CN107820084A - A kind of video-aware coding method and device - Google Patents
A kind of video-aware coding method and device Download PDFInfo
- Publication number
- CN107820084A CN107820084A CN201610822438.7A CN201610822438A CN107820084A CN 107820084 A CN107820084 A CN 107820084A CN 201610822438 A CN201610822438 A CN 201610822438A CN 107820084 A CN107820084 A CN 107820084A
- Authority
- CN
- China
- Prior art keywords
- coding
- layer
- video
- subjective quality
- parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/154—Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/146—Data rate or code amount at the encoder output
- H04N19/147—Data rate or code amount at the encoder output according to rate distortion criteria
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/189—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding
- H04N19/19—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding using optimisation based on Lagrange multipliers
-
- 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
- H04N19/30—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
A kind of video-aware coding method and device are announced herein, including：Obtain video information；Initiation parameter configures；Select level and its object currently to be encoded；The object is encoded, and the subjective quality model based on the level verifies the coding result of the coding, and the coding result is selected.It not only can strictly meet the subjective coding quality demand of user in each level, more excellent code efficiency can also be realized in each rank, and then realize that overall coding efficiency is more excellent, in the case where ensureing the subjective loss of subjective free of losses or given degree, bigger video compress, video code conversion efficiency are realized.
Description
Technical field
The application is related to computer realm, espespecially a kind of video-aware coding method and device.
Background technology
With the continuous development of digital video service in multimedia application, the demand of video cloud computing is improved constantly, it is former
The big data volume of beginning video source is that existing wired, wireless-transmission network bandwidth and storage resource are unaffordable.Thus,
Data compression of the vision signal in transmission, storing process turns into one of focus studied and applied at present.Video data compression
The purpose of (also referred to as Video coding) is to eliminate the various data redundancies of vision signal.
Video coding often brings certain coding distortion (mass loss).And distribution, compression lattice in such as video
In the environment such as the application of the transcoding such as formula conversion and HD video recording, the video observed by human eye is in subjective quality
(Subjective video quality) is unchanged and is frequently necessary to be satisfied, and then, is carried based on the perceptual coding of human vision model
Go out.Perceptual coding is to establish human vision model and carry out Video coding using the model, is also defined as based on minimum people
The encryption algorithm for the collimation error that eye can perceive.
For perceptual coding, how to optimize subjective quality model and the premise for being able to ensure in subjective quality using the model
Lower completion higher efficiency is collapsed into current study hotspot.Common perceptual coding method mainly uses subjective quality mould
Type substitutes the distortion evaluation method during current encoder.Specifically, by the distortion computation mistake in original mode selection processes
Journey replaces with subjective distortion computation, and original code check calculating process is constant, finally calculates one according to subjective distortion and code check
Individual new rate-distortion model replaces original rate-distortion model J=D+ λ * R, wherein, D is original distortion, and λ is rate distortion ginseng
Number, R is block of code bits number.There are the following problems for the above method：(1) most of instrument of existing Video coding is all to use model
J=D+ λ * R train what is come, and the situation that single code efficiency is decreased obviously can be produced by directly establishing new model；(2) simple one
The cataloged procedure of the modified block of secondary property still can not ensure the final subjective quality of video, it is necessary to establish overall Video coding frame
Structure, more complicated cataloged procedure is completed to ensure that the subjective quality of video meets user's request.
The content of the invention
In order to solve the above-mentioned technical problem, the invention provides a kind of video-aware approach and device.
The embodiments of the invention provide following scheme：
A kind of video-aware coding method, including：
Obtain the information of video；
Parameter configuration is carried out according to the information of the video, the parameter includes coding parameter and subjective quality parameter, institute
Stating subjective quality parameter includes subjective quality model and subjective quality condition；
The video is encoded according to the coding parameter step by step, and according to subjective quality model checking
The coding result of coding, it is met the coding result of the subjective quality condition.
Wherein, the information for obtaining video, including such as the next item down or two：
Directly read each hierarchical information of video；
Video specifying information is obtained by fast coding process.
Wherein, each hierarchical information or video specifying information of the information of the video or video include following one or more
Information：
Video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio.
Wherein, the carry out parameter configuration, including：The subjective quality parameter is configured, the subjective quality parameter is used for
Control subjective loss caused by cataloged procedure；The coding parameter of at least one level according to the information configuration of the video.
Wherein, when configuring the subjective quality parameter, if subjective quality model corresponding to the level of the video and institute
It is different to state model corresponding to subjective quality parameter, then the subjective quality parameter is converted into and meets the level subjective quality mould
The quality-controlling parameters of type.
Wherein, the coding parameter includes coding frame per second, speed class, quality scale, quantization parameter, code check, Yi Jiqi
He predicts, converted, quantifying and entropy code instrument.
Wherein, it is described that the video is encoded according to the coding parameter step by step, and according to the subjective quality mould
Type verifies the coding result of the coding, is met the coding result of the subjective quality condition, including：
It is determined that the level and its object currently to be encoded；
The object is encoded based on the coding parameter, described in the subjective quality model checking based on the level
The coding result of coding, it is met the coding result of the subjective quality condition.
Wherein, the level for determining currently to be encoded and its object, including：
If all objects of first layer have been met the coding result of the subjective quality condition, and the second layer is true
It is set to the level currently to be encoded, and the second object of the second layer is defined as the object currently to be encoded,
If the coding result that first layer has object is unsatisfactory for the subjective quality condition, the object is defined as currently
The object to be encoded, the first layer is defined as the level currently to be encoded；
If there is object to be not being met the coding result of the subjective quality condition in the first layer, by described
Two layers of level for being defined as currently being encoded, and the first object of the second layer is defined as the object currently to be encoded；
Wherein, the second layer is the last layer of the first layer, and the first layer corresponds to the institute of the second layer
The first object is stated, second object is next object of first object.
Wherein, it is described that the second layer is defined as the level currently to be encoded, and by the first object of the second layer
It is defined as after the object currently to be encoded, in addition to：
Adjust coding parameter and the first object to the second layer recompiles；Or
To the first object of the second layer, detect whether that coding result be present meets the subjective quality condition；Or
Third layer is returned, the third layer is defined as the level currently to be encoded, and institute will be corresponded in the third layer
The object for stating the second layer is defined as the object currently to be encoded, wherein, the third layer is the last layer of the second layer.
Wherein, the object is encoded, including：Using the one or more coding parameters specified to described right
As being encoded, the decoding and rebuilding data of the object are obtained.
Wherein, the executive mode encoded to the object is following any one：
A variety of coding parameters are run simultaneously, and obtain multigroup decoding and rebuilding data simultaneously；
Order performs, and terminates when there is a kind of coding result of coding parameter to meet the subjective quality condition, and obtain
The decoding and rebuilding data of the coding parameter.
Wherein, the subjective quality model based on the level verifies the coding result of the coding, including：
It is calculated to the decoding and rebuilding data of each object using the subjective quality model of level where the object
With the subjective quality difference between initial data.
Wherein, the coding result for being met subjective quality condition, including：
The quality-controlling parameters configured according to the result of the checking, during using the parameter configuration, are met subjectivity
The coding result of quality requirements.
Wherein, at least one level includes from low to high：Predicting unit, coding unit, band, image, video-frequency band,
Video.
Wherein, in the coding result of object in obtaining the predicting unit and coding unit, using following any one
Kind or several ways are met the coding result of the subjective quality condition：
Select the coding result of the more preferable predictive mode of subjective quality；
Selection meets the coding result of the predictive mode that number of coded bits is minimum under the conditions of subjective quality；
A determination pattern is made to reduce predicted distortion.
Wherein, when being encoded to the coding unit, band, image, video-frequency band or videl stage, corresponding coding ginseng
Number includes as follows any one or a few：Lagrangian, Rate Control, the volume used when quantization parameter, model selection
Code instrument.
A kind of device for video-aware coding, including：
Acquiring unit, for obtaining the information of video；
Dispensing unit, for carrying out parameter configuration according to the information of the video, the parameter includes coding parameter and master
Appearance quality parameter, the subjective quality parameter include subjective quality model and subjective quality condition；
Execution unit, for being encoded step by step according to the coding parameter to the video, and according to the subjective matter
The coding result that model verifies the coding is measured, is met the coding result of the subjective quality condition.
Wherein, the acquiring unit is used for the information for obtaining video, including such as the next item down or two：
Directly read each hierarchical information of video；
Video specifying information is obtained by fast coding process.
Wherein, each hierarchical information or video specifying information of the information of the video or video include following one or more
Information：
Video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio.
Wherein, the dispensing unit is used to carry out parameter configuration, including：
The subjective quality parameter is configured, the subjective quality parameter is used to control subjective loss caused by cataloged procedure；
The coding parameter of at least one level according to the information configuration of the video.
Wherein, the dispensing unit, it is additionally operable to when configuring the subjective quality parameter, if the level pair of the video
The subjective quality model answered is different from model corresponding to the subjective quality parameter, then the subjective quality parameter is converted into symbol
Close the quality-controlling parameters of the level subjective quality model.
Wherein, the coding parameter includes coding frame per second, speed class, quality scale, quantization parameter, code check, Yi Jiqi
He predicts, converted, quantifying and entropy code instrument.
Wherein, the execution unit, is specifically used for：
It is determined that the level and its object currently to be encoded；
The object is encoded based on the coding parameter, described in the subjective quality model checking based on the level
The coding result of coding, it is met the coding result of the subjective quality condition.
Wherein, the execution unit is used to determine level and its object currently to be encoded, including：
If all objects of first layer have been met the coding result of the subjective quality condition, and the second layer is true
It is set to the level currently to be encoded, and the second object of the second layer is defined as the object currently to be encoded,
If the coding result that first layer has object is unsatisfactory for the subjective quality condition, the object is defined as currently
The object to be encoded, the first layer is defined as the level currently to be encoded；
If there is object to be not being met the coding result of the subjective quality condition in the first layer, by described
Two layers of level for being defined as currently being encoded, and the first object of the second layer is defined as the object currently to be encoded；
Wherein, the second layer is the last layer of the first layer, and the first layer corresponds to the institute of the second layer
The first object is stated, second object is next object of first object.
Wherein, the execution unit is used to the second layer being defined as the level currently to be encoded, and by described second
First object of layer is defined as after the object currently to be encoded, in addition to：
Adjust coding parameter and the first object to the second layer recompiles；Or
To the first object of the second layer, detect whether that coding result be present meets the subjective quality condition；Or
Third layer is returned, the third layer is defined as the level currently to be encoded, and institute will be corresponded in the third layer
The object for stating the second layer is defined as the object currently to be encoded, wherein, the third layer is the last layer of the second layer.
Wherein, the execution unit is used to encode the object, including：Using described in the one or more specified
Coding parameter encodes to the object, obtains the decoding and rebuilding data of the object.
Wherein, the executive mode that the execution unit is used to encode the object is following any one：
A variety of coding parameters are run simultaneously, and obtain multigroup decoding and rebuilding data simultaneously；
Order performs, and terminates when there is a kind of coding result of coding parameter to meet the subjective quality condition, and obtain
The decoding and rebuilding data of the coding parameter.
Wherein, the execution unit is used for the coding knot that the subjective quality model based on the level verifies the coding
Fruit, including：
It is calculated to the decoding and rebuilding data of each object using the subjective quality model of level where the object
With the subjective quality difference between initial data.
Wherein, the execution unit is used for the coding result for being met subjective quality condition, including：
The quality-controlling parameters configured according to the result of the checking, during using the parameter configuration, are met subjectivity
The coding result of quality requirements.
Wherein, at least one level includes from low to high：Predicting unit, coding unit, band, image, video-frequency band,
Video.
Wherein, the execution unit is used in the coding result of object in obtaining the predicting unit and coding unit,
The coding result of the subjective quality condition is met using any one or a few following mode：
Select the coding result of the more preferable predictive mode of subjective quality；
Selection meets the coding result of the predictive mode that number of coded bits is minimum under the conditions of subjective quality；
A determination pattern is made to reduce predicted distortion.
Wherein, the execution unit is used to compile to the coding unit, band, image, video-frequency band or videl stage
During code, corresponding coding parameter includes as follows any one or a few：The Lagrange used when quantization parameter, model selection
Operator, Rate Control, coding tools.
Video-aware coding method provided in an embodiment of the present invention and device, make present coding unit and its predicting unit,
Multiple video levels such as band, image, image sets, video-frequency band, video establish subjective quality assessment system, and one-level wherein
Or the multistage comparison and selection completed between a variety of coding parameters, and then in the case where meeting that user gives subjective distortion parameter, realize
More excellent code efficiency, it is possible to achieve the perceptual coding of higher video compression efficiency, not only can strictly meet in each level
The subjective coding quality demand of user, more excellent code efficiency can also be realized in each rank, and then realize overall coding effect
Rate is more excellent, in the case where ensureing the subjective loss of subjective free of losses or given degree, realizes bigger video compress, video code conversion effect
Rate.
The method and device of the embodiment of the present invention, be particularly suitable for use in subjective low the loss even transcoding process of free of losses,
Realized from coding unit and its predicting unit to multiple levels of video on the basis of subjective quality is ensured more excellent
Code efficiency, and then ensure the subjective quality of video and realize that the code efficiency of whole video is more excellent.
Brief description of the drawings
Accompanying drawing is used for providing further understanding technical solution of the present invention, and a part for constitution instruction, with this
The embodiment of application is used to explain technical scheme together, does not form the limitation to technical solution of the present invention.
Fig. 1 is the flow chart of video-aware coding method of the embodiment of the present invention；
Fig. 2 is the example schematic diagram of video hierarchical structure of the embodiment of the present invention.
Fig. 3 is the flow chart of video-aware coding method of the embodiment of the present invention；
Fig. 4 is the specific implementation flow chart of video-aware coding method of the embodiment of the present invention；
Fig. 5 is the composition structural representation of video-aware code device of the embodiment of the present invention.
Embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
As shown in figure 1, the embodiment of the present invention provides a kind of video-aware coding method, mainly may include：
Step 101, the information of video is obtained；
Step 102, parameter configuration is carried out according to the information of the video, the parameter includes coding parameter and subjective quality
Parameter, the subjective quality parameter include subjective quality model and subjective quality condition；
Step 103, the video is encoded according to the coding parameter step by step, and according to the subjective quality model
The coding result of the coding is verified, is met the coding result of the subjective quality condition.
Wherein, the various information for being favorably improved video coding efficiency of video to be encoded can be obtained in step 101.
Wherein, encoded video is treated in step 103 and carries out hierarchical decomposition, and level is numbered.To each level
The selection of interlayer coding and control, level and its object that currently encoded can be determined according to current encoding state, then
The object is encoded, and the subjective quality model based on the level verifies the coding result of the coding, to described
Coding result is selected to find satisfactory coding parameter.
Can be that the existing object of current layer is encoded in step 103 in practical application；Current layer object is
Data after coding are calculated by subjective quality model, calculate the subjective loss of existing object coding, coding result is entered
Row checking；Coding result is selected according to above-mentioned the result.As a specific embodiment, coding result is tested
Card, the result that can be calculated according to subjective quality model, select to be adapted to using the quality-controlling parameters of level where existing object
Coding result.
The above method of the embodiment of the present invention, be particularly suitable for use in subjective low loss even coding, the transcoding process of free of losses,
Using configurable subjective quality parameter as input, the coding result for meeting the subjective quality condition is found.
In some implementations, above-mentioned level can include following one or several video levels from low to high：
Predicting unit, coding unit, band, image, video-frequency band, video.Video hierarchical structure schematic diagram is as shown in Figure 2.
Wherein, the information of video is obtained in step 101, such as the next item down or two can be included：
1) each hierarchical information of video is directly read；
2) video specifying information is obtained by fast coding process.
In practical application, the information of the video or each hierarchical information of video or video specifying information can include as follows
One or more information：Video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio.
Wherein, parameter configuration is carried out in step 102, can be included：Configure the subjective quality parameter, the subjective quality
Parameter is used to control subjective loss caused by cataloged procedure；The volume of at least one level according to the information configuration of the video
Code parameter.
In some implementations, when configuring the subjective quality parameter, if subjective corresponding to the level of the video
Quality model is different from model corresponding to the subjective quality parameter, then is converted into the subjective quality parameter and meets the layer
The quality-controlling parameters of level subjective quality model.
In some implementations, the coding parameter can include coding frame per second, speed class, quality scale, quantization
Parameter, code check and other predictions, conversion, quantization and entropy code instrument.
In practical application, subjective quality parameter may include：Parameter and expression corresponding to the subjective quality model of each level
Parameter of subjective quality condition etc..When configuring subjective quality parameter, if the subjective quality model of a certain level is matched somebody with somebody with initialization
Model corresponding to subjective quality parameter when putting is different, then is converted into the subjective quality parameter of the level and meets current level master
The quality-controlling parameters of appearance quality model.
Wherein, defined in subjective quality model subjective quality assessment specific algorithm and subjective quality condition, pass through master
The calculating of appearance quality module coding result is verified to obtain subjective quality validation value (i.e. hereinafter decoding and rebuilding data with original
Mean square deviation between beginning data etc.), can be by the subjective quality validation value and the quality control of configuration when selecting coding result
Parameter compares, and the coding parameter (i.e. optimal coding parameter) of subjective quality loss reduction is selected according to comparative result；For pre-
Survey for unit, pattern or foundation that predictive mode encodes as it.Each level has its corresponding subjective quality model,
If need to handle multiple levels during perceptual coding, then need to establish corresponding to it this multiple level respectively
Subjective quality model.If a certain layer need to only be handled, then only need to establish the subjective quality model of the level.
Wherein it is possible to configure more set coding parameters, and in configuration codes parameter, include configuration codes frame per second, speed threshold
Secondary, quality scale, quantization parameter, video code rate and other predictions, conversion, quantization and entropy code instrument.Speed class refers to
Desired coding rate.Quality scale then refers to that encoder lays particular stress on the coding quality for optimizing what type of video in coding,
Such as animation, film or face.
Wherein, the video is encoded according to the coding parameter step by step in step 103, and according to the subjective matter
The coding result that model verifies the coding is measured, the coding result of the subjective quality condition is met, can include：It is determined that
The level and its object currently to be encoded；The object is encoded based on the coding parameter, the master based on the level
Appearance quality model verifies the coding result of the coding, is met the coding result of the subjective quality condition.
Wherein, the level for determining currently to be encoded and its object, can include：
If all objects of first layer have been met the coding result of the subjective quality condition, and the second layer is true
It is set to the level currently to be encoded, and the second object of the second layer is defined as the object currently to be encoded,
If the coding result that first layer has object is unsatisfactory for the subjective quality condition, the object is defined as currently
The object to be encoded, the first layer is defined as the level currently to be encoded；
If there is object to be not being met the coding result of the subjective quality condition in the first layer, by described
Two layers of level for being defined as currently being encoded, and the first object of the second layer is defined as the object currently to be encoded；
Wherein, the second layer is the last layer of the first layer, and the first layer corresponds to the institute of the second layer
The first object is stated, second object is next object of first object.
For example, the first layer can be the slice layer in Fig. 2, the first object represents the coding unit in this layer, second
Layer can be image layer, and the second layer is the band in this layer, and third layer can be video-frequency band layer, and the 3rd object can be video-frequency band
Image in layer.
Wherein, it is described that the second layer is defined as the level currently to be encoded, and by the first object of the second layer
It is defined as after the object currently to be encoded, in addition to：
Adjust coding parameter and the first object to the second layer recompiles；Or
To the first object of the second layer, detect whether that coding result be present meets the subjective quality condition；Or
Third layer is returned, the third layer is defined as the level currently to be encoded, and institute will be corresponded in the third layer
The object for stating the second layer is defined as the object currently to be encoded, wherein, the third layer is the last layer of the second layer.
Wherein, the object is encoded, can included：Using the one or more coding parameters specified to institute
State object to be encoded, obtain the decoding and rebuilding data of the object.
In practical application, the executive mode encoded to the object is following any one：
1) while a variety of coding parameters are run, and obtains multigroup decoding and rebuilding data simultaneously；
2) order performs, and terminates when there is a kind of coding result of coding parameter to meet the subjective quality condition, and
To the decoding and rebuilding data of the coding parameter.
Wherein, the subjective quality model based on the level verifies the coding result of the coding, can include：Using institute
The subjective quality model of level calculates it between initial data to the decoding and rebuilding data of each object where stating object
Subjective quality difference.Here, the coding result for being met subjective quality condition, can include：According to the checking
Result, the quality-controlling parameters configured during using the parameter configuration, be met the coding result of subjective quality condition.
In some implementations, in the coding result of object in obtaining the predicting unit and coding unit, use
Any one or a few following mode is met the coding result of the subjective quality condition：
1) coding result of the more preferable predictive mode of subjective quality is selected；
2) selection meets the coding result of the predictive mode that number of coded bits is minimum under the conditions of subjective quality；
3) a determination pattern is made to reduce predicted distortion.
In practical application, when being encoded to the coding unit, band, image, video-frequency band or videl stage, accordingly
Coding parameter can include as follows any one or a few：1) quantization parameter；2) the Lagrange calculation used during model selection
Son；3) Rate Control；4) various coding toolses (such as prediction, conversion, quantifying entropy code mode).
As shown in Fig. 2 be a kind of video hierarchical structure schematic diagram, comprising from low to high：Predicting unit, coding unit,
Band, image, video-frequency band, video.In Fig. 2, video layer is the 0th layer, and video-frequency band layer is the 1st layer, by that analogy, coding unit layer
For the 4th layer.Object in video layer is video, i.e., video 0, be 1 ..., video 5.Each object in video layer is corresponding
One video-frequency band layer, the object of video-frequency band layer corresponding to video 3 include video-frequency band 1, video-frequency band 2 ..., video-frequency band N, video-frequency band
The corresponding image layer of each object (i.e. each video-frequency band) in layer, image layer corresponding to video-frequency band 4 including image 1, image 2,
Image 3 ..., image 6；Each object in each image layer is the corresponding slice layer of each image, as shown in Fig. 2 figure
Slice layer includes as corresponding to 2：Band 1, band 2, band 3 ..., band 7.As shown in Fig. 2 each object is in slice layer
Each corresponding coding unit layer of band, the object of coding unit layer is coding unit, and each coding unit layer may include
Multiple coding units, the corresponding predicting unit layer of each coding unit, the object of predicting unit layer is predicting unit, each pre-
Surveying elementary layer may include multiple predicting units.
As shown in figure 3, the Video coding cognitive method of the embodiment of the present invention may include following steps：
Step 301, the information of video is obtained；
Step 302, parameter configuration is carried out according to the information of the video, the parameter includes coding parameter and subjective quality
Parameter, the subjective quality parameter include subjective quality model and subjective quality condition；
Step 303, it is determined that the level and its object currently to be encoded；
Step 304, the object is encoded based on the coding parameter, the subjective quality model based on the level
The coding result of the coding is verified, is met the coding result of the subjective quality condition；
Wherein, the level that above-mentioned perceptual coding method is related to can be one or multiple, in the embodiment of the present invention
The level being related to includes six layers, is respectively from low to high：Predicting unit, coding unit, band, image, video-frequency band, video.
Wherein, the specific acquisition modes that the information of video can use can be：1) the video tool of each level is directly read
Body information；And/or 2) video specifying information etc. obtained by fast coding process.Here, the video information specifically obtained can be with
Including video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio etc..
Wherein, in the step of step 302 parameter configuration, the parameters configured are mainly needed in cataloged procedure.Specifically
It may include following steps：1) firstly, it is necessary to be inputted according to user or configure subjective quality parameter, the subjective matter according to video information
Amount parameter is used to control subjective loss caused by cataloged procedure；2) video information secondly, obtained according to step 301 is to each level
Coding parameter carries out initial configuration, and these coding parameters can include coding frame per second, speed class, quality scale, quantization ginseng
Number, video code rate, coding tools etc..Here, in step 1), if the subjective quality model and initial configuration of a certain level
Subjective quality parameter corresponding to model it is different, then the subjective quality parameter is also converted into the matter for meeting the hierarchal model
Measure control parameter.In addition, each level coding parameter of configuration has comprised at least a kind of coding parameter, each level configuration of preferred pair
A variety of coding parameters, subsequently therefrom to select the optimal coding parameter of quality.
Wherein, in the step of level and its object currently to be encoded are determined in step 303, selection from high to low currently will
The level of coding, specifically：If when first layer finds the coding result for meeting subjective quality condition of this layer of all object,
The second layer is returned to continue to encode next second object.If first layer is obtained to some first object using a certain coding parameter
To coding result do not meet subjective quality condition, can continue to attempt other coding parameters to first object, and judge it
Whether coding result meets subjective quality condition, and the coding result of subjective quality condition is met until finding first object.When
Any one or several objects of first layer are present when can not find the coding result for meeting subjective quality condition, return to the second layer,
The second object that the first layer is corresponded in the second layer is recompiled.Wherein, it is also an option that performing following steps after return：
1) to corresponded in next second object or the second layer of the second layer first layer the second Object Selection other coding join
Number is encoded and judges whether the coding result of the coding parameter meets subjective quality condition；Or 2) detect other simultaneously
Whether the coding result of the coding parameter of operation meets subjective quality condition；If 3) first corresponding to the existing object of the second layer
The coding result for meeting subjective quality condition is still not present in any one in layer or several first objects, then returns to the 3rd
Layer, the third layer are the last layer of the second layer, the corresponding second layer of each 3rd object in third layer, the second layer first
The last layer of layer, the corresponding first layer of each second object in the second layer, multiple first objects can be included in first layer.
Wherein, when being encoded to the object, the one or more coding parameters specified can be used right to this respectively
The data encoding of elephant, obtain the decoding and rebuilding data of the object data.Wherein, the executive mode of these coding parameters includes：1)
Mutually the data without dependence are run simultaneously, and obtain multigroup reconstructed results (i.e. decoding and rebuilding data) simultaneously；2) order performs, after
Whether continuous object performs whether coding meets subjective quality encoding condition dependent on the coding result of previous object.
Wherein, the subjective quality model based on the level verifies the coding result of the coding, is met the master
The coding result of appearance quality condition, can be specifically：Using the subjective quality model of level where existing object to each coding
Its decoding and rebuilding data of calculation and object are with the subjective quality difference between initial data.Subjective quality difference is bigger, illustrates subjectivity
Loss is bigger, and subjective quality difference is smaller, illustrates that subjective loss is smaller.
Wherein, optimal coding knot is selected using the quality-controlling parameters of level where existing object according to result of calculation
Fruit, and then determine optimal coding parameter.Wherein, selected in predicting unit and the specific coding result of the object of coding unit
When, use quality control parameter intervention mode selection course, comprising：1) select the smaller i.e. subjective quality of subjective quality loss more preferable
Predictive mode；Or, 2) selection meets code efficiency highest predictive mode under the conditions of subjective quality；Or, 3) make a determination mould
Formula reduces predicted distortion.Selected in the specific coding result of coding unit, band, image, video-frequency band or the other object of videl stage
When, specific method, which includes, attempts various coding parameters to select the coding parameter for meeting subjective quality condition, described coding ginseng
Number includes as follows any one or a few：The Lagrangian used when quantization parameter, model selection, Rate Control, its
His various coding toolses (such as various predictions, conversion, quantifying, entropy code instrument), above-mentioned predictive mode is predicting unit coding
One kind of parameter.
Highly preferred embodiment of the present invention as detailed below.After reading the following description and drawings, being familiar with this those skilled in the art can be light
Readily understood spirit of the invention.
Some vocabulary has been used among present specification and claim to censure specific component.In art
Has usually intellectual, it is to be appreciated that hardware manufacturer may call same component with different nouns.The application is said
Bright book and claim are come not in a manner of the difference of title is used as and distinguishes component with the difference of component functionally
Criterion as differentiation."comprising" or " comprising " mentioned in working as in specification in the whole text and claims are open term,
Therefore " include but be not limited to " should be construed to.In addition, " coupling " one word includes any direct and is indirectly electrically connected hand herein
Section.Therefore, if the first device of described in the text one is coupled to a second device, representing the first device can directly be electrically connected in
The second device, or be electrically connected indirectly to the second device through other devices or connection means.
In an embodiment as shown in Figure 4, the coding level chosen includes five layers, is predicting unit, coding respectively
Unit, image, image sets, video.
In embodiment as shown in Figure 4, the specific implementation process of video-aware coding may include：
Step S401, obtain video information；
Specifically, the video information obtained includes：Video code rate, frame per second, video compression format, quantization parameter, key frame
Distribution, image resolution ratio, time span, true code check and key frame number.Wherein, the time span of video, true code check with
And key frame number can be obtained by fast coding process.Other information can be by directly reading the video information of each level
Obtain.
Step S402, carry out following parameter configuration：Set maximum quantization parameter (qpmax) to be no more than 30, use x264
The veryslow class of encoder, subjective quality parameter value are configured to 9.5, using mean square deviation as subjective quality model it is main according to
According to.
Step S403, the selection of interlayer coding and control：When what K+1 layers (current layer) found all objects meets subjective matter
During the coding result of amount condition, return to K layers and continue to encode next object, continue step S404；Looked for not when K+1 layers are present
During to the coding result for meeting subjective quality condition, return to K layers (last layer of current layer) and the existing object to K layers selects
Select other coding parameters to be encoded, continue step S404.These coding parameters can include modification code check, secondary coding is matched somebody with somebody
Put, qpmax, I frame period.
For example, for the hierarchical structure shown in Fig. 2, video layer is the 0th layer, and video-frequency band layer is the 1st layer, by that analogy, is compiled
Code elementary layer is the 5th layer.Object in video layer is video, video 0 as shown in Figure 2, be 1 ..., video 5.In video layer
Each object correspond to a video-frequency band layer, as shown in Fig. 2 the object of video-frequency band layer corresponding to video 3 include video-frequency band 1,
Video-frequency band 2 ..., video-frequency band N, each object is the corresponding image layer of each video-frequency band in video-frequency band layer, as shown in Fig. 2
Image layer corresponding to video-frequency band 4 include image 1, image 2, image 3 ..., image 6；Each each object in image layer is
Each corresponding slice layer of image, as shown in Fig. 2 slice layer corresponding to image 2 includes：Band 1, band 2, band
3rd ..., band 7.As shown in Fig. 2 each object is that each band corresponds to a coding unit layer, coding unit in slice layer
The object of layer is coding unit, and each coding unit layer may include multiple coding units, and corresponding one of each coding unit is in advance
Elementary layer is surveyed, the object of predicting unit layer is predicting unit, and each predicting unit layer may include multiple predicting units.
If necessary to all be encoded to the 0th layer to the 4th layer of the hierarchical structure shown in Fig. 2, then step S403 is extremely
The interlayer iterator mechanism that S406 is formed is as follows：As shown in Fig. 2 video 3 is the 0th layer of the 3rd object, encoder can be from the 0th layer
The coding unit layer positioned at the band 6 of 4 image of video-frequency band 2 is gone straight down to, if (coding is single for all objects of coding unit layer
Member) coding result all meet subjective quality condition, then the last layer i.e. slice layer for returning to coding unit layer continues under coding
One object is band 7；Otherwise, then the last layer slice layer of coding unit layer is returned, is recompiled using other coding parameters
Band 6, until band 6 meets subjective quality condition, if the coding result of the various coding parameters of band 6 can not all meet to lead
Appearance quality condition, then the last layer for returning to slice layer is that image layer attempts other Image Coding parameters and re-starts coding, with this
Analogize, until choosing one group of video layer parameter, video-frequency band layer parameter, image layer parameter, band layer parameter to cause band 6 to meet master
Appearance quality condition.If all objects of current slice layer, which are the coding result of all bands of image 2, all meets subjective quality
Condition, then the last layer for returning to slice layer is that image layer continues to encode next image i.e. image 3.If the institute of present image layer
There are object i.e. all images of video-frequency band 4 to be satisfied by subjective quality condition, then the last layer for returning to image layer be video-frequency band layer after
The next video-frequency band of code of sequeling i.e. video-frequency band 5.
Step S404, a variety of coding parameters are sequentially performed respectively to the data encoding of existing object, latter coding parameter
Whether the coding result dependent on former coding parameter meets subjective quality condition；
If specifically, in a variety of coding parameters, first coding parameter meets subjective quality condition, then order performs
Second,
Step S405, using the subjective quality model of level where existing object to the decoding and rebuilding data meter of each object
It is calculated with the mean square deviation between initial data；
Step S406, selected for predicting unit and coding unit and meet most efficient predictive mode under the conditions of subjective quality
Coding result；It is full to select that various coding parameters are attempted for coding unit, band, image, video-frequency band or the other object of videl stage
The coding result of sufficient subjective quality condition.
Step S407, judge the whether all encoded completion of all objects of current layer, if it is, return to step S403, such as
Fruit is no, then return to step S405.
In this way, until all objects for all levels that need to be encoded all encode completion, and be video layer to highest level
Under each object have selected one group of video layer parameter, video-frequency band layer parameter, image layer parameter, band layer parameter and cause its volume
Code result meets subjective quality condition, then, this flow terminates.
As shown in figure 5, the application also provides a kind of device for video-aware coding, including：
Acquiring unit 51, for obtaining the information of video；
Dispensing unit 52, for according to the information of the video carry out parameter configuration, the parameter include coding parameter and
Subjective quality parameter, the subjective quality parameter include subjective quality model and subjective quality condition；
Execution unit 53, for being encoded step by step according to the coding parameter to the video, and according to the subjectivity
Quality model verifies the coding result of the coding, is met the coding result of the subjective quality condition.
Wherein, the acquiring unit 51 is used for the information for obtaining video, can include such as the next item down or two：
Directly read each hierarchical information of video；
Video specifying information is obtained by fast coding process.
Wherein, each hierarchical information or video specifying information of the information of the video or video include following one or more
Information：Video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio.
Wherein, the dispensing unit 52 is used to carry out parameter configuration, can include：Configure the subjective quality parameter, institute
State subjective quality parameter and be used to control subjective loss caused by cataloged procedure；At least one according to the information configuration of the video
The coding parameter of individual level.The coding parameter include coding frame per second, speed class, quality scale, quantization parameter, code check, with
And other predictions, conversion, quantization and entropy code instrument.
Wherein, the dispensing unit 52, it may also be used for when configuring the subjective quality parameter, if the layer of the video
Subjective quality model corresponding to level is different from model corresponding to the subjective quality parameter, then converts the subjective quality parameter
To meet the quality-controlling parameters of the level subjective quality model.
Wherein, the execution unit 53, is particularly used in：It is determined that the level and its object currently to be encoded；Based on described
Coding parameter encodes to the object, and the subjective quality model based on the level verifies the coding result of the coding,
It is met the coding result of the subjective quality condition.
Specifically, the execution unit 53 can be used for determining level and its object currently to be encoded, including：
If all objects of first layer have been met the coding result of the subjective quality condition, and the second layer is true
It is set to the level currently to be encoded, and the second object of the second layer is defined as the object currently to be encoded,
If the coding result that first layer has object is unsatisfactory for the subjective quality condition, the object is defined as currently
The object to be encoded, the first layer is defined as the level currently to be encoded；
If there is object to be not being met the coding result of the subjective quality condition in the first layer, by described
Two layers of level for being defined as currently being encoded, and the first object of the second layer is defined as the object currently to be encoded；
Wherein, the second layer is the last layer of the first layer, and the first layer corresponds to the institute of the second layer
The first object is stated, second object is next object of first object.
Wherein, the execution unit 53 is used to the second layer being defined as the level currently to be encoded, and by described the
Two layers of the first object is defined as after the object currently to be encoded, in addition to：
Adjust coding parameter and the first object to the second layer recompiles；Or
To the first object of the second layer, detect whether that coding result be present meets the subjective quality condition；Or
Third layer is returned, the third layer is defined as the level currently to be encoded, and institute will be corresponded in the third layer
The object for stating the second layer is defined as the object currently to be encoded, wherein, the third layer is the last layer of the second layer.
Wherein, the execution unit 53 is used to encode the object, including：Use the one or more institutes specified
State coding parameter to encode the object, obtain the decoding and rebuilding data of the object.Here, the execution unit 53 is used
In the executive mode encoded to the object be following any one：
A variety of coding parameters are run simultaneously, and obtain multigroup decoding and rebuilding data simultaneously；
Order performs, and terminates when there is a kind of coding result of coding parameter to meet the subjective quality condition, and obtain
The decoding and rebuilding data of the coding parameter.
Wherein, the execution unit 53 is used for the coding knot that the subjective quality model based on the level verifies the coding
Fruit, including：It is calculated to the decoding and rebuilding data of each object using the subjective quality model of level where the object
With the subjective quality difference between initial data.
Wherein, the execution unit 53 can be used for the coding result for being met subjective quality condition, including：
The quality-controlling parameters configured according to the result of the checking, during using the parameter configuration, are met subjectivity
The coding result of quality requirements.
In practical application, at least one level includes from low to high：Predicting unit, coding unit, band, image,
Video-frequency band, video.
In practical application, the execution unit 53 is used for the coding of the object in the predicting unit and coding unit is obtained
When as a result, the coding result of the subjective quality condition is met using any one or a few following mode：
Select the coding result of the more preferable predictive mode of subjective quality；
Selection meets the coding result of the predictive mode that number of coded bits is minimum under the conditions of subjective quality；
A determination pattern is made to reduce predicted distortion.
In practical application, the execution unit 53 is used for the coding unit, band, image, video-frequency band or video
When level is encoded, corresponding coding parameter includes as follows any one or a few：Used when quantization parameter, model selection
Lagrangian, Rate Control, coding tools.
The above-mentioned video-aware coding method of the embodiment of the present invention and device, the subjectivity for the video that needs encode can be controlled
Mass loss degree, code efficiency can be obviously improved compared to correlation technique under the loss of identical subjective quality, and may be used also
Subjective approximate lossless code efficiency is obviously improved, the video for encoding generation meets existing video encoding standard, can direct quilt
Conventional player decoding.
The above-mentioned video-aware coding method of the embodiment of the present invention and device, the code efficiency lifting example that encoding efficiency is brought
Such as, approximate lossless coding efficiency corresponding to the subjective quality parameter of threshold value 9.5 is as shown in table 1 below, averagely obtains 31.1% pressure
Contracting efficiency, video time and the ratio of transcoding time are 1.05.Thus it will be evident that the embodiment of the present invention can significantly improve
Code efficiency of the general video under given subjective quality parameter.
Table 1
Above-mentioned flow chart meant for illustration is according to the demonstration example of adaptive reference image selection technique of the present invention, art technology
Personnel can change each step, rearrange above-mentioned steps under the premise of without departing from the spirit of the present invention, split certain step,
Or some steps of combination, to realize the present invention.
Above description can allow those skilled in the art according to the Content Implementation of the application-specific and its needs present invention.Institute
The various modifications of embodiment are stated it will become apparent to those skilled in the art that and can will be defined above substantially former
Then it is applied to other embodiment.Therefore, the present invention is not limited to described specific embodiment, but meet with the principle that discloses and
The consistent widest range of novel feature.In above-mentioned detailed description, in order to provide thorough understanding of the present invention, describe each
Kind specific detail.In addition, it will be appreciated by those skilled in the art that the present invention is enforceable.
The above-mentioned embodiment of the present invention can be implemented in various hardware, Software Coding or both combination.For example, this hair
Bright embodiment can be to be integrated into the circuit of video compress chip or be integrated into video compression software to perform the program of said process
Code.Embodiments of the invention are alternatively to be performed in digital signal processor (Digital Signal Processor, DSP)
The above-mentioned processing of execution program code.The present invention can also refer to computer processor, DSP, microprocessor or scene can program
Design the multiple functions that gate array (Field Programmable Gate Array, FPGA) performs.It can be configured according to the present invention
Above-mentioned computing device particular task, it defines the machine-readable software code for the ad hoc approach that the present invention discloses by performing
Or firmware code is completed.Software code or firmware code can be developed into different program languages and different forms or shape
Formula, different target platforms has been alternatively it and has compiled solution software code.However, according to the software code of execution task of the present invention and other
Different code pattern, type and the language of type configuration code do not depart from spirit and scope of the invention.
Although the present invention is disclosed above with regard to preferred embodiment and its advantage, in the technical field of the invention ordinary skill
Personnel, without departing from the spirit and scope of the present invention, when can be used for a variety of modifications and variations.Protection scope of the present invention, which is worked as, to be regarded
Claims before, which define, to be defined.
Claims (32)
- A kind of 1. video-aware coding method, it is characterised in that including：Obtain the information of video；Parameter configuration is carried out according to the information of the video, the parameter includes coding parameter and subjective quality parameter, the master Appearance quality parameter includes subjective quality model and subjective quality condition；The video is encoded according to the coding parameter step by step, and the coding is verified according to the subjective quality model Coding result, be met the coding result of the subjective quality condition.
- 2. according to the method for claim 1, it is characterised in that the information for obtaining video, including such as the next item down or two ：Directly read each hierarchical information of video；Video specifying information is obtained by fast coding process.
- 3. method according to claim 1 or 2, it is characterised in that the information of the video or each hierarchical information of video Or video specifying information includes following one or more information：Video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio.
- 4. according to the method for claim 1, it is characterised in that the carry out parameter configuration, including：The subjective quality parameter is configured, the subjective quality parameter is used to control subjective loss caused by cataloged procedure；The coding parameter of at least one level according to the information configuration of the video.
- 5. according to the method for claim 4, it is characterised in thatWhen configuring the subjective quality parameter, if subjective quality model corresponding to the level of the video and the subjective quality Model corresponding to parameter is different, then the subjective quality parameter is converted into the quality control for meeting the level subjective quality model Parameter processed.
- 6. according to the method for claim 4, it is characterised in thatThe coding parameter includes coding frame per second, speed class, quality scale, quantization parameter, code check and other predictions, become Change, quantify and entropy code instrument.
- 7. according to the method for claim 1, it is characterised in that described that the video is entered according to the coding parameter step by step Row coding, and according to the coding result of the subjective quality model checking coding, it is met the subjective quality condition Coding result, including：It is determined that the level and its object currently to be encoded；The object is encoded based on the coding parameter, the subjective quality model based on the level verifies the coding Coding result, be met the coding result of the subjective quality condition.
- 8. according to the method for claim 7, it is characterised in that the level for determining currently to be encoded and its object, bag Include：If all objects of first layer have been met the coding result of the subjective quality condition, the second layer is defined as The level currently to be encoded, and the second object of the second layer is defined as the object currently to be encoded,If the coding result that first layer has object is unsatisfactory for the subjective quality condition, the object is defined as currently compiling The object of code, is defined as the level currently to be encoded by the first layer；If there is object to be not being met the coding result of the subjective quality condition in the first layer, by the second layer It is defined as the level currently to be encoded, and the first object of the second layer is defined as the object currently to be encoded；Wherein, the second layer is the last layer of the first layer, and the first layer corresponds to described the of the second layer One object, second object are next object of first object.
- 9. according to the method for claim 7, it is characterised in that described that the second layer is defined as the layer currently to be encoded Level, and after the first object of the second layer is defined as into the object currently to be encoded, in addition to：Adjust coding parameter and the first object to the second layer recompiles；OrTo the first object of the second layer, detect whether that coding result be present meets the subjective quality condition；OrThird layer is returned, the third layer is defined as the level currently to be encoded, and described the will be corresponded in the third layer Two layers of object is defined as the object currently to be encoded, wherein, the third layer is the last layer of the second layer.
- 10. according to the method for claim 7, it is characterised in that the object is encoded, including：Use what is specified One or more coding parameters encode to the object, obtain the decoding and rebuilding data of the object.
- 11. according to the method for claim 10, it is characterised in that the executive mode that is encoded to the object is Following any one：A variety of coding parameters are run simultaneously, and obtain multigroup decoding and rebuilding data simultaneously；Order performs, and terminates when there is a kind of coding result of coding parameter to meet the subjective quality condition, and obtain the volume The decoding and rebuilding data of code parameter.
- 12. the method according to claim 10 or 11, it is characterised in that the subjective quality model checking based on the level The coding result of the coding, including：It is calculated with original to the decoding and rebuilding data of each object using the subjective quality model of level where the object Subjective quality difference between beginning data.
- 13. according to the method for claim 12, it is characterised in that the coding knot for being met subjective quality condition Fruit, including：The quality-controlling parameters configured according to the result of the checking, during using the parameter configuration, are met subjective quality The coding result of condition.
- 14. according to the method described in any one of claim 1 to 13, it is characterised in that at least one level is from low to high Including：Predicting unit, coding unit, band, image, video-frequency band, video.
- 15. according to the method for claim 14, it is characterised in that the object in the predicting unit and coding unit is obtained Coding result when, the coding result of the subjective quality condition is met using any one or a few following mode：Select the coding result of the more preferable predictive mode of subjective quality；Selection meets the coding result of the predictive mode that number of coded bits is minimum under the conditions of subjective quality；A determination pattern is made to reduce predicted distortion.
- 16. according to the method for claim 14, it is characterised in thatWhen being encoded to the coding unit, band, image, video-frequency band or videl stage, corresponding coding parameter includes such as Under any one or a few：The Lagrangian used when quantization parameter, model selection, Rate Control, coding tools.
- A kind of 17. device for video-aware coding, it is characterised in that including：Acquiring unit, for obtaining the information of video；Dispensing unit, for carrying out parameter configuration according to the information of the video, the parameter includes coding parameter and subjective matter Parameter is measured, the subjective quality parameter includes subjective quality model and subjective quality condition；Execution unit, for being encoded step by step according to the coding parameter to the video, and according to the subjective quality mould Type verifies the coding result of the coding, is met the coding result of the subjective quality condition.
- 18. device according to claim 17, it is characterised in that the acquiring unit is used for the information for obtaining video, bag Include such as the next item down or two：Directly read each hierarchical information of video；Video specifying information is obtained by fast coding process.
- 19. the device according to claim 17 or 18, it is characterised in that the information of the video or each level letter of video Breath or video specifying information include following one or more information：Video code rate, frame per second, video compression format, quantization parameter, key frame distribution, image resolution ratio.
- 20. device according to claim 17, it is characterised in that the dispensing unit is used to carry out parameter configuration, including：The subjective quality parameter is configured, the subjective quality parameter is used to control subjective loss caused by cataloged procedure；The coding parameter of at least one level according to the information configuration of the video.
- 21. device according to claim 20, it is characterised in that the dispensing unit, be additionally operable to configuring the subjectivity During mass parameter, if the model corresponding with the subjective quality parameter of subjective quality model corresponding to the level of the video is not Together, then the subjective quality parameter is converted into the quality-controlling parameters for meeting the level subjective quality model.
- 22. device according to claim 20, it is characterised in thatThe coding parameter includes coding frame per second, speed class, quality scale, quantization parameter, code check and other predictions, become Change, quantify and entropy code instrument.
- 23. device according to claim 17, it is characterised in that the execution unit, be specifically used for：It is determined that the level and its object currently to be encoded；The object is encoded based on the coding parameter, the subjective quality model based on the level verifies the coding Coding result, be met the coding result of the subjective quality condition.
- 24. device according to claim 7, it is characterised in that the execution unit is used to determine the layer currently to be encoded Level and its object, including：If all objects of first layer have been met the coding result of the subjective quality condition, the second layer is defined as The level currently to be encoded, and the second object of the second layer is defined as the object currently to be encoded,If the coding result that first layer has object is unsatisfactory for the subjective quality condition, the object is defined as currently compiling The object of code, is defined as the level currently to be encoded by the first layer；If there is object to be not being met the coding result of the subjective quality condition in the first layer, by the second layer It is defined as the level currently to be encoded, and the first object of the second layer is defined as the object currently to be encoded；Wherein, the second layer is the last layer of the first layer, and the first layer corresponds to described the of the second layer One object, second object are next object of first object.
- 25. device according to claim 24, it is characterised in that the execution unit is used to the second layer being defined as The level currently to be encoded, and after the first object of the second layer is defined as into the object currently to be encoded, in addition to：Adjust coding parameter and the first object to the second layer recompiles；OrTo the first object of the second layer, detect whether that coding result be present meets the subjective quality condition；OrThird layer is returned, the third layer is defined as the level currently to be encoded, and described the will be corresponded in the third layer Two layers of object is defined as the object currently to be encoded, wherein, the third layer is the last layer of the second layer.
- 26. device according to claim 24, it is characterised in that the execution unit is used to compile the object Code, including：The object is encoded using the one or more coding parameters specified, obtains the decoding of the object Rebuild data.
- 27. device according to claim 26, it is characterised in that the execution unit is used to encode the object Executive mode be following any one：A variety of coding parameters are run simultaneously, and obtain multigroup decoding and rebuilding data simultaneously；Order performs, and terminates when there is a kind of coding result of coding parameter to meet the subjective quality condition, and obtain the volume The decoding and rebuilding data of code parameter.
- 28. device according to claim 26, it is characterised in that the execution unit is used for the subjectivity based on the level Quality model verifies the coding result of the coding, including：It is calculated with original to the decoding and rebuilding data of each object using the subjective quality model of level where the object Subjective quality difference between beginning data.
- 29. device according to claim 28, it is characterised in that the execution unit is used to be met subjective quality bar The coding result of part, including：The quality-controlling parameters configured according to the result of the checking, during using the parameter configuration, are met subjective quality The coding result of condition.
- 30. according to the device described in any one of claim 17 to 29, it is characterised in that at least one level is from low to high Including：Predicting unit, coding unit, band, image, video-frequency band, video.
- 31. device according to claim 30, it is characterised in that the execution unit is used to obtain the predicting unit With during the coding result of object, the subjective quality is met using any one or a few following mode in coding unit The coding result of condition：Select the coding result of the more preferable predictive mode of subjective quality；Selection meets the coding result of the predictive mode that number of coded bits is minimum under the conditions of subjective quality；A determination pattern is made to reduce predicted distortion.
- 32. device according to claim 30, it is characterised in thatThe execution unit is used for when being encoded to the coding unit, band, image, video-frequency band or videl stage, accordingly Coding parameter include following any one or a few：The Lagrangian used when quantization parameter, model selection, code check Control, coding tools.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610822438.7A CN107820084B (en) | 2016-09-13 | 2016-09-13 | Video perception coding method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610822438.7A CN107820084B (en) | 2016-09-13 | 2016-09-13 | Video perception coding method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107820084A true CN107820084A (en) | 2018-03-20 |
| CN107820084B CN107820084B (en) | 2020-02-07 |
Family
ID=61601408
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610822438.7A Active CN107820084B (en) | 2016-09-13 | 2016-09-13 | Video perception coding method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107820084B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109788316A (en) * | 2019-02-27 | 2019-05-21 | 腾讯科技（深圳）有限公司 | Code rate control, video transcoding method and device, computer equipment and storage medium |
| CN110139101A (en) * | 2019-06-20 | 2019-08-16 | 电子科技大学 | A kind of frame level bit distribution method based on the control of the domain λ code rate |
| CN111107395A (en) * | 2019-12-31 | 2020-05-05 | 广州市百果园网络科技有限公司 | Video transcoding method, device, server and storage medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5214507A (en) * | 1991-11-08 | 1993-05-25 | At&T Bell Laboratories | Video signal quantization for an mpeg like coding environment |
| CN1492685A (en) * | 2002-10-03 | 2004-04-28 | ���µ�����ҵ��ʽ���� | Image coding method and image coding device |
| DE10200901B4 (en) * | 2001-09-14 | 2004-09-16 | Siemens Ag | Efficient coding of video signals for scalable simul-cast storage and transmission as well as the associated codec |
| CN101729897A (en) * | 2005-03-16 | 2010-06-09 | 株式会社东芝 | Image encoding/decoding method and apparatus therefor |
| CN101902296A (en) * | 2010-06-23 | 2010-12-01 | 中兴通讯股份有限公司 | Coding/decoding method and device for fountain codes |
| CN103596011A (en) * | 2013-11-20 | 2014-02-19 | 北京中星微电子有限公司 | Storage processing method and device for image data |
| CN103888762A (en) * | 2014-02-24 | 2014-06-25 | 西南交通大学 | Video coding framework based on HEVC standard |
| US20140286422A1 (en) * | 2013-03-22 | 2014-09-25 | Fujitsu Limited | Video coding apparatus and method |
| CN105338357A (en) * | 2015-09-29 | 2016-02-17 | 湖北工业大学 | Distributed video compressed sensing coding technical method |
| CN105828069A (en) * | 2016-04-06 | 2016-08-03 | 上海交通大学 | Encoder adaptive adjustment method based on subjective quality evaluation |
-
2016
- 2016-09-13 CN CN201610822438.7A patent/CN107820084B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5214507A (en) * | 1991-11-08 | 1993-05-25 | At&T Bell Laboratories | Video signal quantization for an mpeg like coding environment |
| DE10200901B4 (en) * | 2001-09-14 | 2004-09-16 | Siemens Ag | Efficient coding of video signals for scalable simul-cast storage and transmission as well as the associated codec |
| CN1492685A (en) * | 2002-10-03 | 2004-04-28 | ���µ�����ҵ��ʽ���� | Image coding method and image coding device |
| CN101729897A (en) * | 2005-03-16 | 2010-06-09 | 株式会社东芝 | Image encoding/decoding method and apparatus therefor |
| CN101902296A (en) * | 2010-06-23 | 2010-12-01 | 中兴通讯股份有限公司 | Coding/decoding method and device for fountain codes |
| US20140286422A1 (en) * | 2013-03-22 | 2014-09-25 | Fujitsu Limited | Video coding apparatus and method |
| CN103596011A (en) * | 2013-11-20 | 2014-02-19 | 北京中星微电子有限公司 | Storage processing method and device for image data |
| CN103888762A (en) * | 2014-02-24 | 2014-06-25 | 西南交通大学 | Video coding framework based on HEVC standard |
| CN105338357A (en) * | 2015-09-29 | 2016-02-17 | 湖北工业大学 | Distributed video compressed sensing coding technical method |
| CN105828069A (en) * | 2016-04-06 | 2016-08-03 | 上海交通大学 | Encoder adaptive adjustment method based on subjective quality evaluation |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109788316A (en) * | 2019-02-27 | 2019-05-21 | 腾讯科技（深圳）有限公司 | Code rate control, video transcoding method and device, computer equipment and storage medium |
| CN109788316B (en) * | 2019-02-27 | 2020-08-14 | 腾讯科技（深圳）有限公司 | Code rate control method and device, video transcoding method and device, computer equipment and storage medium |
| CN110139101A (en) * | 2019-06-20 | 2019-08-16 | 电子科技大学 | A kind of frame level bit distribution method based on the control of the domain λ code rate |
| CN110139101B (en) * | 2019-06-20 | 2022-03-25 | 电子科技大学 | Frame-level bit distribution method based on lambda domain code rate control |
| CN111107395A (en) * | 2019-12-31 | 2020-05-05 | 广州市百果园网络科技有限公司 | Video transcoding method, device, server and storage medium |
| CN111107395B (en) * | 2019-12-31 | 2021-08-03 | 广州市百果园网络科技有限公司 | Video transcoding method, device, server and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107820084B (en) | 2020-02-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101919250B (en) | Pixel block processing | |
| US9183643B2 (en) | Apparatus and method for generating typical colors for image coding | |
| CN103188493B (en) | Image encoding apparatus and image encoding method | |
| CN104247422A (en) | Angular Table for Improving Intra Prediction | |
| CN105959698A (en) | Method and apparatus for performing interpolation based on transform and inverse transform | |
| CN104539949B (en) | The method and device of quick partitioning based on edge direction in HEVC screen codings | |
| CN104303504B (en) | Stablize different coding and the decoding of information and instantaneous/random information | |
| JP2017085603A (en) | Region-based image compression | |
| CN107820084A (en) | A kind of video-aware coding method and device | |
| JP6035026B2 (en) | Image coding method | |
| US20220277491A1 (en) | Method and device for machine learning-based image compression using global context | |
| CN102651815A (en) | Method of compression of digital images using a fixed number of bits per block | |
| CN108353175A (en) | The method and apparatus of prediction processing vision signal caused by coefficient of utilization | |
| CN109923860A (en) | Quantization parameter determination and layered code | |
| WO2022028197A1 (en) | Image processing method and device thereof | |
| CN107770525A (en) | A kind of method and device of Image Coding | |
| CN103716623A (en) | Video compression encoding-and-decoding method and encoder-decoder on the basis of weighting quantification | |
| CN107454397A (en) | The image coding/decoding method and device of color space conversion are performed to predicted value | |
| CN107454414A (en) | Image coding/decoding method and relevant apparatus | |
| KR20200138079A (en) | Method and apparatus for image compressing based on machine-learning using global context | |
| CN105556850B (en) | Encoder, decoder and its operation method | |
| JP5223150B1 (en) | Image processing apparatus and image processing method | |
| CN107155364A (en) | The method and apparatus that the signal based on figure is handled using geometric graphic element | |
| CN109547788B (en) | Image compression method, equipment and image transmission system | |
| JP2015019364A (en) | Image compression apparatus and 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |