CN109377757A - The vehicle driving track extraction method of license plate identification data based on the rough error containing multi-source - Google Patents
The vehicle driving track extraction method of license plate identification data based on the rough error containing multi-source Download PDFInfo
- Publication number
- CN109377757A CN109377757A CN201811363220.5A CN201811363220A CN109377757A CN 109377757 A CN109377757 A CN 109377757A CN 201811363220 A CN201811363220 A CN 201811363220A CN 109377757 A CN109377757 A CN 109377757A
- Authority
- CN
- China
- Prior art keywords
- point
- license plate
- identification data
- plate identification
- vehicle driving
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
- G08G1/0112—Measuring and analyzing of parameters relative to traffic conditions based on the source of data from the vehicle, e.g. floating car data [FCD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses the vehicle driving track extraction methods of the license plate identification data based on the rough error containing multi-source, comprising steps of the first step, the license plate identification data of the network interface card port system that satisfies the need capture carries out Detection of Gross Errors and rejecting;The Detection of Gross Errors and rejecting further comprise: identifying and reject discrete point and jump, the rough error point verifying based on running speed value and the rough error point verifying based on running distance value;Second step extracts vehicle driving track based on the license plate identification data after elimination of rough difference.The method of the present invention can accurately and effectively calculate vehicle driving track, be applicable in and be related to the excavation and analysis of traffic big data, can be well used in fields such as city road network traffic statistics, odd-and-even license plate rule, traffic monitorings.
Description
Technical field
The invention belongs to Traffic Systems operations, planning application field relevant to management etc., and in particular to Yi Zhongji
In the vehicle driving track extraction method of the license plate identification data of the rough error containing multi-source.
Background technique
With sharply increasing for Urban vehicles poputation, automobile has become the preferred vehicles of people's trip, vehicle
Trip track because it implies traffic trip abundant and status information, and space division when becoming research urban traffic flow therewith
The basis of cloth characteristic, the history planning driving path for how extracting vehicle already become instantly popular research topic.It is traditional based on
Not only time-consuming for the method for manual research, at high cost, the result for needing to put into a large amount of human and material resources and financial resources, and obtaining
With actual conditions are uncertain is consistent completely, can also cause certain interference to the operation of road grid traffic sometimes.Then there is research
Scholar attempts to be equipped with real time positioning data caused by the vehicle of vehicle-mounted GPS apparatus to be fitted the driving path of vehicle.But
For the car ownership huge for city, the vehicle specific gravity equipped with GPS device is relatively small, can not accomplish to each vehicle
Real-time tracing.Therefore, it is necessary to find a kind of more efficiently information collection and analysis means to obtain real-time, accurate row
Car data.
The bayonet system of road network is a kind of traffic that the motor vehicles to by bayonet position are shot, recorded and handled
Monitoring system.Compared with GPS gathers data, the bayonet system of road network can not only round-the-clock, a wide range of entire metropolitan district of detection
In domain road network vehicle by situation, and the real-time property being collected into is good, acquisition rate is high.But similarly there is license plates
There are the problems such as error in identification error rate height, bayonet time disunity, bayonet position.Currently, each city bayonet is layouted
The accuracy of rate and equipment is irregular, needs the actual conditions for each city to formulate a set of effective method
To make good use of these license plate identification datas to the greatest extent, and then provide for traffic department's formulation decision-making policy more structurally sound
Data are supported.
Summary of the invention
The vehicle driving trajectory extraction of the object of the present invention is to provide a kind of license plate identification data based on the rough error containing multi-source
Method.
In order to achieve the above objectives, the vehicle driving track of the license plate identification data provided by the invention based on the rough error containing multi-source
Extracting method, comprising steps of
The license plate identification data of the first step, the network interface card port system that satisfies the need capture carries out Detection of Gross Errors and rejecting;
The Detection of Gross Errors and rejecting further comprise:
(1.1) it identifies and rejects discrete point and jump, specifically:
(1.1a) constructs bayonet website adjacency list using the spatial relation between each bayonet website in road network;
(1.1b) obtains the point that same vehicle single trip is passed through using SQL statement from license plate identification data, will be obtained
Point is obtained the point collection of the vehicle secondary trip by the vehicle by time-sequencing;
(1.1c) combines adjacency list, judges that point concentrates whether each point is discrete point or jump one by one;
The discrete point, which refers to, is much larger than S with the reach distance of its front and back point interdigit0Point, S0For with place city consecutive points
Interdigit is apart from relevant distance threshold, distance threshold S0Setting be for reject with consecutive points interdigit distance completely it is unreasonable
Discrete type point;The jump refers to and the upper non-conterminous point in adjacent but spatial position of previous point time;
(1.1d) rejects discrete point and the corresponding license plate identification data of jump;
(1.2) the rough error point verifying based on running distance value, specifically:
Traversal eliminate the point collection of discrete point and jump, calculate current point and its front and back point interdigit reach distance compared with
The most short reach distance S ' of small value S and its front and back point interdigit, when S is much larger than S ', then current point is rough error point, rejects it
Corresponding license plate identification data;
(1.3) the rough error point verifying based on running speed value, specifically:
The point collection of running distance value verifying is traversed through, obtains the driving speed that vehicle concentrates adjacent two o'clock interdigit in point one by one
Angle value, when running speed value is unreasonable, then latter point is rough error point in adjacent two point, rejects its corresponding Car license recognition number
According to;
Second step extracts vehicle driving track based on the license plate identification data after elimination of rough difference, specifically:
(2.1) it to the point collection after the first step carries out elimination of rough difference, obtains point and concentrates the most short of every adjacent two o'clock interdigit
Path, the overlay tree of all shortest path compositions vehicle driving track;
(2.2) longest path, i.e. vehicle driving track are calculated to overlay tree using path computation algorithm.
Further, judge one kind of discrete point method particularly includes:
One empirical value N is set, judges whether point and the reach distance of its front and back two o'clock interdigit are all larger than N* S0If being all larger than
N* S0, then the point is discrete point;It is on the contrary then be not;
N value need to consider in road network distance between set bayonet, and be debugged repeatedly to obtain desired value.
Further, in sub-step (1.2), judge that running speed is worth unreasonable one kind method particularly includes:
Consider the practical driving situation in vehicle driving speed limit and city, a running speed threshold value is set, and vehicle is adjacent two
The running speed value of point interdigit is greater than running speed threshold value, then it is assumed that running speed value is unreasonable.
Further, in sub-step (1.3), judge whether S is much larger than one kind of S ' method particularly includes:
One empirical value N ' is set and judges formula S > N ' * S ', when meeting judgement formula, then it is assumed that S is much larger than S ';
N ' value need to consider the practical trip situation and road conditions in city, and be debugged repeatedly to obtain desired value.
Further, in sub-step (2.1), point concentrates the shortest path per adjacent two o'clock interdigit to obtain with the following method
:
Judge whether section where adjacent two point is adjacent segments, if adjacent segments, is directly taken between adjacent two point
Path is as shortest path;Otherwise, shortest path is calculated using path computation algorithm.
Further, in sub-step (2.2), longest path is calculated to overlay tree using path computation algorithm, specifically:
A road section s is arbitrarily selected from overlay tree, from section s, longest path is asked using path computation algorithm, by the longest
Section where another endpoint in path is denoted as s ';From ' the s of section, longest path, i.e. vehicle are asked using path computation algorithm again
Trip track.
The path computation algorithm is dijkstra's algorithm or A* algorithm.
The present invention have following features and the utility model has the advantages that
It is special the invention proposes a kind of method that license plate identification data using the rough error containing multi-source extracts city vehicle trip track
Not in the excavation and analysis for being related to traffic big data, the history of city vehicle can be more effectively calculated using the method for the present invention
Trip track, calculated history trip track can not only reproduce to comprehensive system complicated traffic circulation scene, can also be
The investigation and update work of basic OD matrix provide thinking and effective technological means, further can also formulate phase for traffic department
Decision, the regulation of pass provide more structurally sound data supporting.
It should be noted that OD represents traffic start-stop point, an OD matrix described in text i.e. two-dimensional table, element generation in table
Traffic trip amount between table start-stop point.OD matrix is also referred to as OD table, OD survey result.
Detailed description of the invention
Fig. 1 is adjacency list schematic diagram in embodiment 1, wherein figure (a) show the letter of position between road network and bayonet website
Single schematic diagram, figure (b) are the adjacency list structures established based on figure (a);
Fig. 2 is the overlay tree of Vehicle emission track calculated in embodiment 1;
Fig. 3 is in embodiment 1 based on overlay tree Vehicle emission track calculated;
Fig. 4 ~ 5 are the overviews of involved project in embodiment 2, wherein Fig. 4 top half is that the vehicle in current survey region is logical
Market condition, mainly based on traffic above-ground, in conjunction with Fig. 5 it can be found that ground shares 9 lanes at present (comprising 2 bus zones)
For motor-driven vehicle going;Improved section becomes 4 lane of ground (comprising 2 bus zones) and 4 lane of underground, is specifically shown in figure
Shown in the lower half portion of 4 and Fig. 5.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Embodiment 1
The present invention carries out rough error spy firstly the need of to initial data (i.e. the original license plate identification data of road network bayonet system capture)
It surveys and rejects;Then, based on the license plate identification data after elimination of rough difference, the trip of city vehicle is extracted in conjunction with path computation algorithm
Track.
The present embodiment will be described in detail in conjunction with principle of the attached drawing to specific implementation process of the present invention.
The first step, the Detection of Gross Errors of license plate identification data and rejecting.
The initial data that the present invention utilizes is the license plate identification data of road network bayonet system capture, wherein comprising largely because of visitor
See multi-source rough error data caused by reason.For the accuracy for ensuring Vehicle emission trajectory extraction, it should identify and reject original
Rough error data in data.
The detection of rough error data includes following two parts with the principle rejected:
(1) it based on the spatial topotaxy between the spatial relation and bayonet website between bayonet website and road network, rejects not
Reasonable license plate identification data.
Specifically, the space topological between the spatial relation and bayonet website first based on bayonet website and road network
Relationship constructs the adjacency list of each bayonet website in road network, as shown in Figure 1.In Fig. 1 (a), A, B, C, D, E, F respectively indicate difference
Bayonet website, be abbreviated as point.According to common sense, it is assumed that current time license plate is present in E, then subsequent time vehicle
Board can only appear at adjacent point D, B, F of E;If appearing in other points, then it is assumed that point where subsequent time is rough error
Point, the license plate identification data of point is rough error data where the subsequent time, should be rejected.Based on Fig. 1 (a) structure figures 1(b), tool
For body, first to A, B, C, D, E, F number consecutively in Fig. 1 (a) be 1,2,3,4,5,6, then respectively with A, B, C, D, E, F be top
Point, such as using E as vertex, then certain license plate of current time present in E, subsequent time can only appear at D, B, F, then may be used
It is expressed as the 5th row of Fig. 1 (b).
The point that same vehicle single trip is passed through is obtained using SQL query statement, obtained point is passed through by the vehicle
Cross the point collection that time-sequencing obtains the vehicle secondary trip.Each point is concentrated to identify one by one point in conjunction with the adjacency list of building
It whether is rough error point.Rough error point further comprises discrete point and jump in the present invention.
Discrete point refers to the point isolated with other points, and the corresponding license plate identification data in discrete point place is rough error data,
It should reject.Whether current point is discrete point, can be judged according to the distance between current point and its former and later two point, point
Concentrate former and later two points, that is, current point front and back moment point of current point in position.Assuming that current point is B, former and later two
Point is A and C, and A, B, C are then sequentially continuous three points on the time.If SAB>> S0And SBC>> S0, then point B be from
Scatterplot.Wherein, SABIndicate the reach distance between point A and B, SBCIndicate the reach distance between point B and C, S0For with institute
In city, consecutive points interdigit is apart from relevant distance threshold, distance threshold S0Setting be for rejecting and consecutive points interdigit distance
Completely unreasonable discrete point.S0A kind of specific obtaining value method are as follows: be taken as effective point concentrate between adjacent two point away from
From average value, effective point collection be by being verified as the set that correct point is constituted through the method for the present invention, here, correct point
It is not point that discrete point is also not jump that position, which refers to,.First point is concentrated for point, first assumes that it is correct point, benefit
Second point is judged with the proof rule based on running distance value and based on running speed value, if second point
It is judged as rough error point, then denies the correctness of first point;Otherwise, first point is correct point.It accepts if not
The correctness of first point, then using second point as first point, using based on running distance value and based on driving
The proof rule of velocity amplitude judges second point, will be correct until finding first and second correct point
Point is put into effective point collection.In the detection process for carrying out discrete point and jump to point collection, by the non-discrete point detected and
Non-hopping point is also placed in effective point collection.
In the one kind for judging whether to be much larger than method particularly includes: by the way that an empirical value N is arranged, to judge SABAnd SBCWhether
Much larger than S0.Settable judgement formula SAB>N* S0And SBC>N* S0, when two judgement formulas meet, then B is judged as discrete
Point.N value is arranged according to distance between bayonet set in city road network, is substantially empirical value, needs repeatedly in practical application
Debugging is to obtain desired value.
Jump refer in time there are successive neighbouring relations, the non-conterminous point on spatial position.For example, current
Moment license plate present in E, certain license plate last moment present in C, C and E it is adjacent in time and on spatial position
It is non-adjacent, then it is jump at current time E, the corresponding license plate identification data in current time E place should be rejected.Judge jump
When, point collection need to be traversed, current point and its previous point are if it is temporally adjacent point, it is only necessary to judge whether the two is empty
Between position it is adjacent;If the two spatial position is non-conterminous, then it is assumed that current point is rough error point.
(2) since the recognition accuracy of bayonet camera is lower, a large amount of other kinds of rough errors are still had, only according to
Adjacency list is also far from enough to reject the rough error in initial data.So the present invention passes through to the Car license recognition number being continuously shot
According to corresponding proof rule is formulated, further to remove rough error data.
The proof rule is as follows:
Proof rule of the 2-1 based on running distance value:
Each point that traversal point is concentrated in chronological order, takes the previous moment of current point and later moment in time point, calculating to work as
The smaller value of reach distance between preceding point and its front and back moment point, is denoted as S;Calculate again previous moment point and it is latter when
The most short reach distance of punctum interdigit, is denoted as S ';If the minimum of S > > S ', i.e. point midway and its front and back two o'clock interdigit distance
Value there will naturally be unreasonable at this time much larger than the distance between two point of front and back, then it is assumed that current point is rough error point, is rejected thick
Not good enough corresponding license plate identification data;Otherwise it is assumed that current point is correct point.
It can be by the way that an empirical value N ' be arranged, to judge whether S is much larger than S '.Specifically, judgement formula S > N ' * is set
S ', when meeting judgement formula, then it is assumed that S is much larger than S '.N ' value need to consider the practical trip situation and road conditions in city, and through anti-
Polyphony tries to obtain desired value.
Proof rule of the 2-2 based on running speed value:
Specifically, running speed threshold value is set, and the setting of the running speed threshold value is considered as vehicle driving speed limit and city
The practical driving situation in city.Running speed value of the only same vehicle between adjacent two point is not more than running speed threshold value
When, it is reasonable to be just considered.So, the running speed value when vehicle between adjacent two point is greater than running speed threshold value, then recognizes
It is unreasonable for running speed value, later moment in time point in adjacent two point is thought into rough error point, later moment in time point is corresponding
License plate identification data reject.The present embodiment middle rolling car threshold speed is set as 70km/h.
When it is implemented, using SQL statement, look into the license plate identification data after an elimination of rough difference was carried out
Ask, obtain the point collection that same vehicle is sequentially arranged, according to point collection to same vehicle between two adjacent points
Running speed value verified.
Second step extracts vehicle driving track based on the license plate identification data after elimination of rough difference.
To each vehicle, it is based on its license plate, the trip of its single is inquired from the license plate identification data after elimination of rough difference and is passed through
The set for the point crossed, is denoted as point collection, and point concentrates each point to arrange sequentially in time.Using path computation algorithm to point
Position is concentrated carries out shortest path calculating per adjacent two point respectively.Since there are deviations, i.e. point for the point of road network bayonet system
It can not fit like a glove with road net data, all shortest paths can only form the overlay tree of vehicle driving track, see deep in Fig. 2
Black line.It should be noted that whether section where first judging two adjacent points is adjacent before carrying out shortest path calculating, if
Place section is adjacent, directly takes shortest path of the path between two point as two point;Otherwise path computing is used
Algorithm calculates longest path to overlay tree, obtains final vehicle driving track, sees pitch black line in Fig. 3.Specifically, it is calculating
When longest path, a road section s is arbitrarily selected first from overlay tree, s seeks longest path from section, and the longest path is another
Section where end point is denoted as s ', asks longest path, i.e., final Vehicle emission track again ' from section s.
In the present embodiment, dijkstra's algorithm or A* algorithm etc. is can be used in path computation algorithm.
Embodiment 2
The present embodiment is applicable cases of 1 method of embodiment in three Jiangkou business core space project of Ningbo municipal government development, is used
To verify the feasibility of modification scheme.
According to the traffic condition of current Ningbo City, intend the walking space that adjustment expands city major trunk roads ground, by ground
Face traffic is transformed into network of continuously going slowly, and border will be passed through under the traffic of main city zone arterial highway, turns main city zone to light car city
Type, improved lane distribution condition become 4 lane of ground (comprising 2 public affairs from original 9 lanes (including 2 bus zones)
Hand over lane) and 4 lane of underground, as shown in Fig. 4 ~ 5.For the feasibility for verifying the modification scheme, need to count daily by trunk
The through trip situation in section where road.
The present embodiment is extracted in the time in one week on the 7th June 1 day to 2018 June in 2018 using 1 method of embodiment,
Vehicle flowrate is carried out by the trip track of each motor vehicle of relevant road segments, and according to trip track.Through counting, Chang Chunlu
To the west of enter the street Liu Ting vehicle flowrate in, averagely there is 15.2% vehicle flowrate to be driven out to from the hilllock Ling Dong Huo Lingxi;In the same period
It is interior, from the vehicle flowrate driven into the east of the hilllock Ling Donglingxi, averagely have 18.8% vehicle flowrate from the hilllock Ma Yuan Huo Liuting (Changchun road with
West) it is driven out to, specifically it is shown in Table 1.
The vehicle flowrate of 1 relevant road segments of table
Time | 2018/6/1 | 2018/6/2 | 2018/6/3 | 2018/6/4 | 2018/6/5 | 2018/6/6 | 2018/6/7 |
Enter the vehicle number in the street Liu Ting to the west of the road of Changchun | 5961 | 8974 | 7588 | 7517 | 8087 | 8274 | 7770 |
The vehicle number that above-mentioned vehicle is driven out to from the hilllock Ling Dong Huo Lingxi | 710 | 1520 | 1197 | 1003 | 1280 | 1357 | 1257 |
Percentage | 11.9% | 16.9% | 15.8% | 13.3% | 15.8% | 16.4% | 16.2% |
The vehicle number driven into the east of the hilllock Ling Donglingxi | 10594 | 14635 | 10920 | 9706 | 13031 | 14148 | 14003 |
The vehicle number that above-mentioned vehicle is driven out to from the hilllock Ma Yuan Huo Liuting (to the west of the road of Changchun) | 1446 | 3091 | 2231 | 1947 | 2585 | 2659 | 2505 |
Percentage | 13.6% | 21.1% | 20.4% | 20.1% | 19.8% | 18.8% | 17.9% |
By table 1 it can be found that on the major trunk roads, the volume of traffic less than 20% is through trip, i.e. these vehicles are to can be used
What tunnel passed through, and the remaining volume of traffic is to need to hair traffic using ground section on the section.Therefore, the road Xia Chuan
The utilization rate of section is not high, does not play the role of road pavement traffic diverging.On the other hand, due to the reduction in ground lane,
Two lane of ground can not undertake arriving for the region and send out transport need;Moreover, after the under-traverse tunnel traffic capacity generates redundancy, meeting
Attract more through trips current using under-traverse tunnel, so that huge traffic pressure is brought at the related crossing to tunnel upstream and downstream
Power.In conclusion two lane of ground of under-traverse tunnel scheme can not undertake the public vehicles in the region to hair transport need, reduce
The accessibilities of three Jiangkou core space public vehicles.The program should give rejection.
The invention proposes the method for extracting Vehicle emission track using license plate identification data, but license plate identification datas
There are larger rough errors, and there are certain position deviations for road net data and bayonet data, and therefore, the method for the present invention is first with road network
Spatial relation building bayonet between adjacency list, to reject the discrete point and jump in original license plate identification data.Together
When, less accurately go on a journey track, i.e. overlay tree are calculated using the license plate identification data after excluding gross error;On the basis of overlay tree
On using the longest path algorithm in graph theory, obtain complete and accurate Vehicle emission track.Experiment shows the method for the present invention
It can be well used in fields such as city road network traffic statistics, odd-and-even license plate rule, traffic monitorings.
For those skilled in the art, can also be made on the basis of above description other it is various forms of variation or
Change, and these belong to true spirit and derive other variation or change still fall within the scope of the present invention.
Claims (7)
1. the vehicle driving track extraction method of the license plate identification data based on the rough error containing multi-source, characterized in that comprising steps of
The license plate identification data of the first step, the network interface card port system that satisfies the need capture carries out Detection of Gross Errors and rejecting;
The Detection of Gross Errors and rejecting further comprise:
(1.1) it identifies and rejects discrete point and jump, specifically:
(1.1a) constructs bayonet website adjacency list using the spatial relation between each bayonet website in road network;
(1.1b) obtains the point that same vehicle single trip is passed through using SQL statement from license plate identification data, will be obtained
Point is obtained the point collection of the vehicle secondary trip by the vehicle by time-sequencing;
(1.1c) combines adjacency list, judges that point concentrates whether each point is discrete point or jump one by one;
The discrete point, which refers to, is much larger than S with the reach distance of its front and back point interdigit0Point, S0For with place city consecutive points
Interdigit is apart from relevant distance threshold, distance threshold S0Setting be for reject with consecutive points interdigit distance completely it is unreasonable
Discrete type point;The jump refers to and the upper non-conterminous point in adjacent but spatial position of previous point time;
(1.1d) rejects discrete point and the corresponding license plate identification data of jump;
(1.2) the rough error point verifying based on running distance value, specifically:
Traversal eliminate the point collection of discrete point and jump, calculate current point and its front and back point interdigit reach distance compared with
The most short reach distance S ' of small value S and its front and back point interdigit, when S is much larger than S ', then current point is rough error point, rejects it
Corresponding license plate identification data;
(1.3) the rough error point verifying based on running speed value, specifically:
The point collection of running distance value verifying is traversed through, obtains the driving speed that vehicle concentrates adjacent two o'clock interdigit in point one by one
Angle value, when running speed value is unreasonable, then latter point is rough error point in adjacent two point, rejects its corresponding Car license recognition number
According to;
Second step extracts vehicle driving track based on the license plate identification data after elimination of rough difference, specifically:
(2.1) it to the point collection after the first step carries out elimination of rough difference, obtains point and concentrates the most short of every adjacent two o'clock interdigit
Path, the overlay tree of all shortest path compositions vehicle driving track;
(2.2) longest path, i.e. vehicle driving track are calculated to overlay tree using path computation algorithm.
2. the vehicle driving track extraction method of the license plate identification data based on the rough error containing multi-source as described in claim 1,
It is characterized in:
Judge one kind of discrete point method particularly includes:
One empirical value N is set, judges whether point and the reach distance of its front and back two o'clock interdigit are all larger than N* S0If being all larger than N*
S0, then the point is discrete point;It is on the contrary then be not;
N value need to consider in road network distance between set bayonet, and be debugged repeatedly to obtain desired value.
3. the vehicle driving track extraction method of the license plate identification data based on the rough error containing multi-source as described in claim 1,
It is characterized in:
In sub-step (1.2), judge that running speed is worth unreasonable one kind method particularly includes:
Consider the practical driving situation in vehicle driving speed limit and city, a running speed threshold value is set, and vehicle is adjacent two
The running speed value of point interdigit is greater than running speed threshold value, then it is assumed that running speed value is unreasonable.
4. the vehicle driving track extraction method of the license plate identification data based on the rough error containing multi-source as described in claim 1,
It is characterized in:
In sub-step (1.3), judge whether S is much larger than one kind of S ' method particularly includes:
One empirical value N ' is set and judges formula S > N ' * S ', when meeting judgement formula, then it is assumed that S is much larger than S ';
N ' value need to consider the practical trip situation and road conditions in city, and be debugged repeatedly to obtain desired value.
5. the vehicle driving track extraction method of the license plate identification data based on the rough error containing multi-source as described in claim 1,
It is characterized in:
In sub-step (2.1), point concentrates the shortest path per adjacent two o'clock interdigit to obtain with the following method:
Judge whether section where adjacent two point is adjacent segments, if adjacent segments, is directly taken between adjacent two point
Path is as shortest path;Otherwise, shortest path is calculated using path computation algorithm.
6. the vehicle driving track extraction method of the license plate identification data based on the rough error containing multi-source as described in claim 1,
It is characterized in:
In sub-step (2.2), longest path is calculated to overlay tree using path computation algorithm, specifically:
A road section s is arbitrarily selected from overlay tree, from section s, longest path is asked using path computation algorithm, by the longest
Section where another endpoint in path is denoted as s ';From ' the s of section, longest path, i.e. vehicle are asked using path computation algorithm again
Trip track.
7. such as the vehicle driving track extraction method of the license plate identification data described in claim 5 or 6 based on the rough error containing multi-source,
It is characterized in that:
The path computation algorithm is dijkstra's algorithm or A* algorithm.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811363220.5A CN109377757A (en) | 2018-11-16 | 2018-11-16 | The vehicle driving track extraction method of license plate identification data based on the rough error containing multi-source |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811363220.5A CN109377757A (en) | 2018-11-16 | 2018-11-16 | The vehicle driving track extraction method of license plate identification data based on the rough error containing multi-source |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109377757A true CN109377757A (en) | 2019-02-22 |
Family
ID=65389219
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811363220.5A Withdrawn CN109377757A (en) | 2018-11-16 | 2018-11-16 | The vehicle driving track extraction method of license plate identification data based on the rough error containing multi-source |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109377757A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110033051A (en) * | 2019-04-18 | 2019-07-19 | 杭州电子科技大学 | A kind of trawler behavior method of discrimination based on multistep cluster |
CN110751837A (en) * | 2019-10-23 | 2020-02-04 | 招商华软信息有限公司 | Method, device and equipment for determining high-speed driving path and storage medium |
CN110796060A (en) * | 2019-10-23 | 2020-02-14 | 招商华软信息有限公司 | Method, device and equipment for determining high-speed driving route and storage medium |
CN112309118A (en) * | 2020-11-03 | 2021-02-02 | 广州市交通规划研究院 | Vehicle trajectory calculation method based on space-time similarity |
CN112748452A (en) * | 2020-12-11 | 2021-05-04 | 上海城市交通设计院有限公司 | GPS track cleaning method based on road network data |
CN114944083A (en) * | 2022-05-13 | 2022-08-26 | 公安部交通管理科学研究所 | Method for judging distance between running vehicle on expressway and front vehicle |
CN115790640A (en) * | 2023-02-02 | 2023-03-14 | 广州斯沃德科技有限公司 | Vehicle mileage correction method and device based on multidimensional data intelligent analysis |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06323862A (en) * | 1993-05-12 | 1994-11-25 | Alpine Electron Inc | Mobile navigation system |
KR100690279B1 (en) * | 2005-09-12 | 2007-03-09 | 주식회사 리트코 | Multipurpose video image detection system |
CN101958046A (en) * | 2010-09-26 | 2011-01-26 | 隋亚刚 | Vehicle track recognition system and method |
CN102542789A (en) * | 2010-12-28 | 2012-07-04 | 财团法人工业技术研究院 | Driving path reconstruction method, system and computer program product |
CN105913668A (en) * | 2016-07-04 | 2016-08-31 | 中国电子科技集团公司第二十八研究所 | Directional fake-licensed car detection method based on vast traffic data statistics |
CN107240264A (en) * | 2017-06-26 | 2017-10-10 | 安徽四创电子股份有限公司 | A kind of non-effective driving trace recognition methods of vehicle and urban road facility planing method |
-
2018
- 2018-11-16 CN CN201811363220.5A patent/CN109377757A/en not_active Withdrawn
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06323862A (en) * | 1993-05-12 | 1994-11-25 | Alpine Electron Inc | Mobile navigation system |
KR100690279B1 (en) * | 2005-09-12 | 2007-03-09 | 주식회사 리트코 | Multipurpose video image detection system |
CN101958046A (en) * | 2010-09-26 | 2011-01-26 | 隋亚刚 | Vehicle track recognition system and method |
CN102542789A (en) * | 2010-12-28 | 2012-07-04 | 财团法人工业技术研究院 | Driving path reconstruction method, system and computer program product |
CN105913668A (en) * | 2016-07-04 | 2016-08-31 | 中国电子科技集团公司第二十八研究所 | Directional fake-licensed car detection method based on vast traffic data statistics |
CN107240264A (en) * | 2017-06-26 | 2017-10-10 | 安徽四创电子股份有限公司 | A kind of non-effective driving trace recognition methods of vehicle and urban road facility planing method |
Non-Patent Citations (2)
Title |
---|
孙玉砚等: "基于车牌识别系统车辆轨迹的行为异常检测", 《计算机研究与发展》 * |
胡正华等: "基于路网卡口系统车牌识别数据的机动车出行轨迹提取方法", 《测绘科学技术》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110033051A (en) * | 2019-04-18 | 2019-07-19 | 杭州电子科技大学 | A kind of trawler behavior method of discrimination based on multistep cluster |
CN110751837A (en) * | 2019-10-23 | 2020-02-04 | 招商华软信息有限公司 | Method, device and equipment for determining high-speed driving path and storage medium |
CN110796060A (en) * | 2019-10-23 | 2020-02-14 | 招商华软信息有限公司 | Method, device and equipment for determining high-speed driving route and storage medium |
CN110751837B (en) * | 2019-10-23 | 2021-12-14 | 招商华软信息有限公司 | Method, device and equipment for determining high-speed driving path and storage medium |
CN110796060B (en) * | 2019-10-23 | 2023-09-01 | 招商华软信息有限公司 | High-speed driving route determining method, device, equipment and storage medium |
CN112309118A (en) * | 2020-11-03 | 2021-02-02 | 广州市交通规划研究院 | Vehicle trajectory calculation method based on space-time similarity |
CN112748452A (en) * | 2020-12-11 | 2021-05-04 | 上海城市交通设计院有限公司 | GPS track cleaning method based on road network data |
CN112748452B (en) * | 2020-12-11 | 2022-09-23 | 上海城市交通设计院有限公司 | GPS track cleaning method based on road network data |
CN114944083A (en) * | 2022-05-13 | 2022-08-26 | 公安部交通管理科学研究所 | Method for judging distance between running vehicle on expressway and front vehicle |
CN114944083B (en) * | 2022-05-13 | 2023-03-24 | 公安部交通管理科学研究所 | Method for judging distance between running vehicle on expressway and front vehicle |
CN115790640A (en) * | 2023-02-02 | 2023-03-14 | 广州斯沃德科技有限公司 | Vehicle mileage correction method and device based on multidimensional data intelligent analysis |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109377757A (en) | The vehicle driving track extraction method of license plate identification data based on the rough error containing multi-source | |
CN109448370B (en) | Traffic control subarea division method based on vehicle track data | |
CN109410577B (en) | Self-adaptive traffic control subarea division method based on space data mining | |
CN103646187B (en) | Method for obtaining vehicle travel path and OD (Origin-Destination) matrix in statistic period | |
CN107134142A (en) | A kind of urban road method for predicting based on multisource data fusion | |
CN102779410B (en) | Parallel implementation method of multi-source heterogeneous traffic data fusion | |
CN104700646B (en) | A kind of taxi exception track real-time detection method based on online gps data | |
CN102595323B (en) | Method for obtaining resident travel characteristic parameter based on mobile phone positioning data | |
CN110579219B (en) | Track data processing method and device, storage medium and computer equipment | |
CN110111574B (en) | Urban traffic imbalance evaluation method based on flow tree analysis | |
CN105448092B (en) | The analysis method and device of a kind of associated vehicle | |
CN106971534A (en) | Commuter characteristic analysis method based on number plate data | |
CN102855760B (en) | On-line queuing length detection method based on floating vehicle data | |
CN100468481C (en) | Intelligent analysis system for municipal traffic journey time | |
CN106856049B (en) | Key intersection demand aggregation analysis method based on bayonet number plate identification data | |
CN105206048A (en) | Urban resident traffic transfer mode discovery system and method based on urban traffic OD data | |
CN105825669A (en) | System and method for identifying urban expressway traffic bottlenecks | |
CN111862606A (en) | Illegal operating vehicle identification method based on multi-source data | |
CN105303854A (en) | Travel route data processing method and device | |
CN107958031A (en) | Resident trip OD distribution extracting methods based on fused data | |
CN106651027A (en) | Internet regular bus route optimization method based on social network | |
CN109859495A (en) | A method of overall travel speed is obtained based on RFID data | |
CN102968900A (en) | Method for processing RFID (Radio Frequency Identification) traffic data | |
CN103093625B (en) | City road traffic condition real-time estimation method based on reliability verification | |
CN110021161B (en) | Traffic flow direction prediction method and system |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190222 |