CN109766777A - Detection method, device, storage medium and the electronic equipment of abnormal track - Google Patents
Detection method, device, storage medium and the electronic equipment of abnormal track Download PDFInfo
- Publication number
- CN109766777A CN109766777A CN201811554843.0A CN201811554843A CN109766777A CN 109766777 A CN109766777 A CN 109766777A CN 201811554843 A CN201811554843 A CN 201811554843A CN 109766777 A CN109766777 A CN 109766777A
- Authority
- CN
- China
- Prior art keywords
- track
- section
- orderly
- characteristic information
- road
- 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)
- Train Traffic Observation, Control, And Security (AREA)
Abstract
This disclosure relates to which a kind of detection method, device, storage medium and the electronic equipment of exception track, detect accuracy rate not high technical problem in abnormal track using the prior art for solving.This method comprises: track to be detected to be divided into the orderly section being made of multiple sections, wherein the road type in two neighboring section is different in orderly section;Obtain the track characteristic information of every a road section in orderly section;Judge in the orderly section of track to be detected and in the orderly section of reference trajectory, whether the track characteristic information in two sections of same position matches, and it records in the orderly section of track to be detected and the orderly section of reference trajectory, the section number that track characteristic information matches;If in the orderly section of track to be detected and the orderly section of reference trajectory, the section number that track characteristic information matches is less than preset threshold, it is determined that track to be detected is abnormal track.
Description
Technical field
This disclosure relates to data mining technology field, and in particular, to it is a kind of exception track detection method, device, deposit
Storage media and electronic equipment.
Background technique
With can positioning intelligent equipment universal and wireless communication technique development, the location data of extensive quantity
Collected and persistence saves, and forms the track data of magnanimity, how to have passed through the trajectory data mining exception information of magnanimity
Important subject as data mining technology field.In the prior art, when carrying out abnormal track detection, usually by track
Several mutually independent global properties are expressed as, are then detected using the abnormal point method of determining and calculating based on distance.However it adopts
In aforementioned manners, the local difference between track is had ignored, causes the accuracy rate of testing result lower.
Summary of the invention
Purpose of this disclosure is to provide detection method, device, storage medium and the electronic equipments of a kind of abnormal track, are used for
It solves to detect accuracy rate not high technical problem in abnormal track using the prior art.
To achieve the goals above, disclosure first aspect provides a kind of detection method of abnormal track, the method packet
It includes:
Track to be detected is divided into the orderly section being made of multiple sections, wherein adjacent two in the orderly section
The road type in a section is different;
Obtain the track characteristic information of every a road section in the orderly section;
Judge in the orderly section of the track to be detected and in the orderly section of reference trajectory, same position
Whether the track characteristic information in two sections matches, and record the track to be detected orderly section and the reference trajectory
Orderly section in, section number that track characteristic information matches;
If in the orderly section of the track to be detected and the orderly section of the reference trajectory, track characteristic information phase
The section number matched is less than preset threshold, it is determined that the track to be detected is abnormal track.
It is optionally, described that track to be detected is divided into the orderly section being made of multiple sections, comprising:
The track to be detected is divided into multiple sections in advance;
The track data of every a road section is obtained, the track data includes the longitude and latitude of tracing point, plane characteristic information, indulges
Disconnected characteristic information;
For every a road section, the track data in the section is inputted to the disaggregated model of road type for identification, is somebody's turn to do
The road type information in the section of disaggregated model output;
Merge the identical adjacent segments of road type information described in the multiple section, obtains the orderly section.
Optionally, the method also includes establishing the method for the disaggregated model, comprising:
The track data of each historical track is obtained, the track data includes the longitude and latitude of tracing point, plane characteristic letter
Breath, vertical disconnected characteristic information;
The road type for marking each historical track obtains the road class of the track data of each historical track
Type label;
According to the track data train classification models for having road type label, to obtain road type for identification
Object-class model.
Optionally, the basis has the track data train classification models of road type label, to be used for
Identify the object-class model of road type, comprising:
The track data is divided into training set and forecast set, and by the tracing point of track data in the training set
Longitude and latitude, plane characteristic information, vertical disconnected characteristic information are as input training sample data, by track number described in the training set
According to road type label as output training sample data, the disaggregated model is trained;
Every training obtains a disaggregated model, then the longitude and latitude of the tracing point of track data, plane in the forecast set is special
Reference breath, vertical disconnected characteristic information input the disaggregated model, and output obtains prediction road type label;
It is predicted in road type label and the forecast set between the real road type label of track data according to described
Coincidence rate, determine the object-class model.
Optionally, the track characteristic information for obtaining every a road section in the orderly section, comprising:
The track data of every a road section in the orderly section is carried out vacuating processing, obtains target trajectory data;
The track characteristic information of the target trajectory data is extracted, the track characteristic information includes in the orderly section
The start-stop point, distance, direction and the gradient of every a road section.
Disclosure second aspect provides a kind of detection device of abnormal track, and described device includes:
Section division module, for track to be detected to be divided into the orderly section being made of multiple sections, wherein described
The road type in two neighboring section is different in orderly section;
Data obtaining module, for obtaining the track characteristic information of every a road section in the orderly section;
Section matching module, for judging in the orderly section of the track to be detected and in the orderly of reference trajectory
In section, whether the track characteristic information in two sections of same position matches, and records the orderly of the track to be detected
Section and the section number that in the orderly section of the reference trajectory, track characteristic information matches;
Abnormal track determining module, if for the orderly section of the track to be detected and the orderly road of the reference trajectory
Duan Zhong, the section number that track characteristic information matches are less than preset threshold, it is determined that the track to be detected is abnormal track.
Optionally, the section division module includes:
Submodule is divided, for the track to be detected to be divided into multiple sections in advance;
Acquisition submodule, for obtaining the track data of every a road section, the track data include tracing point longitude and latitude,
Plane characteristic information, vertical disconnected characteristic information;
It identifies submodule, for every a road section, the track data in the section is inputted to the classification of road type for identification
Model obtains the road type information in the section of disaggregated model output;
Merge submodule to obtain for merging the identical adjacent segments of road type information described in the multiple section
The orderly section.
Optionally, described device further includes model building module, and the model building module is used for:
The track data of each historical track is obtained, the track data includes the longitude and latitude of tracing point, plane characteristic letter
Breath, vertical disconnected characteristic information;
The road type for marking each historical track obtains the road class of the track data of each historical track
Type label;
According to the track data train classification models for having road type label, to obtain road type for identification
Object-class model.
Optionally, the model building module is used for:
The track data is divided into training set and forecast set, and by the tracing point of track data in the training set
Longitude and latitude, plane characteristic information, vertical disconnected characteristic information are as input training sample data, by track number described in the training set
According to road type label as output training sample data, the disaggregated model is trained;
Every training obtains a disaggregated model, then the longitude and latitude of the tracing point of track data, plane in the forecast set is special
Reference breath, vertical disconnected characteristic information input the disaggregated model, and output obtains prediction road type label;
It is predicted in road type label and the forecast set between the real road type label of track data according to described
Coincidence rate, determine the object-class model.
Optionally, the data obtaining module includes:
Submodule is vacuated, carries out vacuating processing for the track data to every a road section in the orderly section, obtains mesh
Mark track data;
Extracting sub-module, for extracting the track characteristic information of the target trajectory data, the track characteristic packet
Include the start-stop point, distance, direction and the gradient of every a road section in the orderly section.
The disclosure third aspect provides a kind of computer readable storage medium, is stored thereon with computer program, the program
The step of detection method for the abnormal track that disclosure first aspect provides is realized when being executed by processor.
Disclosure fourth aspect provides a kind of electronic equipment, comprising:
Memory is stored thereon with computer program;
Processor, for executing the computer program in the memory, to realize that disclosure first aspect provides
Abnormal track detection method the step of.
In the above-mentioned technical solutions, by the way that track to be detected is divided into the orderly section being made of multiple sections, this has
The road type in two neighboring section is different in sequence section, then obtains the track characteristic information of every a road section in orderly section,
And analyze the matching feelings of the track characteristic information in each section in track to be detected and the track characteristic information of reference trajectory respective stretch
Condition can determine whether track to be detected is abnormal track.In this way, being orderly section by trajectory map to be detected, and successively right
The track characteristic information of every a road section carries out the matching analysis in orderly section, remains the abnormality degree of track local feature information,
The accuracy rate of abnormal track detection can be promoted.
Other feature and advantage of the disclosure will the following detailed description will be given in the detailed implementation section.
Detailed description of the invention
Attached drawing is and to constitute part of specification for providing further understanding of the disclosure, with following tool
Body embodiment is used to explain the disclosure together, but does not constitute the limitation to the disclosure.In the accompanying drawings:
Fig. 1 is a kind of flow chart of the detection method for abnormal track that the embodiment of the present disclosure provides;
Fig. 2 is a kind of flow chart of the detection method for abnormal track that the embodiment of the present disclosure provides;
Fig. 3 is the schematic diagram that a kind of track that the embodiment of the present disclosure provides is divided into multiple sections in advance;
Fig. 4 is a kind of block diagram of the detection device for abnormal track that the embodiment of the present disclosure provides;
Fig. 5 is a kind of block diagram of the detection device for abnormal track that the embodiment of the present disclosure provides;
Fig. 6 is the block diagram for a kind of electronic equipment that the embodiment of the present disclosure provides.
Specific embodiment
It is described in detail below in conjunction with specific embodiment of the attached drawing to the disclosure.It should be understood that this place is retouched
The specific embodiment stated is only used for describing and explaining the disclosure, is not limited to the disclosure.
Track is usually to be made of several static points connected each other, and be expressed as several mutually independent global properties,
Such as track present position, the track direction of motion, track movement velocity, path length etc..In the prior art, between track
Difference degree (distance) can be indicated by above-mentioned attribute difference, can detecte hair in conjunction with the abnormal point method of determining and calculating based on distance
Now abnormal track.However use the above method, have ignored the local difference between track, cause the accuracy rate of testing result compared with
It is low.
In order to solve to detect accuracy rate not high technical problem in abnormal track using the prior art, the embodiment of the present disclosure is provided
A kind of detection method of exception track, is combined with specific embodiments below described in detail technical solution provided by the present disclosure.
Fig. 1 is a kind of flow chart of the detection method for abnormal track that the embodiment of the present disclosure provides, as shown in Figure 1, the party
Method the following steps are included:
S11, track to be detected is divided into the orderly section being made of multiple sections.
The spatial form of road axis is referred to as linear, linear to be divided into horizontal alignment and vertical broken string shape, further planar line
Shape may include straight line, circular curve, three kinds of easement curve it is linear, indulge broken string shape indicate road direction of advance go up a slope, descending it is vertical
To the gradient.It therefore, can be forthright, loop, bending road, upward trend, downhill path by classification of road according to road alignment.In step
In S11, track to be detected is divided into multiple there are the orderly sections of succession relation, wherein two neighboring road in orderly section
The road type of section is different.
S12, the track characteristic information for obtaining every a road section in orderly section.
Illustratively, track characteristic information includes in orderly section in the start-stop point of every a road section, distance, direction and the gradient
At least one.Wherein, beginning and end is first point and the last one point of every a road section respectively, and distance is every a road section
From the distance of origin-to-destination, the gradient be every a road section altitude change amount mean value, section direction can according to road type into
Row determines, the direction in forthright section is the azimuth mean value of tracing point, the direction on loop and bending road be expressed as being not at [0 °,
360 °] and far from [0 °, 360 °] value, such as a certain loop section direction be -360 °, a certain bending section direction be 720 °.
S13, judge in the orderly section of track to be detected and in the orderly section of reference trajectory, same position
Whether the track characteristic information in two sections matches, and records the orderly section of track to be detected and the orderly road of reference trajectory
Duan Zhong, the section number that track characteristic information matches.
Wherein, reference trajectory can be user in prepared desired path before travel, be also possible to system according to user
The traffic path of the beginning and end invocation map plane-generating of setting.Such as in fleet management, reference trajectory is to set in advance
Fixed a plurality of traffic route, by being carried out abnormality detection to track to achieve the purpose that avoid bus private;In shared trip clothes
In business, reference trajectory can be system according to passenger's order beginning and end invocation map traffic route generated, by right
Driver's traffic route carries out abnormality detection, and can find that driver's driving states are abnormal, in time to guarantee the personal safety of passenger.And
And reference trajectory orderly section and track characteristic information can pre-generate and store, planning identical base convenient for subsequent
When standard gauge mark, the orderly section of the reference trajectory and the track characteristic information in each orderly section can be directly obtained.
Specifically, every a road section in orderly section is basic comparing unit, will be in the orderly section of track to be detected
The track characteristic information of every a road section, respectively with the track characteristic information in same position section in the orderly section of reference trajectory into
Row compares one by one, judges whether the track characteristic information in the two sections matches.
In the case that in track, characteristic information only includes single features attribute, the Xiang Te in two sections of position if they are the same
It is consistent to levy ATTRIBUTE INDEX, it is determined that two sections match;Or pre-determined distance threshold value, if they are the same between the two of position section
Distance (mismatched degree) be less than the distance threshold, then the two sections are mutually matched.For example, single features attribute is road
Section start-stop point, then when the section start-stop point in the two of same position section is consistent, can determine that the two sections match.It is in-orbit
In the case that mark characteristic information includes various features attribute, two sections of position have the feature greater than preset quantity if they are the same
Attribute is consistent, it is determined that two sections match;For example, there are four types of characteristic attributes for the track characteristic information tool in section, respectively
Start-stop point, distance, direction and the gradient in section, if preset quantity is three kinds, two section characteristic attributes of same position
When consistent type is no less than three kinds, it can determine that the two sections match.Accordingly, it counts and records having for track to be detected
In sequence section and the orderly section of reference trajectory, section number that track characteristic information matches.
If track characteristic information matches in the orderly section of S14, the orderly section of track to be detected and reference trajectory
Section number is less than preset threshold, it is determined that track to be detected is abnormal track.
Two sections for defining the same position of track to be detected and reference trajectory match for local matching;To be checked
It surveys in the orderly section of track and in the orderly section of reference trajectory, the track characteristic in two sections of each same position
Information successively local matching, then can determine track to be detected and reference trajectory global registration;If the orderly section of track to be detected
It is less than preset outlier threshold with the number to match in the orderly section of reference trajectory, then can determine that track to be detected is abnormal
Track.
For example, by the orderly section of the track to be detected counted step S13 and the orderly section of reference trajectory,
The 70% of the section section the Shuo Weizong quantity that track characteristic information matches, being less than preset threshold is 85%, then can determine to be checked
Surveying track is abnormal track.
Using the above method, by the way that track to be detected is divided into the orderly section being made of multiple sections, the orderly road
The road type in two neighboring section is different in section, then obtains the track characteristic information of every a road section in orderly section, and divide
The match condition for analysing the track characteristic information in each section in track to be detected and the track characteristic information of reference trajectory respective stretch, can
It whether is abnormal track with determination track to be detected.In this way, being orderly section by trajectory map to be detected, and successively to orderly road
The track characteristic information of every a road section carries out the matching analysis in section, remains the abnormality degree of track local feature information, Ke Yiti
The accuracy rate for rising abnormal track detection expands the application scenarios of abnormal track detection technology so that testing result is relatively reliable.
Fig. 2 is a kind of flow chart of the detection method for abnormal track that the embodiment of the present disclosure provides, as shown in Fig. 2, the party
Method the following steps are included:
S21, track to be detected is divided into multiple sections in advance.
Illustratively, track to be detected is divided into multiple sections in advance for unit according to desired guiding trajectory length, as shown in figure 3,
Track to be detected is divided into section A, section B, five part section C, section D and section E.
In the specific implementation, which can be the current driving trace of vehicle, for it is pre-planned
Reference trajectory is compared, to judge whether the current driving trace of vehicle deviates the reference trajectory.It in this case, will be to
Detection track be divided into multiple sections refer to vehicle from beginning track record play the track run between current time
It is divided.
S22, the track data for obtaining every a road section.
The track data in section can be the geographic position data of each tracing point in the section, for example, vehicle is travelling
The geographic position data of position (tracing point) can be presently according to a certain frequency collection in the process.
Specifically, track data may include the longitude and latitude of tracing point, plane characteristic information, vertical disconnected characteristic information.Example
Ground, longitude and latitude data are for calculating section distance;Plane characteristic information includes the bearing data of tracing point, such as azimuth is equal
The information such as value, orientation mean angular deviation or azimuth sectors type, for characterizing the plane characteristic in section, to distinguish road segment classification
For forthright, loop or which kind of being bent in road;Vertical disconnected characteristic information includes the elevation data of tracing point, such as height above sea level standard deviation,
Altitude change amount mean value, for characterizing the vertical disconnected feature in section, to distinguish section as upward slope or or downhill path.
Optionally, every a road section vacuates tracing point using evacuating algorithm, rising for every a road section can be retained
Point, terminal and the tracing point for being greater than or equal to length threshold with upper retention point distance, are equivalent to the section being divided into n
Sub- section, wherein n-1 is that the section vacuates the tracing point quantity retained between Origin And Destination after processing, and it is n that size, which is arranged,
The analysis window of+1 tracing point, in this way, sliding backward window with step-length h, h is less than or equal to n+1, successively obtains analysis window
Longitude and latitude, plane characteristic information, the vertical disconnected characteristic information for the tracing point for inside including, can be obtained the track data of every a road section.
S23, it is directed to every a road section, the track data in the section is inputted to the disaggregated model of road type for identification, is obtained
The road type information in the section exported to the disaggregated model.
In a kind of possible embodiment, the disaggregated model for road type classification can be built by the following method
It is vertical: to obtain the track data of each historical track;The road type for marking each historical track obtains the rail of each historical track
The road type label of mark data;According to the track data train classification models for having road type label, to obtain for knowing
The object-class model of other road type.
Specifically, the track data of different types of road is obtained from historical track, which includes tracing point
Longitude and latitude, plane characteristic information, vertical disconnected characteristic information, such as track data is longitude, latitude, azimuth, the height above sea level of tracing point
Etc. information.Track data is marked according to road type, obtains the road type mark of the track data of each historical track
Label.Then, track data is divided into training set and forecast set, and by the longitude and latitude of the tracing point of track data in training set,
Plane characteristic information, vertical disconnected characteristic information are as input training sample data, by the road type mark of track data in training set
Label are trained disaggregated model as output training sample data.Every training obtains a disaggregated model, then by rail in forecast set
Longitude and latitude, plane characteristic information, the vertical disconnected characteristic information of the tracing point of mark data input the disaggregated model, and output obtains prediction road
Road type label, and according to the symbol in prediction road type label and forecast set between the real road type label of track data
Conjunction rate, determines object-class model.
Illustratively, object-class model can be determined as follows: if the disaggregated model institute that this training obtains is defeated
Coincidence rate in prediction road type label and forecast set out between the real road type label of track data is greater than default
Threshold value (for example, preset threshold is 95%) can then determine that the disaggregated model that this training obtains is object-class model;Or
Compare the practical road of track data in the prediction road type label and forecast set that multiple disaggregated models that training obtains are exported
Coincidence rate between the type label of road determines that wherein the highest disaggregated model of coincidence rate is object-class model.
Optionally, before being trained disaggregated model, the historical trajectory data got can be carried out vacuating place
Reason retains valid data, to reduce influence of the track data amount difference to modeling under different roads, and promotes modeling efficiency.
Specifically, for a historical track, it is first determined length threshold, and retain starting point, terminal and the intermediate point of the historical track,
Intermediate point is the tracing point for being not less than length threshold in track with upper retention point distance.Such as length threshold is L/n, reservation is gone through
The starting point of history track, terminal, the point for being greater than or equal to a upper retention point distance L/n, the tracing point after vacuating be starting point, in
Between point 1, intermediate point 2 ..., intermediate point n-1, terminal, wherein intermediate point 1 be apart from starting point distance be greater than or equal to L/n point,
Intermediate point 2 is the point for being greater than or equal to L/n apart from starting point distance, and so on.It is instructed the track data after vacuating as input
Practice sample data and input disaggregated model, using the road type label of the track data as output training sample data, to classification
Model is trained, until training obtains object-class model.
For all sections of track shown in Fig. 3, the track data in each section is inputted into object-class model respectively, is passed through
Object-class model can export to obtain the road type of section A to section E successively are as follows: forthright-forthright-forthright-loop-bending
Road.
S24, merge the identical adjacent segments of road type information in multiple sections, obtain orderly section.
Wherein, the road type in two neighboring section is different in orderly section.By step S23, it is known that section A, section B
And section C road type is identical, is forthright type, section D and section E road type are identical, are non-straight type.That
, merge the identical adjacent segments of road type information in track shown in Fig. 3, obtained orderly section includes, by section A,
Section I made of section B and section C merges, and the section II as made of section D and section E merging.Correspondingly, road
The track data of section I includes the merging of non-duplicate data in section A, section B and section C, and the track data of section II includes
The merging of non-duplicate data in section D and section E, in this way, track shown in Fig. 3 will be mapped as two adjacent road types
Different orderly section: forthright-non-straight.
S25, the track data of every a road section in orderly section is carried out vacuating processing, obtains target trajectory data.
The method for vacuating processing is carried out to the track data of every a road section in orderly section, during train classification models
The method for vacuate processing to historical trajectory data is identical, and details are not described herein again.
S26, the track characteristic information for extracting target trajectory data.
Illustratively, the track characteristic information of target trajectory data include the start-stop point of every a road section in orderly section, distance,
Direction and the gradient.Wherein, beginning and end is ordered into first point and the last one point of every a road section in section respectively, away from
From being every a road section with a distance from origin-to-destination, the gradient is the altitude change amount mean value of every a road section, and section direction can root
It is determined according to road type.
S27, judge in the orderly section of track to be detected and in the orderly section of reference trajectory, same position
Whether the track characteristic information in two sections matches, and records the orderly section of track to be detected and the orderly road of reference trajectory
Duan Zhong, the section number that track characteristic information matches.
If track characteristic information matches in the orderly section of S28, the orderly section of track to be detected and reference trajectory
Section number is less than preset threshold, it is determined that track to be detected is abnormal track.
It using the above method, is divided in advance by carrying out section to track to be detected, and establishes road type for identification
Track identification to be detected can be the section of multiple known road types by disaggregated model.Further, section is divided
Section, combination obtain orderly section, and carry out vacuating processing to track data, it is possible to reduce the processing of subsequent track characteristic information
Amount achievees the purpose that the efficiency for promoting road type identification.In addition, successively believing the track characteristic of every a road section in orderly section
Breath carries out the matching analysis, remains the abnormality degree of track local feature information, can promote the accuracy rate of abnormal track detection, make
It is relatively reliable to obtain testing result, expands the application scenarios of abnormal track detection technology.
Fig. 4 is a kind of block diagram of the detection device for abnormal track that the embodiment of the present disclosure provides, which can pass through
Software, hardware or both are implemented in combination with as some or all of of electronic equipment.Referring to Fig. 4, which includes:
Section division module 401, for track to be detected to be divided into the orderly section being made of multiple sections, wherein
The road type in two neighboring section is different in orderly section;
Data obtaining module 402, for obtaining the track characteristic information of every a road section in orderly section;
Section matching module 403, for judging in the orderly section of track to be detected and in the orderly of reference trajectory
In section, whether the track characteristic information in two sections of same position matches, and records the orderly section of track to be detected
In the orderly section of reference trajectory, section number that track characteristic information matches;
Abnormal track determining module 404, if in the orderly section of track to be detected and the orderly section of reference trajectory,
The section number that track characteristic information matches is less than preset threshold, it is determined that track to be detected is abnormal track.
Using above-mentioned apparatus, by the way that track to be detected is divided into the orderly section being made of multiple sections, the orderly road
The road type in two neighboring section is different in section, then obtains the track characteristic information of every a road section in orderly section, and divide
The match condition for analysing the track characteristic information in each section in track to be detected and the track characteristic information of reference trajectory respective stretch, can
It whether is abnormal track with determination track to be detected.In this way, being orderly section by trajectory map to be detected, and successively to orderly road
The track characteristic information of every a road section carries out the matching analysis in section, remains the abnormality degree of track local feature information, Ke Yiti
The accuracy rate for rising abnormal track detection expands the application scenarios of abnormal track detection technology so that testing result is relatively reliable.
Fig. 5 is a kind of block diagram of the detection device for abnormal track that the embodiment of the present disclosure provides, which can pass through
Software, hardware or both are implemented in combination with as some or all of of electronic equipment.Referring to Fig. 5, the detection of abnormal track is filled
Setting 400 further includes model building module 405, and model building module 405 is used for: obtaining the track data of each historical track, rail
Mark data include the longitude and latitude of tracing point, plane characteristic information, vertical disconnected characteristic information;Mark the road class of each historical track
Type obtains the road type label of the track data of each historical track;According to the track data instruction for having road type label
Practice disaggregated model, to obtain the object-class model of road type for identification.
Specifically, model building module 405 is used for: track data being divided into training set and forecast set, and by training set
Longitude and latitude, plane characteristic information, the vertical disconnected characteristic information of the tracing point of middle track data are used as input training sample data, will instruct
Practice and concentrate the road type label of track data as output training sample data, disaggregated model is trained;It is every trained
It is to a disaggregated model, then the longitude and latitude of the tracing point of track data, plane characteristic information, vertical disconnected characteristic information in forecast set is defeated
Enter the disaggregated model, output obtains prediction road type label;According to track data in prediction road type label and forecast set
Real road type label between coincidence rate, determine object-class model.
Optionally, section division module 401 includes: division submodule, for track to be detected to be divided into multiple roads in advance
Section;Acquisition submodule, for obtaining the track data of every a road section, track data includes the longitude and latitude of tracing point, plane characteristic
Information, vertical disconnected characteristic information;It identifies submodule, for every a road section, the track data in the section is inputted into road for identification
The disaggregated model of type obtains the road type information in the section of disaggregated model output;Merge submodule, it is multiple for merging
The identical adjacent segments of road type information in section, obtain orderly section.
Optionally, data obtaining module 402 includes: to vacuate submodule, for the track to every a road section in orderly section
Data carry out vacuating processing, obtain target trajectory data;Extracting sub-module, the track characteristic for extracting target trajectory data are believed
Breath, track characteristic information includes the start-stop point of every a road section, distance, direction and the gradient in orderly section.
It using above-mentioned apparatus, is divided in advance by carrying out section to track to be detected, and establishes road type for identification
Track identification to be detected can be the section of multiple known road types by disaggregated model.Further, section is divided
Section, combination obtain orderly section, and carry out vacuating processing to track data, it is possible to reduce the processing of subsequent track characteristic information
Amount achievees the purpose that the efficiency for promoting road type identification.In addition, successively believing the track characteristic of every a road section in orderly section
Breath carries out the matching analysis, remains the abnormality degree of track local feature information, can promote the accuracy rate of abnormal track detection, make
It is relatively reliable to obtain testing result, expands the application scenarios of abnormal track detection technology.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
Fig. 6 is the block diagram for a kind of electronic equipment 600 that the embodiment of the present disclosure provides.As shown in fig. 6, the electronic equipment 600
It may include: processor 601, memory 602.The electronic equipment 600 can also include multimedia component 603, input/output
(I/O) one or more of interface 604 and communication component 605.
Wherein, processor 601 is used to control the integrated operation of the electronic equipment 600, to complete above-mentioned abnormal track
All or part of the steps in detection method.Memory 602 is for storing various types of data to support in the electronic equipment
600 operation, these data for example may include any application or method for operating on the electronic equipment 600
Instruction and the relevant data of application program, such as contact data, the message of transmitting-receiving, picture, audio, video etc..This is deposited
Reservoir 602 can realize by any kind of volatibility or non-volatile memory device or their combination, for example, it is static with
Machine accesses memory (Static Random Access Memory, abbreviation SRAM), electrically erasable programmable read-only memory
(Electrically Erasable Programmable Read-Only Memory, abbreviation EEPROM), erasable programmable
Read-only memory (Erasable Programmable Read-Only Memory, abbreviation EPROM), programmable read only memory
(Programmable Read-Only Memory, abbreviation PROM), and read-only memory (Read-Only Memory, referred to as
ROM), magnetic memory, flash memory, disk or CD.Multimedia component 603 may include screen and audio component.Wherein
Screen for example can be touch screen, and audio component is used for output and/or input audio signal.For example, audio component may include
One microphone, microphone is for receiving external audio signal.The received audio signal can be further stored in storage
Device 602 is sent by communication component 605.Audio component further includes at least one loudspeaker, is used for output audio signal.I/O
Interface 604 provides interface between processor 601 and other interface modules, other above-mentioned interface modules can be keyboard, mouse,
Button etc..These buttons can be virtual push button or entity button.Communication component 605 is for the electronic equipment 600 and other
Wired or wireless communication is carried out between equipment.Wireless communication, such as Wi-Fi, bluetooth, near-field communication (Near Field
Communication, abbreviation NFC), 2G, 3G or 4G or they one or more of combination, therefore corresponding communication
Component 605 may include: Wi-Fi module, bluetooth module, NFC module.
In one exemplary embodiment, electronic equipment 600 can be by one or more application specific integrated circuit
(Application Specific Integrated Circuit, abbreviation ASIC), digital signal processor (Digital
Signal Processor, abbreviation DSP), digital signal processing appts (Digital Signal Processing Device,
Abbreviation DSPD), programmable logic device (Programmable Logic Device, abbreviation PLD), field programmable gate array
(Field Programmable Gate Array, abbreviation FPGA), controller, microcontroller, microprocessor or other electronics member
Part is realized, for executing the detection method of above-mentioned abnormal track.
In a further exemplary embodiment, a kind of computer readable storage medium including program instruction is additionally provided, it should
The step of detection method of above-mentioned abnormal track is realized when program instruction is executed by processor.For example, this computer-readable is deposited
Storage media can be the above-mentioned memory 602 including program instruction, and above procedure instruction can be by the processor of electronic equipment 600
601 execute to complete the detection method of above-mentioned abnormal track.
The preferred embodiment of the disclosure is described in detail in conjunction with attached drawing above, still, the disclosure is not limited to above-mentioned reality
The detail in mode is applied, in the range of the technology design of the disclosure, a variety of letters can be carried out to the technical solution of the disclosure
Monotropic type, these simple variants belong to the protection scope of the disclosure.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case where shield, it can be combined in any appropriate way.In order to avoid unnecessary repetition, the disclosure to it is various can
No further explanation will be given for the combination of energy.
In addition, any combination can also be carried out between a variety of different embodiments of the disclosure, as long as it is without prejudice to originally
Disclosed thought equally should be considered as disclosure disclosure of that.
Claims (10)
1. a kind of detection method of exception track, which is characterized in that the described method includes:
Track to be detected is divided into the orderly section being made of multiple sections, wherein two neighboring road in the orderly section
The road type of section is different;
Obtain the track characteristic information of every a road section in the orderly section;
Judge in the orderly section of the track to be detected and in the orderly section of reference trajectory, two of same position
Whether the track characteristic information in section matches, and records the orderly section of the track to be detected and having for the reference trajectory
In sequence section, section number that track characteristic information matches;
If track characteristic information matches in the orderly section of the track to be detected and the orderly section of the reference trajectory
Section number is less than preset threshold, it is determined that the track to be detected is abnormal track.
2. the method according to claim 1, wherein described be divided into track to be detected is made of multiple sections
Orderly section, comprising:
The track to be detected is divided into multiple sections in advance;
The track data of every a road section is obtained, the track data includes the longitude and latitude of tracing point, plane characteristic information, vertical disconnected spy
Reference breath;
For every a road section, the track data in the section is inputted to the disaggregated model of road type for identification, obtains the classification
The road type information in the section of model output;
Merge the identical adjacent segments of road type information described in the multiple section, obtains the orderly section.
3. according to the method described in claim 2, it is characterized in that, the method also includes establishing the side of the disaggregated model
Method, comprising:
The track data of each historical track is obtained, the track data includes the longitude and latitude of tracing point, plane characteristic information, indulges
Disconnected characteristic information;
The road type for marking each historical track obtains the road type mark of the track data of each historical track
Label;
According to the track data train classification models for having road type label, to obtain the mesh of road type for identification
Mark disaggregated model.
4. according to the method described in claim 3, it is characterized in that, the basis has the track number of road type label
According to train classification models, to obtain the object-class model of road type for identification, comprising:
The track data is divided into training set and forecast set, and by the longitude and latitude of the tracing point of track data in the training set
Degree, plane characteristic information, vertical disconnected characteristic information are as input training sample data, by track data described in the training set
Road type label is trained the disaggregated model as output training sample data;
Every training obtains a disaggregated model, then believes the longitude and latitude of the tracing point of track data, plane characteristic in the forecast set
Breath, vertical disconnected characteristic information input the disaggregated model, and output obtains prediction road type label;
According to the symbol in the prediction road type label and the forecast set between the real road type label of track data
Conjunction rate determines the object-class model.
5. method according to claim 1 to 4, which is characterized in that described to obtain in the orderly section often
The track characteristic information of a road section, comprising:
The track data of every a road section in the orderly section is carried out vacuating processing, obtains target trajectory data;
The track characteristic information of the target trajectory data is extracted, the track characteristic information includes each in the orderly section
Start-stop point, distance, direction and the gradient in section.
6. a kind of detection device of exception track, which is characterized in that described device includes:
Section division module, for track to be detected to be divided into the orderly section being made of multiple sections, wherein described orderly
The road type in two neighboring section is different in section;
Data obtaining module, for obtaining the track characteristic information of every a road section in the orderly section;
Section matching module, for judging in the orderly section of the track to be detected and in the orderly section of reference trajectory
In, whether the track characteristic information in two sections of same position matches, and records the orderly section of the track to be detected
In the orderly section of the reference trajectory, section number that track characteristic information matches;
Abnormal track determining module, if for the orderly section of the track to be detected and the orderly section of the reference trajectory
In, the section number that track characteristic information matches is less than preset threshold, it is determined that the track to be detected is abnormal track.
7. device according to claim 6, which is characterized in that the section division module includes:
Submodule is divided, for the track to be detected to be divided into multiple sections in advance;
Acquisition submodule, for obtaining the track data of every a road section, the track data includes the longitude and latitude of tracing point, plane
Characteristic information, vertical disconnected characteristic information;
It identifies submodule, for every a road section, the track data in the section is inputted to the disaggregated model of road type for identification,
Obtain the road type information in the section of disaggregated model output;
Merge submodule, for merging the identical adjacent segments of road type information described in the multiple section, obtains described
Orderly section.
8. device according to claim 7, which is characterized in that described device further includes model building module, the model
Module is established to be used for:
The track data of each historical track is obtained, the track data includes the longitude and latitude of tracing point, plane characteristic information, indulges
Disconnected characteristic information;
The road type for marking each historical track obtains the road type mark of the track data of each historical track
Label;
According to the track data train classification models for having road type label, to obtain the mesh of road type for identification
Mark disaggregated model.
9. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is held by processor
The step of any one of claims 1 to 5 the method is realized when row.
10. a kind of electronic equipment characterized by comprising
Memory is stored thereon with computer program;
Processor, for executing the computer program in the memory, to realize any one of claims 1 to 5 institute
The step of stating method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811554843.0A CN109766777B (en) | 2018-12-18 | 2018-12-18 | Abnormal track detection method and device, storage medium and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811554843.0A CN109766777B (en) | 2018-12-18 | 2018-12-18 | Abnormal track detection method and device, storage medium and electronic equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109766777A true CN109766777A (en) | 2019-05-17 |
CN109766777B CN109766777B (en) | 2021-08-13 |
Family
ID=66451547
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811554843.0A Active CN109766777B (en) | 2018-12-18 | 2018-12-18 | Abnormal track detection method and device, storage medium and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109766777B (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110726414A (en) * | 2019-10-25 | 2020-01-24 | 百度在线网络技术(北京)有限公司 | Method and apparatus for outputting information |
CN110795467A (en) * | 2019-09-26 | 2020-02-14 | 腾讯大地通途(北京)科技有限公司 | Traffic rule data processing method and device, storage medium and computer equipment |
CN111475544A (en) * | 2020-03-30 | 2020-07-31 | 智慧航海(青岛)科技有限公司 | Method and device for detecting outliers in ship track data |
CN111505681A (en) * | 2020-04-23 | 2020-08-07 | 上海普适导航科技股份有限公司 | Method and system for correcting rarefaction of positioning track |
CN111862586A (en) * | 2019-12-11 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining abnormal road section of road area and storage medium |
CN111882873A (en) * | 2020-07-22 | 2020-11-03 | 平安国际智慧城市科技股份有限公司 | Track anomaly detection method, device, equipment and medium |
CN112147658A (en) * | 2019-06-27 | 2020-12-29 | 财付通支付科技有限公司 | Method, device and equipment for judging moving direction of vehicle and storage medium |
CN112533208A (en) * | 2019-08-27 | 2021-03-19 | 中国移动通信有限公司研究院 | Model training method, false terminal identification method and device, and electronic device |
CN112862248A (en) * | 2021-01-06 | 2021-05-28 | 山东旗帜信息有限公司 | Road section rarefying method, equipment and medium |
CN113535504A (en) * | 2020-04-17 | 2021-10-22 | 华为技术有限公司 | Data thinning method and device |
CN113593218A (en) * | 2021-06-28 | 2021-11-02 | 北京百度网讯科技有限公司 | Method and device for detecting traffic abnormal event, electronic equipment and storage medium |
CN114353810A (en) * | 2022-01-10 | 2022-04-15 | 河海大学 | HMM efficient map matching method based on R tree and track segmentation |
CN115457769A (en) * | 2022-08-31 | 2022-12-09 | 东风商用车有限公司 | Data correction method, device, equipment and readable storage medium |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103489010A (en) * | 2013-09-25 | 2014-01-01 | 吉林大学 | Fatigue driving detecting method based on driving behaviors |
CN104331422A (en) * | 2014-10-14 | 2015-02-04 | 广州市香港科大霍英东研究院 | Road section type presumption method |
CN106153064A (en) * | 2015-04-17 | 2016-11-23 | 高德信息技术有限公司 | The display packing of a kind of intersection and device |
CN106573624A (en) * | 2014-08-19 | 2017-04-19 | 奥迪股份公司 | Method for operating a motor vehicle, motor vehicle, and computer program |
CN106767867A (en) * | 2016-12-08 | 2017-05-31 | 努比亚技术有限公司 | Vehicle navigation apparatus and method |
CN107038884A (en) * | 2017-05-17 | 2017-08-11 | 北汽福田汽车股份有限公司 | Push the methods, devices and systems of road conditions |
CN107153931A (en) * | 2016-03-03 | 2017-09-12 | 重庆邮电大学 | A kind of Express Logistics dispense method for detecting abnormality |
CN107316459A (en) * | 2017-07-21 | 2017-11-03 | 武汉依迅北斗空间技术有限公司 | A kind of track of vehicle method for detecting abnormality and system |
CN107399333A (en) * | 2017-07-24 | 2017-11-28 | 清华大学 | A kind of accurate parking method of train towards train automatic Pilot |
CN108133611A (en) * | 2016-12-01 | 2018-06-08 | 中兴通讯股份有限公司 | Vehicle driving trace monitoring method and system |
US10088324B2 (en) * | 2016-05-02 | 2018-10-02 | International Business Machines Corporation | Trajectory data compression |
CN108922173A (en) * | 2018-06-20 | 2018-11-30 | 青岛海信网络科技股份有限公司 | A kind of vehicle deviation detection method and device |
-
2018
- 2018-12-18 CN CN201811554843.0A patent/CN109766777B/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103489010A (en) * | 2013-09-25 | 2014-01-01 | 吉林大学 | Fatigue driving detecting method based on driving behaviors |
CN106573624A (en) * | 2014-08-19 | 2017-04-19 | 奥迪股份公司 | Method for operating a motor vehicle, motor vehicle, and computer program |
CN104331422A (en) * | 2014-10-14 | 2015-02-04 | 广州市香港科大霍英东研究院 | Road section type presumption method |
CN106153064A (en) * | 2015-04-17 | 2016-11-23 | 高德信息技术有限公司 | The display packing of a kind of intersection and device |
CN107153931A (en) * | 2016-03-03 | 2017-09-12 | 重庆邮电大学 | A kind of Express Logistics dispense method for detecting abnormality |
US10088324B2 (en) * | 2016-05-02 | 2018-10-02 | International Business Machines Corporation | Trajectory data compression |
CN108133611A (en) * | 2016-12-01 | 2018-06-08 | 中兴通讯股份有限公司 | Vehicle driving trace monitoring method and system |
CN106767867A (en) * | 2016-12-08 | 2017-05-31 | 努比亚技术有限公司 | Vehicle navigation apparatus and method |
CN107038884A (en) * | 2017-05-17 | 2017-08-11 | 北汽福田汽车股份有限公司 | Push the methods, devices and systems of road conditions |
CN107316459A (en) * | 2017-07-21 | 2017-11-03 | 武汉依迅北斗空间技术有限公司 | A kind of track of vehicle method for detecting abnormality and system |
CN107399333A (en) * | 2017-07-24 | 2017-11-28 | 清华大学 | A kind of accurate parking method of train towards train automatic Pilot |
CN108922173A (en) * | 2018-06-20 | 2018-11-30 | 青岛海信网络科技股份有限公司 | A kind of vehicle deviation detection method and device |
Non-Patent Citations (2)
Title |
---|
任杰等: "基于划分检测模型的终端区异常轨迹检测方法", 《航空计算技术》 * |
李梅: "高速公路安全服务水平分级方法研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112147658B (en) * | 2019-06-27 | 2023-08-11 | 财付通支付科技有限公司 | Method, device, equipment and storage medium for judging moving direction of vehicle |
CN112147658A (en) * | 2019-06-27 | 2020-12-29 | 财付通支付科技有限公司 | Method, device and equipment for judging moving direction of vehicle and storage medium |
CN112533208A (en) * | 2019-08-27 | 2021-03-19 | 中国移动通信有限公司研究院 | Model training method, false terminal identification method and device, and electronic device |
CN110795467A (en) * | 2019-09-26 | 2020-02-14 | 腾讯大地通途(北京)科技有限公司 | Traffic rule data processing method and device, storage medium and computer equipment |
CN110795467B (en) * | 2019-09-26 | 2024-02-27 | 腾讯大地通途(北京)科技有限公司 | Traffic rule data processing method, device, storage medium and computer equipment |
CN110726414A (en) * | 2019-10-25 | 2020-01-24 | 百度在线网络技术(北京)有限公司 | Method and apparatus for outputting information |
CN110726414B (en) * | 2019-10-25 | 2021-07-27 | 百度在线网络技术(北京)有限公司 | Method and apparatus for outputting information |
CN111862586A (en) * | 2019-12-11 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining abnormal road section of road area and storage medium |
CN111475544A (en) * | 2020-03-30 | 2020-07-31 | 智慧航海(青岛)科技有限公司 | Method and device for detecting outliers in ship track data |
CN111475544B (en) * | 2020-03-30 | 2023-06-30 | 智慧航海(青岛)科技有限公司 | Method and device for detecting outliers in ship track data |
CN113535504A (en) * | 2020-04-17 | 2021-10-22 | 华为技术有限公司 | Data thinning method and device |
CN113535504B (en) * | 2020-04-17 | 2023-12-29 | 华为云计算技术有限公司 | Data thinning method and device |
CN111505681A (en) * | 2020-04-23 | 2020-08-07 | 上海普适导航科技股份有限公司 | Method and system for correcting rarefaction of positioning track |
CN111882873B (en) * | 2020-07-22 | 2022-01-28 | 平安国际智慧城市科技股份有限公司 | Track anomaly detection method, device, equipment and medium |
CN111882873A (en) * | 2020-07-22 | 2020-11-03 | 平安国际智慧城市科技股份有限公司 | Track anomaly detection method, device, equipment and medium |
CN112862248A (en) * | 2021-01-06 | 2021-05-28 | 山东旗帜信息有限公司 | Road section rarefying method, equipment and medium |
CN113593218A (en) * | 2021-06-28 | 2021-11-02 | 北京百度网讯科技有限公司 | Method and device for detecting traffic abnormal event, electronic equipment and storage medium |
CN114353810A (en) * | 2022-01-10 | 2022-04-15 | 河海大学 | HMM efficient map matching method based on R tree and track segmentation |
CN115457769A (en) * | 2022-08-31 | 2022-12-09 | 东风商用车有限公司 | Data correction method, device, equipment and readable storage medium |
CN115457769B (en) * | 2022-08-31 | 2023-10-13 | 东风商用车有限公司 | Data correction method, device, equipment and readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN109766777B (en) | 2021-08-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109766777A (en) | Detection method, device, storage medium and the electronic equipment of abnormal track | |
CN105528359B (en) | For storing the method and system of travel track | |
CN106878951B (en) | User trajectory analysis method and system | |
CN100555355C (en) | The method and system that the passage rate of road traffic calculates and mates | |
CN109697221A (en) | Method for digging, device, storage medium and the electronic equipment of track rule | |
Goh et al. | Online map-matching based on hidden markov model for real-time traffic sensing applications | |
CN109993969A (en) | A kind of road conditions determine information acquisition method, device and equipment | |
CN106574975A (en) | Trajectory matching using peripheral signal | |
CN107228677A (en) | Driftage recognition methods and device | |
CN105335597B (en) | For obtaining the method and system of the trajectory model of route | |
CN107871400B (en) | Road network information updating method and device | |
CN105910612A (en) | Personalized navigation method and system | |
CN109272032A (en) | Trip mode recognition methods, device, computer equipment and storage medium | |
CN108766031A (en) | A kind of method and apparatus of detection lane obstructions object | |
CN111862605B (en) | Road condition detection method and device, electronic equipment and readable storage medium | |
CN108663065A (en) | A kind of recognition methods of road type and device | |
CN105608919B (en) | The determination method and device of interchange of position | |
Moosavi et al. | Characterizing driving context from driver behavior | |
CN109756845A (en) | The recognition methods of cheating order and device in network about parking lot scape | |
CN107133689B (en) | Position marking method | |
CN108286973B (en) | Running data verification method and device and hybrid navigation system | |
CN108538054A (en) | A kind of method and system obtaining traffic information based on mobile phone signaling data | |
Jiang et al. | A framework of travel mode identification fusing deep learning and map-matching algorithm | |
CN113837442A (en) | Method and system for generating correction track oscillation route, terminal device and storage medium | |
CN108320184A (en) | For the method, apparatus of vehicle charging, equipment, system and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |