CN110198456A - Video pushing method, device and computer readable storage medium based on live streaming - Google Patents
Video pushing method, device and computer readable storage medium based on live streaming Download PDFInfo
- Publication number
- CN110198456A CN110198456A CN201910341639.9A CN201910341639A CN110198456A CN 110198456 A CN110198456 A CN 110198456A CN 201910341639 A CN201910341639 A CN 201910341639A CN 110198456 A CN110198456 A CN 110198456A
- Authority
- CN
- China
- Prior art keywords
- video
- video frame
- short
- live
- sighted frequency
- 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
- 238000000034 method Methods 0.000 title claims abstract description 91
- 238000003860 storage Methods 0.000 title abstract description 15
- 208000001491 myopia Diseases 0.000 claims abstract description 178
- 238000001514 detection method Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 30
- 238000010586 diagram Methods 0.000 description 29
- 238000012545 processing Methods 0.000 description 25
- 238000012015 optical character recognition Methods 0.000 description 12
- 230000004083 survival effect Effects 0.000 description 12
- 230000000694 effects Effects 0.000 description 10
- 230000006870 function Effects 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 8
- 238000013527 convolutional neural network Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 5
- 238000009331 sowing Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000002716 delivery method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 241000272171 Scolopacidae Species 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000005498 polishing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/21—Server components or server architectures
- H04N21/218—Source of audio or video content, e.g. local disk arrays
- H04N21/2187—Live feed
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/234—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/234—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs
- H04N21/23418—Processing of video elementary streams, e.g. splicing of video streams or manipulating encoded video stream scene graphs involving operations for analysing video streams, e.g. detecting features or characteristics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/27—Server based end-user applications
- H04N21/274—Storing end-user multimedia data in response to end-user request, e.g. network recorder
- H04N21/2743—Video hosting of uploaded data from client
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/85—Assembly of content; Generation of multimedia applications
- H04N21/854—Content authoring
- H04N21/8549—Creating video summaries, e.g. movie trailer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/85—Assembly of content; Generation of multimedia applications
- H04N21/854—Content authoring
- H04N21/8547—Content authoring involving timestamps for synchronizing content
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Databases & Information Systems (AREA)
- Computer Security & Cryptography (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention discloses a kind of video pushing method based on live streaming, device and computer readable storage medium, this method identifies the video frame in the second live video stream during to live streaming the second live video stream of client push；If the first video frame identified meets short-sighted frequency pushing condition, since the timestamp of first video frame, the short-sighted frequency of target is pushed, wherein the short-sighted frequency of target is to generate based on the first live video stream pushed.By during to the second live video stream of push, play the short-sighted frequency of the excellent collection of choice specimens, the prior art is not only avoided by reviewing the case where mode plays target short-sighted frequency generation, and solves the problems, such as that specific broadcasting staff is needed to play high labor cost caused by excellent collection of choice specimens video in the prior art.
Description
Technical field
The present invention relates to technical field of video processing more particularly to a kind of video pushing method based on live streaming, device and
Computer readable storage medium.
Background technique
The game live streaming of current network platform is more and more, and the excellent collection of choice specimens of game on existing live streaming platform is live streaming knot
Shu Hou is carried out editing arrangement to the live video played by main broadcaster or platform operation later period and is uploaded on live streaming platform,
Then it is presented to the user in a manner of reviewing, with reference to shown in Fig. 1 a, user needs editing in program request Fig. 1 a to arrange video to review
The excellent collection of choice specimens of game.However in practical application, user seldom watches on-demand content in live streaming platform, and user is for game essence
The color collection of choice specimens reviews demand also than stronger, and traditional way is similar to TV competition broadcast, and specific broadcasting staff is needed to come
Realtime interpretation simultaneously plays, and causes cost of labor higher.
Therefore, how to realize and play the excellent collection of choice specimens video during live streaming to solve to need in the prior art specifically to turn
It is one of the technical issues of being worthy of consideration that the personnel of broadcasting, which play the problem of high labor cost caused by excellent collection of choice specimens video,.
Summary of the invention
The embodiment of the present invention provides a kind of video pushing method based on live streaming, device and computer readable storage medium,
Excellent collection of choice specimens video is played during live streaming to realize, to solve that specific broadcasting staff is needed to play essence in the prior art
The problem of high labor cost caused by color collection of choice specimens video.
On the one hand, the embodiment of the present invention provides a kind of video pushing method based on live streaming, comprising:
During to live streaming the second live video stream of client push, the video in the second live video stream is identified
Frame；
If the first video frame identified meets short-sighted frequency pushing condition, opened from the timestamp of first video frame
Begin, push the short-sighted frequency of target, wherein the short-sighted frequency of target is to be generated based on the first live video stream pushed.
On the other hand, the embodiment of the present invention provides a kind of video push device based on live streaming, comprising:
Identification module, for during to live streaming the second live video stream of client push, the second live streaming of identification to be regarded
Video frame in frequency stream；
Video push module, if the first video frame for identifying meets short-sighted frequency pushing condition, from described first
The timestamp of video frame starts, and pushes the short-sighted frequency of target, wherein the short-sighted frequency of target is based on the first live streaming view pushed
Frequency stream generates.
In another aspect, the embodiment of the present invention provides a kind of computer equipment, including memory, processor and it is stored in storage
On device and the computer program that can run on a processor, the processor realize base provided by the invention when executing described program
In the video pushing method of live streaming.
In another aspect, the embodiment of the present invention provides a kind of computer-readable medium, it is stored with processor-executable instruction, institute
Processor-executable instruction is stated for executing the video pushing method provided by the invention based on live streaming.
Video pushing method based on live streaming, device and computer readable storage medium provided in an embodiment of the present invention,
During live streaming the second live video stream of client push, the video frame in the second live video stream is identified；If identifying
The first video frame meet short-sighted frequency pushing condition, then since the timestamp of first video frame, push the short-sighted frequency of target,
Wherein, the short-sighted frequency of the target is to be generated based on the first live video stream pushed.By the way that view is being broadcast live to push second
During frequency flows, the short-sighted frequency of the excellent collection of choice specimens is played, the prior art is not only avoided by reviewing mode and plays the short-sighted frequency of target
The case where occur, and solve and specific broadcasting staff needed to play people caused by excellent collection of choice specimens video in the prior art
Work problem at high cost.
In addition, after generating the short-sighted frequency of target, the present invention is proposed to live streaming the second live video stream of client push
In the process, the short-sighted frequency of target is pushed, to also avoid the prior art by reviewing the case where mode plays target short-sighted frequency
Occur.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
It obtains it is clear that understand through the implementation of the invention.The objectives and other advantages of the invention can be by written explanation
Specifically noted structure is achieved and obtained in book, claims and attached drawing.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes a part of the invention, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 a is the schematic diagram for playing excellent collection of choice specimens video in the prior art；
Fig. 1 b is the effect diagram for launching advertisement at the position being manually specified in the prior art；
Fig. 2 is the application scenarios schematic diagram of the video pushing method provided in an embodiment of the present invention based on live streaming；
Fig. 3 is one of the flow diagram of video pushing method provided in an embodiment of the present invention based on live streaming；
Fig. 4 is the flow diagram that determining candidate video frame provided in an embodiment of the present invention meets short video intercepting condition；
Fig. 5 is candidate video frame schematic diagram provided in an embodiment of the present invention；
Fig. 6 is the candidate video frame schematic diagram provided in an embodiment of the present invention obtained based on heroic alliance's live video stream；
Fig. 7 a is the two of the flow diagram of the video pushing method provided in an embodiment of the present invention based on live streaming；
Fig. 7 b is one of the schematic diagram of the push short-sighted frequency of target provided in an embodiment of the present invention；
Fig. 8 is the two of the schematic diagram of the push short-sighted frequency of target provided in an embodiment of the present invention；
Fig. 9 is the short-sighted frequency of target played main broadcaster's lower sowing time provided in an embodiment of the present invention --- excellent collection of choice specimens video signal
Figure；
Figure 10 is the three of the schematic diagram of the push short-sighted frequency of target provided in an embodiment of the present invention；
Figure 11 a is main broadcaster's frame region schematic diagram provided in an embodiment of the present invention；
Figure 11 b be it is provided in an embodiment of the present invention can carrying advertisement region schematic diagram；
Figure 12 is event-monitoring schematic diagram provided in an embodiment of the present invention；
The structural schematic diagram of Figure 13 video push device provided in an embodiment of the present invention based on live streaming；
Figure 14 is the execution building-block of logic of the video pushing method provided in an embodiment of the present invention based on live streaming；
Figure 15 is the structural schematic diagram of video processing module 134 provided in an embodiment of the present invention；
Figure 16 is that the structure of the calculating equipment provided in an embodiment of the present invention for implementing the video pushing method based on live streaming is shown
It is intended to.
Specific embodiment
Video pushing method based on live streaming, device and computer readable storage medium provided in an embodiment of the present invention are used
Excellent collection of choice specimens video is played during live streaming to realize, it is excellent to solve to need specific broadcasting staff to play in the prior art
The problem of high labor cost caused by collection of choice specimens video.
Below in conjunction with Figure of description, preferred embodiment of the present invention will be described, it should be understood that described herein
Preferred embodiment only for the purpose of illustrating and explaining the present invention and is not intended to limit the present invention, and in the absence of conflict, this hair
The feature in embodiment and embodiment in bright can be combined with each other.
To facilitate the understanding of the present invention, in technical term of the present invention:
1, optical character identification, (Optical Character Recognition, OCR) are defeated by optics such as scannings
Enter mode and convert image information for the text of various bills, newpapers and periodicals, books, manuscript and other printed matters, text is recycled to know
Other technology converts image information to the computer input technology that can be used.
2, content distributing network, (Content Delivery Network, CDN), is the content constructed on network
Distribution network passes through the function such as the load balancing, content distribution, scheduling of central platform by the Edge Server for being deployed in various regions
It can module, content needed for obtaining user nearby, reduction network congestion, raising user's access response speed and hit rate.
3, Software Development Kit, (Software Development Kit, SDK) is usually specific software package,
The set of developing instrument when software frame, hardware platform, operating system etc. establish application software.
4, plug-flow: the process the good content transmission of acquisition phase package to server is referred to.Being exactly in fact will be live
Vision signal pass to the process of network." plug-flow " is relatively high to network requirement, if unstable networks, live streaming effect will be very
Phenomena such as difference, when watching live streaming Caton will occur for spectators, influences user and watches live-experience.
5, transcoding technology is a kind of a kind of technology that audio-video signal is converted into another format from format.As sound
Video real-time messages transport protocol (RTMP) stream format is transcoded into FLV file format.
Inventors have found that the prior art needs artificial editing live content to generate the excellent collection of choice specimens, but this method needs
Expend higher human cost and time；In addition, launching the position one of advertisement when the prior art launches advertisement during live streaming
As by being manually set in advance, then the advertising information placement position with reference to shown in Fig. 1 b carries in live stream and shows advertisement etc.
Information, the information such as advertisement will be showed in setting position during live streaming, can thus generate and launch advertisement etc. regardless of occasion
The problem of information, the case where not blocking live content there is only information such as the advertisements of dispensing, and will affect user and watch live streaming
Experience.
It such as to solve the above problems, then need to change existing live streaming platform, to generate excellent collection during live streaming
Bright and beautiful video is obtained without manually passing through later period editing processing.In consideration of it, the present invention devises the video push side based on live streaming
Case can specifically identify in the second live video stream during to live streaming the second live video stream of client push
Video frame；If the first video frame identified meets short-sighted frequency pushing condition, opened from the timestamp of first video frame
Begin, pushes the short-sighted frequency of target, and the short-sighted frequency of the target is to be generated based on the first live video stream pushed, to realize
Target short-sighted frequency is pushed during live streaming, also just solves need specific broadcasting staff's broadcasting excellent in the prior art in this way
The problem of high labor cost caused by collection of choice specimens video；In addition, the short-sighted frequency of target in the present invention is generated based on following methods
: the candidate video frame in the first live video stream pushed is obtained, when above-mentioned candidate video frame belongs to the excellent collection of choice specimens, from
Interception includes the video-frequency band of above-mentioned candidate video frame in above-mentioned first live stream, is then based on above-mentioned video-frequency band and generates short-sighted frequency,
Thus i.e. excellent collection of choice specimens video is also just not necessarily to artificial later period editing to realize the effect for automatically generating excellent collection of choice specimens video
Processing, and then save cost of labor and time.
After having introduced the design philosophy of the embodiment of the present invention, the technical solution of the embodiment of the present invention can be fitted below
Application scenarios do some simple introductions, it should be noted that application scenarios introduced below are merely to illustrate of the invention real
Apply example and non-limiting.In the specific implementation, technical side provided in an embodiment of the present invention can be neatly applied according to actual needs
Case.
Please refer to a kind of application scenarios shown in Fig. 2, in the scene, including terminal device 101, terminal device 102, terminal
Equipment 103 and server 104, terminal device 101, terminal device 102, terminal device 103 respectively with user 1, user 2, user 3
It is corresponding.Terminal device 101, terminal device 102 can install same live streaming client in terminal device 103.With user 1
It initiates to be illustrated for live streaming, user 2 and user 3 watch the live streaming that user 1 initiates, when user 1 passes through live streaming client hair
When playing live streaming, live video stream is pushed to by server 104 by terminal device 101, then server 104 pushes live stream
To terminal device 102 and terminal device 103, such user 2 and user 3 can watch the live streaming of the initiation of user 1, and work as this always
It broadcasts after stream is pushed to user 2 and user 3, in order to realize the push of the short-sighted frequency of target, server 104 can pass through and implement the present invention
There is provided based on the video pushing method of live streaming in the live streaming client push into terminal device 102 and terminal device 103 the
The short-sighted frequency of target is pushed during two live video streams, thereby realizes the effect that target short-sighted frequency is pushed during live streaming
Fruit so also just without the special open function of reviewing the excellent collection of choice specimens, and then also solves and needs specifically to turn in the prior art
The personnel of broadcasting play the problem of high labor cost caused by excellent collection of choice specimens video.
Terminal device 101, terminal device 102 and terminal device 103 above-mentioned can be mobile phone, tablet computer, palm electricity
Brain (Personal Digital Assistant, PDA), laptop, intelligent wearable equipment (such as smartwatch and intelligence
Can bracelet), personal computer, etc., terminal device no matter what type can be equipped with same in the terminal device
The corresponding client of one social platform, and corresponding social account can be logged in by each terminal device.And it is aforementioned
Server 104 can be personal computer, large and medium-sized computer, computer cluster, etc..Be broadcast live client can for web,
The live streaming client that android system or iOS system provide.
It should be noted that the live video stream in the present invention can be game live video stream, for ease of description, after
It is continuous to be illustrated by taking game live video stream as an example.
Certainly, it method provided in an embodiment of the present invention and is not exclusively in application scenarios shown in Fig. 2, can be also used for it
Its possible application scenarios, the embodiment of the present invention are simultaneously not limited.For each equipment institute energy of application scenarios shown in Fig. 2
The function of realization will be described together in subsequent embodiment of the method, not repeat excessively first herein.
Below with reference to application scenarios shown in Fig. 2, with reference to following figures description, illustrative embodiments are mentioned according to the present invention
The video pushing method based on live streaming supplied.It should be noted that above-mentioned application scenarios be merely for convenience of understanding it is of the invention
Spirit and principle and show, embodiments of the present invention are not limited in this respect.On the contrary, embodiments of the present invention can
To be applied to applicable any scene.
It please refers to shown in Fig. 3, is the flow diagram of the video pushing method provided in an embodiment of the present invention based on live streaming,
It is illustrated by taking the server that the above method is applied in Fig. 2 as an example, the above method of server implementation may include following
Step:
S31, candidate video frame is obtained from the first live video stream pushed.
In this step, in order to generate the excellent collection of choice specimens so that live streaming user watches again, then need from the first live streaming view
Candidate video frame is intercepted in frequency stream.It is alternatively possible to timing from the first live video stream obtain candidate video frame, such as it is per second from
A candidate video frame is intercepted in first live video stream.
Optionally, before obtaining candidate video frame in the first live video stream, further includes:
Confirm and carries video identification mark in first live video stream.
Specifically, live streaming platform has thousands of live streaming room being broadcast live, and the live content in some rooms
It is relatively uninteresting, it is not enough to transfer the interest of user, therefore the live streaming room of live streaming platform can be configured, by being one
A little live streaming room configuration video identification marks, while some excellent live streamings can be assigned to configured with view by the server that platform is broadcast live
The live streaming room of frequency identification marking.Based on this, when based on live streaming room push live video stream, in the live video stream
Carrying is that the video identification that the live streaming room configures identifies, when getting live video stream, if the live video stream carries view
Frequency identification marking can then obtain candidate video frame from the live video stream that this has been pushed.
S32, when candidate video frame meets short video intercepting condition, interception includes candidate view from the first live video stream
The video-frequency band of frequency frame.
In this step, the short video intercepting condition of different live video stream settings is different, it is intended to intercept it is some it is excellent,
The relatively high short-sighted frequency of temperature.For example, can generally be intercepted very smart when the first live video stream is game live video stream
Color, the worth video-frequency band watched again, therefore short video intercepting condition is set based on this purpose, in this way, when candidate video frame is full
When the short video intercepting condition of foot, show that the content of the candidate video frame is more excellent, therefore intercepts the video comprising candidate video frame
Section are as follows: on the basis of the candidate video frame, interception time stamp is later than several video frames of the timestamp of the candidate video frame, with
And timestamp is earlier than several video frames of the timestamp of candidate video frame.
S33, the video-frequency band based on interception generate the short-sighted frequency of target.
In this step, after intercepting video-frequency band, which is exactly excellent collection of choice specimens video；In order to allow viewing live streaming user to obtain
Know that currently watched live streaming is collection of choice specimens video or live video, the video-frequency band of interception can be handled, such as add and regarding
Plus " the excellent collection of choice specimens " to identify currently playing video as the excellent collection of choice specimens in the first video frame of frequency range；Further, it is also possible to again
The configuration one short-sighted frequency of beginning for indicating will to play excellent collection of choice specimens video and one, which are used to characterize excellent collection of choice specimens video, to be broadcast
The short-sighted frequency of end for putting end by the video-frequency band for starting short-sighted frequency, interception and terminates the short-sighted frequency composition short-sighted frequency of target.It is based on
This, can automatically generate excellent collection of choice specimens video without artificial later period editing live video stream, not only save cost of labor and
And save the generation time of excellent collection of choice specimens video.
It should be noted that the present invention is to the short-sighted frequency for starting short-sighted frequency, interception and terminates the video for including in short-sighted frequency
The quantity of frame without limit, specifically can according to the actual situation depending on.
Based on any of the above-described embodiment, it can determine that candidate video frame meets short video intercepting according to method shown in Fig. 4
Condition:
S41, the content type for identifying candidate video frame.
Specifically, it can use trained content type identification model to identify the content type of candidate video frame.Tool
When body is implemented, the picture off-line training content type identification model of various content types can use, obtain candidate view when online
When frequency frame, trained content type identification model is called, it, should using candidate video frame as the input of content type identification model
Content type identification model can export the matching probability between the candidate video frame and each content type, according to matching probability
It can determine the content type of the candidate video frame.
Optionally, content type identification model can be the disaggregated model based on convolutional neural networks, by live scene point
Class has a large amount of live streaming room to need to carry out type identification, Er Qieji at multiple content types and content status, due to per second
Bright and beautiful video also has higher requirement to the speed of identification, based on these consider the present invention can use to VGG16 disaggregated model into
Row training obtains content type identification model.
When candidate video frame is game video frame, then content type is that type of play, content type identification model are corresponding
For type of play identification model, when identifying that the type of play of candidate video frame is sought survival with danger spot using type of play identification model
Matching probability maximum when, it is determined that the type of play of candidate video frame is sought survival for danger spot.
If S42, the content type are setting content type, event detection is carried out to the candidate video frame, and work as
Determine that there are the event types for when event, determining existing event in the candidate video frame.
Specifically, some setting content types can be preset and be stored in setting list, identifying candidate view
When the content type of frequency frame, tentatively judge that the candidate video frame may be if the content type identified is in setting list
Featured videos frame can then carry out event detection in order to further determine whether being featured videos frame to the candidate video frame, and
There are the event types for when event, determining existing event in determining the candidate video frame.
Event in the present invention is used to characterize the content in candidate video frame, when live streaming is game live streaming, then snipes field
Event in scape may include killing event etc..
When the content type identified is not setting content type, then without executing follow-up process.
When carrying out event detection to candidate video frame, The present invention gives three kinds of methods, are certainly not limited to these three calculations
Method specifically can choose suitable algorithm according to the difference of application scenarios and carry out event detection, next examine to each event
Method of determining and calculating is introduced:
Mode one: the event detection based on optical character recognition algorithms.
It is alternatively possible to carry out event monitoring according to following processes using optical character recognition algorithms: from candidate video frame
Middle interception key area carries out Text region to the key area, determines the candidate video frame based on Text region result
In whether there is event.
Specifically, due to will lead to OCR service delay and accuracy rate by shadow when excessive video frame carries out OCR identification
It rings, therefore in order to improve OCR recognition speed and accuracy rate, key area can be intercepted from candidate video frame, in this way, only to key
Region carries out OCR identification rather than carries out OCR identification to entire candidate video frame, so as to improve the text of candidate video frame
Recognition speed.It should be noted that the location information of key area can be substantially in different candidate video frames in practical application
It determines, is such as based on empirically determined above-mentioned location information, is then based on these location informations to intercept key area, with reference to Fig. 5 institute
Show, the schematic diagram of 5 key areas in right side is intercepted out from the candidate video frame on the left of Fig. 5.
When it is implemented, may include specific character in image where some events, and the different corresponding texts of event
Word is different, therefore is based on this principle, after intercepting key area, can include text in these key areas, then can use light
The text in character recognition algorithm identification key area is learned, it is right using the rule between event and text according to Text region result
Differentiated in candidate video frame with the presence or absence of event.
Mode two: the event detection based on convolutional neural networks sorting algorithm.
Convolutional neural networks can be based on according to following processes and event monitoring is carried out to candidate video frame: according to candidate video
Frame and trained event detection model determine candidate video frame with the presence or absence of event, wherein the instruction of above-mentioned event detection model
Practice sample are as follows: there are the event image pattern of event and the image patterns in addition to event image pattern.
Specifically, the event detection model in the present invention is trained convolutional neural networks, is distinguished using various events
Corresponding event image pattern and there is no the image pattern of event to convolutional neural networks be iterated training obtain.
Herein on basis, candidate video frame is input in trained event detection model, the event detection model
The event detection outcome of candidate video frame can be exported, if there are events in candidate video frame, existing thing can be exported by exporting result
Part type identification, if event is not present in candidate video frame, output is for characterizing the mark there is no event, to be different from thing
Part type identification.It is alternatively possible to indicate event type identification with number, and the different corresponding events of event type identification
Also not identical.For example, digital " 1 " mark kills event, digital " 1 " mark opens sniperscope event, and digital " 0 " mark is not deposited
In event.
Mode three: it is based on the matched event detection of feature templates.
It is alternatively possible to according to following process embodiments two, comprising: by the character modules of candidate video frame and each event
Plate carries out characteristic matching；According to matching result, determine the candidate video frame with the presence or absence of event.
Specifically, the feature templates of candidate video frame and each event are input in feature templates matching algorithm, the mould
Plate matching algorithm can export the result that whether there is event in candidate video frame.
When it is implemented, being pre-configured with the corresponding feature templates of a variety of events, feature templates matching algorithm pair is then utilized
Candidate video frame and the corresponding feature templates of a variety of events carry out Rapid matching, determine candidate video frame and each thing respectively
The quantity for the pixel that the feature templates of part match, if the feature templates based on a certain event determine the pixel to match
Quantity be greater than setting amount threshold, it is determined that there are the events in candidate video frame.For example, if the feature based on the event that kills
Template determines that the quantity of the pixel to match is greater than setting amount threshold, it is determined that there are events in candidate video frame, then
Existing event is to kill event, and then can determine the corresponding event type of the event of killing.
Specifically, multiple events can correspond to an event type, and event type may include killing type, gunslinging class
Type, group's war type etc., specifically can according to the actual situation depending on.
If S43, the event type are setting event type, it is determined that candidate video frame meets short video intercepting condition.
Specifically, it is pre-configured with a setting event type list, when determining event type that step S43 is identified and set
When determining the setting event type in event type list and matching, it is determined that candidate video frame meets short video intercepting condition.
Next process shown in Fig. 4 in order to better understand is illustrated with following two examples:
Herein in connection with shown in Fig. 5, Fig. 5 is the candidate video frame obtained from the first live video stream, if being known based on step S41
Not Chu the content type (type of play) of candidate video frame seek survival for danger spot, and danger spot is sought survival as setting content type, then from time
It selects and intercepts key area in video frame, and Text region, the text identified are carried out to the key area based on OCR recognition methods
Event is killed to determine to exist in the candidate video frame for " 3 times kill ", and identifies that the event type for the event of killing is to kill
Type, and this kills type as setting event type, it is determined that candidate video frame shown in fig. 5 meets short video intercepting condition,
The video-frequency band comprising the candidate video frame is then intercepted out from the first live video stream, then adds collection of choice specimens video for the video-frequency band
Mark, and configure and start short-sighted frequency and terminate short-sighted frequency, to constitute the short-sighted frequency of target namely excellent game collection of choice specimens video.
As shown in connection with fig. 6, Fig. 6 is the candidate video obtained from the first live video stream (heroic alliance's live video stream)
Frame, when identifying the content type of the candidate video frame for heroic alliance using VGG16 disaggregated model, and heroic alliance belongs to and sets
Determine content type, then intercepts key area from candidate video frame, and carry out based on key area of the OCR recognizer to interception
Text region, such as identify texts such as " enemy are killed ", it is determined that there are event and event type is in the candidate video frame
Type is killed, and kills type as setting event type, then can intercept out from heroic alliance's live video stream includes Fig. 6 institute
The video-frequency band of the candidate video frame shown is then based on the video-frequency band and generates the short-sighted frequency of target, namely the excellent collection of choice specimens view of heroic alliance
Frequently.
Optionally, can also there be the candidate video frame of the texts such as " double play ", " super refreshing " in heroic alliance's game, if getting
Such candidate video frame simultaneously identifies above-mentioned text, then can decide whether as a war event type, and for heroic alliance
For game, it includes the texts such as " double play ", " super refreshing " when getting that once extensive group's war, which is the key that choose excellent collection of choice specimens video,
When the candidate video frame of word, then the video-frequency band comprising the candidate video frame intercepted can be assumed that as excellent collection of choice specimens video.In addition,
It is also based on object detection to determine game residue number, enemy's residue number, our remaining number and kill event etc., knot
It closes and states Text region result and further determine whether and then to generate essence according to the method provided by the invention for a war event type
Color collection of choice specimens video.Based on this, by such excellent collection of choice specimens video be pushed to during live streaming user viewing, user not but not
Dislike can improve the excitement levels that user watches live streaming instead.
After the product process for describing the short-sighted frequency of target, the playing flow of the short-sighted frequency of target is next introduced, clothes are worked as
Device be engaged in when the second live video stream of client push is broadcast live, it is straight that the short-sighted frequency of target can be mixed into second during push
It broadcasts and is pushed to live streaming client in video flowing, be also achieved that during live streaming in this way and play excellent collection of choice specimens video, thus
The page just is reviewed without being specially arranged one on live streaming platform, reduces the operation cost of live streaming platform, while without spy
Fixed broadcasting staff plays excellent collection of choice specimens video, and then solves and specific broadcasting staff is needed to play excellent collection in the prior art
The problem of high labor cost caused by bright and beautiful video.The short video push process of target shown in Fig. 7 a can be specifically referred to, including
Following steps:
S71, view during to live streaming the second live video stream of client push, in the second live video stream of identification
Frequency frame.
If S72, the first video frame identified meet short-sighted frequency pushing condition, from the timestamp of first video frame
Start, pushes the short-sighted frequency of target.
Optionally, the first live video stream and the second live video stream can for same live streaming room live video stream and
Meet incidence relation in time, i.e. the push time of the first live video stream and the push time of the second live video stream meet
Incidence relation, for example, when live streaming is danger spot survival game live streaming, if the second live video stream is the 5th innings of danger spot survival game
Live video stream, then the first live video stream can be the 4th innings of danger spot survival game live video stream, or it is preceding
The live video stream of any innings of danger spot survival game in four innings of danger spot survival game.
Optionally, the first live video stream and the second live video stream may not be the live video in same live streaming room
Stream or the first live video stream and the second live video stream can also be unsatisfactory for association in time relationship in time.For example, working as
When live streaming is broadcast live for shopping, the first live video stream is shopping live video stream, then passes through reality based on first live video stream
After applying the process generation more short-sighted frequency of shopping target of bright spot shown in Fig. 3, in other live streaming rooms and/or can not have
When the time point of association in time relationship pushes the second live video stream, the short-sighted frequency of shopping target is pushed, can be allowed so more
User watches the short-sighted frequency of shopping target, and then live streaming user is allowed to recognize the article in short-sighted frequency of doing shopping, to improve
The diffusance of the article.
Optionally, when pushing the short-sighted frequency of target the timestamp since the first video frame, the first video can not be pushed
Timestamp is later than the video frame of the timestamp of the first video frame in frame and the second live video stream, and when the short video push of target is complete
Cheng Hou pushes what the timestamp completed in the second live video stream with push matched then since the timestamp that push is completed
Video frame.
Optionally, when the first video frame meets short-sighted frequency pushing condition, start to push away in the timestamp of first video frame
When sending the short-sighted frequency of target, the second live video stream of push can also continue to, to reach in the first video frame and subsequent video
The effect of the short-sighted frequency of target is shown on frame with wicket.
In step S71 and step S72, during the push of the second live video stream, target cannot be pushed at any time
Short-sighted frequency, it is also desirable to the content in the second live video stream be identified, therefore can be identified in real time in the second live video stream
Video frame, when identifying that the first video frame meets short-sighted frequency pushing condition in the second live video stream, just can from this first
The timestamp of video frame starts to push the short-sighted frequency of target, then during push target short-sighted frequency, second live video
Several video frames that timestamp is later than the timestamp of first video frame in stream will not be pushed, the view that will not be specifically pushed
The quantity of frequency frame can be depending on the playing duration of the short-sighted frequency of target.If the short video playing of target 2 minutes, the second live streaming view
In frequency stream since the timestamp of first video frame, the video frame in 2 minutes will not be pushed.
Specifically, when stating process on the implementation, the second live video stream is decoded in advance, then will decoding must the
Two live videos are streamed to Filter Chain encoder and carry out special effect processing, if encoder increases a time window, with
Guarantee the fluency of pushing video streaming.When identifying that the first video frame in the second live video stream meets video playing condition
When, really start the first video frame in the switching push short-sighted frequency of target after encoder output first video frame, then successively
The subsequent video frame in the short-sighted frequency of target is pushed, is not the second live video of direct push after the short video push of target is complete
Video frame in stream, and be to continue with and read frame from encoder, when encountering the video frame in the second live video stream, refresh
Then the timestamp of encoder and the timestamp polishing frame according to the video frame encountered starts to push in the second live video stream
Video frame is also achieved that the push short-sighted frequency of target during pushing the second live video stream in this way, also achieves simultaneously
Continue to push the video frame in the second live video stream after the completion of target short video push.
The process for pushing the short-sighted frequency of target in the second live video stream in order to better understand, is said in conjunction with Fig. 7 b
Bright, it is RTMP lattice that t1~t7, which is the format of the successive video frames and the second live video stream in the second live video stream, in Fig. 7 b
Formula, the short-sighted frequency of target includes two video frames of T1 and T2, when identifying that t3 video frame meets short-sighted frequency pushing condition, then by T1
Replace t3 and t4 video frame with T2 video frame, the second live video stream replaced in this way are as follows: t1, t2, T1, T2, t5, t6 and
T7, not only complete in this way the broadcasting of the short-sighted frequency of target and also when the short-sighted frequency of target finish playing subsequent continued broadcasting put the second live streaming view
Video in frequency stream.
Optionally, in the short-sighted frequency of push target, just inevitably there are some video frames not in the second live video stream
It can be pushed to live streaming client, therefore in order to avoid video frame important and/or excellent in the second live video stream is not pushed away
Live streaming client is given, present invention proposition further includes following processes:
When pushing the short-sighted frequency of the target, continue the video frame for identifying second live video stream, if identifying the
Two video frames meet short-sighted frequency and stop pushing condition, then since the timestamp of second video frame, it is short to stop push target
Video, and second live video stream is pushed since second video frame.
Specifically, when the second live video stream is game live video stream, can exist in game live video stream
More excellent video frame, and the short-sighted frequency of target pushed may result in the live streaming visitor that more excellent video frame does not push
The user at family end, in this way live streaming client also could not just see excellent live video stream, and then it is straight to second to will cause user
The interest-degree for broadcasting video flowing is not high；In order to avoid this problem, the embodiment of the present invention is proposed, to meet short-sighted frequency pushing condition
Video frame is the i-th video frame, then during pushing the short-sighted frequency of target in the second live video stream, continues the second live streaming of identification
Timestamp is later than the video frame of the timestamp of above-mentioned video frame in video flowing, such as will continue to identification i+1 video frame, the i-th+2 view
Frequency frame ..., when identifying that there are the i-th+5 video frames (the second video frame) to meet short-sighted frequency stop condition, i.e. the i-th+5 video frame
Content be splendid contents when, then can stop push the short-sighted frequency of target, then start the i-th+5 video frame of push and i-th+6 view
Frequency frame ..., such user will not both miss the featured videos in the second live video stream and watch target short-sighted
Frequently.As shown in connection with fig. 8, t1~t7 is successive video frames in the second live video stream in Fig. 8, the short-sighted frequency of target include T1~
T8 video frame then since the timestamp of t3 video frame, pushes mesh when identifying that t3 video frame meets short-sighted frequency pushing condition
T1 video frame, the T2 video frame etc. of short-sighted frequency are marked, the video frames such as t4, t5 will not be pushed at this time, but will continue to regard t4, t5 etc.
Frequency frame continues to identify, when identifying that t5 video frame (the second video frame) meets short-sighted frequency stop condition, then would not push at this time
T3 video frame (is illustrated) so that video frame interval time is identical as an example, then continues the push t5 video in the second live video stream
Frame, t6 video frame etc..
It should be noted that the short-sighted frequency pushing condition and short-sighted frequency stopping pushing condition in the embodiment of the present invention can roots
It is set according to actual conditions, it is intended to not influence the live streaming of splendid contents in the second live video stream.Optionally, in identification second
When video frame in live video stream, it can be identified using existing method, can also determine thing present in video frame
Part, then judges whether existing event meets short-sighted frequency pushing condition etc..Or it is pre-configured with and some meets short video push
The video frame of condition, then by the second live video stream video frame and the above-mentioned video frame for meeting short-sighted frequency pushing condition into
Row matches one by one, determines that the video frame met in the second live video stream meets short-sighted frequency pushing condition if successful match.
Optionally, if second live video stream is game live video stream, meet the of short-sighted frequency pushing condition
One video frame is included at least with the next item down: being belonged in video frame and video frame between adjacent two innings of game in setting regions
There is no target objects.
Specifically, when the second live video stream be game live video stream when, then can between two innings of game (game play
Between) the short-sighted frequency of target is pushed, target short-sighted frequency can also be pushed in sowing time under main broadcaster, this purpose is based on, if identifying game
When video frame in live video stream belongs to the video frame between two innings of game, shows to be currently entering game interoffice, then may be used
To push the short-sighted frequency of target；If identifying the region (setting regions) in the video frame in game live video stream where main broadcaster's frame
In be not present target object when, then show to broadcast under main broadcaster, the short-sighted frequency of target can be pushed at this time, refering to what is shown in Fig. 9, based on Fig. 9
Sow the short-sighted frequency of target --- the excellent collection of choice specimens video schematic diagram of sowing time broadcasting.
Optionally, when the second live video stream is game live video stream, then it is based on any of the above-described embodiment, is being obtained
When the short-sighted frequency of target, following processes can also be implemented: confirming that the short-sighted frequency of the target meets game interoffice pushing condition.
Specifically, the game interoffice pushing condition of Game Setting different in the present invention may be different or the same, and have
Body can be set according to the actual situation.Since the game interoffice time is shorter, therefore if the short-sighted frequency of target generated is too long
Or excellent degree is not enough, then will not play the short-sighted frequency of target in game interoffice.And it broadcasts and shows in current live video under main broadcaster
Appearance is not excellent to be explained without main broadcaster, then can arbitrarily push the short-sighted frequency of target at this time.
Specifically, when candidate video frame meets short video intercepting condition, game in the candidate video frame can also be identified
Status information, if game state information meets game interoffice pushing condition, when detecting the video in game live video stream
When frame belongs to the video frame between two innings of game, then allow to push the mesh generated based on the candidate video frame in game interoffice
Short-sighted frequency is marked, does not otherwise allow to push the short-sighted frequency of target generated based on the candidate video frame in game interoffice.Different game
Corresponding game state information is different, specifically can according to the actual situation depending on.
Shown in Figure 5, the game state information of the danger spot survival game in Fig. 5 can be, but not limited to are as follows: game is surplus
Remaining number and total number of persons etc. is killed, the game interoffice pushing condition of the danger spot survival game being then arranged can be with are as follows: game is surplus
Remaining number is not higher than the first volume residual threshold value and/or kills total number of persons and kills amount threshold etc. not less than first, then when being based on
Fig. 5 identifies game residue number and after killing total number of persons, it can be determined that the game residue number that identifies and kills total number of persons
The game interoffice pushing condition for whether meeting above-mentioned danger spot survival game enters trip when determining based on the above embodiment if meeting
When play interoffice, according to the short-sighted frequency of target for candidate video frame of the method for pushing push comprising Fig. 5 that any of the above-described embodiment provides.
Shown in Figure 6, the game state information of heroic alliance's game in Fig. 6 can be, but not limited to are as follows: game is surplus
Remaining number, our remaining number, assists, kills number etc. at enemy's residue number, the heroic alliance travel based on this setting
The game interoffice pushing condition of play can be, but not limited to are as follows: game residue number is surplus not higher than the second volume residual threshold value, enemy
Remainder amount is not low not less than the 4th volume residual threshold value, secondary attack number not higher than third volume residual threshold value, our volume residual
In preset quantity threshold value and/or kills number and kill amount threshold etc. not less than second.Heroic alliance travel is identified when being based on Fig. 6
After the game state information of play, the upper right corner " 5/5/10 " respectively indicates the number of killing/death number/assists in Fig. 6, can be with
Judge whether above-mentioned game state information meets the game interoffice pushing condition of heroic alliance's game, when based on upper if meeting
It states embodiment and determines that the method for pushing push provided according to any of the above-described embodiment includes the candidate of Fig. 6 when entering game interoffice
The short-sighted frequency of the target of video frame.
It should be noted that above-mentioned game residue number, enemy's residue number, our remaining number, secondary attack number, killing
Number can be based on digital Matching Model and identify, the training process of the model and the training process class of content type identification model
Seemingly, it no longer describes herein.
As shown in connection with fig. 10, game live video stream includes t1~t8 video frame in Figure 10, when according to the present invention above-mentioned
After one embodiment generates the short-sighted frequency of heroic alliance target, which includes T1~T5 video frame, however, it is determined that
The heroic short-sighted frequency of alliance target meets game interoffice playing condition, then t3 video frame (in identifying game live video stream
One video frame) belong to video frame between adjacent two innings of game, that is, meet short-sighted frequency pushing condition, then in t3 video frame pair
The timestamp answered starts to push T1 video frame, T2 video frame；When identify t5 video frame (the second video frame) be featured videos frame,
When meeting short-sighted frequency stopping pushing condition, then T3 video frame is not pushed, start to push t5 video frame, t6 video frame, t7 video
Frame；When identifying that there is no target objects, i.e. main broadcaster's lower sowing time in setting regions in t8 video frame (the first video frame), then at this time
It can continue to push T1~T5 video frame that the short-sighted frequency of heroic alliance target includes, the short-sighted frequency of other targets can also be pushed.Tool
Body, when the short-sighted frequency of heroic alliance target is unsatisfactory for game interoffice playing condition, then detecting that t3 video frame is game play
Between video frame when, then the above-mentioned heroic short-sighted frequency of alliance target will not be pushed at this time, only when detecting that t8 video frame is main broadcaster
Under when broadcasting video frame, can just push the short-sighted frequency of above-mentioned heroic alliance target.
It further include following mistakes after the video-frequency band based on interception generates the short-sighted frequency of target based on any of the above-described embodiment
Journey:
Determine that the dispensing area information of the short-sighted frequency of the target, above-mentioned dispensing area information include in the short-sighted frequency of the target
The area information for launching data can be loaded in video frame.
Herein on basis, in the short-sighted frequency of the push target, the dispensing area information is pushed to the live streaming
Client so that after the live streaming client receives the dispensing area information, obtain data to be put and based on it is described can
The area information that data are launched in load launches the data to be put.
Specifically, the dispensing area information in the present invention may include that can load the area information for launching data；Above-mentioned throwing
Putting area information can also include dispensing duration and dispensing form etc..Specifically, it is determined that is gone out loads the region for launching data
Information can indicate are as follows: the coordinate of central point can load the length and width for launching data area.It is, of course, also possible to there is other expressions
Method, specifically can according to the actual situation depending on.
Specifically, after obtaining the short-sighted frequency of target, when identifying whether candidate video frame meets short video intercepting condition, also
Can identify can load the area information for launching data in the candidate video frame, can similarly identify its in the short-sighted frequency of target
The area information for launching data can be loaded in his video frame, can filter out the number of the video frame with same area information in this way
Amount, the quantity determination for being then based on video frame can launch the dispensing duration of data.Above-mentioned dispensing form can be dispensing dynamic number
According to, or launch static data etc..
Specifically, loading in the present invention is launched data and be can be, but not limited to as advertisement, text etc., is if launching data
Advertisement, then can load launch data area be can carrying advertisement region.For convenience, the subsequent throwing to be embedded in live streaming
Data are put to be illustrated for advertisement.
Specifically, it can be loaded according to main broadcaster position in the content of video frame in the short-sighted frequency of target, video frame to determine
Launch the area information of data.Can also be set according to the demand of live streaming platform can load the area information etc. for launching data.
It is alternatively possible to which dispensing data can be loaded by determining in the video frame in the short-sighted frequency of target according to following flow implementations
Area information, by determine candidate video frame in the short-sighted frequency of target load the area information for launching data for said
It is bright, comprising the following steps: according to candidate video frame and trained zone location model, determining in candidate video frame can not be blocked
The area information and confidence level of candidate region；And according to the confidence level that can not block candidate region, determination can not occlusion area；So
Afterwards according to can not occlusion area area information, determine can be loaded in candidate video frame launch data area information.
By implementing the process of the above-mentioned area information for loading dispensing data, main broadcaster's frame region as can not in Figure 11 a
Occlusion area, then based on this can not occlusion area area information, selection close to this can not occlusion area and be not belonging to this can not
The fringe region of occlusion area, which is determined as to load, launches data area, and then output can load the region letter for launching data area
Breath.It please refers to shown in Figure 11 b, is given in Figure 11 b and load dispensing data area based on what main broadcaster's frame region was determined, when embedding
The data entered be advertisement when, then can be loaded in Figure 11 b launch data area be exactly can carrying advertisement region.
It optionally, can also be according to following when can load the area information for launching data in determining the candidate video frame
Implementation Process: determining the variation degree between the candidate video frame and previous video frame, if variation degree is greater than predeterminable level threshold
Value, then the area information for launching data can be loaded by being determined based on above-mentioned process in the candidate video frame.If variation degree is less than pre-
If degree threshold value, then the area information for launching data will can be directly loaded in previous video frame as can add in the candidate video frame
Carry the area information for launching data.The position for loading data area for being sent to live streaming client in this way is relatively fixed, thus
So that live streaming client launches advertisement in fixed position, while important content in live video will not be blocked.
Optionally, in same application scene, it is solid that the area information in the region of advertisement can be launched in some video frames
Fixed, for example, the surface of main broadcaster's frame region or left side fixed position are typically used for throwing in some scene of game
Put advertisement, therefore each interface in a game can be analyzed in advance, then filter out it is some can be in fixation
The video frame of advertisement is launched in position, and records the area information that can launch advertising area in these video frames and stored, lattice
Formula is shaped like (video frame images can load the area information for launching data in video frame images).In this way, working as from the short-sighted frequency of target
In get candidate video frame after, which can be matched with the above-mentioned video frame filtered out, if matching at
The area information for loading dispensing data of the video frame of successful match is then determined as loading in the candidate video frame by function
Launch the area information of data；If matching unsuccessful, loading in the candidate video frame is determined according still further to above-mentioned process
Launch the area information of data.
It is determining that the area information for launching data can be loaded in the short-sighted frequency of target in video frame according to any of the above-described method
Afterwards, it can be carried and be sent to live streaming client together with the short-sighted frequency of target of push in launching area information, it is straight in this way
Client is broadcast when watching live streaming, the live streaming of viewing is the short-sighted frequency of target if identifying, based on the throwing launched in area information
Duration and form are put from acquisition in ad system not higher than the form phase in the dispensing duration and form and dispensing area information
Then the advertisement matched is believed based on the area information that can load dispensing data in area information in video frame is launched in the region again
Indicated the loading of breath launches data area and launches the advertisement got, and with reference to shown in Figure 11 b, Figure 11 b, which gives, to be added
Carry the effect diagram that advertising area launches advertisement.This way it is possible to realize live streaming client launches advertisement in suitable region
Purpose, while content important in video frame will not be blocked.
Optionally, live streaming client can also implement following processes: determine currently playing before launching advertisement to be put
Time is not in data frequency control range.
Specifically, it is illustrated so that data frequency control range is advertisement frequency control range as an example, in order to avoid frequently launching advertisement
The case where occur, before getting advertisement to be put, if current play time not advertisement frequency control range in if can work as
Advertisement to be put is launched in the video frame of the preceding short-sighted frequency of target being played on, otherwise abandons launching advertisement to be put.
Optionally, it can also be controlled in the frequency that server side launch advertisement, such as be pushed away to live streaming client
When sending the short-sighted frequency of target, the last time for sending and launching area information can be determined, however, it is determined that above-mentioned time and push target
Difference between the time of short-sighted frequency is less than time interval threshold value, then it is short-sighted not send the target when the short-sighted frequency of this push target
The dispensing area information of frequency, specifically, time interval threshold value can be, but not limited to be 30 minutes etc..
Based on any of the above-described embodiment, after implementing process shown in Fig. 4, further includes: can by candidate video frame and
Recognition result, such as the content type, the event type that identify are stored, and event-monitoring is then carried out, and please refer to Figure 12 institute
The event-monitoring schematic diagram shown, by being monitored to the event in the short-sighted frequency of target (excellent collection of choice specimens video), it can be determined that mesh
Mark whether short-sighted frequency (excellent collection of choice specimens video) meets desired effect, to obtain the preferably short-sighted frequency of target (excellent collection of choice specimens video).
Method provided by the invention, by capturing the candidate view in the first live video stream pushed during live streaming
Frequency frame, interception includes the view of candidate short-sighted frequency from the first live video stream if candidate video frame meets short video intercepting condition
Frequency range, and then the short-sighted frequency of target is generated based on the video-frequency band, when short video intercepting condition is that the excellent collection of choice specimens intercepts condition, thus
Also it is achieved that the effect for automatically generating the excellent collection of choice specimens video during live streaming, also avoids the prior art by manually cutting
Volume mode the problem of making high labor cost caused by excellent collection of choice specimens generation；In addition, after generating the short-sighted frequency of target, this hair
It is bright to propose to push the short-sighted frequency of target during to live streaming the second live video stream of client push, realize and be broadcast live
The effect that excellent collection of choice specimens video is played in journey not only avoids the prior art by reviewing the case where mode plays target short-sighted frequency
Occur, and solves and specific broadcasting staff is needed to play cost of labor caused by excellent collection of choice specimens video in the prior art
High problem.
Based on the same inventive concept, a kind of video push device based on live streaming is additionally provided in the embodiment of the present invention, it should
Device can be set in the server 104 in Fig. 2, since the principle that above-mentioned apparatus solves the problems, such as is pushed away with the video based on live streaming
Delivery method is similar, therefore the implementation of above-mentioned apparatus may refer to the implementation of method, and overlaps will not be repeated.
It as shown in figure 13, is the structural schematic diagram of the video push device provided in an embodiment of the present invention based on live streaming, packet
It includes: identification module 131 and video push module 132, in which:
Identification module 131, for during to live streaming the second live video stream of client push, identification second to be broadcast live
Video frame in video flowing；
Video push module 132, if the first video frame for identifying meets short-sighted frequency pushing condition, from described
The timestamp of one video frame starts, and pushes the short-sighted frequency of target, wherein the short-sighted frequency of target is based on the first live streaming pushed
What video flowing generated.
Optionally, the identification module 131 is also used to continue to identify described second in the short-sighted frequency of the push target
The video frame of live video stream；
The video push module 132, if the second video frame for being also used to identify meets short-sighted frequency and stops pushing condition,
Then since the timestamp of second video frame, stop the push short-sighted frequency of target, and push since second video frame
Second live video stream.
Optionally, the video push device provided by the invention based on live streaming, further includes:
Module 133 is obtained, for obtaining candidate video frame from the first live video stream pushed；
Video processing module 134 is straight from described first for when the candidate video frame meets short video intercepting condition
Broadcast the video-frequency band that interception in video flowing includes the candidate video frame；
Short-sighted frequency generation module 135 generates the short-sighted frequency of target for the video-frequency band based on interception.
Optionally, the video push device provided by the invention based on live streaming, further includes:
Determining module 136, for determining that the dispensing area information of the short-sighted frequency of the target, the dispensing area information include
The area information for launching data can be loaded in the short-sighted frequency of target in video frame；
Herein on basis, video push module 132 is also used in the short-sighted frequency of the push target, by the dispensing
Area information is pushed to the live streaming client, so as to obtain after the live streaming client receives the dispensing area information
Data to be put simultaneously launch the data to be put based on the area information for loading dispensing data.
Optionally, the video processing module 134, specifically for identifying the content type of the candidate video frame；If institute
Stating content type is setting content type, then carries out event detection to the candidate video frame, and when the determining candidate video
There are the event types for when event, determining existing event in frame；If the event type is setting event type, it is determined that institute
It states candidate video frame and meets short video intercepting condition.
Optionally, the video processing module 134, specifically for intercepting key area from the candidate video frame；It is right
The key area carries out Text region, is determined in the candidate video frame based on Text region result with the presence or absence of event.
Optionally, the video processing module 134 is specifically used for being examined according to the candidate video frame and trained event
Model is surveyed, determines the candidate video frame with the presence or absence of event, wherein the training sample of the event detection model are as follows: exist
The event image pattern and the image pattern in addition to event image pattern of event.
Optionally, the video push device provided by the invention based on live streaming, further includes:
Confirmation module 137, for the acquisition module 133 obtained from the first live video stream candidate video frame it
Before, confirm and carries video identification mark in first live video stream.
Optionally, described to meet short-sighted frequency pushing condition if second live video stream is game live video stream
The first video frame include at least with the next item down: set area in the video frame and video frame belonged between adjacent two innings of game
Target object is not present in domain.
For convenience of description, above each section is divided by function describes respectively for each module (or unit).Certainly, exist
Implement to realize the function of each module (or unit) in same or multiple softwares or hardware when the present invention.
Specifically, after having introduced the video push device provided in an embodiment of the present invention based on live streaming, in conjunction with Fig. 2 and
Shown in Figure 14, the video push device provided by the invention based on live streaming further includes control module, transcoding module, content distribution
CDN module, and video processing module 134 and determining module 136 are set to video and understand in engine, are described based on above-mentioned apparatus
Video pushing method provided in an embodiment of the present invention based on live streaming, substantially process are as follows: when main broadcaster user 1 is based on terminal device
During the live streaming room that live streaming client in 101 provides is broadcast live, the video push module 132 in server is being obtained
To after the first live video stream pushed, first live video stream is then sent to control module, since platform is broadcast live
The live streaming room for needing to generate excellent collection of choice specimens video is relatively more, therefore the short-sighted frequency generation module 135 in the multiple Figure 13 of possible configuration,
And the first live video stream received is sent to the opposite short-sighted frequency reduced of load for carrying out load balancing by control module
Generation module 135；Video identification request, such video processing are sent from short-sighted frequency generation module 135 to video processing module 134
Module 134 sends video frame acquisition request upon receiving the request, to module 133 is obtained, and is connecing in this way, obtaining module 133
After receiving the request, candidate video frame is obtained from the first live video stream, is then sent to video and is understood video in engine
Processing module 134, by video processing module 134 when identifying that the candidate video frame meets short video intercepting condition, from described
Interception includes the video-frequency band of the candidate video frame and is sent to video generation module 135 in first live video stream, by video
Generation module 135 generates the short-sighted frequency of target based on the video-frequency band of interception.Enter and 1 phase of user when user 2 is based on live streaming client
With live streaming room when, can understand that the video processing module 134 in engine sends notice, to video to allow video processing module
134 know that user 2 watches the live streaming of user 1, at this time after generation target short-sighted frequency, since main broadcaster user 1 is still carrying out directly
It broadcasts, therefore the second live video stream can be generated, in this way, short-sighted frequency generation module 135 can give short-sighted take place frequently of target to video push
Module 132, from video push module 132 during to live streaming the second live video stream of client push, push target is short
Video；Specifically, it during push, needs video stream carrying out transcoding to transcoding module, then by the view after transcoding
Frequency stream is sent to CDN module, by CDN module by pushing video streaming to the live streaming client in the terminal device 102 in Fig. 2, this
The user 2 of sample terminal device 102 is perceived by the second live streaming when opening respective live room when entering live streaming client
The video flowing of video flowing or the short-sighted frequency of target.In addition, video understands that the determining module 136 in engine can determine the short-sighted frequency of target
Dispensing area information, then by launch area information be sent to CDN module, when user 2 live streaming client terminal playing live streaming when,
If it is determined that currently playing video frame belongs to the short-sighted frequency of target, then can timing per second obtained to CDN and launch area information, then from
Ad system obtains the advertisement to match with dispensing area information, and the location information for then including in launching area information is signified
The advertisement got is launched at the position shown.
Optionally, the CDN module in Figure 14 can be replaced by long link control module, i.e., used by long link control module
The form of websocket long connection sends order to live streaming client, launches area information based on order transmitting, may be used also certainly
To use other modes, the invention does not limit this.
It further, is the structural schematic diagram of video processing module 134 with reference to shown in Figure 15, video processing module 134 wraps
Content type identification module 1341, picture preprocessing module 1342, event type identification module 1343 are included, then when acquisition module
After 133 get the candidate video frame in the first live video stream, candidate video frame is sent to content type identification module
1341, the content type of candidate video frame is identified by content type identification module 1341, in identifying content type to set
When holding type, candidate video frame is sent to picture preprocessing module 1342, by picture preprocessing module 1342 from candidate video
Key area is intercepted in frame, is then sent to event type identification module 1343, when event type identification module 1343 receives
After key area, OCR service can be called to identify the text of key area, be then based on the text that identifies to identify candidate
The event type of event present in video frame, when the event type identified is setting event type, from the first live streaming view
In frequency stream then video-frequency band is sent to video generation module 135 by video-frequency band of the interception comprising candidate video frame；In addition, working as content
When type identification module 1341 identifies that the content type of candidate video frame is not setting content type, then process terminates.
In the video pushing method based on live streaming, the device and computer-readable for describing exemplary embodiment of the invention
After storage medium, next, introducing the calculating equipment of another exemplary embodiment according to the present invention.
Person of ordinary skill in the field it is understood that various aspects of the invention can be implemented as system, method or
Program product.Therefore, various aspects of the invention can be embodied in the following forms, it may be assumed that complete hardware embodiment, complete
The embodiment combined in terms of full Software Implementation (including firmware, microcode etc.) or hardware and software, can unite here
Referred to as circuit, " module " or " system ".
In some possible embodiments, calculating equipment according to the present invention can include at least at least one processing
Device and at least one processor.Wherein, the memory is stored with program code, when said program code is by the processing
When device executes, so that the processor executes the base of the illustrative embodiments various according to the present invention of this specification foregoing description
Step in the video pushing method of live streaming.For example, the processor can execute step S31~step as shown in Figure 3
The video push process based on live streaming in S33.
The calculating equipment 160 of this embodiment according to the present invention is described referring to Figure 16.The meter that Figure 16 is shown
Calculating equipment 160 is only an example, should not function to the embodiment of the present invention and use scope bring any restrictions.
As shown in figure 16, equipment 160 is calculated to show in the form of universal computing device.The component for calculating equipment 160 can be with
Including but not limited to: at least one above-mentioned processor 161, above-mentioned at least one processor 162, connection different system components (packet
Include memory 162 and processor 161) bus 163.
Bus 163 indicates one of a few class bus structures or a variety of, including memory bus or Memory Controller,
Peripheral bus, processor or the local bus using any bus structures in a variety of bus structures.
Memory 162 may include the readable medium of form of volatile memory, such as random access memory (RAM)
1621 and/or cache memory 1622, it can further include read-only memory (ROM) 1623.
Memory 162 can also include program/utility 1625 with one group of (at least one) program module 1624,
Such program module 1624 includes but is not limited to: operating system, one or more application program, other program modules and
It may include the realization of network environment in program data, each of these examples or certain combination.
Calculating equipment 160 can also communicate with one or more external equipments 164 (such as keyboard, sensing equipment etc.), also
Can be enabled a user to one or more with calculate equipment 160 interact equipment communication, and/or with make the calculating equipment
The 160 any equipment (such as router, modem etc.) that can be communicated with one or more of the other calculating equipment are led to
Letter.This communication can be carried out by input/output (I/O) interface 165.Also, calculating equipment 160 can also be suitable by network
Orchestration 166 and one or more network (such as local area network (LAN), wide area network (WAN) and/or public network, such as because of spy
Net) communication.As shown, network adapter 166 is communicated by bus 163 with other modules for calculating equipment 16.It should
Understand, although not shown in the drawings, other hardware and/or software module can be used in conjunction with equipment 160 is calculated, including but unlimited
In: microcode, device driver, redundant processor, external disk drive array, RAID system, tape drive and data
Backup storage system etc..
Based on the same inventive concept, the embodiment of the present invention also provides a kind of storage medium, which is stored with calculating
Machine instruction, when the computer instruction is run on computers, so that computer executes the video based on live streaming as the aforementioned and pushes away
The step of delivery method.
Based on the same inventive concept, the embodiment of the present invention also provides a kind of video process apparatus, the video process apparatus packet
At least one processor and readable storage medium storing program for executing are included, when the instruction for including in the readable storage medium storing program for executing is by least one processor
When execution, the step of video pushing method based on live streaming as the aforementioned can be executed.
Based on the same inventive concept, the embodiment of the present invention also provides a kind of chip system, which includes processor,
It can also include memory, the step of for realizing video pushing method based on live streaming as the aforementioned.The chip system can be with
It is made of chip, also may include chip and other discrete devices.
In some possible embodiments, the video pushing method provided in an embodiment of the present invention based on live streaming is each
Aspect is also implemented as a kind of form of program product comprising program code, when described program product is transported on computers
When row, said program code is used to that the computer to be made to execute the illustrative embodiments various according to the present invention for stating description above
Letter based on the step in the video pushing method of live streaming.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications can be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (12)
1. a kind of video pushing method based on live streaming characterized by comprising
During to live streaming the second live video stream of client push, the video frame in the second live video stream is identified；
If the first video frame identified meets short-sighted frequency pushing condition, since the timestamp of first video frame, push away
Send target short-sighted frequency, wherein the short-sighted frequency of target is to generate based on the first live video stream pushed.
2. the method as described in claim 1, which is characterized in that further include:
In the short-sighted frequency of the push target, continue the video frame for identifying second live video stream；
If the second video frame identified, which meets short-sighted frequency, stops pushing condition, opened from the timestamp of second video frame
Begin, stops the push short-sighted frequency of target, and push second live video stream since second video frame.
3. method according to claim 1 or 2, which is characterized in that the short-sighted frequency of target is straight based on first pushed
Video flowing generation is broadcast, is specifically included:
Candidate video frame is obtained from the first live video stream pushed；
When the candidate video frame meets short video intercepting condition, interception includes the time from first live video stream
Select the video-frequency band of video frame；
Video-frequency band based on interception generates the short-sighted frequency of target.
4. method as claimed in claim 3, which is characterized in that further include:
Determine that the dispensing area information of the short-sighted frequency of the target, the dispensing area information include video in the short-sighted frequency of the target
The area information for launching data can be loaded in frame；And
In the short-sighted frequency of the push target, the dispensing area information is pushed to the live streaming client, so that described straight
After podcast family termination receives the dispensing area information, data to be put are obtained and based on the region for loading dispensing data
Information launches the data to be put.
5. method as claimed in claim 3, which is characterized in that it is short-sighted to determine that the candidate video frame meets by the following method
Frequency interception condition:
Identify the content type of the candidate video frame；
If the content type is setting content type, event detection is carried out to the candidate video frame, and work as described in determination
There are the event types for when event, determining existing event in candidate video frame；
If the event type is setting event type, it is determined that the candidate video frame meets short video intercepting condition.
6. method as claimed in claim 5, which is characterized in that carry out event detection to the candidate video frame, specifically include:
Key area is intercepted from the candidate video frame；
Text region is carried out to the key area, is determined in the candidate video frame based on Text region result with the presence or absence of thing
Part.
7. method as claimed in claim 5, which is characterized in that carry out event detection to the candidate video frame, specifically include:
According to the candidate video frame and trained event detection model, determine that the candidate video frame whether there is event,
Wherein, the training sample of the event detection model are as follows: there are the event image pattern of event and in addition to event image pattern
Image pattern.
8. method as claimed in claim 3, which is characterized in that from the first live video stream obtain candidate video frame it
Before, further includes:
Confirm and carries video identification mark in first live video stream.
9. method according to claim 1 or 2, which is characterized in that if second live video stream is game live video
When stream, first video frame for meeting short-sighted frequency pushing condition is included at least with the next item down: being belonged between adjacent two innings of game
Between video frame and video frame in target object is not present in setting regions.
10. a kind of video push device based on live streaming characterized by comprising
Identification module, for identifying the second live video stream during to live streaming the second live video stream of client push
In video frame；
Video push module, if the first video frame for identifying meets short-sighted frequency pushing condition, from first video
The timestamp of frame starts, and pushes the short-sighted frequency of target, wherein the short-sighted frequency of target is based on the first live video stream pushed
It generates.
11. a kind of computer equipment including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, which is characterized in that the processor is realized when executing described program such as claim 1 to 9 any claim institute
The method stated.
12. a kind of computer-readable medium, is stored with processor-executable instruction, which is characterized in that the processor is executable
Instruction is for executing the method as described in claim 1 to 9 any claim.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910341639.9A CN110198456B (en) | 2019-04-26 | 2019-04-26 | Live broadcast-based video pushing method and device and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910341639.9A CN110198456B (en) | 2019-04-26 | 2019-04-26 | Live broadcast-based video pushing method and device and computer-readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110198456A true CN110198456A (en) | 2019-09-03 |
| CN110198456B CN110198456B (en) | 2023-02-07 |
Family
ID=67752178
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910341639.9A Active CN110198456B (en) | 2019-04-26 | 2019-04-26 | Live broadcast-based video pushing method and device and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110198456B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110688498A (en) * | 2019-09-27 | 2020-01-14 | 北京奇艺世纪科技有限公司 | Information processing method and device, electronic equipment and storage medium |
| CN111083515A (en) * | 2019-12-31 | 2020-04-28 | 广州华多网络科技有限公司 | Method, device and system for processing live broadcast content |
| CN111405302A (en) * | 2020-03-06 | 2020-07-10 | 北京达佳互联信息技术有限公司 | Live broadcast room attention reminding method and device, electronic equipment and storage medium |
| CN111428660A (en) * | 2020-03-27 | 2020-07-17 | 腾讯科技（深圳）有限公司 | Video editing method and device, storage medium and electronic device |
| CN111814640A (en) * | 2020-06-30 | 2020-10-23 | 北京玩在一起科技有限公司 | Method and system for detecting live event of electric competition and extracting information |
| CN112235587A (en) * | 2020-10-12 | 2021-01-15 | 北京字节跳动网络技术有限公司 | Live broadcast method and equipment |
| CN112348649A (en) * | 2020-12-07 | 2021-02-09 | 福州市长乐区三互信息科技有限公司 | Live broadcast category recommendation method |
| CN113542803A (en) * | 2021-07-09 | 2021-10-22 | 上海哔哩哔哩科技有限公司 | Live content processing method and device |
| CN115022654A (en) * | 2022-05-18 | 2022-09-06 | 北京达佳互联信息技术有限公司 | Video editing method and device in live scene |
| CN115022663A (en) * | 2022-06-15 | 2022-09-06 | 北京奇艺世纪科技有限公司 | Live stream processing method and device, electronic equipment and medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103618950A (en) * | 2013-11-11 | 2014-03-05 | 李达航 | Inter-cut method of content of smart television |
| CN104065982A (en) * | 2014-06-19 | 2014-09-24 | 腾讯科技（深圳）有限公司 | Streaming media live broadcast method and device |
| WO2016000429A1 (en) * | 2014-06-30 | 2016-01-07 | 中兴通讯股份有限公司 | Method and device for detecting video conference hotspot scenario |
| CN107682717A (en) * | 2017-08-29 | 2018-02-09 | 百度在线网络技术（北京）有限公司 | Service recommendation method, device, equipment and storage medium |
| CN108810615A (en) * | 2017-04-28 | 2018-11-13 | 阿里巴巴集团控股有限公司 | The method and apparatus for determining spot break in audio and video |
| CN108833936A (en) * | 2018-05-25 | 2018-11-16 | 广州虎牙信息科技有限公司 | Direct broadcasting room information-pushing method, device, server and medium |
| CN109074828A (en) * | 2016-04-01 | 2018-12-21 | 乐威指南公司 | For providing the system and method for the playlist for the user's related content for replacing ad content to be played back |
| CN109155136A (en) * | 2016-04-01 | 2019-01-04 | 奥誓公司 | Computerized system and method for automatically detecting and rendering highlights from video |
-
2019
- 2019-04-26 CN CN201910341639.9A patent/CN110198456B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103618950A (en) * | 2013-11-11 | 2014-03-05 | 李达航 | Inter-cut method of content of smart television |
| CN104065982A (en) * | 2014-06-19 | 2014-09-24 | 腾讯科技（深圳）有限公司 | Streaming media live broadcast method and device |
| WO2016000429A1 (en) * | 2014-06-30 | 2016-01-07 | 中兴通讯股份有限公司 | Method and device for detecting video conference hotspot scenario |
| CN109074828A (en) * | 2016-04-01 | 2018-12-21 | 乐威指南公司 | For providing the system and method for the playlist for the user's related content for replacing ad content to be played back |
| CN109155136A (en) * | 2016-04-01 | 2019-01-04 | 奥誓公司 | Computerized system and method for automatically detecting and rendering highlights from video |
| CN108810615A (en) * | 2017-04-28 | 2018-11-13 | 阿里巴巴集团控股有限公司 | The method and apparatus for determining spot break in audio and video |
| CN107682717A (en) * | 2017-08-29 | 2018-02-09 | 百度在线网络技术（北京）有限公司 | Service recommendation method, device, equipment and storage medium |
| CN108833936A (en) * | 2018-05-25 | 2018-11-16 | 广州虎牙信息科技有限公司 | Direct broadcasting room information-pushing method, device, server and medium |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110688498A (en) * | 2019-09-27 | 2020-01-14 | 北京奇艺世纪科技有限公司 | Information processing method and device, electronic equipment and storage medium |
| CN110688498B (en) * | 2019-09-27 | 2022-07-22 | 北京奇艺世纪科技有限公司 | Information processing method and device, electronic equipment and storage medium |
| CN111083515B (en) * | 2019-12-31 | 2021-07-23 | 广州华多网络科技有限公司 | Method, device and system for processing live broadcast content |
| CN111083515A (en) * | 2019-12-31 | 2020-04-28 | 广州华多网络科技有限公司 | Method, device and system for processing live broadcast content |
| CN111405302A (en) * | 2020-03-06 | 2020-07-10 | 北京达佳互联信息技术有限公司 | Live broadcast room attention reminding method and device, electronic equipment and storage medium |
| CN111428660A (en) * | 2020-03-27 | 2020-07-17 | 腾讯科技（深圳）有限公司 | Video editing method and device, storage medium and electronic device |
| CN111428660B (en) * | 2020-03-27 | 2023-04-07 | 腾讯科技（深圳）有限公司 | Video editing method and device, storage medium and electronic device |
| CN111814640A (en) * | 2020-06-30 | 2020-10-23 | 北京玩在一起科技有限公司 | Method and system for detecting live event of electric competition and extracting information |
| CN112235587A (en) * | 2020-10-12 | 2021-01-15 | 北京字节跳动网络技术有限公司 | Live broadcast method and equipment |
| CN112348649A (en) * | 2020-12-07 | 2021-02-09 | 福州市长乐区三互信息科技有限公司 | Live broadcast category recommendation method |
| CN112348649B (en) * | 2020-12-07 | 2022-07-29 | 福州灵和晞科技有限公司 | Live broadcast category recommendation method |
| CN113542803A (en) * | 2021-07-09 | 2021-10-22 | 上海哔哩哔哩科技有限公司 | Live content processing method and device |
| CN115022654A (en) * | 2022-05-18 | 2022-09-06 | 北京达佳互联信息技术有限公司 | Video editing method and device in live scene |
| CN115022654B (en) * | 2022-05-18 | 2024-01-19 | 北京达佳互联信息技术有限公司 | Video editing method and device in live broadcast scene |
| CN115022663A (en) * | 2022-06-15 | 2022-09-06 | 北京奇艺世纪科技有限公司 | Live stream processing method and device, electronic equipment and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110198456B (en) | 2023-02-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110198456A (en) | Video pushing method, device and computer readable storage medium based on live streaming | |
| CN110784759B (en) | Bullet screen information processing method and device, electronic equipment and storage medium | |
| US8487866B2 (en) | Method and system for managing an interactive video display system | |
| CN102467661B (en) | Multimedia device and method for controlling the same | |
| CN101354745B (en) | Method and apparatus for recognizing video document | |
| CN110099283A (en) | Information-pushing method, device, equipment and storage medium | |
| CN110703913B (en) | Object interaction method and device, storage medium and electronic device | |
| CN107851103A (en) | Automated content identification fingerprint sequence matching | |
| CN113438493B (en) | Popup animation generation method, server, system and storage medium | |
| CN103365936A (en) | Video recommendation system and method thereof | |
| CN106534946A (en) | Control method and device for video playing | |
| CN106878767A (en) | Video broadcasting method and device | |
| CN109783686A (en) | Behavioral data processing method, device, terminal device and storage medium | |
| CN103955846B (en) | The control method and device of multiple terminal intelligent feedbacks are controlled in information processing system | |
| CN106168956A (en) | data statistical analysis method and system for intelligent terminal | |
| CN109168043B (en) | Method, equipment and system for displaying recommendation information | |
| CN108495072A (en) | A kind of video system and method applied to sports training and teaching | |
| WO2022116545A1 (en) | Interaction method and apparatus based on multi-feature recognition, and computer device | |
| US20200211061A1 (en) | Method and system for targeted advertising in an over-the-top (ott) ecosystem | |
| CN107172482A (en) | The generation method and device of graphic interchange format picture | |
| US20240323457A1 (en) | System and Method for Analyzing Videos in Real-Time | |
| CN101751630A (en) | Method for remotely controlling vehicle mounted LED screen to play text advertisements | |
| CN109218840A (en) | Living broadcast interactive method, storage medium, electronic equipment and system in iOS mobile device | |
| CN110351575A (en) | A kind of processing method and relevant device of virtual pet | |
| CN105979331A (en) | Smart television data recommend method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |