CN108874979A - A kind of suspect vehicle discovery algorithm based on space-time unique constraint - Google Patents
A kind of suspect vehicle discovery algorithm based on space-time unique constraint Download PDFInfo
- Publication number
- CN108874979A CN108874979A CN201810589430.XA CN201810589430A CN108874979A CN 108874979 A CN108874979 A CN 108874979A CN 201810589430 A CN201810589430 A CN 201810589430A CN 108874979 A CN108874979 A CN 108874979A
- Authority
- CN
- China
- Prior art keywords
- stability condition
- space
- cfl stability
- time
- cfl
- 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 claims abstract description 32
- 235000013399 edible fruits Nutrition 0.000 claims description 4
- 230000003014 reinforcing effect Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 5
- 230000000007 visual effect Effects 0.000 abstract description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000012544 monitoring process Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 241000396836 Trinodes Species 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A kind of suspect vehicle based on space-time unique constraint of the present invention finds algorithm, is related to the application fields such as intelligent transportation, space-time database and artificial intelligence, especially a kind of suspect vehicle discovery method for not depending on visual signature.Include the following steps：1）It submitskA CFL stability condition；kIt is natural number,kValue range be 1≤k≤10；kA CFL stability condition iskA relatively unambiguous room and time range；2）It successively searches and meets in track databasekThe vehicle of a CFL stability condition, and obtain final query result.The space-time characteristic moved using suspect vehicle, the time range of suspect vehicle movement and spatial dimension are expressed as CFL stability condition, the inquiry that suspect vehicle is carried out based on space-time restriction can support multi-space constraint condition and distributed processing mode to improve the efficiency of suspect vehicle discovery.
Description
Technical field
A kind of suspect vehicle based on space-time unique constraint of the present invention finds algorithm, is related to intelligent transportation, space-time database
Method is found with application fields, especially a kind of suspect vehicles for not depending on visual signature such as artificial intelligence.
Background technique
Suspect vehicle identification is the major issue in intelligent transportation, it is intended to help public security organ's tracking violation, illegal vehicle.
Suspect vehicle in real life generally includes two classes：First is that with Typical external feature vehicle, if any witness or certain
The monitoring system at place has recorded the Typical external feature of suspect vehicle, such as license plate number, brand and model and color；Second is that external special
Levy unknown vehicle.Current suspect vehicle discovery algorithm mainly has the suspect vehicle of Typical external feature, generallys use base
In the discovery algorithm of image procossing, main includes the license plate number for determining suspect vehicle and the traffic route two for determining suspect vehicle
Class.Determine that the license plate number of suspect vehicle, i.e., the features such as vehicle brand, color and model provided for eye witness pass through prison
All information of vehicles for meeting feature in traffic injury time section near the system queries scene of the accident are controlled, investigation is straight one by one
To the license plate number for determining suspect vehicle, so that it is determined that the identity information of suspect.Determining the traffic route of suspect vehicle is
It shows and determines license plate number and time interval, the motion profile of given vehicle is inquired by monitoring system, i.e., vehicle is in given time area
It is interior by which section and wherein stop how long etc. information.The identification of suspect vehicle license plate number and the inquiry of traffic route
It can combine, the efficiency and precision of license plate number identification can be improved according to traffic route.
For license plate number identification mainly by the method for image procossing, each vehicle is extracted in the video of monitoring system shooting
Feature, with live witness provide vehicle characteristics match, find accident spot attachment, accident generation before and after
Time interval in all vehicles for meeting feature.Current traffic surveillance and control system has been able to the passing vehicle of identification rapidly and efficiently
License plate number, thus may determine that the license plate number of all vehicles for meeting feature.Then according to license plate number, in conjunction with each car
Traffic route further reduces the range of suspect vehicle.
For the suspect vehicle of no Typical external feature, generally use the mode of inquiry, according to the police speculate time and
The vehicle for the condition that meets is searched in traffic system in position.Existing suspect vehicle querying method uses the data of relational database
Storage and querying method, including two classes：It is when being retrieved in relational database first is that determining the time and determining the inquiry of position
Between and position all vehicles equal with specified criteria；Second is that the inquiry of time range and determining position, i.e. inquiry given time
Pass through all vehicles of specific position in range.Due to the character position that location information generallys use in such methods, only
It supports the inquiry of determining position, and is difficult the inquiry suitable for spatial dimension.However, the inquiry of suspect vehicle in practical applications
It is normally based on the inquiry of time range and spatial dimension, i.e. the vehicle within a certain period of time Jing Guo certain panel region is as suspicious
Vehicle.Moreover, in order to increase inquiry precision, it usually needs the pact of multiple groups time range and spatial dimension is arranged according to case clue
Beam.
The discovery of suspect vehicle facilitates the police and chases suspect, and suspect vehicle chases work in order to interfere the police
Make, the traffic route that various modes hide itself is generallyd use, so that the police are difficult to be accurately positioned a certain moment suspect vehicle position
In which position, needs to put into a large amount of manpower and material resources and screen the monitor video of traffic system by manual type, and gradually expand
Time and spatial dimension are checked.Therefore need to provide a kind of suspect vehicle discovery method for not depending on visual signature, it can
Automatically, efficiently find suspect vehicle, provide decision support for police's behavior of fighting crime.
Existing suspect vehicle discovery method mainly passes through image recognition technology, and discovery has the suspicion of certain surface
Vehicle, or utilize vehicle of the relational database query scientific discovery by locality, can not efficiently solve in the time and
The suspect vehicle without clear surface finds the problem in the uncertain situation in place.
Summary of the invention
The purpose of the present invention is provide a kind of suspect vehicle discovery based on space-time unique constraint in view of the above shortcomings
Algorithm can find the suspect vehicle without clear surface in the uncertain situation of when and where, solve to based on figure
As knowledge method for distinguishing and the method based on relational database can not efficiently find no Typical external feature and location information is not
The inquiry problem of specific suspect vehicle.
The present invention adopts the following technical solutions to achieve：
A kind of suspect vehicle discovery algorithm based on space-time unique constraint, includes the following steps,
1）It submitskA CFL stability condition
kIt is natural number,kValue range be 1≤k≤10；
kA CFL stability condition iskThe relatively unambiguous room and time range of group；
2）It successively searches and meets in track databasekThe vehicle of a CFL stability condition, and obtain final query result.
Step 2）Detailed process, include the following steps：
It is described belowiFor natural number, 1≤i<k；
2-1）If toiThe query result of a CFL stability condition is sky, then illustrates that there is no meet institute in traffic control system
There is CFL stability condition measuring car, query result is sky, enters step 2-3）；If toiThe inquiry knot of a CFL stability condition
Fruit is not sky, then enters step 2-2）；
2-2）Update source database and current space-time constraint condition；
If theiA CFL stability condition is the last one CFL stability condition, then current queries result is final inquiry knot
Fruit enters step 2-3）；It otherwise, will be for theiThe query result of a CFL stability condition, asi+ 1 space-time restriction item
The data source of part inquiry, enters step 2-1）Carry out theiThe inquiry of+1 CFL stability condition；
2-3）User is confirmed whether to need to adjust CFL stability condition, if not adjust, exports query result, terminates algorithm and holds
Row process；If desired it adjusts, then it is on the high side or on the low side to determine current results by user, if current results are on the low side, enters step 2-
4）；If current results are on the high side, 2-5 is entered step）；
2-4）According to the actual situation by user, most unessential space-time is deleted in the CFL stability condition for currently having calculated completion
Constraint condition or time and space limitation in the strongest CFL stability condition of relaxed constraints；Or by algorithms selection current
Query result and the smallest CFL stability condition of input source data volume ratio in the CFL stability condition of completion have been calculated, it is loosened
Time and space limitation；If deleting thejA CFL stability condition, then using the source data of j-th of CFL stability condition asj+1
The source data of a CFL stability condition, enters step 2-1）Carry out thejThe inquiry of+1 CFL stability condition；If looseningjIt is a
CFL stability condition then keepsjThe source data of+1 CFL stability condition is constant, enters step 2-1）Carry out thejA space-time
The inquiry of constraint condition；It is describedjFor natural number, 1≤j < k；
2-5）Completion is currently being calculated according to the actual situation by userkWhen reinforcing constraining most weak in a CFL stability condition
Time and space in empty constraint condition limit or add new CFL stability condition；Or it is calculated by algorithms selection currently
It completeskQuery result and the maximum CFL stability condition of input source data volume ratio in a CFL stability condition, reinforce at that time
Between and space limitation, regard current queries result as source data, by new CFL stability condition or reinforce time and space limit
CFL stability condition after system as systemk+ 1 CFL stability condition, enters step 2-1）Carry out thek+ 1 space-time is about
The inquiry of beam condition.
Invention advantage：Present invention is mainly used for efficiently find no Typical external in the uncertain situation of when and where
The suspect vehicle of feature.When suspect vehicle is without clear surface, the existing method based on image recognition is unavailable；Work as position
When confidence breath is uncertain, the querying method based on relational database can not efficiently find suspect vehicle.It is transported using suspect vehicle
The time range of suspect vehicle movement and spatial dimension are expressed as CFL stability condition, about based on space-time by dynamic space-time characteristic
The inquiry of Shu Jinhang suspect vehicle can support multiple CFL stability conditions and distributed processing mode to improve suspect vehicle
It was found that efficiency.
Detailed description of the invention
Below with reference to attached drawing, the invention will be further described：
Fig. 1 is the logical flow chart of suspect vehicle discovery algorithm of the present invention；
Fig. 2 is that the constraint space of suspect vehicle discovery algorithm of the present invention divides embodiment schematic diagram；
Fig. 3 is the distributed method flow chart divided based on data that the present invention uses；
Fig. 4 is the distributed method flow chart based on data backup that the present invention uses；
Fig. 5 is that the present invention is based on the implementation flow charts that the distributed algorithm suspect vehicle of cloud platform discovery algorithm is realized.
Specific embodiment
Illustrate the present invention program below by specific implementation example with reference.
Assuming that each carobjDuring the motiontMoment is by positionpWhen system can generate a record (obj,t, p), the format of each CFL stability condition isC=(T 1,T 2,pos 1,pos 2), whereinT 1WithT 2Indicate the time, andT 1≤T 2,pos 1=
(x 1,y 1) andpos 2= (x 2,y 2),x 1Withx 2Indicate positionpos 1Withpos 2Longitude,y 1Withy 2Indicate positionpos 1Withpos 2's
Latitude, with positionpos 1Withpos 2It is denoted as to be formed by rectangular area to angular vertexArea(C).If objectobj?tMoment position
In regionArea(C) in, i.e.,, then claim objectobjMeet condition CFL stability conditionC,
It is denoted as。
For some suspect, its general slip time and route usually can be inferred that, in its escape route
Upper selectionkA relatively unambiguous room and time range, i.e., it is givenkA CFL stability condition, each CFL stability conditionC i 's
Format is：C i =(T i1,T i2,pos i1,pos i2), 1≤i≤k, indicate suspected vehicles time interval [T i1,T i2] in can by with
Positionpos i1Withpos i2It is denoted as to be formed by rectangular area to angular vertexArea(C i ).Each carobjPath locus be one
The set of series of samples point, if vehicle existst 1Moment is by positionp 1,t 2Moment is by positionp 2...,t m Moment is by position
It setsp m , then its path bePath(obj)={(obj,t 1,p 1), (obj,t 2,p 2),…,(obj,t m ,p m )}.If certain vehicleobj
Path can meet and given all may be usedkA CFL stability condition, i.e.,,
Then think vehicleobjFor suspect vehicle.
Present invention discover that the process of suspect vehicle be searched in traffic control system it is eligiblekA space-time restriction item
All vehicles of part（1≤k≤10）, the logic flow of this method is as shown in Figure 1.It successively searches and meets in track databasek
The vehicle of a CFL stability condition, wherein for thei（1≤i<k）The query result of a CFL stability condition is asiAt+1
The data source of empty constraint condition inquiry.If query result is that empty or user thinks that current query result is more or less, explanation
The setting of current space-time constraint condition is unreasonable, the CFL stability condition for needing user's adjustment given at this time.CFL stability condition
Adjustment there are two kinds of situations：First is that CFL stability condition causes by force very much query result less for empty or result, need to delete at this time
Except certain CFL stability conditions or loosen the time in certain CFL stability conditions and space limitation；Second is that CFL stability condition
It is too weak to cause query result quantity more, need to increase more CFL stability conditions at this time or by certain CFL stability conditions
Time and space limitation reinforce.The concrete scheme of CFL stability condition adjustment should combine practical application to select by user.
kA CFL stability condition iskA relatively unambiguous room and time range.
It submitskThe specific steps of a CFL stability condition, including：
1-1）For each CFL stability conditionC i =(T i1,T i2,pos i1,pos i2), construct rectangular spatial areasArea(C i ), the square
Shape area of space conductC i Constraint space, two is respectively to angular vertexpos i1=(x i1,y i1) andpos i2=(x i2,y i2),
WhereiniIt is the serial number of CFL stability condition,T i1、T i2、pos i1、pos i2Respectively indicate suspect vehicle time interval [T i1,T i2]
It is interior to pass throughArea(C i ) region, 1≤i≤k；
1-2）By constraint spaceArea(C i ) according to the different grain size of GeoHash multiple subspaces are divided into, specifically it is divided into：
1-2-1）Using more granularity index structures of GeoHash, from the subspace for finding out covering constraint space compared with coarseness；
1-2-2）Find out the subspace that constraint space is covered in more fine granularity；
1-2-3）It is successively searched in thinner dynamics, until constraint space is completely coveredArea(C i )。
Fig. 2 is that the constraint space used in the present invention divides embodiment schematic diagram, in this embodiment, constraint spaceArea
(C i ) can be divided intoarea 1、area 2Witharea 3、area 4~area 6Three kinds of varigrained 6 sub-spaces.
It successively searches and meets in track databasekThe vehicle of a CFL stability condition, i.e.,：
2-1）In track databaseTDB i It is middle search time interval [T i1,T i2] interior by every sub-spacesarea j All vehicles
?Lobj(area j )；
2-2）Merge step 2-1）What is generated is all as a result, obtaining meeting CFL stability conditionC i All suspect vehicles
；
2-3）If all CFL stability conditions are processed at exporting suspect vehicle setGobj(C i )；Otherwise rail is updated
Mark databaseTDB i+1=Gobj(C i ), jump to step 2-1）.
Suspect vehicle discovery algorithm of the present invention can improve effect under distributed environment in a manner of parallel processing
Rate.
It illustrates below, mainly using the distributed method divided based on data and the distribution side based on data backup
Method.
1）The distributed method divided based on data：
If original track dataTDBBe stored in a manner of horizontal division multiple nodesn 1,n 2,…,n q On, i.e.,And, whereinqIndicate node
Quantity, then each noden i Final result is generated according to above-mentioned algorithmGobj(n i ).The suspect vehicle set of final output is, whereinResultFor meet all CFL stability conditions suspect vehicle set, such as
Shown in Fig. 3.
2）Distributed method based on data backup：
If original track dataTDBMultiple nodesn 1,n 2,…,n q On store original track dataTDBOne
Complete copy, i.e., each noden i Upper equal storing data collectionTDB, wherein 1≤i≤q.At this point, ifq≥k, then selectkA node,
Each noden i 1 CFL stability condition of upper processingC i Suspect vehicle inquiry, obtain resultResult(C i ), then by these
As a result it seeks common ground and the suspect vehicle for meeting all CFL stability conditions can be obtained, i.e.,,
As shown in Figure 4.Ifq<k, then each noden i Upper processingThe suspect vehicle of a CFL stability condition is inquired, and result is obtainedResult(n i ), then these results, which seek common ground, can be obtained the suspect vehicle for meeting all CFL stability conditions, i.e.,。
It is based on centralized approach based on the distributed method that data divide, calculation amount is less；And point based on data backup
Cloth method improves efficiency of algorithm using redundant node, can return the result in a relatively short period of time.Under cloud environment, due to number
Horizontal division is pressed according to general, and each deblocking stores a certain number of backups, therefore can draw using based on data
Divide and based on the distributed method that data backup combines, the distributed method divided based on data used between data block,
The distributed method based on data backup is used in each data backup inside data block.
Inventive algorithm requires to indicate the absolute position of vehicle with longitude and latitude in the collection process of vehicle position information, and
It is not based on the relative position of symbol.In traffic system, since data volume is huge, generallys use cloud platform and realize to vehicle location
The storage and analysis of information.Centralized algorithm of the present invention is deployed on each back end, and distributed strategy is disposed
In master（Host）On.When user submits CFL stability condition to carry out suspect vehicle inquiry, master is first according to first number
It is judged that how using cloud platform characteristic carry out distributed deployment, on each node carry out centralized algorithm inquiry generate
Local result, then local result is merged and generates final all suspect vehicles.
Assuming that existing track database TDB is divided horizontally into two data blocks of TDB1 and TDB2, each data block storage 3
Three copies of a copy, TDB1 are respectively stored on N1, N2 and N3 node, and the copy of TDB2 data block is stored in N4, N5
On tri- nodes of N6.When carrying out distributed treatment, it is respectively adopted on data block TDB1 and TDB2 based on data copy
Distributed method, and the suspect vehicle discovery result on TDB is the union of the result on TDB1 and TDB2, as shown in Figure 5.
Meet the local suspect vehicle of CFL stability condition C1, C2 and C3 in the solution respectively of three copies to TDB1, generates result
R11, R21 and R31, and then generate the suspect vehicle for being directed to data block TDB1；Needle can similarly be obtained
To the suspect vehicle of data block TDB1.Therefore it can obtain and be directed to when meeting of entire data set TDB
All suspect vehicles of empty constraint condition C1, C2 and C3。
Through the foregoing embodiment and analyze, compared with traditional suspect vehicle finds method, the present invention：
It 1, i.e., can without providing that the surfaces such as license plate, color, the model of suspect vehicle can find independent of visual signature
Doubt vehicle；
2, the suspect vehicle in support uncertain time and space discovery, i.e. given time range and position range can carry out can
Vehicle is doubted to search；
3, it supports distributed deployment to improve efficiency, both support the distributed deployment divided based on data or supports standby based on data
The distributed deployment of part.
Claims (4)
1. a kind of suspect vehicle based on space-time unique constraint finds algorithm, which is characterized in that include the following steps：
1）It submitskA CFL stability condition；
kIt is natural number,kValue range be 1≤k≤10；
kA CFL stability condition iskA relatively unambiguous room and time range；
2）It successively searches and meets in track databasekThe vehicle of a CFL stability condition, and obtain final query result.
2. the suspect vehicle according to claim 1 based on space-time unique constraint finds algorithm, which is characterized in that step 2）
It successively searches and meets in track databasekThe specific steps of the vehicle of a CFL stability condition include：
It is described belowiFor natural number, 1≤i<k；
2-1）If toiThe query result of a CFL stability condition is sky, then illustrates that there is no meet institute in traffic control system
There is CFL stability condition measuring car, query result is sky, enters step 2-3）；If toiThe inquiry knot of a CFL stability condition
Fruit is not sky, then enters step 2-2）；
2-2）Update source database and current space-time constraint condition；
If theiA CFL stability condition is the last one CFL stability condition, then current queries result is final inquiry knot
Fruit enters step 2-3）；It otherwise, will be for theiThe query result of a CFL stability condition, asi+ 1 space-time restriction item
The data source of part inquiry, enters step 2-1）Carry out theiThe inquiry of+1 CFL stability condition；
2-3）User is confirmed whether to need to adjust CFL stability condition, if not adjust, exports query result, terminates algorithm and holds
Row process；If desired it adjusts, then it is on the high side or on the low side to determine current results by user, if current results are on the low side, enters step 2-
4）；If current results are on the high side, 2-5 is entered step）；
2-4）According to the actual situation by user, most unessential space-time is deleted in the CFL stability condition for currently having calculated completion
Constraint condition or time and space limitation in the strongest CFL stability condition of relaxed constraints；Or by algorithms selection current
Query result and the smallest CFL stability condition of input source data volume ratio in the CFL stability condition of completion have been calculated, it is loosened
Time and space limitation；If deleting thejA CFL stability condition, then using the source data of j-th of CFL stability condition asj+1
The source data of a CFL stability condition, enters step 2-1）Carry out thejThe inquiry of+1 CFL stability condition；If looseningjIt is a
CFL stability condition then keepsjThe source data of+1 CFL stability condition is constant, enters step 2-1）Carry out thejA space-time
The inquiry of constraint condition；It is describedjFor natural number, 1≤j < k；
2-5）Completion is currently being calculated according to the actual situation by userkWhen reinforcing constraining most weak in a CFL stability condition
Time and space in empty constraint condition limit or add new CFL stability condition；Or it is calculated by algorithms selection currently
It completeskQuery result and the maximum CFL stability condition of input source data volume ratio in a CFL stability condition, reinforce at that time
Between and space limitation, regard current queries result as source data, by new CFL stability condition or reinforce time and space limit
CFL stability condition after system as systemk+ 1 CFL stability condition, enters step 2-1）Carry out thek+ 1 space-time is about
The inquiry of beam condition.
3. the suspect vehicle according to claim 1 based on space-time unique constraint finds algorithm, which is characterized in that step 1）
It submitskThe specific steps of a CFL stability condition, including：
1-1）For each CFL stability conditionC i =(T i1,T i2,pos i1,pos i2), construct rectangular spatial areasArea(C i ), the square
Shape area of space conductC i Constraint space, two is respectively to angular vertexpos i1=(x i1,y i1) andpos i2=(x i2,y i2),
WhereiniIt is the serial number of CFL stability condition,T i1、T i2、pos i1、pos i2Respectively indicate suspect vehicle time interval [T i1,T i2]
It is interior to pass throughArea(C i ) region, 1≤i≤k；
1-2）By constraint spaceArea(C i ) according to the different grain size of GeoHash multiple subspaces are divided into, specifically it is divided into：
1-2-1）Using more granularity index structures of GeoHash, from the subspace for finding out covering constraint space compared with coarseness；
1-2-2）Find out the subspace that constraint space is covered in more fine granularity；
1-2-3）It is successively searched in thinner dynamics, until constraint space is completely coveredArea(C i )。
4. the suspect vehicle according to claim 3 based on space-time unique constraint finds algorithm, which is characterized in that in track
It successively searches and meets in databasekThe vehicle of a CFL stability condition, i.e.,：
2-1）In track databaseTDB i It is middle search time interval [T i1,T i2] interior by every sub-spacesarea j All vehicles
?Lobj(area j )；
2-2）Merge step 2-1）What is generated is all as a result, obtaining meeting CFL stability conditionC i All suspect vehicles
；
2-3）If all CFL stability conditions are processed at exporting suspect vehicle setGobj(C i )；Otherwise track is updated
DatabaseTDB i+1=Gobj(C i ), jump to step 2-1）.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810589430.XA CN108874979B (en) | 2018-06-08 | 2018-06-08 | Suspicious vehicle discovery method based on space-time range constraint |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810589430.XA CN108874979B (en) | 2018-06-08 | 2018-06-08 | Suspicious vehicle discovery method based on space-time range constraint |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108874979A true CN108874979A (en) | 2018-11-23 |
| CN108874979B CN108874979B (en) | 2021-04-30 |
Family
ID=64337573
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810589430.XA Expired - Fee Related CN108874979B (en) | 2018-06-08 | 2018-06-08 | Suspicious vehicle discovery method based on space-time range constraint |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108874979B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110083599A (en) * | 2019-03-28 | 2019-08-02 | 华东师范大学 | A kind of track of vehicle data index method based on temporal-spatial interpolating |
| CN110334232A (en) * | 2019-06-28 | 2019-10-15 | 深圳市商汤科技有限公司 | Records application method and device, storage medium |
| CN110888884A (en) * | 2019-11-26 | 2020-03-17 | 杭州三汇数字信息技术有限公司 | Vehicle code fitting method and system based on geohash matching |
| CN113742606A (en) * | 2020-05-29 | 2021-12-03 | 京东城市(北京)数字科技有限公司 | Object identification method and device, electronic equipment and readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007058608A (en) * | 2005-08-24 | 2007-03-08 | Ntt Communications Kk | Parking lot guiding device |
| CN103294775A (en) * | 2013-05-10 | 2013-09-11 | 苏州祥益网络科技有限公司 | Police service cloud image recognition vehicle management and control system based on geographic space-time constraint |
| CN104581803A (en) * | 2015-02-03 | 2015-04-29 | 杨奎武 | Urban vehicle-mounted network mobile model of time-space restriction |
| US20160203522A1 (en) * | 2014-03-22 | 2016-07-14 | Retailmenot, Inc. | Peer-to-peer geotargeting content with ad-hoc mesh networks |
| CN107730889A (en) * | 2017-10-30 | 2018-02-23 | 昆明理工大学 | A kind of target vehicle search method based on traffic video |
| CN108021561A (en) * | 2016-10-28 | 2018-05-11 | 沈阳建筑大学 | A kind of abnormal mobile object detection method based on track data stream |
-
2018
- 2018-06-08 CN CN201810589430.XA patent/CN108874979B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007058608A (en) * | 2005-08-24 | 2007-03-08 | Ntt Communications Kk | Parking lot guiding device |
| CN103294775A (en) * | 2013-05-10 | 2013-09-11 | 苏州祥益网络科技有限公司 | Police service cloud image recognition vehicle management and control system based on geographic space-time constraint |
| US20160203522A1 (en) * | 2014-03-22 | 2016-07-14 | Retailmenot, Inc. | Peer-to-peer geotargeting content with ad-hoc mesh networks |
| CN104581803A (en) * | 2015-02-03 | 2015-04-29 | 杨奎武 | Urban vehicle-mounted network mobile model of time-space restriction |
| CN108021561A (en) * | 2016-10-28 | 2018-05-11 | 沈阳建筑大学 | A kind of abnormal mobile object detection method based on track data stream |
| CN107730889A (en) * | 2017-10-30 | 2018-02-23 | 昆明理工大学 | A kind of target vehicle search method based on traffic video |
Non-Patent Citations (1)
| Title |
|---|
| 朱美玲 等: "基于车牌识别流数据的车辆伴随模式发现方法", 《软件学报》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110083599A (en) * | 2019-03-28 | 2019-08-02 | 华东师范大学 | A kind of track of vehicle data index method based on temporal-spatial interpolating |
| CN110083599B (en) * | 2019-03-28 | 2023-06-23 | 华东师范大学 | Vehicle track data indexing method based on space-time interpolation |
| CN110334232A (en) * | 2019-06-28 | 2019-10-15 | 深圳市商汤科技有限公司 | Records application method and device, storage medium |
| CN110888884A (en) * | 2019-11-26 | 2020-03-17 | 杭州三汇数字信息技术有限公司 | Vehicle code fitting method and system based on geohash matching |
| CN110888884B (en) * | 2019-11-26 | 2022-08-23 | 杭州三汇数字信息技术有限公司 | Vehicle code fitting method and system based on geohash matching |
| CN113742606A (en) * | 2020-05-29 | 2021-12-03 | 京东城市(北京)数字科技有限公司 | Object identification method and device, electronic equipment and readable storage medium |
| CN113742606B (en) * | 2020-05-29 | 2024-10-22 | 京东城市(北京)数字科技有限公司 | Object recognition method, device, electronic equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108874979B (en) | 2021-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108874979A (en) | A kind of suspect vehicle discovery algorithm based on space-time unique constraint | |
| CN112396027B (en) | Vehicle re-identification method based on graph convolution neural network | |
| CN109766868B (en) | Real scene pedestrian-sheltered detection network based on body key point detection and detection method thereof | |
| KR101395094B1 (en) | Method and system for detecting object in input image | |
| CN110516586A (en) | A kind of facial image clustering method, system, product and medium | |
| Chen et al. | S-CNN: Subcategory-aware convolutional networks for object detection | |
| Gong et al. | Object detection based on improved YOLOv3-tiny | |
| Huang et al. | Multiple objects tracking in the UAV system based on hierarchical deep high-resolution network | |
| Mohan et al. | Environment selection and hierarchical place recognition | |
| CN111524140B (en) | Medical image semantic segmentation method based on CNN and random forest method | |
| CN107688830A (en) | It is a kind of for case string and show survey visual information association figure layer generation method | |
| CN109993061A (en) | A kind of human face detection and tracing method, system and terminal device | |
| Kordopatis-Zilos et al. | Leveraging efficientnet and contrastive learning for accurate global-scale location estimation | |
| Abbott et al. | Deep object classification in low resolution lwir imagery via transfer learning | |
| CN117036897A (en) | Method for detecting few sample targets based on Meta RCNN | |
| Chen et al. | Svqnet: Sparse voxel-adjacent query network for 4d spatio-temporal lidar semantic segmentation | |
| Zhao et al. | BUPT-MCPRL at TRECVID 2016. | |
| Fofana et al. | Optimal Flame Detection of Fires in Videos Based on Deep Learning and the Use of Various Optimizers | |
| CN112990048B (en) | Vehicle pattern recognition method and device | |
| Yang et al. | Self-ordering point clouds | |
| Tong et al. | Research on object detection in campus scene based on faster R-CNN | |
| Zhu et al. | Real-time traffic sign detection based on YOLOv2 | |
| Cheng et al. | Class attendance checking system based on deep learning and global optimization | |
| Baldwin et al. | Automatic fuzzy Cartesian granule feature discovery using genetic programming in image understanding | |
| Xiaozhou et al. | Recognition method of outdoor design scene based on support vector machine and feature fusion |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230105 Address after: 221000 7th floor, Building 8, Xuzhou Software Park, No. 6, Software Park Road, Quanshan District, Xuzhou City, Jiangsu Province Patentee after: Jiangsu Tianze Zhilian Information Technology Co.,Ltd. Address before: No. 80, yunlongshan Road, Jianye District, Nanjing City, Jiangsu Province, 210019 Patentee before: TIANZE INFORMATION INDUSTRY Corp. |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210430 |