CN103327329A - Method and system for quantitative adjustment of image quality jump - Google Patents

Abstract

The invention discloses a method and system for quantitative adjustment of image quality jump. According to the method and system for quantitative adjustment of the image quality jump, on one hand, a method for reducing a difference between a quantization parameter of a preset I frame and a quantization parameter of a before P frame in an image which possibly produces the image quality jump according to the phenomena that the problem of the image quality jump is only caused by the situations that the difference between the quantization parameter of the preset I frame and the quantization parameter of the before P frame is large, and the preset I frame and the before P frame belong to the same scene, thus, a flicker feeling produced due to the image quality jump is reduced, and the subjective effect of an encoder is improved; on the other hand, when the preset I frame and the before P frame belong to different scenes, the image cannot produce the flicker phenomenon even when the quantization parameter difference is large, and the preset I frame can serve as a starting reference frame of a new scene. According to the method and system for quantitative adjustment of the image quality jump, due to the fact that movement intensity is used for adjusting the quantization parameters, the encoder can further adjust the quantization parameters according to the complexity of an encoding film source, and optimization of the encoding objective performance can be achieved.

Description

A kind of quantification method of adjustment and system of picture quality saltus step

Technical field

The present invention relates to the coding and decoding video field, relate in particular to a kind of quantification method of adjustment and system of picture quality saltus step.

Background technology

Prior art when last two field picture is poor quality P frame, is right after when a frame is the I frame of the fine Same Scene of picture quality thereafter in video sequence, can produce saltus step, the flickering of picture quality.To many scenes of complicated movement switching sequence, when adopting little I frame period to carry out Low Bit-rate Coding, the problems referred to above highlight maximization.And when front and back two frame image quality differed greatly but belong to different scene, flickering then can be alleviated.Conventional quantization strategy is not made special design to this, be the unified upper GOP(Group of Pictures of employing) in the average quantisation parameter of P frame determine the quantization parameter (QP) of I frame among the current GOP.

Summary of the invention

The purpose of the embodiment of the invention is to propose a kind of quantification method of adjustment of picture quality saltus step, is intended to solve prior art when the picture quality saltus step occurs, and can produce saltus step, the flickering problem of picture quality.

The embodiment of the invention is achieved in that a kind of quantification method of adjustment of picture quality saltus step, and described method comprises:

Step1: judge whether that current encoded frame is that P frame and next coded frame are the I frame, if then enter Step2;

Step2: judge that whether occurrence scene switches next coded frame with respect to current encoded frame, if not, then makes scene switching mark variable flag=0, if then make flag=1;

Step3: the first correction value (QP that calculates the current encoded frame quantization parameter _t ^M1);

Step4: utilize the first correction value of described quantization parameter that current encoded frame is encoded;

Step5: next coded frame is set to current encoded frame;

Step6: judge whether flag=0, if then enter Step7;

Step7: the second correction value (QP that calculates the current encoded frame quantization parameter _t ^M2), then utilize the second correction value of quantization parameter that current encoded frame is encoded, enter Step8;

Step8: judge whether next coded frame frame exists, if there is no, then finish; Otherwise next coded frame is set to current encoded frame, then reenters Step1.

Further, among the described step Step1, judge whether that current encoded frame is that P frame and next coded frame are the I frame, if not, then: at first, current encoded frame is encoded;

Then, judge whether next coded frame exists, if there is no, then finish, otherwise,

Make that current encoded frame is next coded frame, reenter again Step1.

Further, among the described step Step6, judge whether flag=0, if not, then enter Step9,

Step9: judge that whether next coded frame frame exists, if do not exist, then enters step Step10;

Step10: the 3rd correction value (QP that calculates the current encoded frame quantization parameter _t ^M3), then utilize the 3rd correction value of quantization parameter that current encoded frame is encoded, then finish.

4, the quantification method of adjustment of picture quality saltus step as claimed in claim 3 is characterized in that, among the described step Step9, judges whether next coded frame frame exists, and then enters Step11 if exist;

Step11; Utilize current encoded frame and next coded frame to calculate the 4th correction value of current encoded frame quantization parameter;

Step12: utilize the 4th correction value of quantization parameter that current encoded frame is encoded, then next coded frame is set to current encoded frame, then reenters Step1.

Another purpose of the embodiment of the invention is to propose a kind of quantification Adjustment System of picture quality saltus step, described system comprises: the first judge module, the first coding processing module, the second judge module, the first correction value computing module, the second coding processing module, the 3rd judge module, the second correction value computing module, the 3rd coding processing module

The first judge module is used for judging whether that current encoded frame is that P frame and next coded frame are the I frame, then enters if not the first coding processing module, if then enter the second judge module;

The first coding processing module is used for current encoded frame is encoded; Then, judge whether next coded frame exists, if there is no, then finish; Otherwise, make that current encoded frame is next coded frame, reenter again the first judge module;

The second judge module is used for judging whether occurrence scene switches next coded frame with respect to current encoded frame, and if not, scene switching mark variable flag=0 then is if then flag=1 enters the first correction value computing module;

The first correction value computing module is for the first correction value of calculating the current encoded frame quantization parameter;

The second coding processing module is used for utilizing the first correction value of described quantization parameter that current encoded frame is encoded, and then next coded frame is set to current encoded frame;

The 3rd judge module is used for judging whether scene switching mark variable flag=0, if then enter the second correction value computing module;

The second correction value computing module is for the second correction value (QP that calculates the current encoded frame quantization parameter _t ^M2), enter the 3rd coding processing module;

The 3rd coding processing module is used for utilizing the second correction value of quantization parameter that current encoded frame is encoded, and judges then whether next coded frame frame exists, and if there is no, then finishes; Otherwise next coded frame is set to current encoded frame, then reenters the first judge module.

Further, described system also comprises: the 4th judge module, the 3rd correction value computing module, the 4th coding processing module,

The 3rd judge module is used for judging whether scene switching mark variable flag=0, then enters if not the 4th judge module;

The 4th judge module is used for judging whether next coded frame frame exists, if do not exist, then enters the 3rd correction value computing module;

The 3rd correction value computing module for three correction values of calculating the current encoded frame quantization parameter, enters the 4th coding processing module;

The 4th coding processing module is used for utilizing three correction values of quantization parameter that current encoded frame is encoded, and then finishes.

Further, described system also comprises: the 4th correction value computing module, the 5th coding processing module,

The 4th judge module is used for judging whether next coded frame frame exists, and then enters the 4th correction value computing module if exist,

The 4th correction value computing module is used for utilizing current encoded frame and next coded frame to calculate the 4th correction value of current encoded frame quantization parameter, enters the 5th coding processing module;

The 5th coding processing module is used for utilizing the 4th correction value of quantization parameter that current encoded frame is encoded, and then next coded frame is set to current encoded frame, then reenters the first judge module.

Beneficial effect of the present invention

In order to address the above problem, the embodiment of the invention proposes a kind of quantification method of adjustment and system of picture quality saltus step.Embodiment of the invention method on the one hand, only result from default I frame quantization parameter and last P frame quantization parameter differs greatly and both belong to the phenomenon of a scene together according to picture quality saltus step problem, designed at the image that might produce the picture quality saltus step, reduce both poor methods of quantization parameter, thereby reduce the flickering that produces owing to the picture quality saltus step, promote the subjective effect of encoder.On the other hand, when default I frame and last P frame did not belong to a scene together, image can not produce scintillation yet even quantization parameter differs greatly, and then default I frame will be as the initial reference frame of new scene.Embodiment of the invention method utilizes exercise intensity to adjust quantization parameter, makes encoder further adjust quantization parameter according to the complexity of coding film source, thereby reaches the optimization of the objective performance of coding.

Description of drawings

Fig. 1 is the quantification method of adjustment flow chart of a kind of picture quality saltus step of the preferred embodiment of the present invention;

Fig. 2 is the method flow diagram that step Step11 comprises in Fig. 1 flow chart;

Fig. 3 is the quantification Adjustment System structure chart of a kind of picture quality saltus step of the preferred embodiment of the present invention;

Fig. 4 is the 4th correction value computing module modular structure figure in Fig. 3 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 with the embodiment of the 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.

Embodiment of the invention method on the one hand, only result from default I frame quantization parameter and last P frame quantization parameter differs greatly and both belong to the phenomenon of a scene together according to picture quality saltus step problem, designed at the image that might produce the picture quality saltus step, reduce both poor methods of quantization parameter, thereby reduce the flickering that produces owing to the picture quality saltus step, promote the subjective effect of encoder.On the other hand, when default I frame and last P frame did not belong to a scene together, image can not produce scintillation yet even quantization parameter differs greatly, and then default I frame will be as the initial reference frame of new scene.Embodiment of the invention method utilizes exercise intensity to adjust quantization parameter, makes encoder further adjust quantization parameter according to the complexity of coding film source, thereby reaches the optimization of the objective performance of coding.

Embodiment one

Fig. 1 is the quantification method of adjustment flow chart of a kind of picture quality saltus step of the preferred embodiment of the present invention; Said method comprising the steps of:

Step1: be the I frame if current encoded frame is P frame and next coded frame, then enter Step2; Otherwise, { at first, current encoded frame is encoded; Then, judge whether next coded frame exists, if there is no, then finish; Otherwise, make that current encoded frame is next coded frame, reenter again Step1.}

Step2: judge next coded frame with respect to current encoded frame whether occurrence scene switch (judging namely whether next coded frame is the scene switch frame) if not, flag=0 then, if, flag=1 then.

In full, flag is scene switching mark variable; The scene change detection method can adopt disclosed any method in the industry.

Step3: the first correction value of calculating the current encoded frame quantization parameter.

QP _t ^m1=clip(QP _t+temp，QP _min，QP _max)，

Wherein temp is intermediate variable, and its computational methods are as follows:

If flag=0, then

Otherwise, then $\mathrm{temp}=\left\{\begin{array}{ccc}Q{P}_{t-1}+\mathrm{Thres}1-Q{P}_t& ,& Q{P}_t-Q{P}_{t-1}<\mathrm{<figure-callout\; id="1"\; label="Thres"\; filenames="HDA00003330772900011.png"\; state="\{\{state\}\}">Thres</figure-callout>}1\\ 0& ,& \mathrm{else}\end{array}\right.$

In full, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{ccc}x& ,& a\&le;x\&le;b\\ a& ,& x<a\\ b& ,& x>b\end{array}\right.;$ QP _MinThe presentation code device is set the minimum quantization parameter; QP _MaxThe maximum quantization parameter that the presentation code device is set; Thres1 represents the first decision threshold, general Thres1 〉=3; Thres2 represents the second decision threshold, general Thres2〉38; QP _tThe quantization parameter of expression current encoded frame; QP _t ^M1The first correction value of expression present frame quantization parameter; QP _T-1The quantization parameter that represents previous coded frame; Frame _jRepresent j coded frame; QP _jRepresent j coded frame quantization parameter; GOP _tThe GOP at expression current encoded frame place is also referred to as current GOP;

Represent current GOP quantization parameter; Expression is averaged to the variable that satisfies condition; Weight ₁Represent the first weight factor; Weight ₂Represent the second weight factor, general 0＜weight ₁＜weight ₂≤ 0.5.

Step4: utilize the first correction value of described quantization parameter that current encoded frame is encoded.

Step5: next coded frame is set to current encoded frame.

Step6: if flag=0 then enters Step7; If not, then enter Step9.

Step7: the second correction value QP that calculates the current encoded frame quantization parameter _t ^M2, then utilize the second correction value of quantization parameter that current encoded frame is encoded, enter Step8.

QP _t ^M2=clip (QP _t+ Thres1-2+temp, QP _Min, QP _Max), wherein temp is intermediate variable, has been calculated by Step3 and has obtained.

Step8: judge whether next coded frame frame exists, if there is no, then finish; Otherwise next coded frame is set to current encoded frame, then reenters Step1.

Step9: judge that whether next coded frame frame exists, if do not exist, then enters step Step10; Then enter Step11 if exist;

Step10: the 3rd correction value QP that calculates the current encoded frame quantization parameter _t ^M3, then utilize the 3rd correction value of quantization parameter that current encoded frame is encoded, then finish.

QP _t ^M3=clip (QP _t-temp, QP _Min, QP _Max), wherein temp is intermediate variable, has been calculated by Step3 and has obtained.

Step11: utilize current encoded frame and next coded frame to calculate the 4th correction value of current encoded frame quantization parameter.

(Fig. 2 is the method flow diagram that step Step11 comprises in Fig. 1 flow chart specifically to comprise following steps; ):

The first step is carried out down-sampling to the monochrome information frame of current encoded frame and is processed, and obtains current processed frame; Again the monochrome information frame of next coded frame carried out the down-sampling processing, obtain next processed frame (Downsapling method can adopt known nearest neighbor algorithm, bilinear interpolation, cubic convolution etc.).

Second step is divided into piece with current processed frame, more next processed frame also is divided into piece; And calculate the piece statistical variable of each piece.

$\mathrm{ti}\_\mathrm{bloc}{k}_{t,n}=\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}(i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$

Wherein, block size can determine as required, and is commonly used such as 16x16,8x8 etc.; When block size is 16x16, then be called macro block; When the macro block Further Division is that size is n ₁Xn ₂Piece then the smaller piece of these sizes be called sub-block (n ₁≤ 16 _Perhapsn ₂≤ 16, but both can not equal 16 simultaneously); N piece of current processed frame is designated as

N piece of next processed frame is designated as

With

The palpus in the same size;

Pixel value for the capable j row of current processed frame i;

Pixel value for the capable j row of next processed frame i; Ti_block _{T, n}Expression piece statistical variable;

Represent next processed frame; $\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$ It is right to represent

With

Carry out subtraction, wherein requirement

Belong to And

Belong to

Then all are asked for

Ask mean square deviation.

The 3rd step, calculate the frame variable, be designated as TI_frame _t

$\mathrm{TI}\_{\mathrm{frame}}_t=\underset{n=\mathrm{1,2},...,{\mathrm{number}}_{\mathrm{block}}}{\mathrm{sum}}\left(\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)\right),$ Wherein

$\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)=\left\{\begin{array}{ccc}1& ,& \mathrm{ti}\_{\mathrm{block}}_{t,n}>\mathrm{Thres}3\\ 0& ,& \mathrm{else}\end{array}\right.$

In full, Thres3 represents the 3rd decision threshold, and it is the decreasing function of fps, general desirable Thres3=8* (1+24/fps); The frame per second of fps presentation code film source; Expression all variable summations to satisfying condition; Number _BlockRepresent the piece sum that a two field picture comprises.

The 4th step, the 4th correction value QP of calculating current encoded frame quantization parameter _t ^M4

QP _t ^m4=clip(QP _t-clip(const-(TI_frame _t>>bit),ΔQP _min,ΔQP _max),QP _min,QP _max)

In full,〉〉 the right displacement computing of expression; Δ QP _MinThe expression quantization parameter is adjusted lower limit, and is general-10＜Δ QP _Min＜Δ QP _MaxΔ QP _MaxExpression expression quantization parameter is adjusted the upper limit, general 0＜Δ QP _Max＜10; Const represents quantization parameter adjustment constant, general const=(ceil (number _Block* weight ₃) bit)+1; Weight ₃Represent the 3rd weight factor, general 0.6≤weight ₃＜1; Ceil (x) expression is greater than the smallest positive integral of x; Bit represents the displacement constant, general 1＜bit≤log ₂Number _Block-1.

Step12: utilize the 4th correction value of quantization parameter that current encoded frame is encoded, then next coded frame is set to current encoded frame, then reenters Step1.

Embodiment two

Fig. 3 is the quantification Adjustment System structure chart of a kind of picture quality saltus step of the preferred embodiment of the present invention; Described system comprises: the first judge module, the first coding processing module, the second judge module, the first correction value computing module, the second coding processing module, the 3rd judge module, the second correction value computing module, the 3rd coding processing module,

The first judge module is used for judging whether that current encoded frame is that P frame and next coded frame are the I frame, then enters if not the first coding processing module, if then enter the second judge module;

The first coding processing module is used for current encoded frame is encoded; Then, judge whether next coded frame exists, if there is no, then finish; Otherwise, make that current encoded frame is next coded frame, reenter again the first judge module;

The second judge module is used for judging whether occurrence scene switches next coded frame with respect to current encoded frame, and if not, scene switching mark variable flag=0 then is if then flag=1 enters the first correction value computing module;

The first correction value computing module is for the first correction value of calculating the current encoded frame quantization parameter;

The second coding processing module is used for utilizing the first correction value of described quantization parameter that current encoded frame is encoded, and then next coded frame is set to current encoded frame;

The 3rd judge module is used for judging whether scene switching mark variable flag=0, if then enter the second correction value computing module;

The second correction value computing module is for the second correction value QP that calculates the current encoded frame quantization parameter _t ^M2, enter the 3rd coding processing module;

The 3rd coding processing module is used for utilizing the second correction value of quantization parameter that current encoded frame is encoded, and judges then whether next coded frame frame exists, and if there is no, then finishes; Otherwise next coded frame is set to current encoded frame, then reenters the first judge module.

Further, described system also comprises: the 4th judge module, the 3rd correction value computing module, the 4th coding processing module,

The 3rd judge module is used for judging whether scene switching mark variable flag=0, then enters if not the 4th judge module;

The 4th judge module is used for judging whether next coded frame frame exists, if do not exist, then enters the 3rd correction value computing module;

The 3rd correction value computing module for three correction values of calculating the current encoded frame quantization parameter, enters the 4th coding processing module;

The 4th coding processing module is used for utilizing three correction values of quantization parameter that current encoded frame is encoded, and then finishes.

Further, described system also comprises: the 4th correction value computing module, the 5th coding processing module,

The 4th judge module is used for judging whether next coded frame frame exists, and then enters the 4th correction value computing module if exist,

The 4th correction value computing module is used for utilizing current encoded frame and next coded frame to calculate the 4th correction value of current encoded frame quantization parameter, enters the 5th coding processing module;

The 5th coding processing module is used for utilizing the 4th correction value of quantization parameter that current encoded frame is encoded, and then next coded frame is set to current encoded frame, then reenters the first judge module.

Further, in the first correction value computing module, the first correction value of described calculating current encoded frame quantization parameter is specially:

QP _t ^m1=clip(QP _t+temp，QP _min，QP _max)，

Wherein temp is intermediate variable, and its computational methods are as follows:

If flag=0, then

Otherwise, then $\mathrm{temp}=\left\{\begin{array}{ccc}Q{P}_{t-1}+\mathrm{Thresl}-Q{P}_t& ,& Q{P}_t-Q{P}_{t-1}<\mathrm{Thresl}\\ 0& ,& \mathrm{else}\end{array}\right.$

In full, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{ccc}x& ,& a\&le;x\&le;b\\ a& ,& x<a\\ b& ,& x>b\end{array}\right.;$ QP _MinThe presentation code device is set the minimum quantization parameter; QP _MaxThe maximum quantization parameter that the presentation code device is set; Thres1 represents the first decision threshold, general Thres1 〉=3; Thres2 represents the second decision threshold, general Thres2〉38; QP _tThe quantization parameter of expression current encoded frame; QP _t ^M1The first correction value of expression present frame quantization parameter; QP _T-1The quantization parameter that represents previous coded frame; Frame _jRepresent j coded frame; QP _jRepresent j coded frame quantization parameter; GOP _tThe GOP at expression current encoded frame place is also referred to as current GOP; Represent current GOP quantization parameter;

Expression is averaged to the variable that satisfies condition; Weight ₁Represent the first weight factor; Weight ₂Represent the second weight factor, general 0＜weight ₁＜weight ₂≤ 0.5.

Further, in the second correction value computing module, the second correction value QP of described calculating current encoded frame quantization parameter _t ^M2Be specially: QP _t ^M2=clip (QP _t+ Thres1-2+temp, QP _Min, QP _Max);

Further, in the 3rd correction value computing module, calculate the 3rd correction value QP of current encoded frame quantization parameter _t ^M3Be specially: QP _t ^M3=clip (QP _t-temp, QP _Min, QP _Max);

Further, the 4th correction value computing module also comprises down-sampling processing module, piece statistical variable computing module, frame variable computing module, the 4th correction value calculating sub module (Fig. 4 is the 4th correction value computing module modular structure figure in Fig. 3 structure chart),

The down-sampling processing module is used for the monochrome information frame of current encoded frame is carried out the down-sampling processing, obtains current processed frame; Again the down-sampling processing is carried out in the monochrome information frame of next coded frame, obtain next processed frame;

Piece statistical variable computing module is used for current processed frame is divided into piece, more next processed frame also is divided into piece; And calculate the piece statistical variable of each piece.

Frame variable computing module is used for calculating the frame variable, is designated as TI_frame _t

The 4th correction value calculating sub module is for the 4th correction value of calculating the current encoded frame quantization parameter.

Further, in the piece statistical variable computing module, the piece statistical variable that calculates each piece is specially:

$\mathrm{ti}\_\mathrm{bloc}{k}_{t,n}=\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}(i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$

Wherein, block size can determine as required, and is commonly used such as 16x16,8x8 etc.; When block size is 16x16, then be called macro block; When the macro block Further Division is that size is n ₁Xn ₂Piece then the smaller piece of these sizes be called sub-block (n ₁≤ 16 _Perhapsn ₂≤ 16, but both can not equal 16 simultaneously); N piece of current processed frame is designated as

N piece of next processed frame is designated as With

The palpus in the same size;

Pixel value for the capable j row of current processed frame i;

Pixel value for the capable j row of next processed frame i; Ti_block _{T, n}Expression piece statistical variable;

Represent next processed frame;

$\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$ It is right to represent

With Carry out subtraction, wherein requirement

Belong to

And

Belong to

Then all are asked for

Ask mean square deviation.

Further, frame variable computing module, calculate frame variable TI_framet and be specially:

$\mathrm{TI}\_{\mathrm{frame}}_t=\underset{n=\mathrm{1,2},...,{\mathrm{number}}_{\mathrm{block}}}{\mathrm{sum}}\left(\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)\right),$ Wherein

$\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)=\left\{\begin{array}{ccc}1& ,& \mathrm{ti}\_{\mathrm{block}}_{t,n}>\mathrm{Thres}3\\ 0& ,& \mathrm{else}\end{array}\right.$

In full, Thres3 represents the 3rd decision threshold, and it is the decreasing function of fps, general desirable Thres3=8* (1+24/fps); The frame per second of fps presentation code film source;

Expression all variable summations to satisfying condition; Number _BlockRepresent the piece sum that a two field picture comprises.

Further, in the 4th correction value calculating sub module, calculate the 4th correction value of current encoded frame quantization parameter

Be specially: QP _t ^M4=clip (QP _t-clip (const-(TI_frame _tBit), Δ QP _Min, Δ QP _Max), QP _Min, QP _Max)

In full,〉〉 the right displacement computing of expression; Δ QP _MinThe expression quantization parameter is adjusted lower limit, and is general-10＜Δ QP _Min＜Δ QP _MaxΔ QP _MaxExpression expression quantization parameter is adjusted the upper limit, general 0＜Δ QP _Max＜10; Const represents quantization parameter adjustment constant, general const=(ceil (number _Block* weight ₃) bit)+1; Weight ₃Represent the 3rd weight factor, general 0.6≤weight ₃＜1; Ceil (x) expression is greater than the smallest positive integral of x; Bit represents the displacement constant, general 1＜bit≤log ₂Number _Block-1.

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

The above only is 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., all should be included within protection scope of the present invention.

Claims (14)

1. the quantification method of adjustment of a picture quality saltus step is characterized in that, described method comprises:

Step1: judge whether that current encoded frame is that P frame and next coded frame are the I frame, if then enter Step2;

Step2: judge that whether occurrence scene switches next coded frame with respect to current encoded frame, if not, then makes scene switching mark variable flag=0, if then make flag=1;

Step3: the first correction value (QP that calculates the current encoded frame quantization parameter _t ^M1);

Step4: utilize the first correction value of described quantization parameter that current encoded frame is encoded;

Step5: next coded frame is set to current encoded frame;

Step6: judge whether flag=0, if then enter Step7;

Step7: the second correction value (QP that calculates the current encoded frame quantization parameter _t ^M2), then utilize the second correction value of quantization parameter that current encoded frame is encoded, enter Step8;

Step8: judge whether next coded frame frame exists, if there is no, then finish; Otherwise next coded frame is set to current encoded frame, then reenters Step1.

2. the quantification method of adjustment of picture quality saltus step as claimed in claim 1 is characterized in that, among the described step Step1, judges whether that current encoded frame is that P frame and next coded frame are the I frame, if not, and then: at first, current encoded frame is encoded;

Then, judge whether next coded frame exists, if there is no, then finish, otherwise,

Make that current encoded frame is next coded frame, reenter again Step1.

3. the quantification method of adjustment of picture quality saltus step as claimed in claim 1 is characterized in that, among the described step Step6, judges whether flag=0, if not, then enters Step9,

Step9: judge that whether next coded frame frame exists, if do not exist, then enters step Step10;

Step10: the 3rd correction value (QP that calculates the current encoded frame quantization parameter _t ^M3), then utilize the 3rd correction value of quantization parameter that current encoded frame is encoded, then finish.

4. the quantification method of adjustment of picture quality saltus step as claimed in claim 3 is characterized in that, among the described step Step9, judges whether next coded frame frame exists, and then enters Step11 if exist;

Step11; Utilize current encoded frame and next coded frame to calculate the 4th correction value of current encoded frame quantization parameter;

Step12: utilize the 4th correction value of quantization parameter that current encoded frame is encoded, then next coded frame is set to current encoded frame, then reenters Step1.

5. the quantification method of adjustment of picture quality saltus step as claimed in claim 1 is characterized in that,

Described " calculating the first correction value of current encoded frame quantization parameter " is specially:

QP _t ^m1=clip(QP _t+temp，QP _min，QP _max)，

Wherein temp is intermediate variable, and its computational methods are as follows:

If flag=0, then

Otherwise, then $\mathrm{temp}=\begin{array}{cc}& \end{array}\left\{\begin{array}{ccc}{\mathrm{QP}}_{t-1}+\mathrm{Thres}1-{\mathrm{QP}}_t& ,& {\mathrm{QP}}_t-{\mathrm{QP}}_{t-1}<\mathrm{Thres}1\\ 0& ,& \mathrm{else}\end{array}\right.$

Wherein, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{ccc}x& ,& a\&le;x\&le;b\\ a& ,& x<a\\ b& ,& x>b\end{array}\right.;$ QP _MinThe presentation code device is set the minimum quantization parameter; QP _MaxThe maximum quantization parameter that the presentation code device is set; Thres1 represents the first decision threshold, Thres1 〉=3; Thres2 represents the second decision threshold, Thres2〉38; QP _tThe quantization parameter of expression current encoded frame; QP _t ^M1The first correction value of expression present frame quantization parameter; QP _T-1The quantization parameter that represents previous coded frame; Frame _jRepresent j coded frame; QP _jRepresent j coded frame quantization parameter; GOP _tThe GOP at expression current encoded frame place,

Represent current GOP quantization parameter;

Expression is averaged to the variable that satisfies condition; Weight ₁Represent the first weight factor; Weight ₂Represent the second weight factor, 0＜weight ₁＜weight ₂≤ 0.5;

Described " the second correction value QP of calculating current encoded frame quantization parameter _t ^M2" be specially

QP _t ^m2=clip(QP _t+Thres1-2+temp,QP _min,QP _max)。

6. the quantification method of adjustment of picture quality saltus step as claimed in claim 3 is characterized in that,

Described " the 3rd correction value (QP of calculating current encoded frame quantization parameter _t ^M3) " be specially:

QP _t ^m3=clip(QP _t-temp,QP _min,QP _max)。

7. the quantification method of adjustment of picture quality saltus step as claimed in claim 4 is characterized in that,

Described " calculating the 4th correction value of current encoded frame quantization parameter " comprising:

The first step is carried out down-sampling to the monochrome information frame of current encoded frame and is processed, and obtains current processed frame; Again the down-sampling processing is carried out in the monochrome information frame of next coded frame, obtain next processed frame;

Second step is divided into piece with current processed frame, more next processed frame also is divided into piece; And calculate the piece statistical variable of each piece,

$\mathrm{ti}\_\mathrm{bloc}{k}_{t,n}=\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}(i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$

Wherein, n piece with current processed frame is designated as

N piece of next processed frame is designated as

In the same size; f _t ^Deal(i, j) is the pixel value of the capable j row of current processed frame i;

Pixel value for the capable j row of next processed frame i; Ti_block _{T, n}Expression piece statistical variable;

Represent next processed frame;

$\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$ Expression is to f _t ^Deal(i, j) with

Carry out subtraction, wherein require f _t ^Deal(i, j) belongs to

And

Belong to

Then all are asked for

Ask mean square deviation;

In the 3rd step, calculate frame variable TI_frame _t,

$\mathrm{TI}\_{\mathrm{frame}}_t=\underset{n=\mathrm{1,2},...,{\mathrm{number}}_{\mathrm{block}}}{\mathrm{sum}}\left(\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)\right),$ Wherein

$\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)=\left\{\begin{array}{ccc}1& ,& \mathrm{ti}\_{\mathrm{block}}_{t,n}>\mathrm{Thres}3\\ 0& ,& \mathrm{else}\end{array}\right.$

