CN104778245A - Similar trajectory mining method and device on basis of massive license plate identification data - Google Patents
Similar trajectory mining method and device on basis of massive license plate identification data Download PDFInfo
- Publication number
- CN104778245A CN104778245A CN201510167058.XA CN201510167058A CN104778245A CN 104778245 A CN104778245 A CN 104778245A CN 201510167058 A CN201510167058 A CN 201510167058A CN 104778245 A CN104778245 A CN 104778245A
- Authority
- CN
- China
- Prior art keywords
- track
- license plate
- data
- vehicle
- identification data
- 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
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention discloses a similar trajectory mining method and a similar trajectory mining device on the basis of massive license plate identification data. The method comprises three main steps of trajectory organization and screening, point escort relationship calculation and trajectory similarity judgment. According to the similar trajectory mining method and the similar trajectory mining device on the basis of massive license plate identification data, the problem of lagging of calculation of response time under massive data sets is solved; calculation accuracy is improved on the basis of analysis on the license plate identification data; due to use of a Hadoop Map Reduce distributed processing mode, calculation efficiency is improved; similar trajectories are efficiently and rapidly mined; the similar trajectory mining method and the similar trajectory mining device can be used for finding escort vehicles in the field of the traffic service.
Description
Technical field
The invention belongs to the method for areas of information technology, can based on the magnanimity license plate identification data similar track of discovery vehicle rapidly and efficiently in intelligent transportation field, and then for identifying escort vehicle.
The present invention relates to again the device of large data analysis, uses described method to complete similar trajectory calculation in Hadoop MapReduce cluster environment, for the data mining of intelligent transportation field, can be public security police and handles a case and provide auxiliary.
Background technology
In city road network, the similar track of vehicle excavates, it is the important service computation of intelligent transportation field, correlative study is emphasis and focus always, may be used for the bus route design of intelligent transportation system, smart city environment, suspicion of crime vehicle is deployed to ensure effective monitoring and control of illegal activities, also can provide reference proposition based on vehicle driving rule for city road planning, there is far-reaching social and economic significance.Similar track excavates to be needed to find the similarity of track of vehicle on Spatial Dimension and time dimension, is that a kind of typical large data analysis calculates.Concrete, the similar track of the vehicle based on magnanimity license plate identification data excavates, and for any period of history or current slot, with different similarity definitions and constraint condition, finds the similar track of vehicle.License plate identification data described here is from city traffick information acquisition, relevant license plate recognition technology is a class technology emerging in recent years: the vehicle image information analyzing road camera collection, identify license plate number wherein, and by there is position, shooting time, to take pictures and the Information encapsulation such as direction of traffic is license plate identification data.Perfect along with monitoring technique, car plate capture rate and recognition accuracy significantly improve, and the vehicle driving information acquiring technology based on license plate identification data is extensively adopted in numerous city, have driven the mining analysis technology of association area business and the development of calculation element.Vehicle is as mobile object, and position changes along with Time Continuous.Compare the Floating Car vehicle data collection technology based on GPS technology, the collecting vehicle information technology based on license plate identification data has the advantages such as work continuity is strong, data accuracy is high, detection sample size is large, covering vehicle range is wide.Derive from the license plate identification data of urban road Real-Time Monitoring, comprise the typical time-space attribute such as monitoring time, geographic position, and the association attributes of vehicle itself, have typical temporal and spatial correlations, time Order continuous, the position feature that can survey.In addition, being connected and dispose extensively due to front-end equipment by private network, license plate identification data has magnanimity, feature that renewal frequency is high.A large size city can accumulate more than 10,000,000,000 license plate identification data records for 1 year, and the scale of data set will substantially exceed traditional sampling method, so magnanimity license plate identification data excavates service computation to similar track it is also proposed higher requirement.
Based in the vehicle adjoint mode excavation of similar track, the people such as the Lu-An Tang of University of Illinois give a kind of companion's candidate collection model at flagship meeting International Conference on Data Engineering and accelerate the intelligent crossover method of process, for the trip companion of the acquisition vehicle of low expense from the dynamic stream data of vehicle GPS.The people such as the Fang Aifen of National Traffic Management Research Institute, Ministry of Public Security write articles at " computer utility and software ", a kind of discovery algorithm of escort vehicle is given based on mistake car data, the inquiry problem of escort vehicle is converted into the association rule mining problem of data mining, method has the advantages that efficiency is high and extendability is strong.The Zhao Xinyong of Harbin Institute of Technology publishes an article in " Traffic transport system engineering and information ", analyzes vehicle driving feature based on vehicle identification data, and filters out possible escort vehicle according to specific experience desired value.In the similarity measurement of track data stream, the people such as the Zhao Hongbin of Harbin Engineering University, analyze the character that road network space track is similar, the space-time method for expressing of a kind of mobile object track modeling is proposed in " computer engineering and application ", track is transformed into Euclidean space from road network space, and provide a kind of similarity measurement method between track based on point of interest POI (Points Of Interesting) distance, effectively abbreviation carried out to track and reduce the number of track interior joint, thus reduction Algorithms T-cbmplexity, avoid mobile object to be in actual applications often limited to spatial network and track and distance treatment technology thereof in existing Euclidean space cannot be utilized.Similar, the people such as Zhang Yanling, based on the characteristic of road network space mobile object, consider space-time similarity but not are only spatial simlanty, propose a kind of similar track search method of mobile object in road network space at " software " magazine.
Can see from above related work both at home and abroad, the research that the similar track of vehicle excavates still is in developing stage, and technology is relatively immature.The problem of two aspects below main existence or defect:
First aspect, work at present research is most based on limited data set, as the GPS monitor data of Floating Car in a couple of days, but not the vehicle identification data of several months rank, bring complicacy to consider to current huge traffic data not enough, cause the low or inquiry for the treatment of effeciency under large data environment slowly.
Second aspect, manyly during the algorithm setting parameter such as support does not consider the feature of vehicle as mobile object, along with setting value is higher or on the low side, cause Query Result too much or very few, and accuracy is low.
For the mining analysis of magnanimity license plate identification data, propose the setting parameter of applicability, there is not yet relevant report up to now.
Summary of the invention
The object of the invention is to overcome above-mentioned technological deficiency, thus similar track digging efficiency and the not high problem of accuracy under solving magnanimity license plate identification data.
The present invention, by some accompanying relationship definition track of vehicle similarity, proposes a kind of similar track method for digging calculated based on multistage tasks in parallel.Calculating, by Hadoop MapReduce distributed environment, is carried out Task-decomposing, scheduling and executed in parallel, is realized efficient optimization process by described method.
Specifically, the invention discloses following technical scheme:
1., based on a similar track method for digging for magnanimity license plate identification data, it is characterized in that, described method comprises:
(1) track tissue and screening step, for removing invalid redundant data;
(2) with calculation procedure, for Maintenance Intermediate Point spinoff collection;
(3) track similarity determination, for adding up and calculating track of vehicle similarity, obtains the escort vehicle pair with similarity track.
2. the similar track method for digging based on magnanimity license plate identification data described in, it is characterized in that, wherein said track tissue and screening step, utilize the vehicle identification data related in a kind of track list structure stores processor process, this step is completed by the first order of three class pipeline, is a MapReduce computational tasks, input magnanimity license plate identification data collection, Output rusults is track chained list set 1, and passes to second level streamline use.
3. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, described point with calculation procedure, for obtaining intermediate point spinoff collection; The result of calculation of the first order of the track chained list set 1 described in this step utilizes, being completed by the second level of three class pipeline, is the realization of another MapReduce operation, the track list structure 1 described in input, Output rusults is track chained list set 2, and passes to the use of third level streamline.
4. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, described track similarity determination step, for calculating track of vehicle similarity, obtains the escort vehicle pair with similar track; The result of calculation of the second level of the track chained list set 2 described in this step utilizes, being completed by the third level of three class pipeline, is the realization of another MapReduce operation, the track list structure 2 described in input, Output rusults is track chained list set 3, and final writing in files system or database.
5. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, described magnanimity license plate identification data collection L represents, refers to all vehicle information data that on tested road network, each monitoring point is caught; Every bar license plate identification data l ∈ L can be expressed as
wherein v
irepresent the number-plate number (can unique identification vehicle),
represent vehicle v
ithrough monitoring point n
k; Further,
wherein
represent the monitoring point n of vehicle process
kgeographic position,
represent that vehicle is through monitoring point n
ktime.
6. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, described track of vehicle t
irepresenting, is vehicle v
ione group of monitoring point sequence of process in chronological order in a time range;
Further, t
ican be expressed as:
wherein, to any p<q, have
t
iin the monitoring point number that comprises be called and the length of track be designated as l
i.
7. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, the described adjoint sim of point
n(v
i, v
j) represent, refer to two vehicle v
iand v
jat certain hour threshold value δ
tinterior priority is through certain monitoring point n
pand its a kind of relation met the following conditions:
only may to exist in this monitoring point through two of same monitoring point vehicles within the scope of certain hour and once put accompanying relationship.
8. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, judge similar track by track similarity; Track similarity refers to the similarity degree of two tracks of vehicle, with simD (t
i, t
j) represent;
Further,
wherein l
iand l
jbe respectively vehicle i and vehicle j course length, m is that described two cars is by way of the monitoring point number with an accompanying relationship.
9. the similar track method for digging based on magnanimity license plate identification data described in, is characterized in that, described track of vehicle t
iwith track of vehicle t
jfor similar track, refer to given trace similarity threshold δ
d, course length threshold value δ
lwith in time range dur, t
iand t
jthe track pair simultaneously meeting following two conditions:
(1) track t
iand t
jsimilarity simD (t
i, t
j)>=δ
d;
(2) track t
iand t
jcourse length l
i>=δ
l, l
j>=δ
l.
10. the similar track method for digging based on magnanimity license plate identification data described in, it is characterized in that, the calculating of similar track is obtained in described method, can according to the description of step 5-9, the threshold value of setting different parameters is for limiting design conditions, avoid shorter track of vehicle as the erroneous judgement of similar track, and invalid data is filtered; Specific as follows:
Suppose that set point is with time threshold δ
t, track similarity threshold δ
dwith course length threshold value δ
l, utilize existing license plate identification data collection L, find out all similar track S set T of vehicle meeting described definition in given time range dur;
Concrete,
The similar track method for digging based on magnanimity license plate identification data described in 11., it is characterized in that, described first order streamline MapReduce operation process, the carrying out track of vehicle length being less than to given trace length threshold is deleted, and sets up the track chained list of all vehicles further; Concrete, following two steps can be decomposed into further:
(1) Map task burst from file system reads license plate identification data, obtains the license plate identification data of scope dur preset time; Described license plate identification data presses monitoring time dividing data burst; Described Map task license plate identification data record is converted to license plate number be key, the data item that is value with time and monitoring point; The data item of same keys will be sent to same Reduce task;
(2) license plate identification data is organized as track of vehicle chained list by license plate number by Reduce task, forms the track in described time range dur; Described Reduce task judges length to each track of vehicle chained list, deletes and is less than course length threshold value δ
lchained list, all the other qualified tracks are exported as described track chained list set 1.
The similar track method for digging based on magnanimity license plate identification data described in 12., is characterized in that, described second level streamline MapReduce operation process can be divided into following two steps:
(1) Map task reads the result of first order operation, and to be converted to monitoring point be key, monitoring time and license plate number is the data item of value, passes to Reduce task; The data item of same keys will be sent to same Reduce task;
(2) groups of data items of same monitoring point is woven to a blockchain table excessively by Reduce task, forms several described blockchain table excessively; Described Reduce task according to monitoring time successively sequence and calculation level adjoint, export the result data item meeting an accompanying relationship; Described result data item comprises the license plate number of two cars, adjoint time range and respective course length.
The similar track method for digging based on magnanimity license plate identification data described in 13., is characterized in that, described point, with calculating, carries out as follows:
(1) obtain untreated blockchain table excessively, from gauge outfit, obtain data item 1;
(2) if data item 1 exists the follow-up data item 2 do not scanned, flag data item 2 is for scan, judge whether the difference of time attribute contained by two data item 1 and data item 2 is less than time threshold: if meet threshold value, the license plate number 1 that output data item 1 comprises and the license plate number 2 that data item 2 comprises are combined as key, with time range for being worth, turn (2); If do not meet threshold value, turn (2);
(3) if data item 1 does not exist do not scan follow-up, if data item 1 be not described chained list last, then the direct follow-up data item of data item 1 is labeled as data item 1, turns (2); If data item 1 be described chained list last, then mark this chained list processed, turn (1).
The similar track method for digging based on magnanimity license plate identification data described in 14., is characterized in that, described third level streamline MapReduce operation process can be divided into following two steps:
(1) Map task reads the result that second level streamline MapReduce operation exports, and license plate number combination identical data will be sent to the process of same Reduce task;
(2) Reduce task counts license plate number combination, also namely forms the some accompanying relationship counting of two; Calculate the track similarity of described two cars, export the result data item meeting similarity threshold; Described result data item, with the car plate number key of described two cars, with adjoint time range, the adjoint number of times of point and track similarity for value.
15. 1 kinds, based on the similar track excavating gear of magnanimity license plate identification data, is characterized in that, comprise as lower component:
(1) data memory module: this module divides module with digital independent, deblocking computing module is connected with similar trajectory computation module, adopts distributed file system framework; This module deposits magnanimity license plate identification data, the intermediate data of operation excavates net result with result, road network monitoring point basic data with similar track;
(2) digital independent divides module: this module is connected with data memory module, deblocking computing module, for track tissue and screening; This module reads the magnanimity license plate identification data in data memory module, divides and removes invalid redundant data, the data subset of each piecemeal met the demands is passed to deblocking computing module and calculates by monitoring point;
(3) deblocking computing module: this module divides module with data memory module, digital independent and is connected, receives the block data subset that digital independent divides module, for putting with calculating; Result of calculation is stored in the intermediate result collection of data memory module;
(4) similar trajectory computation module: this module is connected with data memory module, Man Machine Interface, for track Similarity measures and judgement; The parameter that this module is transmitted according to Man Machine Interface, reads the intermediate result collection of deblocking computing module stored in data memory module, calculates the vehicle pair obtaining and have similar track, and result is returned Man Machine Interface;
(5) Man Machine Interface: this module and similar track are calculated module and be connected, this module provides interactive interface for user, and support that user inputs calculating parameter, described parameter comprises course length threshold value, the adjoint time threshold of point, similarity threshold and time range; Described parameter will pass to similar trajectory computation module, and described Man Machine Interface receives the result of calculation of described similar trajectory computation module, and by the map of this result of calculation in described interactive interface for user presents.
The similar track that the present invention effectively can be applied to magnanimity license plate identification data excavates, the high and strong adaptability of counting yield.This can according to following experiment test explanation.Such as, device section of the present invention is deployed on the cluster of ten machines, and every platform machines configurations is 4 core CPU, 4G internal memories, cluster can distributed storage capacity be 800G.The device of above-mentioned configuration, based on certain city true license plate identification data of 80 days (data volume is more than 400,000,000), the calculating of excavating wherein certain day similar track needs about 2 minutes.Different according to the threshold value inputted in Man Machine Interface module, result of calculation can show tens to several thousand different similar tracks in the map page.
Accompanying drawing explanation
The description that the present invention can carry out with reference to following Figure and being better understood, and in all of the figs, employ identical or similar Reference numeral to identify.Described accompanying drawing comprises in this manual together with detailed description below and forms the part of this instructions, and is used for illustrating preferred embodiment of the present invention further and explaining principle and advantage of the present invention.In the accompanying drawings:
Fig. 1 is invalid redundant data screening schematic diagram of the present invention;
Fig. 2 is the basic procedure that the similar track of vehicle of the present invention excavates;
Fig. 3 is that point of the present invention is with calculation flow chart;
Fig. 4 is the Organization Chart of similar track excavating gear of the present invention.
Embodiment
For making the technical problem to be solved in the present invention, technical scheme and advantage clearly, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.It is to be understood that; following specific embodiment only schematically illustrates; unless otherwise specified; even if the Different Optimization means in the different embodiment of the application and each embodiment are not set forth in the same embodiment; also should be understood as that and can be applied to any other side described here, embodiment or example; unless incompatible or be explicitly excluded outside in this article with it, otherwise combination between all embodiments not having limit of the application or sub-portfolio are all in record of the present invention and protection domain.Purport of the present invention and preferred implementation thereof is elaborated below with reference to instructions and accompanying drawing.
Embodiment 1
The invention provides a kind of similar track excavating gear, it mainly comprises data memory module, reads Data Placement module, deblocking computing module, similar trajectory computation module and Man Machine Interface.Below with reference to the accompanying drawings 4 modules is described in detail.
Data memory module: this module divides module with digital independent, deblocking computing module is connected with similar trajectory computation module, adopts distributed file system framework; This module deposits magnanimity license plate identification data, the intermediate data of operation excavates net result with result, road network monitoring point basic data with similar track;
Digital independent divides module: this module is connected with data memory module, deblocking computing module, for track tissue and screening; This module reads the magnanimity license plate identification data in data memory module, divides and removes invalid redundant data, the data subset of each piecemeal met the demands is passed to deblocking computing module and calculates by monitoring point;
Deblocking computing module: this module divides module with data memory module, digital independent and is connected, receives the block data subset that digital independent divides module, for putting with calculating; Result of calculation is stored in the intermediate result collection of data memory module;
Similar trajectory computation module: this module is connected with data memory module, Man Machine Interface, for track Similarity measures and judgement; The parameter that this module is transmitted according to Man Machine Interface, reads the intermediate result collection of deblocking computing module stored in data memory module, calculates the vehicle pair obtaining and have similar track, and result is returned Man Machine Interface;
Man Machine Interface: this module and similar track are calculated module and be connected, this module provides interactive interface for user, and support that user inputs calculating parameter, described parameter comprises course length threshold value, the adjoint time threshold of point, similarity threshold and time range; Described parameter will pass to similar trajectory computation module, and described Man Machine Interface receives the result of calculation of described similar trajectory computation module, and by the map of this result of calculation in described interactive interface for user presents.
Road camera sensing device catches identification data in real time, transfers to described similar track excavating gear by private network.The data item of these data comprises 22 attributes such as monitoring time, monitoring point ID, license plate number, with structurized data item form, is stored to the data memory module of device, as history license plate identification data.These data form a text in units of sky.Based on the magnanimity history license plate identification data of described storage, device of the present invention can carry out calculating and the excavation of the similar track of vehicle.
Device carries out redundancy Screening Treatment to license plate identification data, removes invalid redundant data information.May occur following situation in reality scene, certain car has only occurred seldom (being such as less than or equal to 2 times) several times at license plate identification data collection, and make course length too short, correlation calculation result is nonsensical.So need to screen out this kind of invalid redundant data.Fig. 1 illustrates and removes the situation of data: once, being less than the threshold value 2 of setting, can being removed after first order Activity Calculation only appears in the identification data of vehicle 1,3,7 in the historical data, and result of calculation is as shown on the right of Fig. 1.Remove invalid redundant data by screening, reduce the scale of data set, the computing velocity of two-stage operation after can improving and accuracy rate.
The present invention also comprises a kind of similar track method for digging based on magnanimity license plate identification data:
(1) track tissue and screening step, for removing invalid redundant data;
(2) with calculation procedure, for Maintenance Intermediate Point spinoff collection;
(3) track similarity determination, for adding up and calculating track of vehicle similarity, obtains the escort vehicle pair with similarity track.
Described track tissue and screening step, utilize the vehicle identification data related in a kind of track list structure stores processor process, this step is completed by the first order of three class pipeline, it is a MapReduce computational tasks, input magnanimity license plate identification data collection, Output rusults is track chained list set 1, and passes to second level streamline use.
Described point with calculation procedure, for obtaining intermediate point spinoff collection; The result of calculation of the first order of the track chained list set 1 described in this step utilizes, being completed by the second level of three class pipeline, is the realization of another MapReduce operation, the track list structure 1 described in input, Output rusults is track chained list set 2, and passes to the use of third level streamline.
Described track similarity determination step, for calculating track of vehicle similarity, obtains the escort vehicle pair with similar track; The result of calculation of the second level of the track chained list set 2 described in this step utilizes, being completed by the third level of three class pipeline, is the realization of another MapReduce operation, the track list structure 2 described in input, Output rusults is track chained list set 3, and final writing in files system or database.
Described magnanimity license plate identification data collection L represents, refers to all vehicle information data that on tested road network, each monitoring point is caught; Every bar license plate identification data l ∈ L can be expressed as
wherein v
irepresent the number-plate number (can unique identification vehicle),
represent vehicle v
ithrough monitoring point n
k; Further,
wherein
represent the monitoring point n of vehicle process
kgeographic position,
represent that vehicle is through monitoring point n
ktime.
Described track of vehicle t
irepresenting, is vehicle v
ione group of monitoring point sequence of process in chronological order in a time range; Further, t
ican be expressed as:
wherein, to any p<q, have
t
iin the monitoring point number that comprises be called and the length of track be designated as l
i.
The described adjoint sim of point
n(v
i, v
j) represent, refer to two vehicle v
iand v
jat certain hour threshold value δ
tinterior priority is through certain monitoring point n
pand its a kind of relation met the following conditions:
only may to exist in this monitoring point through two of same monitoring point vehicles within the scope of certain hour and once put accompanying relationship.
Method judges similar track by track similarity; Track similarity refers to the similarity degree of two tracks of vehicle, with simD (t
i, t
j) represent;
Further,
wherein l
iand l
jbe respectively vehicle i and vehicle j course length, m is that described two cars is by way of the monitoring point number with an accompanying relationship.
Described track of vehicle t
iwith track of vehicle t
jfor similar track, refer to given trace similarity threshold δ
d, course length threshold value δ
lwith in time range dur, t
iand t
jthe track pair simultaneously meeting following two conditions:
(1) track t
iand t
jsimilarity simD (t
i, t
j)>=δ
d;
(2) track t
iand t
jcourse length l
i>=δ
l, l
j>=δ
l.
Obtain the calculating of similar track in described method, the threshold value that can set different parameters, for limiting design conditions, is avoided shorter track of vehicle as the erroneous judgement of similar track, and is filtered invalid data; Specific as follows:
Suppose that set point is with time threshold δ
t, track similarity threshold δ
dwith course length threshold value δ
l, utilize existing license plate identification data collection L, find out all similar track S set T of vehicle meeting described definition in given time range dur;
Concrete,
Described first order streamline MapReduce operation process, the carrying out track of vehicle length being less than to given trace length threshold is deleted, and sets up the track chained list of all vehicles further; Concrete, following two steps can be decomposed into further:
(1) Map task burst from file system reads license plate identification data, obtains the license plate identification data of scope dur preset time; Described license plate identification data presses monitoring time dividing data burst; Described Map task license plate identification data record is converted to license plate number be key, the data item that is value with time and monitoring point; The data item of same keys will be sent to same Reduce task;
(2) license plate identification data is organized as track of vehicle chained list by license plate number by Reduce task, forms the track in described time range dur; Described Reduce task judges length to each track of vehicle chained list, deletes and is less than course length threshold value δ
lchained list, all the other qualified tracks are exported as described track chained list set 1.
Described second level streamline MapReduce operation process can be divided into following two steps:
(1) Map task reads the result of first order operation, and to be converted to monitoring point be key, monitoring time and license plate number is the data item of value, passes to Reduce task; The data item of same keys will be sent to same Reduce task;
(2) groups of data items of same monitoring point is woven to a blockchain table excessively by Reduce task, forms several described blockchain table excessively; Described Reduce task according to monitoring time successively sequence and calculation level adjoint, export the result data item meeting an accompanying relationship; Described result data item comprises the license plate number of two cars, adjoint time range and respective course length.
Described point, with calculating, carries out as follows:
(1) obtain untreated blockchain table excessively, from gauge outfit, obtain data item 1;
(2) if data item 1 exists the follow-up data item 2 do not scanned, flag data item 2 is for scan, judge whether the difference of time attribute contained by two data item 1 and data item 2 is less than time threshold: if meet threshold value, the license plate number 1 that output data item 1 comprises and the license plate number 2 that data item 2 comprises are combined as key, with time range for being worth, turn (2); If do not meet threshold value, turn (2);
(3) if data item 1 does not exist do not scan follow-up, if data item 1 be not described chained list last, then the direct follow-up data item of data item 1 is labeled as data item 1, turns (2); If data item 1 be described chained list last, then mark this chained list processed, turn (1).
Described third level streamline MapReduce operation process can be divided into following two steps:
(1) Map task reads the result that second level streamline MapReduce operation exports, and license plate number combination identical data will be sent to the process of same Reduce task;
(2) Reduce task counts license plate number combination, also namely forms the some accompanying relationship counting of two; Calculate the track similarity of described two cars, export the result data item meeting similarity threshold; Described result data item, with the car plate number key of described two cars, with adjoint time range, the adjoint number of times of point and track similarity for value.
Embodiment 2
Composition graphs 2 basic procedure excavates flow process to similar track and is described.First, read history license plate identification data, reject invalid redundant data and realize data screening; Then carry out an accompanying relationship to the data after screening to calculate, a spinoff is write similar track Candidate Set; Finally according to the threshold calculations track similarity of setting, return the similar track and relevant escort vehicle that satisfy condition.In fig. 2,
S1 is history license plate identification data, is used for ensuing calculating as raw data.After first order operation track tissue with the process of screening, result forms the data memory module of track chained list S set 2 writing station.
S2 experienced by the track tissue of first order operation and the track chained list set after screening, and is the track of vehicle data set after rejecting invalid redundant data.For S2, device, by the calculating through second level operation, by monitoring point dividing data, and sorts by the time order and function through monitoring point, all data item identical for monitoring point is organized into one and crosses blockchain table; From linked list head node, all data item successively after contrast in time range threshold value, and judge whether that there is an accompanying relationship.S3 is formed, the data memory module of writing station after calculating completes.
S3 be experienced by second level operation some accompanying relationship calculate after result, be similar track Candidate Set.Track similarity will be judged through third level Activity Calculation track similarity for S3 device.In this process, for S3 extract car plate to, time range, point be with the information such as number and course length; By car plate pair, by two license plate numbers and vehicle with the time stored in chained list.Each data item of this chained list, have recorded an accompanying relationship of two cars at the appointed time scope certain monitoring point inherent.Afterwards, to each record judge similarity whether meet before the threshold value of setting, if meet this record of threshold value by the data memory module of writing station.Such as, following output record, in two hours of 8 o'clock to 10 o'clock morning of on November 13rd, 2012, there is an accompanying relationship at monitoring point JNC88888, respective course length is 12 and 15 to embody two cars (capital 888888 and capital 999999).
< capital 888888, capital 999999, <2012-11-1308:00:00,2012-11-1310:00:00>, JNC88888,12,15>
S4 is the result after the track similarity determination that experienced by third level operation, have recorded the two cars with similar track.Device, by similarity right for the different vehicle of each monitoring point of calculating, will meet the result write data memory module of judgment threshold.Such as, following output record, in two hours of 8 o'clock to 10 o'clock morning of on November 13rd, 2012, putting adjoint number of times is 15, and track similarity degree is 88% to embody two cars (capital 888888 and capital 999999).
< capital 888888, capital 999999, <2012-11-1308:00:00,2012-11-1310:00:00>, 0.88,15>.
Embodiment 3
Second level operation organizes vehicle to cross blockchain table as follows.Assuming that the license plate identification data that certain monitoring point obtains, time span was from 0 o'clock to 24 o'clock on the 13rd November in 2012.These data read in the track chained list set after first order Activity Calculation screening removal redundancy, the method of the invention extracts monitoring point in every bar data, license plate number, the attribute items such as writing time, sort by the time order and function through monitoring point, be formed in the blockchain table excessively in described time range for each monitoring point.Certain monitoring point list structure described is as shown in following output record.
< monitoring point ID< car plate 1, time 1, course length 1; Car plate 2, time 2, course length 2; ...; Car plate n, time n, course length n>>
Embodiment 4
The workflow management point companion pressed shown in Fig. 3 is done in the second level.Wherein, if a pair vehicle i and j at the appointed time in scope has an accompanying relationship, now using license plate number right for escort vehicle, separately course length, write described S3 with time range as intermediate result collection.
Embodiment 5
Concrete, the result of second level operation is added up by monitoring point by third level operation, calculate all vehicles right a little with the time range that number of times is similar with two wheel paths; The track similarity of described two cars can be calculated subsequently, all results are write described S4.
Finally should be noted that above only in order to technical scheme of the present invention to be described and unrestricted.Although to invention has been detailed description, those of ordinary skill in the art is to be understood that, modify to technical scheme of the present invention or equivalent replacement, do not depart from the spirit and scope of technical solution of the present invention, it all should be encompassed in the middle of right of the present invention.
Claims (15)
1., based on a similar track method for digging for magnanimity license plate identification data, it is characterized in that, described method comprises:
(1) track tissue and screening step, for removing invalid redundant data;
(2) with calculation procedure, for Maintenance Intermediate Point spinoff collection;
(3) track similarity determination, for adding up and calculating track of vehicle similarity, obtains the escort vehicle pair with similarity track.
2. the similar track method for digging based on magnanimity license plate identification data according to claim 1, it is characterized in that, wherein said track tissue and screening step, utilize the vehicle identification data related in a kind of track list structure stores processor process, this step is completed by the first order of three class pipeline, is a MapReduce computational tasks, input magnanimity license plate identification data collection, Output rusults is track chained list set 1, and passes to second level streamline use.
3. the similar track method for digging based on magnanimity license plate identification data according to claim 1, is characterized in that, described point with calculation procedure, for obtaining intermediate point spinoff collection; The result of calculation of the first order of the track chained list set 1 described in this step utilizes, being completed by the second level of three class pipeline, is the realization of another MapReduce operation, the track list structure 1 described in input, Output rusults is track chained list set 2, and passes to the use of third level streamline.
4. the similar track method for digging based on magnanimity license plate identification data according to claim 1, is characterized in that, described track similarity determination step, for calculating track of vehicle similarity, obtains the escort vehicle pair with similar track; The result of calculation of the second level of the track chained list set 2 described in this step utilizes, being completed by the third level of three class pipeline, is the realization of another MapReduce operation, the track list structure 2 described in input, Output rusults is track chained list set 3, and final writing in files system or database.
5. the similar track method for digging based on magnanimity license plate identification data according to claim 1, is characterized in that, described magnanimity license plate identification data collection L represents, refers to all vehicle information data that on tested road network, each monitoring point is caught; Every bar license plate identification data l ∈ L can be expressed as
wherein v
irepresent the number-plate number (can unique identification vehicle),
represent vehicle v
ithrough monitoring point n
k; Further,
wherein
represent the monitoring point n of vehicle process
kgeographic position,
represent that vehicle is through monitoring point n
ktime.
6. the similar track method for digging based on magnanimity license plate identification data according to claim 1, is characterized in that, described track of vehicle t
irepresenting, is vehicle v
ione group of monitoring point sequence of process in chronological order in a time range; Further, t
ican be expressed as:
wherein, to any p<q, have
t
iin the monitoring point number that comprises be called and the length of track be designated as l
i.
7. the similar track method for digging based on magnanimity license plate identification data according to claim 3, is characterized in that, the described adjoint sim of point
n(v
i, v
j) represent, refer to two vehicle v
iand v
jat certain hour threshold value δ
tinterior priority is through certain monitoring point n
pand its a kind of relation met the following conditions:
only may to exist in this monitoring point through two of same monitoring point vehicles within the scope of certain hour and once put accompanying relationship.
8. the similar track method for digging based on magnanimity license plate identification data according to claim 4, is characterized in that, judge similar track by track similarity; Track similarity refers to the similarity degree of two tracks of vehicle, with simD (t
i, t
j) represent; Further,
wherein l
iand l
jbe respectively vehicle i and vehicle j course length, m is that described two cars is by way of the monitoring point number with an accompanying relationship.
9. the similar track method for digging based on magnanimity license plate identification data according to claim 4, is characterized in that, described track of vehicle t
iwith track of vehicle t
jfor similar track, refer to given trace similarity threshold δ
d, course length threshold value δ
lwith in time range dur, t
iand t
jthe track pair simultaneously meeting following two conditions:
(1) track t
iand t
jsimilarity simD (t
i, t
j)>=δ
d;
(2) track t
iand t
jcourse length l
i>=δ
l, l
j>=δ
l.
10. the similar track method for digging based on magnanimity license plate identification data according to claim 1, it is characterized in that, the calculating of similar track is obtained in described method, can according to the description of claim 5-9, the threshold value of setting different parameters is for limiting design conditions, avoid shorter track of vehicle as the erroneous judgement of similar track, and invalid data is filtered; Specific as follows:
Suppose that set point is with time threshold δ
t, track similarity threshold δ
dwith course length threshold value δ
l, utilize existing license plate identification data collection L, find out all similar track S set T of vehicle meeting described definition in given time range dur;
Concrete,
The 11. similar track method for digging based on magnanimity license plate identification data according to claim 2, it is characterized in that, described first order streamline MapReduce operation process, the carrying out track of vehicle length being less than to given trace length threshold is deleted, and sets up the track chained list of all vehicles further; Concrete, following two steps can be decomposed into further:
(1) Map task burst from file system reads license plate identification data, obtains the license plate identification data of scope dur preset time; Described license plate identification data presses monitoring time dividing data burst; Described Map task license plate identification data record is converted to license plate number be key, the data item that is value with time and monitoring point; The data item of same keys will be sent to same Reduce task;
(2) license plate identification data is organized as track of vehicle chained list by license plate number by Reduce task, forms the track in described time range dur; Described Reduce task judges length to each track of vehicle chained list, deletes and is less than course length threshold value δ
lchained list, all the other qualified tracks are exported as described track chained list set 1.
The 12. similar track method for digging based on magnanimity license plate identification data according to claim 3, is characterized in that, described second level streamline MapReduce operation process can be divided into following two steps:
(1) Map task reads the result of first order operation, and to be converted to monitoring point be key, monitoring time and license plate number is the data item of value, passes to Reduce task; The data item of same keys will be sent to same Reduce task;
(2) groups of data items of same monitoring point is woven to a blockchain table excessively by Reduce task, forms several described blockchain table excessively; Described Reduce task according to monitoring time successively sequence and calculation level adjoint, export the result data item meeting an accompanying relationship; Described result data item comprises the license plate number of two cars, adjoint time range and respective course length.
The 13. similar track method for digging based on magnanimity license plate identification data according to claim 12, is characterized in that, described point, with calculating, carries out as follows:
(1) obtain untreated blockchain table excessively, from gauge outfit, obtain data item 1;
(2) if data item 1 exists the follow-up data item 2 do not scanned, flag data item 2 is for scan, judge whether the difference of time attribute contained by two data item 1 and data item 2 is less than time threshold: if meet threshold value, the license plate number 1 that output data item 1 comprises and the license plate number 2 that data item 2 comprises are combined as key, with time range for being worth, turn (2); If do not meet threshold value, turn (2);
(3) if data item 1 does not exist do not scan follow-up, if data item 1 be not described chained list last, then the direct follow-up data item of data item 1 is labeled as data item 1, turns (2); If data item 1 be described chained list last, then mark this chained list processed, turn (1).
The 14. similar track method for digging based on magnanimity license plate identification data according to claim 4, is characterized in that, described third level streamline MapReduce operation process can be divided into following two steps:
(1) Map task reads the result that second level streamline MapReduce operation exports, and license plate number combination identical data will be sent to the process of same Reduce task;
(2) Reduce task counts license plate number combination, also namely forms the some accompanying relationship counting of two; Calculate the track similarity of described two cars, export the result data item meeting similarity threshold; Described result data item, with the car plate number key of described two cars, with adjoint time range, the adjoint number of times of point and track similarity for value.
15. 1 kinds, based on the similar track excavating gear of magnanimity license plate identification data, is characterized in that, comprise as lower component:
(1) data memory module: this module divides module with digital independent, deblocking computing module is connected with similar trajectory computation module, adopts distributed file system framework; This module deposits magnanimity license plate identification data, the intermediate data of operation excavates net result with result, road network monitoring point basic data with similar track;
(2) digital independent divides module: this module is connected with data memory module, deblocking computing module, for track tissue and screening; This module reads the magnanimity license plate identification data in data memory module, divides and removes invalid redundant data, the data subset of each piecemeal met the demands is passed to deblocking computing module and calculates by monitoring point;
(3) deblocking computing module: this module divides module with data memory module, digital independent and is connected, receives the block data subset that digital independent divides module, for putting with calculating; Result of calculation is stored in the intermediate result collection of data memory module;
(4) similar trajectory computation module: this module is connected with data memory module, Man Machine Interface, for track Similarity measures and judgement; The parameter that this module is transmitted according to Man Machine Interface, reads the intermediate result collection of deblocking computing module stored in data memory module, calculates the vehicle pair obtaining and have similar track, and result is returned Man Machine Interface;
(5) Man Machine Interface: this module and similar track are calculated module and be connected, this module provides interactive interface for user, and support that user inputs calculating parameter, described parameter comprises course length threshold value, the adjoint time threshold of point, similarity threshold and time range; Described parameter will pass to similar trajectory computation module, and described Man Machine Interface receives the result of calculation of described similar trajectory computation module, and by the map of this result of calculation in described interactive interface for user presents.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510167058.XA CN104778245B (en) | 2015-04-09 | 2015-04-09 | Similar track method for digging and device based on magnanimity license plate identification data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510167058.XA CN104778245B (en) | 2015-04-09 | 2015-04-09 | Similar track method for digging and device based on magnanimity license plate identification data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104778245A true CN104778245A (en) | 2015-07-15 |
| CN104778245B CN104778245B (en) | 2018-11-27 |
Family
ID=53619709
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510167058.XA Active CN104778245B (en) | 2015-04-09 | 2015-04-09 | Similar track method for digging and device based on magnanimity license plate identification data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104778245B (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105225494A (en) * | 2015-11-03 | 2016-01-06 | 中兴软创科技股份有限公司 | Based on the Vehicle tracing method and apparatus of electronic police data |
| CN105261218A (en) * | 2015-10-27 | 2016-01-20 | 杭州电子科技大学 | Floating car accompany behavior mode digging method based on large data analysis |
| CN105912709A (en) * | 2016-04-28 | 2016-08-31 | 泰华智慧产业集团股份有限公司 | Big data-based accompanying vehicle analysis method and system |
| CN106528865A (en) * | 2016-12-02 | 2017-03-22 | 航天科工智慧产业发展有限公司 | Quick and accurate cleaning method of traffic big data |
| CN106611499A (en) * | 2015-10-21 | 2017-05-03 | 北京计算机技术及应用研究所 | Method of detecting vehicle hotspot path |
| WO2017162084A1 (en) * | 2016-03-25 | 2017-09-28 | 阿里巴巴集团控股有限公司 | Method and device for analyzing data similarity |
| CN108564788A (en) * | 2018-06-07 | 2018-09-21 | 重庆邮电大学 | A kind of colleague's vehicle discovery method based on streaming big data |
| CN108764197A (en) * | 2018-06-06 | 2018-11-06 | 中兴智能交通股份有限公司 | With vehicle identification method, device, terminal and computer readable storage medium |
| CN108984619A (en) * | 2018-06-13 | 2018-12-11 | 东莞市盟大塑化科技有限公司 | A method of the self-defined report based on big data |
| CN109118766A (en) * | 2018-09-04 | 2019-01-01 | 华南师范大学 | A kind of colleague's vehicle discriminating method and device based on traffic block port |
| CN109711385A (en) * | 2019-01-09 | 2019-05-03 | 宽凳（北京）科技有限公司 | A kind of Lane detection method, apparatus, equipment and storage medium |
| CN109785614A (en) * | 2018-12-17 | 2019-05-21 | 北京掌行通信息技术有限公司 | A kind of monitoring method and device of magnanimity mobile position data |
| CN109918395A (en) * | 2019-02-19 | 2019-06-21 | 北京明略软件系统有限公司 | One kind of groups method for digging and device |
| CN109947793A (en) * | 2019-03-20 | 2019-06-28 | 深圳市北斗智能科技有限公司 | Analysis method, device and the storage medium of accompanying relationship |
| CN110889422A (en) * | 2018-09-10 | 2020-03-17 | 百度在线网络技术（北京）有限公司 | Method, device and equipment for judging vehicles in same driving and computer readable medium |
| CN111291776A (en) * | 2018-12-07 | 2020-06-16 | 北方工业大学 | Channel information extraction method based on crowd-sourced trajectory data |
| CN111444294A (en) * | 2019-01-17 | 2020-07-24 | 杭州海康威视系统技术有限公司 | Track completion method and device and electronic equipment |
| CN112040413A (en) * | 2020-08-06 | 2020-12-04 | 杭州数梦工场科技有限公司 | User track calculation method and device and electronic equipment |
| CN112328649A (en) * | 2020-10-09 | 2021-02-05 | 福建亿榕信息技术有限公司 | Multi-track data similarity calculation method and storage device |
| CN112395277A (en) * | 2020-12-09 | 2021-02-23 | 招商华软信息有限公司 | Vehicle information screening method, device, equipment and storage medium |
| CN112766267A (en) * | 2021-02-24 | 2021-05-07 | 人机互联网络(深圳)有限公司 | Diversified automobile service O2O system based on deep operation management |
| CN113723316A (en) * | 2021-09-01 | 2021-11-30 | 杭州智诚惠通科技有限公司 | Vehicle identification method, device, equipment and storage medium |
| CN115840857A (en) * | 2023-02-22 | 2023-03-24 | 昆明理工大学 | Group behavior pattern mining method combining multivariate space-time trajectories |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103237045A (en) * | 2013-02-22 | 2013-08-07 | 北方工业大学 | Parallel processing system and parallel processing method for large-scale real-time traffic data |
| JP2013235515A (en) * | 2012-05-10 | 2013-11-21 | Nippon Telegr & Teleph Corp <Ntt> | Data distribution management system, apparatus, method and program |
| CN103646541A (en) * | 2013-12-16 | 2014-03-19 | 电子科技大学 | Vehicle congestion degree acquiring method based on Hadoop |
| CN104200669A (en) * | 2014-08-18 | 2014-12-10 | 华南理工大学 | Fake-licensed car recognition method and system based on Hadoop |
-
2015
- 2015-04-09 CN CN201510167058.XA patent/CN104778245B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013235515A (en) * | 2012-05-10 | 2013-11-21 | Nippon Telegr & Teleph Corp <Ntt> | Data distribution management system, apparatus, method and program |
| CN103237045A (en) * | 2013-02-22 | 2013-08-07 | 北方工业大学 | Parallel processing system and parallel processing method for large-scale real-time traffic data |
| CN103646541A (en) * | 2013-12-16 | 2014-03-19 | 电子科技大学 | Vehicle congestion degree acquiring method based on Hadoop |
| CN104200669A (en) * | 2014-08-18 | 2014-12-10 | 华南理工大学 | Fake-licensed car recognition method and system based on Hadoop |
Non-Patent Citations (2)
| Title |
|---|
| 卢帅: "一种车辆移动对象相似轨迹查询算法", 《计算机与数字工程》 * |
| 郑苏杭: "面向海量交通信息流的分布式序列模式挖掘研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106611499A (en) * | 2015-10-21 | 2017-05-03 | 北京计算机技术及应用研究所 | Method of detecting vehicle hotspot path |
| CN105261218A (en) * | 2015-10-27 | 2016-01-20 | 杭州电子科技大学 | Floating car accompany behavior mode digging method based on large data analysis |
| CN105225494A (en) * | 2015-11-03 | 2016-01-06 | 中兴软创科技股份有限公司 | Based on the Vehicle tracing method and apparatus of electronic police data |
| WO2017162084A1 (en) * | 2016-03-25 | 2017-09-28 | 阿里巴巴集团控股有限公司 | Method and device for analyzing data similarity |
| CN107229940A (en) * | 2016-03-25 | 2017-10-03 | 阿里巴巴集团控股有限公司 | Data adjoint analysis method and device |
| CN105912709A (en) * | 2016-04-28 | 2016-08-31 | 泰华智慧产业集团股份有限公司 | Big data-based accompanying vehicle analysis method and system |
| CN106528865A (en) * | 2016-12-02 | 2017-03-22 | 航天科工智慧产业发展有限公司 | Quick and accurate cleaning method of traffic big data |
| CN108764197A (en) * | 2018-06-06 | 2018-11-06 | 中兴智能交通股份有限公司 | With vehicle identification method, device, terminal and computer readable storage medium |
| CN108564788A (en) * | 2018-06-07 | 2018-09-21 | 重庆邮电大学 | A kind of colleague's vehicle discovery method based on streaming big data |
| CN108984619A (en) * | 2018-06-13 | 2018-12-11 | 东莞市盟大塑化科技有限公司 | A method of the self-defined report based on big data |
| CN109118766A (en) * | 2018-09-04 | 2019-01-01 | 华南师范大学 | A kind of colleague's vehicle discriminating method and device based on traffic block port |
| CN110889422A (en) * | 2018-09-10 | 2020-03-17 | 百度在线网络技术（北京）有限公司 | Method, device and equipment for judging vehicles in same driving and computer readable medium |
| CN111291776B (en) * | 2018-12-07 | 2023-06-02 | 北方工业大学 | Channel information extraction method based on crowd source track data |
| CN111291776A (en) * | 2018-12-07 | 2020-06-16 | 北方工业大学 | Channel information extraction method based on crowd-sourced trajectory data |
| CN109785614A (en) * | 2018-12-17 | 2019-05-21 | 北京掌行通信息技术有限公司 | A kind of monitoring method and device of magnanimity mobile position data |
| CN109711385A (en) * | 2019-01-09 | 2019-05-03 | 宽凳（北京）科技有限公司 | A kind of Lane detection method, apparatus, equipment and storage medium |
| CN111444294A (en) * | 2019-01-17 | 2020-07-24 | 杭州海康威视系统技术有限公司 | Track completion method and device and electronic equipment |
| CN111444294B (en) * | 2019-01-17 | 2023-10-10 | 杭州海康威视系统技术有限公司 | Track complement method and device and electronic equipment |
| CN109918395A (en) * | 2019-02-19 | 2019-06-21 | 北京明略软件系统有限公司 | One kind of groups method for digging and device |
| CN109947793A (en) * | 2019-03-20 | 2019-06-28 | 深圳市北斗智能科技有限公司 | Analysis method, device and the storage medium of accompanying relationship |
| CN112040413A (en) * | 2020-08-06 | 2020-12-04 | 杭州数梦工场科技有限公司 | User track calculation method and device and electronic equipment |
| CN112328649A (en) * | 2020-10-09 | 2021-02-05 | 福建亿榕信息技术有限公司 | Multi-track data similarity calculation method and storage device |
| CN112395277A (en) * | 2020-12-09 | 2021-02-23 | 招商华软信息有限公司 | Vehicle information screening method, device, equipment and storage medium |
| CN112395277B (en) * | 2020-12-09 | 2024-01-09 | 招商华软信息有限公司 | Vehicle information screening method, device, equipment and storage medium |
| CN112766267A (en) * | 2021-02-24 | 2021-05-07 | 人机互联网络(深圳)有限公司 | Diversified automobile service O2O system based on deep operation management |
| CN113723316A (en) * | 2021-09-01 | 2021-11-30 | 杭州智诚惠通科技有限公司 | Vehicle identification method, device, equipment and storage medium |
| CN113723316B (en) * | 2021-09-01 | 2024-04-16 | 杭州智诚惠通科技有限公司 | Vehicle identification method, device, equipment and storage medium |
| CN115840857A (en) * | 2023-02-22 | 2023-03-24 | 昆明理工大学 | Group behavior pattern mining method combining multivariate space-time trajectories |
| CN115840857B (en) * | 2023-02-22 | 2023-05-09 | 昆明理工大学 | Group behavior pattern mining method combining multiple space-time tracks |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104778245B (en) | 2018-11-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104778245A (en) | Similar trajectory mining method and device on basis of massive license plate identification data | |
| Mostafa et al. | Review of image-based analysis and applications in construction | |
| CN109448370B (en) | Traffic control subarea division method based on vehicle track data | |
| CN102799897B (en) | Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling | |
| CN108416985B (en) | Geological disaster monitoring and early warning system and method based on image recognition | |
| Kumar et al. | Application of land change modeler for prediction of future land use land cover: a case study of Vijayawada City | |
| CN114428828A (en) | Method and device for digging new road based on driving track and electronic equipment | |
| CN106649809A (en) | Internet of Things analysis method | |
| Elghaish et al. | Deep learning for detecting distresses in buildings and pavements: a critical gap analysis | |
| CN108257385B (en) | Method for discriminating abnormal events based on public transportation | |
| Opara et al. | Defect detection on asphalt pavement by deep learning | |
| CN114120428A (en) | Graph code joint detection correlation method and device, computer equipment and storage medium | |
| Wang et al. | Construction and application of artificial intelligence crowdsourcing map based on multi-track GPS data | |
| Khan et al. | Short-term traffic prediction using deep learning long short-term memory: Taxonomy, applications, challenges, and future trends | |
| CN111047860A (en) | Vehicle running track extraction method | |
| Benallal et al. | An image-based convolutional neural network system for road defects detection | |
| CN106355537A (en) | Smart analysis method and system for interrelated cases | |
| Li et al. | Learning to holistically detect bridges from large-size vhr remote sensing imagery | |
| CN114780591B (en) | Calculation method and system for detecting travel license plate recognition errors | |
| Zhao et al. | A localization method for stagnant water in city road traffic image | |
| CN112365495A (en) | Method for identifying illegal building based on remote sensing | |
| Chen et al. | Self-intersection attention pooling based classification for rock recognition | |
| CN111177294A (en) | Method for solving intersection intermediary centrality based on vehicle track data | |
| Liu et al. | YOLO-Based Efficient Vehicle Object Detection | |
| Dekker et al. | Point Cloud Analysis of Railway Infrastructure: A Systematic Literature Review |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |