CN102420980B - Frame-layer and macroblock-layer quantization parameter adjusting method and system - Google Patents

Abstract

The invention discloses a frame-layer and macroblock-layer quantization parameter adjusting method and system. On one hand, the adjustment range of frame-layer quantization parameters is determined through dynamic and static contrast information of an image; on the other hand, further adjustment is carried out through a current bit or QP (Quantization Parameter) conditions of RC (Rate Control) output, which can effectively control the actual bit rate to close a target bit rate. In addition, the invention further provides a novel macroblock-layer quantization parameter adjusting method, wherein quantization parameters of a current macroblock are adaptively adjusted by utilizing static and dynamic contrast information of the current macroblock and a current frame as well as the frame-layer quantization parameters, so as to keep the stability of the fluctuating performance of code rate while the rate-distortion performance of an encoder is improved.

Description

A kind of frame layer and macroblock layer quantization parameter method of adjustment and system

Technical field

The present invention relates to field of video encoding, relate in particular to a kind of frame layer and macroblock layer quantization parameter method of adjustment and system.

Background technology

One of key problem in Video coding is exactly RC (Rate Control), determines all kinds of coding modes by it, as macroblock partitions, motion vector, quantization parameter etc.Wherein, because quantification can cause distortion, make the structure of quantification and the selection of quantization parameter particularly important.Meanwhile, in Video coding, first coding frame reconstruct after using the reference frame as next code frame, thereby make the quality of next code frame and the quality of reference frame closely bound up.

Prior art generally adopts the frame layer quantization parameter adjustment modes of fixed cycle, because it does not consider the characteristic of image, thus cannot online adaptive adjustment, thus its performance restricted.

Summary of the invention

The object of the embodiment of the present invention is to propose a kind of frame layer quantization parameter method of adjustment, and being intended to solve prior art frame layer quantization parameter adjustment modes cannot online adaptive adjustment, and the distortion performance of encoder is not high, the unsettled problem of code check fluctuating nature.

Described frame layer quantization parameter method of adjustment comprises the following steps:

Extract present frame Static and dynamic contrast information;

According to the quantization parameter of the dynamic static contrast information of current frame image and Rate Control output, determine the final adjustment function of frame layer quantization parameter;

According to the quantization parameter of present frame Rate Control output and described final adjustment function and draw the quantization parameter that frame layer is new.

Another object of the embodiment of the present invention is to propose a kind of frame layer quantization parameter adjustment System, and described system comprises:

Static and dynamic contrast information extraction module, for extracting present frame Static and dynamic contrast information;

The final adjustment function acquisition module of frame layer quantization parameter, for according to the quantization parameter of the dynamic static contrast information of current frame image and Rate Control output, determines the final adjustment function of frame layer quantization parameter;

The new quantization parameter acquisition module of frame layer, for according to the quantization parameter of present frame Rate Control output and described final adjustment function and draw the quantization parameter that frame layer is new.

Another object of the embodiment of the present invention is to propose a kind of macroblock layer quantization parameter method of adjustment, and described method comprises:

Extract the dynamic and static contrast information of current macro and present frame;

By the dynamic and static contrast information of frame layer quantization parameter and current macro, present frame, determine the quantization parameter of current macro.

Another object of the embodiment of the present invention is to propose a kind of macroblock layer quantization parameter adjustment System,, described system comprises:

Dynamic and the static contrast information extraction module of macro block and frame, for extracting the dynamic and static contrast information of current macro and present frame;

The quantization parameter acquisition module of macro block, for the dynamic and static contrast information by frame layer quantization parameter and current macro, present frame, determines the quantization parameter of current macro.

Beneficial effect of the present invention:

The embodiment of the present invention is determined the adjusting range of frame layer quantization parameter on the one hand by dynamic, the static contrast information of image, on the other hand by the QP situation of current bit or RC output, for further adjustments, this can effectively control actual bit rate and draw close to target bit rate.Utilize frame layer quantization parameter of the present invention to adjust strategy, not only can enhancing rate distortion performance but also can make the code check fluctuation of encoder output code flow more mild, thereby lifting network transmission performance, can make the more approaching coding target bit rate of the actual output bit rate of encoder in addition.

In addition, the embodiment of the present invention is utilized Static and dynamic contrast information and the frame layer quantization parameter of current macro, present frame, self adaptation is adjusted the quantization parameter of current macro, thereby in promoting the distortion performance of encoder, keeps the stability of code check fluctuating nature.

Brief description of the drawings

Fig. 1 is a kind of frame layer of embodiment of the present invention quantization parameter method of adjustment flow chart;

Fig. 2 is the acquisition methods flow chart that the frame layer quantization parameter in Fig. 1 method finally adjusted function;

Fig. 3 is a kind of frame layer of embodiment of the present invention quantization parameter adjustment System structural representation;

Fig. 4 is the detailed structure schematic diagram that the frame layer quantization parameter in Fig. 3 system finally adjusted function;

Fig. 5 is a kind of macroblock layer quantization parameter of embodiment of the present invention method of adjustment flow chart;

Fig. 6 is a kind of macroblock layer quantization parameter of embodiment of the present invention adjustment System structural representation.

Embodiment

In order to make object 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 for explaining the present invention, not in order to limit the present invention.

The embodiment of the present invention adopts a kind of new frame layer quantization parameter method of adjustment, determine on the one hand the adjusting range of frame layer quantization parameter by dynamic, the static contrast information of image, on the other hand by the QP situation of current bit or RC output, for further adjustments, this can effectively control actual bit rate and draw close to target bit rate; In addition, the embodiment of the present invention has also proposed a kind of new macroblock layer from quantizing parameter regulation means, utilize Static and dynamic contrast information and the frame layer quantization parameter of current macro, present frame, self adaptation is adjusted the quantization parameter of current macro, thereby in promoting the distortion performance of encoder, keep the stability of code check fluctuating nature.

Fig. 1 is a kind of frame layer of embodiment of the present invention quantization parameter method of adjustment flow chart, and described method comprises:

S101: extract present frame Static and dynamic contrast information

Contrast_static _t＝mean(Contrast_static _n，t|Mb _n，t∈region)

Contrast_moving _t＝mean(Contrast_movingj _n，t)|Mb _n，t∈region，j＝1or2)

The wherein time sequence number of t presentation video in video sequence, n is the label of macro block in image, Contrast_static _tit is the static contrast information of t two field picture; Contrast_static _{n, t}it is the static contrast information of n macro block of t frame; Mb _{n, t}be n macro block of t frame; Contrast_moving _tbe the dynamic contrast information of t two field picture; Contrast_movingj _{n, t}represent the dynamic contrast information of n the macro block of t frame of asking for distinct methods, when j=1, n macro block Mb of t frame while representing not use motion vector mv information _{n, t}dynamic contrast, when j=2, n macro block Mb of t frame while representing to use motion vector mv information _{n, t}dynamic contrast; Region is the image-region of adding up, and mean is for getting mean value function.

(1) the static contrast information extracting method of present frame is as follows:

If current macro is Mb _{n, t}, p _{i, j, t}for Mb _{n, t}in pixel, the time sequence number of t presentation video in video sequence, n is the label of macro block in image, i, j represents respectively the row, column label in the young image of current pixel, Mb _{n, t}be that n macro block of t frame is also current macro, p _{i, j, t}for Mb _{n, t}in the capable j of t two field picture i row pixel be also current pixel.

P _{i, j, t}8 adjacent pixels be listed as follows:

p i-1，j-1，t	p i-1，j，t	p i-1，j+1，t
			p i，j-1，t	p i，j，t	p i，j+1，t
p i+1，j-1，t	p i+1，j，t	p i+1，j+1，t

Note sum _{i, j, t}=p _{i-1, j-1}, t+p _{i-1, j, t}+ p _{i-1, j+1, t}+ p _{i, j-1, t}+ p _{i, j+1, t}+ p _{i+1, j-1, t}+ p _{i+1, j, t}+ p _{i+1, j+1}, t, p _{i, j}three in abutting connection with being:

P _{i, j, t}three in abutting connection with the first pattern: Contrast_3 _{i, j, t, 1}=p _{i-1, j-1, t}+ p _{i-1, j, t}+ p _{i-1, j+1, t}

P _{i, j, t}three in abutting connection with the second pattern: Contrast_3 _{i, j, t, 2}=p _{i-1, j, t}+ p _{i-1, j+1, t}+ p _{i, j+1, t}

P _{i, j, t}three in abutting connection with the third pattern: Contrast_3 _{i, j, t, 3}=p _{i-1, j+1, t}+ p _{i, j+1, t}+ p _{i+1, j+1, t}

P _{i, j, t}three in abutting connection with the 4th kind of pattern: Contrast_3 _{i, j, t, 4}=p _{i, j+1, t}+ p _{i+1, j+1, t}+ p _{i+1, j, t}

P _{i, j, t}three in abutting connection with the 5th kind of pattern: Contrast_3 _{i, j, t, 5}=p _{i+1, j+1, t}+ p _{i+1, j, t}+ p _{i+1, j-1, t}

P _{i, j, t}three in abutting connection with the 6th kind of pattern: Contrast_3 _{i, j, t, 6}=p _{i+1, j, t}+ p _{i+1, j-1, t}+ p _{i, j-1, t}

P _{i, j, t}three in abutting connection with the 7th kind of pattern: Contrast_3 _{i, j, t, 7}=p _{i+1, j-1, t}+ p _{i, j-1, t}+ p _{i-1, j-1, t}

P _{i, j, t}three in abutting connection with the 8th kind of pattern: Contrast_3 _{i, j, t, 8}=p _{i, j-1, t}+ p _{i-1, j-1, t}+ p _{i-1, j, t}

Five in abutting connection with being:

P _{i, j, t}five in abutting connection with k kind pattern: Contrast_5 _{i, j, t, k}=sum _{i, j, t}-Contrast_3 _{i, j, t, k}current pixel p _{i, j, t}contrast be:

${\mathrm{Contrast}}_{i,j,t}=\underset{1\≤k\≤8}{\max }\left(\mathrm{abs}(5*\mathrm{Contrast}\_3_{i,j,t,k}-3*\mathrm{Contrast}\_5_{i,j,t,k})\right)/15$

Contrast _{i, j, t}be the capable j of t two field picture i row pixel be also current pixel p _{i, j, t}contrast, Contrast_3 _{i, j, t, k}for p _{i, j, t}three in abutting connection with k kind patterns, Contrast_5 _{i, j, t, k}for p _{i, j, t}five in abutting connection with k kind patterns.

Current macro Mb _{n, t}the static contrast of (n macro block of t frame) is:

Contrast_static _n，t＝sum(δ(Contrast _i，j，t)|p _i，j，t∈M _n，t)，

Wherein and Thres _{f (x)}for corresponding decision threshold.

Contrast_static _{n, t}be the static contrast information of n macro block of t frame, Contrast _{i, j, t}be the capable j of t two field picture i row pixel be also the contrast of current pixel, p _{i, j, t}for Mb _{n, t}in the capable j of t two field picture i row pixel be also current pixel.

The static contrast of present frame is:

Contrast_static _t＝mean(Contrast_static _n，t|Mb _n，t∈region)

(2) present frame dynamic contrast information extracting method is as follows:

The embodiment of the present invention provides two kinds of methods of obtaining video image macro-block multidate information contrast.

Method 1: do not use motion vector mv information, current macro Mb _{n, t}dynamic contrast be:

Contrast_moving1 _n，t＝sum(δ(Contrast _i，j，t)|p _i，j，t∈Mb _n，t&&p _i，j，t＝p _i，j，t-p _i，j，t-1)

The method is without carrying out motion search, and amount of calculation is little.

Method 2: use motion vector mv information, current macro Mb _{n, t}dynamic contrast be:

Contrast_moving2 _n，t＝sum(δ(mvx _n，t ²+mvy _n，t ²))|mv _n，t∈Mb _n，t)

Mvx _{n, t}represent the horizontal component of the motion vector of n macro block of t frame, mvy _{n, t}represent the vertical component of the motion vector of n macro block of t frame.

The motion vector that the method is obtained based on motion search, precision is high.

Based on the above two kinds methods of obtaining video image macro-block multidate information contrast, show that present frame dynamic contrast is:

Contrast_moving _t＝mean(Contrast_movingj _n，t)|Mb _n，t∈region，j＝1or2)

Contrast_movingj _{n, t}represent the dynamic contrast information of n the macro block of t frame of asking for distinct methods; When j=1, n macro block Mb of t frame while representing not use motion vector mv information _{n, t}dynamic contrast, when j=2, n macro block Mb of t frame while representing to use motion vector mv information _{n, t}dynamic contrast.

S102: by the dynamic static contrast information of current frame image and the QP of RC output _t, determine the final adjustment function of frame layer quantization parameter QP wherein RC is Rate Control, and QP is quantization parameter, QP _tfor the quantization parameter of original t frame, it is the adjustment function of t frame quantization parameter.

Be illustrated in figure 2 described frame layer quantization parameter and finally adjust the acquisition methods flow chart of function, the acquisition methods that described frame layer quantization parameter finally adjusted function comprises the following steps:

S1021: first by the coding sequence number id of frame per second fps and image _t, determine image sets period of waves.

period＝f ₀(id _t，fps))

Wherein, period is image sets period of waves, id _tbe the coding sequence number of t frame, f ₀(id _t, fps)) be the coded sequence id by frame per second fps and image _tdefinite image sets function period of waves.

S1022: determine that by described image sets frame layer quantization parameter initially adjust dQP period of waves _{1, t}function dQP _{1, t}=f ₁(id _t, period),

Wherein f ₁(x, y) meets, f ₁(x+ky, y)=f ₁(x, y),

DQP _{1, t}be that t frame quantization parameter is initially adjusted function, k is phase constant, to ask local derviation.

S1023: dynamic, static contrast information and described initial adjustment function in conjunction with current frame image obtain frame layer quantization parameter correction dQP _{2, t}function

dQP _2，t＝f ₂(Contrast_static _t，Contrast_moving _t，dQP _1，t)

DQP _{2, t}represent t frame quantization parameter correction function.

S1024: according to the QP situation of current bit or RC output, adjust frame layer quantization parameter correction function dQP _{2, t}, obtain frame layer QP and finally adjust function

${\mathrm{dQP}}_t^{\mathrm{new}}=f_3({\mathrm{dQP}}_{2,t},{\mathrm{QP}}_t,{\mathrm{bit}}_t)$

be that t frame quantization parameter is finally adjusted function, Bitt is current bit value

S103: obtain the quantization parameter that frame layer is new

Be a kind of frame layer of embodiment of the present invention quantization parameter adjustment System structural representation as shown in Figure 3, described system comprises:

Frame Static and dynamic contrast information extraction module, for extracting present frame Static and dynamic contrast information;

The final adjustment function acquisition module of frame layer quantization parameter, for according to the quantization parameter of the dynamic static contrast information of current frame image and Rate Control output, determines the final adjustment function of frame layer quantization parameter;

The new quantization parameter acquisition module of frame layer, for according to the quantization parameter of present frame Rate Control output and described final adjustment function and draw the quantization parameter that frame layer is new.

Be the detailed structure schematic diagram that frame layer quantization parameter in Fig. 3 system finally adjusted function acquisition module as shown in Figure 4, described frame layer quantization parameter finally adjusted function acquisition module and comprised:

Image sets determination module period of waves, for according to the coding sequence number of frame per second and image, determines image sets period of waves;

Frame layer quantization parameter initially adjusted function determination module, for determining that according to image sets frame layer quantization parameter initially adjust function period of waves;

Frame layer quantization parameter correction function determination module, for obtaining frame layer quantization parameter correction function according to dynamic, the static contrast information of current frame image and described initial adjustment function;

Frame layer quantization parameter finally adjusted function determination module, for according to the quantization parameter situation of current bit or Rate Control output, adjusts described frame layer quantization parameter correction function, determines that frame layer quantization parameter finally adjust function.

Use the frame layer quantization parameter method of adjustment of the embodiment of the present invention, not only can promote the distortion performance of encoder, and can make the code check fluctuation of encoder more mild, promote the network transmission performance of code stream, can make in addition the more approaching coding target bit rate of the actual output bit rate of encoder.

Fig. 5 is a kind of macroblock layer quantization parameter of embodiment of the present invention method of adjustment flow chart.

Definite various modes that exists of current macroblock layer quantization parameter is as determined method according to the energy of AC coefficient, and macroblock layer Rate Control is determined method, adjusts method based on visual characteristic.Wherein the method based on AC coefficient energy is calculated simply, and detail of the high frequency is had to good preservation characteristics.The embodiment of the present invention, in conjunction with the Static and dynamic contrast information of macro block and the Static and dynamic contrast information of frame, frame layer quantization parameter, is determined the quantization parameter of current macro.Wherein, Static and dynamic contrast information is respectively used to protect the high-frequency information of motion macro block and static macro block.

Described macroblock layer quantization parameter method of adjustment comprises the following steps:

S501: the dynamic and static contrast information that extracts current macro and present frame

The static contrast C ontrast_static of current macro _{n, t}, current macro dynamic contrast Contrast_movingj _{n, t}the static contrast C ontrast_static of (j=1 or 2) and present frame _t, present frame dynamic contrast Contrast_moving _tacquisition methods above in detail record, do not repeat them here.

S502: by frame layer quantization parameter QP _iand the dynamic and static contrast information of current macro, present frame, determine the quantization parameter QP of current macro _{n, t}

QP _n，t＝QP _t+f _{mb_static}*weight _{mb_static}+f _{mb_moving}*weight _{mb_moving}

Wherein QP _tfor the quantization parameter of original t frame, weight _{mb_static}, weight _{mb_moving}, be respectively the weight of static contrast and dynamic contrast information; f _{mb_static}(Contrast_static _{n, t}, Contrast_static _t), for the quantization parameter based on static contrast is adjusted function,

F _{mb_moving}(Contrast_movingj _{n, t}, Contrast_moving _t), for the quantization parameter based on dynamic contrast information is adjusted function,

Be a kind of macroblock layer quantization parameter of embodiment of the present invention adjustment System structural representation as shown in Figure 6, described system comprises:

Dynamic and the static contrast information extraction module of macro block and frame, for extracting the dynamic and static contrast information of current macro and present frame;

The quantization parameter acquisition module of macro block, for according to the dynamic and static contrast information of frame layer quantization parameter and current macro, present frame, determines the quantization parameter of current macro.

Use the macroblock layer quantization parameter method of adjustment of the embodiment of the present invention can, in promoting the distortion performance of encoder, keep the stability of code check fluctuating nature.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.

Claims (9)

1. a frame layer quantization parameter method of adjustment, is characterized in that, said method comprising the steps of:

Extract present frame Static and dynamic contrast information;

According to the quantization parameter of the dynamic static contrast information of current frame image and Rate Control output, determine the final adjustment function of frame layer quantization parameter;

According to the quantization parameter of present frame Rate Control output and described final adjustment function and draw the quantization parameter that frame layer is new;

Wherein, the static contrast information of described present frame is:

Contrast_static _t=mean(Contrast_static _n,t|Mb _n,t∈region)

The wherein time sequence number of t presentation video in video sequence, n is the label of macro block in image, Contrast_static _tit is the static contrast information of t two field picture; Contrast_static _n,tbe the static contrast information of n macro block of t frame, Mb _n,tbe n macro block of t frame, region is the image-region of adding up, and mean is for getting mean value function.

2. frame layer quantization parameter method of adjustment as claimed in claim 1, is characterized in that,

Described Contrast_static _n,t=sum (d (Contrast _{i, j, t}) | p _{i, j, t}∈ M _n,t),

Wherein and Thres _{f (x)}for corresponding decision threshold,

Contrast_static _n,tbe the static contrast information of n macro block of t frame, Contrast _{i, j, t}be the capable j of t two field picture i row pixel be also the contrast of current pixel, p _{i, j, t}for Mb _n,tin the capable j of t two field picture i row pixel be also current pixel;

Wherein

Contrast _{i, j, t}be the capable j of t two field picture i row pixel be also current pixel p _{i, j, t}contrast, Contrast_3 _{i, j, t, k}for p _{i, j, t}three in abutting connection with k kind patterns, Contrast_5 _{i, j, t, k}for p _{i, j, t}five in abutting connection with k kind patterns.

3. frame layer quantization parameter method of adjustment as claimed in claim 2, is characterized in that described Contrast_3 _{i, j, t, k}be specially:

P _{i, j, t}three in abutting connection with the first pattern: Contrast_3 _{i, j, t, 1}=p _{i-1, j-1, t}+ p _{i-1, j, t}+ p _{i-1, j+1, t}

P _{i, j, t}three in abutting connection with the second pattern: Contrast_3 _{i, j, t, 2}=p _{i-1, j, t}+ p _{i-1, j+1, t}+ p _{i, j+1, t}

P _{i, j, t}three in abutting connection with the third pattern: Contrast_3 _{i, j, t, 3}=p _{i-1, j+1, t}+ p _{i, j+1, t}+ p _{i+1, j+1, t}

P _{i, j, t}three in abutting connection with the 4th kind of pattern: Contrast_3 _{i, j, t, 4}=p _{i, j+1, t}+ p _{i+1, j+1, t}+ p _{i+1, j, t}

P _{i, j, t}three in abutting connection with the 5th kind of pattern: Contrast_3 _{i, j, t, 5}=p _{i+1, j+1, t}+ p _{i+1, j, t}+ p _{i+1, j-1, t}

P _{i, j, t}three in abutting connection with the 6th kind of pattern: Contrast_3 _{i, j, t, 6}=p _{i+1, j, t}+ p _{i+1, j-1, t}+ p _{i, j-1, t}

P _{i, j, t}three in abutting connection with the 7th kind of pattern: Contrast_3 _{i, j, t, 7}=p _{i+1, j-1, t}+ p _{i, j-1, t}+ p _{i-1, j-1, t}

P _{i, j, t}three in abutting connection with the 8th kind of pattern: Contrast_3 _{i, j, t, 8}=p _{i, j-1, t}+ p _{i-1, j-1, t}+ p _{i-1, j, t}

Described Contrast_5 _{i, j, t, k}be specially:

P _{i, j, t}five in abutting connection with k kind pattern: Contrast_5 _{i, j, t, k}=sum _{i, j, t}-Contrast_3 _{i, j, t, k}

Wherein, sum _{i, j, t}=p _{i-1, j-1, t}+ p _{i-1, j, t}+ p _{i-1, j+1, t}+ p _{i, j-1, t}+ p _{i, j+1, t}+ p _{i+1, j-1, t}+ p _{i+1, j, t}+ p _{i+1, j+1, t}, the time sequence number of t presentation video in video sequence, i, j represent respectively the row, column label in the young image of current pixel, p _{i, j, t}for the capable j of t two field picture i in current macro row pixel be also current pixel.

4. frame layer quantization parameter method of adjustment as claimed in claim 1, is characterized in that,

Described present frame dynamic contrast information is:

Contrast_moving _t=mean(Contrast_movingj _n,t)|Mb _n,t∈region,j=1or2)

Wherein, Contrast_moving _tbe the dynamic contrast information of t two field picture; Mb _n,tbe n macro block of t frame, region is the image-region of adding up, and mean is for getting mean value function; Contrast_movingj _n,trepresent the dynamic contrast information of n the macro block of t frame of asking for distinct methods, when j=1, n macro block Mb of t frame while representing not use motion vector mv information _n,tdynamic contrast, when j=2, n macro block Mb of t frame while representing to use motion vector mv information _n,tdynamic contrast;

Contrast_moving1 _n,t=sum(d(Contrast _i,j,t)|p _i,j,t∈Mb _n,t&&p _i,j,t=p _i,j,t-p _i,j,t-1)

Wherein,

Contrast _{i, j, t}be the capable j of t two field picture i row pixel be also current pixel p _{i, j, t}contrast, Contrast_3 _{i, j, t, k}for p _{i, j, t}three in abutting connection with k kind patterns, Contrast_5 _{i, j, t, k}p _{i, j, t}five in abutting connection with k kind patterns;

Contrast_moving2 _n,t=sum(d(mvx _n,t ²+mvy _n,t ²))|mv _n,t∈Mb _n,t)

Wherein mvx _n,trepresent the horizontal component of the motion vector of n macro block of t frame, mvy _n,trepresent the vertical component of the motion vector of n macro block of t frame.

5. frame layer quantization parameter method of adjustment as claimed in claim 1, is characterized in that, the final adjustment function of described definite frame layer quantization parameter specifically comprises the following steps:

According to the coding sequence number of frame per second and image, determine image sets period of waves;

Determine that according to image sets frame layer quantization parameter initially adjust function period of waves;

Obtain frame layer quantization parameter correction function according to dynamic, the static contrast information of current frame image and described initial adjustment function;

According to the quantization parameter situation of current bit or Rate Control output, adjust described frame layer quantization parameter correction function, obtain frame layer quantization parameter and finally adjust function.

6. frame layer quantization parameter method of adjustment as claimed in claim 5, is characterized in that,

Described image sets is period of waves:

period=f ₀(id _t,fps))

Wherein, period is image sets period of waves, id _tbe the coding sequence number of t frame, f ₀(id _t, fps)) be the coded sequence id by frame per second fps and image _tdefinite image sets function period of waves;

Described frame layer quantization parameter initially adjusted function and is:

dQP _1,t=f ₁(id _t,period)，

Wherein f ₁(x, y) meets, f ₁(x+ky, y)=f ₁(x, y),

DQP _{1, t}be that t frame quantization parameter is initially adjusted function, k is phase constant, to ask local derviation; Described frame layer quantization parameter correction function is:

dQP _2,t=f ₂(Contrast_static _t,Contrast_moving _t,dQP _1,t)

DQP _{2, t}represent t frame quantization parameter correction function;

Described frame layer quantization parameter finally adjusted function and is:

be that t frame quantization parameter is finally adjusted function, Bitt is current bit value.

7. a frame layer quantization parameter adjustment System, is characterized in that, described system comprises:

Static and dynamic contrast information extraction module, for extracting present frame Static and dynamic contrast information;

The final adjustment function acquisition module of frame layer quantization parameter, for according to the quantization parameter of the dynamic static contrast information of current frame image and Rate Control output, determines the final adjustment function of frame layer quantization parameter;

The new quantization parameter acquisition module of frame layer, for according to the quantization parameter of present frame Rate Control output and described final adjustment function and draw the quantization parameter that frame layer is new

Wherein, the static contrast information of described present frame is:

Contrast_static _t=mean(Contrast_static _n,t|Mb _n,t∈region)

The wherein time sequence number of t presentation video in video sequence, n is the label of macro block in image, Contrast_static _tit is the static contrast information of t two field picture; Contrast_static _n,tbe the static contrast information of n macro block of t frame, Mb _n,tbe n macro block of t frame, region is the image-region of adding up, and mean is for getting mean value function.

8. frame layer quantization parameter adjustment System as claimed in claim 7, is characterized in that, the final adjustment function acquisition module of described frame layer quantization parameter comprises:

Image sets determination module period of waves, for according to the coding sequence number of frame per second and image, determines image sets period of waves;

Frame layer quantization parameter initially adjusted function determination module, for determining that according to image sets frame layer quantization parameter initially adjust function period of waves;

Frame layer quantization parameter correction function determination module, for obtaining frame layer quantization parameter correction function according to dynamic, the static contrast information of current frame image and described initial adjustment function;

Frame layer quantization parameter finally adjusted function determination module, for according to the quantization parameter situation of current bit or Rate Control output, adjusts described frame layer quantization parameter correction function, determines that frame layer quantization parameter finally adjust function.

9. a macroblock layer quantization parameter method of adjustment, is characterized in that, described method comprises:

Extract the dynamic and static contrast information of current macro and present frame;

By the dynamic and static contrast information of frame layer quantization parameter and current macro, present frame, determine the quantization parameter of current macro;

Wherein, the quantization parameter of described current macro is:

QP _n,t=QP _t+f _{mb_static}*weight _{mb_static}+f _{mb_moving}*weight _{mb_moving}

Wherein QP _n,tfor the quantization parameter of current macro, QP _tfor the quantization parameter of original t frame, weight _{mb_static}, weight _{mb_moving}, be respectively the weight of static contrast and dynamic contrast information; f _{mb_static}(Contrast_static _n,t, Contrast_static _t), for the quantization parameter based on static contrast is adjusted function, f _{mb_moving}(Contrast_movingj _n,t, Contrast_moving _t), for the quantization parameter based on dynamic contrast information is adjusted function, the wherein time sequence number of t presentation video in video sequence, n is the label of macro block in image, Contrast_static _tbe the static contrast information of t two field picture, Contrast_static _n,tbe the static contrast information of n macro block of t frame, Contrast_moving _tbe the dynamic contrast information of t two field picture, Contrast_movingj _n,trepresent the dynamic contrast information of n the macro block of t frame of asking for distinct methods, when j=1, n macro block Mb of t frame while representing not use motion vector mv information _n,tdynamic contrast, when j=2, n macro block Mb of t frame while representing to use motion vector mv information _n,tdynamic contrast.