CN109993049A - A kind of video image structure analysis system towards intelligent security guard field - Google Patents
A kind of video image structure analysis system towards intelligent security guard field Download PDFInfo
- Publication number
- CN109993049A CN109993049A CN201811428069.9A CN201811428069A CN109993049A CN 109993049 A CN109993049 A CN 109993049A CN 201811428069 A CN201811428069 A CN 201811428069A CN 109993049 A CN109993049 A CN 109993049A
- Authority
- CN
- China
- Prior art keywords
- analysis
- pedestrian
- video image
- vehicle
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of video image structure analysis systems towards intelligent security guard field, comprising: video acquisition module, for acquiring video image information；Picture structure analysis module for testing and analyzing to object in the video image information, and carries out classification and character recognition；Database query and search after is written in database writing module, the data for obtaining described image structured analysis module analysis；And structured message retrieval module carries out efficient matching retrieval to the structured message inside database, and search result is returned for the text or image information according to input.The video information that security protection video monitoring collection arrives can be carried out structured analysis, and classification storage by the present invention, efficiently can quickly be matched to required video information content when required by gopher.
Description
Technical field
The present invention relates to the artificial intelligence judgment technology fields of video image structureization analysis, in particular to one kind is towards intelligence
The video image structure analysis system of energy safety-security area.
Background technique
It is constantly improve with living standards of the people with the continuous social and economic development, intelligent security guard for traffic intersection,
The safety monitoring demand in the place such as residential property and market fairground is increasing.The function that existing electronic police is mainly realized
Can be mainly the candid photograph violating the regulations of vehicular traffic, but its application field is limited, not to other information, for example, pedestrian attribute into
Row analysis, is not also detected and is analyzed to vehicle on all roads.Intelligent security guard can recorde lower monitoring camera institute
All multi informations of acquisition image are retrieved after being provided with, and are provided efficiently for fields such as public security, criminal investigations, are easily ensured.
Summary of the invention
The object of the present invention is to provide a kind of video image structure analysis systems towards intelligent security guard field, are used for
Video image information acquired in structured analysis, and it is stored, it is matched convenient for later retrieval.
The technical solution adopted by the present invention to solve the technical problems is:
A kind of video image structure analysis system towards intelligent security guard field, comprising:
Video acquisition module, for acquiring video image information；
Picture structure analysis module for testing and analyzing to object in the video image information, and is divided
Class and character recognition；
Database writing module, the data write-in database for obtaining described image structured analysis module analysis wait for
Later query and search；And
Structured message retrieval module, for the text or image information according to input, to the structuring inside database
Information carries out efficient matching retrieval, and search result is returned.
Further, described image structured analysis module includes:
Pedestrian and vehicle detection unit retrieve the position coordinates in image where pedestrian and vehicle by the image of input
Information；
Pedestrian's attributive analysis unit carries out pedestrian's attributive analysis to the pedestrian detected；And
Vehicle attribute analytical unit carries out vehicle attribute analysis to the vehicle detected.
Further, pedestrian's attributive analysis unit includes: pedestrian's weight identification feature extraction unit, pedestrian's dressing attribute mark
Remember unit and face characteristic extraction unit.
Further, the vehicle attribute analytical unit includes: car plate detection unit, Car license recognition unit and vehicle body identification
Unit.
Further, the concrete analysis content of pedestrian's attributive analysis module includes: the extraction of pedestrian's weight identification feature, jacket
UNEVEN LENGTH OF SLEEVE analysis, long shorts analysis, whether wear a pair of spectacles, whether shaven head, whether band mask and face characteristic extract；
Further, the concrete analysis content of the vehicle attribute analysis module includes: car plate detection identification, the knowledge of vehicle brand
Not, manufacturer and body color identification.
The beneficial effects of the present invention are: the video information that security protection video monitoring collection arrives can be carried out structuring by the present invention
Analysis, and classification storage efficiently can quickly be matched to required video information when required by gopher
Content.
Detailed description of the invention
Fig. 1 is present system structural schematic diagram.
Fig. 2 is picture structure analysis module structural schematic diagram.
Fig. 3 is pedestrian's attributive analysis cellular construction schematic diagram.
Fig. 4 is vehicle attributive analysis cellular construction schematic diagram.
Fig. 5 is work disposal flow chart of the invention.
Specific embodiment
The present invention is mainly based upon the safety-security area video structure analyzing of deep learning, and specific system structure is as schemed
Shown in 1, including video acquisition module, picture structure analysis module, database writing module and structured message retrieval module.
Wherein, video acquisition module, for acquiring video image information；Picture structure analysis module, for collecting
Video image information in object tested and analyzed, and carry out classification and character recognition；Database writing module is used for institute
It states the data that the analysis of picture structure analysis module obtains and database query and search after is written；Structured message retrieves mould
Block carries out efficient matching retrieval to the structured message inside database for the text or image information according to input, and
Search result is returned.
It more specifically says: as shown in Fig. 2, picture structure analysis module includes: pedestrian and vehicle detection unit, pedestrian
Attributive analysis unit and vehicle attribute analytical unit.
Wherein, pedestrian and vehicle detection unit, the object detection technology being based primarily upon in deep learning and example divide skill
Art, on the basis of classical object detection algorithms (such as faster rcnn, rfcn or mask rcnn), by marking a large amount of necks
Domain sample is trained optimization to algorithm model as training set.The module by input image, retrieve image in pedestrian and
Location coordinate information where vehicle；
Pedestrian's attributive analysis unit, as shown in figure 3, the unit is used to carry out pedestrian's attributive analysis to the pedestrian detected,
The unit includes: pedestrian's weight identification feature extraction unit, pedestrian's dressing attribute marking unit and face characteristic extraction unit.Specifically
It is extracted for pedestrian's weight identification feature, identifies that sample carries out network training again by a large amount of pedestrian, extract the last net of gained
Network feature.On the basis of pedestrian's picture, to other attribute (upper and lower clothes attribute, such as UNEVEN LENGTH OF SLEEVE, length library, clothes of pedestrian
Color etc.) it is marked, it forms a large amount of training dataset and is trained for deep learning network.
Vehicle attribute analytical unit, as shown in figure 4, the unit is used to carry out vehicle attribute analysis to the vehicle detected,
It specifically includes that car plate detection unit, Car license recognition unit and vehicle body recognition unit.Deep learning of the car plate detection based on design
Detection of License is trained from the position of a large amount of vehicle pictures acceptance of the bid caravan board as training set.Car license recognition is based on
OCR identification technology carries out letter symbol identification to the license plate of detection.Car license recognition is based on acquisition largely from each province
License plate picture is modeled using two-way LSTM network and CTC Loss.Vehicle body analysis is based primarily upon the vapour of all big enterprises' production
Simultaneously its brand, manufacturer and body color are marked for vehicle signboard, carry out Classification and Identification using deep neural network.
Then work disposal process of the invention carries out people as shown in figure 5, by video acquisition device acquisition video image
And vehicle detection carries out pedestrian's attributive analysis to it if detecting pedestrian, has including detecting the number of people, carries out people to the number of people
Head attributive analysis, detects face, carries out face characteristic extraction to face, and write data into database and save；Pedestrian's analysis
After the completion, vehicle is detected, vehicle attribute analysis then is carried out to each vehicle, and carry out car plate detection and identification, and will test
It is saved in recognition result write-in database.
The advantages of basic principles and main features and this programme of this programme have been shown and described above.The technology of the industry
Personnel are it should be appreciated that this programme is not restricted to the described embodiments, and the above embodiments and description only describe this
The principle of scheme, under the premise of not departing from this programme spirit and scope, this programme be will also have various changes and improvements, these changes
Change and improvement is both fallen within the scope of claimed this programme.This programme be claimed range by appended claims and its
Equivalent thereof.
Claims (6)
1. a kind of video image structure analysis system towards intelligent security guard field characterized by comprising
Video acquisition module, for acquiring video image information；
Picture structure analysis module, for being tested and analyzed to object in the video image information, and carry out classification and
Character recognition；
Database writing module, the data write-in database for obtaining described image structured analysis module analysis is after
Query and search；And
Structured message retrieval module, for the text or image information according to input, to the structured message inside database
Efficient matching retrieval is carried out, and search result is returned.
2. video image structure analysis system as described in claim 1, which is characterized in that described image structured analysis mould
Block includes:
Pedestrian and vehicle detection unit retrieve the location coordinate information in image where pedestrian and vehicle by the image of input；
Pedestrian's attributive analysis unit carries out pedestrian's attributive analysis to the pedestrian detected；And
Vehicle attribute analytical unit carries out vehicle attribute analysis to the vehicle detected.
3. video image structure analysis system as claimed in claim 2, which is characterized in that pedestrian's attributive analysis unit
It include: pedestrian's weight identification feature extraction unit, pedestrian's dressing attribute marking unit and face characteristic extraction unit.
4. video image structure analysis system as claimed in claim 2, which is characterized in that the vehicle attribute analytical unit
It include: car plate detection unit, Car license recognition unit and vehicle body recognition unit.
5. video image structure analysis system as described in claim 1, which is characterized in that pedestrian's attributive analysis module
Concrete analysis content include: pedestrian weight identification feature extract, jacket UNEVEN LENGTH OF SLEEVE analysis, long shorts analysis, whether wear a pair of spectacles, be
Whether no shaven head extracts with mask and face characteristic.
6. video image structure analysis system as described in claim 1, which is characterized in that the vehicle attribute analysis module
Concrete analysis content include: car plate detection identification, vehicle brand recognition, manufacturer and body color identification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811428069.9A CN109993049A (en) | 2018-11-27 | 2018-11-27 | A kind of video image structure analysis system towards intelligent security guard field |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811428069.9A CN109993049A (en) | 2018-11-27 | 2018-11-27 | A kind of video image structure analysis system towards intelligent security guard field |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109993049A true CN109993049A (en) | 2019-07-09 |
Family
ID=67128379
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811428069.9A Pending CN109993049A (en) | 2018-11-27 | 2018-11-27 | A kind of video image structure analysis system towards intelligent security guard field |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109993049A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111209807A (en) * | 2019-12-25 | 2020-05-29 | 航天信息股份有限公司 | Yolov 3-based video structuring method and system |
| CN111860523A (en) * | 2020-07-28 | 2020-10-30 | 上海兑观信息科技技术有限公司 | Intelligent recording system and method for sound image file |
| CN112565717A (en) * | 2021-02-18 | 2021-03-26 | 深圳市安软科技股份有限公司 | Video structuring method, related device, system and storage medium |
| CN113255477A (en) * | 2021-05-08 | 2021-08-13 | 深圳市安软科技股份有限公司 | Comprehensive management system and method for pedestrian video images |
| CN114494978A (en) * | 2022-02-21 | 2022-05-13 | 浪潮云信息技术股份公司 | Pipeline-based parallel video structured reasoning method and system |
| CN114860993A (en) * | 2021-02-04 | 2022-08-05 | 宏碁股份有限公司 | Method for screening objects and computer program product |
-
2018
- 2018-11-27 CN CN201811428069.9A patent/CN109993049A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111209807A (en) * | 2019-12-25 | 2020-05-29 | 航天信息股份有限公司 | Yolov 3-based video structuring method and system |
| CN111860523A (en) * | 2020-07-28 | 2020-10-30 | 上海兑观信息科技技术有限公司 | Intelligent recording system and method for sound image file |
| CN111860523B (en) * | 2020-07-28 | 2024-04-30 | 上海兑观信息科技技术有限公司 | Intelligent recording system and method for sound image files |
| CN114860993A (en) * | 2021-02-04 | 2022-08-05 | 宏碁股份有限公司 | Method for screening objects and computer program product |
| CN112565717A (en) * | 2021-02-18 | 2021-03-26 | 深圳市安软科技股份有限公司 | Video structuring method, related device, system and storage medium |
| CN112565717B (en) * | 2021-02-18 | 2021-05-25 | 深圳市安软科技股份有限公司 | Video structuring method, related device, system and storage medium |
| CN113255477A (en) * | 2021-05-08 | 2021-08-13 | 深圳市安软科技股份有限公司 | Comprehensive management system and method for pedestrian video images |
| CN114494978A (en) * | 2022-02-21 | 2022-05-13 | 浪潮云信息技术股份公司 | Pipeline-based parallel video structured reasoning method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109993049A (en) | A kind of video image structure analysis system towards intelligent security guard field | |
| CN104200669B (en) | Fake-licensed car recognition method and system based on Hadoop | |
| CN109740420B (en) | Vehicle law violation identification method and related product | |
| CN107730905A (en) | Multitask fake license plate vehicle vision detection system and method based on depth convolutional neural networks | |
| CN101261722A (en) | Electronic police background intelligent management and automatic implementation system | |
| CN101587643B (en) | Identification method of fake-licensed cars | |
| CN103714326B (en) | One-sample face identification method | |
| CN103810505A (en) | Vehicle identification method and system based on multilayer descriptors | |
| CN111078946A (en) | Bayonet vehicle retrieval method and system based on multi-target regional characteristic aggregation | |
| CN103279738B (en) | Automatic identification method and system for vehicle logo | |
| CN111814751A (en) | Vehicle attribute analysis method and system based on deep learning target detection and image recognition | |
| Patil et al. | Vehicle number plate detection using yolov8 and easyocr | |
| CN113673527A (en) | License plate recognition method and system | |
| CN109086778A (en) | A kind of license plate deblurring recognition methods based on convolutional neural networks | |
| CN110222660A (en) | A signature authentication method and system based on the fusion of dynamic and static features | |
| Ghadage et al. | A review paper on automatic number plate recognition system using machine learning algorithms | |
| CN115272222A (en) | Method, device and equipment for processing road detection information and storage medium | |
| CN108427718A (en) | A kind of vehicle traveling information temporal index and big data analysis method | |
| CN112115800A (en) | Vehicle combination recognition system and method based on deep learning target detection | |
| CN105844286A (en) | Newly added vehicle logo identification method and apparatus | |
| CN119048990A (en) | Control system for urban data management | |
| CN111881792A (en) | Mobile micro-bayonet system and working method thereof | |
| CN110807415B (en) | Traffic checkpoint vehicle intelligent retrieval system and method based on annual inspection marks | |
| CN113283303A (en) | License plate recognition method and device | |
| Kodwani et al. | Automatic license plate recognition in real time videos using visual surveillance techniques |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190709 |