CN108134917A - A kind of video-meeting method based on Intel QSV technologies - Google Patents
A kind of video-meeting method based on Intel QSV technologies Download PDFInfo
- Publication number
- CN108134917A CN108134917A CN201711381327.8A CN201711381327A CN108134917A CN 108134917 A CN108134917 A CN 108134917A CN 201711381327 A CN201711381327 A CN 201711381327A CN 108134917 A CN108134917 A CN 108134917A
- Authority
- CN
- China
- Prior art keywords
- vpp
- video
- qsv
- intel
- technologies
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/14—Systems for two-way working
- H04N7/15—Conference systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/222—Studio circuitry; Studio devices; Studio equipment
- H04N5/262—Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects ; Cameras specially adapted for the electronic generation of special effects
- H04N5/265—Mixing
Abstract
The present invention provides a kind of video-meeting method based on Intel QSV technologies, including a plurality of decoders, encoder and VPP, specifically comprises the following steps：N terminal opens video conference, and wherein n is the integer more than or equal to 2；After the video data of each terminal is decoded by decoder, there is n surface picture；If n is 4, VPP synthesizes 4 surface pictures；It is no, then it is destructed to VPP progress, and VPP is reconfigured according to terminal number, surface pictures are synthesized later；By the surface pictures after synthesis by encoder output to each terminal, the image quality of meeting picture is improved, reduces the utilization rate of CPU.
Description
Technical field
The present invention relates to a kind of video-meeting methods based on Intel QSV technologies.
Background technology
Distinguish two kinds in existing video conferencing system, a kind of is to walk the mode that DSP is solved firmly, but the price of DSP occupies height
Under not；Also a kind of is the mode for using the soft solutions of CPU, but the encoding and decoding limited capacity of CPU, can not be expired using a CPU
The demand of foot one or more meeting, causes meeting picture to have some setbacks.
Invention content
The technical problem to be solved in the present invention is to provide a kind of video-meeting method based on Intel QSV technologies, carry
The image quality of high meeting picture reduces the utilization rate of CPU.
The invention is realized in this way：A kind of video-meeting method based on Intel QSV technologies, including a plurality of decodings
Device, encoder and VPP, specifically comprise the following steps：
Step 1, n terminal open video conference, and wherein n is the integer more than or equal to 2；
Step 2, each terminal video data be decoded by decoder after, have n surface picture；
If step 3, n are 4, VPP synthesizes 4 surface pictures；It is no, then it is destructed to VPP progress, and root
VPP is reconfigured according to terminal number, later synthesizes surface pictures；
Surface pictures after synthesis are passed through encoder output to each terminal by step 4.
Further, the step 2 is specially further：The video data of each terminal is decoded by decoder,
According to the output frame rate of VPP, the surface for having decoded is stored in respectively in corresponding buffer queue.
Further, building-up process is further in the step 3：If one or more terminals do not decode normally,
A picture is set as its corresponding surface picture, the surface for having decoded is taken out from buffer queue later and draws
Face is synthesized.
Further, the surface pictures taken away in buffer queue are deleted.
The invention has the advantages that：A kind of video-meeting method based on Intel QSV technologies of the present invention, this method base
In Intel QSV technologies, since the CPU of Intel has integrated GPU functions, using QSV technologies, efficiently use integrated GPU come into
The encoding and decoding of row video, so as to effectively reduce the utilization rate of CPU, and the processing for image of GPU is also significantly larger than the energy of CPU
Power can effectively improve the number formulary of meeting in this way, and CPU then can preferably do operational operation.
Description of the drawings
The present invention is further illustrated in conjunction with the embodiments with reference to the accompanying drawings.
Fig. 1 is the schematic diagram of the Conference Framework the present invention is based on QSV.
Fig. 2 is that the surface of the video-meeting method the present invention is based on Intel QSV technologies synthesizes schematic diagram.
Fig. 3 is reconfigures VPP flow charts the present invention is based on the video-meeting method of Intel QSV technologies.
Fig. 4 is the synthetic schemes of the video-meeting method the present invention is based on Intel QSV technologies.
Specific embodiment
The present invention is based on the video-meeting method of Intel QSV technologies, including a plurality of decoders, encoder and VPP,
Specifically comprise the following steps：
Step 1, n terminal open video conference, and wherein n is the integer more than or equal to 2；
Step 2, each terminal video data be decoded by decoder, according to the output frame rate of VPP, will solve
The surface of code out is stored in respectively in corresponding buffer queue；
If step 3, n are 4, VPP synthesizes 4 surface pictures；It is no, then it is destructed to VPP progress, and root
VPP is reconfigured according to terminal number, later synthesizes surface pictures, if one or more terminals do not decode normally,
A picture is then set as its corresponding surface picture, takes out the surface for having decoded and from buffer queue later
Picture is synthesized；The surface pictures taken away in buffer queue are deleted；
Surface pictures after synthesis are passed through encoder output to each terminal by step 4.
A kind of specific embodiment of the present invention：
1st, the Conference Framework based on QSV
As shown in Figure 1, it is a frame of QSV, wherein decoder, encoder, VPP is to use QSV technologies.Wherein
VPP (Video Processing) is the API set that QSV is provided, which is mainly used as the place of some video datas
Reason, for example synthesize, enhance, scaling, cutting etc..QSV is coded/and the data of decoding/VPP operation are all conceptualized as surface.
Inside QSV, the interaction of all resources, being all based on surface, this is abstract.For the realization of meeting, mainly using QSV
Coding and decoding is carried out, reaches quick coding-decoding operation.Synthesis to conference layout and meeting picture are realized using VPP
The operations such as enhancing.
2nd, VPP synthesis explanations
For meeting, maximum demand is exactly to synthesize.Therefore the API of VPP set, has just played compared here
Important function.As shown in Fig. 2, with the layout of a 2x2, mono- picture synthesis of VPP is briefly described：
After decoder, there is the surface of 4, be surface1~surface4 respectively.Successively according to the suitable of 2x2
Surface1~surface4 is carried out VPP operation by sequence.When the synthesis for meeting 4 sprites, VPP can then export surface
out.The picture being namely finally synthesizing.Lower explanation can be carried out with following pseudocode：
3rd, layout switching explanation
Above, we can find out from pseudocode, and VPP is set layout in the dust when initialization,
Exactly initialize 4 layouts, then the surface of needs reaches 4, will export the surface after synthesis.And in meeting
In, it may be desired to layout is switched over.So do when switching, VPP will be carried out destructed, and reconfigure VPP, such as
Shown in Fig. 3.
4th, frame per second synchronizes
After being accessed due to each terminal, frame per second is not quite similar.Possible terminal access is 15 frames, the other is 20
Frame also has higher 30 frame.And VPP is synthesized at this time, can only be fixed same frame per second.We are closed with VPP
Into the frame per second of output a simple illustration is done for 20.And there are 3 sprites to be synthesized, frame per second is respectively 15 frames, 20 frames,
25 frames.With 1 second, that is, being synthesized to this picture for 1000ms parsed.
| VPP (20 frame) | 15 frames | 20 frames | 25 frames | |
| One frame time | 50ms | 66.7ms | 50ms | 40ms |
From upper table, it can be seen that in addition to input frame rate is 20 frames, can be exported with VPP synchronous.In addition, all
It cannot synchronize.Therefore relate to the different input of these frame per second of how desynchronizing here.As shown in figure 4, processing
Flow is as follows：
Time shaft is subject to VPP's.Since VPP must have enough inputs, could normally export (under 3 people layout, it must
There must be the input of 3 surface).And due to the input of 15 frames, VPP is 50ms in the time of first time synthesis, at this time 15 frames
Picture, there are no normal decodings to come out, therefore there is no available surface.Therefore we realize LOGO's
Surface substitutes this gap.So first picture of VPP synthesis, the input of 15 frames is by the surface of a LOGO
Picture substitutes.And the input of 20 frames then just successfully decoded, first frame can be got.25 frames, 40ms when
It waits, successfully decoded first frame.All decodings, are all first put into buffer queue, fall to ensure that queue is discontented with, so
The queue of each input is carried out cleaning operation by VPP after synthetic operation has been done：In addition to synthetic frame, remaining is all cleaned up.
And decoder is after the decoding, this synthetic frame can be all placed on to queue end, new picture is in the head of queue.Ensure so follow-up
VPP is queue heads, that is, newest picture come what is when synthesis, taken.Input, is closed at the 100ms moment as 15 frames
Into, it is decoded to come there is no new picture, it is synthesized in this way, that is, the picture of first time successfully decoded.And as 25
Frame input, decode the second frame in 100ms.So it is used, be in fact the picture of the second frame (namely most
New picture).It can simply be summarized as：When there is no successfully decoded, exported using the surface of LOGO to replace frame.And
Successfully decoded, then its nearest decoded surface is obtained to carry out VPP synthesis.
Although specific embodiments of the present invention have been described above, those familiar with the art should manage
Solution, our described specific embodiments are merely exemplary rather than for the restriction to the scope of the present invention, are familiar with this
The equivalent modification and variation that the technical staff in field is made in the spirit according to the present invention, should all cover the present invention's
In scope of the claimed protection.
Claims (4)
1. a kind of video-meeting method based on Intel QSV technologies, it is characterised in that：Including a plurality of decoders, encoder
And VPP, specifically comprise the following steps：
Step 1, n terminal open video conference, and wherein n is the integer more than or equal to 2；
Step 2, each terminal video data be decoded by decoder after, have n surface picture；
If step 3, n are 4, VPP synthesizes 4 surface pictures；It is no, then it is destructed to VPP progress, and according to end
End number reconfigures VPP, later synthesizes surface pictures；
Surface pictures after synthesis are passed through encoder output to each terminal by step 4.
2. a kind of video-meeting method based on Intel QSV technologies according to claim 1, it is characterised in that：It is described
Step 2 is specially further：The video data of each terminal is decoded by decoder, will according to the output frame rate of VPP
Decoded surface out is stored in respectively in corresponding buffer queue.
3. a kind of video-meeting method based on Intel QSV technologies as described in claim 1, it is characterised in that：The step
Building-up process is further in rapid 3：If one or more terminals do not decode normally, it is corresponding as its to set a picture
Surface pictures take out the surface pictures for having decoded and from buffer queue, are synthesized later.
4. a kind of video-meeting method based on Intel QSV technologies as claimed in claim 3, it is characterised in that：It will caching
The surface pictures taken away in queue are deleted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711381327.8A CN108134917A (en) | 2017-12-20 | 2017-12-20 | A kind of video-meeting method based on Intel QSV technologies |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711381327.8A CN108134917A (en) | 2017-12-20 | 2017-12-20 | A kind of video-meeting method based on Intel QSV technologies |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108134917A true CN108134917A (en) | 2018-06-08 |
Family
ID=62392027
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711381327.8A Pending CN108134917A (en) | 2017-12-20 | 2017-12-20 | A kind of video-meeting method based on Intel QSV technologies |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108134917A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108924471A (en) * | 2018-09-13 | 2018-11-30 | 福建星网智慧科技股份有限公司 | A kind of conference system based on QSV fast video encoding and decoding |
| CN109451313A (en) * | 2018-12-14 | 2019-03-08 | 深圳市网心科技有限公司 | A kind of method for video coding, system and electronic equipment and storage medium |
| CN109862362A (en) * | 2018-12-29 | 2019-06-07 | 深圳市网心科技有限公司 | A kind of MPEG2 coding method, system, medium and the equipment of video |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101257607A (en) * | 2008-03-12 | 2008-09-03 | 中兴通讯股份有限公司 | Multiple-picture processing system and method for video conference |
| US20140321554A1 (en) * | 2013-04-26 | 2014-10-30 | Kin-Hang Cheung | Transcoding management techniques |
| CN104768063A (en) * | 2015-04-07 | 2015-07-08 | 天脉聚源（北京）教育科技有限公司 | Video coding method and device |
| CN105162809A (en) * | 2015-10-23 | 2015-12-16 | 北京理工大学 | Real-time video transcoding system based on intermodule communication |
| CN106789853A (en) * | 2015-11-24 | 2017-05-31 | 中国移动通信集团公司 | The dynamic dispatching method and device of a kind of transcoder |
| CN107277430A (en) * | 2017-07-27 | 2017-10-20 | 福建星网智慧科技股份有限公司 | A kind of video flowing carousel method of video conferencing system |
-
2017
- 2017-12-20 CN CN201711381327.8A patent/CN108134917A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101257607A (en) * | 2008-03-12 | 2008-09-03 | 中兴通讯股份有限公司 | Multiple-picture processing system and method for video conference |
| US20140321554A1 (en) * | 2013-04-26 | 2014-10-30 | Kin-Hang Cheung | Transcoding management techniques |
| CN104768063A (en) * | 2015-04-07 | 2015-07-08 | 天脉聚源（北京）教育科技有限公司 | Video coding method and device |
| CN105162809A (en) * | 2015-10-23 | 2015-12-16 | 北京理工大学 | Real-time video transcoding system based on intermodule communication |
| CN106789853A (en) * | 2015-11-24 | 2017-05-31 | 中国移动通信集团公司 | The dynamic dispatching method and device of a kind of transcoder |
| CN107277430A (en) * | 2017-07-27 | 2017-10-20 | 福建星网智慧科技股份有限公司 | A kind of video flowing carousel method of video conferencing system |
Non-Patent Citations (1)
| Title |
|---|
| 郭献: "基于 Intel平台的H.264硬件加速", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108924471A (en) * | 2018-09-13 | 2018-11-30 | 福建星网智慧科技股份有限公司 | A kind of conference system based on QSV fast video encoding and decoding |
| CN109451313A (en) * | 2018-12-14 | 2019-03-08 | 深圳市网心科技有限公司 | A kind of method for video coding, system and electronic equipment and storage medium |
| CN109862362A (en) * | 2018-12-29 | 2019-06-07 | 深圳市网心科技有限公司 | A kind of MPEG2 coding method, system, medium and the equipment of video |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108134917A (en) | A kind of video-meeting method based on Intel QSV technologies | |
| JP2521010B2 (en) | Method and apparatus for displaying multiple video windows | |
| US20050147166A1 (en) | Decoding device, electronic apparatus, computer, decoding method, program, and recording medium | |
| CN105165009B (en) | A kind of system of more visualization component Screen sharings, device and method | |
| RU2014105187A (en) | METHODS, DEVICES AND SYSTEMS FOR PARALLEL CODING AND DECODING OF VIDEO INFORMATION | |
| CN102388616A (en) | Image signal decoding device, image signal decoding method, image signal encoding device, image signal encoding method, and program | |
| CN105245880B (en) | Video file cutting method and distributed trans-coding method | |
| CN105681826B (en) | Video-frequency data transmission system and method and video data decoding method | |
| US20200413079A1 (en) | System and method for temporal differencing with variable complexity | |
| CN109862357A (en) | Cloud game image encoding method, device, equipment and the storage medium of low latency | |
| US20060062311A1 (en) | Graceful degradation of loop filter for real-time video decoder | |
| CN102215399B (en) | Frame buffer compression method, video encoder and system of video encoder | |
| Bauer et al. | The MPEG-4 multimedia coding standard: Algorithms, architectures and applications | |
| CN103503444A (en) | Signaling number of active layers in video coding | |
| CN104041050B (en) | Multithreading texture decoder | |
| CA2585157A1 (en) | Video processing | |
| BR9813605A (en) | Process and apparatus for cleaning compressed mpeg video bit streams | |
| CN105681893A (en) | Method and device for decoding stream media video data | |
| Iwasaki et al. | Single-chip MPEG-2 422P@ HL CODEC LSI with multi-chip configuration for large scale processing beyond HDTV level | |
| CN102883213B (en) | Subtitle extraction method and device | |
| CN101822053A (en) | Encoding method and device for cartoonizing natural video, corresponding video signal comprising cartoonized natural video and decoding method and device therefore | |
| JP2011019126A (en) | Image processor and image processing method | |
| CN101710994B (en) | Method and system for video decoding | |
| CN113301290B (en) | Video data processing method and video conference terminal | |
| CN110113658A (en) | A kind of ultrahigh resolution video broadcasting method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180608 |
|
| WD01 | Invention patent application deemed withdrawn after publication |