CN104333754A - SHVC (scalable high efficiency video coding) enhancement layer video coding method based on rapid prediction mode selection - Google Patents
SHVC (scalable high efficiency video coding) enhancement layer video coding method based on rapid prediction mode selection Download PDFInfo
- Publication number
- CN104333754A CN104333754A CN201410609900.6A CN201410609900A CN104333754A CN 104333754 A CN104333754 A CN 104333754A CN 201410609900 A CN201410609900 A CN 201410609900A CN 104333754 A CN104333754 A CN 104333754A
- Authority
- CN
- China
- Prior art keywords
- video
- layer
- prediction mode
- enhancement layer
- enhancement
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention discloses an SHVC (scalable high efficiency video coding) enhancement layer video coding method based on rapid prediction mode selection, mainly solving the problem of low code rate of an enhancement layer in SHVC standards. The SHVC enhancement layer video coding method includes: (1), determining a correlation between a basic layer and the enhancement layer, and acquiring a prediction mode probability statistical table by probability statistics; (2), selecting a rough candidate prediction mode of the enhancement layer according to the probability statistical table; (3) respectively coding previous n frames of a video sequence by means of the rough candidate prediction mode and the Yin method to acquire a rate-distortion performance estimator BD-PSNR; (4), adjusting the rough candidate prediction mode according to the rate-distortion performance estimator BD-PSNR to acquire a final candidate prediction mode; (5), coding the video sequence according to the final candidate prediction mode. The SHVC enhancement layer video coding method has the advantages that on the premise of guaranteeing video rate-distortion performance, coding difficulty is effectively lowered and coding time is reduced; the SHVC enhancement layer video coding method can be applied to real-time video application.
Description
Technical field
The invention belongs to field of video processing, particularly a kind of predictive mode fast selecting method, can be used for video compression.
Background technology
In January, 2013, the Video coding joint working group JCT-VC set up by the international Video coding tissue of dynamic image expert group MPEG and Video Coding Experts group VCEG Liang great develops the International video coding new standard HEVC having formulated and replaced H.264/AVC.But coding standard lacks flexibility, when different terminals has different requirements for the resolution, frame per second etc. of video, just repeatedly must encode to the video flowing inputted.In order to address this problem, the basis of efficient video coding HEVC standard expands scalable video technology.
Scalable efficient video coding SHVC can by disposable for the vision signal of the input code stream being encoded into different code check, spatial resolution and video quality, to adapt to different bandwidth and different clients to the requirement of video, lower to network requirement, use share wires, the real-time, interactive of video conference can also be realized, apply comparatively general in video conference.
The conbined usage of the retractilities such as time domain, spatial domain, quality and bit-depth and various retractility is had in SHVC.Wherein, spatial domain is scalable to be referred on the basis of source code flow, produces the video flowing of multiple spatial resolution through first encoding, and the resolution of enhancement layer is than the height of Primary layer, but the picture material of each layer is identical, just spatial resolution is higher, the more clear exquisiteness of video.Original input video through down-sampling, and adopts the Video coding AVC of efficient video coding HEVC or advanced person to be encoded to become Primary layer bit stream.The inter-layer reference image that the reconstructing video of Primary layer obtains through up-sampling can be used for inter-layer prediction, uses inter-layer prediction encoding enhancement layer image can improve enhancement layer coding quality.
Because enhancement layer in SHVC with the addition of inter-layer prediction mode ILR, and all code tree unit of enhancement layer will carry out the coding of four degree of depth, each coding depth has multiple predictive mode, comprise SKIP, Inter_2N × 2N, Inter_2N × N, Inter_N × 2N, Inter_2N × nU, Inter_2N × nD, Inter_nL × 2N, Inter_nR × 2N, Intra_2N × 2N, Intra_N × N and ILR, a large amount of rate distortion costs is needed to calculate, just optimum prediction mode can be obtained after rate-distortion optimization, therefore whole process computation complexity is very high, reduce code rate, limit the practical application of standard, need to accelerate predictive mode selection course further.
So far, the predictive mode fast selecting method proposed during SHVC spatial domain is scalable mainly Peng Yin, the people such as Taoran Lu and Tao Chen are in January, 2013, what the 12nd the International video meeting that JCT-VC organizes proposed is entitled as the enhancement layer mode quick selecting method mentioned in the JCTVC-L0174 file of " Inter-layer reference picture placement ", is called Yin method in the present invention.Particular content is as follows:
(1) be I frame at cycle tests Primary layer, under enhancement layer is the condition of P frame, have two kinds of systems of selection.First method is: only select intra pattern and ILR pattern, other all inter-frame modes are not selected, and second method is: Inter_2N × 2N pattern of a choice for use merge, intra pattern and ILR pattern;
(2) cycle tests Primary layer be I frame, under enhancement layer be B frame or Primary layer is the condition of non-I frame, relative to Primary layer forecasting process, enhancement layer prediction has two changes: one is in Inter_2N × 2N pattern, got rid of the selection of inter-layer reference image by estimation, two is do ILR pattern after intra pattern.
In above-mentioned Yin method, for the special frames of enhancement layer and the P frame of black under these three kinds of configuration condition of low time delay B, low time delay P and Stochastic accessing in Fig. 1, Yin method is adopted to encode, other frames of enhancement layer adopt scalable efficient video coding SHVC method to encode, although this method can reduce the scramble time under the prerequisite of fraction distortion performance.But the method still has multitude of video frame not adopt fast method, cataloged procedure still needs to consume the plenty of time.
Summary of the invention
The object of the invention is to the deficiency for above-mentioned prior art, propose a kind of method that SHVC enhancement-layer video selected fast based on predictive mode is encoded, under the prerequisite of fraction distortion performance, reduce encoder complexity and the scramble time of all picture formats configuration.
Realize the technical scheme of the object of the invention, comprise the steps:
1., based on the SHVC enhancement-layer video coding method that predictive mode is selected fast, comprise the steps:
2. determine the corresponding relation of Primary layer coding unit and enhancement layer coding unit:
1.1) in the spatial domain of scalable efficient video coding SHVC is scalable, provide the quantization step value of Primary layer and enhancement layer, input a video sequence and carry out down-sampling, obtain two groups of videos that resolution is different, wherein, the video that resolution is little is Primary layer, and the video that resolution is large is enhancement layer;
1.2) be the lower Item unit of 4 N × N by the unitary Item dividing elements of Primary layer 2N × 2N, the coding unit of the coding unit of these 4 N × N 2N × 2N that corresponding enhancement layer 4 is adjacent respectively, N=4,8,16,32;
(2) the Primary layer optimum prediction mode of n frame and the probabilistic relation of enhancement layer optimum prediction mode before a statistics video sequence, 40 ﹤ n ﹤ 60:
2.1) N × N coding unit encoded in Primary layer is divided into 4 × 4 region, there is a kind of predictive mode in each 4 × 4 regions, the predictive mode that in statistics 4 × 4 region, access times are maximum, and using the optimum prediction mode of this pattern as Primary layer N × N coding unit;
2.2) record the optimum prediction mode of N × N coding unit of Primary layer in n frame before a video sequence and the optimum prediction mode of corresponding enhancement layer 2N × 2N coding unit, obtain the probabilistic relation of Primary layer optimum prediction mode and enhancement layer optimum prediction mode;
(3) step (1) and step (2) is repeated, obtain the Primary layer optimum prediction mode of multiple video sequence and the probabilistic relation of enhancement layer optimum prediction mode respectively, and the probabilistic relation of all video sequences is done on average, obtain predictive mode probability statistics table;
(4) in predictive mode probability statistics table, often kind of corresponding 11 kinds of enhancement layer predictive modes of Primary layer optimum prediction mode, for often kind of Primary layer optimum prediction mode, by 11 of its correspondence kinds of enhancement layer predictive modes according to probability order arrangement from big to small, if probability is greater than the predictive mode of 5% and number is no more than 7 kinds, then directly it can be used as rough candidate modes, otherwise the high front 7 kinds of candidate pattern of select probability are as rough candidate modes;
(5) with rough candidate modes, n frame before video sequence is encoded, record code check and brightness peak signal to noise ratio;
(6) by Yin method, n frame before video is encoded, record code check and brightness peak signal to noise ratio;
(7) change the quantization step value of Primary layer and enhancement layer, repeat step (1) to step (6);
(8) according to step (5) to the code check in step (7) and brightness peak signal to noise ratio, obtain distortion performance estimator BD-PSNR;
(9) judge whether rough candidate modes is final candidate modes according to the result of step (8): if meet-0.055dB ﹤ BD-PSNR ﹤-0.045dB, then rough candidate modes is exactly final candidate modes, otherwise, rough candidate modes is adjusted, its BD-PSNR is met the demands, and using the rough candidate modes after adjustment as final candidate modes;
(10) according to final candidate modes, video sequence is encoded:
10.1) input a video sequence and carry out down-sampling, obtaining Primary layer and enhancement-layer video;
10.2) first frame of Yin method to Primary layer and enhancement layer is utilized to encode;
10.3) from the second frame, the video of Primary layer and enhancement layer is encoded with diverse ways respectively: if Primary layer is non-I frame, efficient video coding HEVC method is then utilized to encode to Primary layer video, record the optimum prediction mode of all coding units, and according to the candidate modes that Primary layer optimum prediction mode finds enhancement layer final, enhancement-layer video is encoded; If when Primary layer is I frame, then perform step 10.4);
10.4) configuration condition of video sequence is judged: if the configuration condition of video sequence is low time delay, efficient video coding HEVC method is then utilized to encode to Primary layer video, the optimum prediction mode of all coding units of record Primary layer, according to the candidate modes that Primary layer optimum prediction mode finds enhancement layer final, enhancement-layer video is encoded; If the configuration condition of video sequence is Stochastic accessing, then utilizes efficient video coding HEVC method to encode to Primary layer video, utilize Yin method to encode to enhancement-layer video.
The present invention's tool compared with existing Yin method has the following advantages:
A () the present invention utilizes the scalable middle Primary layer in SHVC spatial domain this feature identical with enhancement-layer video content, the Primary layer optimum prediction mode obtained by the method for probability statistics and the probabilistic relation of enhancement layer optimum prediction mode, final candidate modes can be determined, decrease the predictive mode number of enhancement layer, under the prerequisite of fraction distortion performance, reduce encoder complexity, improve code rate;
B () the present invention owing to all having done predictive mode probability statistics to the various picture formats of video sequence, and obtains corresponding final candidate modes, therefore, it is possible to be widely used in low time delay B, low time delay P and Stochastic accessing three kinds of video sequence configuration condition.
Accompanying drawing explanation
Fig. 1 is the video frame type of existing Primary layer under three kinds of video sequence configuration condition and enhancement layer;
Fig. 2 is realization flow figure of the present invention.
Embodiment
Below in conjunction with accompanying drawing, embodiments of the present invention is described in detail.The present embodiment is implemented premised on technical solution of the present invention, gives detailed execution mode and specific operation process, but protection scope of the present invention is not limited to following embodiment.
The present invention carries out on SHM6.1, and the configuration condition of video sequence has low time delay B, low time delay P and Stochastic accessing three kinds, and the quantization step QP in often kind of situation is as shown in table 1:
Quantization step under table 1 three kinds of video sequence configuration condition
| Video sequence configuration condition | Primary layer quantization step QP0 | Enhancement layer quantization step-length QP1 |
| Low time delay B | 20～25，26～29，30～33，34～40 | QP0+0～QP0+1，QP0+2～QP0+3 |
| Low time delay P | 20～25，26～29，30～33，34～40 | QP0+0～QP0+1，QP0+2～QP0+3 |
| Stochastic accessing | 20～25，26～30，31～34，35～40 | QP0+0～QP0+1，QP0+2～QP0+3 |
With reference to Fig. 2, performing step of the present invention is as follows:
Step one: the corresponding relation determining Primary layer coding unit and enhancement layer coding unit.
1.1) provide video test sequence information, it is as shown in the table:
The specifying information of table 2 video test sequence
1.2) when video sequence configuration condition is low time delay B, quantization step QP0=(20 ~ 25), QP1=(QP0+0 ~ QP0+1), Traffic video sequence in input table 2 also carries out down-sampling, obtain two groups of videos that resolution is respectively 1280 × 800 and 2560 × 1600, wherein, resolution be 1280 × 800 video be Primary layer, resolution be 2560 × 1600 video be enhancement layer;
1.3) by the unitary Item dividing elements of Primary layer 2N × 2N be the lower Item unit of 4 N × N, the coding unit of these 4 N × N is the coding unit of 4 adjacent 2N × 2N in corresponding enhancement layer respectively, N=4,8,16,32.
Step 2: the Primary layer optimum prediction mode of 50 frames and the probabilistic relation of enhancement layer optimum prediction mode before statistics Traffic video sequence.
2.1) N × N coding unit encoded in Primary layer is divided into 4 × 4 region, there is a kind of predictive mode in each 4 × 4 regions, the predictive mode that in statistics 4 × 4 region, access times are maximum, and using the optimum prediction mode of this pattern as Primary layer N × N coding unit;
2.2) record the optimum prediction mode of N × N coding unit and the optimum prediction mode of corresponding enhancement layer 2N × 2N coding unit in the Primary layer of 50 frames before Traffic video sequence, obtain the probabilistic relation of Primary layer optimum prediction mode and enhancement layer optimum prediction mode.
Step 3: repeat step one and step 2, obtain the Primary layer optimum prediction mode of other six video sequences and the probabilistic relation of enhancement layer optimum prediction mode in table 2 respectively, i.e. PeopleOnStreet, Kimono, ParkScene, Cactus, BasketballDrive, BQTerrace, and do on average to the probabilistic relation of seven video sequences, obtain predictive mode probability statistics, as table 3:
Under table 3 low time delay B configuration condition, the probabilistic relation (%) during QP0=(20 ~ 25) QP1=(QP0+0 ~ QP0+1) between Primary layer optimum prediction mode and enhancement layer optimum prediction mode
In table 3, pattern 0 ~ pattern 10 represents these 11 kinds of predictive modes of SKIP, Inter_2N × 2N, Inter_2N × N, Inter_N × 2N, Inter_2N × nU, Inter_2N × nD, Inter_nL × 2N, Inter_nR × 2N, Intra_2N × 2N, Intra_N × N and ILR respectively.
Step 4: in the predictive mode probability statistics of table 3, for often kind of Primary layer optimum prediction mode, by 11 of its correspondence kinds of enhancement layer predictive modes according to probability order arrangement from big to small, if probability is greater than the predictive mode of 5% and number is no more than 7 kinds, then directly it can be used as rough candidate modes, otherwise the high front 7 kinds of candidate pattern of select probability are as rough candidate modes.
Step 5: encode with front 50 frames of seven video sequences in rough candidate modes respectively his-and-hers watches 2.
5.1) with efficient video coding HEVC, Primary layer video is encoded, record the optimum prediction mode of all coding units;
5.2) according to size and the index value of enhancement layer current coded unit, in the region that Primary layer finds video content corresponding, the optimum prediction mode in this region is obtained;
5.3) according to the optimum prediction mode of Primary layer, with rough candidate modes, enhancement-layer video is encoded;
5.4) code check and the brightness peak signal to noise ratio of seven video sequences is recorded respectively;
Step 6: encode with front 50 frames of seven video sequences in Yin method his-and-hers watches 2, records code check and the brightness peak signal to noise ratio of seven video sequences respectively.
Step 7: in Table 1 under other seven groups of quantization step conditions, repeats step 1.2) to step 6, obtain code check and the brightness peak signal to noise ratio of seven video sequences respectively.
Step 8: according to step 5 to the code check in step 7 and brightness peak signal to noise ratio, the description in the VCEG-AE07 meeting document of group JCT-VC proposition is combined according to Video coding, by typing two group code rate and brightness peak smnr data, and load macro document, obtain the distortion performance estimator BD-PSNR of seven video sequences, and the distortion performance estimator BD-PSNR of these seven video sequences is averaged.
Step 9: according to average rate distortion performance estimator BD-PSNR, the candidate modes that adjustment enhancement layer is rough.
9.1) judge whether rough candidate modes is final candidate modes according to the result of step 7: if meet-0.055dB ﹤ BD-PSNR ﹤-0.045dB, then rough candidate modes is exactly final candidate modes, otherwise, rough candidate modes is adjusted, performs step 9.2);
9.2) if distortion performance estimator BD-PSNR ﹤ is-0.055dB, then choose the pattern that other probability of non-rough candidate modes are the highest, as wherein a kind of rough candidate modes, but candidate modes number can not more than 7 kinds; If distortion performance estimator BD-PSNR ﹥ is-0.045dB, remove the pattern that in rough candidate modes, probability is minimum, reduce rough candidate modes number, its BD-PSNR is met the demands, and using the rough candidate modes after adjustment as final candidate modes, the final candidate modes obtained is as shown in table 4 ~ table 11:
The candidate modes that during table 4 QP0=(20 ~ 25) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 5 QP0=(20 ~ 25) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 6 QP0=(26 ~ 29) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 7 QP0=(26 ~ 29) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 8 QP0=(30 ~ 33) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 9 QP0=(30 ~ 33) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 10 QP0=(34 ~ 40) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 11 QP0=(34 ~ 40) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
9.3), when video sequence configuration condition is low time delay P, step one is performed to step 9.2), obtain final candidate modes as shown in table 12 ~ table 19:
The candidate modes that during table 12 QP0=(20 ~ 25) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 13 QP0=(20 ~ 25) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 14 QP0=(26 ~ 29) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 15 QP0=(26 ~ 29) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 16 QP0=(30 ~ 33) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 17 QP0=(30 ~ 33) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 18 QP0=(34 ~ 40) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 19 QP0=(34 ~ 40) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
9.4), when video sequence configuration condition is Stochastic accessing, step one is performed to step 9.2), obtain final candidate modes as shown in table 20 ~ table 27.
The candidate modes that during table 20 QP0=(20 ~ 25) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 21 QP0=(20 ~ 25) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 22 QP0=(26 ~ 30) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 23 QP0=(26 ~ 30) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 24 QP0=(31 ~ 34) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 25 QP0=(31 ~ 34) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
The candidate modes that during table 26 QP0=(35 ~ 40) QP1=(QP0+0 ~ QP0+1), enhancement layer is final
The candidate modes that during table 27 QP0=(35 ~ 40) QP1=(QP0+2 ~ QP0+3), enhancement layer is final
In table 4 ~ table 27, QP0 represents the value of the quantization step QP of Primary layer, QP1 represents the value of the quantization step QP of enhancement layer, and pattern 0 ~ pattern 10 represents these 11 kinds of predictive modes of SKIP, Inter_2N × 2N, Inter_2N × N, Inter_N × 2N, Inter_2N × nU, Inter_2N × nD, Inter_nL × 2N, Inter_nR × 2N, Intra_2N × 2N, Intra_N × N and ILR respectively.
Step 10: according to final candidate modes, each video sequence in his-and-hers watches 2 is encoded.
10.1), when video sequence configuration condition is low time delay B, quantization step QP0=(20 ~ 25), QP1=(QP0+0 ~ QP0+1), input video sequence also carries out down-sampling, obtains Primary layer and enhancement-layer video;
10.2) first frame of Yin method to Primary layer and enhancement layer is utilized to encode;
10.3) from the second frame, the video of Primary layer and enhancement layer is encoded with diverse ways respectively: if Primary layer is non-I frame, efficient video coding HEVC method is then utilized to encode to Primary layer video, record the optimum prediction mode of all coding units, and according to the candidate modes that Primary layer optimum prediction mode finds enhancement layer final, enhancement-layer video is encoded; If when Primary layer is I frame, then perform step 10.4);
10.4) configuration condition of video sequence is judged: if the configuration condition of video sequence is low time delay, efficient video coding HEVC method is then utilized to encode to Primary layer video, the optimum prediction mode of all coding units of record Primary layer, according to the candidate modes that Primary layer optimum prediction mode finds enhancement layer final, enhancement-layer video is encoded; If the configuration condition of video sequence is Stochastic accessing, then utilizes efficient video coding HEVC method to encode to Primary layer video, utilize Yin method to encode to enhancement-layer video, record enhancement layer coding time and whole video sequence coding time respectively;
10.5) other seven groups of quantization steps in his-and-hers watches 1, perform step 10.1 respectively) to step 10.4);
10.6) be low time delay P and Stochastic accessing two kinds of situations to video sequence configuration condition, perform step 10.1 respectively) to step 10.5), the result of the scramble time obtained compared with existing Yin method, shown in table 28.
The distortion performance of the relative Yin method of table 28 the present invention and time decreased amount
In table 28, BD-PSNR representation rate distortion performance estimator, unit is dB, and Δ Time represents the time variations that the present invention compared with the time of Yin method, and EL represents that the scramble time of enhancement layer compares, and Total represents that the scramble time of whole video sequence compares.
As can be seen from Table 28, when the configuration condition of video sequence is low time delay B, when distortion performance estimator BD-PSNR decreased average 0.066dB, enhancement layer coding time decreased 37.77%, whole process code time decreased 30.14%; When the configuration condition of frequency sequence is low time delay P, when distortion performance estimator BD-PSNR decreased average 0.059dB, enhancement layer coding time decreased 41.22%, whole process code time decreased 32.98%; When the configuration condition of frequency sequence is Stochastic accessing, when distortion performance estimator BD-PSNR decreased average 0.052dB, enhancement layer coding time decreased 45.90%, whole process code time decreased 36.58%.Wherein, under the configuration condition of Stochastic accessing, the distortion performance of video compression is best, and cataloged procedure speed-raising is the fastest, and illustrate under Stochastic accessing configuration condition, the correlation of Primary layer and enhancement layer is maximum; And under the configuration condition of low time delay B, the correlation of Primary layer and enhancement layer is minimum.
In sum, the present invention utilizes the correlation of Primary layer and enhancement-layer video, selects the candidate modes that enhancement layer is final, encodes to enhancement-layer video, is finally made comparisons the scramble time of this fast method and Yin method.Obtain conclusion by experiment, when the excursion of average BD-PSNR is-0.066dB ~-0.052dB, this fast method makes enhancement layer average coding time reduce 37.77% ~ 45.90%, and the ensemble average scramble time of video reduces 30.14% ~ 36.58%.Therefore, under three kinds of video sequence configuration condition, the present invention on the basis of fraction distortion performance, can reduce encoder complexity effectively, improves code rate, can be used for real-time video application.
Foregoing description is preferred embodiment of the present invention, and obvious researcher in this field can make various amendment and replacement with reference to preferred embodiment of the present invention and accompanying drawing to the present invention, and these amendments and replacement all should fall within protection scope of the present invention.
Claims (3)
1., based on the SHVC enhancement-layer video coding method that predictive mode is selected fast, comprise the steps:
(1) corresponding relation of Primary layer coding unit and enhancement layer coding unit is determined:
1.1) in the spatial domain of scalable efficient video coding SHVC is scalable, provide the quantization step value of Primary layer and enhancement layer, input a video sequence and carry out down-sampling, obtain two groups of videos that resolution is different, wherein, the video that resolution is little is Primary layer, and the video that resolution is large is enhancement layer;
1.2) be the lower Item unit of 4 N × N by the unitary Item dividing elements of Primary layer 2N × 2N, the coding unit of the coding unit of these 4 N × N 2N × 2N that corresponding enhancement layer 4 is adjacent respectively, N=4,8,16,32;
(2) the Primary layer optimum prediction mode of n frame and the probabilistic relation of enhancement layer optimum prediction mode before a statistics video sequence, 40 ﹤ n ﹤ 60:
2.1) N × N coding unit encoded in Primary layer is divided into 4 × 4 region, there is a kind of predictive mode in each 4 × 4 regions, the predictive mode that in statistics 4 × 4 region, access times are maximum, and using the optimum prediction mode of this pattern as Primary layer N × N coding unit;
2.2) record the optimum prediction mode of N × N coding unit of Primary layer in n frame before a video sequence and the optimum prediction mode of corresponding enhancement layer 2N × 2N coding unit, obtain the probabilistic relation of Primary layer optimum prediction mode and enhancement layer optimum prediction mode;
(3) step (1) and step (2) is repeated, obtain the Primary layer optimum prediction mode of multiple video sequence and the probabilistic relation of enhancement layer optimum prediction mode respectively, and the probabilistic relation of all video sequences is done on average, obtain predictive mode probability statistics table;
(4) in predictive mode probability statistics table, often kind of corresponding 11 kinds of enhancement layer predictive modes of Primary layer optimum prediction mode, for often kind of Primary layer optimum prediction mode, by 11 of its correspondence kinds of enhancement layer predictive modes according to probability order arrangement from big to small, if probability is greater than the predictive mode of 5% and number is no more than 7 kinds, then directly it can be used as rough candidate modes, otherwise the high front 7 kinds of candidate pattern of select probability are as rough candidate modes;
(5) with rough candidate modes, n frame before video sequence is encoded, record code check and brightness peak signal to noise ratio;
(6) by Yin method, n frame before video is encoded, record code check and brightness peak signal to noise ratio;
(7) change the quantization step value of Primary layer and enhancement layer, repeat step (1) to step (6);
(8) according to step (5) to the code check in step (7) and brightness peak signal to noise ratio, obtain distortion performance estimator BD-PSNR;
(9) judge whether rough candidate modes is final candidate modes according to the result of step (8): if meet-0.055dB ﹤ BD-PSNR ﹤-0.045dB, then rough candidate modes is exactly final candidate modes, otherwise, rough candidate modes is adjusted, its BD-PSNR is met the demands, and using the rough candidate modes after adjustment as final candidate modes;
(10) according to final candidate modes, video sequence is encoded:
10.1) input a video sequence and carry out down-sampling, obtaining Primary layer and enhancement-layer video;
10.2) first frame of Yin method to Primary layer and enhancement layer is utilized to encode;
10.3) from the second frame, the video of Primary layer and enhancement layer is encoded with diverse ways respectively: if Primary layer is non-I frame, efficient video coding HEVC method is then utilized to encode to Primary layer video, record the optimum prediction mode of all coding units, and according to the candidate modes that Primary layer optimum prediction mode finds enhancement layer final, enhancement-layer video is encoded; If when Primary layer is I frame, then perform step 10.4);
10.4) configuration condition of video sequence is judged: if the configuration condition of video sequence is low time delay, efficient video coding HEVC method is then utilized to encode to Primary layer video, the optimum prediction mode of all coding units of record Primary layer, according to the candidate modes that Primary layer optimum prediction mode finds enhancement layer final, enhancement-layer video is encoded; If the configuration condition of video sequence is Stochastic accessing, then utilizes efficient video coding HEVC method to encode to Primary layer video, utilize Yin method to encode to enhancement-layer video.
2. the SHVC enhancement-layer video coding method selected fast based on predictive mode according to claim 1, the rough candidate modes of the utilization wherein described in step (5) is encoded to n frame before video sequence, carries out as follows:
2a) with efficient video coding HEVC, Primary layer video is encoded, record the optimum prediction mode of all coding units;
2b) according to size and the index value of enhancement layer current coded unit, in the region that Primary layer finds video content corresponding, obtain the optimum prediction mode in this region;
2c) according to the optimum prediction mode of Primary layer, with rough candidate modes, enhancement-layer video is encoded.
3. the SHVC enhancement-layer video coding method selected fast based on predictive mode according to claim 1, adjusting rough candidate modes wherein described in step (9) is compared distortion performance estimator BD-PSNR and setting threshold:
If distortion performance estimator BD-PSNR ﹤ is-0.055dB, then choose the pattern that other probability of non-rough candidate modes are the highest, as wherein a kind of rough candidate modes, but candidate modes number can not more than 7 kinds;
If distortion performance estimator BD-PSNR ﹥ is-0.045dB, remove the pattern that in rough candidate modes, probability is minimum, reduce rough candidate modes number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410609900.6A CN104333754B (en) | 2014-11-03 | 2014-11-03 | Based on the SHVC enhancement-layer video coding methods that predictive mode is quickly selected |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410609900.6A CN104333754B (en) | 2014-11-03 | 2014-11-03 | Based on the SHVC enhancement-layer video coding methods that predictive mode is quickly selected |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104333754A true CN104333754A (en) | 2015-02-04 |
| CN104333754B CN104333754B (en) | 2017-06-13 |
Family
ID=52408393
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410609900.6A Active CN104333754B (en) | 2014-11-03 | 2014-11-03 | Based on the SHVC enhancement-layer video coding methods that predictive mode is quickly selected |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104333754B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104853188A (en) * | 2015-03-12 | 2015-08-19 | 康佳集团股份有限公司 | Selection control method and system for SVC rapid coding mode in wireless network |
| CN107948649A (en) * | 2016-10-12 | 2018-04-20 | 北京金山云网络技术有限公司 | A kind of method for video coding and device based on subjective quality model |
| CN108235026A (en) * | 2018-02-01 | 2018-06-29 | 重庆邮电大学 | A kind of fast encoding method of spatial scalable |
| CN108259898A (en) * | 2018-02-01 | 2018-07-06 | 重庆邮电大学 | Fast encoding method in frame based on Quality scalable video coding QSHVC |
| CN109120926A (en) * | 2017-06-23 | 2019-01-01 | 腾讯科技（深圳）有限公司 | Predicting mode selecting method, device and video encoder |
| CN113225553A (en) * | 2021-04-18 | 2021-08-06 | 南京理工大学 | Method for predicting optimal threshold point in high dynamic video double-layer backward compatible coding system |
| CN113709492A (en) * | 2021-08-25 | 2021-11-26 | 重庆邮电大学 | SHVC (scalable video coding) spatial scalable video coding method based on distribution characteristics |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109040753B (en) * | 2018-08-17 | 2020-12-22 | 广州酷狗计算机科技有限公司 | Prediction mode selection method, device and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101562739A (en) * | 2009-06-02 | 2009-10-21 | 北京大学 | Video coding processing method and video coding processing device |
| CN102025995A (en) * | 2010-12-10 | 2011-04-20 | 浙江大学 | Spatial enhancement layer rapid mode selection method of scalable video coding |
| CN103402085A (en) * | 2013-07-19 | 2013-11-20 | 哈尔滨工业大学深圳研究生院 | Mode selection method and system based on video scalable coding compression |
| US20140205009A1 (en) * | 2013-01-21 | 2014-07-24 | The Regents Of The University Of California | Method and apparatus for spatially scalable video compression and transmission |
-
2014
- 2014-11-03 CN CN201410609900.6A patent/CN104333754B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101562739A (en) * | 2009-06-02 | 2009-10-21 | 北京大学 | Video coding processing method and video coding processing device |
| CN102025995A (en) * | 2010-12-10 | 2011-04-20 | 浙江大学 | Spatial enhancement layer rapid mode selection method of scalable video coding |
| US20140205009A1 (en) * | 2013-01-21 | 2014-07-24 | The Regents Of The University Of California | Method and apparatus for spatially scalable video compression and transmission |
| CN103402085A (en) * | 2013-07-19 | 2013-11-20 | 哈尔滨工业大学深圳研究生院 | Mode selection method and system based on video scalable coding compression |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104853188A (en) * | 2015-03-12 | 2015-08-19 | 康佳集团股份有限公司 | Selection control method and system for SVC rapid coding mode in wireless network |
| CN107948649B (en) * | 2016-10-12 | 2020-07-03 | 北京金山云网络技术有限公司 | Video coding method and device based on subjective quality model |
| CN107948649A (en) * | 2016-10-12 | 2018-04-20 | 北京金山云网络技术有限公司 | A kind of method for video coding and device based on subjective quality model |
| US10986332B2 (en) | 2017-06-23 | 2021-04-20 | Tencent Technology (Shenzhen) Company Limited | Prediction mode selection method, video encoding device, and storage medium |
| CN109120926A (en) * | 2017-06-23 | 2019-01-01 | 腾讯科技（深圳）有限公司 | Predicting mode selecting method, device and video encoder |
| CN109120926B (en) * | 2017-06-23 | 2019-08-13 | 腾讯科技（深圳）有限公司 | Predicting mode selecting method, device and medium |
| CN108259898A (en) * | 2018-02-01 | 2018-07-06 | 重庆邮电大学 | Fast encoding method in frame based on Quality scalable video coding QSHVC |
| CN108235026A (en) * | 2018-02-01 | 2018-06-29 | 重庆邮电大学 | A kind of fast encoding method of spatial scalable |
| CN108259898B (en) * | 2018-02-01 | 2021-09-28 | 重庆邮电大学 | Intra-frame fast coding method based on quality scalable video coding QSHVC |
| CN108235026B (en) * | 2018-02-01 | 2021-09-28 | 重庆邮电大学 | Space-scalable quick coding method |
| CN113225553A (en) * | 2021-04-18 | 2021-08-06 | 南京理工大学 | Method for predicting optimal threshold point in high dynamic video double-layer backward compatible coding system |
| CN113225553B (en) * | 2021-04-18 | 2022-09-06 | 南京理工大学 | Method for predicting optimal threshold point in high dynamic video double-layer backward compatible coding system |
| CN113709492A (en) * | 2021-08-25 | 2021-11-26 | 重庆邮电大学 | SHVC (scalable video coding) spatial scalable video coding method based on distribution characteristics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104333754B (en) | 2017-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104333754A (en) | SHVC (scalable high efficiency video coding) enhancement layer video coding method based on rapid prediction mode selection | |
| CN100461867C (en) | Inage predicting encoding method in frame | |
| CN100348051C (en) | An enhanced in-frame predictive mode coding method | |
| CN100417228C (en) | Method of selecting in frame prediction mode based on H.264/AVC standard frame image | |
| He et al. | H. 26L-based fine granularity scalable video coding | |
| CN103248893B (en) | From H.264/AVC standard to code-transferring method and transcoder thereof the fast frame of HEVC standard | |
| CN100568952C (en) | Based on real-time fine granular scalability method H.264 | |
| CN105681797B (en) | A kind of DVC-HEVC video transcoding methods based on prediction residual | |
| CN103916675B (en) | A kind of low latency inner frame coding method divided based on band | |
| CN101895756A (en) | Method and system for coding, decoding and reconstructing video image blocks | |
| CN100440975C (en) | Method and apparatus for improving the average image refresh rate in a compressed video bitstream | |
| KR20090066035A (en) | Method of constituting intra prediction mode using spatial edge detection | |
| Pan et al. | Frame-level Bit Allocation Optimization Based on<? brk?> Video Content Characteristics for HEVC | |
| CN104394409A (en) | Space-domain correlation based rapid HEVC (High Efficiency Video Coding) predication mode selection method | |
| CN104853191A (en) | HEVC fast coding method | |
| CN103442228A (en) | Quick frame inner transcoding method from H.264/AVC standard to HEVC standard and transcoder thereof | |
| CN100502507C (en) | Video data compaction coding method | |
| CN106101709A (en) | A kind of Primary layer inter-frame prediction method of the SHVC quality scalability combining enhancement layer | |
| KR100757208B1 (en) | Fast multiple reference frame selection method in motion estimation for H.264 video encoding | |
| CN110913232B (en) | Selection method and device of TU division mode and readable storage medium | |
| Chiu et al. | Hybrid distributed video coding with frame level coding mode selection | |
| CN114189688B (en) | Chrominance component prediction method based on luminance template matching | |
| CN110113601B (en) | HEVC intra-frame rapid algorithm selection method based on video picture texture features | |
| Al-Khafaji¹ et al. | Color image compression of inter-prediction base | |
| KR100718468B1 (en) | Method and device for video down-sampling transcoding |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |