CN109788291A - A kind of digital video transform method, device, equipment and storage medium - Google Patents
A kind of digital video transform method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN109788291A CN109788291A CN201910111024.7A CN201910111024A CN109788291A CN 109788291 A CN109788291 A CN 109788291A CN 201910111024 A CN201910111024 A CN 201910111024A CN 109788291 A CN109788291 A CN 109788291A
- Authority
- CN
- China
- Prior art keywords
- transformation
- residual block
- kernel
- transformation kernel
- block
- 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.)
- Pending
Links
Abstract
This application discloses a kind of digital video transform method, device, equipment and storage mediums, are related to digital signal processing technique field.The described method includes: dividing digital video, multiple luminance blocks are obtained；The luminance block is predicted, corresponding first residual block is obtained；First transformation kernel is overturn, the second transformation kernel is obtained；Pre-transform is successively carried out to the first residual block using third transformation kernel, the first transformation kernel and the second transformation kernel, obtains corresponding transformation coefficient；According to transformation coefficient, optimal transformation core is determined；The first residual block is checked using optimal transformation to be converted.In this method, it is current it is believed that on the basis of most suitable mapping mode in field of digital video remaining, by introducing new transformation kernel, and after to this, new transformation kernel is overturn, therefrom selection optimal transformation is checked the first residual block and is converted, and is greatly improved energy accumulating effect and property.
Description
Technical field
The present invention relates to digital signal processing technique field more particularly to a kind of digital video transform method, device, equipment
And storage medium.
Background technique
In digital video coding field, prediction, transformation, quantization, entropy coding, the several portions of loop filtering are divided on big frame
Point.Conversion section therein, DCT2 transformation kernel is by it is believed that be most suitable mapping mode all the time.But DCT2 is
A kind of ideal KLT transformation is approximate, does not represent it and is suitable for all residual blocks.In the small size residual block of discrete space
Above, the correlation in block between coefficient will not perfectly approach 1, have the mode of signal distributions quite different for transformation.Therefore, certain
Residual block has special regular governed conversion characteristics, then it is more suitable for other mapping modes.
Summary of the invention
It is a primary object of the present invention to propose a kind of digital video transform method, device, equipment and storage medium, it is intended to
One of the technical problem in above-mentioned technical problem is solved at least in the first degree.
To achieve the above object, first aspect present invention provides a kind of digital video transform method, the method includes with
Lower step:
Digital video is divided, multiple luminance blocks are obtained；
The luminance block is predicted, corresponding first residual block is obtained；
First transformation kernel is overturn, the second transformation kernel is obtained；
Successively first residual block is carried out using third transformation kernel, first transformation kernel and second transformation kernel
Pre-transform obtains corresponding transformation coefficient；
According to the transformation coefficient, optimal transformation core is determined；
First residual block is checked using the optimal transformation to be converted.
Digital video transform method provided by the invention, remain it is current it is believed that in field of digital video it is most suitable
Mapping mode on the basis of, by introducing new transformation kernel, and after the transformation kernel new to this is overturn, successively use each change
It changes the first residual block of verification to be converted, so as to determine it most according to the distribution character of signal in the first different residual blocks
Excellent transformation kernel, rather than it is unalterable each first residual block is converted using the same transformation kernel, greatly promoted
Energy accumulating effect and property；Furthermore direction overturning is carried out by the new transformation kernel to introducing, can not only be guaranteed
Transformed each band information is still assembled, while being conducive to the energy accumulating that opposite signal is distributed with current demand signal variance.
In addition, digital video transform method provided by the invention can also have following additional technical feature:
Preferably, the division digital video, obtains multiple luminance blocks, comprising:
Digital video is divided, multiple coding units are obtained；
The multiple coding unit YUV component is separated, multiple luminance blocks are obtained.
Preferably, described that first transformation kernel is overturn, obtain the second transformation kernel, comprising:
Vertical direction overturning and/or horizontal direction overturning are carried out to the first transformation kernel, obtain the second transformation kernel.
Preferably, described according to the transformation coefficient, determine optimal transformation core, comprising:
Respectively according to the transformation coefficient, the rate distortion costs of corresponding first residual block are calculated；
Compare the rate distortion costs, obtains minimum rate distortion costs；
Using the corresponding transformation kernel of the minimum rate distortion costs as optimal transformation core.
Preferably, method further include: save divide digital video division information, the luminance block is predicted it is pre-
The identification information of measurement information, each transformation kernel；
Correspondingly, it is described according to the transformation coefficient, calculate separately the rate distortion costs of corresponding first residual block, comprising:
The transformation coefficient is quantified, quantization parameter is obtained；
Inverse quantization and inverse transformation are carried out to the quantization parameter, obtain corresponding second residual block；
Compare corresponding first residual block of the transformation coefficient and the second residual block, obtains the distortion of corresponding residual error；
Entropy coding is carried out respectively to the quantization parameter, the division information, the predictive information and the identification information,
Obtain the whole code rate of corresponding bright block；
According to the distortion of the residual error and the whole code rate of corresponding luminance block, the rate distortion of corresponding first residual block is calculated
Cost.
Preferably, described that the luminance block is predicted, obtain corresponding first residual block, comprising: to the brightness
Block carries out intra prediction or inter-prediction, obtains corresponding first residual block；
When carrying out pre-transform to first residual block using the third transformation kernel, to the corresponding transformation coefficient amount of progress
Change, after obtaining quantization parameter, further includes: if the quantization parameter is full null matrix, using the third transformation kernel as most
Excellent transformation kernel.
Preferably, described that the luminance block is predicted, obtain corresponding first residual block, comprising: to the brightness
Block carries out intra prediction, obtains corresponding first residual block；
When carrying out pre-transform to first residual block using the third transformation kernel, according to corresponding transformation coefficient, meter
After the rate distortion costs for calculating corresponding first residual block, further includes:
Determine the current minimum rate distortion costs in each rate distortion costs currently obtained；
Calculate the corresponding rate distortion threshold of first residual block；
If the ratio of the rate distortion costs currently calculated and the current minimum rate distortion costs is more than that the rate is distorted threshold
Value, then using the third transformation kernel as optimal transformation core.
Preferably, method further include: record the information of the optimal transformation core of each first residual block.
In addition, to achieve the above object, second aspect of the present invention provides a kind of digital video converting means, comprising:
Division module obtains multiple luminance blocks for dividing digital video；
Prediction module obtains corresponding first residual block for predicting the luminance block；
Flip module obtains the second transformation kernel for carrying out direction overturning to the first transformation kernel；
Pre-transform module, for successively using third transformation kernel, first transformation kernel and second transformation kernel to institute
It states the first residual block and carries out pre-transform, obtain corresponding transformation coefficient；
Determining module, for determining optimal transformation core according to the transformation coefficient；
Conversion module is converted for checking first residual block using the optimal transformation.
In addition, to achieve the above object, third aspect present invention provides a kind of equipment, comprising: memory and processor；
It is stored with computer program on the memory, when the computer program is executed by the processor, realizes such as
Method described in first aspect present invention.
In addition, to achieve the above object, fourth aspect present invention provides a kind of computer readable storage medium, the calculating
It is stored with computer program on machine readable storage medium storing program for executing, realizes when the computer program is executed by processor such as the present invention first
Method described in aspect.
Additional aspect and advantage of the invention can be set forth in part in the description, partially will from the following description
Become obvious, or practice through the invention is recognized.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the application
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Attached drawing 1 is a kind of flow diagram of digital video transform method provided by one embodiment of the present invention；
Attached drawing 2 is a kind of schematic diagram that direction overturning is carried out to the first transformation kernel provided by one embodiment of the present invention；
Attached drawing 3 is the record schematic diagram of the upset information of the first transformation kernel provided by one embodiment of the present invention；
Attached drawing 4 is a kind of module composition block diagram of digital video converting means provided by one embodiment of the present invention.
Specific embodiment
The illustrative embodiments of the disclosure are more fully described below with reference to accompanying drawings.Although showing this public affairs in attached drawing
The illustrative embodiments opened, it being understood, however, that may be realized in various forms the disclosure without the reality that should be illustrated here
The mode of applying is limited.It is to be able to thoroughly understand the disclosure on the contrary, providing these embodiments, and can be by this public affairs
The range opened is fully disclosed to those skilled in the art.
Fig. 1 is a kind of flow diagram of digital video transform method provided in an embodiment of the present invention；As shown in Figure 1, institute
State method the following steps are included:
Step S11: digital video is divided, multiple luminance blocks are obtained；
In one embodiment of the application, step S11 includes:
Digital video is divided, multiple coding units are obtained；
The multiple coding unit YUV component is separated, multiple luminance blocks are obtained.
Preferably, the luminance block obtained is the block of 4x4 to 64x64 size.
In the present embodiment, step S11 further include: save the division information for dividing digital video；Wherein, division information packet
Include the information for each coding unit of mark that maximum coding unit is divided by recurrence.
Step S12: predicting each luminance block, obtains corresponding first residual block；
In the present embodiment, step S12 can for each luminance block carry out inter-prediction, can also for each luminance block into
Row intra prediction obtains corresponding first residual block；
In the present embodiment, it is illustrated for carrying out inter-prediction to each luminance block, step S12 includes:
Using the luminance block of current predictive as present intensity block；
If present intensity block is first luminance block, pixel value and preset pixel value to present intensity block compare
It is right, obtain the first residual block of present intensity block；
If present intensity block is non-first luminance block, pixel value and present intensity block to present intensity block it is previous
The pixel value of luminance block is compared, and obtains the first residual block of present intensity block.
It is noted that predicting in the application each luminance block, the method for obtaining corresponding first residual block is unlimited
For the above method, any existing prediction technique is applicable, is no longer described in detail one by one herein.
In the present embodiment, step S12 further include: save the predictive information predicted each luminance block；Wherein, it predicts
Information includes the prediction mode information that each coding unit uses, such as inter prediction or infra-frame prediction.
Step S13: overturning the first transformation kernel, obtains the second transformation kernel；
It is common, DCT2 transformation kernel by it is believed that be most suitable mapping mode in field of digital video, however for
Different signal distributions situation, certain conversion characteristics with particular law, is more suitable for other mapping modes；Therefore, this reality
It applies in example, on the basis of retaining DCT2 transformation kernel, new transformation kernel is introduced, to become to the residual block with particular law
It changes；DCT2 transformation kernel to be denoted as third transformation kernel in the present embodiment, and the transformation kernel newly introduced is denoted as first convenient for distinguishing
Transformation kernel, the transformation kernel after overturning to the transformation kernel newly introduced are denoted as the second transformation kernel；Wherein, the first transformation kernel can be
DST7 transformation kernel can also be DST8 transformation kernel；It is illustrated so that the first transformation kernel is DST7 transformation kernel as an example below.
Further, vertical direction overturning is carried out to the first transformation kernel and/or horizontal direction is overturn, obtain the second transformation
Core；Attached drawing 2, which is shown, carries out vertical direction overturning, the second obtained transformation kernel to the first transformation kernel of 4*4 size.Wherein,
One transformation kernel is followed successively by low frequency transform base system from top to bottom and counts to high frequency conversion base system number, by each frequency of the first transformation kernel
Coefficient do once inside out operation, without exchange frequency range sequence, can not only guarantee that transformed low-frequency information is still assembled,
Be conducive to the energy accumulating that opposite signal is distributed with current demand signal variance simultaneously.
Step S14: successively each first residual block is carried out using third transformation kernel, the first transformation kernel and the second transformation kernel pre-
Transformation, obtains corresponding transformation coefficient；
In the present embodiment, third transformation kernel is specially transformation kernel used by DCT2 mapping mode；Further, it uses
Transformation kernel carries out the process of pre-transform to the first residual block, with the existing process phase converted using transformation kernel to residual block
Together, this will not be detailed here.
Further, method further include: be each transformation kernel allocation identification information, correspondingly, using each transformation kernel to the
One residual block carries out pre-transform, after obtaining corresponding transformation coefficient, further includes: save the identification information of corresponding transformation kernel.
Step S15: according to obtained transformation coefficient, optimal transformation core is determined；
In one embodiment of the application, step S15 includes:
Step S15-1: respectively according to transformation coefficient, the rate distortion costs of corresponding first residual block are calculated；
In one embodiment of the application, step S15-1 includes:
Step A1: quantifying transformation coefficient, obtains quantization parameter；
Step A2: inverse quantization and inverse transformation are carried out to quantization parameter, obtain corresponding second residual block；
Step A3: compare corresponding first residual block of transformation coefficient and the second residual block, obtain the distortion of corresponding residual error；
Step A4: entropy volume is carried out respectively to the division information of quantization parameter, preservation, the predictive information of preservation and identification information
Code, obtains the whole code rate of corresponding bright block；
Step A5: according to the distortion of obtained residual error and the whole code rate of corresponding luminance block, corresponding first residual error is calculated
The rate distortion costs of block.
Specifically, by RDO (Rate-Distortion Optimization, rate-distortion optimization) to obtained residual error
The whole code rate of distortion and corresponding luminance block is calculated, and the rate distortion costs of corresponding first residual block are obtained.
It is noted that above-mentioned carry out quantization to transformation coefficient and carry out the side of inverse quantization and inverse transformation to quantization parameter
Method can be used existing quantization and inverse quantization and inverse transform method, no longer be described in detail in the present embodiment.
Step S15-2: compare each rate distortion costs being calculated, obtain minimum rate distortion costs；
Specifically, to calculating to each rate distortion costs be ranked up, obtain minimum rate distortion costs；Wherein, it sorts
It can be ascending sort, or descending sort.
Step S15-3: using the corresponding transformation kernel of minimum rate distortion costs as optimal transformation core.
All the time, DCT2 transformation kernel is by it is believed that be most suitable mapping mode in field of digital video, and DST7
To a kind of special vector, i.e. signal variance is from left to right gradually increased with vector position, has good energy accumulating effect,
The second transformation kernel after overturning to DST7 transformation kernel, the block that signal variance can be made bigger to the lower right corner have better energy
Measure congregational rate；In the present embodiment, be respectively adopted third transformation kernel, the first transformation kernel and the second transformation kernel to the first residual block into
Row pre-transform, and corresponding rate distortion costs are calculated, so as to according to the distribution character of signal in the first different residual blocks
Determine its optimal transformation kernel, rather than it is unalterable each first residual block is converted using the same transformation kernel, pole
Big improves energy accumulating effect and property.
Step S16: corresponding first residual block is checked using optimal transformation and is converted.
In one embodiment of the application, no matter intra prediction is carried out to luminance block in step s 12 or carries out frame
Between predict, using third transformation kernel to the first residual block carry out pre-transform when, quantify to corresponding transformation coefficient, obtain
To after quantization parameter, the type for the quantization parameter that can also be comprised determining that, if obtained quantization parameter is full zero moment
Battle array, then using the third transformation kernel as optimal transformation core.
In the present embodiment, due in sequence successively using third transformation kernel, the first transformation kernel and the second transformation kernel to same
One the first residual block is converted, and after being converted using third transformation kernel, passes through the class to obtained quantization parameter
Type is judged, if quantization parameter is full null matrix, then it is assumed that the rate of having reached is distorted relatively good tradeoff, and there is no need to make
The first current residual block is converted with the first transformation kernel and the second transformation kernel, improves optimal transformation core constant speed really
Rate.
In one embodiment of the application, if carrying out intra prediction to luminance block in step S12, become using third
When changing verification the first residual block progress pre-transform, according to corresponding transformation coefficient, the rate for calculating corresponding first residual block is distorted generation
After valence, can also include:
Step B1: the current minimum rate distortion costs in each rate distortion costs currently obtained are determined；
Specifically, being ranked up to each rate distortion costs currently obtained, current minimum rate distortion costs are obtained.
Step B2: the corresponding rate distortion threshold of the first residual block is calculated；
Specifically, calculating each first according to formula Threshold=1.0+0.05* (14-log 2 (Width*Height))
The corresponding rate distortion threshold of residual block, wherein Width is the width of the first residual block, and Height is the height of the first residual block；
For example, if 12, then threshold value is 1.1 for the value of different log 2 (Width*Height)；When if 11, then
Threshold value is 1.15.Will uniformly because the distribution of bulk residual error is opposite, threshold value can be set smaller, skip as far as possible more
Some mapping modes；And fritter residual error is more uneven due to being distributed, then needs to be arranged big threshold value, guarantees more become as far as possible
It changes in the selection for participating in RDO.
Step B3: if the ratio of the rate distortion costs currently calculated and current minimum rate distortion costs is more than that rate is distorted threshold
Value, then using third transformation kernel as optimal transformation core.
In the present embodiment, if the ratio of the rate distortion costs currently calculated and current minimum rate distortion costs is distorted more than rate
Threshold value, then it is assumed that the rate of having reached is distorted relatively good tradeoff, and there is no need to use the first transformation kernel and the second transformation kernel pair
The first current residual block is converted；As a result, by being converted and being calculated accordingly using third transformation kernel every time
Rate distortion costs after, execute step B1-B3, improve optimal transformation core constant speed rate really.
Based on any of the above-described embodiment, in one embodiment of the application, method further include: record each first residual block
Optimal transformation core information.
Wherein, the information of optimal transformation core includes the type information of optimal transformation core or the type letter of optimal transformation core
Breath and upset information etc., wherein type information is the character of 1 bit, it is preferred that character 0 indicates to use third transformation kernel pair
The transformation that first residual block carries out, character 1 indicate the change carried out using the first transformation kernel or the second transformation kernel to the first residual block
It changes；It further, further include the upset information (Index) of optimal transformation core when type information is 1, which is 2 ratios
The character of special position, wherein the character of first bit, for indicating whether horizontal direction overturns, the character of the second bit is used for table
Show whether vertical direction overturns；Preferably, the expression of character 1 is not overturn, and character 0 indicates overturning；In the present embodiment, with the first transformation
For core is DST7 transformation kernel, upset information be can be found in such as Fig. 3, two character representation upset informations in bracket, example
Such as (11), indicate that horizontal direction is not overturn, vertical direction is not overturn, i.e., optimal transformation core is the first transformation kernel, for another example (10) table
Show horizontal direction do not overturn, vertical direction overturning, i.e. optimal transformation core is the carried out after vertical direction overturning to the first transformation
Two transformation kernels；It is noted that the number " 0,1,2,3 " before 3 bracket of attached drawing only indicates different upset informations, in reality
Cataloged procedure in without in all senses.
As a result, remaining current it is believed that passing through on the basis of most suitable mapping mode in field of digital video
New transformation kernel is introduced, and after the transformation kernel new to this is overturn, successively residual block is converted using each transformation kernel, from
And its optimal transformation kernel can be determined according to the distribution character of signal in different residual blocks, rather than unalterable use
The same transformation kernel converts each residual block, is greatly improved energy accumulating effect and property；Furthermore by right
The new transformation kernel introduced carries out direction overturning, can not only guarantee that transformed each band information is still assembled, while advantageous
In the energy accumulating for being distributed opposite signal with current demand signal variance.
It is a kind of digital video transform method provided by the embodiments of the present application above, corresponds to the above method, this hair
It is bright that a kind of digital video converting means is also provided, since the implementation that described device solves the problems, such as is similar to the above method, because
The content of this and method partial response, the detailed description that reference can be made to the above method embodiment are subsequent not repeat them here.It is understood that
, device provided by the present application may include the unit or module for being able to carry out each step in above method example, these
Unit or module can realize that the present invention does not limit by way of hardware, software or soft or hard combination.Below with reference to Fig. 4
It specifically describes；
Specifically, as shown in figure 4, digital video converting means provided by the embodiments of the present application includes:
Division module 21 obtains multiple luminance blocks for dividing digital video；
Prediction module 22 obtains corresponding first residual block for predicting the luminance block；
Flip module 23 obtains the second transformation kernel for overturning to the first transformation kernel；
Pre-transform module 24, for successively using third transformation kernel, first transformation kernel and second transformation kernel pair
First residual block carries out pre-transform, obtains corresponding transformation coefficient；
First determining module 25, for determining optimal transformation core according to the transformation coefficient；
Conversion module 26 is converted for checking first residual block using the optimal transformation.
In one embodiment of the application, division module 21 is specifically used for:
Digital video is divided, multiple coding units are obtained；
The multiple coding unit YUV component is separated, multiple luminance blocks are obtained.
In one embodiment of the application, flip module 23 is specifically used for:
Vertical direction overturning and/or horizontal direction overturning are carried out to the first transformation kernel, obtain the second transformation kernel
In one embodiment of the application, determining module 25 includes:
Computational submodule, for according to the transformation coefficient, calculating the rate distortion costs of corresponding first residual block respectively；
Comparative sub-module is used for the rate distortion costs, obtains minimum rate distortion costs；
As submodule, for using the corresponding transformation kernel of the minimum rate distortion costs as optimal transformation core.
In one embodiment of the application, device further include:
Preserving module, for saving the division information for dividing digital video, the prediction predicted the luminance block is believed
The identification information of breath, the corresponding transformation kernel of each transformation coefficient；
Correspondingly, computational submodule is specifically used for:
The transformation coefficient is quantified, quantization parameter is obtained；
Inverse quantization and inverse transformation are carried out to the quantization parameter, obtain corresponding second residual block；
Compare corresponding first residual block of the transformation coefficient and the second residual block, obtains the distortion of corresponding residual error；
Entropy coding is carried out respectively to the quantization parameter, the division information, the predictive information and the identification information,
Obtain the whole code rate of corresponding bright block；
According to the distortion of the residual error and the whole code rate of corresponding luminance block, the rate distortion of corresponding first residual block is calculated
Cost.
In one embodiment of the application, prediction module 22 is specifically used for: to the luminance block carry out intra prediction or
Inter-prediction obtains corresponding first residual block；
Correspondingly, device further include: first is used as module, residual to first using third transformation kernel for pre-transform module 21
When poor block carries out pre-transform, computational submodule quantifies corresponding transformation coefficient, after obtaining quantization parameter, if the amount
Changing coefficient is full null matrix, then using the third transformation kernel as optimal transformation core.
In one embodiment of the application, prediction module 22 is specifically used for: carrying out intra prediction to the luminance block, obtains
To corresponding first residual block；
Correspondingly, device further include:
Second determining module, when carrying out pre-transform to the first residual block using third transformation kernel for pre-transform module 21,
Computational submodule is according to corresponding transformation coefficient, and after the rate distortion costs for calculating corresponding first residual block, determination has currently been obtained
To each rate distortion costs in current minimum rate distortion costs；
Computing module, for calculating the corresponding rate distortion threshold of first residual block；
Second is used as module, if the ratio of rate distortion costs and the current minimum rate distortion costs for currently calculating
More than the rate distortion threshold, then using the third transformation kernel as optimal transformation core.
In one embodiment of the application, device further include: logging modle, for recording the optimal of each first residual block
The information of transformation kernel.
Digital video converting means provided by the embodiments of the present application, the digital video transformation side provided with previous embodiment
Method has the same effect for identical inventive concept.
In addition, to realize above-described embodiment, the embodiment of the present application also provides a kind of equipment, comprising: memory and processor；
It is stored with computer program on the memory, when the computer program is executed by the processor, is realized
Digital video transform method described in aforementioned any embodiment.
In addition, to realize above-described embodiment, the embodiment of the present application also provides a kind of computer readable storage medium, the meter
It is stored with computer program on calculation machine readable storage medium storing program for executing, when the computer program is executed by processor, realizes aforementioned any
Digital video transform method described in embodiment.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, method, article or the system that include a series of elements not only include those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or system institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including being somebody's turn to do
There is also other identical elements in the process, method of element, article or system.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can be realized by means of software and necessary general hardware platform, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on this understanding, technical solution of the present invention substantially in other words does the prior art
The part contributed out can be embodied in the form of software products, which is stored in one as described above
In storage medium (such as ROM/RAM, magnetic disk, CD), including some instructions are used so that terminal device (it can be mobile phone,
Computer, server or network equipment etc.) execute method described in each embodiment of the present invention.
The preferable specific embodiment of the above, only the application, but the protection scope of the application is not limited thereto,
Within the technical scope of the present application, any changes or substitutions that can be easily thought of by anyone skilled in the art,
Should all it cover within the scope of protection of this application.Therefore, the protection scope of the application should be with scope of protection of the claims
It is quasi-.
Claims (10)
1. a kind of digital video transform method characterized by comprising
Digital video is divided, multiple luminance blocks are obtained；
The luminance block is predicted, corresponding first residual block is obtained；
First transformation kernel is overturn, the second transformation kernel is obtained；
Successively first residual block is become in advance using third transformation kernel, first transformation kernel and second transformation kernel
It changes, obtains corresponding transformation coefficient；
According to the transformation coefficient, optimal transformation core is determined；
First residual block is checked using the optimal transformation to be converted.
2. being wrapped the method according to claim 1, wherein the division digital video, obtains multiple luminance blocks
It includes:
Digital video is divided, multiple coding units are obtained；
The multiple coding unit YUV component is separated, multiple luminance blocks are obtained.
3. obtaining the second change the method according to claim 1, wherein described overturn the first transformation kernel
Change core, comprising:
Vertical direction overturning and/or horizontal direction overturning are carried out to the first transformation kernel, obtain the second transformation kernel.
4. the method according to claim 1, wherein described determine optimal transformation core according to the transformation coefficient,
Include:
Respectively according to the transformation coefficient, the rate distortion costs of corresponding first residual block are calculated；
Compare the rate distortion costs, obtains minimum rate distortion costs；
Using the corresponding transformation kernel of the minimum rate distortion costs as optimal transformation core.
5. according to the method described in claim 4, it is characterized in that, method further include: save the division letter for dividing digital video
Breath, the predictive information that the luminance block is predicted, the corresponding transformation kernel of each transformation coefficient identification information；
It is described according to the transformation coefficient, calculate separately the rate distortion costs of corresponding first residual block, comprising:
The transformation coefficient is quantified, quantization parameter is obtained；
Inverse quantization and inverse transformation are carried out to the quantization parameter, obtain corresponding second residual block；
Compare corresponding first residual block of the transformation coefficient and the second residual block, obtains the distortion of corresponding residual error；
Entropy coding is carried out to the quantization parameter, the division information, the predictive information and the identification information respectively, is obtained
The whole code rate of corresponding bright block；
According to the distortion of the residual error and the whole code rate of corresponding luminance block, the rate for calculating corresponding first residual block is distorted generation
Valence.
6. the method according to claim 5, which is characterized in that
It is described that the luminance block is predicted, obtain corresponding first residual block, comprising: carry out to the luminance block pre- in frame
Survey or inter-prediction, obtain corresponding first residual block；
When carrying out pre-transform to first residual block using the third transformation kernel, corresponding transformation coefficient is quantified,
After obtaining quantization parameter, further includes: if the quantization parameter is full null matrix, using the third transformation kernel as optimal change
Change core.
7. according to the method described in claim 5, it is characterized in that,
It is described that the luminance block is predicted, obtain corresponding first residual block, comprising: carry out to the luminance block pre- in frame
It surveys, obtains corresponding first residual block；
When carrying out pre-transform to first residual block using the third transformation kernel, according to corresponding transformation coefficient, phase is calculated
After the rate distortion costs for the first residual block answered, further includes:
Determine the current minimum rate distortion costs in each rate distortion costs currently obtained；
Calculate the corresponding rate distortion threshold of first residual block；
If the ratio of the rate distortion costs currently calculated and the current minimum rate distortion costs is more than the rate distortion threshold,
Using the third transformation kernel as optimal transformation core.
8. method according to claim 1-7, which is characterized in that method further include: record each first residual block
Optimal transformation core information.
9. a kind of digital video converting means characterized by comprising
Division module obtains multiple luminance blocks for dividing digital video；
Prediction module obtains corresponding first residual block for predicting the luminance block；
Flip module obtains the second transformation kernel for carrying out direction overturning to the first transformation kernel；
Pre-transform module, for successively using third transformation kernel, first transformation kernel and second transformation kernel to described the
One residual block carries out pre-transform, obtains corresponding transformation coefficient；
Determining module, for determining optimal transformation core according to the transformation coefficient；
Conversion module is converted for checking first residual block using the optimal transformation.
10. a kind of equipment characterized by comprising memory and processor；
It is stored with computer program on the memory, when the computer program is executed by the processor, is realized as weighed
Benefit requires the described in any item methods of 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910111024.7A CN109788291A (en) | 2019-02-12 | 2019-02-12 | A kind of digital video transform method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910111024.7A CN109788291A (en) | 2019-02-12 | 2019-02-12 | A kind of digital video transform method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109788291A true CN109788291A (en) | 2019-05-21 |
Family
ID=66504355
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910111024.7A Pending CN109788291A (en) | 2019-02-12 | 2019-02-12 | A kind of digital video transform method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109788291A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111988618A (en) * | 2019-05-22 | 2020-11-24 | 杭州海康威视数字技术股份有限公司 | Decoding and encoding method, decoding end and encoding end |
| CN114339230A (en) * | 2022-03-03 | 2022-04-12 | 杭州未名信科科技有限公司 | Transformation core selection method and device for video coding, storage medium and terminal |
| CN114827629A (en) * | 2019-08-23 | 2022-07-29 | 杭州海康威视数字技术股份有限公司 | Encoding method, decoding method, device and machine readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014036278A (en) * | 2012-08-07 | 2014-02-24 | Nippon Hoso Kyokai <Nhk> | Image encoder, image decoder and program |
| CN104604224A (en) * | 2012-09-04 | 2015-05-06 | 高通股份有限公司 | Transform basis adjustment in scalable video coding |
| CN106464878A (en) * | 2014-05-14 | 2017-02-22 | 联发科技（新加坡）私人有限公司 | Method of alternative transform for data compression |
| CN107005695A (en) * | 2014-11-28 | 2017-08-01 | 联发科技股份有限公司 | Method and device for the alternative transforms of Video coding |
| US20180249179A1 (en) * | 2017-02-28 | 2018-08-30 | Google Inc. | Transform Kernel Selection and Entropy Coding |
| CN109089117A (en) * | 2017-05-11 | 2018-12-25 | 联发科技股份有限公司 | The method and device of coding or decoding video data |
-
2019
- 2019-02-12 CN CN201910111024.7A patent/CN109788291A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014036278A (en) * | 2012-08-07 | 2014-02-24 | Nippon Hoso Kyokai <Nhk> | Image encoder, image decoder and program |
| CN104604224A (en) * | 2012-09-04 | 2015-05-06 | 高通股份有限公司 | Transform basis adjustment in scalable video coding |
| CN106464878A (en) * | 2014-05-14 | 2017-02-22 | 联发科技（新加坡）私人有限公司 | Method of alternative transform for data compression |
| CN107005695A (en) * | 2014-11-28 | 2017-08-01 | 联发科技股份有限公司 | Method and device for the alternative transforms of Video coding |
| US20180249179A1 (en) * | 2017-02-28 | 2018-08-30 | Google Inc. | Transform Kernel Selection and Entropy Coding |
| CN109089117A (en) * | 2017-05-11 | 2018-12-25 | 联发科技股份有限公司 | The method and device of coding or decoding video data |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111988618A (en) * | 2019-05-22 | 2020-11-24 | 杭州海康威视数字技术股份有限公司 | Decoding and encoding method, decoding end and encoding end |
| CN111988618B (en) * | 2019-05-22 | 2022-05-20 | 杭州海康威视数字技术股份有限公司 | Decoding and encoding method, decoding end and encoding end |
| CN114827629A (en) * | 2019-08-23 | 2022-07-29 | 杭州海康威视数字技术股份有限公司 | Encoding method, decoding method, device and machine readable storage medium |
| CN114827629B (en) * | 2019-08-23 | 2024-02-27 | 杭州海康威视数字技术股份有限公司 | Encoding method, decoding method, apparatus, and machine-readable storage medium |
| CN114339230A (en) * | 2022-03-03 | 2022-04-12 | 杭州未名信科科技有限公司 | Transformation core selection method and device for video coding, storage medium and terminal |
| CN114339230B (en) * | 2022-03-03 | 2022-09-02 | 杭州未名信科科技有限公司 | Transformation core selection method and device for video coding, storage medium and terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101971631B (en) | Adaptive intra mode selection | |
| CN109788291A (en) | A kind of digital video transform method, device, equipment and storage medium | |
| CN102415097B (en) | Distortion weighing | |
| CN106716997A (en) | Video encoding method using in-loop filter parameter prediction and apparatus therefor, and video decoding method and apparatus therefor | |
| CN108965869A (en) | The method and apparatus of image prediction | |
| US20170230656A1 (en) | Sample adaptive offset systems and methods | |
| CN108055544A (en) | SAO coding methods and equipment and SAO coding/decoding methods and equipment | |
| CN103067704B (en) | A kind of method for video coding of skipping in advance based on coding unit level and system | |
| CN111741297B (en) | Inter-frame prediction method, video coding method and related devices | |
| CN103596004A (en) | Intra-frame prediction method and device based on mathematical statistics and classification training in HEVC | |
| CN106961604B (en) | Data Rate Distribution, coding unit code rate allocation method and the device of Video coding | |
| CN112672149B (en) | Video processing method and device, storage medium and server | |
| CN106162176A (en) | Method for choosing frame inner forecast mode and device | |
| CN112055203A (en) | Inter-frame prediction method, video coding method and related devices thereof | |
| Medhat et al. | Adaptive low‐complexity motion estimation algorithm for high efficiency video coding encoder | |
| CN110087075B (en) | Image coding method, coding device and computer storage medium | |
| CN110381312B (en) | HEVC (high efficiency video coding) -based depth partition range prediction method and device | |
| CN115484464A (en) | Video coding method and device | |
| CN111669595A (en) | Screen content coding method, device, equipment and medium | |
| CN108391132A (en) | Word block coding method and device | |
| CN108924551A (en) | The prediction technique and relevant device of video image coding pattern | |
| CN110557621B (en) | Parameter acquisition method, pixel point pair selection method and related equipment | |
| CN103051896B (en) | Mode skipping-based video frequency coding method and mode skipping-based video frequency coding system | |
| Schaefer et al. | A hybrid color quantization algorithm incorporating a human visual perception model | |
| CN109640087A (en) | A kind of intra prediction mode decision method, device and equipment |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190521 |
|
| RJ01 | Rejection of invention patent application after publication |