CN104796694B - Optimization intraframe video coding method based on video texture information - Google Patents
Optimization intraframe video coding method based on video texture information Download PDFInfo
- Publication number
- CN104796694B CN104796694B CN201510216023.0A CN201510216023A CN104796694B CN 104796694 B CN104796694 B CN 104796694B CN 201510216023 A CN201510216023 A CN 201510216023A CN 104796694 B CN104796694 B CN 104796694B
- Authority
- CN
- China
- Prior art keywords
- coding
- unit
- video
- gradient
- coding unit
- 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.)
- Expired - Fee Related
Links
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
A kind of optimization intraframe video coding method based on video texture information of field of video encoding, calculated by I frames gradient information with mapping, intracoded frame all pixels are calculated using video image gradient information and its physical meaning, and mapped for efficient video intra prediction mode；Coding unit is divided, coding unit attribute is judged according to the gradient information of pixel in current coded unit, and combination large code unit includes four attributes and orientation consistency compared with lower Item unit and judges dividing mode from the bottom to top, unnecessary division traversal is skipped；Fast mode decision, using training method under line, is predicted and optimizes to the predictive mode selection under the conditions of different coding cell attribute.The present invention combines video texture characteristic, and the model selection and coding unit division to intraframe video coding are optimized, and improve coding rate under the premise of coding quality is ensured, and ensure that stability of the algorithm to different video sequential coding.
Description
Technical field
The present invention relates to a kind of method of the intraframe video coding of technical field of video coding, specifically one kind is based on
The optimization intraframe video coding method of video texture information.
Background technology
Video coding technique, is that according to compression standard video is converted into video code flow by certain technique compresses.Mesh
Before, develop for different video application field and technology, there is various video coding standard, mainly have MPEG-x series and
H.26x it is serial, such as MPEG-2, MPEG-4 standard, H.263, H.264 standard etc., in addition, also domestic digital audio/video encoding/decoding
The audio/video encoding standard AVS that technical standard working group formulates.With multimedia development, people are for 3D, high definition, ultra high-definition
Demand Deng video strengthens.The thing followed is the requirements at the higher level to video coding efficiency, and be then born HEVC (high
Efficiency video coding, efficient video coding), realize the code efficiency lifting compared to H.264 nearly 50%.
AVS working groups have also made efficient video coding standard of new generation, AVS2.Reached it is close with HEVC, even some
The more efficient video encoding standard of video scene application aspect.In order to pursue the lifting of coding quality, efficient video coding draws
A large amount of new coding techniques are entered, the computation complexity of coding are also considerably increased while quality is lifted, is particularly concentrated on
Intraframe coding module in Video coding so that coding rate is reduced.Efficient video coding standard is optimized, is that video is compiled
One study hotspot of code research.
Spatial coherence and the important evidence that temporal correlation is video compress.In intraframe coding, spatial coherence is past
It is past to play a significant role in an encoding process.The embodiment of video texture information exactly spatial coherence, in image procossing with regarding
Frequency is widely used in encoding.According to the retrieval understanding to prior art, marginal information, spatial coherence statistics, gradient letter
The modes such as breath are the important application to spatial coherence.In video coding process, using video texture information to coding staff
Formula is predicted, and can rationally reduce the prediction that traversal is needed in cataloged procedure and Fractionation regimen, on the premise of Accurate Prediction,
A large amount of computation complexities for reducing efficient video coding, so as to realize the reasonably optimizing of efficient video coding device.To cataloged procedure
Simplification often bring the decline of video encoding quality with substituting, irrational coding framework modification will also result in encoder
It is unstable.How video texture information is accurately applied, with Video coding reasonable combination, realize Accurate Prediction, reduce Video coding
Complexity is an important topic of present encoding research.
By the retrieval discovery to prior art, Chinese patent literature CN103517069A, open (bulletin) day
2014.01.15, a kind of HEVC intra-frame prediction quick mode selection method based on texture analysis is disclosed, this method is to compiling
Code tree unit is carried out before infra-frame prediction, according to the gradient absolute value on the four directions such as level, vertical, lower-left, bottom right and really
The main grain direction and Texture complication of each Unit 4 × 4 in code tree of delimiting the organizational structure unit, and according to texture smooth region use compared with
Big coding unit, texture complex region determines the division of present encoding tree unit using the principle compared with lower Item unit.In prediction
When, according to the main grain direction of predicting unit, some predictive modes most unlikely are excluded, then according to HEVC coding standards
Carry out coarse mode selection and R-D optimized mode selection.The HEVC infra-frame predictions based on texture analysis that the technology is proposed
Fast schema selection method can significantly improve coding rate on the premise of coding quality is ensured.But the prior art and sheet
Invention is compared, and its insurmountable technical problem includes directly and accurately prediction direction estimation, the model selection list simplified
And for the predictive mode selection adjustment and coding unit division of the processing of different video sequence signature.To different video sequence
In terms of intraframe coding stability, the present invention has stronger stability and accuracy.
Chinese patent literature CN103096090A, open (bulletin) day 2013.05.08, discloses a kind of for video
The method that encoding block in compression is divided, it is characterised in that comprise the following steps：The pixel value in whole LCU is read, block is completed
Merge search table；Into the CU of each depth, according to depth and positional information, the depth of the combined block of search table correspondence position is obtained
Degree and positional information；If current CU depth is consistent with combined block depth, the judgement flow of block division methods from the bottom to top is carried out；
Otherwise, the block carried out from top to bottom divides the judgement flow of fast algorithm.The block stroke that two aspects are proposed is combined using the technology
Divide fast algorithm on the premise of the video quality and bit rate output for ensureing HEVC encoders are basically unchanged, greatly accelerate coding
Speed, improves code efficiency.But the prior art is compared with the present invention, its insurmountable technical problem is included fast intuitively
Sdi video correlation and texture information application.Search for table method calculating complexity and optimization property is relatively low, the prior art is even more only
Depth is divided and is modified.The present invention will be predicted to divide with depth and is combined using more concise intuitively computational methods, excellent
Change speed and performance is all greatly promoted.
The content of the invention
The present invention is directed to deficiencies of the prior art, proposes that a kind of optimization frame in based on video texture information is regarded
Frequency coding method, employs the gradient information of video sequence and is screened come the predictive mode to intraframe coding, reduce needs
The rate distortion for carrying out large amount of complex computing judges candidate's amount.And employ and divided for recurrence coding unit in efficient video coding
The Partitioning optimization from the bottom to top of framework.Coding unit is divided into by smooth and coarse two attribute according to video texture complexity,
The merging and division of coding unit are targetedly judged again.While in order to ensure coding framework to different video sequential coding efficiency
Stability, based on experiment, the predictive mode candidate of varying number is selected to different attribute coding unit, so as to ensure coding
The stabilization of quality with it is excellent under the conditions of, greatly improve the speed of Video coding.
The present invention is achieved by the following technical solutions：
The present invention comprises the following steps：
The first step, all pixels to intracoded frame, i.e. I frames carry out gradient calculation, obtain representing pixel motion direction
Grad；Gradient and predictive mode mapping relations figure are drawn by contrasting the relation of intra prediction direction figure and gradient, so that
The optimal prediction modes of each pixel are drawn in an encoding process；
Second step, carries out attributive classification, then with regard to the coding unit of same alike result according to video texture information to encoding block
Merge, then adjustable coding unit division is carried out to the coding unit after optimization；
Described attributive classification refers to：The optimal prediction modes of each pixel of current coded unit are counted, work as system
There is a kind of model selection rate more than 80% in meter result, then judge in current coded unit, pixel has unified operation
Direction, image texture is relatively simple, is smooth unit by the attribute setup of the coding unit；Otherwise by the attribute setup of the unit
For coarse cell.
Described merging refers to：Analysis judgement is carried out from minimum coding unit 8x8 modules, when continuous four identical big
Lower Item unit is smoothing properties and its main gradient direction has uniformity, then this four units can be combined into one it is larger
Unit, and the large code unit is also set to smooth unit, gradient direction is consistent therewith；Otherwise selection current coded unit is big
It is small, operation is not merged to it.
3rd step, sets different mode candidate values according to the different attribute of coding unit and is predicted, then in gradient
Map the pattern that respective amount is selected according to the pattern frequency of occurrences from big to small in the gradient direction counted and carry out RDcost
Calculate, pattern minimum selection RDcost, as optimum prediction mode.
Described prediction, first by way of being trained under line, the different sequences of selection different resolution are surveyed respectively
Examination, coding rate and the optimal balance point of quality are selected according to coding result, then carry out grouping comparison test around equalization point,
Experimental selection goes out the optimal mode number of candidates that smooth unit and coarse cell should be selected respectively.
Technique effect
Compared with prior art, the present invention combines video texture information and the predictive mode of intraframe coding is selected and encoded
Dividing elements are optimized, so as to reduce encoder complexity, improve coding rate.Whole cataloged procedure is divided into two main portions
Point, fast mode decision and fast coding dividing elements.By the gradient information of video texture, and itself and prediction direction and volume
The relation of code dividing elements, simplifies to the calculating process that predictive mode is selected, and the division of coding unit is predicted,
Avoid unnecessary coding from predicting to calculate, so as to improve transcoding speed, save the time.
Brief description of the drawings
Fig. 1 is schematic flow sheet of the present invention.
Fig. 2 is that sequence B Q Square performance curves compare figure；
In figure：QP is respectively 27,32,38,45.
Embodiment
Embodiments of the invention are elaborated below, the present embodiment is carried out lower premised on technical solution of the present invention
Implement, give detailed embodiment and specific operating process, but protection scope of the present invention is not limited to following implementations
Example.
Embodiment 1
As shown in figure 1, the present embodiment comprises the following steps：
Step 1: all pixels to intracoded frame, i.e. I frames carry out gradient calculation, obtain representing pixel motion direction
Grad；Gradient and predictive mode mapping relations figure are drawn by contrasting the relation of intra prediction direction figure and gradient, so that
The optimal prediction modes of each pixel are drawn in an encoding process.
HEVC introduces the brightness mould in substantial amounts of intra prediction direction, such as HEVC to ensure the accuracy of prediction
Block employs in 33 directional prediction patterns and two special predictive modes (DC, planar), AVS2 and employs 30 directions
Property predictive mode and three special predictive modes (DC, plane, bilinearity), above-mentioned Grad and the infra-frame prediction side in HEVC
There are one-to-one mapping relations to sexual norm.
Step 2: analysis judgement is carried out to encoding block according to video texture information, so that quick dividing mode selection is made,
Specially：
2.1) optimal prediction modes of each pixel of current coded unit are counted, it is a kind of when existing in statistical result
Model selection rate then judges in current coded unit that pixel has unified traffic direction, and image texture is more more than 80%
Simply, it is smooth unit by the attribute setup of the coding unit；Otherwise it is coarse cell by the attribute setup of the unit.
2.2) analysis judgement is carried out from minimum coding unit 8x8 modules, when continuous four formed objects coding units
It is smoothing properties and its main gradient direction has uniformity, then this four units can be combined into a larger unit, and should
Large code unit is also set to smooth unit, and gradient direction is consistent therewith；Otherwise current coded unit size is selected, not to it
Merge.
2.3) adjustable coding unit division is carried out to the coding unit after optimization；Based on the recursive calculation side in HEVC
Formula, main computation complexity starts from inner most minimum coding unit, i.e. 8x8 coding units, therefore designed in the present invention
Division determinating mode from the bottom to top can farthest reduce amount of calculation.
Step 3: fast mode decision, specific steps include：
3.1) by the way of being trained under line, the different sequences of selection different resolution are tested respectively, according to coding
As a result selection coding rate and the optimal balance point of quality, then carry out grouping comparison test around equalization point, experimental selection goes out flat
The optimal mode number of candidates that sliding unit should be selected respectively with coarse cell.
3.2) correspondence number is selected according to the pattern frequency of occurrences from big to small in the gradient direction that gradient map is counted
The pattern of amount carries out RDcost (Rate Distortion, rate distortion costs) and calculated, pattern minimum selection RDcost, is
Optimum prediction mode.
Described RDcost, which is calculated, to be referred to：Rate distortion value calculation, i.e., the image fault value in the case of a kind of code check of correspondence.
In summary, the advantage of the invention is that：
1) video texture characteristic is combined, using the spatial coherence of video image, to optimization intraframe video coding device
Predictive mode selection is optimized, and reduces prediction module computation complexity, is saved in the case where ensureing predictablity rate
Time.
2) coding unit is divided with reference to recurrence framework and be optimized, reduce unnecessary traversal, improve coding speed
Degree.
3) determine that optimum prediction mode quantity is selected according to training experiment under different video sequence line, it is ensured that algorithm is not to
With the code efficiency stability of video sequence.
Algorithm proposed by the present invention is tested on AVS2 identifying codes RD_9.0, is configured using full frame intrinsic parameter.Tool
Stereoscopic frequency sequence is selected and parameter meets the setting of its Platform Requirements：
For assessment algorithm performance, analyzed using three below parameter：
PSNR (dB)=PSNRYpro-PSNRYAVS2
BR, ET, PSNR are respectively code check increase percentage, coding reduction percentage of time, the decline of Y-channel Y-PSNR
Value.
Experimental result is as shown in table 1, Fig. 2.
Table 1, QP=43 algorithm experimental result
As can be seen from Table 1, algorithm proposed by the present invention reduces 48% volume to existing newest AVS2 encoder algos
The code time, and influence smaller in terms of code check with signal to noise ratio, it is sufficiently stable to different video sequential coding quality.
In addition, the present invention is not limited only to the example above application, it is equal for all method for video coding based on coding unit
It can be achieved.For under other coding standards according to present invention explanation quoted or convert all should belong to the present invention appended by power
The protection domain that profit is required.
Claims (1)
1. a kind of optimization intraframe video coding method based on video texture information, it is characterised in that comprise the following steps：
The first step, all pixels to intracoded frame, i.e. I frames carry out gradient calculation, obtain representing the ladder in pixel motion direction
Angle value；Gradient and predictive mode mapping relations figure are drawn by contrasting the relation of intra prediction direction figure and gradient, so as to compile
The optimal prediction modes of each pixel are drawn during code；
Second step, attributive classification is carried out according to video texture information to encoding block, is then carried out with regard to the coding unit of same alike result
Merge, then adjustable coding unit division is carried out to the coding unit after optimization；
3rd step, sets different mode candidate values according to the different attribute of coding unit and is predicted, then in gradient map
The pattern for selecting respective amount from big to small according to the pattern frequency of occurrences in the gradient direction counted carries out RDcost calculating,
The pattern for selecting RDcost minimum, as optimum prediction mode；
Described prediction, first by way of being trained under line, the different sequences of selection different resolution are tested respectively, root
Coding rate and the optimal balance point of quality are selected according to coding result, then grouping comparison test, experiment is carried out around equalization point
Select the optimal mode number of candidates that smooth unit and coarse cell should be selected respectively；
Described attributive classification refers to：The optimal prediction modes of each pixel of current coded unit are counted, when statistics knot
There is a kind of model selection rate in fruit more than 80%, then judge in current coded unit, pixel has unified traffic direction,
Image texture is relatively simple, is smooth unit by the attribute setup of the coding unit；Otherwise it is thick by the attribute setup of the unit
Rough unit；
Described merging refers to：Analysis judgement is carried out from minimum coding unit 8x8 modules, when continuous four formed objects are compiled
Code unit is smoothing properties and its main gradient direction has uniformity, then this four units can be combined into a larger list
Member, and the large code unit is also set to smooth unit, gradient direction is consistent therewith；Otherwise selection current coded unit is big
It is small, without union operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510216023.0A CN104796694B (en) | 2015-04-30 | 2015-04-30 | Optimization intraframe video coding method based on video texture information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510216023.0A CN104796694B (en) | 2015-04-30 | 2015-04-30 | Optimization intraframe video coding method based on video texture information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104796694A CN104796694A (en) | 2015-07-22 |
| CN104796694B true CN104796694B (en) | 2017-08-15 |
Family
ID=53561172
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510216023.0A Expired - Fee Related CN104796694B (en) | 2015-04-30 | 2015-04-30 | Optimization intraframe video coding method based on video texture information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104796694B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106101711B (en) * | 2016-08-26 | 2019-03-01 | 成都杰华科技有限公司 | A kind of quick real-time video codec compression algorithm |
| CN108259897B (en) * | 2018-01-23 | 2021-08-27 | 北京易智能科技有限公司 | Intra-frame coding optimization method based on deep learning |
| CN109660793B (en) * | 2018-10-26 | 2021-03-16 | 西安科锐盛创新科技有限公司 | Prediction method for bandwidth compression |
| CN109413421B (en) * | 2018-10-26 | 2021-01-19 | 张豪 | Video encoding method, video encoding apparatus, video decoding method, and video decoding apparatus |
| CN109587481B (en) * | 2018-10-26 | 2020-07-10 | 江苏怀业信息技术股份有限公司 | Video encoding method and apparatus |
| CN109510983B (en) * | 2018-10-26 | 2020-08-11 | 苏州龙盈软件开发有限公司 | Multi-mode selection prediction method for complex texture in bandwidth compression |
| CN109982079B (en) * | 2019-04-03 | 2021-04-02 | 北京电子工程总体研究所 | Intra-frame prediction mode selection method combined with texture space correlation |
| CN110213584A (en) * | 2019-07-03 | 2019-09-06 | 北京电子工程总体研究所 | Coding unit classification method and coding unit sorting device based on Texture complication |
| CN110708551A (en) * | 2019-10-22 | 2020-01-17 | 腾讯科技（深圳）有限公司 | Video encoding method, video encoding device, computer-readable storage medium, and computer apparatus |
| CN111953987B (en) * | 2020-07-15 | 2022-08-09 | 广州柯维新数码科技有限公司 | Video transcoding method, computer device and storage medium |
| CN111953988B (en) * | 2020-07-15 | 2022-08-09 | 广州柯维新数码科技有限公司 | Video transcoding method, computer device and storage medium |
| CN114449273B (en) * | 2020-11-06 | 2023-07-21 | 北京大学 | HEVC-based enhanced block division searching method and device |
| CN113194311B (en) * | 2021-04-27 | 2022-12-27 | 北京京东方技术开发有限公司 | Image processing method, image processor, and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010132977A1 (en) * | 2009-05-21 | 2010-11-25 | Ecole De Technologie Superieure | Method and system for efficient video transcoding |
| JP2011525094A (en) * | 2008-06-16 | 2011-09-08 | ドルビー・ラボラトリーズ・ライセンシング・コーポレーション | Rate control model adaptation based on slice dependency for video coding |
| CN102984521A (en) * | 2012-12-12 | 2013-03-20 | 四川大学 | High-efficiency video coding inter-frame mode judging method based on temporal relativity |
| CN103096090A (en) * | 2013-02-20 | 2013-05-08 | 广州柯维新数码科技有限公司 | Method of dividing code blocks in video compression |
| CN103888762A (en) * | 2014-02-24 | 2014-06-25 | 西南交通大学 | Video coding framework based on HEVC standard |
| 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 |
-
2015
- 2015-04-30 CN CN201510216023.0A patent/CN104796694B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011525094A (en) * | 2008-06-16 | 2011-09-08 | ドルビー・ラボラトリーズ・ライセンシング・コーポレーション | Rate control model adaptation based on slice dependency for video coding |
| WO2010132977A1 (en) * | 2009-05-21 | 2010-11-25 | Ecole De Technologie Superieure | Method and system for efficient video transcoding |
| CN102984521A (en) * | 2012-12-12 | 2013-03-20 | 四川大学 | High-efficiency video coding inter-frame mode judging method based on temporal relativity |
| CN103096090A (en) * | 2013-02-20 | 2013-05-08 | 广州柯维新数码科技有限公司 | Method of dividing code blocks in video compression |
| CN103888762A (en) * | 2014-02-24 | 2014-06-25 | 西南交通大学 | Video coding framework based on HEVC standard |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104796694A (en) | 2015-07-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104796694B (en) | Optimization intraframe video coding method based on video texture information | |
| CN102186070B (en) | Method for realizing rapid video coding by adopting hierarchical structure anticipation | |
| Lei et al. | Fast intra prediction based on content property analysis for low complexity HEVC-based screen content coding | |
| CN103561263B (en) | Based on motion vector constraint and the motion prediction compensation method of weighted motion vector | |
| CN103873861B (en) | Coding mode selection method for HEVC (high efficiency video coding) | |
| CN104378643B (en) | A kind of 3D video depths image method for choosing frame inner forecast mode and system | |
| CN103546749B (en) | Method for optimizing HEVC (high efficiency video coding) residual coding by using residual coefficient distribution features and bayes theorem | |
| CN101783957B (en) | Method and device for predictive encoding of video | |
| CN103248895B (en) | A kind of quick mode method of estimation for HEVC intraframe coding | |
| CN102984521A (en) | High-efficiency video coding inter-frame mode judging method based on temporal relativity | |
| CN103596004A (en) | Intra-frame prediction method and device based on mathematical statistics and classification training in HEVC | |
| CN104837019B (en) | AVS to HEVC optimization video transcoding methods based on SVMs | |
| CN101710993A (en) | Block-based self-adaptive super-resolution video processing method and system | |
| CN103067704B (en) | A kind of method for video coding of skipping in advance based on coding unit level and system | |
| CN107277506B (en) | Motion vector accuracy selection method and device based on adaptive motion vector precision | |
| CN104853191B (en) | A kind of HEVC fast encoding method | |
| CN106303521B (en) | A kind of HEVC Rate-distortion optimization method based on sensitivity of awareness | |
| CN105898332B (en) | For the fast deep figure frame mode decision method of 3D-HEVC coding standards | |
| CN105187826B (en) | For the fast intra mode decision method of high efficiency video encoding standard | |
| CN107018412B (en) | A kind of DVC-HEVC video transcoding method based on key frame coding unit partition mode | |
| Chen et al. | A novel fast intra mode decision for versatile video coding | |
| CN110213584A (en) | Coding unit classification method and coding unit sorting device based on Texture complication | |
| CN109587503A (en) | A kind of 3D-HEVC depth map intra-frame encoding mode high-speed decision method based on edge detection | |
| CN107318016A (en) | A kind of HEVC inter-frame forecast mode method for rapidly judging based on zero piece of distribution | |
| CN108769696A (en) | A kind of DVC-HEVC video transcoding methods based on Fisher discriminates |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170815 Termination date: 20200430 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |