unit CU is divided into 16 fritters that block size is identical;
    K, according to the motion vector information obtaining in code stream H.264, extract the motion vector of each fritter in coding unit CU, and calculate the phase place of the motion vector of the fritter described in each ;
    By the variance sum D of the phase place of each fritter in coding unit CU under the dividing mode of 7 kinds of predicting unit PU of following formula II calculating ψ, the dividing mode of 7 kinds of described predicting unit PU is the asymmetric pattern of cutting apart of the symmetry division pattern of 2N × 2N, 2N × N, N × 2N and 2N × nU, 2N × nD, nL × 2N, nR × 2N;
    D ψ＝D ψ1+D ψ2 ⅱ
    In formula II, ψ 1, ψ 2represent respectively two segmented areas in predicting unit PU dividing mode; wherein, (i, j) is ψ 1the transverse and longitudinal coordinate of piecemeal; (i ', j ') be ψ 2the transverse and longitudinal coordinate of piecemeal;
    Relatively D in the dividing mode situation of 7 kinds of described predicting unit PU ψ, choose and make D ψthe value dividing mode of hour predicting unit PU is defined as the dividing mode of predicting unit PU;
    (4) directly dope the motion vector of predicting unit PU in current HEVC by the motion vector information extracting in code stream H.264, concrete steps are:
    L, basis at the motion vector of the H.264 code stream that H.264 decoder terminal extracts, find respectively the corresponding best matching blocks in reference frame unit of each fritter comprising in the predicting unit PU that determines dividing mode in step (l); Described best matching blocks refer to described predicting unit PU in the nearest reference frame unit of each fritter of comprising; Calculate respectively the variance distance B of each fritter comprising in reference frame unit and predicting unit PU by following formula III,
    In formula III, for the position of t fritter in predicting unit PU; for the position of best matching blocks corresponding to t fritter in PU unit;
    Wherein, with described predicting unit PU in the variance of each fritter of comprising be described best matching blocks apart from the reference frame unit of sum minimum;
    If m obtains a best matching blocks by step (l), directly calculate skew between predicting unit PU and described best matching blocks and be the motion vector of described predicting unit PU; If obtain two or more best matching blocks by step (l), the summation absolute error criterion SAD of the predicting unit PU of calculative determination dividing mode and described two or more best matching blocks respectively, choose the best matching blocks of sad value minimum as best matching blocks, directly calculate skew between predicting unit PU and described best matching blocks and be the motion vector of described predicting unit PU.
  2. One kind realize claimed in claim 1 based on motion vector analysis by the transcoding device to code-transferring method the fast frame of HEVC H.264, it is characterized in that, a kind of realize based on motion vector analysis by the transcoding device to code-transferring method the fast frame of HEVC H.264, comprise: H.264 decoder module, motion vector information extraction module, storage module, computing module and HEVC coding module, described H.264 decoder module connects described motion vector information extraction module, described motion vector information extraction module connects described storage module, described storage module connects described computing module, described computing module connects described HEVC coding module.
  3. According to claim 2 a kind of realize claimed in claim 1 based on motion vector analysis by the transcoding device to code-transferring method the fast frame of HEVC H.264, it is characterized in that, described H.264 decoder module is decoded to code stream H.264 in decoder terminal H.264; Described motion vector information extraction module is for the reference frame information that extracts described H.264 code stream in decoder terminal H.264 and comprise, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block; Described storage module, for storing reference frame information that described H.264 code stream that motion vector information extraction module extracts comprises, macroblock size, macro block position, the motion vector MV of macro block, the prediction mode of macro block; Described computing module is for the variance distance of the motion vector of interior each macro block comprising of calculation code unit CU, be used for the variance sum of the phase place of calculation code unit each fritter of CU, the variance of each fritter comprising for computing reference frame unit distance and predicting unit PU is apart from sum, also for calculating the skew between predicting unit PU and described best matching blocks; Described HEVC coding module, for according to the data result obtaining from motion vector information that H.264 code stream extracts and computing module, is optimized coding to present frame.
CN201410340334.3A 2014-07-16 2014-07-16 It is a kind of based on motion vector analysis by H.264 to the quick interframe code-transferring method and transcoding device of HEVC Active CN104079937B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410340334.3A CN104079937B (en) 2014-07-16 2014-07-16 It is a kind of based on motion vector analysis by H.264 to the quick interframe code-transferring method and transcoding device of HEVC

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410340334.3A CN104079937B (en) 2014-07-16 2014-07-16 It is a kind of based on motion vector analysis by H.264 to the quick interframe code-transferring method and transcoding device of HEVC

Publications (2)

Publication Number Publication Date
CN104079937A true CN104079937A (en) 2014-10-01
CN104079937B CN104079937B (en) 2017-06-23

Family

ID=51600946

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410340334.3A Active CN104079937B (en) 2014-07-16 2014-07-16 It is a kind of based on motion vector analysis by H.264 to the quick interframe code-transferring method and transcoding device of HEVC

Country Status (1)

Country Link
CN (1) CN104079937B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104320658A (en) * 2014-10-20 2015-01-28 南京邮电大学 HEVC (High Efficiency Video Coding) fast encoding method
CN104837019A (en) * 2015-04-30 2015-08-12 上海交通大学 AVS-to-HEVC optimal video transcoding method based on support vector machine
CN105430418A (en) * 2015-11-13 2016-03-23 山东大学 Method for rapidly transcoding from H.264/AVC to HEVC
WO2017101349A1 (en) * 2015-12-14 2017-06-22 乐视控股（北京）有限公司 Real-time transcoding method and device with encoding information inheritance
CN107018412A (en) * 2017-04-20 2017-08-04 四川大学 A kind of DVC HEVC video transcoding methods based on key frame coding unit partition mode
CN107426573A (en) * 2017-08-08 2017-12-01 鄂尔多斯应用技术学院 Adaptive fast prediction dividing elements method and device based on motion homogeney
CN107465923A (en) * 2016-06-06 2017-12-12 谷歌公司 Adaptive overlapping block prediction in the video coding of variable block length
CN109246432A (en) * 2018-09-26 2019-01-18 深圳市网心科技有限公司 A kind of video transcoding method, system, equipment and computer readable storage medium

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060072665A1 (en) * 2004-10-02 2006-04-06 Jin-Soo Cho Methods and transcoders that estimate an output macroblock and motion vector for video transcoding
EP1838106A2 (en) * 2006-03-23 2007-09-26 Sony Corporation Video signal coding system and method of coding video signal for network transmission, video output apparatus, and signal conversion apparatus
WO2008039857A2 (en) * 2006-09-26 2008-04-03 Dilithium Networks Pty Ltd. Method and apparatus for compressed video bitstream conversion with reduced-algorithmic-delay
CN103491334A (en) * 2013-09-11 2014-01-01 浙江大学 Video transcode method from H264 to HEVC based on region feature analysis
CN103546749A (en) * 2013-10-14 2014-01-29 上海大学 Method for optimizing HEVC (high efficiency video coding) residual coding by using residual coefficient distribution features and bayes theorem
CN103796023A (en) * 2014-03-05 2014-05-14 武汉拓宝电子系统有限公司 H.265/HEVC transcoding method and transcoder based on H.264/AVC macro block structure and texture

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060072665A1 (en) * 2004-10-02 2006-04-06 Jin-Soo Cho Methods and transcoders that estimate an output macroblock and motion vector for video transcoding
EP1838106A2 (en) * 2006-03-23 2007-09-26 Sony Corporation Video signal coding system and method of coding video signal for network transmission, video output apparatus, and signal conversion apparatus
WO2008039857A2 (en) * 2006-09-26 2008-04-03 Dilithium Networks Pty Ltd. Method and apparatus for compressed video bitstream conversion with reduced-algorithmic-delay
CN103491334A (en) * 2013-09-11 2014-01-01 浙江大学 Video transcode method from H264 to HEVC based on region feature analysis
CN103546749A (en) * 2013-10-14 2014-01-29 上海大学 Method for optimizing HEVC (high efficiency video coding) residual coding by using residual coefficient distribution features and bayes theorem
CN103796023A (en) * 2014-03-05 2014-05-14 武汉拓宝电子系统有限公司 H.265/HEVC transcoding method and transcoder based on H.264/AVC macro block structure and texture

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
李乐: "针对HEVC视频编码标准的帧间快速算法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 *
蒋炜，田翔，陈耀武: "H.264到HEVC的低复杂度视频转码算法", 《浙江大学学报（工学版）》 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104320658B (en) * 2014-10-20 2017-12-26 南京邮电大学 A kind of HEVC fast encoding methods
CN104320658A (en) * 2014-10-20 2015-01-28 南京邮电大学 HEVC (High Efficiency Video Coding) fast encoding method
CN104837019A (en) * 2015-04-30 2015-08-12 上海交通大学 AVS-to-HEVC optimal video transcoding method based on support vector machine
CN104837019B (en) * 2015-04-30 2018-01-02 上海交通大学 AVS to HEVC optimization video transcoding methods based on SVMs
CN105430418A (en) * 2015-11-13 2016-03-23 山东大学 Method for rapidly transcoding from H.264/AVC to HEVC
CN105430418B (en) * 2015-11-13 2018-04-10 山东大学 H.264/AVC, one kind arrives HEVC fast transcoding methods
WO2017101349A1 (en) * 2015-12-14 2017-06-22 乐视控股（北京）有限公司 Real-time transcoding method and device with encoding information inheritance
CN107465923A (en) * 2016-06-06 2017-12-12 谷歌公司 Adaptive overlapping block prediction in the video coding of variable block length
CN107465923B (en) * 2016-06-06 2020-09-25 谷歌有限责任公司 Adaptive overlapped block prediction in variable block size video coding
CN107018412A (en) * 2017-04-20 2017-08-04 四川大学 A kind of DVC HEVC video transcoding methods based on key frame coding unit partition mode
CN107018412B (en) * 2017-04-20 2019-09-10 四川大学 A kind of DVC-HEVC video transcoding method based on key frame coding unit partition mode
CN107426573A (en) * 2017-08-08 2017-12-01 鄂尔多斯应用技术学院 Adaptive fast prediction dividing elements method and device based on motion homogeney
CN109246432A (en) * 2018-09-26 2019-01-18 深圳市网心科技有限公司 A kind of video transcoding method, system, equipment and computer readable storage medium

Also Published As

Publication number Publication date
CN104079937B (en) 2017-06-23

Similar Documents

Publication Publication Date Title
CN103248893B (en) From H.264/AVC standard to code-transferring method and transcoder thereof the fast frame of HEVC standard
CN104079937A (en) H.264-to-HEVC fast inter-frame transcoding method and device based on motion vector analysis
US10778998B2 (en) Moving picture coding device, moving picture coding method, moving picture coding program, moving picture decoding device, moving picture decoding method, and moving picture decoding program
CN103069800B (en) For the method and apparatus of encoded video and coding/decoding method and equipment
US9197903B2 (en) Method and system for determining a macroblock partition for data transcoding
US11025949B2 (en) Moving picture coding device, moving picture coding method and moving picture coding program, and moving picture decoding device, moving picture decoding method and moving picture decoding program
CN114009040B (en) Encoder, decoder and corresponding methods
EP1618744B1 (en) Video transcoding
CN101621687B (en) Methodfor converting video code stream from H. 264 to AVS and device thereof
CN103442228B (en) Code-transferring method and transcoder thereof in from standard H.264/AVC to the fast frame of HEVC standard
WO2006073116A1 (en) Video encoding method and device, video decoding method and device, program thereof, and recording medium containing the program
US20100310184A1 (en) Dual prediction video encoding and decoding method and device
CN112673633B (en) Encoder, decoder and corresponding methods for merging modes
US20050276331A1 (en) Method and apparatus for estimating motion
CN104038764A (en) H.264-to-H.265 video transcoding method and transcoder
CN113170202B (en) Encoder, decoder and corresponding methods for constructing MPM list of block applying multi-hypothesis prediction
CN105681797A (en) Prediction residual based DVC-HEVC (Distributed Video Coding-High Efficiency Video Coding) video transcoding method
CN110351552B (en) Fast coding method in video coding
CN102696226B (en) The H264 code-transferring method of multiplexing code stream information
US20060222251A1 (en) Method and system for frame/field coding
CN101854554A (en) Video encoding and decoding system based on image inpainting predication
CN100586185C (en) Mode selection method for transcoding 264 video to reduce resolving capability
CN105306946A (en) Quality scalable video coding method based on mean square error threshold
CN101262607A (en) Two-folded prediction video coding and decoding method and device
Zhou et al. Disaster transmission network based on next-generation video standard

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