Wherein, Thres3 represents the 3rd decision threshold, and it is the decreasing function of fps; Thres3=8* (1+24/fps); The frame per second of fps presentation code film source;

Expression all variable summations to satisfying condition; Number _BlockRepresent the piece sum that a two field picture comprises;

The 4th step, the 4th correction value QP of calculating current encoded frame quantization parameter _t ^M4,

QP _t ^m4=clip(QP _t-clip(const-(TI_frame _t>>bit),ΔQP _min,ΔQP _max),QP _min,QP _max)

Wherein,〉〉 the right displacement computing of expression; Δ QP _MinThe expression quantization parameter is adjusted lower limit, and is general-10＜Δ QP _Min＜Δ QP _MaxΔ QP _MaxExpression expression quantization parameter is adjusted the upper limit, 0＜Δ QP _Max＜10; Const represents quantization parameter adjustment constant, const=(ceil (number _Block* weight ₃) bit)+1; Weight ₃Represent the 3rd weight factor, 0.6≤weight ₃＜1; Ceil (x) expression is greater than the smallest positive integral of x; Bit represents the displacement constant, 1＜bit≤log ₂Number _Block-1.

8. the quantification Adjustment System of a picture quality saltus step, it is characterized in that, described system comprises: the first judge module, the first coding processing module, the second judge module, the first correction value computing module, the second coding processing module, the 3rd judge module, the second correction value computing module, the 3rd coding processing module

The first judge module is used for judging whether that current encoded frame is that P frame and next coded frame are the I frame, then enters if not the first coding processing module, if then enter the second judge module;

The first coding processing module is used for current encoded frame is encoded; Then, judge whether next coded frame exists, if there is no, then finish; Otherwise, make that current encoded frame is next coded frame, reenter again the first judge module;

The second judge module is used for judging whether occurrence scene switches next coded frame with respect to current encoded frame, and if not, scene switching mark variable flag=0 then is if then flag=1 enters the first correction value computing module;

The first correction value computing module is for the first correction value of calculating the current encoded frame quantization parameter;

The second coding processing module is used for utilizing the first correction value of described quantization parameter that current encoded frame is encoded, and then next coded frame is set to current encoded frame;

The 3rd judge module is used for judging whether scene switching mark variable flag=0, if then enter the second correction value computing module;

The second correction value computing module is for the second correction value (QP that calculates the current encoded frame quantization parameter _t ^M2), enter the 3rd coding processing module;

The 3rd coding processing module is used for utilizing the second correction value of quantization parameter that current encoded frame is encoded, and judges then whether next coded frame frame exists, and if there is no, then finishes; Otherwise next coded frame is set to current encoded frame, then reenters the first judge module.

9. the quantification Adjustment System of picture quality saltus step as claimed in claim 8 is characterized in that,

Described system also comprises: the 4th judge module, the 3rd correction value computing module, the 4th coding processing module,

The 3rd judge module is used for judging whether scene switching mark variable flag=0, then enters if not the 4th judge module;

The 4th judge module is used for judging whether next coded frame frame exists, if do not exist, then enters the 3rd correction value computing module;

The 3rd correction value computing module for three correction values of calculating the current encoded frame quantization parameter, enters the 4th coding processing module;

The 4th coding processing module is used for utilizing three correction values of quantization parameter that current encoded frame is encoded, and then finishes.

10. the quantification Adjustment System of picture quality saltus step as claimed in claim 9 is characterized in that, described system also comprises: the 4th correction value computing module, the 5th coding processing module,

The 4th judge module is used for judging whether next coded frame frame exists, and then enters the 4th correction value computing module if exist,

The 4th correction value computing module is used for utilizing current encoded frame and next coded frame to calculate the 4th correction value of current encoded frame quantization parameter, enters the 5th coding processing module;

The 5th coding processing module is used for utilizing the 4th correction value of quantization parameter that current encoded frame is encoded, and then next coded frame is set to current encoded frame, then reenters the first judge module.

11. the quantification Adjustment System of picture quality saltus step as claimed in claim 8 is characterized in that, in described the first correction value computing module, the first correction value of described calculating current encoded frame quantization parameter is specially:

QP _t ^m1=clip(QP _t+temp，QP _min，QP _max)，

Wherein temp is intermediate variable, and its computational methods are as follows:

If flag=0, then

Otherwise, then $\mathrm{temp}=\left\{\begin{array}{ccc}{\mathrm{QP}}_{t-1}+\mathrm{Thres}1-{\mathrm{QP}}_t& ,& {\mathrm{QP}}_t-{\mathrm{QP}}_{t-1}<\mathrm{Thres}1\\ 0& ,& \mathrm{else}\end{array}\right.$

Wherein, $\mathrm{clip}(x,a,b)=\left\{\begin{array}{ccc}x& ,& a\&le;x\&le;b\\ a& ,& x<a\\ b& ,& x>b\end{array}\right.;$ QP _MinThe presentation code device is set the minimum quantization parameter; QP _MaxThe maximum quantization parameter that the presentation code device is set; Thres1 represents the first decision threshold, Thres1 〉=3; Thres2 represents the second decision threshold, Thres2〉38; QP _tThe quantization parameter of expression current encoded frame; QP _t ^M1The first correction value of expression present frame quantization parameter; QP _T-1The quantization parameter that represents previous coded frame; Frame _jRepresent j coded frame; QP _jRepresent j coded frame quantization parameter; GOP _tThe GOP at expression current encoded frame place;

Represent current GOP quantization parameter; Expression is averaged to the variable that satisfies condition; Weight ₁Represent the first weight factor; Weight ₂Represent the second weight factor, 0＜weight ₁＜weight ₂≤ 0.5;

In the second correction value computing module, the second correction value QP of described calculating current encoded frame quantization parameter _t ^M2Be specially: QP _t ^M2=clip (QP _t+ Thres1-2+temp, QP _Min, QP _Max).

12. the quantification Adjustment System of picture quality saltus step as claimed in claim 9 is characterized in that,

In the 3rd correction value computing module, calculate the 3rd correction value of current encoded frame quantization parameter

Be specially: QP _t ^M3=clip (QP _t-temp, QP _Min, QP _Max).

13. the quantification Adjustment System of picture quality saltus step as claimed in claim 10 is characterized in that,

Further, the 4th correction value computing module also comprises down-sampling processing module, piece statistical variable computing module, frame variable computing module, the 4th correction value calculating sub module,

The down-sampling processing module is used for the monochrome information frame of current encoded frame is carried out the down-sampling processing, obtains current processed frame; Again the down-sampling processing is carried out in the monochrome information frame of next coded frame, obtain next processed frame;

Piece statistical variable computing module is used for current processed frame is divided into piece, more next processed frame also is divided into piece; And calculate the piece statistical variable of each piece;

Frame variable computing module is used for calculating frame variable (TI_frame _t);

The 4th correction value calculating sub module is for the 4th correction value of calculating the current encoded frame quantization parameter.

14. the quantification Adjustment System of picture quality saltus step as claimed in claim 13 is characterized in that,

In the piece statistical variable computing module, the piece statistical variable that calculates each piece is specially:

$\mathrm{ti}\_\mathrm{bloc}{k}_{t,n}=\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}(i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$

Wherein, n piece of current processed frame is designated as

N piece of next processed frame is designated as

With

In the same size;

Pixel value for the capable j row of current processed frame i;

Pixel value for the capable j row of next processed frame i; Ti_block _{T, n}Expression piece statistical variable;

Represent next processed frame;

$\mathrm{std}({f_t}^{\mathrm{deal}}(i,j)-f_{t+1}^{\mathrm{deal}}i,j)|{f_t}^{\mathrm{deal}}(i,j)\&Element;\mathrm{bloc}{k}_{t,n}^{\mathrm{deal}}\mathrm{and}{f}_{t+1}^{\mathrm{deal}}(i,j)\&Element;{\mathrm{block}}_{t+1,n}^{\mathrm{deal}})$ Expression is to f _t ^Deal(i, j) with Carry out subtraction, wherein require f _t ^Deal(i, j) belongs to

And

Belong to

Then all are asked for

Ask mean square deviation;

In the frame variable computing module, calculate frame variable TI_frame _tBe specially:

$\mathrm{TI}\_{\mathrm{frame}}_t=\underset{n=\mathrm{1,2},...,{\mathrm{number}}_{\mathrm{block}}}{\mathrm{sum}}\left(\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)\right),$ Wherein

$\mathrm{sign}(\mathrm{ti}\_{\mathrm{block}}_{t,n},\mathrm{Thres}3)=\left\{\begin{array}{ccc}1& ,& \mathrm{ti}\_{\mathrm{block}}_{t,n}>\mathrm{Thres}3\\ 0& ,& \mathrm{else}\end{array}\right.$

Wherein, Thres3 represents the 3rd decision threshold, and it is the decreasing function of fps, Thres3=8* (1+24/fps); The frame per second of fps presentation code film source;

Expression all variable summations to satisfying condition; Number _BlockRepresent the piece sum that a two field picture comprises;

In the 4th correction value calculating sub module, calculate the 4th correction value QP of current encoded frame quantization parameter _t ^M4Be specially:

QP _t ^m4=clip(QP _t-clip(const-(TI_frame _t>>bit),ΔQP _min,ΔQP _max),QP _min,QP _max)

Wherein,〉〉 the right displacement computing of expression; Δ QP _MinThe expression quantization parameter is adjusted lower limit ,-10＜Δ QP _Min＜Δ QP _MaxΔ QP _MaxExpression expression quantization parameter is adjusted the upper limit, 0＜Δ QP _Max＜10; Const represents quantization parameter adjustment constant, const=(ceil (number _Block* weight ₃) bit)+1; Weight ₃Represent the 3rd weight factor, 0.6≤weight ₃＜1; Ceil (x) expression is greater than the smallest positive integral of x; Bit represents the displacement constant, general 1＜bit≤log ₂Number _Block-1.

Images