CN106131562A - A kind of code stream importance computational methods and device - Google Patents

Abstract

The present invention is applicable to field of video transmission, a kind of code stream importance computational methods and device are provided, the present invention is by obtaining the code stream of video to be transmitted, and this code stream is multiple packets flow at transmitted per unit time of this video to be transmitted after scalable video；In this code stream, the distortion factor to this packet carries out minimum operation and draws rate distortion importance；This packet is carried out motion vector intensity computing and draws kinetic characteristic importance；The rate distortion importance of this packet and the kinetic characteristic importance of this packet are done multiplying and obtains the importance of this packet to determine the priority that this packet transmits；According to the importance of this packet, this packet is transmitted.In the present invention, rate distortion importance according to this packet and kinetic characteristic importance determine the importance of this packet, and according to the importance of this packet, the packet that the importance of prioritised transmission packet is high, decrease the phenomenons such as packet loss, error code, delay, improve the video quality that user broadcasts.

Description

A kind of code stream importance computational methods and device

Technical field

The present invention is applicable to field of video transmission, particularly relates to a kind of code stream importance computational methods and device.

Background technology

Scalable video to be carried out (Scalable Video Coding SVC) before video is transmitted, And uploaded to transmission channel by transmitting terminal, and after network transmits, receiving terminal receive and decode, user just can be broadcast by terminal Put device and play video.

Due to restriction and the transmission characteristic of self of Internet bandwidth, when transmitting video data amount exceedes channel width Time, it may appear that the phenomenons such as packet loss, error code, delay, the video quality causing user to broadcast substantially reduces.Accordingly, it would be desirable to video The code stream of transmission carries out importance calculating.The code stream that prioritised transmission is bigger on video quality impact, to adapt to network state, reduces Distortion.

Summary of the invention

The present invention provides a kind of code stream importance computational methods and device, it is intended to solve to exceed letter when transmitting video data amount During road bandwidth, the phenomenons such as packet loss, error code, delay occur, the problem that the video quality causing user to broadcast reduces.

For solving above-mentioned technical problem, the present invention is achieved in that a kind of code stream importance computational methods, described method Including:

Obtaining the code stream of video to be transmitted, described code stream is the multiple of described video to be transmitted after scalable video Packet is at the flow of transmitted per unit time；In described code stream, the distortion factor to described packet carries out minimum operation and obtains Go out the rate distortion importance of described packet；The frame at packet place described in described code stream is carried out motion vector intensity computing Draw the kinetic characteristic importance of described packet；By rate distortion importance and the described number of described packet in described code stream Carry out multiplying according to the kinetic characteristic importance of bag, obtain the importance of described packet, to determine that described packet transmits Priority；In described code stream, according to the importance of described packet, described packet is transmitted.

A kind of code stream importance calculates device, and described device includes:

Acquisition module is for obtaining the code stream of video to be transmitted, and described code stream is described to be passed after scalable video Multiple packets of defeated video are at the flow of transmitted per unit time；First computing module is used in described code stream described number Carry out minimum operation according to the distortion factor of bag and draw the rate distortion importance of described packet；Second computing module is for described The frame of packet place described in code stream carries out motion vector intensity computing and draws the kinetic characteristic importance of described packet；The Three computing modules are used for rate distortion importance and the kinetic characteristic weight of described packet of described packet in described code stream The property wanted carries out multiplying, obtains the importance of described packet, to determine the priority that described packet transmits；Transport module For in described code stream, according to the importance of described packet, described packet is transmitted.

The present invention compared with prior art, has the beneficial effects that: the present invention is by obtaining the code stream of video to be transmitted, this code Flow for multiple packets of this video to be transmitted after scalable video at the flow of transmitted per unit time；At this code stream In the distortion factor of this packet carried out minimum operation draw rate distortion importance；To the frame at this packet place in this code stream Carry out motion vector intensity computing and draw the kinetic characteristic importance of this packet；To the rate distortion importance of this packet and should The kinetic characteristic importance of packet does multiplying, and to obtain the importance of this packet preferential to determine that this packet transmits Level；According to the importance of this packet, this packet is transmitted.In the present invention, the rate distortion according to this packet is important Property and kinetic characteristic importance determine the importance of this packet, and according to the importance of this packet, these data of prioritised transmission The high packet of importance of bag, decreases the phenomenons such as packet loss, error code, delay, improves the video quality that user broadcasts.

Accompanying drawing explanation

In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing In having technology to describe, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is only this Some embodiments of invention.

Fig. 1 is that a kind of code stream importance computational methods that first embodiment of the invention provides realize schematic flow sheet；

Fig. 2 is that a kind of code stream importance computational methods that second embodiment of the invention provides realize schematic flow sheet；

Fig. 3 is that a kind of code stream importance that third embodiment of the invention provides calculates device schematic diagram；

Fig. 4 is that a kind of code stream importance that fourth embodiment of the invention provides calculates device schematic diagram.

Detailed description of the invention

In order to make the purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, right The present invention is further elaborated.Obviously, described embodiment is only a part of embodiment of the present invention, and not all Embodiment.Based on the embodiment in the present invention, those skilled in the art are obtained under not making creative work premise Every other embodiment, broadly falls into the scope of protection of the invention.

It should be noted that code stream to be those skilled in the art many to the video to be transmitted after scalable video The proprietary call of one of the flow that individual data wrap in transmitted per unit time.

The code stream importance computational methods that the embodiment of the present invention provides can apply to various terminal unit, such as computer, hands The terminal such as machine, panel computer and other terminal.

As first embodiment of the present invention, as it is shown in figure 1, Fig. 1 is a kind of code that first embodiment of the invention provides Stream importance computational methods realize schematic flow sheet, and the step of the code stream importance computational methods shown in Fig. 1 includes:

S101, obtain the code stream of video to be transmitted；

This code stream is that multiple packets of this video to be transmitted after scalable video are at transmitted per unit time Flow.

The approach of the code stream obtaining this video to be transmitted is not construed as limiting, and can obtain the code of the video to be transmitted of terminal storage Stream, it is also possible to obtain the code stream of the video to be transmitted downloaded from server, it is also possible to obtain the video to be transmitted of other terminal Code stream.

S102, in this code stream, the distortion factor to this packet carries out minimum operation and draws the rate distortion weight of this packet The property wanted；

This rate distortion importance is the priority of this packet transmission determined by the distortion factor of this packet, this rate distortion Importance values is the biggest, and the priority of this packet transmission is the highest.

S103, the frame at this packet place in this code stream is carried out motion vector intensity computing draw the motion of this packet Characteristic importance；

The motion vector intensity of this frame is the human eye sensitivity to different video content change, and the embodiment of the present invention is passed through The motion vector intensity of the frame calculating this packet place obtains the kinetic characteristic importance of this packet, to obtain according to human eye The priority of this packet transmission obtaining the sensitivity of different video content change, wherein, the motion of this packet is special Property importance the biggest, this packet transmission priority the highest.

It should be noted that the order of step S102 and step 103 can mutually be replaced.

S104, in this code stream, the rate distortion importance of this packet and the kinetic characteristic importance of this packet are carried out Multiplying, obtains the importance of this packet；

S105, in this code stream according to the importance of this packet, this packet is transmitted.

The importance of this packet is for determining the priority that this packet transmits, and the importance of this packet is the biggest, should The priority of packet transmission is the highest, then be preferentially transmitted this packet.

The embodiment of the present invention is by obtaining the code stream of video to be transmitted, and this code stream is that this is to be passed after scalable video Multiple packets of defeated video are at the flow of transmitted per unit time；In this code stream, the distortion factor of this packet is carried out minimum Value computing draws the rate distortion importance of this packet；The frame at this packet place is carried out motion vector intensity computing and draws this Packet kinetic characteristic importance；The rate distortion importance of this packet and the kinetic characteristic importance of this packet are done multiplication Computing obtains the importance of this packet, to determine the priority that this packet transmits；According to the importance of this packet, to this Packet is transmitted.In the present invention, according to rate distortion importance and the kinetic characteristic importance of this packet of this packet Determine the importance of this packet, and according to the importance of this packet, the data that the importance of this packet of prioritised transmission is high Bag, decreases the phenomenons such as packet loss, error code, delay, improves the video quality that user broadcasts.

As second embodiment of the present invention, as in figure 2 it is shown, Fig. 2 is a kind of code that second embodiment of the invention provides Stream importance computational methods realize schematic flow sheet, and the step of a kind of code stream importance computational methods shown in Fig. 2 includes:

S201, obtain the code stream of video to be transmitted；

This code stream is that multiple packets of this video to be transmitted after scalable video are at transmitted per unit time Flow.

The approach of the code stream obtaining video to be transmitted is not construed as limiting, and can obtain the code of the video to be transmitted of terminal storage Stream, it is also possible to obtain the code stream of the video to be transmitted downloaded from server, it is also possible to obtain the video to be transmitted of other terminal Code stream.

S202, the reconstruction distortion calculating this packet and drift distortion；

The distortion of rebuilding of this packet is the decoder distortion causing this frame after this data-bag lost.The drift of this packet is lost The true distortion caused for causing the error diffusion of other frame in same picture group after this data-bag lost.

Calculate and lose after this packet, the frame at this packet place and other reconstruction frames mean square with reference to this frame when encoding Error sum obtains the drift distortion of this frame, and this reconstruction frames is the frame rebuild after this LOF packet.

If B_kRepresent the code stream of the i-th frame kth packet, D_k ^rec(B_k) represent the reconstruction distortion of the i-th frame kth packet, D_k ^dra(B_k) represent the drift distortion of the i-th frame kth packet, then

$D_k^{rec}\left(B_k\right)=E{\[F(i,k,m)-F(i-1,k,m)\]}^2$

$D_k^{dra}\left(B_k\right)=\underset{i=n}{\overset{l}{\Σ}}E{\[F(i,k,m)-\hat{F}(i,k,m)\]}^2$

Wherein, F (i, k, m), F (i-1, k, m) be respectively the m-th of kth packet in i frame and i-1 frame and rebuild macro block, E[x]²Represent the mean square error of adjacent two frames,Represent the reconstruction frames after losing m macro block.

It should be noted that this video to be transmitted after scalable video comprises I frame, P frame in transmitting procedure With B frame, in the embodiment of the present invention, the i-th frame of setting only represents the order of frame, is not restricted the type of frame.

S203, this packet is rebuild the drift distortion of distortion and this packet carry out additive operation and obtain this packet The distortion factor；

The drift distortion rebuilding distortion and this packet of this packet calculated by step 202, calculates this packet The distortion factor, then the distortion factor of the i-th frame kth packet is:

D_k(B_k)=D_k ^rec(B_k)+D_k ^dra(B_k)

S204, in this code stream, the distortion factor to this packet carries out minimum operation and draws the rate distortion weight of this packet The property wanted；

It is then C when transmission bandwidth_BTime, the rate distortion importance of the i-th frame kth packet can be expressed as:

Wherein,

Utilizing method of Lagrange multipliers, this packet importance fortran is:

$\[P_D(i,k)\]=\underset{B_1,B_2,...B_N}{min}\underset{k=1}{\overset{N}{\Σ}}{D}_k\left(B_k\right)+\mathrm{\λ}(\underset{k=1}{\overset{N}{\Σ}}{B}_k\≤C_B),$

Wherein,Then

[P_D(i, k)]=λ

S205, the background motion vector intensity calculating each frame and object motion vector strength；

This background motion vector intensity represents the background change intensity of this frame.This object motion vector strength represents this frame Object variation intensity.

This frame is carried out background motion vector intensity computing and obtains the background motion vector intensity of this frame；It is right to carry out this frame As motion vector intensity computing obtains the object motion vector strength of this frame；If the background motion vector intensity of the i-th frame is M_iB, the The object motion vector strength of i frame is M_iO, then,

$M_{iB}=\frac{\underset{n}{\Σ}\underset{m}{\Σ}\sqrt{v_{xib}^2(m,n)+v_{yib}^2(m,n)}}{V_{\max }*N_{ib}},$

$M_{iO}=\frac{\underset{n}{\Σ}\underset{m}{\Σ}\sqrt{v_{xio}^2(m,n)+v_{yio}^2(m,n)}}{V_{\max }*N_{io}},$

Wherein, the background (v of this frame_xib,v_yib) it is that (m n) vows macro block from the motion of the reference frame of estimation to present frame Amount, V_maxFor the maximum of this frame motion vector, N_ibFor the macroblock number of background non-zero motion vectors, the object (v of this frame_xio,v_yio) For macro block, (m, n) from the motion vector of the reference frame of estimation to present frame, N_ioMacroblock number for object non-zero motion vectors.

S206, background motion vector intensity and object motion vector strength to this frame compute weighted, and obtain this frame Kinetic characteristic importance；

Background motion vector intensity and object motion vector strength to this frame compute weighted and obtain the motion of this frame Vector strength, this motion vector intensity is the human eye sensitivity to different video content change, if the motion vector of the i-th frame is strong Degree is: M_i, then,

M_i=ω_BM_iB+ω_OM_iO

Wherein, ω_B、ω_OIt is respectively weight and the weight of object motion vector of background motion vector, wherein, ω_B, ω_O ∈ [0,1], and ω_B+ω_O=1.

Make the weights omega of this background motion vector_BBy spatial distribution ω₁, exercise intensity ω₂With motion complexity ω₃Three Individual factors composition, then,

ω_B=ω₁+ω₂+ω₃,

${\mathrm{\ω}}_1=\frac{N_B}{Num\left(s\right)},$

Wherein, N_BFor the macroblock number of motion vector non-zero in background, Num (s) is the macroblock number of this frame；

${\mathrm{\ω}}_2=\frac{\underset{i=1}{\overset{N_B}{\Σ}}\left(\right|\left|{\mathrm{mv}}_x\right||+|\left|{\mathrm{mv}}_y\right|\left|\right)}{\underset{i=1}{\overset{Num\left(s\right)}{\Σ}}\left(\right|\left|{\mathrm{mv}}_x\right||+|\left|{\mathrm{mv}}_y\right|\left|\right)},$

Wherein, mv_x, mv_yIt is respectively abscissa and vertical coordinate, the ω of motion vector₂Represent exercise intensity, i.e. kinergety Size, this kinergety is the biggest, and movable information is the abundantest.

${\mathrm{\ω}}_3=PE=-(\underset{l=1}{\overset{m}{\Σ}}\frac{Num\left(s_l\right)}{Num\left(s\right)}*\log (\frac{Num\left(s_l\right)}{Num\left(s\right)}\left)\right)/\log \left(36\right)$

Wherein, s=[s is made₁,s₂..., s_m], wherein m 36, s_lIt it is the dimension of each non-NULL in motion vector direction rectangular histogram Degree, Num (s_l) it is the macroblock number of motion vector non-zero, then ω in each dimension₃It is motion complexity.

The kinetic characteristic importance of this frame be according to human eye in this frame different content change sensitivity and obtain The priority of this frame transmission.

The embodiment of the present invention obtains the kinetic characteristic importance P of this frame by the motion vector intensity calculating this frame_V(i), To obtain the priority of this frame transmission sensitivity of different video content change obtained according to human eye, the then fortune of the i-th frame Dynamic characteristic importance is:

P_V(i)=M_i

S207, the kinetic characteristic importance of the frame at this packet place is defined as the kinetic characteristic importance of packet.

The importance of the i-th frame kth packet is P_V(i, k)=M_i。

S208, in this code stream, rate distortion importance and this packet kinetic characteristic importance of this packet are taken advantage of Method computing, obtains the importance of this frame；

By the formula shown in step S202 to S207, the importance of the i-th frame kth packet is:

PRI (i, k)=[P_D(i,k)]^α×[P_V(i,k)]^β

Further, the importance of the i-th frame is:

PRI (i, k)=λ^α×M_i ^β

Wherein, α, β ∈ [0,1], and alpha+beta=1, α, β are empirical value, and its value is according to the content character of this video to be transmitted Determine, for the video scene that kinetic characteristic is complicated, such as football match etc., or picture comprise the video of multiple moving region Scene, such as, perform live etc., then need to arrange bigger β value and typically larger than 0.5；Video field simple for kinetic characteristic Scape, such as news report etc., then needs to arrange bigger α value and typically larger than 0.5.

S209, in this code stream according to the importance of this packet, this packet is transmitted.

This packet importance is for determining the priority that this packet transmits, and this packet importance is the biggest, these data The priority of bag transmission is the highest, then be preferentially transmitted this packet.

The embodiment of the present invention is by obtaining the code stream of video to be transmitted, and this code stream is that this is to be passed after scalable video Multiple packets of defeated video are at the flow of transmitted per unit time；Calculate reconstruction distortion and the drift distortion of this packet, right Reconstruction distortion and the drift distortion of this packet carry out additive operation and draw the distortion factor of this packet, the distortion to this packet Degree carries out minimum operation and draws this data packet rate distortion importance；Calculate background motion vector intensity and the object motion of this frame Vector strength, and the background motion vector intensity and object motion vector strength to this frame computes weighted and draws the fortune of this frame Dynamic characteristic importance；Kinetic characteristic importance according to this frame determines the kinetic characteristic importance of this packet, to this packet Rate distortion importance and the kinetic characteristic importance of this packet do multiplying and obtain the importance of this packet, to determine The priority of this packet transmission；According to the importance of this packet, this packet is transmitted.

In the present invention, rate distortion importance and kinetic characteristic importance according to this packet determine the important of this packet Property, and according to the importance of this packet, the packet that the importance of prioritised transmission packet is high, decrease packet loss, error code, prolong The phenomenon such as late, improves the video quality that user broadcasts.

As the 3rd embodiment of the present invention, as it is shown on figure 3, Fig. 3 is a kind of code that third embodiment of the invention provides Stream importance calculates device schematic diagram, and a kind of code stream importance shown in Fig. 1 calculates device, and this device specifically includes that acquisition mould Block the 31, first computing module the 32, second computing module the 32, the 3rd computing module 34 and transport module 35；

Wherein, acquisition module 31, for obtaining the code stream of video to be transmitted；

This code stream be after scalable video multiple packets of described video to be transmitted at transmitted per unit time Flow.

First computing module 32, carries out minimum operation for the distortion factor to this packet in this code stream and draws this number Rate distortion importance according to bag；

This rate distortion importance is the priority of this packet transmission determined by the distortion factor of this packet, this rate distortion Importance values is the biggest, and the priority of this packet transmission is the highest.

Second computing module 33, carries out motion vector intensity computing for frame to this packet place in this code stream and obtains Go out the kinetic characteristic importance of this packet；

The motion vector intensity of this frame is the human eye sensitivity to different video content change, and the embodiment of the present invention is passed through The motion vector intensity of the frame calculating this packet place obtains the kinetic characteristic importance of this packet, to obtain according to human eye The priority of this packet transmission obtaining the sensitivity of different video content change, wherein, the motion of this packet is special Property importance values the biggest, this packet transmission priority the highest.

3rd computing module 34, is used for the motion of the rate distortion importance of this packet He this packet in this code stream Characteristic importance carries out multiplying, obtains the importance of this packet, to determine the priority that this packet transmits；

Transport module 35, in this code stream, is transmitted this packet according to the importance of this packet.

The importance of this packet is for determining the priority that this packet transmits, and the importance values of this packet is the biggest, The priority of this packet transmission is the highest, then be preferentially transmitted this packet.

The embodiment of the present invention is not use up details and refer to the first embodiment shown in Fig. 1, do not repeat them here.

The embodiment of the present invention obtains the code stream of video to be transmitted by acquisition module 31, and this code stream is to compile through telescopic video After Ma, multiple packets of this video to be transmitted are at the flow of transmitted per unit time；First computing module 32 is right in this code stream The distortion factor of this packet carries out minimum operation and draws rate distortion importance；Second computing module 33 is to this packet place Frame carries out motion vector intensity computing and draws this packet kinetic characteristic importance；3rd computing module 34 rate to this packet The kinetic characteristic importance of distortion importance and this packet does multiplying and obtains the importance of this packet, to determine this number Priority according to bag transmission；This packet, according to the importance of this packet, is transmitted by transport module 35.In the present invention, Rate distortion importance according to this packet and kinetic characteristic importance determine the importance of this packet, and according to this packet Importance, the packet that the importance of prioritised transmission packet is high, decrease the phenomenons such as packet loss, error code, delay, improve use The video quality that family is broadcasted.

As the 4th embodiment of the present invention, as shown in Figure 4, Fig. 4 is a kind of code that fourth embodiment of the invention provides Stream importance calculates device schematic diagram, and a kind of code stream importance shown in Fig. 4 calculates device, and this device specifically includes that acquisition mould Block the 41, first computing module the 42, second computing module the 43, the 3rd computing module 44, transport module 45, rebuild computing module 46, Drift calculation module 47, background computing module 48 and calculation and object module 49；Wherein, the 3rd computing module 44 includes: first To module 441, second obtain module 442 and multiplication computing module 443；Details are as follows for the function of modules:

Acquisition module 41, for obtaining the code stream of video to be transmitted；

This code stream be after scalable video multiple packets of described video to be transmitted at transmitted per unit time Flow.

First computing module 42, carries out minimum operation for the distortion factor to this packet in this code stream and draws this number Rate distortion importance according to bag；

The rate distortion importance of this packet is the priority of this packet transmission determined by the distortion factor of this packet, This rate distortion importance values is the biggest, and the priority of this packet transmission is the highest.

When transmission bandwidth is C_BTime, the rate distortion importance of the i-th frame kth packet can be expressed as:

Wherein,

Utilizing method of Lagrange multipliers, the rate distortion importance fortran of the i-th frame kth packet is:

$\[P_D(i,k)\]=\underset{B_1,B_2,...B_N}{min}\underset{k=1}{\overset{N}{\Σ}}{D}_k\left(B_k\right)+\mathrm{\λ}(\underset{k=1}{\overset{N}{\Σ}}{B}_k\≤C_B),$

Wherein,Then

[P_D(i, k)]=λ

Second computing module 43, carries out motion vector intensity computing for frame to this packet place in this code stream and obtains Go out the kinetic characteristic importance of this packet；

The motion vector intensity of the frame at this packet place is the human eye sensitivity to different video content change, this Bright embodiment obtains the kinetic characteristic importance of this packet by the motion vector intensity calculating the frame at this packet place, with Obtain the priority of this packet transmission sensitivity of different video content change obtained according to human eye, wherein, this number Kinetic characteristic importance values according to bag is the biggest, and the priority of this packet transmission is the highest.

If the motion vector intensity of the i-th frame is: M_i, then the i-th frame kth packet kinetic characteristic importance be:

P_V(i, k)=M_i

3rd computing module 44, is used for the described rate distortion importance of this packet and this packet in this code stream Kinetic characteristic importance carries out multiplying, obtains the importance of this packet, to determine the priority that this packet transmits；

Wherein, the 3rd computing module 44 includes:

First obtains module 441, and for the rate distortion importance of this packet carries out exponent arithmetic, to obtain first important Property parameter；

Second obtains module 442, obtains the second weight for the kinetic characteristic importance of this packet is carried out exponent arithmetic The property wanted parameter；

Multiplication computing module 443, obtains for this first importance parameter is done multiplying with this second importance parameter Importance to this packet.

The importance of the i-th frame kth packet is:

PRI (i, k)=[P_D(i,k)]^α×[P_V(i,k)]^β

Further, the importance of the i-th frame kth packet is:

PRI (i, k)=λ^α×M_i ^β

Wherein, α, β ∈ [0,1], and alpha+beta=1, α, β are empirical value, and its value is according to the content character of this video to be transmitted Determine, for the video scene that kinetic characteristic is complicated, such as football match etc., or picture comprise the video of multiple moving region Scene, such as, perform live etc., then need to arrange bigger β value and typically larger than 0.5；Video field simple for kinetic characteristic Scape, such as news report etc., then needs to arrange bigger α value and typically larger than 0.5.

Transport module 45, for according to the importance of this packet, the data that the importance of this packet of prioritised transmission is high Bag.

The importance of this packet is for determining the priority that this packet transmits, and the importance values of this packet is the biggest, The priority of this packet transmission is the highest, then be preferentially transmitted this packet.

First computing module 42, the drift distortion being additionally operable to rebuild this packet distortion and this packet carries out addition Computing obtains the distortion factor of this packet；

The reconstruction distortion of the i-th frame kth packet is D_k ^rec(B_k), the drift distortion of the i-th frame kth packet is D_k ^dra (B_k), then the distortion factor of the i-th frame kth packet is:

D_k(B_k)=D_k ^rec(B_k)+D_k ^dra(B_k)

Rebuild computing module 46, after being used for calculating this data-bag lost, the frame at this packet place and adjacent with this frame The mean square error of previous frame obtains the reconstruction distortion of this packet；

This packet rebuild the decoder distortion that distortion is the frame causing this packet place after this data-bag lost.

If B_kRepresent the code stream of the i-th frame kth packet, D_k ^rec(B_k) represent the reconstruction distortion of the i-th frame kth packet, Then

$D_k^{rec}\left(B_k\right)=E{\[F(i,k,m)-F(i-1,k,m)\]}^2$

Wherein, F (i, k, m), F (i-1, k, m) be respectively the m-th of kth packet in i frame and i-1 frame and rebuild macro block, E[x]²Represent the mean square error of adjacent two frames.

Drift calculation module 47, after being used for calculating this data-bag lost, with reference to being somebody's turn to do when the frame at this packet place and coding The mean square error sum of other reconstruction frames of frame obtains the drift distortion of this packet；

The drift distortion of this packet be cause after this data-bag lost other frame in same picture group error diffusion and The distortion caused.This reconstruction frames is the frame rebuild after this data-bag lost macro block.

If B_kRepresent the code stream of the i-th frame kth packet, D_k ^dra(B_k) represent the drift distortion of the i-th frame kth packet, Then

$D_k^{dra}\left(B_k\right)=\underset{i=n}{\overset{l}{\Σ}}E{\[F(i,k,m)-\hat{F}(i,k,m)\]}^2$

Wherein, F (i, k, m) be i frame kth packet m-th rebuild macro block, E [x]²Represent the mean square error of two frames,Represent the reconstruction frames after losing m macro block.

It should be noted that this video to be transmitted after scalable video comprises I frame, P frame in transmitting procedure With B frame, in the embodiment of the present invention, the i-th frame of setting only represents the order of frame, is not restricted the type of frame.

Second computing module 43, is additionally operable to the background motion vector intensity to this frame and object motion vector strength adds Power computing, obtains the kinetic characteristic importance of this frame.

Background motion vector intensity and object motion vector strength to this frame compute weighted and obtain the motion of this frame Vector strength, this motion vector intensity is the human eye sensitivity to different video content change, if the motion vector of the i-th frame is strong Degree is: M_i, then,

M_i=ω_BM_iB+ω_OM_iO

Wherein, ω_B、ω_OIt is respectively weight and the weight of object motion vector of background motion vector, wherein, ω_B, ω_O ∈ [0,1], and ω_B+ω_O=1.

Make the weights omega of this background motion vector_BBy spatial distribution ω₁, exercise intensity ω₂With motion complexity ω₃Three Individual factors composition, then,

ω_B=ω₁+ω₂+ω₃,

${\mathrm{\ω}}_1=\frac{N_B}{Num\left(s\right)},$

Wherein, N_BFor the macroblock number of motion vector non-zero in background, Num (s) is the macroblock number of this frame；

${\mathrm{\ω}}_2=\frac{\underset{i=1}{\overset{N_B}{\Σ}}\left(\right|\left|{\mathrm{mv}}_x\right||+|\left|{\mathrm{mv}}_y\right|\left|\right)}{\underset{i=1}{\overset{Num\left(s\right)}{\Σ}}\left(\right|\left|{\mathrm{mv}}_x\right||+|\left|{\mathrm{mv}}_y\right|\left|\right)},$

Wherein, mv_x, mv_yIt is respectively abscissa and vertical coordinate, the ω of motion vector₂Represent exercise intensity, i.e. kinergety Size, this kinergety is the biggest, and movable information is the abundantest.

${\mathrm{\ω}}_3=PE=-(\underset{l=1}{\overset{m}{\Σ}}\frac{Num\left(s_l\right)}{Num\left(s\right)}*\log (\frac{Num\left(s_l\right)}{Num\left(s\right)}\left)\right)/\log \left(36\right)$

Wherein, s=[s is made₁,s₂..., s_m], wherein m 36, s_lIt it is the dimension of each non-NULL in motion vector direction rectangular histogram Degree, Num (s_l) it is the macroblock number of motion vector non-zero, then ω in each dimension₃It is motion complexity.

The kinetic characteristic importance of this frame be according to human eye in this frame different content change sensitivity and obtain The priority of this frame transmission.

The embodiment of the present invention obtains the kinetic characteristic importance P of this frame by the motion vector intensity calculating this frame_V(i), To obtain the priority of this frame transmission sensitivity of different video content change obtained according to human eye, the then fortune of the i-th frame Dynamic characteristic importance is:

P_V(i)=M_i

Second computing module 43, is additionally operable to the kinetic characteristic importance of this packet place frame is defined as this packet Kinetic characteristic importance.

The importance of the i-th frame kth packet is:

P_V(i, k)=M_i。

Background computing module 48, the background motion obtaining this frame for this frame carries out background motion vector intensity computing is vowed Amount intensity；

This background motion vector intensity represents the background change intensity of this frame.If the background motion vector intensity of the i-th frame is M_iB, then,

$M_{iB}=\frac{\underset{n}{\Σ}\underset{m}{\Σ}\sqrt{v_{xib}^2(m,n)+v_{yib}^2(m,n)}}{V_{\max }*N_{ib}}$

Wherein, the background (v of this frame_xib,v_yib) it is that (m n) vows macro block from the motion of the reference frame of estimation to present frame Amount, V_maxFor the maximum of this frame motion vector, N_ibMacroblock number for background non-zero motion vectors.

Calculation and object module 49, the object motion obtaining this frame for this frame carries out object motion vector strength computing is vowed Amount intensity.

This object motion vector strength represents the object variation intensity of this frame.If the object motion vector strength of the i-th frame is M_iO, then,

$M_{iO}=\frac{\underset{n}{\Σ}\underset{m}{\Σ}\sqrt{v_{xio}^2(m,n)+v_{yio}^2(m,n)}}{V_{\max }*N_{io}},$

Wherein, the object (v of this frame_xio,v_yio) it is that (m n) vows macro block from the motion of the reference frame of estimation to present frame Amount, V_maxFor the maximum of this frame motion vector, N_ioMacroblock number for object non-zero motion vectors.

The embodiment of the present invention do not use up details, refer to the second embodiment shown in Fig. 2, do not repeat them here.

The embodiment of the present invention obtains the code stream of video to be transmitted by acquisition module 41, and this code stream is to compile through telescopic video After Ma, multiple packets of this video to be transmitted are at the flow of transmitted per unit time；Rebuild computing module 46 and calculate this packet Reconstruction distortion, Drift calculation module 47 calculates the drift distortion of this packet, first computing module 42 weight to this packet Build distortion and drift distortion carries out additive operation and draws the distortion factor of this packet, the mistake to this packet of first computing module 42 True degree carries out minimum operation and draws rate distortion importance；Background computing module 48 calculate this frame background motion vector intensity and Calculation and object module 48 calculates object motion vector strength, and the second computing module 43 is to the background motion vector intensity of this frame and right The kinetic characteristic importance of this frame is drawn as motion vector intensity computes weighted；Second computing module 43 is according to this packet The kinetic characteristic importance of the frame at place determines the kinetic characteristic importance of this packet；3rd computing module 44 is to this packet Rate distortion importance and the kinetic characteristic importance of this packet do multiplying and obtain the importance of this packet, to determine The priority of this packet transmission；This packet, according to the importance of this packet, is transmitted by transport module 45.

In the present invention, rate distortion importance and kinetic characteristic importance according to this packet determine the important of this packet Property, and according to the importance of this packet, the packet that the importance of prioritised transmission packet is high, decrease packet loss, error code, prolong The phenomenon such as late, improves the video quality that user broadcasts.

In multiple embodiments provided herein, it should be understood that disclosed system, apparatus and method, permissible Realize by another way.Such as, device embodiment described above is only schematically, such as, and described module Dividing, be only a kind of logic function and divide, actual can have other dividing mode, the most multiple modules or assembly when realizing Can in conjunction with or be desirably integrated into another system, or some features can be ignored, or does not performs.Another point, shown or The coupling each other discussed or direct-coupling or communication linkage can be the indirect couplings by some interfaces, device or module Close or communication linkage, can be electrical, machinery or other form.

The described module illustrated as separating component can be or may not be physically separate, shows as module The parts shown can be or may not be physical module, i.e. may be located at a place, or can also be distributed to multiple On mixed-media network modules mixed-media.Some or all of module therein can be selected according to the actual needs to realize the mesh of the present embodiment scheme 's.

It addition, each functional module in each embodiment of the present invention can be integrated in a processing module, Ke Yishi Modules is individually physically present, it is also possible to two or more modules are integrated in a module.Above-mentioned integrated module Both can realize to use the form of hardware, it would however also be possible to employ the form of software function module realizes.

It should be noted that for aforesaid each method embodiment, in order to simplicity describes, therefore it is all expressed as a series of Combination of actions, but those skilled in the art should know, the present invention is not limited by described sequence of movement because According to the present invention, some step can use other order or carry out simultaneously.Secondly, those skilled in the art also should know Knowing, it might not be all this that embodiment described in this description belongs to preferred embodiment, involved action and module Bright necessary.

In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and does not has the portion described in detail in certain embodiment Point, may refer to the associated description of other embodiments.

The foregoing is only presently preferred embodiments of the present invention, not in order to limit invention, all spirit in the present invention Any amendment, equivalent and improvement etc. with being made within principle, should be included within the scope of the present invention.

Claims (14)

1. code stream importance computational methods, it is characterised in that described method includes:

Obtaining the code stream of video to be transmitted, described code stream is multiple data of described video to be transmitted after scalable video Wrap in the flow of transmitted per unit time；

In described code stream, the distortion factor to described packet carries out minimum operation and show that the rate distortion of described packet is important Property；

The frame at packet place in described code stream is carried out motion vector intensity computing and draws the kinetic characteristic weight of described packet The property wanted；

The rate distortion importance of described packet and the kinetic characteristic importance of described packet are taken advantage of by described code stream Method computing, obtains the importance of described packet, to determine the priority that described packet transmits；

In described code stream, according to the importance of described packet, described packet is transmitted.

2. the method for claim 1, it is characterised in that described in described code stream, the distortion factor of described packet is entered Include before the rate distortion importance that row minimum operation draws described packet:

Described packet is rebuild the drift distortion of distortion and described packet carry out additive operation and obtain described packet The distortion factor.

3. method as claimed in claim 2, it is characterised in that the described reconstruction distortion to described packet and described packet Drift before distortion carries out the distortion factor that additive operation obtains described packet, including:

After calculating described data-bag lost, the frame at this packet place and the previous frame adjacent with the frame at described packet place Mean square error obtain the reconstruction distortion of described packet；

After calculating described data-bag lost, the frame at this packet place obtains described packet with the mean square error sum of reconstruction frames Drift distortion, described reconstruction frames is the frame rebuild after described data-bag lost macro block.

4. the method for claim 1, it is characterised in that described in described code stream, described frame to be carried out motion vector strong Degree computing show that the kinetic characteristic importance of described packet includes:

Background motion vector intensity and object motion vector strength to described frame compute weighted, and obtain the motion of described frame Characteristic importance；

The kinetic characteristic importance of described frame is defined as the kinetic characteristic importance of described packet.

5. method as claimed in claim 4, it is characterised in that the described background motion vector intensity to described frame and object fortune Dynamic vector intensity computes weighted, before obtaining the kinetic characteristic importance of described frame, including:

Described frame is carried out background motion vector intensity computing and obtains the described background motion vector intensity of described frame；

Described frame is carried out object motion vector strength computing and obtains the described object motion vector strength of described frame.

6. the method for claim 1, it is characterised in that described in described code stream by the rate distortion weight of described packet The kinetic characteristic of the property wanted and described packet is important carries out product calculation, and the importance obtaining described packet includes:

The rate distortion importance of described packet is carried out exponent arithmetic and obtains the first importance parameter；

The kinetic characteristic importance of described packet is carried out exponent arithmetic and obtains the second importance parameter；

Described first importance parameter and described second importance parameter are done multiplying and obtains the importance of described packet.

7. the method for claim 1, it is characterised in that the described importance according to described packet, to described data Bag is transmitted including:

According to the importance of described packet, the packet that the importance of packet described in prioritised transmission is high.

8. a code stream importance calculates device, it is characterised in that described device includes:

Acquisition module, for obtaining the code stream of video to be transmitted, described code stream is described to be transmitted after scalable video Multiple packets of video are at the flow of transmitted per unit time；

First computing module, carries out minimum operation for the distortion factor to described packet in described code stream and draws described number Rate distortion importance according to bag；

Second computing module, draws described number for the frame at packet place in described code stream is carried out motion vector intensity computing Kinetic characteristic importance according to bag；

3rd computing module, is used for rate distortion importance and the motion of described packet of described packet in described code stream Characteristic importance carries out multiplying, obtains the importance of described packet, to determine the priority that described packet transmits；

Transport module, in described code stream, according to the importance of described packet, is transmitted described packet.

9. device as claimed in claim 8, it is characterised in that

Described first computing module, the drift distortion being additionally operable to rebuild described packet distortion and described packet adds Method computing obtains the distortion factor of described packet.

10. device as claimed in claim 9, it is characterised in that described device also includes:

Rebuild computing module, after being used for calculating this data-bag lost, the frame at described packet place and with described packet place The mean square error of the adjacent previous frame of frame obtain the reconstruction distortion of described packet；

Drift calculation module, after being used for calculating this data-bag lost, the frame at described packet place and the mean square error of reconstruction frames Sum obtains the drift distortion of described packet, and described reconstruction frames is the frame rebuild after described data-bag lost macro block.

11. devices as claimed in claim 8, it is characterised in that

Described second computing module, is additionally operable to the background motion vector intensity to described frame and object motion vector strength adds Power computing, obtains the described kinetic characteristic importance of described frame.

Described second computing module, the kinetic characteristic importance being additionally operable to the frame by described packet place is defined as described data The kinetic characteristic importance of bag.

12. devices as claimed in claim 11, it is characterised in that described device also includes:

Background computing module, obtains the described background motion of described frame for described frame is carried out background motion vector intensity computing Vector strength；

Calculation and object module, obtains the described object motion of described frame for described frame is carried out object motion vector strength computing Vector strength.

13. devices as claimed in claim 8, it is characterised in that described 3rd computing module includes:

First obtains module, obtains the first importance ginseng for the rate distortion importance of described packet carries out exponent arithmetic Number；

Second obtains module, obtains the second importance ginseng for the kinetic characteristic importance of described packet carries out exponent arithmetic Number；

Multiplication computing module, obtains institute for described first importance parameter and described second importance parameter are done multiplying State the importance of packet.

14. devices as claimed in claim 8, it is characterised in that

Described transport module, is additionally operable to the importance according to described packet, and described in prioritised transmission, the importance of packet is high Packet.