CN103167290A - Method and device for quantizing video coding movement intensity - Google Patents

Abstract

An embodiment of the invention discloses a method and a device for quantizing video coding movement intensity. The method includes the steps: firstly, providing a macro-block movement intensity and fuzzy intensity acquisition mechanism; and secondly, designing quantizing adjustment strategies on different coding structures and different frame categories according to the movement intensity and the fuzzy intensity of each macro-block. By designing the subsequent quantizing adjustment strategies, code rates can be secondarily adjusted according to the movement intensity, namely, lower code rates are distributed for the macro-blocks violently moving, the saved code rates are distributed to portions sensitive to subjective effect and rate-distortion performance improvement of a video image, and finally, rate-distortion performances and subjective effects are optimized when a coder violently or non-violently moves.

Description

A kind of Video coding exercise intensity quantization method and device

Technical field

The present invention relates to the coding and decoding video field, relate in particular to a kind of Video coding exercise intensity quantization method and device.

Background technology

The violent film source of moving has the little characteristics of inter-frame information redundancy, and therefore, this type of film source is compressed just needs encoder to have well behaved inter-frame coding.And when under low code check, during the violent film source of compression movement, the overall inadequate contradiction of code check will all the more show especially, and is at this moment, also just higher at the performance requirement of inter-frame information redundancy technology for eliminating to encoder accordingly.

On the other hand,, when the movement velocity of object reaches certain limit, the problems such as video image is fuzzy, diplopia often appear in the deficiency due to the equipment of IMAQ.Like this, concerning encoder, its original film source just has defective.At this moment, if encoder distributes larger code check to this type of video image, on the one hand, can not promote the subjective effect of video image, also can further produce the waste of code check on the other hand under low code check, aggravate the inadequate contradiction of overall code check, its to the lifting of distortion performance also without large help.And in motion slowly or in the collection of rest image, the problems such as this image blurring, diplopia can not appear.Yet, for this phenomenon, encoder commonly used is not generally all considered both differences, only designs general coding strategy, this is not done special design, this just makes encoder can not be issued in all kinds of situations the optimization realization of distortion performance, subjective effect.

Summary of the invention

The purpose of the embodiment of the present invention is to propose a kind of Video coding exercise intensity quantization method, be intended to solve the prior art encoder when encoding, do not consider in film source the difference of the violent and image that motion is slow or static of motion, adopt universal coding side rate, can not reach the optimized problem of distortion performance, subjective effect.

The embodiment of the present invention is achieved in that a kind of Video coding exercise intensity quantization method, and described method comprises:

Judgement frame _tWhether be the I frame, if not,

Make n=1;

Select current coding macro block Mb _t,nCoding mode;

Calculate current coding macro block Mb _t,nExercise intensity and vague intensity;

According to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

According to current coding macro block Mb _t,nCoding mode and revise quantization parameter

To macro block Mb _t,nAdvance

The row coding;

Judge whether n 〉=width*height/256 sets up, make if not n=n+1, return to and enter step " selection current coding macro block Mb _t,nCoding mode ", if,

Judgement min (FrameNum (frame _mm) | frame _mmDo not encode) whether exist, finish if do not exist, otherwise

Judgement FrameNum (frame _{m_i})=min (FrameNum (frame _mm) | frame _mmDo not encode) whether set up, if, make t=m_i, reenter step " judgement frame _tWhether be the I frame ", finish if not;

Wherein, frame _tIt is t frame coded frame; T presentation code frame frame _tTime sequence number in current source sequence; N represents macro block Mb _t,nAt frame _tIn position number; Mb _t,nRepresent t frame coded frame frame _tN macro block: width, height represents respectively columns and the line number of a two field picture take whole pixel as unit;

Frame _{m_i}Represent m_i frame coded frame; Frame _mmRepresent mm frame coded frame; M_i presentation code frame frame _{m_i}Time sequence number in current source sequence; Mm presentation code frame frame _mmTime sequence number in current source sequence; FrameNum (frame _mm) expression frame _mmThe coding sequence number; FrameNum (frame _{m_i}) expression frame _{m_i}The coding sequence number; Min (variable | condition) represent satisfying condition

All variablees minimize.

Further, described step " judgement frame _tWhether be the I frame " also comprise before step

Make t=1.

Further, described step " judgement frame _tWhether be the I frame ", if, to frame _tThen coding enters step " judgement min (FrameNum (frame _mm) | frame _mmDo not encode) whether exist ".

The purpose of the embodiment of the present invention is to propose a kind of Video coding exercise intensity quantization device, and described device comprises:

The first initialization module, frame category judging treatmenting module, the first coding module, the second initialization module, coding mode are selected module, exercise intensity and vague intensity computing module, are revised quantification parameter calculating module, the second coding module, the first judging treatmenting module, the second judging treatmenting module, the 3rd judging treatmenting module

The first initialization module is used for initialization t=1; T presentation code frame frame _tTime sequence number in current source sequence;

The frame category judging treatmenting module is used for judgement frame _tWhether be the I frame, if enter the first coding module, otherwise enter the second initialization module;

The first coding module is used for frame _tEncode, enter the second judging treatmenting module;

The second initialization module is used for initialization and makes n=1; N represents macro block Mb _t,nAt frame _tIn position number;

Coding mode is selected module, is used for selecting current coding macro block Mb _t,nCoding mode;

Exercise intensity and vague intensity computing module are used for calculating Mb _t,nExercise intensity and vague intensity;

Revise the quantification parameter calculating module, be used for according to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

The second coding module is used for according to current coding macro block Mb _t,nCoding mode and revise quantization parameter

, to macro block Mb _t,nEncode;

The first judging treatmenting module if be used for judgement n 〉=width*height/256, enters the second judging treatmenting module, otherwise, make n=n+1, reenter coding mode and select module; Width, height represent respectively columns and the line number of a two field picture take whole pixel as unit;

The second judging treatmenting module is if be used for judgement min (FrameNum (frame _mm) | frame _mmDo not encode) do not exist, finish, otherwise, enter the 3rd judging treatmenting module;

The 3rd judging treatmenting module is used for judgement

If FrameNum is (frame _{m_i})=min (FrameNum (frame _mm) | frame _mmDo not encode),

, make t=m_i, reenter the frame category judging treatmenting module, otherwise, finish;

Wherein, frame _{m_i}Represent m_i frame coded frame; Frame _mmRepresent mm frame coded frame; M_i presentation code frame frame _{m_i}Time sequence number in current source sequence; Mm presentation code frame frame _mmTime sequence number in current source sequence; FrameNum (frame _mm) expression frame _mmThe coding sequence number; FrameNum (frame _{m_i}) expression frame _{m_i}The coding sequence number; Min (variable | condition) represent all variablees that satisfy condition are minimized.

Beneficial effect of the present invention

The embodiment of the present invention proposes a kind of Video coding exercise intensity quantization method and device.At first propose macro block exercise intensity, vague intensity securing mechanism, then according to exercise intensity and the vague intensity of each macro block, design quantizes to adjust strategy on different coding structure and different frame classification.Adjust the design of strategy by this subsequent quantizatiion, can realize that code check carries out the secondary adjustment according to exercise intensity, namely reach for the violent less code check of macroblock allocation of motion, the Data Rate Distribution that this part is saved arrives the part that video image subjective effect, distortion performance is promoted sensitivity, finally, reach encoder and acutely reach the distortion performance in the violent situation of non-motion, the optimization realization of subjective effect in motion.

Description of drawings

Fig. 1 is a kind of Video coding exercise intensity of preferred embodiment of the present invention quantization method flow chart;

Fig. 2 is a kind of Video coding exercise intensity of preferred embodiment of the present invention quantization device structure chart.

Embodiment

In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated, for convenience of explanation, only show the part relevant to the embodiment of the present invention.Should be appreciated that the specific embodiment that this place is described, only be used for explaining the present invention, not in order to limit the present invention.

The embodiment of the present invention proposes a kind of Video coding exercise intensity quantization method and device.At first propose macro block exercise intensity, vague intensity securing mechanism, then according to exercise intensity and the vague intensity of each macro block, design quantizes to adjust strategy on different coding structure and different frame classification.Adjust the design of strategy by this subsequent quantizatiion, can realize that code check carries out the secondary adjustment according to exercise intensity, namely reach for the violent less code check of macroblock allocation of motion, the Data Rate Distribution that this part is saved arrives the part that video image subjective effect, distortion performance is promoted sensitivity, finally, reach encoder and acutely reach the distortion performance in the violent situation of non-motion, the optimization realization of subjective effect in motion.

Embodiment one

Fig. 1 is a kind of Video coding exercise intensity of preferred embodiment of the present invention quantization method flow chart, said method comprising the steps of:

S101 makes t=1.

T presentation code frame frame _tTime sequence number in current source sequence, being called for short t is coded frame frame _tThe time sequence number;

S102, judgement frame _tWhether be the I frame, if, to frame _tThen coding enters step S108; Otherwise enter step S103;

Frame _tBe called t frame coded frame, also referred to as current encoded frame;

S103 makes n=1;

S104 selects current coding macro block Mb _t,nCoding mode, calculate Mb _t,nExercise intensity and vague intensity.

S1041: calculate Mb _t,nExercise intensity, that is:

S1042: calculate Mb _t,nVague intensity, that is:

If Mb_intensity _t,n=max (width, height) * 4,

$\mathrm{Fuzzy}\_{\mathrm{intensity}}_{t,n}=\left\{\begin{array}{c}\frac{\underset{y_t(i,j)\∈{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }},{\mathrm{std}}_{t,\mathrm{uv}}^{\max }\≠0\\ \frac{\underset{y_t(i,j)\∈{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }+{10}^{-8}},\mathrm{else}\end{array}\right.$ , wherein,

${\mathrm{std}}_{t,\mathrm{uv}}^{\max }=\max (\underset{u_t(i,j)\∈{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(u_t(i,j)\right),\underset{v_t(i,j)\∈{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(v_t(i,j)\right))$

In the embodiment of the present invention, when block size is 16x16, be called macro block; The size that is when the macro block Further Division is n ₁xn ₂Piece (n ₁≤ 16 or n ₂≤ 16, but both can not equal 16 simultaneously), the smaller piece of these sizes is called sub-block;

Mb _t,nRepresent t frame coded frame frame _tN macro block (being current coding macro block); N represents macro block Mb _t,nAt frame _tIn position number; Mb_intensity _t,nBe Mb _t,nExercise intensity; Block _{T, n, m}Represent t frame coded frame frame _tN macro block Mb _t,nM sub-block; M represents sub-block block _{T, n, m}At Mb _t,nIn position number;

Expression macro block Mb _t,nMotion vector,

Expression sub-block block _{T, n, m}Motion vector, first representation in components motion vector of above-mentioned motion vector is at the component on the x direction of principal axis, the component of second representation in components motion vector on the y direction of principal axis; Abs represents to ask absolute value; Width, height represent respectively columns and the line number of a two field picture take whole pixel as unit;

Expression all variable maximizings to satisfying condition; Max (a, b) represents a, the b maximizing;

Fuzzy_intensity _t,nBe Mb _t,nVague intensity;

Expression is asked mean square deviation to all variablees that satisfy condition;

(y _t(i, j)) expression Mb _t,nThe brightness mean square deviation;

(u _t(i, j)) expression Mb _t,nU colourity mean square deviation;

(v _t(i, j)) expression Mb _t,nV colourity mean square deviation;

Expression Mb _t,nThe maximum of u colourity mean square deviation v colourity mean square deviation; y _t(i, j), u _t(i, j), v _t(i, j) represents respectively frame _tThe capable j row of i brightness value, frame _tThe capable j row of i colourity u component chromatic value, frame _tThe capable j row of i colourity v component chromatic value.

S105 is according to different coding structures and different frame categories, computing macro block Mb _t,nCorrection quantize

Parameter

, specific as follows:

Situation 1: when coding structure is IP ... P(is namely without the B frame) and frame _tFrame category when being the P frame

If

(Mb_intensity _t,nThres ₁And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or, (Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y _t(i, j))＜10)

${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{<figure-callout\; id="1"\; label="Thres"\; filenames="HDA00002998778400011.png"\; state="\{\{state\}\}">Thres</figure-callout>}}_1}{{\mathrm{const}}_2}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

In full, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{c}x,a\&le;x\&le;b\\ a,x<a\\ b,x>b\end{array}\right.;$ QP _minFor encoder is set the minimum quantization parameter;

QP _maxMaximum quantization parameter for the encoder setting; QP _t,nBe Mb _t,nInitial quantization parameters;

Be Mb _t,nCorrection quantize the parameter increase value; Be Mb _t,nThe correction quantization parameter; Thres ₁Decision threshold for situation 1; Const ₁, const ₂, be respectively quantification and adjust parameter 1, quantize to adjust parameter 2; General const ₁=10, const ₂=20, Thres ₁=25* (1+24/fps), the frame per second of fps presentation code film source; Int (x) expression rounds calculating.

Situation 2: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the P frame

If

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{Thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

Wherein, const ₃For quantizing to adjust parameter 3, general, const ₃=const ₂+ 8* (1+bframe), the B frame period that bframe presentation code device is set; Thres ₂Be the decision threshold of situation 2, general, Thres ₂=Thres ₁+ 25* (1+bframe).

Situation 3: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the B frame

If

(frame _tAs the reference frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode),

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{Thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, if

(frame _tConduct is not with reference to frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{<figure-callout\; id="1"\; label="Thres"\; filenames="HDA00002998778400011.png"\; state="\{\{state\}\}">Thres</figure-callout>}}_1}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

S106: according to current coding macro block Mb _t,nCoding mode and revise quantization parameter

, to macro block Mb _t,nEncode;

S107: if n 〉=width*height/256 enters S108; Otherwise n=n+1, reenter S104.

S108: if min is (FrameNum (frame _mm) | frame _mmDo not encode) do not exist, finish; Otherwise, enter step S109;

S109: if FrameNum is (frame _{m_i})=min (FrameNum (frame _mm) | frame _mmDo not encode),

, t=m_i reenters S102, otherwise, finish.

Wherein, frame _{m_i}Represent m_i frame coded frame; Frame _mmRepresent mm frame coded frame; M_i presentation code frame frame _{m_i}Time sequence number in current source sequence, being called for short m_i is coded frame frame _{m_i}The time sequence number; Mm presentation code frame frame _mmTime sequence number in current source sequence, being called for short mm is coded frame frame _mmThe time sequence number; FrameNum (frame _mm) expression frame _mmThe coding sequence number; FrameNum (frame _{m_i}) expression frame _{m_i}The coding sequence number; Min (variable | condition) represent all variablees that satisfy condition are minimized.

Embodiment two

Fig. 2 is a kind of Video coding exercise intensity of preferred embodiment of the present invention quantization device structure chart, and described device comprises: the first initialization module, frame category judging treatmenting module, the first coding module, the second initialization module, coding mode are selected module, exercise intensity and vague intensity computing module, are revised quantification parameter calculating module, the second coding module, the first judging treatmenting module, the second judging treatmenting module, the 3rd judging treatmenting module.

The first initialization module is used for initialization t=1; T presentation code frame frame _tTime sequence number in current source sequence, being called for short t is coded frame frame _tThe time sequence number;

The frame category judging treatmenting module is used for judgement frame _tWhether be the I frame, if enter the first coding module, otherwise enter the second initialization module;

The first coding module is used for frame _tEncode, enter the second judging treatmenting module;

The second initialization module is used for initialization and makes n=1; N represents macro block Mb _t,nAt frame _tIn position number;

Coding mode is selected module, is used for selecting current coding macro block Mb _t,nCoding mode;

Exercise intensity and vague intensity computing module are used for calculating Mb _t,nExercise intensity and vague intensity;

Revise the quantification parameter calculating module, be used for according to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

The second coding module is used for according to current coding macro block Mb _t,nCoding mode and revise quantization parameter

, to macro block Mb _t,nEncode;

The first judging treatmenting module if be used for judgement n 〉=width*height/256, enters the second judging treatmenting module, otherwise, make n=n+1, reenter coding mode and select module; Width, height represent respectively columns and the line number of a two field picture take whole pixel as unit;

The second judging treatmenting module is if be used for judgement min (FrameNum (frame _mm) | frame _mmDo not encode) do not exist, finish, otherwise, enter the 3rd judging treatmenting module;

The 3rd judging treatmenting module is used for judgement

If FrameNum is (frame _{m_i})=min (FrameNum (frame _mm) | frame _mmDo not encode),

, make t=m_i, reenter the frame category judging treatmenting module, otherwise, finish.

Wherein, frame _{m_i}Represent m_i frame coded frame; Frame _mmRepresent mm frame coded frame; M_i presentation code frame frame _{m_i}Time sequence number in current source sequence, being called for short m_i is coded frame frame _{m_i}The time sequence number; Mm presentation code frame frame _mmTime sequence number in current source sequence, being called for short mm is coded frame frame _mmThe time sequence number; FrameNum (frame _mm) expression frame _mmThe coding sequence number; FrameNum (frame _{m_i}) expression frame _{m_i}The coding sequence number; Min (variable | condition) represent all variablees that satisfy condition are minimized.

Described " calculating Mb _t,nExercise intensity and vague intensity " be specially:

Calculate Mb _t,nExercise intensity, that is:

Calculate Mb _t,nVague intensity, that is:

If Mb_intensity _t,n=max (width, height) * 4,

$\mathrm{Fuzzy}\_{\mathrm{intensity}}_{t,n}=\left\{\begin{array}{c}\frac{\underset{y_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }},{\mathrm{std}}_{t,\mathrm{uv}}^{\max }\&NotEqual;0\\ \frac{\underset{y_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }+{10}^{-8}},\mathrm{else}\end{array}\right.$ , wherein,

${\mathrm{std}}_{t,\mathrm{uv}}^{\max }=\max (\underset{u_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(u_t(i,j)\right),\underset{v_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(v_t(i,j)\right))$

In the embodiment of the present invention, when block size is 16x16, be called macro block; The size that is when the macro block Further Division is n ₁xn ₂Piece (n ₁≤ 16 or n ₂≤ 16, but both can not equal 16 simultaneously), the smaller piece of these sizes is called sub-block;

Mb _t,nRepresent t frame coded frame frame _tN macro block (being current coding macro block); N represents macro block Mb _t,nAt frame _tIn position number; Mb_intensity _t,nBe Mb _t,nExercise intensity; Block _{T, n, m}Represent t frame coded frame frame _tN macro block Mb _t,nM sub-block; M represents sub-block block _{T, n, m}At Mb _t,nIn position number;

Expression macro block Mb _t,nMotion vector,

Expression sub-block block _{T, n, m}Motion vector, first representation in components motion vector of above-mentioned motion vector is at the component on the x direction of principal axis, the component of second representation in components motion vector on the y direction of principal axis; Abs represents to ask absolute value; Width, height represent respectively columns and the line number of a two field picture take whole pixel as unit; Expression all variable maximizings to satisfying condition; Max (a, b) represents a, the b maximizing;

Fuzzy_intensity _t,nBe Mb _t,nVague intensity;

Expression is asked mean square deviation to all variablees that satisfy condition;

(y _t(i, j)) expression Mb _t,nThe brightness mean square deviation;

(u _t(i, j)) expression Mb _t,nU colourity mean square deviation; (v _t(i, j)) expression Mb _t,nV colourity mean square deviation;

Expression Mb _t,nThe maximum of u colourity mean square deviation v colourity mean square deviation; y _t(i, j), u _t(i, j), v _t(i, j) represents respectively frame _tThe capable j row of i brightness value, frame _tThe capable j row of i colourity u component chromatic value, frame _tThe capable j row of i colourity v component chromatic value.

Described " according to different coding structures and different frame categories, computing macro block Mb _t,nCorrection quantize

Parameter

N ", specific as follows:

Situation 1: when coding structure is IP ... P(is namely without the B frame) and frame _tIf frame category when being the P frame

(Mb_intensity _t,nThres ₁And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or,

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y _t(i, j))＜10)

${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{<figure-callout\; id="1"\; label="Thres"\; filenames="HDA00002998778400011.png"\; state="\{\{state\}\}">Thres</figure-callout>}}_1}{{\mathrm{const}}_2}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

In full, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{c}x,a\&le;x\&le;b\\ a,x<a\\ b,x>b\end{array}\right.;$ QP _minFor encoder is set the minimum quantization parameter;

QP _maxMaximum quantization parameter for the encoder setting; QP _t,nBe Mb _t,nInitial quantization parameters;

Be Mb _t,nCorrection quantize the parameter increase value; Be Mb _t,nThe correction quantization parameter; Thres ₁Decision threshold for situation 1; Const ₁, const ₂, be respectively quantification and adjust parameter 1, quantize to adjust parameter 2; General const ₁=10, const ₂=20, Thres ₁=25* (1+24/fps), the frame per second of fps presentation code film source; Int (x) expression rounds calculating.

Situation 2: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the P frame

If

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{Thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

Wherein, const ₃For quantizing to adjust parameter 3, general, const ₃=const ₂+ 8* (1+bframe), the B frame period that bframe presentation code device is set; Thres ₂Be the decision threshold of situation 2, general, Thres ₂=Thres ₁+ 25* (1+bframe).

Situation 3: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the B frame

If

(frame _tAs the reference frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode),

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{Thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, if

(frame _tConduct is not with reference to frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{<figure-callout\; id="1"\; label="Thres"\; filenames="HDA00002998778400011.png"\; state="\{\{state\}\}">Thres</figure-callout>}}_1}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

Those having ordinary skill in the art will appreciate that, realize that all or part of step in above-described embodiment method can complete by the program command related hardware, described program can be stored in a computer read/write memory medium, and described storage medium can be ROM, RAM, disk, CD etc.

The above is only preferred embodiment of the present invention, not in order to limiting the present invention, all any modifications of doing within the spirit and principles in the present invention, is equal to and replaces and improvement etc., within all should being included in protection scope of the present invention.

Claims (10)

1. a Video coding exercise intensity quantization method, is characterized in that, described method comprises:

Judgement frame _tWhether be the I frame, if not,

Make n=1;

Select current coding macro block Mb _t,nCoding mode;

Calculate current coding macro block Mb _t,nExercise intensity and vague intensity;

According to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

According to current coding macro block Mb _t,nCoding mode and revise quantization parameter

To macro block Mb _t,nEncode;

Judge whether n 〉=width*height/256 sets up, make if not n=n+1, return to and enter step " selection current coding macro block Mb _t,nCoding mode ", if,

Judgement min (FrameNum (frame _mm) | frame _mmDo not encode) whether exist, finish if do not exist, otherwise

Judgement FrameNum (frame _{m_i})=min (FrameNum (frame _mm) | frame _mmDo not encode) whether set up, if, make t=m_i, reenter step " judgement frame _tWhether be the I frame ", finish if not;

Wherein, frame _tIt is t frame coded frame; T presentation code frame frame _tTime sequence number in current source sequence; N represents macro block Mb _t,nAt frame _tIn position number; Mb _t,nRepresent t frame coded frame frame _tN macro block: width, height represents respectively columns and the line number of a two field picture take whole pixel as unit;

Frame _{m_i}Represent m_i frame coded frame; Frame _mmRepresent mm frame coded frame; M_i presentation code frame frame _{m_i}Time sequence number in current source sequence; Mm presentation code frame frame _mmTime sequence number in current source sequence; FrameNum (frame _mm) expression frame _mmThe coding sequence number; FrameNum (frame _{m_i}) expression frame _{m_i}The coding sequence number; Min (variable | condition) represent all variablees that satisfy condition are minimized.

2. Video coding exercise intensity quantization method as claimed in claim 1, is characterized in that described step " judgement frame _tWhether be the I frame " also comprise before step

Make t=1.

3. Video coding exercise intensity quantization method as claimed in claim 1, is characterized in that described step " judgement frame _tWhether be the I frame ", if, to frame _tThen coding enters step " judgement min (FrameNum (frame _mm)| frame _mmDo not encode) whether exist ".

4. Video coding exercise intensity quantization method as claimed in claim 1, is characterized in that, calculates current coding macro block Mb _t,nExercise intensity be specially:

Wherein, Mb_intensity _t,nBe Mb _t,nExercise intensity; Block _{T, n, m}Represent t frame coded frame frame _tN macro block Mb _t,nM sub-block; M represents sub-block block _{T, n, m}At Mb _t,nIn position number;

Expression macro block Mb _t,nMotion vector,

Expression sub-block block _{T, n, m}Motion vector, first representation in components motion vector of above-mentioned motion vector is at the component on the x direction of principal axis, the component of second representation in components motion vector on the y direction of principal axis; Abs represents to ask absolute value;

Expression all variable maximizings to satisfying condition; Max (a, b) represents a, the b maximizing.

5. Video coding exercise intensity quantization method as claimed in claim 1, is characterized in that, calculates current coding macro block Mb _t,nVague intensity be specially:

If Mb_intensity _t,n=max (width, height) * 4,

$\mathrm{Fuzzy}\_{\mathrm{intensity}}_{t,n}=\left\{\begin{array}{c}\frac{\underset{y_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }},{\mathrm{std}}_{t,\mathrm{uv}}^{\max }\&NotEqual;0\\ \frac{\underset{y_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }+{10}^{-8}},\mathrm{else}\end{array}\right.$ , wherein,

${\mathrm{std}}_{t,\mathrm{uv}}^{\max }=\max (\underset{u_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(u_t(i,j)\right),\underset{v_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(v_t(i,j)\right))$

Fuzzy_intensity _t,nBe Mb _t,nVague intensity;

Expression is asked mean square deviation to all variablees that satisfy condition;

(y _t(i, j)) expression Mb _t,nThe brightness mean square deviation;

(u _t(i, j)) expression Mb _t,nU colourity mean square deviation;

(v _t(i, j)) expression Mb _t,nV colourity mean square deviation; Expression Mb _t,nThe maximum of u colourity mean square deviation v colourity mean square deviation; y _t(i, j), u _t(i, j), v _t(i, j) represents respectively frame _tThe capable j row of i brightness value, frame _tThe capable j row of i colourity u component chromatic value, frame _tThe capable j row of i colourity v component chromatic value.

6. Video coding exercise intensity quantization method as claimed in claim 1, is characterized in that, according to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

, specific as follows:

Situation 1: when coding structure is IP ... P(is namely without the B frame) and frame _tIf frame category when being the P frame

(Mb_intensity _t,nThres ₁And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or, (Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and (y _t(i, j))＜10)

${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{Thres}}_1}{{\mathrm{const}}_2}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n},$

Wherein, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{c}x,a\&le;x\&le;b\\ a,x<a\\ b,x>b\end{array}\right.;$ QP _minFor encoder is set the minimum quantization parameter; QP _maxMaximum quantization parameter for the encoder setting; QP _t,nBe Mb _t,nInitial quantization parameters;

Be Mb _t,nCorrection quantize the parameter increase value;

Be Mb _t,nThe correction quantization parameter; Thres ₁Decision threshold for situation 1; Const ₁, const ₂, be respectively quantification and adjust parameter 1, quantize to adjust parameter 2; General const ₁=10, const ₂=20, Thres ₁=25* (1+24/fps), the frame per second of fps presentation code film source; Int (x) expression rounds calculating;

Situation 2: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the P frame

If

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{\mathrm{Mb}\_\mathrm{intensit}{y}_{t,n}-{\mathrm{Thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n},$

Wherein, const ₃For quantizing to adjust parameter 3, const ₃=const ₂+ 8* (1+bframe), the B frame period that bframe presentation code device is set; Thres ₂Be the decision threshold of situation 2, Thres ₂=Thres ₁+ 25* (1+bframe);

Situation 3: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the B frame

If

(frame _tAs the reference frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode),

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{\mathrm{Mb}\_\mathrm{int}{\mathrm{ensity}}_{t,n}-{\mathrm{thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, if

(frame _tConduct is not with reference to frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{\mathrm{Mb}\_\mathrm{intensit}{y}_{t,n}-{\mathrm{Thres}}_1}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

7. a Video coding exercise intensity quantization device, is characterized in that, described device comprises:

The first initialization module, frame category judging treatmenting module, the first coding module, the second initialization module, coding mode are selected module, exercise intensity and vague intensity computing module, are revised quantification parameter calculating module, the second coding module, the first judging treatmenting module, the second judging treatmenting module, the 3rd judging treatmenting module

The first initialization module is used for initialization t=1; T presentation code frame frame _tTime sequence number in current source sequence;

The frame category judging treatmenting module is used for judgement frame _tWhether be the I frame, if enter the first coding module, otherwise enter the second initialization module;

The first coding module is used for frame _tEncode, enter the second judging treatmenting module;

The second initialization module is used for initialization and makes n=1; N represents macro block Mb _t,nAt frame _tIn position number;

Coding mode is selected module, is used for selecting current coding macro block Mb _t,nCoding mode;

Exercise intensity and vague intensity computing module are used for calculating Mb _t,nExercise intensity and vague intensity;

Revise the quantification parameter calculating module, be used for according to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

The second coding module is used for according to current coding macro block Mb _t,nCoding mode and revise quantization parameter , to macro block Mb _t,nEncode;

The first judging treatmenting module if be used for judgement n 〉=width*height/256, enters the second judging treatmenting module, otherwise, make n=n+1, reenter coding mode and select module; Width, height represent respectively columns and the line number of a two field picture take whole pixel as unit;

The second judging treatmenting module is if be used for judgement min (FrameNum (frame _mm) | frame _mmDo not encode) do not exist, finish, otherwise, enter the 3rd judging treatmenting module;

The 3rd judging treatmenting module is used for judgement

If FrameNum is (frame _{m_i})=min (FrameNum (frame _mm) | frame _mmDo not encode),

, make t=m_i, reenter the frame category judging treatmenting module, otherwise, finish;

Wherein, frame _{m_i}Represent m_i frame coded frame; Frame _mmRepresent mm frame coded frame; M_i presentation code frame frame _{m_i}Time sequence number in current source sequence; Mm presentation code frame frame _mmTime sequence number in current source sequence; FrameNum (frame _mm) expression frame _mmThe coding sequence number; FrameNum (frame _{m_i}) expression frame _{m_i}The coding sequence number; Min (variable | condition) represent all variablees that satisfy condition are minimized.

8. Video coding exercise intensity quantization device as claimed in claim 7, is characterized in that, in " exercise intensity and vague intensity computing module ", calculates Mb _t,nExercise intensity be specially:

Calculate Mb _t,nExercise intensity, namely

Mb_intensity _t,nBe Mb _t,nExercise intensity; Block _{T, n, m}Represent t frame coded frame frame _tN macro block Mb _t,nM sub-block; M represents sub-block block _{T, n, m}At Mb _t,nIn position number;

Expression macro block Mb _t,nMotion vector,

Expression sub-block block _{T, n, m}Motion vector, first representation in components motion vector of above-mentioned motion vector is at the component on the x direction of principal axis, the component of second representation in components motion vector on the y direction of principal axis; Abs represents to ask absolute value; Width, height represent respectively columns and the line number of a two field picture take whole pixel as unit;

Expression all variable maximizings to satisfying condition; Max (a, b) represents a, the b maximizing.

9. Video coding exercise intensity quantization device as claimed in claim 7, is characterized in that, in " exercise intensity and vague intensity computing module ", calculates Mb _t,nVague intensity is specially:

Calculate Mb _t,nVague intensity, namely

If Mb_intensity _t,n=max (width, height) * 4,

$\mathrm{Fuzzy}\_{\mathrm{intensity}}_{t,n}=\left\{\begin{array}{c}\frac{\underset{y_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }},{\mathrm{std}}_{t,\mathrm{uv}}^{\max }\&NotEqual;0\\ \frac{\underset{y_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(y_t(i,j)\right)}{{\mathrm{std}}_{t,\mathrm{uv}}^{\max }+{10}^{-8}},\mathrm{else}\end{array}\right.$ , wherein,

${\mathrm{std}}_{t,\mathrm{uv}}^{\max }=\max (\underset{u_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(u_t(i,j)\right),\underset{v_t(i,j)\&Element;{\mathrm{Mb}}_{t,n}}{\mathrm{std}}\left(v_t(i,j)\right))$

Fuzzy_intensity _t,nBe Mb _t,nVague intensity;

Expression is asked mean square deviation to all variablees that satisfy condition;

(y _t(i, j)) expression Mb _t,nThe brightness mean square deviation;

(u _t(i, j)) expression Mb _t,nU colourity mean square deviation;

(v _t(i, j)) expression Mb _t,nV colourity mean square deviation; Expression Mb _t,nThe maximum of u colourity mean square deviation v colourity mean square deviation; y _t(i, j), u _t(i, j), v _t(i, j) represents respectively frame _tThe capable j row of i brightness value, frame _tThe capable j row of i colourity u component chromatic value, frame _tThe capable j row of i colourity v component chromatic value.

10. Video coding exercise intensity quantization device as claimed in claim 7, is characterized in that, " correction quantification parameter calculating module " in, described " according to different coding structures and different frame categories, computing macro block Mb _t,nThe correction quantization parameter

" be specially:

Situation 1: when coding structure is IP ... P(is namely without the B frame) and frame _tIf frame category when being the P frame

(Mb_intensity _t,nThres ₁And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or, (Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y _t(i, j))＜10)

${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{{\mathrm{Mb}\_\mathrm{intensity}}_{t,n}-{\mathrm{Thres}}_1}{{\mathrm{const}}_2}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n};$

Wherein, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{c}x,a\&le;x\&le;b\\ a,x<a\\ b,x>b\end{array}\right.;$ QP _minFor encoder is set the minimum quantization parameter; QP _maxMaximum quantization parameter for the encoder setting; QP _t,nBe Mb _t,nInitial quantization parameters;

Be Mb _t,nCorrection quantize the parameter increase value; Be Mb _t,nThe correction quantization parameter; Thres ₁Decision threshold for situation 1; Const ₁, const ₂, be respectively quantification and adjust parameter 1, quantize to adjust parameter 2; General const ₁=10, const ₂=20, Thres ₁=25* (1+24/fps), the frame per second of fps presentation code film source; Int (x) expression rounds calculating;

Situation 2: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the P frame

If

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{\mathrm{Mb}\_\mathrm{intensit}{y}_{t,n}-{\mathrm{Thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

Wherein, const ₃For quantizing to adjust parameter 3, const ₃=const ₂+ 8* (1+bframe), the B frame period that bframe presentation code device is set; Thres ₂Be the decision threshold of situation 2, Thres ₂=Thres ₁+ 25* (1+bframe);

Situation 3: when coding structure is IB ... BPB ... BP(namely contains the B frame) and frame _tWhen frame category is the B frame

If

(frame _tAs the reference frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _t,n≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode),

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{\mathrm{Mb}\_\mathrm{int}{\mathrm{ensity}}_{t,n}-{\mathrm{thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, if

(frame _tConduct is not with reference to frame)

And

(

(Mb_intensity _t,nThres ₂And Mb_intensity _{T, n1}≠ max (width, height) * 4 and Mb _t,nBe inter-frame forecast mode)

Or

(Mb_intensity _t,n=max (width, height) * 4 and Fuzzy_intensity _t,n＜2.5 and

(y (i, j))＜10)

)

, ${\mathrm{\&Delta;QP}}_{t,n}^m=\mathrm{int}\left(\frac{{\mathrm{QP}}_{\max }-{\mathrm{QP}}_{t,n}}{{\mathrm{const}}_1}\right)*\mathrm{int}\left(\frac{\mathrm{Mb}\_\mathrm{int}{\mathrm{ensity}}_{t,n}-{\mathrm{thres}}_2}{{\mathrm{const}}_3}\right)$

${\mathrm{QP}}_{t,n}^m=\mathrm{clip}({\mathrm{QP}}_{t,n}+{\mathrm{\&Delta;QP}}_{t,n}^m,{\mathrm{QP}}_{\min },{\mathrm{QP}}_{\max });$

Otherwise, ${\mathrm{QP}}_{t,n}^m={\mathrm{QP}}_{t,n}.$

Images