CN108012152B - Fast HEVC coding method - Google Patents
Fast HEVC coding method Download PDFInfo
- Publication number
- CN108012152B CN108012152B CN201711289816.0A CN201711289816A CN108012152B CN 108012152 B CN108012152 B CN 108012152B CN 201711289816 A CN201711289816 A CN 201711289816A CN 108012152 B CN108012152 B CN 108012152B
- Authority
- CN
- China
- Prior art keywords
- rdcost
- merge
- skip
- 2nx2n
- coding
- 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
Images
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
- H04N19/124—Quantisation
-
- 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
- H04N19/114—Adapting the group of pictures [GOP] structure, e.g. number of B-frames between two anchor frames
-
- 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
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/177—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being a group of pictures [GOP]
-
- 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
-
- 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
- H04N19/70—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses a fast HEVC coding method. The method specifically comprises the following steps: (1) establishing two threshold arrays T _8x8 and T _16x16 indexed by quantization parameter QP, and initializing T _8x8 and T _16x16 to 0; (2) setting every N GOPs as a GOP group, setting the first GOP in the GOP group as a statistical GOP, normally encoding all frames in the statistical GOP, and calculating values of array elements with a quantization parameter QP as an index in a threshold value T _8x8 or T _16x 16; (3) the 2 nd GOP to the Nth GOP in the GOP group are quick coding GOPs, frames in the quick coding GOPs use a quick algorithm, if the condition is met, Merge & Skip _2Nx2N is directly selected, and the calculation of other subsequent prediction units is skipped; otherwise, RDCost of all prediction modes needs to be calculated, and RDCost minimum is selected as the prediction unit of the CU. The invention has the beneficial effects that: the calculation amount can be saved to the maximum extent, and meanwhile, the coding quality is basically not lost; applying the invention to the 265 coder common coding quality level of ArcVideo, the coding time is saved by 7%, while the BDRate is increased by only 0.5%.
Description
Technical Field
The invention relates to the technical field related to video coding, in particular to a fast HEVC coding method.
Background
HEVC, as a latest generation video coding standard, is doubled in compression efficiency compared with the previous generation coding standard H.264, and meanwhile, the coding complexity is greatly improved. This is mainly due to the more complex coding techniques introduced by HEVC: including larger and more flexible coding units, more kinds of prediction units, more elaborate prediction modes, etc. Specifically, the method comprises the following steps: (1) HEVC promotes the coding unit of h.264 from 16x16 to 64x64, and furthermore, HEVC supports quadtree recursion from the largest coding unit to divide it into four sub-coding units, and the smallest coding unit is 8x8, so HEVC can select the optimal coding unit for the target scene according to the complexity of scene texture, motion amplitude, etc., thereby improving compression efficiency, but at the same time, the selection of more flexible coding units means that the encoder needs to make more coding attempts, and the amount of computation increases. (2) The HEVC can perform partition of prediction units within a coding unit, and for Inter-coded frames, the prediction units include prediction modes such as Merge _2Nx2N, Skip _2Nx2N, Inter _2Nx2N, Inter _2NxN, Inter _ Nx2N, Inter _ NxN, Inter _2NxnU, Inter _2NxnD, Inter _ nLx2N, Inter _ nRx2N, Intra _2Nx2N, and Intra _ NxN, and the coding unit selects a prediction unit with the highest compression rate from a plurality of prediction units, thereby improving coding efficiency and greatly increasing corresponding calculation amount.
The coding unit calculates the coding cost of each prediction unit, and selects the final prediction mode with the lowest cost, so that the selection of the prediction mode is the most calculated part in the coding process of the coder.
In the prediction mode Merge & Skip _2Nx2N, the motion vectors of adjacent blocks in a time domain and a space domain are used as candidate motion vectors of the current PU, the optimal MV is selected, then the index number Merge _ idx of the MV is transmitted in a code stream, and the number of bits consumed by Merge _ idx is small; other inter prediction units than Merge & Skip _2Nx2N require motion search for each PU to find the optimal MV, and MVDs (MV Difference) are transmitted in the code stream to represent MVs, which usually consume more bits than Merge _ idx and are usually more computationally intensive than Merge & Skip _2Nx2N due to the motion search process. From the view of the encoding result, the Inter-coded frame selects the prediction modes of Merge & Skip _2Nx2N with the highest proportion, wherein the calculation amount of the prediction modes of Inter _2Nx2N, Inter _2NxN, Inter _ Nx2N, Inter _ NxN, Intra _2Nx2N and Intra _ NxN is more than the calculation amount, but the final selection proportion is very low, therefore, the prediction modes can be skipped with a very large probability, and the encoding time is saved while the encoding compression efficiency is not influenced.
The existing fast algorithm scheme includes utilizing Skip _2Nx2N to terminate further quadtree division on the coding unit, thereby saving the calculation amount of sub-blocks; there is also a method of using the RDCost of the current PU to skip the subsequent PU computation process. The paper [ J Vanne, "efficiency Mode Decision Schemes for HEVC Inter Prediction", in IEEE Transactions on Circuits and Systems for Video Technology, Vol 24, Sept 2014] analyzed the computational complexity of symmetric Prediction Mode SMP (Inter _2NxN and Inter _ Nx2N) and asymmetric Prediction Mode AMP (Inter _2NxnU, Inter _2NxnD, Inter _ nLx2N, Inter _ nRx2N), investigated the algorithms for reducing SMP and AMP, and effectively saved the encoding time. The paper [ L Shen, "Adaptive Intermode Decision for HEVC Joint adapting Inter-Level and spatial correlation", in IEEE Transactions on Circuits and Systems for Video Technology, Vol 24.OCTOBER 2014] uses PU information and RDCost of neighboring CUs of the current CU to skip the calculation process of SMP and AMP of the current CU, and achieves good results. However, currently, existing schemes generally need to complete several prediction modes, namely Skip & Merge _2Nx2N and Inter _2Nx2N, and in real-time application of an HEVC encoder, SMP and AMP prediction modes are generally closed, so that the effect of the current scheme on real-time encoding configuration is not optimal.
Disclosure of Invention
The invention provides a fast HEVC coding method which greatly saves the calculation amount in order to overcome the defects in the prior art.
In order to achieve the purpose, the invention adopts the following technical scheme:
a fast HEVC coding method specifically comprises the following steps:
(1) establishing two threshold arrays T _8x8 and T _16x16 indexed by quantization parameter QP, and initializing T _8x8 and T _16x16 to 0;
(2) setting every N GOPs as a GOP group, wherein the first GOP in the GOP group is a statistical GOP, normally encoding all frames in the statistical GOP, carrying out classification statistics on Merge & Skip _2Nx2N of coding units CU of 8x8 and 16x16 and RDCost of Inter _2Nx2N in the statistical GOP according to quantization parameters QP, wherein the coding units CU of the same quantization parameters QP are in a group, and if the number of statistical samples under a certain quantization parameter QP is more than M, calculating the value of an array element with the quantization parameter QP as an index in a threshold T _8x8 or T _16x16 by using the group of samples;
(3) the 2 nd to nth GOPs in the GOP group are fast coded GOPs, and the frames in the fast coded GOPs use a fast algorithm: specifically, as follows, the coding quantization parameter of the current coding unit CU is QP, first calculating RDCost of the Merge & Skip _2Nx2N, and then if RDCost of the Merge & Skip _2Nx2N is smaller than the value of the array element indexed by the quantization parameter QP in the threshold T _8x8 or T _16x16, the coding unit CU directly selects the Merge & Skip _2Nx2N, and skips calculation of other subsequent prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and RDCost is selected as the minimum of the RDCost as the prediction unit of the coding unit CU.
Wherein: here T _8x8 and T _16x16 are not a number, but an array indexed by QP, with QP values ranging from 0 to 51. Gop (group of picture) refers to a set of consecutive pictures between two I-frames. RDCost refers to the coding cost. According to the invention, all subsequent prediction mode calculation is skipped directly according to the coding cost of Merge & Skip _2Nx2N, so that the calculation amount can be saved to the maximum extent, and meanwhile, the coding quality is not substantially lost. This is because when the prediction effect of Merge & Skip _2Nx2N is good enough, the distortion and the number of bits for encoding are small, and at this time, for Inter _2Nx2N, because encoding the motion vector consumes more bits than encoding Merge _ Idx, the RDcost of Inter _2Nx2N is hardly lower than that of Merge & Skip _2Nx 2N.
Preferably, in step (2), the value of the array element indexed by the quantization parameter QP in the threshold T _8x8 or T _16x16 is the maximum value satisfying the following condition: when the RDcost of Merge & Skip _2Nx2N is less than the threshold T, the probability that the RDcost of Merge & Skip _2Nx2N is less than the RDcost of Inter _2Nx2N is greater than or equal to 98%.
Preferably, in step (2), N is 10 and M is 1000.
Preferably, in step (3), the frames within the fast coded GOP use the fast algorithm: for a coding unit CU of 16x16, the coding quantization parameter is QP, firstly, RDcost of Merge & Skip _2Nx2N is calculated, and the result is RDcost _ Merge & Skip _16x 16; if RDCost _ Merge & Skip _16x16 < the value of the tuple element indexed by the quantization parameter QP in T _16x16, the 16x16 coding unit CU directly selects Merge & Skip _2Nx2N, and skips the calculation of the other subsequent prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and the prediction unit with the smallest RDCost is selected as the coding unit CU.
Preferably, in step (3), the frames within the fast coded GOP use the fast algorithm: for an 8x8 coding unit CU, the coding quantization parameter is QP, firstly, RDcost of Merge & Skip _2Nx2N is calculated, and the result is RDcost _ Merge & Skip _8x 8; if RDCost _ Merge & Skip _8x8 < T _8x8 value of the array element indexed by the quantization parameter QP, the coding unit CU of 8x8 directly selects Merge & Skip _2Nx2N, skipping the calculation of other prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and RDCost is selected as the minimum of the RDCost as the prediction unit of the coding unit CU.
The invention has the beneficial effects that: the calculation amount can be saved to the maximum extent, and meanwhile, the coding quality is basically not lost; applying the invention to the 265 coder common coding quality level of ArcVideo, the coding time is saved by 7%, while the BDRate is increased by only 0.5%.
Drawings
FIG. 1 is a RDCost distribution characteristic diagram of BasketCalled, QP 32, Merge & Skip _2Nx2N and Inter _2Nx2N according to an embodiment of the present invention;
fig. 2 is an RDCost distribution characteristic diagram of Bcactus, QP 32, Merge & Skip _2Nx2N and Inter _2Nx2N in the embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and detailed description.
A fast HEVC coding method specifically comprises the following steps:
(1) establishing two threshold arrays T _8x8 and T _16x16 indexed by quantization parameter QP, and initializing T _8x8 and T _16x16 to 0;
(2) setting every N GOPs as a GOP group, wherein the first GOP in the GOP group is a statistical GOP, normally encoding all frames in the statistical GOP, carrying out classification statistics on Merge & Skip _2Nx2N of coding units CU of 8x8 and 16x16 and RDCost of Inter _2Nx2N in the statistical GOP according to quantization parameters QP, wherein the coding units CU of the same quantization parameters QP are in a group, and if the number of statistical samples under a certain quantization parameter QP is more than M, calculating the value of an array element with the quantization parameter QP as an index in a threshold T _8x8 or T _16x16 by using the group of samples; wherein: the value of the array element indexed by the quantization parameter QP in the threshold T _8x8 or T _16x16 is the maximum value that satisfies the following condition: when the RDcost of Merge & Skip _2Nx2N is less than the threshold T, the probability that the RDcost of Merge & Skip _2Nx2N is less than the RDcost of Inter _2Nx2N is greater than or equal to 98%; n is 10, M is 1000;
(3) the 2 nd to nth GOPs in the GOP group are fast coded GOPs, and the frames in the fast coded GOPs use a fast algorithm: specifically, as follows, the coding quantization parameter of the current coding unit CU is QP, first calculating RDCost of the Merge & Skip _2Nx2N, and then if RDCost of the Merge & Skip _2Nx2N is smaller than the value of the array element indexed by the quantization parameter QP in the threshold T _8x8 or T _16x16, the coding unit CU directly selects the Merge & Skip _2Nx2N, and skips calculation of other subsequent prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and RDCost is selected as the minimum of the RDCost as the prediction unit of the coding unit CU.
Wherein: for a coding unit CU of 16x16, the coding quantization parameter is QP, firstly, RDcost of Merge & Skip _2Nx2N is calculated, and the result is RDcost _ Merge & Skip _16x 16; if RDCost _ Merge & Skip _16x16 < the value of the tuple element indexed by the quantization parameter QP in T _16x16, the 16x16 coding unit CU directly selects Merge & Skip _2Nx2N, and skips the calculation of the other subsequent prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and RDCost is selected as the minimum of the RDCost as the prediction unit of the coding unit CU. For an 8x8 coding unit CU, the coding quantization parameter is QP, firstly, RDcost of Merge & Skip _2Nx2N is calculated, and the result is RDcost _ Merge & Skip _8x 8; if RDCost _ Merge & Skip _8x8 < T _8x8 value of the array element indexed by the quantization parameter QP, the coding unit CU of 8x8 directly selects Merge & Skip _2Nx2N, skipping the calculation of other prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and RDCost is selected as the minimum of the RDCost as the prediction unit of the coding unit CU.
As shown in fig. 1 and 2, the four subgraphs respectively correspond to the statistical conditions of the coding units CU of 8x8, 16x16, 32x32 and 64x64, the abscissa is RDCost of Merge & Skip _2Nx2N, and the ordinate is (RDCost _ Merge & Skip _2Nx2N-RDCost _ Inter _2Nx2N)/RDCost _ Merge & Skip _2Nx 2N. Therefore, if the ordinate of a certain point is less than 0, the RDCost indicating that the Merge & Skip _2Nx2N is less than the RDCost of Inter _2Nx2N, and the encoding finally selects Merge & Skip _2Nx 2N. As can be seen from fig. 1 and 2, when the RDCost of the Merge & Skip _2Nx2N is smaller than a certain threshold T, all the points fall into the region where y is less than 0 (especially evident for coding units CU of 8x8 and 16x 16). This is because when the prediction effect of Merge & Skip _2Nx2N is good enough, the distortion and the number of bits for encoding are small, and at this time, for Inter _2Nx2N, because encoding the motion vector consumes more bits than encoding Merge _ Idx, the RDcost of Inter _2Nx2N is hardly lower than that of Merge & Skip _2Nx 2N.
It is derived from fig. 1 that when RDCost of Merge & Skip _2Nx2N is smaller than a threshold 2381 for a CU of 8x8 when the quantization parameter QP is 32, 98.82% of points fall in an area where y < 0; for a CU of 16x16, when the RDcost of Merge & Skip is less than the threshold value 3816, 98.75% of the points fall in the region where y < 0. It follows from fig. 2 that when RDCost of Merge & Skip _2Nx2N is smaller than the threshold 2601 for a CU of 8x8 when the quantization parameter QP is 32, 98.82% of the points fall in the region where y < 0; for a CU of 16x16, when the RDcost of Merge & Skip _2Nx2N is less than the threshold 3974, 98.75% of the points fall in the region where y < 0. By counting different QPs of different sequences, the threshold values are relatively close to each other under the same QP for 8x8 and 16x 16; the different QP thresholds are different: the larger the QP, the larger the threshold.
Applying the invention to the 265 coder common coding quality level of ArcVideo, the coding time is saved by 7%, while the BDRate is increased by only 0.5%.
Claims (3)
1. A fast HEVC coding method is characterized by comprising the following steps:
(1) establishing two threshold arrays T _8x8 and T _16x16 indexed by quantization parameter QP, and initializing T _8x8 and T _16x16 to 0;
(2) setting every N GOPs as a GOP group, wherein the first GOP in the GOP group is a statistical GOP, normally encoding all frames in the statistical GOP, carrying out classification statistics on Merge & Skip _2Nx2N of coding units CU of 8x8 and 16x16 and RDCost of Inter _2Nx2N in the statistical GOP according to quantization parameters QP, wherein the coding units CU of the same quantization parameters QP are in a group, and if the number of statistical samples under a certain quantization parameter QP is more than M, calculating the value of an array element with the quantization parameter QP as an index in a threshold array T _8x8 or T _16x16 by using the group of samples; the value of the array element indexed by the quantization parameter QP in the threshold array T _8x8 or T _16x16 is the maximum value that satisfies the following condition: when the RDcost of Merge & Skip _2Nx2N is less than a certain threshold T, the probability that the RDcost of Merge & Skip _2Nx2N is less than the RDcost of Inter _2Nx2N is greater than or equal to 98%;
(3) the 2 nd to nth GOPs in the GOP group are fast coded GOPs, and the frames in the fast coded GOPs use a fast algorithm: specifically, as follows, the coding quantization parameter of the current coding unit CU is QP, first calculating RDCost of the Merge & Skip _2Nx2N, and then if RDCost of the Merge & Skip _2Nx2N is smaller than the value of the array element indexed by the quantization parameter QP in the threshold array T _8x8 or T _16x16, the coding unit CU directly selects the Merge & Skip _2Nx2N, and skips the calculation of other prediction units; otherwise, calculating the RDCost of all prediction modes, and selecting the RDCost as the minimum as the prediction unit of the coding unit CU; wherein the frames within the fast coded GOP use a fast algorithm: for a coding unit CU of 16x16, the coding quantization parameter is QP, firstly, RDcost of Merge & Skip _2Nx2N is calculated, and the result is RDcost _ Merge & Skip _16x 16; if RDCost _ Merge & Skip _16x16 < T _16x16 value of the array element indexed by the quantization parameter QP, the 16x16 coding unit CU directly selects Merge & Skip _2Nx2N, skipping the calculation of other prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and the prediction unit with the smallest RDCost is selected as the coding unit CU.
2. A fast HEVC coding method as claimed in claim 1, wherein in step (2), N is 10 and M is 1000.
3. A fast HEVC coding method as claimed in claim 1, wherein in step (3) the frames in the fast coding GOP are encoded using the fast algorithm: for an 8x8 coding unit CU, the coding quantization parameter is QP, firstly, RDcost of Merge & Skip _2Nx2N is calculated, and the result is RDcost _ Merge & Skip _8x 8; if RDCost _ Merge & Skip _8x8 < T _8x8 value of the array element indexed by the quantization parameter QP, the coding unit CU of 8x8 directly selects Merge & Skip _2Nx2N, skipping the calculation of other prediction units; otherwise, RDCost of all prediction modes needs to be calculated, and the prediction unit with the smallest RDCost is selected as the coding unit CU.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711289816.0A CN108012152B (en) | 2017-12-07 | 2017-12-07 | Fast HEVC coding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711289816.0A CN108012152B (en) | 2017-12-07 | 2017-12-07 | Fast HEVC coding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108012152A CN108012152A (en) | 2018-05-08 |
| CN108012152B true CN108012152B (en) | 2021-04-13 |
Family
ID=62057467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711289816.0A Active CN108012152B (en) | 2017-12-07 | 2017-12-07 | Fast HEVC coding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108012152B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103051896A (en) * | 2012-12-14 | 2013-04-17 | 华中科技大学 | Mode skipping-based video frequency coding method and mode skipping-based video frequency coding system |
| CN104125469A (en) * | 2014-07-10 | 2014-10-29 | 中山大学 | Fast coding method for high efficiency video coding (HEVC) |
| CN105657433A (en) * | 2015-12-30 | 2016-06-08 | 航天恒星科技有限公司 | Image complexity based signal source real-time coding method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130188736A1 (en) * | 2012-01-19 | 2013-07-25 | Sharp Laboratories Of America, Inc. | High throughput significance map processing for cabac in hevc |
-
2017
- 2017-12-07 CN CN201711289816.0A patent/CN108012152B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103051896A (en) * | 2012-12-14 | 2013-04-17 | 华中科技大学 | Mode skipping-based video frequency coding method and mode skipping-based video frequency coding system |
| CN104125469A (en) * | 2014-07-10 | 2014-10-29 | 中山大学 | Fast coding method for high efficiency video coding (HEVC) |
| CN105657433A (en) * | 2015-12-30 | 2016-06-08 | 航天恒星科技有限公司 | Image complexity based signal source real-time coding method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108012152A (en) | 2018-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102106534B1 (en) | Method and apparatus for entropy coding and decoding | |
| CN103997646B (en) | Fast intra-mode prediction mode selecting method in a kind of HD video coding | |
| RU2581553C2 (en) | Method for encoding video using bias control according to pixel classification and device therefor, video decoding method and device therefor | |
| CN110662078B (en) | 4K/8K ultra-high-definition coding inter-frame coding fast algorithm suitable for AVS2 and HEVC | |
| DK2752005T3 (en) | PROCEDURE FOR DERIVING MOVEMENT INFORMATION | |
| Ma et al. | Low complexity rate distortion optimization for HEVC | |
| KR20130047676A (en) | Method and apparatus for determining context model of transform coefficient level in entropy coding and decoding | |
| CN106534849B (en) | A kind of quick HEVC inter-frame encoding methods | |
| CN104601992B (en) | SKIP mode quick selecting methods based on Bayesian Smallest Risk decision | |
| CN104243997A (en) | Method for quality scalable HEVC (high efficiency video coding) | |
| CN103997645A (en) | Quick HEVC intra-frame coding unit and pattern decision method | |
| CN110896481B (en) | Fast intra mode coding method suitable for HEVC | |
| WO2022121787A1 (en) | Method and apparatus for video predictive coding | |
| CN104954787A (en) | HEVC interframe prediction mode selection method and apparatus | |
| CN107613305B (en) | P, B frame rapid motion estimation method in HEVC | |
| CN104333755A (en) | SKIP/Merge RD Cost based CU (Coding Unit) early termination method of B frame in HEVC (High Efficiency Video Coding) | |
| CN107613294B (en) | Method for rapidly skipping P, B frame intra-frame prediction mode in HEVC | |
| CN109547798B (en) | Rapid HEVC inter-frame mode selection method | |
| KR102329129B1 (en) | Image encoding method and apparatus for compensating for sample values, and image decoding method and apparatus for compensating for sample values | |
| CN108012152B (en) | Fast HEVC coding method | |
| Li et al. | Content based fast intra coding for AVS2 | |
| Lee et al. | Improve the HEVC algorithm complexity based on the visual perception | |
| KR101793623B1 (en) | Method and apparatus for detemining coding unit depth based on history | |
| Park et al. | Fast coding unit (CU) depth decision algorithm for high efficiency video coding (HEVC) | |
| CN111212292B (en) | H.266-based adaptive CU partitioning and skip mode method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 310000 E, 16 floor, A block, Paradise software garden, 3 West Gate Road, Xihu District, Hangzhou, Zhejiang. Applicant after: Hangzhou Dang Hong Polytron Technologies Inc Address before: 310030, 16 floor, E software, Paradise Software Park, 3 West Gate Road, Xihu District, Hangzhou, Zhejiang. Applicant before: HANGZHOU DANGHONG TECHNOLOGY CO., LTD. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |