CN109933689A - A kind of video library correlating method and system based on AI algorithm - Google Patents
A kind of video library correlating method and system based on AI algorithm Download PDFInfo
- Publication number
- CN109933689A CN109933689A CN201910180964.1A CN201910180964A CN109933689A CN 109933689 A CN109933689 A CN 109933689A CN 201910180964 A CN201910180964 A CN 201910180964A CN 109933689 A CN109933689 A CN 109933689A
- Authority
- CN
- China
- Prior art keywords
- video
- algorithm
- data resource
- transcoding
- video library
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention discloses a kind of video library correlating methods and system based on AI algorithm.The video library correlating method includes the following steps: that dispatching an at least algorithm service quotient carries out analysis and comparison processing to the data resource in the video library；According to analysis and comparison processing result, the data resource at least one shared visual signature is associated together.The video library correlating method carries out analysis and comparison processing to the data resource in video library based on AI algorithm, to there is the data resource of shared visual signature to be associated automatically, it is convenient that Classification Management, retrieval and comparing etc. are carried out to a large amount of data resource, and be associated display, association push and be associated with the business such as early warning.
Description
Technical field
The present invention relates to video library processing technique more particularly to a kind of video library correlating method based on AI algorithm and it is
System.
Background technique
With the maturation of video monitoring system construction and going deep into for informatization, security industry company and government's machine
The data resource largely taken by video monitoring system is stored in the video library of pass, how so a large amount of data is provided
Source is associated, and carries out an always problem such as Classification Management, retrieval and comparing to facilitate.
Currently, the data resource in video library can only assist being searched by way of semantic search, alternatively, passing through
Manual type is classified by working experience category abundant and keyword etc., is searched for, and work is depended on working efficiency
The working efficiency of the working experience of personnel, more senior staff can be higher, and the efficiency of the staff just to have got started then can
It is much lower.
Summary of the invention
In order to solve above-mentioned the deficiencies in the prior art, the present invention provide a kind of video library correlating method based on AI algorithm and
System carries out analysis and comparison processing to the data resource in video library based on AI algorithm, will have the number of shared visual signature
It is associated automatically according to resource, it is convenient that Classification Management, retrieval and comparing etc. are carried out to a large amount of data resource, and
It is associated display, association push and is associated with the business such as early warning.
The technical problems to be solved by the invention are achieved by the following technical programs:
A kind of video library correlating method based on AI algorithm, includes the following steps:
It dispatches an at least algorithm service quotient and analysis and comparison processing is carried out to the data resource in the video library；
According to analysis and comparison processing result, the data resource at least one shared visual signature is associated together.
Further, the data resource includes video resource；Each algorithm service quotient is being dispatched in the video library
Video resource carry out analysis and comparison processing when, only by coded format be specified standard coding format video resource send
To each algorithm service quotient.
Further, further include following steps:
Further, the original nonstandard video in the video library is subjected to transcoding, is stored after forming transcoding normal video,
Wherein, the original nonstandard video is not specified standard coding format, and the transcoding normal video is specified standard code
Format.
Further, the original nonstandard video in the video library is subjected to transcoding, forms the transcoding normal video
Step includes:
The coded format of original video in the video library is detected, wherein the original video includes primary standard
Video and original nonstandard video, the primary standard video are specified standard coding formats；
If the original video is not specified standard coding format, it is determined as original nonstandard video；
Transcoding is carried out to the original nonstandard video, specified standard coding format is converted into, forms the transcoding standard
Video.
Further, the data resource is parsed and is compared using different AI algorithms between each algorithm service quotient
To processing.
Further, when dispatching each algorithm service quotient progress analysis and comparison processing, scheduling at least two algorithms clothes
The analysis and comparison that business quotient carries out Same Scene to same data resource are handled.
Further, when dispatching each algorithm service quotient progress analysis and comparison processing, according to each algorithm service quotient
Algorithm advantage, dispatch the analysis and comparison that different algorithm service quotient carries out corresponding scene to the data resource and handle.
A kind of video library interconnected system based on AI algorithm, comprising: data resource layer and association service layer, wherein
The data resource layer, for storing data resource, including storing the data handled for the association service layer
The video library of resource；
The association service layer, for carrying out analysis and comparison processing to the data resource in the video library, and according to solution
Analysis and comparison processing result, the data resource at least one shared visual signature is associated together.
Further, the association service layer include: algorithm service module, association service module, dispatch service module and
Program interface module, wherein
The algorithm service module, for carrying out analysis and comparison processing to the data resource in the video library, including at least
One algorithm service quotient；
The association service module will be at least one shared visual signature for foundation analysis and comparison processing result
Data resource is associated together；
The dispatch service module, for carrying out United Dispatching to each algorithm service quotient in the algorithm service module；
Described program interface module, each algorithm service quotient for accessing in the algorithm service module, for the scheduling
Service module carries out United Dispatching to each algorithm service quotient.
Further, the data resource includes video resource；The association service layer further include: transcoding service module,
For the original nonstandard video in the video library to be carried out transcoding, stored after forming transcoding normal video, wherein described
Original nonstandard video is not specified standard coding format, and the transcoding normal video is specified standard coding format.
It is provided the invention has the following beneficial effects: the video library correlating method and system using the algorithm service quotient
AI algorithm carries out analysis and comparison processing to the data resource in the video library, and will there are the data of shared visual signature to provide
Source is associated automatically, and the data resource of different shooting times, different shooting locations and different shooting events etc. is associated in
Together, facilitate according to the association situation between each data resource to a large amount of data resource carry out Classification Management, retrieval and
Comparing etc., and be associated display, association push and be associated with the business such as early warning.
Detailed description of the invention
Fig. 1 is the step schematic diagram of library correlating method provided by the invention；
Fig. 2 is the architecture principle figure of library interconnected system provided by the invention.
Specific embodiment
The present invention will be described in detail with reference to the accompanying drawings and examples.
As shown in Figure 1, a kind of video library correlating method based on AI algorithm, is applied in a video library interconnected system, depending on
Frequency library refers to the video store comprising picture resource and one or more coded format video resources, the same below.Such as Fig. 2 institute
Show, which includes: data resource layer and association service layer, wherein
The data resource layer, for storing data resource, including storing the data handled for the association service layer
The video library of resource, the data resource include picture resource and/or video resource, wherein the video resource includes original
Normal video, original nonstandard video and transcoding normal video, the primary standard video and transcoding normal video are specified
Standard coding format, the original nonstandard video is not specified standard coding format；
The association service layer, for carrying out analysis and comparison processing to the data resource in the video library, and according to solution
Analysis and comparison processing result, the data resource at least one shared visual signature is associated together.
Certainly, image information corresponding with the data resource, described image packet are also stored in the video library
Shooting time, shooting location and the shooting event etc. for including but being not limited to each data resource.
The association service layer includes: algorithm service module, association service module, dispatch service module and routine interface mould
Block, wherein
The algorithm service module, for carrying out analysis and comparison processing to the data resource in the video library, including at least
One algorithm service quotient；
The association service module will be at least one shared visual signature for foundation analysis and comparison processing result
Data resource is associated together；
The dispatch service module, for carrying out United Dispatching to each algorithm service quotient in the algorithm service module；
Described program interface module, each algorithm service quotient for accessing in the algorithm service module, for the scheduling
Service module carries out United Dispatching to each algorithm service quotient.
The video library correlating method includes the following steps:
A dispatch service module schedules at least algorithm service quotient to the data resource in the video library carry out parsing and
Comparison processing；
The association service module will be provided according to analysis and comparison processing result with the data of at least one shared visual signature
Source is associated together.
Wherein, the shared visual signature is not limited to shared face characteristic, shared characteristics of human body or shared motion characteristic
Deng.
The AI algorithm that the video library correlating method and system are provided using the algorithm service quotient is in the video library
Data resource carries out analysis and comparison processing, and will there is the data resource of shared visual signature to be associated automatically, will not
The data resource of same shooting time, different shooting locations and different shooting events etc. is associated together, convenient in business below
Application layer carries out Classification Management, retrieval and data to a large amount of data resource according to the association situation between each data resource
Than equity, and it is associated display, association push and is associated with the business such as early warning.
In addition, the association service layer further include: transcoding service module, for by the original nonstandard view in the video library
Frequency carries out transcoding, is stored after forming the transcoding normal video.
Therefore, which further includes following steps:
Original nonstandard video in the video library is carried out transcoding by the transcoding service module, forms the transcoding normal video
After stored.
Primary standard video and original nonstandard video in the video library are the original of monitoring camera shooting acquisition
Video has one or more coded formats, wherein the original nonstandard video first stores after being shot by the monitoring camera
In the video library, after the transcoding association service layer module row transcoding, it can still be stored in the video library and stay
It backs up, can also be deleted from the video library, depending on specific requirements；The present embodiment is using MP4 format as specified mark
Quasi- coded format.
Wherein, the step of original nonstandard video in the video library being carried out transcoding, forms the transcoding normal video
Include:
The coded format of original video in the video library is detected, wherein the original video includes primary standard
Video and original nonstandard video；
If the original video is not specified standard coding format, it is determined as original nonstandard video；
Transcoding is carried out to the original nonstandard video, specified standard coding format is converted into, forms the transcoding standard
Video.
Wherein, if when judging the coded format of the original video, if the original video is specified standard code
Format is then determined as primary standard video, with no treatment.
When each algorithm service quotient of scheduling carries out analysis and comparison processing to the video resource in the data resource, only
It is video resource (the primary standard video and/or transcoding normal video) hair of specified standard coding format by coded format
Each algorithm service quotient is given, the video resource without by coded format not being specified standard coding format is (described original non-
Mark video) it is sent to each algorithm service quotient.
It is described by the original nonstandard Video Quality Metric in the video library that the video library correlating method and system, which pass through in advance,
Transcoding normal video is stored, and is then again sent to the picture resource, primary standard video and/or transcoding normal video
Each algorithm service quotient individually carries out analysis and comparison processing, and each algorithm service quotient no longer needs to the video to different coding format
Resource carries out video decoding, and only each algorithm service quotient does not save a large amount of decoding time, so that each algorithm service quotient
The research and development of AI algorithm can be absorbed in, the industrial chain division of labor is more clear, and it is more efficient, access threshold is also reduced, and can prop up
It holds some decoded new algorithm service providers of video that are bad to quickly access, so that industry competition is more abundant.
Preferably, described program interface module uses unified api interface standard, in the algorithm service module
Each algorithm service quotient accesses.Each algorithm service quotient accesses the system by unified api interface standard, by the tune
It spends service module and carries out United Dispatching.Due to api interface standard be it is unified, the system compatibility it is bigger, in interface differential technique
Upper need to corresponding with the simple communication exchange of each algorithm service quotient progress and offer resource (such as computing resource and storage money
Source), each algorithm service quotient can first research and develop according to interface document, be connect with the program that exploitation meets interface standard
Mouthful, save the API exploitation and debug time of each algorithm service quotient.
In the present embodiment, described program interface layer include but is not limited to picture parsing, video parsing, vector depositary management reason, to
Buret reason, 1:N are compared, 1:1 is compared or scratch the api interface of at least one function programs such as figure.
Preferably, the algorithm service module includes at least two algorithm service quotient, and is adopted between each algorithm service quotient
Analysis and comparison processing is carried out to the data resource with different AI algorithms.Dispatch each algorithm service quotient carry out parsing and
When comparing processing, the schedulable at least two algorithm service quotient of the dispatch service module are to same data resource (picture resource, original
Beginning normal video and/or transcoding normal video) carry out the analysis and comparison processing of Same Scene, such as recognition of face, the tune
Degree service module can dispatch at least two algorithm service quotient to same data resource (the picture money for needing to carry out recognition of face
Source, primary standard video and/or transcoding normal video) the analysis and comparison processing of recognition of face is carried out, with each family of lateral comparison
Processing capacity of the algorithm service quotient to Same Scene；Alternatively, the dispatch service resource can also be according to each algorithm service quotient's
Algorithm advantage dispatches different algorithm service quotient to the data resource (picture resource, primary standard video and/or transcoding standard
Video) the analysis and comparison processing of corresponding scene, such as recognition of face and action recognition are carried out, the dispatch service module can
With dispatch be good at at least algorithm service quotient of recognition of face to need to carry out recognition of face data resource (picture resource,
Primary standard video and/or transcoding normal video) carry out recognition of face analysis and comparison processing, scheduling be good at action recognition
An at least algorithm service quotient is to data resource (picture resource, primary standard video and/or the transcoding for needing to carry out action recognition
Normal video) the analysis and comparison processing of action recognition is carried out, to comprehensively utilize the algorithm advantage of each algorithm service quotient.
Each algorithm service quotient supports that at least an algorithm service quotient is logical when obtaining the data resource in the video library
It crosses FTP/HTTP to download to the data resource in the video in its local server, then using multithreading to it
Data resource in ground server carries out single cent part, segment processing, is realized with efficiently using the processing capacity of local server
High speed processing.
Embodiments of the present invention above described embodiment only expresses, the description thereof is more specific and detailed, but can not
Therefore limitations on the scope of the patent of the present invention are interpreted as, as long as skill obtained in the form of equivalent substitutions or equivalent transformations
Art scheme should all be fallen within the scope and spirit of the invention.
Claims (10)
1. a kind of video library correlating method based on AI algorithm, which comprises the steps of:
It dispatches an at least algorithm service quotient and analysis and comparison processing is carried out to the data resource in the video library；
According to analysis and comparison processing result, the data resource at least one shared visual signature is associated together.
2. the video library management method according to claim 1 based on AI algorithm, the data resource includes video resource；
When each algorithm service quotient of scheduling carries out analysis and comparison processing to the video resource in the video library, only by coded format
It is that the video resource of specified standard coding format is sent to each algorithm service quotient.
3. the video library management method according to claim 2 based on AI algorithm, further includes following steps:
Original nonstandard video in the video library is subjected to transcoding, is stored after forming transcoding normal video, wherein described
Original nonstandard video is not specified standard coding format, and the transcoding normal video is specified standard coding format.
4. the video library correlating method according to claim 3 based on AI algorithm, which is characterized in that will be in the video library
Original nonstandard video the step of carrying out transcoding, forming the transcoding normal video include:
The coded format of original video in the video library is detected, wherein the original video includes primary standard
Video and original nonstandard video, the primary standard video are specified standard coding formats；
If the original video is not specified standard coding format, it is determined as original nonstandard video；
Transcoding is carried out to the original nonstandard video, specified standard coding format is converted into, forms the transcoding standard
Video.
5. the video library correlating method according to claim 1 based on AI algorithm, which is characterized in that each algorithm service quotient
Between using different AI algorithms to the data resource carry out analysis and comparison processing.
6. the video library correlating method according to claim 5 based on AI algorithm, which is characterized in that dispatching each algorithm
When service provider carries out analysis and comparison processing, scheduling at least two algorithm service quotient carry out Same Scene to same data resource
Analysis and comparison processing.
7. the video library correlating method according to claim 5 or 6 based on AI algorithm, which is characterized in that dispatching each family
When algorithm service quotient carries out analysis and comparison processing, according to the algorithm advantage of each algorithm service quotient, different algorithm clothes are dispatched
The analysis and comparison that business quotient carries out corresponding scene to the data resource are handled.
8. a kind of video library interconnected system based on AI algorithm characterized by comprising data resource layer and association service layer,
Wherein
The data resource layer, for storing data resource, including storing the data handled for the association service layer
The video library of resource；
The association service layer, for carrying out analysis and comparison processing to the data resource in the video library, and according to solution
Analysis and comparison processing result, the data resource at least one shared visual signature is associated together.
9. the video library interconnected system according to claim 8 based on AI algorithm, which is characterized in that the association service layer
It include: algorithm service module, association service module, dispatch service module and program interface module, wherein
The algorithm service module, for carrying out analysis and comparison processing to the data resource in the video library, including at least
One algorithm service quotient；
The association service module will be at least one shared visual signature for foundation analysis and comparison processing result
Data resource is associated together；
The dispatch service module, for carrying out United Dispatching to each algorithm service quotient in the algorithm service module；
Described program interface module, each algorithm service quotient for accessing in the algorithm service module, for the scheduling
Service module carries out United Dispatching to each algorithm service quotient.
10. the video library interconnected system based on AI algorithm according to claim 8 or claim 9, which is characterized in that the data money
Source includes video resource；The association service layer further include: transcoding service module, for will be original nonstandard in the video library
Video carries out transcoding, is stored after forming transcoding normal video, wherein the original nonstandard video is not that specified standard is compiled
Code format, the transcoding normal video is specified standard coding format.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910180964.1A CN109933689A (en) | 2019-03-11 | 2019-03-11 | A kind of video library correlating method and system based on AI algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910180964.1A CN109933689A (en) | 2019-03-11 | 2019-03-11 | A kind of video library correlating method and system based on AI algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109933689A true CN109933689A (en) | 2019-06-25 |
Family
ID=66986676
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910180964.1A Pending CN109933689A (en) | 2019-03-11 | 2019-03-11 | A kind of video library correlating method and system based on AI algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109933689A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014066523A1 (en) * | 2012-10-26 | 2014-05-01 | Google Inc. | Grouping related photographs |
| CN105828030A (en) * | 2016-03-14 | 2016-08-03 | 珠海经济特区远宏科技有限公司 | Video investigation mobile terminal system |
| CN106375721A (en) * | 2016-09-14 | 2017-02-01 | 重庆邮电大学 | Smart video monitoring system based on cloud platform |
| CN107122400A (en) * | 2010-09-24 | 2017-09-01 | 微软技术许可有限责任公司 | The visual cue refining of user's Query Result |
| CN108921918A (en) * | 2018-07-24 | 2018-11-30 | Oppo广东移动通信有限公司 | Video creation method and relevant apparatus |
-
2019
- 2019-03-11 CN CN201910180964.1A patent/CN109933689A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107122400A (en) * | 2010-09-24 | 2017-09-01 | 微软技术许可有限责任公司 | The visual cue refining of user's Query Result |
| WO2014066523A1 (en) * | 2012-10-26 | 2014-05-01 | Google Inc. | Grouping related photographs |
| CN105828030A (en) * | 2016-03-14 | 2016-08-03 | 珠海经济特区远宏科技有限公司 | Video investigation mobile terminal system |
| CN106375721A (en) * | 2016-09-14 | 2017-02-01 | 重庆邮电大学 | Smart video monitoring system based on cloud platform |
| CN108921918A (en) * | 2018-07-24 | 2018-11-30 | Oppo广东移动通信有限公司 | Video creation method and relevant apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105144141B (en) | For using the system and method apart from relevance hashing to media database addressing | |
| KR102531622B1 (en) | Prediction of future viewing of video segments to optimize system resource utilization | |
| Ali et al. | RES: Real-time video stream analytics using edge enhanced clouds | |
| JP6833842B2 (en) | Optimized media fingerprint retention to improve system resource utilization | |
| EP1125245B1 (en) | Image description system and method | |
| US11244232B2 (en) | Feature relationship recommendation method, apparatus, computing device, and storage medium | |
| CN101840435A (en) | Method and mobile terminal for realizing video preview and retrieval | |
| CN104462873A (en) | Picture processing method and picture processing device | |
| US20140207820A1 (en) | Method for parallel mining of temporal relations in large event file | |
| CN106534784A (en) | Acquisition analysis storage statistical system for video analysis data result set | |
| CN106339484A (en) | System and method for intelligent video retrieval processing | |
| CN106331616A (en) | Heterogeneous environment-based massive video structured analysis system and method | |
| CN106713909A (en) | Video encoding method, device and system | |
| CN103020138A (en) | Method and device for video retrieval | |
| CN103187083B (en) | A kind of storage means based on time domain video fusion and system thereof | |
| CN107301245A (en) | A kind of power information video searching system | |
| CN110598042A (en) | Incremental update-based video structured real-time updating method and system | |
| CN109933689A (en) | A kind of video library correlating method and system based on AI algorithm | |
| CN105872731A (en) | Data processing method and device | |
| US20220417540A1 (en) | Encoding Device and Method for Utility-Driven Video Compression | |
| KR101091185B1 (en) | Apparatus and method for analyzing news data | |
| Ying et al. | Exploiting Frame Similarity for Efficient Inference on Edge Devices | |
| CN110941836A (en) | Distributed vertical crawler method and terminal equipment | |
| CN109862319A (en) | A kind of video library processing system and method for supporting more AI algorithms | |
| Lu et al. | TURBO: Opportunistic enhancement for edge video analytics |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190625 |