CN109889749A - A kind of 3D high definition operation recorded broadcast method - Google Patents
A kind of 3D high definition operation recorded broadcast method Download PDFInfo
- Publication number
- CN109889749A CN109889749A CN201910257890.7A CN201910257890A CN109889749A CN 109889749 A CN109889749 A CN 109889749A CN 201910257890 A CN201910257890 A CN 201910257890A CN 109889749 A CN109889749 A CN 109889749A
- Authority
- CN
- China
- Prior art keywords
- video
- gpu
- processing
- audio
- capture card
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a kind of 3D high definition operation recorded broadcast methods, include the following steps: S1, software and hardware: the video card of capture card, support CUDA with GPU, existing operation recorded broadcast equipment and corresponding software；S2, capture card collect audio-video signal, handle audio-video recording function by independent audio-video processing engine calling capture card GPU；Processing Algorithm processing is carried out to delay, reduces delay, the delay of 3D picture is made to be less than 40ms, simulation 3D picture is less than 60ms；S3,3D picture are accelerated to handle by GPU or assist CPU processing by GPU by GPU using CUDA；S4, recording video post-processing using CUDA by GPU assist CPU acceleration handle.This invention takes individual video processing engine, and the delay of 3D picture is made to be less than 40ms, and simulation 3D picture is less than 60ms.The present invention is handled 3D picture by Nvidia GPU using the acceleration of CUDA technology, picture clear and smooth.The present invention can substantially reduce the video processing time.
Description
Technical field
The present invention relates to operation recorded broadcast technologies, more particularly to a kind of 3D high definition operation recorded broadcast method.
Background technique
The medical operating of traditional mode is viewed and emulated and is learnt mainly establishment officer and learnt to operating room scene, either
Video is carried out as teaching to surgery situation to use.The many shortcomings of mode formula: first, since operating room space limits,
Cause it is open ended visit, view and emulate it is fewer in number, teaching, research efficiency it is lower；Second, since personnel can or make in a jumble
The attention for even influencing operation doctor is polluted at operating room air；Third, the operation for capital operation, particular procedure is recorded
Record due to live video personnel and does not know about researcher or student thinks the emphasis of concern, will make some positions the visual field and
Visual field range depth is inadequate；In addition, since video camera is that shielding is poor, when opening the equipment such as ray equipment, electric saw in operation,
Will be heavily disturbed, it causes video effect poor or can not record.
And it is all only need to arrange that a calm direct seeding operation equipment can be completed in operating room and meeting-place by operation recorded broadcast
Audio-visual synchronization live streaming.But current operation recorded broadcast method of crossing has the following deficiencies:
1. the delay that picture has longer time.The existing recorded broadcast method most of the time is delayed over 300ms.Preview when operation
Picture is very high to requirement of real-time, does not pay attention to slightly will cause undesirable consequence.
2. occupying system resource height.Existing recorded broadcast scheme occupies resource for computer system height, is matching the work station set low,
CPU occupation rate highest is more than 90% when recording, will cause computer crash sometimes.This invention takes independent processing engine,
The recording of video is completed by capture card GPU, is not take up cpu resource.
3. image voice owes clear and smooth.Existing recorded broadcast scheme Audio commonly reaches 2Mbps, but image blocks sometimes
, sound noise is bigger, and highest resolution is only supported to 1920x1080.
4. pair video post-processing overlong time recorded.Existing recorded broadcast scheme is general to the video processing in later period to be used
CPU processing, holding time are too long.
Summary of the invention
In view of the above drawbacks of the prior art, technical problem to be solved by the invention is to provide a kind of operations of 3D high definition
Recorded broadcast method, can overcome in the prior art picture have the delay of longer time, occupy that system resource is high, image voice is owed
The disadvantages of clear and smooth.
To achieve the above object, the present invention provides a kind of 3D high definition operation recorded broadcast method, include the following steps:
S1, software and hardware: capture card, support CUDA(Compute Unified Device with NvidiaGPU
Architecture video card), existing operation recorded broadcast equipment and corresponding software (software include 3D high definition operation recording and broadcasting system and its
Support programs.Hardware includes the QP1200 capture card of customization and supports the Nvidia GeForce GTX TITAN of CUDA aobvious
Card.)；It can be understood as the capture card of existing operation recorded broadcast equipment replacing with the capture card with GPU, existing operation recorded broadcast set
Standby video card replaces with the video card of support CUDA.
S2, capture card collect audio-video signal, are called by independent audio-video processing engine (sound, Video processing software)
Capture card GPU, which handles audio-video recording function, also just prevents computer crash to be not take up cpu resource.
Processing Algorithm processing is carried out to delay, this case reduces delay using h.264 mvc algorithm, keeps the delay of 3D picture small
In 40ms, simulates 3D picture and be less than 60ms.
S3,3D picture are handled by GPU using CUDA acceleration, and video processing engine automatic adjustment tone/brightness/contrast/
Saturation degree.
S4, recording video post-processing using CUDA acceleration handle, make GPU aiding CPU handle, reduce the time.
The beneficial effects of the present invention are:
1. making the delay of 3D picture be less than 40ms this invention takes individual video processing engine, simulation 3D picture is less than 60ms.
2. the present invention is handled 3D picture by Nvidia GPU using the acceleration of CUDA technology, picture clear and smooth.
3. the code stream that the present invention uses reaches as high as 50Mbps, general video code flow is set as 12Mbps.Support 4K superelevation
Clear picture, highest resolution support 60fps up to 4096x2160, frame per second.Video processing engine can auto tone/brightness/comparison
Degree/saturation degree adjustment, makes picture clear and smooth.
4. the present invention handles video post-processing using CUDA acceleration, make the processing of Nvidia GPU aiding CPU, when reduction
Between (originally collected buffer dsp, postponed generally to 120ms or more up to 2/3rds or more.Using method of the invention
Delay time only has 40ms afterwards, reduces 2/3rds).
Detailed description of the invention
Fig. 1 is the flow diagram of the processing of present invention acquisition audio-video signal.
Fig. 2 is the process flow schematic diagram that 2D turns 3D processing engine.
Specific embodiment
Present invention will be further explained below with reference to the attached drawings and examples:
Referring to Fig. 1-Fig. 2, the present invention is made of software section and hardware two parts.Software include 3D high definition operation recording and broadcasting system and
Its support programs.Hardware includes the QP1200 capture card of customization and supports the Nvidia GeForce GTX TITAN of CUDA aobvious
Card.
Running environment:
CPU:Intel Core I7-7700
Memory: 16G or more
Operating system: 7 or more windows
Fig. 1 is the flow chart of the processing of present invention acquisition audio-video signal, and detailed process is as follows:
Step 201, the entire operation recording and broadcasting system of starting；
Step 202, acquisition equipment initialization, acquisition equipment includes capture card, GPU etc.；
Step 203, capture card acquire sound, vision signal (this case abbreviation audio-video)；
Step 204, audio-video are by audio-video processing engine (software and hardware of processing sound, video) to the image of acquisition according to setting
Parameter handled, so as to automatically adjust tone/brightness/contrast/saturation degree, to resolution ratio, frame per second, de interlacing side
Formula, color space conversion are handled；
Audio-video handles the time that engine reaches capture card according to audio-video, audio-video is marked with the time, capture card meeting
It beats timestamp twice: being the time that audio-video enters capture card for the first time, second of audio-video is the time completely into caching.
The program (existing operation recorded broadcast software) of user's application will be handled audio-video according to timestamp, it is ensured that sound
Audio video synchronization；
Meanwhile with low latency algorithm (the present embodiment is h.264 mvc algorithm), audio-video is made to use ultralow transmission delay,
This guarantees the delays of 3D picture to be less than 40ms, and simulation 3D picture is less than 60ms.
Step 205 judges whether to need recorded video, this mainly passes through the instruction that operation recorded broadcast software receives and determines；
Step 206, if necessary recorded video, then recording and broadcasting system of performing the operation calls capture card GPU to handle recording function, and is not take up
Cpu resource prevents stuck；
If you do not need to recorded video, then turn to step 207, step 208, step 209 step therein or combinations thereof；
If step 207 only needs to show ordinary video, common single channel 2D video signal transmission into step 212, lead to
It crosses video card and is directly output to display preview screen, into preview step 216.
Step 208, if necessary to synthesis 3D video, then two-way audio video transmission to step 213, by step 213 by
CUDA calls Nvidia GPU to accelerate synthesis 3D video, display preview screen is then communicated to, into preview step 216.
Step 209 simulates 3D program if necessary to call, then single channel 2D video is divided into two-way, respectively left video
Image 302, right wing video image 303, left video image 302 are original untreated common video signal, right wing video figure
As 303 be advanced treating vision signal to be carried out；Then it is transferred in step 214 and 3D processing engine (2D turns 3D software) is turned by 2D
It carries out 2D and turns 3D processing；2D turns 3D processing engine and is assisting CPU to synthesize 3D in video card by Nvidia GPU using CUDA technology,
It is then communicated to display preview screen, into preview step 216.
Step 210, in step 206 by judging whether the video after road needs the later period after capture card GPU recorded video
Processing then gos to step 211 if necessary to post-processing, and by step 211, using CUDA technology, the acceleration in GPU is handled, this
Sample can save the most of the time, and go to step 215 after handling, and keep the video file of processing；After if you do not need to
Phase processing, then go to step 219, end processing.
Step 217 judges whether to need OSD(On-Screen Display), if it is desired, then go to step 218,
Step 218 handles OSD using the GPU of capture card；If it is not required, then going to step 219, end processing.
Fig. 2 is the process flow diagram flow chart that 2D turns 3D processing engine, and detailed process is as follows:
Step 301 is divided into the single channel 2D video in step 209 two-way, respectively left video image 302, right wing video figure
As 303, left video image 302 is original untreated common video signal, and right wing video image 303 is depth to be carried out
Manage vision signal；
Left video image 302 is fed directly to be handled in step 306；Right wing video image 303 is view to be treated
Frequency signal.
Step 303 is transmitted to right wing video image 303 step 304, and step 304 extracts 303 depth of right wing video image
Information forms depth map by step 305 according to algorithm process (h.264 mvc algorithm).
Step 305, treated video signal transmission to step 306 so that treated vision signal and left view
Frequency signal 302 carries out image synchronization matching；
Step 307 is assisting CPU to synthesize 3D in video card the video image matched using CUDA technology by Nvidia GPU
Then video image exports 3D video image.
Place is not described in detail by the present invention, is the well-known technique of those skilled in the art.
The preferred embodiment of the present invention has been described in detail above.It should be appreciated that those skilled in the art without
It needs creative work according to the present invention can conceive and makes many modifications and variations.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical solution, all should be within the scope of protection determined by the claims.
Claims (4)
- A kind of recorded broadcast method 1. 3D high definition is performed the operation, which comprises the steps of:S1, software and hardware: the video card of capture card, support CUDA with GPU, existing operation recorded broadcast equipment and corresponding software；S2, capture card collect audio-video signal, are recorded by independent audio-video processing engine calling capture card GPU processing audio-video Function；Processing Algorithm processing is carried out to delay, reduces delay, the delay of 3D picture is made to be less than 40ms, simulation 3D picture is less than 60ms；S3,3D picture are accelerated to handle by GPU or assist CPU processing by GPU by GPU using CUDA；S4, recording video post-processing using CUDA by GPU assist CPU acceleration handle.
- The recorded broadcast method 2. 3D high definition as described in claim 1 is performed the operation, which comprises the steps of:Step 201, the entire operation recording and broadcasting system of starting；Step 202, acquisition equipment initialization；Step 203, capture card acquire sound, vision signal；Step 204, sound, vision signal are handled the image of acquisition according to the parameter of setting by audio-video processing engine；Audio-video handles the time that engine reaches capture card according to audio-video, audio-video is marked with the time, capture card meeting It beats timestamp twice: being the time that audio-video enters capture card for the first time, second of audio-video is the time completely into caching, Application program will be handled audio-video according to timestamp, it is ensured that audio-visual synchronization；Step 205 judges whether to need recorded video；Step 206, if necessary recorded video, then recording and broadcasting system of performing the operation call the GPU of capture card to handle recording function；If no Recorded video is needed, then turns to step 207, step 208, step 209 step therein or combinations thereof；If step 207 only needs to show ordinary video, the common single channel 2D video signal transmission that capture card acquires is arrived In step 212, step 212 is directly output to display preview screen by video card, into preview step 216；Step 208, if necessary synthesis 3D video, then the two-way audio video transmission of capture card acquisition to step 213, in step It calls GPU acceleration to synthesize 3D video by CUDA in 213, display preview screen is then communicated to, into preview step 216；Step 209, if necessary to call simulation 3D program, then single channel 2D video is divided into two-way, respectively left video image, Right wing video image, left video image are original untreated common video signals, and right wing video image will carry out depth Manage vision signal；Then it is transferred to step 214, step 214 turns 3D processing engine progress 2D by 2D and turns 3D processing；2D turns 3D processing engine benefit It assists CPU to synthesize 3D in video card with CUDA driving GPU, display preview screen is then communicated to, into preview step 216；Step 210, at whether step 206 needs the later period by the video after after capture card GPU recorded video, judging road Reason then gos to step 211 if necessary to post-processing, and by step 211, using CUDA, the acceleration in GPU is handled, and after handling Go to step 215, step 215 keeps the video file of processing；If you do not need to post-processing, then 219 are gone to step, It ends processing.
- The recorded broadcast method 3. 3D high definition as claimed in claim 2 is performed the operation, which is characterized in that further include step 217, step 217 is sentenced It is disconnected whether OSD to be needed to handle, if it is desired, then to go to step 218, step 218 handles OSD using the GPU of capture card；If It does not need, then gos to step 219, end processing.
- The recorded broadcast method 4. 3D high definition as claimed in claim 2 is performed the operation, which is characterized in that in step 214,2D turns the processed of 3D Journey is as follows:Step 301 is fed directly to the left video image in step 209 to be handled in step 306；Step 303, right wing transmission of video images to step 304, step 304 extracts right wing video image depth information, passes through Step 305 forms depth map according to algorithm process；Step 305, treated video signal transmission to step 306 so that treated vision signal and left video letter Number carry out image synchronization matching；Step 307 assists CPU to synthesize 3D video image in video card the video image matched using CUDA driving GPU, so 3D video image is exported afterwards.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910257890.7A CN109889749B (en) | 2019-04-01 | 2019-04-01 | 3D high-definition operation recording and broadcasting method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910257890.7A CN109889749B (en) | 2019-04-01 | 2019-04-01 | 3D high-definition operation recording and broadcasting method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109889749A true CN109889749A (en) | 2019-06-14 |
| CN109889749B CN109889749B (en) | 2021-09-21 |
Family
ID=66935597
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910257890.7A Active CN109889749B (en) | 2019-04-01 | 2019-04-01 | 3D high-definition operation recording and broadcasting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109889749B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110113551A (en) * | 2019-06-19 | 2019-08-09 | 江苏医药职业学院 | A kind of intellectual medical image audio-video recording system and method |
| CN112312071A (en) * | 2019-07-31 | 2021-02-02 | 上海鋆锐医疗器械有限公司 | 4K ultrahigh-definition medical camera system based on 3D camera technology |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102647602A (en) * | 2011-02-17 | 2012-08-22 | 北京大学深圳研究生院 | System for converting 2D (two-dimensional) video into 3D (three-dimensional) video on basis of GPU (Graphics Processing Unit) |
| CN106534672A (en) * | 2016-11-02 | 2017-03-22 | 深圳亿维锐创科技股份有限公司 | Medical operation direct broadcast system |
| US20180101929A1 (en) * | 2016-10-10 | 2018-04-12 | Samsung Electronics Co., Ltd. | Graphics processing devices and graphics processing methods |
| CN108769643A (en) * | 2018-06-15 | 2018-11-06 | 深圳超多维科技有限公司 | Video capture method, apparatus and system |
| CN108937853A (en) * | 2018-05-29 | 2018-12-07 | 华南师范大学 | A kind of three-dimensional opto-acoustic microscopic imaging method and imaging system of multiport parallel processing |
-
2019
- 2019-04-01 CN CN201910257890.7A patent/CN109889749B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102647602A (en) * | 2011-02-17 | 2012-08-22 | 北京大学深圳研究生院 | System for converting 2D (two-dimensional) video into 3D (three-dimensional) video on basis of GPU (Graphics Processing Unit) |
| US20180101929A1 (en) * | 2016-10-10 | 2018-04-12 | Samsung Electronics Co., Ltd. | Graphics processing devices and graphics processing methods |
| CN106534672A (en) * | 2016-11-02 | 2017-03-22 | 深圳亿维锐创科技股份有限公司 | Medical operation direct broadcast system |
| CN108937853A (en) * | 2018-05-29 | 2018-12-07 | 华南师范大学 | A kind of three-dimensional opto-acoustic microscopic imaging method and imaging system of multiport parallel processing |
| CN108769643A (en) * | 2018-06-15 | 2018-11-06 | 深圳超多维科技有限公司 | Video capture method, apparatus and system |
Non-Patent Citations (1)
| Title |
|---|
| 丁莹等: "《复杂环境运动目标检测技术及应用》", 31 January 2014 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110113551A (en) * | 2019-06-19 | 2019-08-09 | 江苏医药职业学院 | A kind of intellectual medical image audio-video recording system and method |
| CN112312071A (en) * | 2019-07-31 | 2021-02-02 | 上海鋆锐医疗器械有限公司 | 4K ultrahigh-definition medical camera system based on 3D camera technology |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109889749B (en) | 2021-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103347163B (en) | Ultra high definition video image processing and transmitting system and method thereof | |
| CN1306796C (en) | Decoder device and receiver using the same | |
| CN106846940A (en) | A kind of implementation method of online live streaming classroom education | |
| CN203118262U (en) | Operation teaching system adopted in hospitals | |
| US20100007773A1 (en) | Video Processing and Telepresence System and Method | |
| CN109889749A (en) | A kind of 3D high definition operation recorded broadcast method | |
| CN109379619A (en) | Sound draws synchronous method and device | |
| CN109729405A (en) | Method for processing video frequency, device, electronic equipment and storage medium | |
| CN106534672A (en) | Medical operation direct broadcast system | |
| CN202736321U (en) | Digital hybrid operation room remote medical teaching system | |
| CN110913273A (en) | Video live broadcasting method and device | |
| US8593575B2 (en) | Video display apparatus for shortened-delay processing of a video signal and video processing method | |
| WO2019080846A1 (en) | Video data processing method and video data processing device | |
| CN107295355A (en) | A kind of audio frequency and video total system and method | |
| CN111479154A (en) | Equipment and method for realizing sound and picture synchronization and computer readable storage medium | |
| CN207926703U (en) | Course recording system | |
| CN106941597A (en) | A kind of video call method and device | |
| CN114339302A (en) | Broadcasting guide method, device, equipment and computer storage medium | |
| CN106791903A (en) | Live processing system, unit and method | |
| CN204168404U (en) | A kind of processing unit of ultra high-definition signal and full HD TV | |
| CN108461157A (en) | A kind of multi-party Tele-consultation System based on APP | |
| CN206313889U (en) | Video switching system | |
| CN205584318U (en) | USB high definition meeting camera | |
| CN113840166A (en) | Method and system for synchronizing audio and video mixing of multi-path streaming media | |
| KR100970992B1 (en) | System and method for multiplexing stereoscopic high-definition video through gpu acceleration and transporting the video with light-weight compression and storage media having program source thereof |
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 | ||
| CB02 | Change of applicant information |
Address after: No.298, Wenhua Road, Sijing Town, Songjiang District, Shanghai, 201601 Applicant after: SHANGHAI QPEIT ELECTRONIC TECHNOLOGY Co.,Ltd. Address before: 201718 room 187, Zone E, floor 1, building 1, No. 53, Lianhu Road, Jinze Town, Qingpu District, Shanghai Applicant before: SHANGHAI QPEIT ELECTRONIC TECHNOLOGY Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |