CN1753500A - Method of selecting in frame prediction mode based on H.264/AVC standard frame image - Google Patents
Method of selecting in frame prediction mode based on H.264/AVC standard frame image Download PDFInfo
- Publication number
- CN1753500A CN1753500A CN 200510086742 CN200510086742A CN1753500A CN 1753500 A CN1753500 A CN 1753500A CN 200510086742 CN200510086742 CN 200510086742 CN 200510086742 A CN200510086742 A CN 200510086742A CN 1753500 A CN1753500 A CN 1753500A
- Authority
- CN
- China
- Prior art keywords
- mode
- value
- macro block
- field picture
- satd
- 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
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
This invention puts forward a selection method for in-frame predication mode based on the H264/AVC standard frame images including a method for a brightness part and a method for color part, among which, the method for the brightness part includes: dividing the frame image to multiple macro-blocks, designing the diversity threshold value to compute the diversity of each block and compare the got diversity and the threshold value, utilizing predicted modes of different sizes based on different compared results, then computing the SATD value of each predicated mode to obtain the optimized modes, deforming and optimizing the optimized mode operation rates to determine the optimum mode finally.
Description
Technical field
The present invention relates to field of video encoding, relate to the system of selection of the intra prediction mode of the two field picture in the standard H.264/AVC particularly.
Background technology
H.264/AVC, nearly 2 years emerging video encoding and decoding standards are the video encoding and decoding standards of being developed jointly together by ITUT and MPEG of new generation.H.264/AVC the initial target of project is to wish that new codec can be under than the much lower code check of the video standard before relatively (such as MPEG-2 or H.263) (such as, half or still less) provides good video quality; Simultaneously, do not increase the coding tools of a lot of complexity, make that hardware is difficult to realize.The another one target is an adaptability, be codec can in a very wide scope, use (such as, both comprise high code check and also comprised low code check, and different video resolution), and can be (such as multicast in diverse network and system, the DVD storage, RTP/IP packet network, ITU-T mediaphone system) work.In order to achieve the above object, many new technology adopt in this standard first, the spatial prediction of intraframe coding for example, the motion estimation and compensation of variable-block size, a plurality of reference frames and adaptive binary coding, these technology have increased the H.264/AVC computational complexity of encoder greatly.In these new characteristics,, enriched the number of intra prediction mode greatly by the technology of utilization variable size block and a plurality of direction predictions.
The article of delivering in " ISO/IEC JTC1/SC29/WG11 and ITU-T SG16 are Q.6; the JVT fourth session " of holding in 2002 " relatively: H.26L intraframe coding and JPEG2000 ", result of study wherein shows, by means of these new technologies H.264/AVC the performance of intraframe coding surpassed the still image compression standard of JPEG-2000.H.264/AVC adopt the method for rate-distortion optimization (RDO) to carry out the selection of intra prediction mode in, thereby reach higher code efficiency.
Yet, in the prior art H.264/AVC coding techniques be defective.This is because foundation is the definition of middle rate-distortion optimization (RDO) H.264/AVC, when video sequence is encoded, must use all intra prediction modes respectively to test one time to two field picture, finds one then from the best predictive mode of RDO angle.
Specifically, be published in the article " making a general survey of H.264/AVC video encoding standard " of 2003 7 monthly magazine 1-19 pages or leaves of periodical " the IEEE journal of video technology circuit and system " and mention, in H.264/AVC, defined the intra prediction mode of 13 kinds of brightness parts and 4 kinds of chrominance sections altogether, they are respectively 9 kind 4 * 4 brightness parts, 4 kind 16 * 16 the brightness part and the intra prediction mode of 4 kind of 8 * 8 chrominance section.In order to select best intra prediction mode, H.264/AVC require when coding, must all carry out a RDO computing to each pattern of each two field picture.And the method for rate-distortion optimization (RDO) is by the calculation rate distortion factor (RDcost), selects optimal mode according to the result of gained, and use therein computing formula is: RDcost=D+R * λ.Wherein, what D represented is the distortion factor of corresponding predictive mode, the stream rate after the R presentation code, and what λ represented is a Lagrangian coefficient.Also just say the RDO calculating that will carry out 13 kinds of predictive modes to the brightness part of image altogether, the RDO that also will carry out 4 kinds of predictive modes to the chrominance section of image calculates.Although the image after the method for rate-distortion optimization (RDO) can guarantee to encode obtains best compression efficiency and picture quality, but because the computation complexity of rate-distortion optimization is very high, amount of calculation is huge, its computing is unusual holding time, therefore, present H.264/AVC coding techniques is not suitable for some needs occasions of using in real time, for example video conference etc.
On the other hand, in the reference coding/decoding module JM98 that JVT provides, also provide another mode selecting method, this method is based on the calculating of absolute transformed sum (SATD).The distortion factor matrix of intra prediction mode is carried out a Hadamard (Hadamard) conversion, and get the absolute value of each element of matrix after the conversion, just can obtain all absolute value additions the value of the SATD corresponding again with this predictive mode.Can find through experiment statistics, based on the candidate's that method obtained who calculates SATD predictive mode be consistent to a great extent based on the predictive mode that method obtained that calculates RDcost.That is to say that by the final optimal mode that obtains of method based on rate-distortion optimization (RDO) very big possibility being arranged is according to one in the predictive mode of the selected several preferential selections of method of calculating SATD.And compare the RDO computing, the calculating of SATD all has very remarkable advantages in amount of calculation with on operation time.
Therefore,, determine the size of predictive mode, tentatively reduce the number of candidate's predictive mode, calculate its SATD again and further obtain preferred predictive mode, at last these patterns are carried out the RDO computing, select optimal mode by the first computing macro block distortion factor.Such method can reduce the number of times of RDO computing, obtain a kind of method for choosing frame inner forecast mode of new two field picture with this, make of the standard method of this new method with respect to prior art, have on the basis that does not reduce image quality, amount of calculation and the saving significantly on operation time might realize.
Summary of the invention
(1) technical problem that will solve
The objective of the invention is to overcome the defective of prior art, a kind of method for video coding H.264/AVC standard, that can effectively reduce operand and operation time that is applied to is proposed, encoder complexity is too high in the solution prior art, operand is excessive, the problem of the application limitations that too much causes consuming time, makes that H.264/AVC the video coding of standard can have more real-time application.
(2) technical scheme
The present invention proposes a kind of method for choosing frame inner forecast mode of H.264/AVC two field picture brightness part, and it may further comprise the steps:
(1) video frame images is divided into macro block into 16 * 16 sizes of respective number;
(2) the diversity factor threshold value of setting video frame images macro block;
(3) diversity factor of each macro block of calculating video frame images;
(4) threshold value of the diversity factor of comparison step (3) gained and step (2) gained;
(5) select the intra prediction mode of different size according to different comparative results, and calculate the value (SATD) of the pairing absolute transformed sum of each predictive mode;
(6) according to the SATD value of step (5) gained, obtain preferred predictive mode;
(7) to the preferred predictive mode of step (6) gained, percent of pass aberration optimizing (RDO) is determined optimal mode.
The method for choosing frame inner forecast mode of above-mentioned two field picture brightness part wherein, for the threshold value of step (2), is not a fixed value, can adjust the occurrence of threshold value according to the requirement of different encoding and decoding picture quality and transmission rate.Wherein a kind of preferred scheme is threshold value D
0=150.
The method for choosing frame inner forecast mode of above-mentioned two field picture brightness part, a kind of preferred scheme are that formula is used in the calculating of diversity factor in the step (3)
L wherein
XyWhat represent is the brightness value of a picture element in the macro block; If with matrix of brightness value structure of the picture element in the macro block, then diversity factor Discrepancy is exactly the standard variance of this matrix.
The method for choosing frame inner forecast mode of above-mentioned two field picture brightness part, a kind of preferred scheme is that the computing formula of diversity factor can also be in the step (3)
L wherein
XyWhat represent is the brightness value of a picture element in the macro block; The value of this diversity factor represents that also if construct a matrix with the brightness value of the picture element in the macro block, then diversity factor D is exactly the variance of this matrix.
The method for choosing frame inner forecast mode of above-mentioned two field picture brightness part, a kind of preferred scheme be, the intra prediction mode of the selected different size of step (5) has 16 * 16 intra prediction mode and 4 * 4 intra prediction mode; For the macro block of diversity factor less than threshold value, select 16 * 16 intra prediction mode for use, for the macro block of diversity factor, select 4 * 4 intra prediction mode for use greater than threshold value.For the predictive mode of these two kinds of different sizes, further calculate the SATD value of each predictive mode correspondence.
The method for choosing frame inner forecast mode of above-mentioned two field picture brightness part, a kind of preferred scheme are according to the SATD value of step (6) gained, to 9 kind of 4 * 4 luma prediction modes, to get wherein 3 kinds of minimum patterns of SATD value, as preferred predictive mode; According to the SATD value of step (6) gained,, get wherein 2 kinds of minimum patterns of SATD value, as preferred predictive mode to 4 kind of 16 * 16 luma prediction modes.
The method for choosing frame inner forecast mode of above-mentioned two field picture brightness part, wherein, the operating sequence of step (1) and step (2) can be exchanged.
The present invention also proposes a kind of method for choosing frame inner forecast mode of H.264/AVC two field picture chrominance section, it is characterized in that it may further comprise the steps:
(1) video frame images is divided into macro block into 16 * 16 sizes of respective number;
(2) the SATD value of each macro block of calculating video frame images;
(3) according to the SATD value of step (2) gained, obtain preferred predictive mode;
(4) to the predictive mode of step (3) gained, percent of pass aberration optimizing (RDO) is determined optimal mode.
The method for choosing frame inner forecast mode of above-mentioned two field picture chrominance section, a kind of preferred scheme is, according to the SATD value of step (3) gained, to 4 kind of 8 * 8 chrominance section intra prediction mode, get wherein 2 kinds of minimum patterns of SATD value, as preferred predictive mode.
The present invention also proposes a kind of mode selecting method of H.264/AVC two field picture, and it comprises the method for choosing frame inner forecast mode of above-mentioned H.264/AVC two field picture brightness part and the method for choosing frame inner forecast mode of above-mentioned H.264/AVC two field picture chrominance section
(3) beneficial effect
Adopt the implementation method of the intra prediction mode selection of H.264/AVC two field picture of the present invention,, tentatively reduced the number of times of operation RDO because the computing macro block distortion factor of going ahead of the rest is determined the size of macroblock prediction pattern; And then adopt calculating further to obtain preferred predictive mode based on SATD, and in these preferred predictive modes, carry out the RDO computing more at last, therefore effectively reduced the number of times of operation RDO.Because the calculating of the distortion factor and SATD is all much simple than the computing of RDO, therefore the present invention can significantly reduce computational complexity, thereby it is consuming time significantly to have reduced coding, and the coding back can be ignored in the negative effect aspect picture quality and the bit rate video sequence.Therefore, it is too complicated and problem that can't extensive use makes that H.264/AVC standard can well be applied to real-time operation, as Network Transmission and video conference or the like well to have solved in the prior art video sequence coding H.264/AVC.
Description of drawings
The flow chart that Fig. 1 selects for the intra prediction mode that one section video sequence is carried out the brightness part;
Fig. 2 is expression, and the optimal mode that draws based on RDO is based on the individual preferably probability of one of predictive mode of n that SATD selects, its function relation figure;
Fig. 3 is for adopting mode selecting method of the present invention and the standard of employing RDO method, at the contrast situation block diagram of coding aspect consuming time.
Fig. 4 is for adopting mode selecting method of the present invention and the standard of employing RDO method, the contrast situation block diagram aspect the average image quality.
Fig. 5 is for adopting mode selecting method of the present invention and the standard of employing RDO method, the contrast situation block diagram aspect coding bit rate.
Embodiment
The method that the intra prediction mode that the H.264/AVC two field picture that the present invention proposes is sent out is selected is described as follows in conjunction with the accompanying drawings and embodiments.Following execution mode only is used to illustrate the present invention; and be not limitation of the present invention; the those of ordinary skill in relevant technologies field; under the situation that does not break away from the spirit and scope of the present invention; can also make various variations and modification; therefore all technical schemes that are equal to also belong to category of the present invention, and scope of patent protection of the present invention should be limited by each claim.
Below the experiment condition of each embodiment be:
(1) the cycle tests totalframes is 300;
(2) frame per second of cycle tests was 30 frame/seconds;
(3) every interval 5 frames are once encoded, and therefore the actual frame number that carries out the two field picture predictive coding is 50;
(4) use the infra-frame prediction mode to encode;
(5) quantization parameter is made as 28;
(6) the standard code test of contrast, (be JVT Reference Software unofficial version JM9.8, the software download address is http://bs.hhi.de/～suehring/tml/download) to use the JM98 coding/decoding module that JVT provides;
(7) video sequence that adopts of test is a JVT test model special topic group in " ISO/IEC JTC1/SC29/WG11 and the ITU-T SG16 Q.6 draft the 4th edition " of in February, 2003 issue, type sequence of recommendation in " assessment of estimation ";
(8) computer configuration of experiment employing is the CPU of P4 Celeron 2.66GHz, the internal memory of 256M.
The test of carrying out in following examples is used to contrast the result of dual mode: wherein, a kind of is to adopt JM98 to be provided with originally whole macro block modes are carried out standard code, and the implementation method of the two field picture model selection that the another kind of this paper of employing proposes is encoded.
(embodiment one)
With reference to Figure of description 1, present embodiment is used to illustrate the method for choosing frame inner forecast mode that adopts two field picture brightness part of the present invention, one section video sequence is carried out the flow process of the intra prediction mode selection of brightness part.
In the present embodiment, at first, two field picture is divided into the macro block of 16 * 16 sizes of respective number, and sets the threshold value of macro block diversity factor.
Then, each macro block is carried out the calculating of diversity factor D, in the present embodiment, the computing formula of use is
The benefit of selecting this formula is to avoid the computing of evolution.
About determining of threshold value, in the starting stage of realizing method of the present invention, carry out a large amount of tests, be used for determining to use different threshold values to adjudicate, the effect that different threshold values is corresponding different, these effects comprise the bit rate behind scramble time, the coding, the image quality value behind the coding.Take all factors into consideration after these aspects, determine threshold value.In most of the cases, can in use backward,, can directly adopt this value, and not need to test again this value as a fixed value if the user does not have special requirement.This threshold value also is adjustable, and for specific situation, for example the user is high especially to the requirement of picture quality, and perhaps the requirement of transmission rate is high especially, can adjust it on the basis of existing threshold value, to adapt to specific situation.
Adjudicate then.For the macro block of diversity factor greater than threshold value, use less size promptly 4 * 4 the size predictive mode that carries out the brightness part select; For the macro block of diversity factor less than threshold value, use bigger size promptly 16 * 16 the size predictive mode that carries out the brightness part select.After such initial option, the number of times of RDO operation obtains reducing for the first time.In the pattern of 9 kind of 4 * 4 brightness part, calculate the SATD value of each pattern, get wherein 3 kinds of patterns of SATD minimum, as preferred predictive mode.In the pattern of 4 kind of 16 * 16 brightness part, calculate the SATD value of each pattern, get wherein 2 kinds of minimum patterns of SATD value, as preferred predictive mode.Come to 2 kinds or 3 kinds through the preferred predictive mode of brightness part that calculates like this.
The chrominance section aspect, owing in H.264/AVC, only support the predictive mode of 4 kind of 8 * 8 chrominance section, therefore there is no need to utilize the diversity factor of each macro block to carry out the selection of predictive mode size, the predictive mode of relative brightness part is selected, and chrominance section must be simply.Can obtain the SATD value of each predictive mode of target macroblock according to the chroma samples value, in the present embodiment, get wherein 2 kinds of minimum predictive modes of SATD value, so preferred predictive mode is 2 kinds.
Then these preferred predictive modes are advanced shape RDO computing, obtain optimization model.Like this, the repeatedly RDO that originally need carry out only needs 2-3 time in the present embodiment.Therefore, significantly reduced operand, it is consuming time to save coding.
Table 1 has provided standard code and has adopted coding of the present invention, the number of times of operation RDO.As can be seen, adopt method of the present invention to encode, the number of times of operation RDO obviously reduces.
| Module type | Standard code | Method of the present invention |
| 4 * 4 |
9 | 3 |
| 16 * 16 brightness modules | 4 | 2 |
| 8 * 8 colourity modules | 4 | 2 |
Table 1
(embodiment two)
Present embodiment is used to illustrate the employing optimum prediction mode that prior art obtained, and just in time is to adopt based on one probability in the several preferred predictive mode that calculating obtained of SATD.Concrete experimental result is referring to Fig. 2, and the abscissa of Fig. 2 is represented the number (n) of the preferred predictive mode selected based on SATD, and ordinate is represented is based on optimum prediction mode that RDO draws and just drops on probability in this n pattern.
Present embodiment has adopted 4 type sequences to comprise " foreman ", " clarie ", " news ", " carphone ".Wherein, the number of the preferred predictive mode that transverse axis is represented to use SATD to calculate and selected, the longitudinal axis is represented under the situation of current number of modes, adopting the optimum prediction mode that prior art obtained, just in time is to adopt based on one probability in the several preferred predictive mode that calculating obtained of SATD.As can be seen from the figure, for 4 different cycle testss, when employing is selected preferred predictive mode based on the calculating of SATD, when the number of modes of selecting more than or equal to 3 the time, these preferred predictive modes and the optimal mode that adopts the prior art gained, the probability more than 90% is equal to, and wherein, for reasonable situation, sequence " news " for example, 97% probability is equal to.That is to say, for method of the present invention, employing obtains preferred predictive mode based on the calculating of SATD, when selected predictive mode number is not less than 3, again these patterns are carried out the RDO computing and obtain optimal mode, compared with the standard method of prior art, both effects are equal on very big degree.Because the calculating of SATD wants much simple with respect to the calculating of RDcost, therefore, adopts method of the present invention, can reduce operand to a great extent, saves operation time.
(embodiment three)
Present embodiment is used for relatively adopting different diversity factor threshold values, to the influential effect of image encoding.Because determining of diversity factor threshold value is very crucial in the method for the invention, so the starting stage of the present invention has been carried out a large amount of tests, obtain a large amount of test datas, at the different different coding effects that threshold value obtained.As space is limited, present embodiment has only been listed 5 different threshold value D
0, with this influence of different threshold values to final coding efficiency being described, concrete evaluating comprises difference, the influence of picture quality, the bit rate variation that coding is consuming time.Wherein, after the negative sign of the variation consuming time of encoding represents to adopt method of the present invention coding, the percentage of time of saving with respect to standard code; After the negative sign of picture quality represents to adopt method of the present invention coding, with respect to standard code at the decline decibel aspect the picture quality.
Table 2 has provided the result of present embodiment, and cycle tests is " Foreman ".As can be seen, threshold value D
0Be provided with greatly more, the scramble time is just few more, and the required time of encoding is short more, and real-time will be good more, but picture quality can variation.Therefore user or system can be provided with concrete threshold value according to the specific requirement of picture quality and the situation of network, and it is required that picture quality is consuming time with coding to take all factors into consideration present case, to reach suitable effect.
| Threshold value D 0 | Variation consuming time (%) | Picture quality changes (dB) | Bit rate variation (%) |
| 50 | -36.2 | -0.02 | 0.0 |
| 150 | -43.3 | -0.04 | 0.4 |
| 250 | -46.6 | -0.06 | 0.7 |
| 300 | -50.8 | -0.08 | 1.2 |
| 400 | -53.3 | -0.14 | 2.4 |
Table 2
As can be seen from Table 2, D
0Under=150 the situation, consuming time can the saving more than 40% of encoding, and decrease in image quality has only 0.04 decibel, and the variation of bit rate also has only 0.4%, and such negative effect is negligible.Therefore, in use backward, as long as the user does not have special requirement, D so to the coding effect
0=150 can be used as a relatively-stationary threshold value, do not need the user again retest to obtain threshold value.If the user has special requirement to the coding effect, for example need to obtain special distinct image quality, then can be at D
0Adjust this threshold value on=150 the basis, for example use D
0=50; If the user is high especially to the requirement of transmission speed, but do not require accurate picture quality, can use higher threshold value, for example D
0=400.
(embodiment four)
Present embodiment is used to illustrate the type sequence at different, the result who adopts mode selecting method of the present invention to obtain, and with respect to the standard method of prior art, the comparison aspect consuming time, the image quality loss of coding, bit rate variation.6 wherein used cycle testss are respectively " News ", " Foreman ", " Bike ", " Mobile ", " Bus ", " Silent ".Used threshold value is D in the present embodiment
0=150, from embodiment 3 as can be seen, this value is a compromise of, picture quality consuming time to encoding and bit rate variation.There is not under the situation of specific (special) requirements D
0=150 can be used as a fixed value in most of the cases uses, so has selected this value in the present embodiment for use.
Concrete test number sees Table 3.Wherein, after the negative sign of variation consuming time represents to adopt method of the present invention coding, the percentage of time of saving with respect to standard code; After the negative sign of picture quality represents to adopt method of the present invention coding, with respect to standard code at the decline decibel aspect the picture quality.
| Cycle tests | Variation consuming time (%) | Picture quality changes (dB) | Bit rate variation (%) |
| News | -30.258 | -0.04 | 0.4 |
| Foreman | -43.356 | -0.04 | 0.4 |
| Bike | -49.462 | -0.03 | 0.4 |
| Mobile | -38.031 | -0.04 | 0.4 |
| Bus | -56.226 | -0.04 | 0.5 |
| Silent | -68.547 | -0.03 | 0.4 |
Table 3
From result data as seen, be under 150 the situation in threshold value, for reasonable situation, encode and consuming timely can save 68.5% (sequence " Silent ") at most, even more unfavorable situation also can be saved 30% (sequence " News ").And no matter for which cycle tests, the influence of their picture quality all is not more than 0.04 decibel, and the influence of such degree is negligible, and the variation of bit rate is not higher than 0.5% yet, is negligible equally.As seen, adopt method of the present invention, can be under the situation of saving operand, it is consuming time significantly to reduce coding, can cause the influence of can not ignore to picture quality or bit rate hardly simultaneously.
Fig. 3 shown cycle tests " Foreman ", after adopting the RDO standard method of prior art and adopting method of the present invention that cycle tests is encoded, for different coding frame numbers, in the comparison of coding aspect consuming time.Wherein, the cylinder that has oblique line to fill is represented the situation of the standard method of prior art, and packless cylinder is represented the situation of method of the present invention.Transverse axis is represented different coding frame numbers, and unit is " frame ", and the longitudinal axis is that coding is consuming time, and unit is " second ".
Fig. 4 shown cycle tests " Foreman ", after adopting the RDO standard method of prior art and adopting method of the present invention that cycle tests is encoded, along with for different coding frame numbers, the comparison aspect picture quality.Wherein, the cylinder that has oblique line to fill is represented the situation of the standard method of prior art, and packless cylinder is represented the situation of method of the present invention, and transverse axis is represented different coding frame numbers, and unit is " frame ", and the longitudinal axis is a picture quality, and unit is " decibel ".
Fig. 5 shown cycle tests " Foreman ", after adopting the RDO standard method of prior art and adopting method of the present invention that cycle tests is encoded, along with for different coding frame numbers, the comparison aspect bit rate.Wherein, the cylinder that has oblique line to fill is represented the situation of the standard method of prior art, and packless cylinder is represented the situation of method of the present invention.Transverse axis is represented different coding frame numbers, and unit is " frame ", and the longitudinal axis is a bit rate, and unit is " Kbps ".
As can be seen, adopt method of the present invention and adopt the method for the standard RDO of prior art that cycle tests is encoded, both results are basic identical aspect picture quality and bit rate, but encode aspect consuming time, adopt method of the present invention obviously can significantly save time, scramble time of 41% to 46% of actual used consuming time method that needs standard RDO.
Claims (10)
1, a kind of method for choosing frame inner forecast mode of H.264/AVC two field picture brightness part is characterized in that it may further comprise the steps:
(1) video frame images is divided into macro block into 16 * 16 sizes of respective number;
(2) the diversity factor threshold value of setting video frame images macro block;
(3) diversity factor of each macro block of calculating video frame images;
(4) threshold value of the diversity factor of comparison step (3) gained and step (2) gained;
(5) select the intra prediction mode of different size according to different comparative results, and calculate the value (SATD) of the pairing absolute transformed sum of each predictive mode;
(6) according to the SATD value of step (5) gained, obtain preferred predictive mode;
(7) to the preferred predictive mode of step (6) gained, percent of pass aberration optimizing (RDO) is determined optimal mode.
2, the method for choosing frame inner forecast mode of two field picture brightness part as claimed in claim 1, it is characterized in that, for the threshold value of step (2), not a fixed value, can adjust the occurrence of threshold value according to the requirement of different encoding and decoding picture quality and transmission rate.
3, the method for choosing frame inner forecast mode of two field picture brightness part as claimed in claim 2 is characterized in that, formula is used in the calculating of diversity factor in the step (3)
L wherein
XyWhat represent is the brightness value of a picture element in the macro block; If with matrix of brightness value structure of the picture element in the macro block, then diversity factor Discrepancy is exactly the standard variance of this matrix.
4, the method for choosing frame inner forecast mode of two field picture brightness part as claimed in claim 2 is characterized in that, the computing formula of diversity factor can also be in the step (3)
As the method for choosing frame inner forecast mode of claim 3 or 4 described two field picture brightness part, it is characterized in that 5, the intra prediction mode of the selected different size of step (5) has 16 * 16 intra prediction mode and 4 * 4 intra prediction mode; For the macro block of diversity factor less than threshold value, select 16 * 16 intra prediction mode for use, for the macro block of diversity factor, select 4 * 4 intra prediction mode for use greater than threshold value; For the predictive mode of these two kinds of sizes, calculate the SATD value of each predictive mode correspondence.
6, the method for choosing frame inner forecast mode of two field picture brightness part as claimed in claim 1, it is characterized in that, according to the SATD value of step (5) gained, to 9 kind of 4 * 4 luma prediction modes, get wherein 3 kinds of minimum patterns of SATD value, as preferred predictive mode; To 4 kind of 16 * 16 luma prediction modes, get wherein 2 kinds of minimum patterns of SATD value, as preferred predictive mode.
7, the method for choosing frame inner forecast mode of two field picture brightness part as claimed in claim 1 is characterized in that the operating sequence of step (1) and step (2) can be exchanged.
8, a kind of method for choosing frame inner forecast mode of H.264/AVC two field picture chrominance section is characterized in that it may further comprise the steps:
(1) video frame images is divided into macro block into 16 * 16 sizes of respective number;
(2) the SATD value of each macro block of calculating video frame images;
(3) according to the SATD value of step (2) gained, obtain preferred predictive mode;
(4) to the predictive mode of step (3) gained, percent of pass aberration optimizing (RDO) is determined optimal mode.
9, the method for choosing frame inner forecast mode of two field picture chrominance section as claimed in claim 8, it is characterized in that, according to the SATD value of step (3) gained, to 4 kind of 8 * 8 chrominance frames inner estimation mode, get wherein 2 kinds of minimum patterns of SATD value, as preferred predictive mode.
10, a kind of mode selecting method of H.264/AVC two field picture, it is characterized in that, it comprises the method for choosing frame inner forecast mode of H.264/AVC two field picture brightness part as claimed in claim 1 and the method for choosing frame inner forecast mode of H.264/AVC two field picture chrominance section as claimed in claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100867421A CN100417228C (en) | 2005-10-31 | 2005-10-31 | Method of selecting in frame prediction mode based on H.264/AVC standard frame image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100867421A CN100417228C (en) | 2005-10-31 | 2005-10-31 | Method of selecting in frame prediction mode based on H.264/AVC standard frame image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1753500A true CN1753500A (en) | 2006-03-29 |
| CN100417228C CN100417228C (en) | 2008-09-03 |
Family
ID=36680149
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100867421A Expired - Fee Related CN100417228C (en) | 2005-10-31 | 2005-10-31 | Method of selecting in frame prediction mode based on H.264/AVC standard frame image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100417228C (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008067762A1 (en) * | 2006-12-04 | 2008-06-12 | Huawei Technologies Co., Ltd. | A prediction compression encoding method and system |
| WO2009015553A1 (en) * | 2007-07-31 | 2009-02-05 | Peking University Founder Group Co., Ltd. | A method and device selecting intra-frame predictive coding best mode for video coding |
| CN101087427B (en) * | 2006-06-06 | 2011-04-06 | 北京大学深圳研究生院 | A H.264 standard in-frame prediction mode selection method |
| CN101325710B (en) * | 2006-11-21 | 2012-02-15 | Vixs***公司 | Motion refinement engine with a plurality of cost calculation methods for use in video encoding and methods for use therewith |
| CN103327325A (en) * | 2013-05-13 | 2013-09-25 | 西安电子科技大学 | Intra-frame prediction mode rapid self-adaptation selection method based on HEVC standard |
| CN103404137A (en) * | 2011-01-09 | 2013-11-20 | 联发科技股份有限公司 | Apparatus and method of efficient sample adaptive offset |
| CN103442235A (en) * | 2013-09-06 | 2013-12-11 | 深圳市融创天下科技股份有限公司 | Image processing method and device |
| US9432670B2 (en) | 2011-01-09 | 2016-08-30 | Mediatek Inc. | Apparatus and method of efficient sample adaptive offset |
| CN106231307A (en) * | 2016-08-09 | 2016-12-14 | 赵羽剑 | A kind of compression of images intra-coding prediction method and hardware thereof realize |
| US9641863B2 (en) | 2011-01-09 | 2017-05-02 | Hfi Innovation Inc. | Apparatus and method of sample adaptive offset for video coding |
| CN106878710A (en) * | 2011-06-03 | 2017-06-20 | 索尼公司 | Image processing apparatus and image processing method |
| CN108777794A (en) * | 2018-06-25 | 2018-11-09 | 腾讯科技（深圳）有限公司 | The coding method of image and device, storage medium, electronic device |
| WO2023005830A1 (en) * | 2021-07-29 | 2023-02-02 | 维沃移动通信有限公司 | Predictive coding method and apparatus, and electronic device |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101888549B (en) * | 2010-06-18 | 2012-05-02 | 浙江大学 | Intra-frame 4*4 forecasting model selection method based on transform domain information |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100474932C (en) * | 2003-12-30 | 2009-04-01 | 中国科学院计算技术研究所 | Video frequency frame image fast coding method based on optimal prediction mode probability |
| CN100401789C (en) * | 2004-06-11 | 2008-07-09 | 上海大学 | Quick selection of prediction modes in H.264/AVC frame |
-
2005
- 2005-10-31 CN CNB2005100867421A patent/CN100417228C/en not_active Expired - Fee Related
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101087427B (en) * | 2006-06-06 | 2011-04-06 | 北京大学深圳研究生院 | A H.264 standard in-frame prediction mode selection method |
| CN101325710B (en) * | 2006-11-21 | 2012-02-15 | Vixs***公司 | Motion refinement engine with a plurality of cost calculation methods for use in video encoding and methods for use therewith |
| WO2008067762A1 (en) * | 2006-12-04 | 2008-06-12 | Huawei Technologies Co., Ltd. | A prediction compression encoding method and system |
| WO2009015553A1 (en) * | 2007-07-31 | 2009-02-05 | Peking University Founder Group Co., Ltd. | A method and device selecting intra-frame predictive coding best mode for video coding |
| US8406286B2 (en) | 2007-07-31 | 2013-03-26 | Peking University Founder Group Co., Ltd. | Method and device for selecting best mode of intra predictive coding for video coding |
| CN103404137B (en) * | 2011-01-09 | 2016-10-19 | 寰发股份有限公司 | The method and apparatus of effective sample adaptive equalization |
| CN103404137A (en) * | 2011-01-09 | 2013-11-20 | 联发科技股份有限公司 | Apparatus and method of efficient sample adaptive offset |
| US9641863B2 (en) | 2011-01-09 | 2017-05-02 | Hfi Innovation Inc. | Apparatus and method of sample adaptive offset for video coding |
| US9432670B2 (en) | 2011-01-09 | 2016-08-30 | Mediatek Inc. | Apparatus and method of efficient sample adaptive offset |
| CN106878710A (en) * | 2011-06-03 | 2017-06-20 | 索尼公司 | Image processing apparatus and image processing method |
| CN103327325B (en) * | 2013-05-13 | 2016-05-25 | 西安电子科技大学 | The quick self-adapted system of selection of intra prediction mode based on HEVC standard |
| CN103327325A (en) * | 2013-05-13 | 2013-09-25 | 西安电子科技大学 | Intra-frame prediction mode rapid self-adaptation selection method based on HEVC standard |
| CN103442235B (en) * | 2013-09-06 | 2017-01-25 | 融创天下（上海）科技发展有限公司 | Image processing method and device |
| CN103442235A (en) * | 2013-09-06 | 2013-12-11 | 深圳市融创天下科技股份有限公司 | Image processing method and device |
| CN106231307A (en) * | 2016-08-09 | 2016-12-14 | 赵羽剑 | A kind of compression of images intra-coding prediction method and hardware thereof realize |
| CN106231307B (en) * | 2016-08-09 | 2019-05-03 | 赵羽剑 | A kind of compression of images intra-coding prediction method and its hardware realization |
| CN108777794A (en) * | 2018-06-25 | 2018-11-09 | 腾讯科技（深圳）有限公司 | The coding method of image and device, storage medium, electronic device |
| CN108777794B (en) * | 2018-06-25 | 2022-02-08 | 腾讯科技（深圳）有限公司 | Image encoding method and apparatus, storage medium, and electronic apparatus |
| WO2023005830A1 (en) * | 2021-07-29 | 2023-02-02 | 维沃移动通信有限公司 | Predictive coding method and apparatus, and electronic device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100417228C (en) | 2008-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1753500A (en) | Method of selecting in frame prediction mode based on H.264/AVC standard frame image | |
| CN100348051C (en) | An enhanced in-frame predictive mode coding method | |
| TWI516095B (en) | Video encoding and decoding using transforms | |
| CN100338957C (en) | Complexity hierarchical mode selection method | |
| CN1809168A (en) | Prediction intra-mode selection in an encoder | |
| CN1909662A (en) | Video compression code ratio control method using uneven distribution of frame target bit number | |
| CN1784015A (en) | Inage predicting encoding method in frame | |
| CN1777283A (en) | Microblock based video signal coding/decoding method | |
| CN1535018A (en) | Coding method of video frequency image jump over macro block | |
| Shen et al. | Ultra fast H. 264/AVC to HEVC transcoder | |
| CN1753501A (en) | Method of selecting H.264/AVC frame to frame motion estimation mode | |
| CN104333754B (en) | Based on the SHVC enhancement-layer video coding methods that predictive mode is quickly selected | |
| CN1774930A (en) | Video transcoding | |
| CN1951122A (en) | Scalable video coding method supporting variable gop size and scalable video encoder | |
| CN1767644A (en) | Non-integer pixel sharing for video encoding | |
| CN1585495A (en) | Quick selection of prediction modes in H.264/AVC frame | |
| CN1589023A (en) | Coding and decoding method and device for multiple coded list lengthening based on context | |
| CN100341334C (en) | Multi-reference frame rapid movement estimation method based on effective coverage | |
| CN1859573A (en) | Predicting coding method and its system in frame | |
| Dissanayake et al. | Performance comparison of HEVC and H. 264/AVC standards in broadcasting environments | |
| CN1842162A (en) | Motion image handling method in video coding | |
| CN1921627A (en) | Video data compaction coding method | |
| CN1268136C (en) | Frame field adaptive coding method based on image slice structure | |
| CN1809167A (en) | Quick inter-frame forecast mode selection method | |
| CN1741612A (en) | Quantization method during video image coding |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080903 Termination date: 20181031 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |