CN102420989A - Intra-frame prediction method and device - Google Patents

Abstract

The invention discloses an intra-frame prediction method and device. The intra-frame prediction method comprises the following steps of: carrying out mode prediction on a 16*16 macro block input to an intra-frame prediction buffer area according to the regulated 4*4 subblock prediction order to obtain a prediction value under each mode, calculating an SAD (Sum of Absolute Differences) value under each prediction mode, selecting a mode with minimum SAD value as an optimal prediction mode of 16 subblocks to obtain an optical prediction mode of the 16 subblocks and a Sun_SAD4 of the SADs of the macro block under the optimal prediction mode of the corresponding subblocks; obtaining the optimal 16*16 prediction mode and the macro block SAD16 under the corresponding optimal mode; and finally comparing the Sum_SAD4 value with the SAD16 value, determining a prediction mode adopted by the macro block, and after the prediction is completed, sending the optical prediction mode and a prediction residual error to a subsequent processing module. According to the invention, reestablishment waiting time in the prediction process is reduced, and intra-frame prediction in a high-resolution video compression process can be processed in real time.

Description

Intra-frame prediction method and device

Technical field

The present invention relates to a kind of infra-prediction techniques, particularly a kind of intra-frame prediction method and device.

Background technology

H.264/AVC the video encoding standard of new generation that the joint video team of being made up of ISO/IEC and ITU-T (JVT) is formulated.H.264/AVC standard is compared with the coded system of MPEG-4 standard, H.263++ standard code, has more outstanding PSNR performance.H.264/AVC PSNR (Y-PSNR) is than the average high 2dB of MPEG-4, than average high 3dB H.263++.Under identical reconstructed image quality, the data volume after H.264/AVC encoding compares H.263++ and MPEG-4 (SP) reduces 60% at least.Under identical reconstructed image quality, H.264/AVC the data behind the coding are stronger to the adaptability of channel delay.H.264/AVC standard both can be applicable to low time delay pattern (like video conferencing) to satisfy real time business, can work in the occasion (like video storage) of no time delay restriction again.Compare former video encoding standard, H.264/AVC standard has been introduced the advanced technology of inter prediction technology etc. of estimation, multi-reference frame and the multiple block size of the infra-frame prediction that comprises in 4 * 4 integer transforms, spatial domain, 1/4 pixel precision.Quoting of new technology brought higher code efficiency, but improved the complexity of algorithm simultaneously greatly.

Infra-frame prediction is as one of key technology of standard H.264; It has very high complexity, and each macro block will carry out 9 kind of 4 * 4 model prediction and 4 kind of 16 * 16 prediction, for high-definition picture; Its amount of calculation is huge; To handle the image of a frame 1920 * 1080, just need carry out several hundred million time add subtract computing, need a large amount of memory access operation in addition.

At present military applications to image compression system require increasingly high; Like multiway images compression and transmission; Traditional infra-frame prediction implementation method based on software also progressively highlights at the bottleneck aspect the processing speed; Research can hard-wired real-time intra-frame prediction method become the important component part of development engineering with device, also is one of the core component of the SoC chip of following realization HD video compressing and coding system.

In the prior art, H.264 intraframe prediction algorithm is having description in the standard, but for how realizing not having unified demand, often based on concrete application requirements to standard H.264 do corresponding delete or change after, realize again.

Through the retrieval of prior art document is found application number is 201010231760.5, name is called the Chinese patent of " a kind of quick 4 * 4 block selection method of intra-frame prediction mode "; This patent proposes 4 * 4 sub-pieces are carried out seven edge direction coefficient calculations earlier; According to fringing coefficient image block is divided into nine kinds of edge pattern again and predicts that the selection rate distortion cost is adjudicated optimal mode.The Chinese patent that application number is 200910100537.4, name is called " method of intra-prediction and device " proposes in prediction process; The not execution predictive mode of selecting reference pixel to rebuild completion in advance carries out prediction processing; Reduce the stand-by period, improved infra-frame prediction efficient.

Have following defective in the prior art: adopt the rate distortion costs function to be unfavorable for that the flowing water of hardware realizes, every kind of model prediction all need be waited for the completion of reconstruction basically; Can effectively reduce the stand-by period though application number is 200910100537.4 Chinese patent, its control procedure is complicated, is unfavorable for that hardware realizes.

Summary of the invention

Technical problem to be solved by this invention provides a kind of intra-frame prediction method and device, and it reduces in the forecasting process rebuilds the stand-by period, and processing sequence is consistent, is convenient to hardware and realizes, can handle the infra-frame prediction in the high-resolution video compression process in real time.

For solveing the technical problem; The invention provides a kind of intra-frame prediction method; Intra-frame prediction method may further comprise the steps: 16 * 16 macro blocks to being input to the infra-frame prediction buffering area carry out model prediction according to adjusted 4 * 4 sub-piece prediction order earlier, obtain the predicted value under the various patterns, again according to the SAD interpretational criteria; Calculate the sad value under each predictive mode; Choose the optimum prediction mode of the minimum pattern of sad value as 16 sub-block, handle 16 sub-block after, obtained the optimum prediction mode of 16 sub-block and the macro block SAD sum Sum_SAD4 under the corresponding sub-piece optimum prediction mode; Carry out the prediction of 16 * 16 predictive modes then, obtain the macro block SAD16 under best 16 * 16 predictive modes and the corresponding optimal mode; Compare Sum_SAD4 value and SAD16 value at last, the predictive mode that the decision macro block should adopt; After prediction finishes, will obtain optimum prediction mode and prediction residual and deliver to the subsequent treatment module.

Preferably, said Sum_SAD4 value is less than the SAD16 value, and then macro block adopts 4 * 4 predictive modes, if the Sum_SAD4 value greater than the SAD16 value, then adopts 16 * 16 predictive modes.

Preferably, said intra-frame prediction method adopts the SAD criterion as the optimal mode evaluation criterion, for 4 * 4 predictive modes, adopts SAD4, specifically suc as formula (1):

$\mathrm{SAD}4=\underset{i=0}{\overset{i=3}{\mathrm{\Σ}}}\underset{j=0}{\overset{j=3}{\mathrm{\Σ}}}|\mathrm{Psrc}(i,j)-\mathrm{Ppre}(i,j)|$ Formula (1)

For 16 * 16 predictive modes, adopt SAD16, specifically suc as formula (2):

$\mathrm{SAD}16=\underset{i=0}{\overset{i=15}{\mathrm{\Σ}}}\underset{j=0}{\overset{j=15}{\mathrm{\Σ}}}|\mathrm{Psrc}(i,j)-\mathrm{Ppre}(i,j)|$ Formula (2)

In formula (1) and the formula (2), Psrc representes source pixel, and Ppre representes predict pixel, | x| representes to ask absolute value.

The present invention also provides a kind of infra-frame prediction device; It comprises control module, predictor computation module, mode adjudging module, order mapping block, residual error output module; Control module is carried out infra-frame prediction control based on adjusted sub-piece processing order to macro block, and controls the water operation of other modules; The mode adjudging module is accomplished the accumulation calculating of sad value; And the judgement optimum prediction mode adopts parallel organization to realize the calculating of SAD; Carry out the SAD judgement of macro block based on the SAD result calculated, upgrade optimal mode register and best sad value, and the predictive mode of conclusive judgement macro block employing; The order mapping block is realized the mapping treatment of sub-piece memory address; The residual error output module is accomplished the output of residual error parameter after prediction is accomplished; Predictor computation module realizes the calculating of pixel predictors under the various predictive modes.

Preferably; Said predictor computation module comprises data distributor, control unit, first processing unit to the, 16 processing units, the first predicted value buffering area, the second predicted value buffering area; Data distributor, first processing unit to the, 16 processing units, the first predicted value buffering area, the second predicted value buffering area are connected with control unit, and first processing unit to the, 16 processing units are connected between data distributor and the first predicted value buffering area, the second predicted value buffering area.

Positive progressive effect of the present invention is: the present invention reduces in the forecasting process and rebuilds the stand-by period, and processing sequence is consistent, is convenient to hardware and realizes, can handle the infra-frame prediction in the high-resolution video compression process in real time.In addition, the present invention improves infra-frame prediction efficient, is easy to hardware and realizes, also can be embedded in the chip.

Description of drawings

Fig. 1 is the exemplary flow sketch map of infra-frame prediction optimization method in the embodiment of the invention.

Fig. 2 is the sub-piece prediction order adjustment sketch map of infra-frame prediction optimization method in the embodiment of the invention.

Fig. 3 is 4 * 4 sub-block prediction mode sketch mapes.

Fig. 4 is the structured flowchart of infra-frame prediction device in the embodiment of the invention.

Fig. 5 is the structured flowchart of predictor computation module in the infra-frame prediction device in the embodiment of the invention.

Embodiment

Lift a preferred embodiment below, and combine accompanying drawing to come the clearer the present invention that intactly explains.

Embodiment

As shown in Figure 1, the intra-frame prediction method in the present embodiment comprises the steps and function: the adjustment of sub-piece prediction order; Brightness 4 * 4 model predictions; Brightness 16 model predictions; Colourity 8 * 8 model predictions; Mode adjudging; Rebuild macro block.

This intra-frame prediction method specifically comprises the steps: 16 * 16 macro blocks that are input to the infra-frame prediction buffering area are carried out model prediction according to adjusted 4 * 4 sub-piece prediction order earlier; Obtain the predicted value under the various patterns; Again according to SAD (absolute error of predict blocks and source images with; Sum of Absolute Differences) interpretational criteria calculates the sad value under each predictive mode, chooses the optimum prediction mode of the minimum pattern of sad value as 4 * 4 sub-pieces (16 sub-block); After handling 16 sub-block, the optimum prediction mode of 16 sub-block and the macro block SAD sum Sum_SAD4 under the corresponding sub-piece optimum prediction mode have been obtained; Carry out the prediction of 16 * 16 predictive modes then, obtain the macro block SAD16 under best 16 * 16 predictive modes and the corresponding optimal mode; Relatively Sum_SAD4 value and SAD16 value determine the predictive mode that macro block should adopt at last, specifically be if the Sum_SAD4 value less than the SAD16 value, macro block employing 4 * 4 predictive modes then are if the Sum_SAD4 value greater than the SAD16 value, then adopts 16 * 16 predictive modes; After prediction finishes, will obtain optimum prediction mode and prediction residual and deliver to the subsequent treatment module.

Luminance macroblock is predicted; Carry out 4 * 4 sub-block mode predictions earlier according to adjusted sub-piece prediction order (see figure 2); From seven kinds of patterns (mode 3 and mode 7 are not done, and see Fig. 3), find the best 4 * 4 patterns of each sub-piece in the macro block by cost minimum (adopting SAD) principle; And then macro block as 16 * 16 macro block (mb) type, from four kinds of predictive modes, find out 16 * 16 best patterns by the minimum principle of cost; The cost value that compares the corresponding optimum prediction mode of two kinds of macro block (mb) types is at last selected the minimum macro block (mb) type of the cost type of coding of current macro the most, and the intraframe predictive coding pattern of the optimum prediction mode of correspondence as macro block.As shown in Figure 2, sub-piece prediction order in the prediction of sub-piece, can significantly reduce the stand-by period of reconstruction after adjustment, effectively improve forecasting efficiency.As shown in Figure 2, the sub-piece label among the figure is the order of sub-piece prediction processing in proper order, and the adjustment pre-treatment is four " Z " font orders in proper order, and adjusted processing sequence then is non-" Z " font.

Former order all need be waited for rebuilding when carrying out sub-piece 1,3,5,7,9,11,13,15 (sequence number before the adjustment) prediction and accomplish; After the order adjustment; Except that carry out sub-piece 1,3 and 15 (adjustment back sequence number) need to wait for rebuild; Other sub-piece predictions all need not waited for reconstruction, have improved the efficient of prediction processing greatly.

The present invention adopts the SAD criterion as the optimal mode evaluation criterion, for 4 * 4 predictive modes, adopts SAD4, specifically suc as formula (1):

$\mathrm{SAD}4=\underset{i=0}{\overset{i=3}{\mathrm{\Σ}}}\underset{j=0}{\overset{j=3}{\mathrm{\Σ}}}|\mathrm{Psrc}(i,j)-\mathrm{Ppre}(i,j)|$ Formula (1)

For 16 * 16 predictive modes, adopt SAD16, specifically suc as formula (2):

$\mathrm{SAD}16=\underset{i=0}{\overset{i=15}{\mathrm{\Σ}}}\underset{j=0}{\overset{j=15}{\mathrm{\Σ}}}|\mathrm{Psrc}(i,j)-\mathrm{Ppre}(i,j)|$ Formula (2)

In formula (1) and the formula (2), Psrc representes source pixel, and Ppre representes predict pixel, | x| representes to ask absolute value.

As far as chrominance macroblock, from four kind of 8 * 8 pattern, find out optimum prediction mode by the minimum principle of cost.

The concrete prediction steps of the present invention is following:

(1) brightness Intra4 * 4 predictions may further comprise the steps:

(1.1) from rebuild macro block, reading 16 left margin pixels puts in the left margin register (LeftB_Reg); Put register Intra16_PreMode=0;

(1.2) carry out 0 sub-block mode 0 prediction; Obtain the minimum sad value of predicted value and SAD_Min(; When pattern 0 prediction; The minimum sad value of initialization is the sad value of pattern 0); Deposit correspondence position among the prediction buffering area A in, put 4x4 predictive mode register in the predictive mode register Intra4x4PredMode(frame) [0]=0;

(1.3) carry out pattern 1 prediction, obtain predicted value and SAD_Cur (current SAD) value, predicted value is temporary in prediction shift register (32bit * 4);

(1.4) compare SAD_Min and SAD_Cur value, if SAD_Min＞SAD_Cur then upgrades SAD_Min=SAD_Cur; Intra4x4PredMode [0]=1; And predicted value shifted out the shift register from prediction, write the corresponding sub-piece position among the prediction buffering area A, replace former predicted value; Otherwise, constant;

(1.5) predictive mode 2～8 is accomplished in repeating step (1.3)～(1.4), obtains sub-piece 0 optimum prediction mode and predicted value;

(1.6) the sub-piece 0 residual sum predictive mode Intra4x4PredMode [0] of output is to DCT (conversion of discrete cosine limit) module;

(1.7) wait for the residual values that IDCT (inverse discrete cosine transformation) returns, rebuild sub-piece 0, deposit in and rebuild the macro block correspondence position, the right margin pixel of sub-piece 0 is deposited in the bounds register (Temp_LeftB1_Reg (32bits)), use when being equipped with sub-piece 1 prediction;

(1.8) prediction of sub-piece 1 is accomplished in repeating step (1.2)～(1.6);

(1.9) prediction and the reconstruction of sub-piece 2～15 are accomplished in repeating step (1.2)～(1.8), obtain SAD4x4 with.

(2) brightness 16 * 16 model predictions may further comprise the steps:

(2.1) carry out the prediction of 16x16 pattern 0, with the value of SAD16 constantly with the SAD4x4 comparison, if SAD16＞SAD4x4; The then prediction of stop mode 0; Otherwise the prediction of completion pattern 0, and predicted value deposited among the prediction buffering area B, Intra16_PreMode is changed to 1;

(2.2) carry out the prediction of 16x16 pattern 1, if Intra16_PreMode=1, then the predicted value with pattern 1 deposits among the prediction buffering area A, otherwise among the PreMB_B, in the forecasting process with best sad value relatively, as greater than then stopping this model prediction;

(2.3) repeating step (2.2) is accomplished the prediction of mode 3～4, obtains Intra16_PreMode and predicted value;

(2.4) if Intra16_PreMode=1, the final macro block in surface adopts the Intra16x16 pattern, gets into step (2.5), otherwise adopts the Intra4x4 pattern, gets into step (2.6); See the Ready signal off, show MB prediction completion.

(2.5) export sub-piece residual error successively to the DCT module, wait for that IDCT returns residual error and rebuilds macro block, write in the reconstruction buffer district, upgrade last line buffer (Up Line Buffer) the correspondence position boundary value of coboundary row buffer;

(2.6) get into colourity Intra8 * 8 predictions.

(3) colourity Intra8 * 8 predictions may further comprise the steps:

(3.1) carry out 0 prediction of chrominance C r macro block mode, predicted value is deposited among the prediction buffering area A, SAD is deposited among the SAD16, the optimal mode register is changed to 0;

(3.2) carry out the prediction of pattern 1, predicted value deposited among the prediction buffering area B, if the SAD of pattern 1 greater than the SAD of pattern 0, the prediction of stop mode 1 is 1 and the value of SAD otherwise upgrade the optimal mode register value;

(3.3) prediction of repeating step (3.1)～(3.2) completion pattern 2,3 draws optimum prediction mode and predicted value;

(3.4) export sub-piece residual error successively to the DCT module, wait for that IDCT returns residual error and rebuilds chrominance macroblock, write in the reconstruction, upgrade coboundary row buffer (Up Line Buffer) corresponding sides dividing value.

As shown in Figure 4, the infra-frame prediction device comprises control module, predictor computation module, mode adjudging module, order mapping block, residual error output module, current macro buffering area, rebuilds macro block buffer, optimal mode buffering area etc.

Wherein, the current macro buffering area is used to deposit the macro block pixels data when pre-treatment; Rebuild the macro block pixels data after macro block buffer is used to deposit reconstruction; The optimal mode buffering area is used to deposit best predictive mode; Control module is accomplished the control and the data dispatch of whole device and is handled, and according to adjusted sub-piece processing order macro block is carried out infra-frame prediction control, and controls the water operation of other modules; The mode adjudging module is accomplished the accumulation calculating of sad value; And the judgement optimum prediction mode adopts parallel organization to realize the calculating of SAD; Carry out the SAD judgement of macro block according to the SAD result calculated, upgrade optimal mode register and best sad value, and the predictive mode of conclusive judgement macro block employing; The order mapping block is realized the mapping treatment of sub-piece memory address, realizes the correct access visit of sub-blocks of data; The residual error output module is accomplished the output of residual error parameter after prediction is accomplished.Predictor computation module realizes the calculating of pixel predictors under the various predictive modes; This module is the nucleus module of infra-frame prediction device; Its structure is as shown in Figure 5; Predictor computation module comprises data distributor, control unit, first processing unit to the, 16 processing units, the first predicted value buffering area A, the second predicted value buffering area B; Data distributor, first processing unit to the, 16 processing units, the first predicted value buffering area A, the second predicted value buffering area B are connected with control unit, and first processing unit to the, 16 processing units are connected between data distributor and the first predicted value buffering area A, the second predicted value buffering area B.Data distributor will calculate six processing units of data allocations to ten that need use according to predictive mode among the figure; Pixel of the corresponding processing of each processing unit; It still is the second predicted value buffering area B that control unit control predicted value specifically stores the first predicted value buffering area A into, and output is handled and produced the first predicted value buffering area A or the second predicted value buffering area B address stored.In predictor computation module, adopted parallel and flowing structure, can in a clock cycle, accomplish the predictor calculation of a sub-block, can handle the infra-frame prediction of high-definition picture in real time.

Though more than described embodiment of the present invention, it will be understood by those of skill in the art that these only illustrate, under the prerequisite that does not deviate from principle of the present invention and essence, can make numerous variations or modification to these execution modes.Therefore, protection scope of the present invention is limited appended claims.

Claims (5)

1. intra-frame prediction method; It is characterized in that; Intra-frame prediction method may further comprise the steps: 16 * 16 macro blocks to being input to the infra-frame prediction buffering area carry out model prediction according to adjusted 4 * 4 sub-piece prediction order earlier, obtain the predicted value under the various patterns, again according to the SAD interpretational criteria; Calculate the sad value under each predictive mode; Choose the optimum prediction mode of the minimum pattern of sad value as 16 sub-block, handle 16 sub-block after, obtained the optimum prediction mode of 16 sub-block and the macro block SAD sum Sum_SAD4 under the corresponding sub-piece optimum prediction mode; Carry out the prediction of 16 * 16 predictive modes then, obtain the macro block SAD16 under best 16 * 16 predictive modes and the corresponding optimal mode; Compare Sum_SAD4 value and SAD16 value at last, the predictive mode that the decision macro block should adopt; After prediction finishes, will obtain optimum prediction mode and prediction residual and deliver to the subsequent treatment module.

2. intra-frame prediction method as claimed in claim 1 is characterized in that, said Sum_SAD4 value is less than the SAD16 value, and then macro block adopts 4 * 4 predictive modes, if the Sum_SAD4 value greater than the SAD16 value, then adopts 16 * 16 predictive modes.

3. intra-frame prediction method as claimed in claim 1 is characterized in that, the preferred SAD criterion of said intra-frame prediction method for 4 * 4 predictive modes, adopts SAD4 as the mode evaluation standard, specifically suc as formula (1):

$\mathrm{SAD}4=\underset{i=0}{\overset{i=3}{\mathrm{\Σ}}}\underset{j=0}{\overset{j=3}{\mathrm{\Σ}}}|\mathrm{Psrc}(i,j)-\mathrm{Ppre}(i,j)|$ Formula (1)

For 16 * 16 predictive modes, adopt SAD16, specifically suc as formula (2):

$\mathrm{SAD}16=\underset{i=0}{\overset{i=15}{\mathrm{\Σ}}}\underset{j=0}{\overset{j=15}{\mathrm{\Σ}}}|\mathrm{Psrc}(i,j)-\mathrm{Ppre}(i,j)|$ Formula (2)

In formula (1) and the formula (2), Psrc representes source pixel, and Ppre representes predict pixel, | x| representes to ask absolute value.

4. infra-frame prediction device; It is characterized in that; It comprises control module, predictor computation module, mode adjudging module, order mapping block, residual error output module; Control module is carried out infra-frame prediction control based on adjusted sub-piece processing order to macro block, and controls the water operation of other modules; The mode adjudging module is accomplished the accumulation calculating of sad value; And the judgement optimum prediction mode adopts parallel organization to realize the calculating of SAD; Carry out the SAD judgement of macro block based on the SAD result calculated, upgrade optimal mode register and best sad value, and the predictive mode of conclusive judgement macro block employing; The order mapping block is realized the mapping treatment of sub-piece memory address; The residual error output module is accomplished the output of residual error parameter after prediction is accomplished; Predictor computation module realizes the calculating of pixel predictors under the various predictive modes.

5. infra-frame prediction device as claimed in claim 4; It is characterized in that; Said predictor computation module comprises data distributor, control unit, first processing unit to the, 16 processing units, the first predicted value buffering area, the second predicted value buffering area; Data distributor, first processing unit to the, 16 processing units, the first predicted value buffering area, the second predicted value buffering area are connected with control unit, and first processing unit to the, 16 processing units are connected between data distributor and the first predicted value buffering area, the second predicted value buffering area.

Images