CN106101699A - Depth modelling mode adjudging method for 3D HEVC depth map encoding - Google Patents

Depth modelling mode adjudging method for 3D HEVC depth map encoding Download PDF

Info

Publication number
CN106101699A
CN106101699A CN201610571479.3A CN201610571479A CN106101699A CN 106101699 A CN106101699 A CN 106101699A CN 201610571479 A CN201610571479 A CN 201610571479A CN 106101699 A CN106101699 A CN 106101699A
Authority
CN
China
Prior art keywords
depth
block
mode
modelling
depth map
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
CN201610571479.3A
Other languages
Chinese (zh)
Other versions
CN106101699B (en
Inventor
张秋闻
赵进超
常化文
蒋斌
吴庆岗
黄琨强
王晓
张文帅
赵小鑫
甘勇
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.)
Zhengzhou University of Light Industry
Original Assignee
Zhengzhou University of Light Industry
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 Zhengzhou University of Light Industry filed Critical Zhengzhou University of Light Industry
Priority to CN201610571479.3A priority Critical patent/CN106101699B/en
Publication of CN106101699A publication Critical patent/CN106101699A/en
Application granted granted Critical
Publication of CN106101699B publication Critical patent/CN106101699B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/103—Selection of coding mode or of prediction mode
    • 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
    • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
    • H04N19/146—Data rate or code amount at the encoder output
    • H04N19/147—Data rate or code amount at the encoder output according to rate distortion criteria
    • 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
    • H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/597—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding specially adapted for multi-view video sequence encoding

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

The invention discloses a kind of depth modelling mode adjudging method for 3D HEVC depth map encoding, depth map macro block is carried out classification and skips judgement, terminate the most in advance according to predicting unit size discrimination outline mode, the rate-distortion optimization carrying out depth modelling pattern obtains the depth modelling pattern of optimum, the steps include: to start the depth modelling mode decision process of depth map macro block；It is near field, middle region and far region current depth block sort；Carry out unnecessary depth block to skip；Size according to current prediction unit carries out the termination in advance of outline mode；Carry out the rate-distortion optimization process of depth modelling mode decision；Judge the optimal depth modeling pattern of depth map.The classification that present invention employs block is skipped and the termination in advance of outline mode, coding efficiency is preferable, and code check the most slightly goes up, and Y-PSNR reduces negligible, averagely save the scramble time of 31.33% than original 3D HEVC coded method, can be applicable to real-time coding.

Description

Depth modelling mode adjudging method for 3D-HEVC depth map encoding
Technical field
The invention belongs to the technical field of Video coding, particularly relate to the degree of depth of intraframe coding in a kind of Video coding Modeling mode adjudging method, a kind of towards the degree of depth of depth map encoding in high efficiency video encoding standard 3D Video Expansion Modeling mode adjudging method.
Background technology
In recent years, universal along with HD video and various Internet video, video compression technology had welcome no small Challenge.The Motion Picture Experts Group (Motion Picture Experts Group, MPEG) of ISO/IEC and the video of ITU-T Coding Experts group (Video Coding Expert Group, VCEG) composition Video coding combines group (Joint Collaborative Team on Video Coding, JCT-VC), and combine and formulated high efficiency video encoding standard H.265/HEVC(High Efficiency Video Coding).A new generation video encoding standard H.265/HEVC in add Many new coding toolses, with previous generation video encoding standard H.264/AVC compared with, coding efficiency H.265/HEVC has had greatly The lifting of amplitude.And, extension 3D-HEVC (high efficiency video based on high efficiency video encoding standard Coding based 3D video coding) also become study hotspot in recent years, 3D-HEVC is capable of three-dimensional and regards more The imaging of point, it uses texture video plus depth form to synthesize virtual view, thus reduces substantial amounts of encoder bit rate.And it is deep The preservation of degree figure is most important to synthesizing high-quality virtual view texture video, therefore, and some new depth map encoding instruments It is introduced into 3D-HEVC, such as depth modelling pattern (DMM) decision-making.But, numerous DMM candidate pattern result in huge meter Calculation amount, this also counteracts that 3D-HEVC application in reality.
At present, there have been many ripe technology for depth map encoding in 3D-HEVC, such as: Ding H et al. proposes A kind of effective depth map encoding algorithm, make use of based on people in discernable depth difference Modeling Theory clever to the perception of the degree of depth Sensitivity feature skips the depth block that some are unnecessary.It is empty that Zhang H B et al. proposes a kind of reference pixel based on content Between the fast deep figure intra mode decision algorithm of sorting technique.But, although the calculating that these technology decrease DMM decision-making is multiple Miscellaneous degree, but code efficiency is the most relatively low, so the coding efficiency of its depth map prediction does not improve.Therefore, how at depth map Coding is accelerated DMM decision method process effectively, it is achieved outstanding code efficiency is the difficulty that current video coding technique faces One of topic.
Depth modelling pattern (DMM) decision-making in depth map encoding, is in order at compression depth image for its essence More preferable Protect edge information part while flat site.Either conventional video encoding and decoding framework HEVC, or 3D video is compiled Decoding framework 3D-HEVC coding framework, their prediction process all refers to block and divides (block partitioning) technology, More than one region (partition) will be divided into by an image block, be predicted in units of described region the most again. And block dividing mode traditional in HEVC is that a square image block is divided into two or four square along the horizontal or vertical direction Shape region (rectangular partition).And DMM pattern introduces a kind of novel block dividing mode, i.e. according to specific Information one depth image block is divided into the region (non-rectangular partition) of two non-rectangles, same Pixel in region represents with identical constant value, and different regional values is encoded separately, and then preferably retains depth image Marginal portion.But the model selection of DMM decision-making needs to travel through all of candidate pattern, which results in huge amount of calculation, meter Evaluation time is longer, and this is necessary for consuming the substantial amounts of scramble time, is unfavorable for carrying out real-time coding.Calculate complexity in this way Spend higher, it is difficult to apply in real time.
Summary of the invention
The problem and shortage existed in view of above prior art, it is an object of the invention to provide a kind of for 3D-HEVC The depth modelling mode adjudging method of depth map encoding, compared with Raw encoder and up-to-date SRPS coded system, not only protects Hold almost identical code efficiency, moreover it is possible to effectively reduce the scramble time.
In order to achieve the above object, the technical scheme is that a kind of degree of depth for 3D-HEVC depth map encoding is built Mould mode adjudging method, first carries out classification to depth block and skips judgement, then according to predicting unit size discrimination outline mode Terminating the most in advance, the rate-distortion optimization finally carrying out depth modelling pattern obtains the depth modelling pattern of optimum, and it specifically walks Suddenly:
(1) the depth modelling mode decision process of depth map macro block, is started；
(2), according to the threshold value of the near field determined and far region, current depth block sort be near field, middle region and Far region；
(3), carry out unnecessary depth block to skip: if depth block belongs near field or middle region, enter step (4)；No Then, depth block depth modelling mode skipping rate-distortion optimization process in far region, enter step (6)；
(4) termination in advance of outline mode, is carried out according to the size of current prediction unit: if current predicting unit chi Very little less than 32 × 32, outline mode is terminated in advance, and enters step (5)；
(5) the rate-distortion optimization process of depth modelling mode decision, is carried out；
(6), judge that the optimal depth of depth map models pattern, enter step (1) and proceed next depth block judgement.
Described step (2) the method that current depth block sort is near field, middle region and far region is:
(2-1) threshold value D of near field, is determined according to below equationnThreshold value D with far regionfValue:
D n = 2 3 ( D m a x - D m i n ) D f = 1 3 ( D m a x - D m i n ) ;
Wherein, DmaxAnd DminRepresent the depth map value that present image is minimum and maximum respectively；
(2-2), by comparing depth map value depth block being divided into three regions, expression formula is:
Wherein, DblockRepresent the depth map value of current prediction unit PU.
Carrying out the method that unnecessary depth block skips in described step (3) is:
According to marginal area depth block blockedgeWith smooth regional depth block blockhomoAccounting for of entire depth number of blocks Than and skip some unnecessary depth block in the motor activity implementations of zones of different.
The method of the rate-distortion optimization process that described step (5) carries out depth modelling mode decision is: according to rate distortion The Lagrange multiplier of cost function and have the decision method that the pattern of minimum rate distortion cost is optimal predictive mode, profit By rate distortion cost function J=Diff+ λ Bit judgement rate distortion situation；Wherein, Diff is present mode and predictive mode Mean difference, λ is Lagrange multiplier, and Bit is bit rate.
Compared with prior art, the present invention only only reduces insignificant PSNR value, merely add a small amount of code check, saves A large amount of scramble times；Shift to an earlier date terminating method owing to have employed comprehensive depth block skipping method and outline mode, obtain fabulous volume While code performance, save the substantial amounts of scramble time, there is actual operation, it is simple to apply in real time.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing In having technology to describe, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is only this Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to Other accompanying drawing is obtained according to these accompanying drawings.
Fig. 1 is the flow chart of the present invention.
Fig. 2 is the RD curve of present invention experimental data compared with original 3D-HEVC encoder, and wherein, (a) is " Dancer ", (b) is " GT_Fly ", and (c) is " Shark ", and (d) is " Poznan_Street ", and (e) is " Poznan_Hall2 ", F () is " Newspaper ", (g) be " Kendo " and (h) be " Balloons ".
Detailed description of the invention
Below in conjunction with accompanying drawing, embodiments of the invention are described in further detail.The present embodiment is with the technology of the present invention Implement under premised on scheme, give detailed embodiment, but protection scope of the present invention is not limited to following enforcement Example.
As it is shown in figure 1, one models mode adjudging method for depth map fast deep, depth block is carried out classification and skips Judging, then terminate the most in advance according to predicting unit size discrimination outline mode, the rate finally carrying out depth modelling pattern is lost True optimization obtains optimum depth modelling pattern.Specifically, it is first near field, middle region and far field current depth block sort Territory, then judges depth block, if belonging near field or middle region, according to current prediction unit PU be smaller in size than 32 × When 32, skipping unnecessary depth block, outline mode is terminated in advance, finally carries out the rate distortion of depth modelling mode decision Optimization process, determines the optimal depth modeling pattern of depth map, and judging process terminates, and the steps include:
(1) the depth modelling mode decision process of depth map macro block, is started.
(2) it is, near field, middle region and far region current depth block sort.
Visual system feature according to the mankind, is generally higher than distant object, target to the sensitivity of target travel nearby Motion activity can be divided into three regions according to depth map value: near field, middle region and far region.And in 3D-HEVC Depth modelling mode decision height rely on the marginal distribution of depth map, we can set up depth block type and depth map value Dependency:
And
Wherein, DblockShow the depth map value of current prediction unit PU, DnAnd DfRepresent near field and far region respectively Threshold value.DmaxAnd DminRepresent the depth map value that present image is minimum and maximum respectively.Threshold value D determined by according tonAnd DfBy the degree of depth Block sort is near field, middle region and far region.
(3), carry out unnecessary depth block to skip: if depth block belongs near field or middle region, enter step (4)；No Then, depth block depth modelling mode skipping rate-distortion optimization process in far region, enter step (6).
Skip unnecessary depth block: according to blockedgeAnd blockhomoIn the accounting of entire depth number of blocks and not Some unnecessary depth block are skipped with the motor activity implementations in region.Because of the depth block of belt edge information in far region Quantity little, DMM decision making process can be skipped in these depth block, during so running in 3D-HEVC depth map encoding Between can be significantly reduced.blockedgeAnd blockhomoRepresent edge and the depth block of flat site respectively.Table 1 is depth block Type is at the distribution of zones of different, blockedgeAnd blockhomoEffect be to indicate two types depth block in three kinds of regions Accounting situation.It is different from depth map value, based on depth block type, utilizes motion activity sensitive by human visual system The characteristic that degree determines, people are relatively low to the relatively low i.e. object of which movement activeness of sensitivity of far region moving object, it means that remote Block in regionedgeAccounting relatively low.Can decide whether to skip marginal zone in current region according to depth block type accounting situation Territory depth block blockedgeCataloged procedure.As shown in Table 1, in far region average the most only 3% blockedgeDepth block, according to Experimental result, skips these depth block negligible on encoding efficiency impact, can adjudicate and skip coding.
Table 1 depth block type is in the distribution of zones of different
(4) termination in advance of outline mode, is carried out according to the size of current prediction unit: if current predicting unit chi Very little less than 32 × 32, outline mode is terminated in advance.
Because generally all selecting tapered mode in 3D-HEVC, and the percentage ratio of tapered mode is along with predicting unit PU chi Very little minimizing can increase.If we can judge predicting unit PU in advance, no matter whether optimum depth modeling pattern is wedge shape Pattern, complicated rate-distortion optimization process can be skipped, and can save the substantial amounts of scramble time.When predicting unit PU Size ratio 32 × 32 hours, outline mode is difficult to be chosen as optimal depth modelling pattern.Therefore, when the size of predicting unit PU When being 8 × 8 and 16 × 16, rate-distortion optimization process can be terminated in advance.Table 2 is different predicting unit PU size and DMM The relation of decision-making, as shown in Table 2: when when being smaller in size than 32 × 32 of PU, can directly determine that the pattern of DMM decision-making is wedge-shaped die Formula.
The different PU size of table 2 and the relation of DMM decision-making
(5) the rate-distortion optimization process of depth modelling mode decision, is carried out.
Lagrange multiplier according to rate distortion (RD) cost function and the pattern having minimum rate distortion cost are optimal The decision method of predictive mode, utilize equation J=Diff+ λ Bit judgement rate distortion situation.Wherein Diff is present mode With the mean difference of predictive mode, λ is Lagrange multiplier, and Bit is bit rate, and J is RD cost function.
(6), judge that the optimal depth of depth map models pattern, enter step (1) and proceed next depth block judgement.
The modeling pattern that depth modelling pattern is optimal is determined according to result of calculation.The result obtaining RD cost function J is the lowest This kind of pattern optimum is described, i.e. selects this optimization model to encode.
In order to verify the effect of the present invention, below several standard testing video sequences are carried out test experiments, at JCT-VC On the high efficiency video coding system platform provided, it is respectively adopted fixed quantisation parameter with HTM.Table 3 provides test data parameters, Table 4 provides test condition, and table 5 provides the result of the present invention.
Standard test sequences:
Data parameters tested by table 3
Experimental configuration:
Table 4 experimental configuration
Central processing unit 2×Intel Xeon E5-2640 v2 2.0GHz
Random access memory 2×16GB DDR3
Operating system Microsoft Windows 10(64-bit)
Experiment porch Microsoft Visual C++2010
Test result:
Table 5 present invention and the contrast and experiment of original 3D-HEVC encoder
Cycle tests ΔPSNR(dB) Δ BR (%) Δ Time (%)
Dancer -0.01 0.47 33.21
GT_Fly 0.00 0.34 34.74
Shark 0.00 0.40 34.17
Poznan_Street -0.03 0.89 25.16
Poznan_Hall2 -0.02 0.78 27.75
Newspaper -0.02 0.55 31.55
Kendo -0.02 0.67 32.41
Balloons -0.01 0.69 31.62
Average -0.01 0.60 31.33
(a) " Dancer ", (b) " GT_Fly ", (c) " Shark ", (d) " Poznan_Street ", (e) in Fig. 2 " Poznan_Hall2 ", (f) " Newspaper ", (g) " Kendo " and (h) " Balloons " at quantization parameter QP to (25,34), (30,39), the RD curve under (35,42) and (40,45).From table 5 and Fig. 2: the present invention and original 3D-HEVC coding staff Formula is compared, and saves the scramble time of average about 31.33% respectively, as Fig. 2 shows, although code check has and goes up by a small margin, PSNR has and declines by a small margin, but saves the substantial amounts of scramble time.
The above, the only present invention preferably detailed description of the invention, but protection scope of the present invention is not limited thereto, Any those familiar with the art in the technical scope that the invention discloses, the change that can readily occur in or replacement, All should contain within protection scope of the present invention.

Claims (4)

1. the depth modelling mode adjudging method for 3D-HEVC depth map encoding, it is characterised in that first to depth block Carry out classification and skip judgement, then terminate the most in advance according to predicting unit size discrimination outline mode, finally carry out the degree of depth and build The rate-distortion optimization of mould pattern obtains the depth modelling pattern of optimum；Its step is as follows:
(1) the depth modelling mode decision process of depth map macro block, is started；
(2), according to the threshold value of the near field determined and far region, it is near field, middle region and far field current depth block sort Territory；
(3), carry out unnecessary depth block to skip: if depth block belongs near field or middle region, enter step (4)；Otherwise, exist Depth block depth modelling mode skipping rate-distortion optimization process in far region, enters step (6)；
(4) termination in advance of outline mode, is carried out according to the size of current prediction unit: if current predicting unit size ratio 32 × 32 is little, and outline mode is terminated in advance, and enters step (5)；
(5) the rate-distortion optimization process of depth modelling mode decision, is carried out；
(6), judge that the optimal depth of depth map models pattern, enter step (1) and proceed next depth block judgement.
Depth modelling mode adjudging method for 3D-HEVC depth map encoding the most according to claim 1, its feature exists In, described step (2) the method that current depth block sort is near field, middle region and far region is:
(2-1) threshold value D of near field, is determined according to below equationnThreshold value D with far regionfValue:
D n = 2 3 ( D m a x - D m i n ) D f = 1 3 ( D m a x - D m i n ) ;
Wherein, DmaxAnd DminRepresent the depth map value that present image is minimum and maximum respectively；
(2-2), by comparing depth map value depth block being divided into three regions, expression formula is:
Wherein, DblockRepresent the depth map value of current prediction unit PU.
Depth modelling mode adjudging method for 3D-HEVC depth map encoding the most according to claim 1, its feature exists In, carrying out the method that unnecessary depth block skips in described step (3) is:
According to marginal area depth block blockedgeWith smooth regional depth block blockhomoEntire depth number of blocks accounting with And skip some unnecessary depth block in the motor activity implementations of zones of different.
Depth modelling mode adjudging method for 3D-HEVC depth map encoding the most according to claim 1, its feature exists In, the method for the rate-distortion optimization process that described step (5) carries out depth modelling mode decision is: according to rate distortion costs letter The Lagrange multiplier counted and the pattern having minimum rate distortion cost are the decision methods of optimal predictive mode, and utilization rate is lost True cost function J=Diff+ λ Bit judgement rate distortion situation；Wherein, Diff is the mean deviation of present mode and predictive mode Different, λ is Lagrange multiplier, and Bit is bit rate.
CN201610571479.3A 2016-07-20 2016-07-20 For the depth modelling mode adjudging method of 3D-HEVC depth map encoding Active CN106101699B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610571479.3A CN106101699B (en) 2016-07-20 2016-07-20 For the depth modelling mode adjudging method of 3D-HEVC depth map encoding

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610571479.3A CN106101699B (en) 2016-07-20 2016-07-20 For the depth modelling mode adjudging method of 3D-HEVC depth map encoding

Publications (2)

Publication Number Publication Date
CN106101699A true CN106101699A (en) 2016-11-09
CN106101699B CN106101699B (en) 2018-12-28

Family

ID=57220751

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610571479.3A Active CN106101699B (en) 2016-07-20 2016-07-20 For the depth modelling mode adjudging method of 3D-HEVC depth map encoding

Country Status (1)

Country Link
CN (1) CN106101699B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103546747A (en) * 2013-09-29 2014-01-29 北京航空航天大学 Color video encoding mode based depth map sequence fractal encoding method
CN104378643A (en) * 2014-12-04 2015-02-25 南京理工大学 Intra-frame prediction mode selection method and system of 3D (3-dimension) video plus depth image

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103546747A (en) * 2013-09-29 2014-01-29 北京航空航天大学 Color video encoding mode based depth map sequence fractal encoding method
CN104378643A (en) * 2014-12-04 2015-02-25 南京理工大学 Intra-frame prediction mode selection method and system of 3D (3-dimension) video plus depth image

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
HONG-BIN ZHANG ET AL.: "《EFFICIENT DEPTH INTRA MODE DECISION BY REFERENCE PIXELS CLASSIFICATION IN 3D-HEVC》", 《2015 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP)》 *
HUI DING ET AL.: "《Depth map pre-processing algorithm for compression based on 3D-HEVC scheme》", 《2015 IEEE 16TH INTERNATIONAL CONFERENCE ON COMMUNICATION TECHNOLOGY (ICCT)》 *

Also Published As

Publication number Publication date
CN106101699B (en) 2018-12-28

Similar Documents

Publication Publication Date Title
CN105049850B (en) HEVC bit rate control methods based on area-of-interest
CN103380620B (en) Picture coding device and method for encoding images
CN104754357B (en) Intraframe coding optimization method and device based on convolutional neural networks
CN104639940B (en) A kind of quick HEVC method for choosing frame inner forecast mode
CN104796694B (en) Optimization intraframe video coding method based on video texture information
CN103686165B (en) Decoding method and Video Codec in depth image frame
CN106604031A (en) Region of interest-based H. 265 video quality improvement method
CN103067704B (en) A kind of method for video coding of skipping in advance based on coding unit level and system
CN104539962A (en) Layered video coding method fused with visual perception features
CN103297781A (en) High efficiency video coding (HEVC) intraframe coding method, device and system based on texture direction
CN105120290B (en) A kind of deep video fast encoding method
CN105898332B (en) For the fast deep figure frame mode decision method of 3D-HEVC coding standards
CN105721866B (en) A kind of coding unit partitioning method and device
CN105120282A (en) Code rate control bit distribution method of temporal dependency
CN106358040A (en) Rate control bit allocation method based on saliency
CN106937116A (en) Low-complexity video coding method based on random training set adaptive learning
Chen et al. A novel fast intra mode decision for versatile video coding
CN103533355A (en) Quick coding method for HEVC (high efficiency video coding)
Lee et al. Fast CU size decision algorithm using machine learning for HEVC intra coding
CN103634601A (en) Structural similarity-based efficient video code perceiving code rate control optimizing method
CN106507106A (en) Video interprediction encoding method based on reference plate
CN106162176A (en) Method for choosing frame inner forecast mode and device
CN107404653A (en) A kind of Parking quick determination method of HEVC code streams
CN103702131B (en) Pattern-preprocessing-based intraframe coding optimization method and system
CN102984524B (en) A kind of video coding-decoding method based on block layer decomposition

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