CN105554595A - Video abstract intelligent extraction and analysis system - Google Patents
Video abstract intelligent extraction and analysis system Download PDFInfo
- Publication number
- CN105554595A CN105554595A CN201410586437.8A CN201410586437A CN105554595A CN 105554595 A CN105554595 A CN 105554595A CN 201410586437 A CN201410586437 A CN 201410586437A CN 105554595 A CN105554595 A CN 105554595A
- Authority
- CN
- China
- Prior art keywords
- video
- search
- video frequency
- moving target
- target
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a video abstract intelligent extraction and analysis system. The system includes steps: A. separating a moving object and a background in an original video; B. reorganizing the extracted moving object and superposing the moving object on the background to form a video abstract; and C. establishing a search index of the moving object according to characteristic parameters of the object, and directly searching the corresponding moving object according to the characteristic parameter.
Description
[technical field]
The present invention relates to video frequency abstract intelligent extraction and analytical system.
[technical background]
Video, as the non-structured data of one, brings very large challenge to the fast browsing of people and retrieval.Especially have the video data of magnanimity in internet video industry and dedicated video monitoring trade, it will be very time-consuming thing that people will browse and search specific event in video.Technology in the urgent need to the A and I structure before similar books is applied in the structuring process of video.Therefore video frequency abstract formation and search technique occur for solving this problem.
Current video summarization technique mainly contains two kinds of patterns: video is slightly look at and video outline.Video is slightly look at and mainly on the basis of shot segmentation, is extracted key frame, is organized into video frequency abstract with the key frame of camera lens.Video outline mainly realizes concentrating of video and ensures continuity and the consistency of video plot.
[summary of the invention]
The object of the present invention is to provide method, system that a kind of video frequency abstract is formed and searches for, after Video Analysis Technology separating background and moving target, reorganize time and order that multiple moving target shows in background video simultaneously, set up video frequency abstract fragment and moving-target search.
The video frequency abstract that this patent proposes is formed and the technology of search has significantly different from existing video frequency abstract pattern:
1) this patent is the foreground target obtained after analyzing video requency frame data by the algorithm process of video analysis in video.2) reorganize time of showing in video of target and sequencing, the time requirement of fragment of can making a summary according to generating video, allow multiple target not block and on overlapping basis, synchronization is simultaneously displayed in video pictures.3) search is set up according to the characteristic parameter of moving target.
In addition, this patent does not relate to the semantic analysis of video content, video frequency abstract has reorganized the time occurred in video and the order of moving target in video, lacks the high-rise technical Analysis to understanding videos such as the semantic analysis between the moving target in video and plot analysis.
[accompanying drawing explanation]
Accompanying drawing illustrates: below in conjunction with drawings and Examples, the present invention is further described.
Fig. 1 is system flow schematic diagram of the present invention
[embodiment]
In order to make those skilled in the art person understand the present invention program better, below in conjunction with drawings and embodiments, the present invention is described in further detail.
Embodiment 1 accompanying drawing is a kind of implementation method flow chart original video files being carried out to video frequency abstract formation and search.As a kind of embodiment of the present invention, original video source can be conventional video multimedia file, also can be the video file of the collection of special video monitoring system and the video file of network monitoring system collection.
Step S101: separating background and moving target from original video.
Moving target can be separated by the method for separating background and moving target from background that is complicated and that change, to upgrade and safeguard simultaneously the background model of dynamic self-adapting, this model algorithm can adapt to light gradient at sunshine, small random motion object interference and boisterous interference.
Step S102: reorganize moving target and the background that is added to formation video frequency abstract.
On the basis of separating background and moving target, moving target is reorganized the time occurred in video and order, and ensure not block between moving target and overlapping.Relativeness between moving target can not cause confusion.This step comprises moving target time of occurrence and order computation method; Moving target is avoided to block and moving target system of selection; Moving target and video background synthetic video method of abstracting.
Step S103: search is set up to foreground target according to clarification of objective parameter.The classification of target can be carried out: 1, the feature of people: face recognition features, the feature of kinematic parameter feature 2, car: Car license recognition feature, color, volume, the feature of travel speed 3, thing: color, volume and the speed of travel according to the moving target extracted.The index database of moving target is set up by characteristic parameter.
Also comprise the characteristic parameter module extracting the target image that will search in step S103 further, according to the index database of the characteristic parameter searching moving target obtained, judge whether the moving target of coupling.
The system that a kind of video frequency abstract that the present invention also provides simultaneously is formed and searches for, this system can obtain the laggard row relax analysis of original video from Third party system, after separating background and moving target, the background that is added to after reorganizing time of moving target and order forms video frequency abstract.
Claims (10)
1. video frequency abstract intelligent extraction and an analytical system, is characterized in that, the method comprising the steps of:
A. separating background and moving target from original video;
B. reorganize moving target occur time and order and the background that is added to forms video frequency abstract;
C. search is set up according to the characteristic parameter of moving target.
2. the method for video frequency abstract formation as claimed in claim 1 and search, it is characterized in that, step a) in original video packets draws together conventional multimedia video frequency file, special supervisory control system gathers video file and/or the video file that network video monitor and control system gathers.
3. video frequency abstract as claimed in claim 2 is formed and the method for search, it is characterized in that, the multimedia video frequency file of described routine comprises TV programme, film, network multimedia file with sound and video.
4. the method for video frequency abstract formation as claimed in claim 1 and search, is characterized in that, from original video, separating background and moving target need set up the background model of renewal automatically and extract the characteristic parameter of moving target.
5. the method for video frequency abstract formation as claimed in claim 1 and search, it is characterized in that, according to the operational objective detected, multiple moving target is organized according to the time of the position in visual field and appearance again, and forms video frequency abstract in the background that is added to.
6. video frequency abstract as claimed in claim 1 is formed and the method for search, it is characterized in that, after separating background and moving target, the characteristic parameter according to moving target sets up target search.
7. the method for video frequency abstract formation as claimed in claim 4 and search, is characterized in that, set up the context update model of dynamic self-adapting, need model algorithm can adapt to light gradient at sunshine, small random motion object interference and boisterous interference.
8. the method for video frequency abstract formation as claimed in claim 5 and search, it is characterized in that, after separate targets and background, the target successively do not occurred in the same time reorganizes in same background, allows the target do not occurred in the same time show with synchronization in the same context.
9. the method for video frequency abstract formation as claimed in claim 6 and search, it is characterized in that, the method that the described characteristic parameter according to moving target sets up target search is:
A) classification of target is set up: 1, the feature of people: recognition of face parameter, the feature of motion feature 2, car: Car license recognition parameter, color, volume, the feature of travel speed 3, thing: color, volume and gait of march;
B) index database of moving target is set up;
C) the target image characteristics parameter extraction will searched for, according to the index database of the characteristic parameter searching moving target obtained, judges whether the target of coupling.
10. the method for video frequency abstract formation as claimed in claim 9 and search, it is characterized in that, step c) comprise the clarification of objective parameter module that will inquire about extracting user's input further, what user inputted can be video or image file, or directly inputs clarification of objective parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410586437.8A CN105554595A (en) | 2014-10-28 | 2014-10-28 | Video abstract intelligent extraction and analysis system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410586437.8A CN105554595A (en) | 2014-10-28 | 2014-10-28 | Video abstract intelligent extraction and analysis system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105554595A true CN105554595A (en) | 2016-05-04 |
Family
ID=55833495
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410586437.8A Pending CN105554595A (en) | 2014-10-28 | 2014-10-28 | Video abstract intelligent extraction and analysis system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105554595A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529406A (en) * | 2016-09-30 | 2017-03-22 | 广州华多网络科技有限公司 | Method and device for acquiring video abstract image |
| CN108600864A (en) * | 2018-04-25 | 2018-09-28 | 中影数字巨幕（北京）有限公司 | A kind of preview generation method and device |
| CN108900792A (en) * | 2018-07-26 | 2018-11-27 | 广州大学 | A kind of ubiquitous video evidence collecting method and system towards car networking |
| CN112335256A (en) * | 2018-04-10 | 2021-02-05 | 脸谱公司 | Automatic decision making based on descriptive model |
-
2014
- 2014-10-28 CN CN201410586437.8A patent/CN105554595A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529406A (en) * | 2016-09-30 | 2017-03-22 | 广州华多网络科技有限公司 | Method and device for acquiring video abstract image |
| CN106529406B (en) * | 2016-09-30 | 2020-02-07 | 广州华多网络科技有限公司 | Method and device for acquiring video abstract image |
| CN112335256A (en) * | 2018-04-10 | 2021-02-05 | 脸谱公司 | Automatic decision making based on descriptive model |
| CN108600864A (en) * | 2018-04-25 | 2018-09-28 | 中影数字巨幕（北京）有限公司 | A kind of preview generation method and device |
| CN108600864B (en) * | 2018-04-25 | 2020-08-28 | 中影数字巨幕（北京）有限公司 | Movie preview generation method and device |
| CN108900792A (en) * | 2018-07-26 | 2018-11-27 | 广州大学 | A kind of ubiquitous video evidence collecting method and system towards car networking |
| CN108900792B (en) * | 2018-07-26 | 2020-07-31 | 广州大学 | Internet of vehicles oriented ubiquitous video evidence obtaining method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102156707A (en) | Video abstract forming and searching method and system | |
| Wiley et al. | Computer vision and image processing: a paper review | |
| CN101021903B (en) | Video caption content analysis system | |
| CN103347167A (en) | Surveillance video content description method based on fragments | |
| CN107087211B (en) | Method and device for detecting lens of host | |
| CN110442747B (en) | Video abstract generation method based on keywords | |
| US20150254231A1 (en) | Method and apparatus for correlating and viewing disparate data | |
| US20190278796A1 (en) | Method and system for searching, publishing and managing the lite cycle of multimedia contents related to public events and the user experience | |
| CN110097026B (en) | Paragraph association rule evaluation method based on multi-dimensional element video segmentation | |
| GB2493580A (en) | Method of searching for a target within video data | |
| CN105554595A (en) | Video abstract intelligent extraction and analysis system | |
| CN101369281A (en) | Retrieval method based on video abstract metadata | |
| CN105138525B (en) | Traffic video processing unit and method and retrieval device and method | |
| CN1851709A (en) | Embedded multimedia content-based inquiry and search realizing method | |
| EP2735984A1 (en) | Video query method, device and system | |
| CN102231820A (en) | Monitoring image processing method, device and system | |
| CN103077236A (en) | System and method for realizing video knowledge acquisition and marking function of portable-type device | |
| CN105554456B (en) | Method for processing video frequency and equipment | |
| CN112989950A (en) | Violent video recognition system oriented to multi-mode feature semantic correlation features | |
| Sumari et al. | Towards practical implementations of person re-identification from full video frames | |
| CN102244769A (en) | Object and key person monitoring system and method thereof | |
| CN115497017A (en) | Broadcast television news stripping method and device based on artificial intelligence | |
| CN111160099B (en) | Intelligent segmentation method for video image target | |
| CN104484457A (en) | Method and system for extracting and searching moving object in parallel video | |
| CN112306985A (en) | Digital retina multi-modal feature combined accurate retrieval method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| DD01 | Delivery of document by public notice |
Addressee: SHANGHAI ZUYUAN SCIENCE AND TECHNOLOGY DEVELOPMENT CO., LTD. Document name: Notification of Publication of the Application for Invention |
|
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160504 |
|
| WD01 | Invention patent application deemed withdrawn after publication |