CN103402087A - Video encoding and decoding method based on gradable bit streams - Google Patents
Video encoding and decoding method based on gradable bit streams Download PDFInfo
- Publication number
- CN103402087A CN103402087A CN2013103117968A CN201310311796A CN103402087A CN 103402087 A CN103402087 A CN 103402087A CN 2013103117968 A CN2013103117968 A CN 2013103117968A CN 201310311796 A CN201310311796 A CN 201310311796A CN 103402087 A CN103402087 A CN 103402087A
- Authority
- CN
- China
- Prior art keywords
- interest
- bit stream
- information
- area
- coding
- 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
Images
Abstract
The invention provides a video encoding and decoding method based on gradable bit streams. The method comprises the following steps of firstly, grading an input original video sequence, extracting an interesting area, and according to the interesting area, dividing an original video into an interesting area, non-interesting information, a divided residual error and an object identifier; secondly, encoding the interesting area, the non-interesting information, the divided residual error and the object identifier into an interesting area bit stream, a non-interesting information bit stream, a divided residual error bit stream and an object identifier bit stream in a gradable mode by different encoding methods; and finally, writing grading information of the original video and identifier information of each bit stream into bit stream, wherein in a decoding process, respectively decoding the interesting area bit stream, the non-interesting information bit stream, the divided residual error bit stream and the object identifier bit stream, and an interesting area image sequence, a background unchanged image sequence, an original texture-respected image sequence and target identifier information-added image sequence can be reconstructed in a gradable mode by combining interesting area position information or target identifier information which are acquired after decoding.
Description
Technical field
The invention belongs to the video coding and decoding technology in the digital media processing technical field, be specifically related to a kind of video coding-decoding method of gradable bit stream.
Background technology
Along with the development of video coding technique and deepening continuously of people's research, video compression coding has had significant progress.But in the face of huge camera chain, growing video data and people are for the high Qinghua of video system, automation, intelligentized demand, and it is unable to do what one wishes that present coding techniques seems again.
The purpose of video compression is to guarantee to remove the video redundancy under certain video quality, facilitating storage, the transmission of video data.But present video coding technique is not treated with a certain discrimination the content in video sequence.In video sequence, monitor video especially, people may be only interested in the object of the motions such as the people in video sequence, car, and can suitably reduce its compression quality as zones such as road, metopes.So not only the quality of area-of-interest can be improved, compression ratio can also be further improved.
For huge video data, if want to find the information that needs, need very large manpower and time.While as the police, solving a case the checking monitoring video recording, need artificially to check one by one in a large amount of videos, waste time and energy.If video sequence can remove the image of the moving objects such as nobody, car when storage, only keep the image that comprises area-of-interest, can save a large amount of manpowers and time.
Summary of the invention
For above-mentioned deficiency of the prior art, the object of the invention is to propose a kind of video coding-decoding method based on gradable bit stream.The method can be divided into different levels by the original video of input, and information at different levels is adopted to different coding methods, and hierarchical coding is different bit stream.In decode procedure, this method is supported the gradable bit stream of decoding not, selectively, with the gradable different video sequence that reconstructs.For video sequence, especially monitor video, can be divided into it area-of-interest and non-area-of-interest.Can be by the original series Delamination Transmission.Being indifferent in the less situation of non-area-of-interest and bandwidth, can only transmit area-of-interest; Also area-of-interest and non-area-of-interest together can be transmitted, decoding end can minute rank, reconstruct selectively different image sequences.
from above-mentioned thought, the present invention proposes a kind of video coding-decoding method based on gradable bit stream, it is characterized in that, at coding side, at first this method carries out classification to the original video sequence of input, extract area-of-interest, according to area-of-interest, original video can be divided into to area-of-interest, non-information of interest, divide residual error, object identity, then to area-of-interest, non-information of interest, divide residual error and adopt different coding methods from object identity, hierarchical coding is the area-of-interest bit stream, non-information of interest bit stream, divide the residual error bit stream, the object identity bit stream.In decoding end, this method is decoded respectively to input code flow, can classification, image sequence that selectively reconstruct area-of-interest image sequence, the constant image sequence of background, the image sequence of respecting original texture and interpolation object id information.
Area-of-interest described in the present invention refers to the zone of the object or person appointment of the motions such as people in video sequence, car; Non-information of interest refers to the image of removing after area-of-interest, and the area-of-interest part of wherein removing can be filled and obtain with the appropriate section of other images or self image, and non-information of interest can be also the background that obtains from training input video sequence; Divide residual error and refer to the difference between original image and non-information of interest; Object identity refers to for representing the mask image of original image object outline.
A kind of video coding-decoding method based on gradable bit stream that the present invention proposes, cataloged procedure comprises:
A) rating information extracts: the original video sequence to input carries out region of interesting extraction, according to area-of-interest, original video is divided into to various level information.
B) rating information coding: the various level information of extracting during rating information is extracted adopts different coding strategies.
C) supplementary coding: the identification information of the rating information of original video and every kind of bit stream is write to bit stream.
The mode that the described rating information of cataloged procedure extracts described region of interesting extraction includes but not limited to: automatically extract area-of-interest, manually mark area-of-interest and manually mark and automatically extract the mode that combines.
It is that original video sequence is classified into to area-of-interest, non-information of interest, divides at least a in residual error, object identity that the described rating information of cataloged procedure extracts.
A kind of optional scheme that the described rating information of cataloged procedure extracts is: the original video sequence to input carries out background modeling, using the background after modeling as non-information of interest; Difference to original image and background image is carried out cluster, obtains area-of-interest; The difference of original image and background image is as dividing residual error; For representing that the mask images of area-of-interest object outline is as object identity.
The data of the described encoding region of interest of cataloged procedure, at least comprise a kind of in following data: the pixel data of area-of-interest, the width of area-of-interest, the height of area-of-interest, the coordinate information of area-of-interest in image, the area-of-interest number, the attribute of object in area-of-interest, the information of event in area-of-interest.
The non-data of interest that described rating information extracts comprises a kind of in following data at least: remove the image after area-of-interest, the area-of-interest of wherein removing part can zero setting, also can use other data stuffings; Non-information of interest can be also the background model that obtains from training input video sequence; Non-information of interest can be also the background image of input.
The division residual error that described rating information extracts, at least comprise a kind of in following data: the difference of the background image of original image and background model or input, the difference of the background image of region of interest area image and background model or input, the difference of the image of removal area-of-interest and the background image of background model or input, the difference of adjacent two width images.
The object identity that described rating information extracts, comprise a kind of in following data: the mask images of object outline in the expression area-of-interest, the coordinate information of indicated object profile at least.
The hierarchical coding process comprises: encoding region of interest, non-information of interest coding, division residual coding and object identity coding.
When the described rating information of cataloged procedure coding is namely encoded respectively to original video at different levels, can adopt different coding parameters, form different bit streams.
The different coding parameter that described each classification to original video of hierarchical coding process is adopted is a kind of in reference frame selection, reference frame number, motion search range, method for searching motion, sub-pixel precision, model selection, quantization parameter at least.
Encoding region of interest described in described rating information cataloged procedure is selected more reference frame, and wherein the reference frame set can comprise background image; Larger hunting zone, the high-quality coding mode that quantization parameter is less.
Non-information of interest coding described in described rating information cataloged procedure is only selected frame mode coding, the high-quality that quantization parameter is less or lossless coding pattern.
Division residual coding described in described rating information cataloged procedure is selected less reference frame, less hunting zone, the low-quality coding mode that quantization parameter is larger.
The less reference frame of object identity codes selection described in described rating information cataloged procedure, less hunting zone, the low-quality coding mode that quantization parameter is larger.
The number that relates to reference frame during coding, the hunting zone of estimation, quantization parameter etc.In the application, this two-stage of area-of-interest and non-information of interest adopts more high-quality coded system, and this two-stage of residual sum object identity adopts medium or low-quality coded system.
Reference frame: find match block in how many width images.The number of reference frame and the coding standard that uses and user's the relation that disposes.Such as standard H.264 allows at most 16 reference frames.Consider compression performance and complexity, generally get 3 reference frames.More can be 5, and less can be 1.Here the more less relativeness when this level Four bit stream coding that refers to.
In like manner hunting zone, be generally 32 or 64 pixels, than I to be 16, more greatly can with 128. quantization parameter scopes be 0-51. than I to be 10-15, medium can be 25-30, can be more greatly 40-50.
In a word, only mean a relative concept here.
In cataloged procedure, encoder, except coded image data, also writes bit stream by the sign of original video rating information and every kind of bit stream.
Decode procedure comprises:
A) decoding incoming bit stream: the bit stream to input is decoded respectively.
B) decoding rating information: decoding obtains the rating information of original video.
C) reconstructed image sequence: gradable, selectable reconstruction video image sequence.
Incoming bit stream in the described decoding incoming bit stream of decode procedure is area-of-interest bit stream, non-information of interest bit stream, divide at least a in residual error bit stream and object identity bit stream.
The described decoding incoming bit stream of decode procedure comprises area-of-interest bitstream decoding, the decoding of non-information of interest, divides residual error bitstream decoding and object identity bitstream decoding.
In decode procedure, the described area-of-interest bitstream decoding of decoding incoming bit stream is except decode image data, can also obtain the width of area-of-interest, the height of area-of-interest, the coordinate information of area-of-interest in image, the area-of-interest number, the attribute of object in area-of-interest, at least a information in area-of-interest in the information of event.
In decode procedure, decoder except decode image data, the rating information of the original video that also needs to decode.
The described reconstruction original image of decode procedure, the coordinate information that decoder need to obtain by decoding in the area-of-interest bit stream or the object mask information in the object identity bit stream, the image that the decoding of area-of-interest and non-information of interest is obtained merges, and obtains the constant image sequence of background.If exist, divide the residual error bit stream, the constant image of background after rebuilding merges with the division residual image that decoding obtains, and obtains to respect the image sequence of original texture.
Described decoding incoming bit stream, comprise the area-of-interest bit stream, and non-information of interest bit stream is divided residual error bit stream and object identity bit stream, can independently decode.
Beneficial effect
The present invention can realize the graduated encoding of video data, is beneficial to the classification storage, and, according to object id information, can do and follow the tracks of fast and detect the high Qinghua of realizing video, automation, intellectuality.
The accompanying drawing explanation
The flow chart of Fig. 1 cataloged procedure of the present invention;
A kind of implementation method schematic diagram of Fig. 2 region of interesting extraction of the present invention;
Fig. 3 decode procedure flow chart of the present invention;
The described image reconstruction flow chart of Fig. 4 decode procedure of the present invention.
Embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.
the present invention proposes a kind of video coding-decoding method based on gradable bit stream (algorithm frame as shown in Figure 1), at first this method carries out classification to the original video sequence of input, extract area-of-interest, according to area-of-interest, original video can be divided into to area-of-interest, non-information of interest, divide residual error, object identity, then to area-of-interest, non-information of interest, divide residual error and adopt different coding methods from object identity, hierarchical coding is the area-of-interest bit stream, non-information of interest bit stream, divide the residual error bit stream, the object identity bit stream, finally the identification information of the rating information of original video and every kind of bit stream is write to bit stream.In decode procedure, this method decode respectively area-of-interest bit stream, non-information of interest bit stream, divide residual error bit stream and object identity bit stream, in conjunction with decoded area-of-interest coordinate information or object id information, can classification, image sequence that selectively reconstruct area-of-interest image sequence, the constant image sequence of background, the image sequence of respecting original texture and interpolation region of interest domain representation information.
As shown in Figure 1, a kind of cataloged procedure of the video coding-decoding method based on gradable bit stream comprises:
Rating information extraction step S1, carry out region of interesting extraction to the original video sequence of input, and extracting mode can be automatic extraction, can be manual mark, or both modes of combining.According to area-of-interest, rating information can be extracted as to area-of-interest, non-information of interest, divide the residual sum object identity.
Rating information coding step S2, encode respectively to area-of-interest, non-information of interest, division residual error and object identity, generates area-of-interest bit stream, non-information of interest bit stream, divides residual error bit stream, object identity bit stream.
Supplementary coding step S3, write bit stream by the identification information of original video rating information and every kind of bit stream.
As shown in Figure 3, a kind of decode procedure of the video coding-decoding method based on gradable bit stream comprises:
Decoding incoming bit stream step S4, this step is decoded respectively to area-of-interest bit stream, non-information of interest bit stream, division residual error bit stream and object identity bit stream.
Supplementary decoding step S5, this step decoding obtains the rating information of original video and the identification information of every kind of bit stream.
Image reconstruction step S6, this step is by coordinate information or the object id information of area-of-interest, reconstructed image.
Reconstructed image step S6 wherein comprises:
Judged whether the object id information step, the rating information according to decoding obtains, judge whether to exist the object identity bit stream.If there is object id information, enter according to interested area information fused images step; If there is no object id information, enter according to region of interest location information domain fused images step.
According to interested area information fused images step, this step is utilized decoded object identity image, according to object id information, region of interest area image and background image is merged, and reconstructs the constant image of background.
According to area-of-interest coordinate information fused images step, this step is utilized decoded region of interest domain representation, according to the area-of-interest coordinate information, region of interest area image and background image is merged, and reconstructs the constant image of background.
Judge whether to divide the residual information step, according to the rating information that decoding obtains, judged whether to exist division residual error bit stream.If exist, divide residual information, enter to merge and divide the residual error step; If there is no divide residual information, current reconstructed image is final reconstructed image.
Merge and divide the residual error step, the division residual image information that this step obtains according to decoding, merge the constant image of background after rebuilding and division residual image the image sequence that obtains respecting original texture.
Below cite an actual example to illustrate a kind of possible implementation of a kind of video coding-decoding method based on gradable bit stream of the present invention.
Be input as the monitor video sequence of YUV4:2:0 form, as shown in Figure 2, obtain the background model of list entries as non-information of interest.Can training with the parts of images of list entries of background model.Ask for original image and background image difference, obtain to divide residual image, to dividing residual image, carry out cluster and threshold decision, obtain the area-of-interest in sequence.Here original video be classified as to area-of-interest, non-information of interest and divide residual error.During decoding, area-of-interest bit stream, non-information of interest bit stream are decoded respectively with division residual error bit stream, according to the coordinate information of area-of-interest in original image of decoding and obtaining in the area-of-interest bit stream, decoded region of interest area image and background image are merged, and then merge with decoded division residual image, obtain to rebuild original image.
For above realization, carried out following test: 4 monitor video sequences, to encoding original series and area-of-interest and non-information of interest difference encoding ratio.Detailed results is in Table 1.
Table 1 comparing result
Finally it should be noted that: obviously, above-described embodiment is only for example of the present invention clearly is described, and is not the restriction to execution mode.For those of ordinary skill in the field, can also make other changes in different forms on the basis of the above description.Here exhaustive without also giving all execution modes.And the apparent variation of being amplified out thus or change still are among protection scope of the present invention.
Claims (21)
1. video coding-decoding method based on gradable bit stream is characterized in that: cataloged procedure is to comprise area-of-interest bit stream, non-information of interest bit stream, divide at least a of residual error bit stream and object identity bit stream by the original video sequence hierarchical coding of input; The decode procedure minute rank of corresponding video bit stream and selectively decoding comprise at least a of area-of-interest bit stream, non-information of interest bit stream, division residual error bit stream and object identity bit stream, to reconstruct the decoded video that meets the different application demand.
2. video coding-decoding method according to claim 1, it is characterized in that: described cataloged procedure comprises:
(1) rating information extracts: the original video sequence to input carries out region of interesting extraction, according to area-of-interest, original video is divided into to the information of different levels;
(2) rating information coding: the various level information of extracting during rating information is extracted adopts different coding strategies;
(3) supplementary coding: the bit stream of the rating information of original video and every one-level sign is write to bit stream.
3. video coding-decoding method according to claim 2, it is characterized in that: in described cataloged procedure, region of interesting extraction mode described in described rating information extracts comprises: automatically extract area-of-interest, manually mark area-of-interest and manually mark and automatically extract the mode that combines.
4. video coding-decoding method according to claim 2, it is characterized in that: in described cataloged procedure, it is that original video sequence is classified into to area-of-interest, non-information of interest, divides at least a in the residual sum object identity that described rating information extracts.
5. video coding-decoding method according to claim 4, it is characterized in that: in described cataloged procedure, the regions of interest data that described rating information extracts, comprise a kind of in following data, the pixel data of area-of-interest at least, the width of area-of-interest, the height of area-of-interest, the area-of-interest coordinate information in image, the area-of-interest number, the attribute of object in area-of-interest, the information of event in area-of-interest.
6. video coding-decoding method according to claim 4, it is characterized in that: the non-data of interest that described rating information extracts, at least comprise a kind of in following data: remove the image after area-of-interest, the area-of-interest part of wherein removing can zero setting, also can enough other data stuffings; Non-information of interest can be also the background model that obtains from training input video sequence; Non-information of interest can be also the background image of input.
7. video coding-decoding method according to claim 4, it is characterized in that: the division residual error that described rating information extracts, at least comprise a kind of in following data: the difference of the background image of original image and background model or input, the difference of the background image of region of interest area image and background model or input, the difference of the image of removal area-of-interest and the background image of background model or input, the difference of adjacent two width images.
8. video coding-decoding method according to claim 4 is characterized in that: the object identity that described rating information extracts comprises a kind of in following data, the mask images of object outline in the expression area-of-interest, the coordinate information of indicated object profile at least.
9. video coding-decoding method according to claim 2, is characterized in that: when described rating information is encoded the data of extracting, can adopt different coding parameters, form different bit streams; These coding parameters are that reference frame selection, reference frame number, motion are searched plain scope, motion and searched at least a in plain method, sub-pixel precision, model selection and quantization parameter at least.
10. video coding-decoding method according to claim 2 is characterized in that: described rating information cataloged procedure comprises encoding region of interest, non-information of interest coding, divide at least a in residual coding and object identity coding.
11. video coding-decoding method according to claim 2, it is characterized in that: the encoding region of interest described in described rating information cataloged procedure is selected more reference frame, wherein the reference frame set comprises background image, larger hunting zone, the high-quality coding mode that quantization parameter is less.
12. video coding-decoding method according to claim 2 is characterized in that: the non-information of interest coding described in described rating information cataloged procedure is only selected frame mode coding, the high-quality that quantization parameter is less or lossless coding pattern.
13. video coding-decoding method according to claim 2 is characterized in that: the division residual coding described in described rating information cataloged procedure is selected less reference frame, less hunting zone, the low-quality coding mode that quantization parameter is larger.
14. video coding-decoding method according to claim 2 is characterized in that: the less reference frame of object identity codes selection described in described rating information cataloged procedure, less hunting zone, the low-quality coding mode that quantization parameter is larger.
15. video coding-decoding method according to claim 2 is characterized in that: encoder is except coded image data, also by the original video rating information and identify each bit stream and write bit stream for the information of which kind of bit stream.
16. video coding-decoding method according to claim 1 is characterized in that: decode procedure comprises,
(1) the decoding incoming bit stream, decode respectively to the bit stream of input;
(2) decoding rating information, decoding obtains the rating information of original video;
(3) reconstructed image sequence, can classification, selectable reconstruction video image sequence.
17. video coding-decoding method according to claim 16 is characterized in that: the incoming bit stream in described decoding incoming bit stream is area-of-interest bit stream, non-information of interest bit stream, divide at least a in residual error bit stream and object identity bit stream.
18. video coding-decoding method according to claim 16, it is characterized in that: the area-of-interest bitstream decoding described in described decoding incoming bit stream is except decode image data, can also obtain the width of area-of-interest, the height of area-of-interest, the coordinate information of area-of-interest in image, the area-of-interest number, the attribute of object in area-of-interest, at least a information in area-of-interest in the information of event.
19. video coding-decoding method according to claim 16 is characterized in that: decoder is except decode image data, and the rating information of the original video that also needs to decode and every kind of bit stream are the information of which kind of bit stream.
20. video coding-decoding method according to claim 16, it is characterized in that: in described reconstruction original image process, the coordinate information that decoder need to obtain by decoding in the area-of-interest bit stream or the object mask information in the object identity bit stream, the image that the decoding of area-of-interest and non-information of interest is obtained merges, and obtains the constant image sequence of background; Exist and divide the residual error bit stream, the constant image of background after rebuilding merges with the division residual image that decoding obtains, and obtains to respect the image sequence of original texture.
21. video coding-decoding method according to claim 16 is characterized in that: described decoding incoming bit stream, comprise the area-of-interest bit stream, non-information of interest bit stream, divide residual error bit stream and object identity bit stream, can both independently decode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013103117968A CN103402087A (en) | 2013-07-23 | 2013-07-23 | Video encoding and decoding method based on gradable bit streams |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013103117968A CN103402087A (en) | 2013-07-23 | 2013-07-23 | Video encoding and decoding method based on gradable bit streams |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103402087A true CN103402087A (en) | 2013-11-20 |
Family
ID=49565602
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013103117968A Pending CN103402087A (en) | 2013-07-23 | 2013-07-23 | Video encoding and decoding method based on gradable bit streams |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103402087A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104168482A (en) * | 2014-06-27 | 2014-11-26 | 中安消技术有限公司 | Method and device for video coding and decoding |
| CN104486625A (en) * | 2014-12-30 | 2015-04-01 | 北京中星微电子有限公司 | Video coding method used for SVAC standard |
| CN105830446A (en) * | 2014-01-03 | 2016-08-03 | 高通股份有限公司 | Methods for coding an inter-layer reference picture set (RPS) and coding end of bitstream (EOB) network access layer (NAL) units in multi-layer coding |
| CN106162191A (en) * | 2015-04-08 | 2016-11-23 | 杭州海康威视数字技术股份有限公司 | A kind of method for video coding based on target and system |
| CN106162190A (en) * | 2015-04-08 | 2016-11-23 | 杭州海康威视数字技术股份有限公司 | A kind of video image compression and the method for reading, Apparatus and system |
| CN103813169B (en) * | 2014-02-19 | 2017-07-21 | 北京大学 | Telescopic object expression method and device in Video Codec |
| CN108833915A (en) * | 2018-06-21 | 2018-11-16 | 杭州雄迈集成电路技术有限公司 | A kind of intelligent coding method of adaptive frame skipping |
| WO2018209549A1 (en) * | 2017-05-16 | 2018-11-22 | 深圳市乃斯网络科技有限公司 | Terminal video interval division method and system |
| CN112187288A (en) * | 2015-03-02 | 2021-01-05 | 三星电子株式会社 | Transmitting apparatus and receiving apparatus |
| WO2021248349A1 (en) * | 2020-06-10 | 2021-12-16 | Plantronics, Inc. | Combining high-quality foreground with enhanced low-quality background |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050175251A1 (en) * | 2004-02-09 | 2005-08-11 | Sanyo Electric Co., Ltd. | Image coding apparatus, image decoding apparatus, image display apparatus and image processing apparatus |
| CN101102495A (en) * | 2007-07-26 | 2008-01-09 | 武汉大学 | A video image decoding and encoding method and device based on area |
| CN101534444A (en) * | 2009-04-20 | 2009-09-16 | 杭州华三通信技术有限公司 | Image processing method, system and device |
-
2013
- 2013-07-23 CN CN2013103117968A patent/CN103402087A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050175251A1 (en) * | 2004-02-09 | 2005-08-11 | Sanyo Electric Co., Ltd. | Image coding apparatus, image decoding apparatus, image display apparatus and image processing apparatus |
| CN101102495A (en) * | 2007-07-26 | 2008-01-09 | 武汉大学 | A video image decoding and encoding method and device based on area |
| CN101534444A (en) * | 2009-04-20 | 2009-09-16 | 杭州华三通信技术有限公司 | Image processing method, system and device |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105830446B (en) * | 2014-01-03 | 2019-03-29 | 高通股份有限公司 | The method and apparatus being decoded for the video information to multilayer bit stream |
| CN105830446A (en) * | 2014-01-03 | 2016-08-03 | 高通股份有限公司 | Methods for coding an inter-layer reference picture set (RPS) and coding end of bitstream (EOB) network access layer (NAL) units in multi-layer coding |
| CN103813169B (en) * | 2014-02-19 | 2017-07-21 | 北京大学 | Telescopic object expression method and device in Video Codec |
| CN104168482A (en) * | 2014-06-27 | 2014-11-26 | 中安消技术有限公司 | Method and device for video coding and decoding |
| CN104486625A (en) * | 2014-12-30 | 2015-04-01 | 北京中星微电子有限公司 | Video coding method used for SVAC standard |
| CN104486625B (en) * | 2014-12-30 | 2019-08-02 | 中星技术股份有限公司 | A kind of method for video coding for SVAC standard |
| CN112187288B (en) * | 2015-03-02 | 2023-09-01 | 三星电子株式会社 | Transmitting apparatus and receiving apparatus |
| CN112187288A (en) * | 2015-03-02 | 2021-01-05 | 三星电子株式会社 | Transmitting apparatus and receiving apparatus |
| CN106162191A (en) * | 2015-04-08 | 2016-11-23 | 杭州海康威视数字技术股份有限公司 | A kind of method for video coding based on target and system |
| CN106162190A (en) * | 2015-04-08 | 2016-11-23 | 杭州海康威视数字技术股份有限公司 | A kind of video image compression and the method for reading, Apparatus and system |
| WO2018209549A1 (en) * | 2017-05-16 | 2018-11-22 | 深圳市乃斯网络科技有限公司 | Terminal video interval division method and system |
| CN108833915A (en) * | 2018-06-21 | 2018-11-16 | 杭州雄迈集成电路技术有限公司 | A kind of intelligent coding method of adaptive frame skipping |
| WO2021248349A1 (en) * | 2020-06-10 | 2021-12-16 | Plantronics, Inc. | Combining high-quality foreground with enhanced low-quality background |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103402087A (en) | Video encoding and decoding method based on gradable bit streams | |
| Guarda et al. | Adaptive deep learning-based point cloud geometry coding | |
| CN102395029B (en) | Video encoding and decoding method and device supporting retractable video browse | |
| CN105306945B (en) | A kind of scalable concentration coding method of monitor video and device | |
| CN106231214A (en) | High-speed cmos sensor image based on adjustable macro block approximation lossless compression method | |
| CN103179402A (en) | Video compression coding and decoding method and device | |
| Wang et al. | Scalable facial image compression with deep feature reconstruction | |
| CN105426883A (en) | Video classified rapid identification method and device | |
| CN116233445B (en) | Video encoding and decoding processing method and device, computer equipment and storage medium | |
| CN110290386B (en) | Low-bit-rate human motion video coding system and method based on generation countermeasure network | |
| CN103338376A (en) | Video steganography method based on motion vector | |
| CN102081795B (en) | Automatic deblocking method based on sparse representation | |
| CN103020138A (en) | Method and device for video retrieval | |
| CN101330611B (en) | Method and apparatus for embedding and erasing video watermark as well as system for processing watermark | |
| CN105704489B (en) | A kind of adaptive video motion vector steganography method based on macro block complexity | |
| CN101237581B (en) | H.264 compression domain real time video object division method based on motion feature | |
| CN106791828A (en) | High performance video code-transferring method and its transcoder based on machine learning | |
| CN105163093B (en) | A kind of compressed domain video abstract extraction method and device towards AVS monitoring shelves | |
| CN110677644B (en) | Video coding and decoding method and video coding intra-frame predictor | |
| CN105933705A (en) | HEVC (High Efficiency Video Coding) decoded video subjective quality evaluation method | |
| CN116600119B (en) | Video encoding method, video decoding method, video encoding device, video decoding device, computer equipment and storage medium | |
| CN110677624B (en) | Monitoring video-oriented foreground and background parallel compression method based on deep learning | |
| Huang et al. | Beyond VVC: Towards perceptual quality optimized video compression using multi-scale hybrid approaches | |
| CN105072445B (en) | A kind of video moving object detection method based on block truncation coding compression domain | |
| CN113691818B (en) | Video target detection method, system, storage medium and computer vision terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20131120 |