CN108399247A - A kind of generation method of virtual identity mark - Google Patents
A kind of generation method of virtual identity mark Download PDFInfo
- Publication number
- CN108399247A CN108399247A CN201810172916.3A CN201810172916A CN108399247A CN 108399247 A CN108399247 A CN 108399247A CN 201810172916 A CN201810172916 A CN 201810172916A CN 108399247 A CN108399247 A CN 108399247A
- Authority
- CN
- China
- Prior art keywords
- vid
- characteristic value
- image
- value data
- virtual identity
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2457—Query processing with adaptation to user needs
- G06F16/24573—Query processing with adaptation to user needs using data annotations, e.g. user-defined metadata
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Processing Or Creating Images (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of generation methods of virtual identity mark, portrait is stamped virtual identity identity property in advance, pass through the comparison score of characteristic value data, same person recognition in a large amount of photos and identify, when later retrieval, it need not be compared again by face characteristic, but high speed can be completed by everyone time-space relationship structural data and retrieve, realization personnel's trajectory analysis, personnel's relationship compares, frequent analysis of haunting, frequency analysis, the collisions such as adjoint analysis compare, to realize the human behavior feature portrait based on video identity archives in the future.Simultaneously under the face characteristic library of same scale, similarly concurrently compares and require, required number of servers is reduced at double, cannot meet the comparison demand of 10,000,000,000 grades of face databases to solve prior art alignments in performance.
Description
Technical field
The present invention relates to recognition of face, the technical fields such as image retrieval, more particularly to extensive fast face identification are answered
Use scene.
Background technology
With being widely used of video monitoring equipment, digital image capture equipment it is universal and network social intercourse website fast
Speed development, the data of facial image increase the scale of magnanimity rapidly, and carrying out effective image retrieval according to picture material has
Important meaning.The recognition of face of the prior art be with video camera or camera acquisition image or video flowing containing face, and
Automatic detect and track face in the picture, and then the face to detecting carries out feature extraction, attributive analysis, identifies comparison
A series of the relevant technologies.Based on face recognition technology search someone event trace when, typically by the face of target person with
Face in database compares one by one, since the face in database is random distribution, when carrying out face comparison, needs
It is compared with face whole in database, it is inefficient.And with safe city, the construction of bright as snow engineering, video monitoring
Construction scale has reached the hundreds thousand of roads in a city, and the human face data acquired daily reaches hundred million grades, and traditional alignments are in property
Can on cannot meet the comparison demands of 10,000,000,000 grades of face databases.
Invention content
It is existing in the prior art to solve it is an object of the invention to provide a kind of generation method of virtual identity mark
Problems.Described method includes following steps:
Step 1 prepares a list for being used for recording VID being initially empty(Hereinafter referred to as VID lists), each includes one
VID and corresponding N characteristic value data.
Step 2 obtains an image data comprising portrait and corresponding adeditive attribute.
Step 3 extracts the characteristic value data of everyone picture in image successively.
Step 4 is successively all characteristic value numbers in the characteristic value data of everyone picture in the image and VID lists
According to being compared, comparison result highest scoring and big Mr. Yu's threshold value are searched.Characteristic value data enters step if not finding
5, otherwise enter step 6.
Step 5 generates a new VID, is not repeated mutually with the VID in the existing record in VID lists, into VID lists
It is inserted into a new record, VID data are the new VID data just generated, and characteristic value data is current portrait in the image
Characteristic value data.Enter step 7.
Step 6 finds the corresponding corresponding VID of this feature Value Data and corresponding N characteristic value data, is set according to system
Set determine whether with current characteristic value replace N characteristic value data in one or with this feature value increase newly a characteristic value number
According to making N become N+1, or do nothing.
Step 7 exports or stores the VID data of the portrait, this feature Value Data, the image data and corresponding attached
Additive attribute is used for subsequent retrieval with analysis, returns to step 2.
The present invention has the advantages that compared with prior art:
Just there is provided a kind of generation methods of virtual identity mark by the present invention, and portrait, which is stamped virtual identity mark, in advance belongs to
Property, when later retrieval:One)It is identified by virtual identity, under the face characteristic library of same scale, similarly concurrently compares and want
It asks, required number of servers is reduced at double;Two)It is identified by virtual identity, is realizing personnel's trajectory analysis, personnel's relationship
When comparison, the collisions such as analysis, frequency analysis, adjoint analysis of frequently haunting compare, it need not be compared again by face characteristic,
But high speed can be completed by everyone time-space relationship structural data and retrieve;Three)It is identified by virtual identity so that will
To realize that the human behavior feature portrait based on video identity archives is possibly realized.
Description of the drawings:
Fig. 1 is flow chart of the method for the present invention.
Fig. 2 is 1 flow chart of steps of present example.
Fig. 3 is the schematic diagram of present example 1.
Specific implementation mode
In order to make the purpose of the present invention, technical solution and advantageous effect be more clearly understood, below in conjunction with example, to this hair
It is bright to be further elaborated.It should be understood as that specific example described herein only to explain the present invention, is not used to
It limits the scope of the invention.
As shown in Fig. 1, the present invention provides a kind of generation methods of virtual identity mark, and the method includes walking as follows
Suddenly:
Step 101 prepares a list for being used for recording VID being initially empty(Hereinafter referred to as VID lists), each includes one
A VID and corresponding N characteristic value data.
Step 102 obtains an image data comprising portrait and corresponding adeditive attribute.
Step 103 extracts the characteristic value data of everyone picture in image successively.
Step S104 is successively all characteristic values in the characteristic value data of everyone picture in the image and VID lists
Data are compared, and search comparison result highest scoring and big Mr. Yu's threshold value.Characteristic value data enters step if not finding
Rapid S105, otherwise enters step S106.
Step S105 generates a new VID, is not repeated mutually with the VID in the existing record in VID lists, is arranged toward VID
A new record is inserted into table, VID data are the new VID data just generated, and characteristic value data is current in the image
The characteristic value data of portrait.Enter step S107.
Step S106 finds the corresponding corresponding VID of this feature Value Data and corresponding N characteristic value data, according to system
Be arranged determine whether with current characteristic value replace N characteristic value data in one or with this feature value increase newly a characteristic value
Data make N become N+1, or do nothing.
Step S107 output stores the VID data of the portrait, this feature Value Data, the image data and corresponding
Adeditive attribute is used for subsequent retrieval with analysis, returns to step S102.
The above is only the preferred embodiment of the present invention.It should be pointed out that coming for those skilled in the art
It says, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications also should be regarded as
Protection scope of the present invention.
Specific example:
An a kind of concrete application example of the generation method of virtual identity mark of the present invention introduced below.
Example 1:
As illustrated in fig. 2, it is assumed that finding the scene for being abducted children for a railway station, according to step S101, we have been set up
VID lists.Face image data and corresponding adeditive attribute are obtained according to the photo of abducted children according to step S102.According to
Step 103 extracts the characteristic value data that children's portrait is abducted in image successively.According to step S104 successively the youngster in the image
The characteristic value data of virgin portrait is compared with all characteristic value datas in existing VID lists, searches comparison result score most
High and big Mr. Yu's threshold value.This example assumes that abducted children are really aerial in VID table datas, then enters step 7, export the children
VID data, this feature Value Data, the image data and corresponding adeditive attribute, using output children VID, comprising
Identity marks when null record in quick-searching go out each space-time position of children, provide useful clue to solve a case.
Claims (9)
1. a kind of generation method of virtual identity mark, which is characterized in that include the following steps:Step 1 prepares one and is initially
The empty list for being used for recording VID(Hereinafter referred to as VID lists), each includes a VID and corresponding N characteristic value
Data;Step 2 obtains an image data comprising portrait and corresponding adeditive attribute;Step 3 is extracted every in image successively
The characteristic value data of one portrait;Step 4 is successively in the characteristic value data of everyone picture in the image and VID lists
All characteristic value datas are compared, and search comparison result highest scoring and big Mr. Yu's threshold value, characteristic value data, if do not looked for
To then entering step 5,6 are otherwise entered step;Step 5 generates a new VID, with the VID in the existing record in VID lists
It does not repeat mutually, a new record is inserted into VID lists, VID data are the new VID data just generated, characteristic value number
According to the characteristic value data for current portrait in the image, 7 are entered step;Step 6 finds the corresponding corresponding VID of this feature Value Data
And corresponding N characteristic value data, determine whether to be replaced in N characteristic value data with current characteristic value according to system setting
One either increases a characteristic value data newly with this feature value and N is made to become N+1 or do nothing;Step 7 export or
VID data, this feature Value Data, the image data and the corresponding adeditive attribute of the portrait are stored for subsequent retrieval and is divided
Analysis uses, and returns to step 2.
2. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that described in step 2
The method for obtaining image can be by capturing camera.
3. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that described in step 2
The method for obtaining image can pass through video interception.
4. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that obtaining described in step 2
The method for taking image can pass through video interception.
5. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that obtaining described in step 2
The method for taking image can be downloaded by picture.
6. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that the people described in step 2
As that can be facial image.
7. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that the people described in step 2
As that can be human body image.
8. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that the figure described in step 2
The adeditive attribute of picture includes but not limited to obtain time and the location information of the image.
9. a kind of generation method of virtual identity mark according to claim 1, which is characterized in that carrying described in step 3
The method for taking the characteristic value data of everyone picture in image can use face, human bioequivalence algorithm based on deep learning.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810172916.3A CN108399247A (en) | 2018-03-01 | 2018-03-01 | A kind of generation method of virtual identity mark |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810172916.3A CN108399247A (en) | 2018-03-01 | 2018-03-01 | A kind of generation method of virtual identity mark |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108399247A true CN108399247A (en) | 2018-08-14 |
Family
ID=63091413
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810172916.3A Pending CN108399247A (en) | 2018-03-01 | 2018-03-01 | A kind of generation method of virtual identity mark |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108399247A (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105138954A (en) * | 2015-07-12 | 2015-12-09 | 上海微桥电子科技有限公司 | Image automatic screening, query and identification system |
CN105488478A (en) * | 2015-12-02 | 2016-04-13 | 深圳市商汤科技有限公司 | Face recognition system and method |
CN106778489A (en) * | 2016-11-14 | 2017-05-31 | 深圳奥比中光科技有限公司 | The method for building up and equipment of face 3D characteristic identity information banks |
-
2018
- 2018-03-01 CN CN201810172916.3A patent/CN108399247A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105138954A (en) * | 2015-07-12 | 2015-12-09 | 上海微桥电子科技有限公司 | Image automatic screening, query and identification system |
CN105488478A (en) * | 2015-12-02 | 2016-04-13 | 深圳市商汤科技有限公司 | Face recognition system and method |
CN106778489A (en) * | 2016-11-14 | 2017-05-31 | 深圳奥比中光科技有限公司 | The method for building up and equipment of face 3D characteristic identity information banks |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107330396B (en) | Pedestrian re-identification method based on multi-attribute and multi-strategy fusion learning | |
CN109344787B (en) | Specific target tracking method based on face recognition and pedestrian re-recognition | |
CN108536851B (en) | User identity recognition method based on moving track similarity comparison | |
US8130285B2 (en) | Automated searching for probable matches in a video surveillance system | |
CN106548164A (en) | The relevance recognition methods of facial image and mobile device | |
CN106998444A (en) | A kind of big data face monitoring system and device | |
CN106845373A (en) | Towards pedestrian's attribute forecast method of monitor video | |
CN108064388A (en) | Personage's method for searching, device, terminal and cloud server | |
CN109544754B (en) | Access control method and system based on face recognition and computer equipment | |
CN103984738A (en) | Role labelling method based on search matching | |
CN107977656A (en) | A kind of pedestrian recognition methods and system again | |
CN102945366A (en) | Method and device for face recognition | |
CN103413295B (en) | A kind of video multi-target long-range tracking | |
CN103399896A (en) | Method and system for recognizing association relationships among users | |
CN109213882A (en) | Picture sort method and terminal | |
CN106663196A (en) | Computerized prominent person recognition in videos | |
CN112232148A (en) | Image clustering method, target track tracking method, electronic device and storage medium | |
CN107103615A (en) | A kind of monitor video target lock-on tracing system and track lock method | |
CN108563651B (en) | Multi-video target searching method, device and equipment | |
CN107230267A (en) | Intelligence In Baogang Kindergarten based on face recognition algorithms is registered method | |
CN105159959A (en) | Image file processing method and system | |
Kaneko et al. | Visual event mining from geo-tweet photos | |
CN109544718A (en) | A kind of Work attendance method and its system | |
Li et al. | Social context-aware person search in videos via multi-modal cues | |
CN114419092A (en) | Personnel movement track tracking and backtracking method and system based on domestic CPU environment |
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 | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20210520 Address after: Room 25b02, Microsoft tech building, 55 Gaoxin South 9th Road, Nanshan District, Shenzhen, Guangdong 518000 Applicant after: Shenzhen antelope video cloud Technology Co.,Ltd. Address before: 1301, Yihua financial building, 2388 Houhai Avenue, Nanshan District, Shenzhen, Guangdong 518000 Applicant before: SHENZHEN ANTELOPE ULTIMATE TECHNOLOGY Co.,Ltd. |
|
TA01 | Transfer of patent application right | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180814 |
|
WD01 | Invention patent application deemed withdrawn after publication |