CN1784009A - Data dividing method for increasing video transmission rubustness and relative error masking method - Google Patents
Data dividing method for increasing video transmission rubustness and relative error masking method Download PDFInfo
- Publication number
- CN1784009A CN1784009A CN 200410098269 CN200410098269A CN1784009A CN 1784009 A CN1784009 A CN 1784009A CN 200410098269 CN200410098269 CN 200410098269 CN 200410098269 A CN200410098269 A CN 200410098269A CN 1784009 A CN1784009 A CN 1784009A
- Authority
- CN
- China
- Prior art keywords
- high frequency
- coefficient
- section
- frequency coefficient
- video
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 96
- 230000005540 biological transmission Effects 0.000 title abstract description 14
- 230000001965 increasing effect Effects 0.000 title abstract description 4
- 230000000873 masking effect Effects 0.000 title description 2
- 239000000284 extract Substances 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 12
- 238000004891 communication Methods 0.000 abstract description 7
- 238000012360 testing method Methods 0.000 description 22
- 230000000694 effects Effects 0.000 description 18
- 230000006835 compression Effects 0.000 description 9
- 238000007906 compression Methods 0.000 description 9
- 238000013461 design Methods 0.000 description 9
- 230000004224 protection Effects 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 5
- 238000004088 simulation Methods 0.000 description 3
- 239000000470 constituent Substances 0.000 description 2
- 238000012827 research and development Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 241001611138 Isma Species 0.000 description 1
- 241000404883 Pisa Species 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 239000007795 chemical reaction product Substances 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000012856 packing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
This invention discloses a data division method for increasing video transmission robustness. It contains intraframe coding video, extracting plurality of important high frequency coefficient of intraframe coding video and combining to one independent first important high frequency coefficient section, locating it behind the dc coefficient section, combining residual high frequency coefficient to one independent second important high frequency coefficient section, locating it behind the first important high frequency coefficient section. For forward predication frame, extracting its dc coefficient and combining to one independent dc coefficient section, combining high coefficient to independent high coefficient section, locating it behind said dc coefficient section. Said invention provides more robust error resisting technology for communication under fallibility environment.
Description
Technical field
Present technique belongs to field of multimedia communication, relate to technology such as video compression, mistake opposing and error concealment and video transmission and terminal equipment design, be applicable to Video service quality (QoS) control of transmission system such as broadband streaming media, radio multimedium and terminal equipment.The present invention is specifically related to a kind of data dividing method that strengthens the video transmission robustness.
Background technology
In recent years, commercial gradually along with the maturation gradually of the broadband streaming media communication technology and 3G wireless communication system just becoming one of research and development focus of industrial quarters based on the multimedia communication technology of video transmission.And original video information generally need take huge bandwidth resources, must compress processing, and the video information of high compression, add shake (Jitter) and the non-QoS assurance (for broadband streaming media) of IP agreement or the decline and the noise (for radio multimedium) of wireless transmission channel of transmission IP network, its reliable transmission also faces many difficult points, need further research and exploration in the actual product research and development.
Video compression standard is one of core technology of field of multimedia communication always, and its international standard mainly is divided into MPEG-1, MPEG-2 and the MPEG-4 of two large divisions: MPEG tissue promulgation
[1]ITU-T issue H.261, H.263/+/++ and up-to-date H.264 standard
[2]These standards are when providing high compression ratio; consider the application of urgent day by day broadband streaming media and radio multimedium business; the mistake opposing technology that has added a series of enhancing video transmission robustnesss; as heavy (Packet Resynchronization) synchronously; data are cut apart (DP; Data Partitioning); reversible entropy coding (RVLC); extended coding (HEC) etc.; wherein; heavy simultaneous techniques and data dividing method are because computation complexity is less; the introducing expense is little; be easy to according to characteristics such as packing of bit stream syntax structure and the error protections of realization lack of balance, at video compression standard such as MPEG-4; H.264 wait and network flow-medium standard such as ISMA1.0
[3]Deng having obtained certain utilization, and be subjected to people's more concern in actual applications.
The normal data dividing method was analyzed in the past
The video compression standard of current main-stream has all adopted the hybrid coding model of piecemeal conversion+motion estimation/compensation+entropy coding, and the thinking of its data dividing method and structure are basic identical in each standard.Therefore, be example at methods analyst and when design with the MPEG-4 standard, but the conclusion of this paper can be applicable to as H.26x waiting other video compression standards based on the hybrid coding pattern.Fig. 1 has provided the structure of the data dividing method under the MPEG-4Simple Profile bit stream syntax.
By Fig. 1 (a)-(f) as can be known, MPEG-4 video packets (VP) is made up of an integer macro block (MB), has approximately equalised bit number, comprises the texture information (TCOEF) understood after harbour information, motion vector information (MV) and the piecemeal conversion etc.Consider that direct current (DC) coefficient occupies critical role in the texture information of intra-coded video frame (IVOP), MPEG-4 extracts the DC coefficient among the IVOP and be divided into independently section, and high frequency coefficient (AC1-63) is arranged in the back of VP continuously, use unique DC decollator (0x6B001) to distinguish between the two.And to forward prediction frame of video (PVOP), MPEG-4 normal data dividing method is divided into motion vector independently section and is arranged in the VP front portion, and all data texturings are placed on the back of VP code stream, middlely cut apart with unique motion vector (MV) decollator (0x1F001).This structure of MPEG-4 normal data dividing method has guaranteed that when the high frequency coefficient section of IVOP was made mistakes, decoder still can rely on the video information of the DC coefficient reconstruction respective quality of correct reception; And make mistakes and the MV section of cutting apart receives when correct when the texture information of PVOP, decoder can further improve the quality of video reconstruction by motion compensation error concealment technology such as (MC), thereby has effectively improved the robustness of video transmission.But the MPEG-4 data dividing method does not fully take into account the high frequency texture of IVOP when design, therefore, when video image contains more detail of the high frequency and the high frequency data texturing of IVOP when makeing mistakes, to lose detailed information in a large number based on the video image that this data dividing method is rebuild, the image recovery effects is not good.Simultaneously, to the DC coefficient of PVOP, MPEG-4 normal data dividing method does not cause enough concerns yet, needs further to improve and improve.
Summary of the invention
The present invention is directed in the past in the video compression standard data dividing method and high frequency texture is equal to treats,, proposed a kind of data dividing method-NewDP of new more robust when the shortcoming of covering poor effect when containing the erroneous frame of enriching detailed information.
According to one aspect of the present invention, texture information after the piecemeal conversion has been carried out test of significance, draw the most important a plurality of of IVOP piece conversion texture information, preferably 1 DC coefficient of 5 AC coefficients and PVOP can improve the effect of erroneous video image reconstruction by a relatively large margin, and can obtain good compromise between luminance component and chromatic component.
According to another aspect of the present invention; the new data dividing method of NewDP has been proposed; be about to a plurality of of IVOP, preferably 5 most important AC coefficients are divided into the independent protection section and the DC coefficient of PVOP are divided into the independent protection section, middlely cut apart with unique code word of knowing.
Cooperate the NewDP method to propose the corresponding error concealing method in addition.
Method of the present invention has not only been inherited in the past the normal data dividing method with the code stream constituent element characteristics by the different storied placements of importance; and with the further ad-hoc location that segments and leave code stream continuously in of the high-frequency information of texture; form new data protection section, and designed the corresponding error concealing technique.The expense that this method is introduced is little, only need to adopt the corresponding error concealing method just can obtain in actual production, to have certain commercial application prospect at decoding end product such as set-top box (STB) or 3G multimedia handset etc. than main, the objective effect of the better video reconstruction of normal data dividing method in the past.
Description of drawings
With reference to the following drawings, the present invention is made an explanation in conjunction with instantiation.
Fig. 1 shows MPEG-4 normal data dividing method structure.Wherein Fig. 1 a shows the overall structure of the video packets VP of IVOP; The motion vector MV that Fig. 1 b shows IVOP is cut apart segment structure; The texture information TCOEF that Fig. 1 c shows IVOP is cut apart segment structure; Fig. 1 d shows the overall structure of the video packets VP of PVOP; The motion vector MV that Fig. 1 e shows PVOP is cut apart segment structure; The texture information TCOEF that Fig. 1 f shows PVOP is cut apart segment structure;
Fig. 2 shows the different scanning method respectively, and wherein Fig. 2 a is Zigzag scanning; Fig. 2 b is alternately horizontal sweep; Fig. 2 c is for replacing vertical scanning;
Fig. 3 shows the texture information test of significance of IVOP and PVOP respectively.Wherein, Fig. 3 a, Fig. 3 c, Fig. 3 e show the result of luminance component (Y), U chromatic component and V chromatic component of the texture test of significance of IVOP respectively; Fig. 3 b, Fig. 3 d, Fig. 3 f show the result of luminance component (Y), U chromatic component and V chromatic component of the texture test of significance of PVOP respectively;
Fig. 4 shows the VP structure according to IVOP in the new data dividing method of the present invention.Wherein, Fig. 4 a shows the structure after IVOP adopts the NewDP method; Fig. 4 b shows most important high frequency coefficient section (AC
1-5Seg.) structure; Fig. 4 c shows time important high frequency coefficient section (TCOEF
I(AC
6-63) Seg.) and structure;
Fig. 5 shows the VP structure according to PVOP in the new data dividing method of the present invention.Wherein, Fig. 5 a shows the VP structure after PVOP adopts the NewDP method; Fig. 5 b shows the structure of DC coefficient segments (DC Seg.); Fig. 5 c shows texture coefficient segments (TCOEF
P(No DC) Seg.) structure;
Fig. 6 show MPEG-4 data dividing method (OldDP) with according to the compression efficiency of new data dividing method of the present invention (NewDP) contrast (PSNR_Y) (mean flow rate PSNR_Y:NewDP-33.85dB; OldDP-33.94dB);
Fig. 7 has provided the result (64kbps) of emulation experiment luminance component (Y); And
Fig. 8 provided the 129th frame the reconstruction visual effect (Foreman, 64kbps, BER=1e-4); Wherein, Fig. 8 a shows the vision reconstruction effect (34.00dB) of NoErr scheme; Fig. 8 b shows the vision of NoEC scheme and rebuilds effect (12.86dB); Fig. 8 c shows the vision of OldDP scheme and rebuilds effect (28.20dB); Fig. 8 e shows the vision of NewDP scheme and rebuilds effect (30.80dB);
Embodiment
Before the design new method, be necessary the test of significance of at first data texturing being segmented.Design following test of significance method, promptly when decoding, decoder only relies on texture constituent element information to be tested to come reconstructed image, the then simple zero setting of other information of texture.When selecting the AC coefficient, what the Intra encoding block adopted is the adaptive coefficient scanning method, promptly when selecting the AC coefficient, if encoder does not adopt AC coefficient prediction (ACpred_flag=0), then adopts conventional Zigzag scan method; If adopt the AC coefficient prediction, the scan mode of the prediction direction decision AC coefficient of DC coefficient, that is when the DC coefficient with its horizontal contiguous block during as the prediction piece, the AC coefficients by using replaces the vertical scanning mode, otherwise, adopt the horizontal sweep mode that replaces.And, then always adopt the Zigzag scan method to the Inter encoding block.Fig. 2 a-2c has provided Zigzag scan method, alternately the horizontal sweep method and the diagram of vertical scanning method alternately respectively.
Select for use the dissimilar test patterns of QCIF size to test repeatedly, Rate Control adopts the TM5 method, and the setting encoder bit rate is 64kbps (10 frame/second) or 128kbps (15 frame/second).The test frame number is 150 frames, and quantification manner adopts the MPEG method, and the refresh rate of IVOP was 15 frame/seconds, gets Y-PSNR (PSNR) mean value the estimating as assessment of 10 tests.Under the 64kbps to the texture information (TCOEF of IVOP
I) and the texture information (TCOEF of PVOP
P) test result see Table 1 and shown in Figure 3, the test result under the 128kbps situation similarly, in this omission.
The mean P SNR of table 1Forman cycle tests is (dB) relatively
| DConly | DC+2AC | DC+5AC | DC+8AC | WholeRec | ||
| IVOP | Y | 21.07 | 22.82 | 25.68 | 28.35 | 33.94 |
| U | 35.67 | 37.40 | 38.31 | 38.60 | 39.15 | |
| V | 35.80 | 37.97 | 39.23 | 39.85 | 40.62 | |
| PVOP | Y | 29.17 | × | 30.59 | × | 33.94 |
| U | 37.90 | × | 38.59 | × | 39.15 | |
| V | 38.49 | × | 39.67 | × | 40.62 |
Attention: DConly-only relies on the DC coefficient reconstruction; DC+nAC-relies on DC coefficient and hithermost n AC coefficient reconstruction (n=2,5,8); WholeRec-rebuilds fully; The experiment of *-not.
Fig. 3 a, Fig. 3 c and Fig. 3 e show the luminance component of IVOP respectively, the test case of U chromatic component and V chromatic component.And Fig. 3 b, Fig. 3 d and Fig. 3 f show the luminance component of PVOP respectively, the test case of U chromatic component and V chromatic component.
By table 1 and Fig. 3 as can be known, to the monochrome information among the IVOP (Y), increase along with the AC coefficient number of the most close selected DC coefficient, its performance improvement is bigger, from mean P SNR, adopt intra-frame macro block DC coefficient (IntraDC)+2AC, IntraDC+5AC, the effect that IntraDC+8AC rebuilds exceeds 1.75dB, 4.61dB and 7.28dB respectively than the effect of simple dependence DC coefficient reconstruction.But along with the increase gradually of AC coefficient number, the amplitude that its gain increases reduces gradually, compares with IntraDC+5AC, and the gain absolute value of IntraDC+2AC and IntraDC+8AC is respectively 2.86dB and 2.67dB, and they all differ 3 AC coefficients.For chromatic component, IntraDC+5AC exceeds 3～4dB than the mean P SNR that IntraDC rebuilds, but it is then very approaching that IntraDC+5AC and IntraDC+8AC rebuild effect, show when the AC number greater than 5 the time, the increase of AC coefficient is little to the performance improvement of chromatic component.For the texture information of PVOP, the increasing of AC coefficient, the effect of image reconstruction is obvious unlike IVOP, but still can improve the reconstruction quality of image.
Therefore; consider the compromise between brightness and the colourity reconstruction effect; as example; IVOP is selected for use most important 5 AC coefficients (can be 5 AC coefficients of the most close DC coefficient), selects for use the DC coefficient as the object of further protecting to PVOP, promptly proposed Fig. 4 and NewDP method structure shown in Figure 5.Wherein, Fig. 4 shows the VP structure of IVOP in the NewDP method.Fig. 4 a is the VP overall structure; Fig. 4 b is TCOEF
IMost important high frequency coefficient section (AC
1-5Seg) structure; Fig. 4 c is TCOEF
IInferior important high frequency coefficient section (AC
6-63) the Seg. structure; And Fig. 5 shows the VP structure of PVOP in the NewDP method.Fig. 5 a is the VP overall structure; Fig. 5 b is the DC coefficient section of cutting apart (DC Seg.) structure; Fig. 4 c is the PVOP high frequency coefficient section of cutting apart (TCOEF
P(NoDC) structure Seg.).
By Figure 4 and 5 as can be known, the NewDP method not only is divided into the DC coefficient of IVOP independently section, also extracts most important 5 AC coefficients, and follows the DC coefficient segments closely and be arranged in independently most important high frequency coefficient section (AC
1-5), the centre is cut apart with the DC decollator, and remaining important high frequency coefficient section (AC
6-63) then to be arranged in VP last.Consider that the MV decollator is unique in IVOP, select for use the MV decollator to cut apart most important high frequency coefficient section (AC
1-5) and time important high frequency coefficient section (AC
6-63).And for PVOP, then adopt and be similar to the processing method of normal data dividing method to IVOP, the DC coefficient is extracted be divided into independently section, and the AC coefficient is not dealt with.Because PVOP do not adopt Intra MB to refresh in cataloged procedure, the DC decollator is unique appearance in the PVOP code stream information of being discussed, and is used as the separator of DC section and AC high-frequency data section so be considered.If adopt Intra MB to refresh, only need to select the identifier of a unique appearance in PVOP to get final product in addition.
When decoding, most important high frequency coefficient section (AC in IVOP
1-5) or PVOP in DC coefficient segments when makeing mistakes, the NewDP method deteriorates to MPEG-4 normal data dividing method, its method for reconstructing also data dividing method with standard is identical.And the inferior important high frequency coefficient section (AC in IVOP
6-63) when makeing mistakes, the NewDP method relies on DC coefficient and most important high frequency coefficient section (AC1-5) reconstructed image that is correctly decoded, for the video image that contains more detailed information, it rebuilds effect will have bigger improvement.Equally, when the high frequency coefficient section of PVOP is made mistakes, can on the video image of motion compensation, superpose the again average of residual error of decoder, thereby also might recover more detailed information, be the characteristics that the NewDP data dividing method not only has MPEG-4 standard dividing method, and taken into full account the important high fdrequency component of texture information, further increased the robustness of video transmission.In addition, this method has only been introduced a spot of code word of additionally cutting apart when realizing, very little to Effect on Performance.Foreman sequence to 150 frames is tested, and adopting the mean P SNR of MPEG-4 normal data dividing method is 33.94dB, is 33.85dB and adopt the mean P SNR after the NewDP method, and both only differ 0.09dB.Its experiment the results are shown in Figure 6.
Experiment and analysis
In l-G simulation test, texture information is carried out wrong scrambling, and keep other information constant.Test channel adopts random bit error code channel, and the error rate is from 1e-3～1e-5.Cycle tests still adopts the Foreman sequence of QCIF form, and its parameter is similar to the parameter of 64kbps code check in the above-mentioned texture information test of significance experiment, and starts self adaptation 4MV estimation, and video packets length is elected 600bits as.Each error rate test 10 times is also averaged, to reduce experimental error.
At the error-detecting method of decoder end employing based on grammatical and semantic.And design following method and carry out error concealment:
(a) if the DC coefficient segments make mistakes, for IVOP, directly copy errors MB top be correctly decoded the MB data, for the situation that does not have top corresponding MB, directly copy the MB that is correctly decoded on its left side; For PVOP, then utilize the MV that is correctly decoded to carry out motion compensated reconstruction;
(b) if the MV section among the PVOP is made mistakes, directly copy the correct reconstruction MB data of last VOP correspondence;
(c) if 5 most important AC coefficient segments (AC among the IVOP
1-5Seg.) make mistakes, the method for reconstructing that then adopts normal data to cut apart directly utilizes the DC coefficient to finish reconstruction tasks;
(d) if the inferior important high-frequency data section (AC among the IVOP
6-63Seg.) make mistakes, then utilize DC and 5 most important AC coefficients to finish reconstruction; If the high-frequency data core dumped (AC among the PVOP
1-63Seg.), the average (DC) of stack residual error on the image that MC rebuilds then;
Above-described error concealment method closely links to each other with the design of NewDP algorithm, therefore, has and the different place of data dividing method in the past.A), b) and c) and conventional error concealment method similar, especially at c) under the situation, the mode of covering of NewDP algorithm promptly deteriorates to the error concealment mode under the MPEG-4 normal data dividing method situation.Because the NewDP algorithm has been made protection to most important AC coefficient segments; so at d) IVOP under the situation; the NewDP algorithm can utilize DC and 5 most important AC coefficients to finish image reconstruction; promptly rely on the basis of DC coefficient reconstruction at MPEG-4 normal data dividing method; 5 the most important AC high-frequency informations that superposeed, thus recovered the more images detail content.Equally, to d) PVOP under the situation, the average of the residual error that superposeed on the basis of motion compensation also might recover more detailed information.
In order to compare, design following different experiments scheme:
The NewDP scheme: the NewDP data dividing method that adopts this paper to propose, and adopt above-mentioned error concealment strategy;
NoEC scheme: do not adopt any error concealment measure, but add error detection,, skip to next VP and continue decoding in case the mistake of detecting is abandoned the decoding of current VP;
OIdDP scheme: adopt MPEG-4 normal data dividing method, in IVOP, do not contain the TCOEF of DC coefficient
IWhen the section of cutting apart is made mistakes, can only rely on the DC coefficient to finish reconstruction; When the DC coefficient segments was made mistakes, concealing method was with the NewDP scheme; And, work as TCOEF for PVOP
PWhen the section of cutting apart is made mistakes, the first motion compensated reconstruction;
NoErr scheme: rebuild without any mistake;
Fig. 7 has provided The simulation experiment result.As seen from the figure, the image that adopts the error concealment technology to rebuild is compared with the image that does not adopt the error concealment technology to rebuild, and performance differs greatly, the gain of the most nearly 20dB, and this shows the necessity that adopts the error concealment technology under the transmission environment of easily makeing mistakes.The NewDP data dividing method of this paper owing to taken into full account important high frequency coefficient, when the video that contains more details is rebuild, obtains to cut apart better master, objective effect than MPEG-4 normal data, at the PSNR of luminance component (PSNR
Y) on, both have the gain of 2～3dB approximately, and on visual effect, the NewDP data dividing method will be seen Fig. 8 significantly better than the reconstruction effect of MPEG-4 normal data dividing method.
Conclusion
At first the texture information to the conversion of MPEG-4 piecemeal has carried out test of significance; and the compromise DC coefficient of having selected 5 most important AC coefficients of IVOP texture information and PVOP is as the further object of protection; form new significant coefficient protection section, promptly proposed a kind of new data dividing method-NewDP.This method is MPEG compatible-4 normal data dividing method on function not only, and has taken into full account the important high fdrequency component of texture information, can design more efficiently easy bugs concealing technique at receiving terminal such as STB etc.L-G simulation test shows that the NewDP method can obtain better video and rebuild effect, thereby for video provides the more mistake opposing technology of robust in the communication under the error-prone environments, has certain commercial prospect.
List of references
[1]Weiping Li，Jens-Rainer Ohm，Mihaela van der Schaar，et al：ISO/IEC JTC1/SC29/WG11，“MPEG-4 Video Verification Model Version18.0(N3908)，”Pisa：MPEG Publisher，January 2001.
[2]Thomas Wiegand，“Text of Final Committee Draft of Joint VideoSpecification(ITU-T Rec.H.264|ISO/IEC 14496-10AVC)，”July 2002.
[3]ISMA，“Internet Streaming Media Alliance ImplementationSpecification version 1.0，”August，2001.
Claims (13)
1, a kind of data dividing method that is used for video data, this method at first is divided into video packets with video data.To the intra-coded video frame, at first the DC coefficient in the video packets is divided into independently section, and, then at first motion vector information (MV) is divided into independently section the forward predictive coded frame of video, it is characterized in that,
For the intra-coded video frame:
Extract a plurality of important high frequency coefficient of intra-coded video frame;
A plurality of important high frequency coefficient (AC with the described intra-coded video frame that extracted
1-5) be combined into an independently most important high frequency coefficient section, and be placed on after the DC coefficient section of described intra-coded video frame, and
With remaining high frequency coefficient (AC
6-63) be combined into an independently time important high frequency coefficient section, and be placed on after the described most important high frequency coefficient section;
For the forward prediction frame of video:
Extract the DC coefficient of this forward prediction frame of video and it is combined into an independently DC coefficient section, and
With high frequency coefficient (AC
1-63) be combined into an independently high frequency coefficient section, and be placed on after the described DC coefficient section.
2, data dividing method according to claim 1, it is characterized in that, in described intra-coded video frame, cut apart this DC coefficient section and described most important high frequency coefficient section with the direct current decollator, and the symbol that is used in unique appearance in the intra-coded video frame is cut apart this most important high frequency coefficient section and time important high frequency coefficient section.
3, data dividing method according to claim 1 is characterized in that, the identifier of unique appearance is the direct current decollator in described intra-coded video frame.
4, data dividing method according to claim 1 is characterized in that, in described forward prediction frame of video, cuts apart the DC coefficient section and the high frequency coefficient section of this forward prediction frame of video with the identifier of a unique appearance in this forward prediction frame of video.
5, data dividing method according to claim 1 is characterized in that, described video standard is the MPEG-4 standard.
6, data dividing method according to claim 4 is characterized in that, the identifier of unique appearance is the direct current decollator in described forward prediction frame of video.
7, data dividing method according to claim 1 is characterized in that, this method is compatible mutually on function with MPEG-4 normal data dividing method.
According to the described data dividing method of one of claim 1 to 7, it is characterized in that 8, the number of the important high frequency coefficient of the intra-coded video frame that is extracted is 5.
9, according to the described data dividing method of one of claim 1 to 7, it is characterized in that, the video packets of described intra-coded video frame comprises a plurality of macro blocks, in the most important high frequency coefficient section of described video packets, the important high frequency coefficient of each piece is placed in regular turn based on macro block and piece, the important high frequency coefficient of different macro blocks is spaced apart by the macro block header, and in described inferior important high frequency coefficient section, the inferior important high frequency coefficient of piece is placed continuously based on macro block and piece in each macro block.
10, according to the described data dividing method of one of claim 1 to 7, it is characterized in that, the video packets of described forward prediction frame of video comprises a plurality of macro blocks, in described DC coefficient section, the DC coefficient of piece is placed in regular turn based on macro block and piece in each macro block, the DC coefficient of different macro blocks is come spaced apart by the macro block header, and in described high frequency coefficient section, the high frequency coefficient of piece is placed continuously based on macro block and piece in each macro block.
According to the described data dividing method of one of claim 1 to 7, it is characterized in that 11, the step of the important high frequency coefficient of described extraction comprises: the degree of approach of DC coefficient is extracted this high frequency coefficient successively as significant coefficient based on high frequency coefficient.
12, a kind of error concealment method that is used for video data, described video data comprises a plurality of video packets, for the video packets of described intra-coded video frame, its DC coefficient is divided into independently section, a plurality of important high frequency coefficient (AC
1-5) be combined into an independently most important high frequency coefficient section, and remaining high frequency coefficient (AC
6-63) be combined into an independently time important high frequency coefficient section; And for the video packets of forward prediction frame of video: its DC coefficient is combined into an independently DC coefficient section, with and high frequency coefficient (AC
1-63) be combined into an independently high frequency coefficient section, it is characterized in that,
(the AC if the inferior important high-frequency data section in the intra-coded video frame makes a mistake
6-63), then utilize the DC coefficient section and the most important high frequency coefficient section (AC of this intra-coded video frame
1-5) finish reconstruction.
13, according to the error concealment method of claim 12, it is characterized in that (AC if the high-frequency data section in the forward prediction frame of video makes a mistake
1-63), the average (DC) of the residual error that then on the image that relies on the motion vector motion compensated reconstruction, superposes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410098269 CN1784009A (en) | 2004-12-01 | 2004-12-01 | Data dividing method for increasing video transmission rubustness and relative error masking method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410098269 CN1784009A (en) | 2004-12-01 | 2004-12-01 | Data dividing method for increasing video transmission rubustness and relative error masking method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1784009A true CN1784009A (en) | 2006-06-07 |
Family
ID=36773686
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410098269 Pending CN1784009A (en) | 2004-12-01 | 2004-12-01 | Data dividing method for increasing video transmission rubustness and relative error masking method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1784009A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008071085A1 (en) * | 2006-12-12 | 2008-06-19 | Huawei Technologies Co., Ltd. | Method and system of error concealment |
| CN109451000A (en) * | 2018-10-26 | 2019-03-08 | 西安科锐盛创新科技有限公司 | Video data management method and its equipment based on cloud storage |
-
2004
- 2004-12-01 CN CN 200410098269 patent/CN1784009A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008071085A1 (en) * | 2006-12-12 | 2008-06-19 | Huawei Technologies Co., Ltd. | Method and system of error concealment |
| CN109451000A (en) * | 2018-10-26 | 2019-03-08 | 西安科锐盛创新科技有限公司 | Video data management method and its equipment based on cloud storage |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102724498B (en) | The coding method of the scalability techniques of content-based information and equipment | |
| KR101016168B1 (en) | Method and apparatus for encoder assisted-frame rate up conversion (ea-fruc) for video compression | |
| Talluri | Error-resilient video coding in the ISO MPEG-4 standard | |
| US6339658B1 (en) | Error resilient still image packetization method and packet structure | |
| JP5007322B2 (en) | Video encoding method | |
| CN101375603B (en) | Method and apparatus for error resilience algorithms in wireless video communication | |
| CN1856112A (en) | Error elimination of compressed video transmission based on H.264 | |
| CN101313589A (en) | Redundant data encoding methods and device | |
| CN108965887A (en) | A kind of video information hiding method and device based on uncoupling between block | |
| CN107343202B (en) | Feedback-free distributed video coding and decoding method based on additional code rate | |
| CN1134174C (en) | Video Error Concealment Method in Terrestrial Digital TV Broadcasting | |
| CN1893663A (en) | Transmission protection method of multi-media communication | |
| CN1784009A (en) | Data dividing method for increasing video transmission rubustness and relative error masking method | |
| CN101064845A (en) | Leaking motion compensation process for video interesting area coding/decoding | |
| Psannis | Efficient redundant frames encoding algorithm for streaming video over error prone wireless channels | |
| Hrušovský et al. | Temporal-spatial error concealment algorithm for intra-frames in H. 264/AVC coded video | |
| Lin et al. | An error resilient coding scheme for H. 26L video transmission based on data embedding | |
| Gao et al. | Early resynchronization, error detection and error concealment for reliable video decoding | |
| Agrafiotis et al. | Hybrid key/Wyner-Ziv frames with flexible macroblock ordering for improved low delay distributed video coding | |
| Fang et al. | Efficient content-based resynchronization approach for wireless video | |
| Zhang et al. | An effective error concealment framework for H. 264 decoder based on video scene change detection | |
| El Beheiry et al. | An adaptive fast and efficient spatial error concealment technique for block-based video coding systems | |
| KR100931269B1 (en) | Real-Time Edge Detection in H.264 / ACC Compression Domain | |
| Xu et al. | A perceptual sensitivity based redundant slices coding scheme for error-resilient transmission H. 264/AVC video | |
| Jang et al. | Object-oriented source and channel coding for video messaging applications over wireless networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |