CN101923781A - Vehicle type recognizing method based on geomagnetic sensing technology - Google Patents
Vehicle type recognizing method based on geomagnetic sensing technology Download PDFInfo
- Publication number
- CN101923781A CN101923781A CN 201010239807 CN201010239807A CN101923781A CN 101923781 A CN101923781 A CN 101923781A CN 201010239807 CN201010239807 CN 201010239807 CN 201010239807 A CN201010239807 A CN 201010239807A CN 101923781 A CN101923781 A CN 101923781A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- sensing technology
- decision tree
- geomagnetic
- recognizing method
- 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
Images
Abstract
The invention discloses a vehicle type recognizing method based on a geomagnetic sensing technology in the technical field of traffic state detection sensors, which is used for solving the problem that the current intelligent traffic system is lack of a vehicle type recognition function. The method comprises the following steps of: acquiring vehicle waveform data through a geomagnetic sensor; enumerating waveform characteristics according to the vehicle waveform data; selecting effective waveform characteristics according to the influence of the waveform characteristics on a vehicle type recognizing result; training by utilizing the effective waveform characteristics and vehicle type sorting functions to obtain a decision tree; and recognizing a vehicle type according to the obtained decision tree. The invention can obtain the vehicle type of the specific vehicle according to the vehicle waveform data acquired by a traffic geomagnetic sensor, thereby providing essential data for an intelligent traffic system.
Description
Technical field
The invention belongs to traffic behavior detecting sensor technical field, relate in particular to a kind of model recognizing method based on geomagnetic sensing technology.
Background technology
It is the important prerequisite of carrying out traffic flow control and traffic administration such as inducing that road traffic state obtains, be the necessary basis of formulating traffic insurance measures such as traffic safety management strategy, traffic hazard detection, the analysis of traffic hazard reason, be the traffic infrastructure management, monitor and safeguard the indispensable firsthand information.Therefore to obtain be the important basic problem that traffic administration, traffic insurance and traffic infrastructure monitoring are safeguarded to traffic behavior.
Obtaining of traffic behavior is that information by the sensor that is laid on the road provides is carried out, and geomagnetic sensor has its special advantages, is not subjected to weather influence, with low cost, data to be easy to handle, be convenient to wireless transmission etc.Obtaining information such as vehicle in addition by fixing traffic sensor is the effective way that traffic administration person grasps road operational vehicle situation.Vehicle identification based on traffic sensor can provide basic data and foundation for the data fusion and the management decision on upper strata.
Along with ITS (Intelligent Transportation System, intelligent transportation system), grasps Traffic Information in real time, particularly road vehicle information in Development in China and application, for vehicle supervision department's decision-making in real time provides foundation, become a problem that needs to be resolved hurrily.
Summary of the invention
The objective of the invention is to, a kind of model recognizing method based on geomagnetic sensing technology is provided, be used to solve current intelligent transportation system and lack vehicle identification, can not real-time decision-making of vehicle supervision department provide the problem of accurate foundation.
Technical scheme is that a kind of model recognizing method based on geomagnetic sensing technology is characterized in that described method comprises:
Step 1: obtain the vehicle Wave data by geomagnetic sensor;
Step 2:, exemplify waveform character according to the vehicle Wave data;
Step 3:, select effective waveform character according to the influence of waveform character to the vehicle recognition result;
Step 4: utilize effective waveform character and vehicle classification function, training obtains decision tree;
Step 5:, carry out vehicle identification according to the decision tree that obtains.
Described geomagnetic sensor is set specific sample frequency.
Described waveform character comprises crest number, maximum crest time ratio, maximum trough time ratio, average, mean square deviation, crest amplitude, trough amplitude, maximum crest value trough value ratio, peak valley distribution series.
Described step 4 is utilized effective waveform character and vehicle classification function, and training obtains before the decision tree, also comprises effective waveform character is carried out the positive number fusion treatment.
Described step 5 specifically is, begins search from the decision tree root that obtains, and up to leaf node, the vehicle of described leaf node correspondence is exactly a recognition result.
The invention has the beneficial effects as follows, the vehicle Wave data that obtains according to the traffic geomagnetic sensor based on the model recognizing method of geomagnetic sensing technology, by extracting effective waveform character, obtain the decision-tree model of vehicle identification, utilize the binary search algorithm of decision tree, obtain the vehicle of concrete vehicle, for ITS provides basic data.
Description of drawings
Fig. 1 is based on the model recognizing method process flow diagram of geomagnetic sensing technology;
Fig. 2 is the comparison of wave shape figure of the different automobile types vehicle that records of geomagnetic sensor;
Fig. 3 extracts vehicle waveform character synoptic diagram;
Fig. 4 is based on the vehicle classification function table of different application;
Fig. 5 is the vehicle classification decision tree synoptic diagram that training obtains;
Fig. 6 is that vehicle is differentiated collection build-in test table as a result;
Fig. 7 is that vehicle is differentiated the outer table with test results of collection.
Embodiment
Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that following explanation only is exemplary, rather than in order to limit the scope of the invention and to use.
Fig. 1 is based on the model recognizing method process flow diagram of geomagnetic sensing technology.Among Fig. 1, a kind of model recognizing method based on geomagnetic sensing technology comprises the following steps:
Step 1: obtain the vehicle Wave data by geomagnetic sensor.
Geomagnetic sensor has self-learning capability, the situation of ground magnetic environment in the time of can learning not have vehicle, and return one through amplifying the corresponding specific amplitude in back, with the size of sign environmental magnetic field.If have vehicle process, the metallics of vehicle to cause the magnetic field fluctuation, by same amplification and processing, set specific sample frequency, obtain the Wave data of vehicle process.By the Wave data of vehicle through sensor, according to certain algorithm, the vehicle Wave data when obtaining each vehicle through sensor.The vehicle Wave data difference that different automobile types obtains; Vehicle of the same race is owing to the vehicle Wave data that factors such as noise obtain also is not quite similar, but shape is roughly constant.Fig. 2 is the comparison of wave shape figure of the different automobile types vehicle that records of geomagnetic sensor, and is among Fig. 2, identical with the vehicle of two oscillogram correspondences of delegation.
Step 2:, exemplify waveform character according to the vehicle Wave data.
As can be seen from Figure 2, the pairing waveform of different vehicle is very different, and the pairing waveform of vehicle of the same race has a lot of similarities, so can distinguish different vehicles according to the feature that waveform possesses.Consideration is from the formation reason of the structure analysis waveform character of vehicle, because it is the situation of change of vehicle through out-of-date earth magnetism that geomagnetic sensor detects, so the metal of vehicle advances the distribution situation of direction along garage, vehicle length, the chassis height, engine structures etc. all can influence the size on the earth magnetism, vehicle is with certain speed process in addition, can make oscillogram have different shapes, can be listed below feature by analysis: the number of Wave crest and wave trough and distribution situation, the minimax amplitude, each characteristic point position etc.
According to analysis, can extract following waveform character:
Feature 1: the number ((n) differs 1 with the trough number) of peak value appears in crest number (f1=m), waveform;
Feature 4: average (f4=average (VehWave));
Feature 5: mean square deviation (f5=var (VehWave));
Feature 6: crest amplitude (f6=max (VehWave));
Feature 7: trough amplitude (f7=min (VehWave));
Feature 8: maximum crest value trough value is than (peak-to valley ratio)
Feature 9: peak valley distribution series (f9=series (VehWave));
Wherein, VehWave is a vehicle Wave data sequence.The peak valley distribution series function of feature 9 can define one 01 sequence, promptly gets 1 during maximum value, gets 0 during minimal value; Also may be defined as partition sequence, when promptly Wave crest and wave trough drops on certain district, get the code name of respective area as Fig. 3.Feature 9 can be meticulousr feature that reflects waveform and trend, but be unfavorable for calculating.
Step 3:, select effective waveform character according to the influence of waveform character to the vehicle recognition result.
According to the feature of said extracted, and be each Feature Fusion a positive number, as proper vector, i.e. an x of this vehicle
i=(f1
i, f2
i... fk
i..., fn
i)
TWherein: x
iIt is the proper vector of i car; Fk
iBe k eigenwert of i car; The characteristic number of n for extracting.
Definition X=(x
1, x
2..., x
k..., x
m)
TBe the eigenmatrix of m car, wherein: the every row among the matrix X is represented n feature of a car.
Definition y=(y
1, y
2..., y
k..., y
m)
TBe vehicle vector, wherein y
i=VehCFun (i), vehicle classification function (VehCFun ()) can be according to different application definitions, and Fig. 4 is based on the vehicle classification function table of different application, and Fig. 4 has exemplified the vehicle classification function definition of different application.
Step 4: utilize effective waveform character and vehicle classification function, training obtains decision tree.
Decision tree is a kind of very effective machine learning classification algorithm, is a kind of instrument of representing processing logic with binary tree figure.Can intuitively, clearly express the logic requirement of processing.
The ultimate principle of decision tree is recursively data to be split into subclass, so that each subclass comprises the similar state of target variable, these target variables are measurable attributes.Each time tree is split, all will estimate of the influence of all input attributes measurable attribute.When this when recursively process finishes, decision tree is also created and is over.Set up the process of decision tree, promptly Shu growth course is the process of constantly data being carried out cutting, the corresponding problem of each cutting, also corresponding node." difference " maximum between the group that each cutting is all required to be divided into.
In the process that travels through from top to bottom along decision tree, all can run into a problem at each node, the difference of problem on each node is answered caused different branches, can arrive a leaf node at last.This process is exactly a process of utilizing decision tree to classify, utilizes several variablees (the corresponding problem of each variable) to judge affiliated classification (corresponding classification of each leaf meeting at last).
Utilize eigenmatrix X and vehicle vector y as given data, training obtains decision tree.
Step 5:, carry out vehicle identification according to the decision tree that obtains.
Each model data for newly obtaining obtains corresponding proper vector according to step 4, the decision tree that utilizes training to obtain, begins search from usage tree root, up to leafy node.The vehicle of leaf node correspondence of this moment is exactly the vehicle of this car of identifying.
Embodiment
The real data that the traffic geomagnetic sensor that utilizes site test to install collects is verified this method, obtains corresponding experiment result.
(1) experimental data and processing
Present embodiment choose collected 264 cars on September 19th, 2009 on-the-spot Wave data as training data, 49 bus data wherein, 215 car data.By to pre-service and vehicle discriminating algorithm, obtain the complete vehicle Wave data of each car, part vehicle Wave data is seen Fig. 2.
(2) feature extraction
Utilize the algorithm in the method, obtain the proper vector of each vehicle Wave data, present embodiment is chosen preceding 8 kinds of features as proper vector, and obtains 264 * 8 eigenmatrix by the proper vector of 264 cars.
(3) training decision tree
For simplicity, the vehicle classification A that chooses the table among Fig. 4 is divided into two classes to vehicle as the vehicle classification mode, compact car and large car.According to the vehicle of known vehicle, obtain the vehicle vector corresponding with eigenmatrix, utilize the treefit function among the matlab, the decision tree that training obtains is as shown in Figure 5.
(4) vehicle is differentiated
By the collection build-in test, the training result that obtains as shown in Figure 6.
As test data, the training result that obtains as shown in Figure 7 to utilize other 281 vehicle datas of collecting September 19 in 2009 (wherein 26 buses, 255 cars).
The above; only for the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.
Claims (5)
1. model recognizing method based on geomagnetic sensing technology is characterized in that described method comprises:
Step 1: obtain the vehicle Wave data by geomagnetic sensor;
Step 2:, exemplify waveform character according to the vehicle Wave data;
Step 3:, select effective waveform character according to the influence of waveform character to the vehicle recognition result;
Step 4: utilize effective waveform character and vehicle classification function, training obtains decision tree;
Step 5:, carry out vehicle identification according to the decision tree that obtains.
2. a kind of model recognizing method based on geomagnetic sensing technology according to claim 1 is characterized in that the specific sample frequency of described geomagnetic sensor setting.
3. a kind of model recognizing method based on geomagnetic sensing technology according to claim 1 is characterized in that described waveform character comprises crest number, maximum crest time ratio, maximum trough time ratio, average, mean square deviation, crest amplitude, trough amplitude, maximum crest value trough value ratio, peak valley distribution series.
4. a kind of model recognizing method according to claim 1 based on geomagnetic sensing technology, it is characterized in that described step 4, utilize effective waveform character and vehicle classification function, training obtains before the decision tree, also comprises effective waveform character is carried out the positive number fusion treatment.
5. a kind of model recognizing method based on geomagnetic sensing technology according to claim 1 is characterized in that described step 5 specifically is, begins search from the decision tree root that obtains, and up to leaf node, the vehicle of described leaf node correspondence is exactly a recognition result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010102398072A CN101923781B (en) | 2010-07-28 | 2010-07-28 | Vehicle type recognizing method based on geomagnetic sensing technology |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010102398072A CN101923781B (en) | 2010-07-28 | 2010-07-28 | Vehicle type recognizing method based on geomagnetic sensing technology |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101923781A true CN101923781A (en) | 2010-12-22 |
CN101923781B CN101923781B (en) | 2013-11-13 |
Family
ID=43338683
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2010102398072A Active CN101923781B (en) | 2010-07-28 | 2010-07-28 | Vehicle type recognizing method based on geomagnetic sensing technology |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101923781B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102779281A (en) * | 2012-06-25 | 2012-11-14 | 同济大学 | Vehicle type identification method based on support vector machine and used for earth inductor |
CN103345842A (en) * | 2013-07-25 | 2013-10-09 | 重庆邮电大学 | Road vehicle classification system and method |
CN104299417A (en) * | 2014-10-09 | 2015-01-21 | 武汉慧联无限科技有限公司 | Vehicle identification method based on waveform detection |
CN104766482A (en) * | 2015-01-29 | 2015-07-08 | 华南理工大学 | Traffic state detection method and device based on earth magnetic field induction |
CN105046946A (en) * | 2014-12-23 | 2015-11-11 | 北京交通大学 | Method for detecting traffic flow parameters based on compound sensor |
CN105321354A (en) * | 2015-09-22 | 2016-02-10 | 中国科学院上海微系统与信息技术研究所 | Simple vehicle detection, classification and identification method based on geomagnetic sensor |
CN107657817A (en) * | 2017-09-29 | 2018-02-02 | 北京航空航天大学 | A kind of vehicle condition detection device based on more geomagnetic sensors |
CN107886727A (en) * | 2017-11-13 | 2018-04-06 | 深圳先进技术研究院 | A kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor |
CN108509962A (en) * | 2017-02-28 | 2018-09-07 | 优信互联(北京)信息技术有限公司 | A kind of method and its device of identification information of vehicles |
CN110310490A (en) * | 2019-04-23 | 2019-10-08 | 深圳市戴升智能科技有限公司 | Vehicle speed estimation method, apparatus, computer equipment and storage medium |
CN111742356A (en) * | 2018-03-27 | 2020-10-02 | 株式会社京三制作所 | Detection system |
CN112419744A (en) * | 2020-08-27 | 2021-02-26 | 宁波大榭招商国际码头有限公司 | Container truck detection method based on waveform detection |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9361798B2 (en) | 2014-06-25 | 2016-06-07 | Global Traffic Technologies, Llc | Vehicle classification system and method |
CN104217612B (en) * | 2014-07-22 | 2016-05-25 | 刘全羽 | A kind of parking lot real-time intelligent vehicle guide system |
CN104200670B (en) * | 2014-08-26 | 2016-04-13 | 长沙理工大学 | A kind of vehicle model recognizing method based on vehicle earth induction feature |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101110107A (en) * | 2007-08-27 | 2008-01-23 | 北京交通大学 | Intelligent integrating and monitoring method for rail traffic vehicle key equipment state |
CN101751787A (en) * | 2008-12-03 | 2010-06-23 | 北京宏德信智源信息技术有限公司 | Multifunctional vehicle intelligent testing sensor and testing method thereof |
-
2010
- 2010-07-28 CN CN2010102398072A patent/CN101923781B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101110107A (en) * | 2007-08-27 | 2008-01-23 | 北京交通大学 | Intelligent integrating and monitoring method for rail traffic vehicle key equipment state |
CN101751787A (en) * | 2008-12-03 | 2010-06-23 | 北京宏德信智源信息技术有限公司 | Multifunctional vehicle intelligent testing sensor and testing method thereof |
Non-Patent Citations (1)
Title |
---|
《北京交通大学硕士学位论文》 20100215 史元超 基于新型传感器的交通状态获取技术的研究 8-51 1-5 , 2 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102779281B (en) * | 2012-06-25 | 2014-10-22 | 同济大学 | Vehicle type identification method based on support vector machine and used for earth inductor |
CN102779281A (en) * | 2012-06-25 | 2012-11-14 | 同济大学 | Vehicle type identification method based on support vector machine and used for earth inductor |
CN103345842A (en) * | 2013-07-25 | 2013-10-09 | 重庆邮电大学 | Road vehicle classification system and method |
CN104299417A (en) * | 2014-10-09 | 2015-01-21 | 武汉慧联无限科技有限公司 | Vehicle identification method based on waveform detection |
CN105046946A (en) * | 2014-12-23 | 2015-11-11 | 北京交通大学 | Method for detecting traffic flow parameters based on compound sensor |
CN105046946B (en) * | 2014-12-23 | 2017-07-14 | 北京交通大学 | Traffic flow parameters detection method based on compound sensor |
CN104766482A (en) * | 2015-01-29 | 2015-07-08 | 华南理工大学 | Traffic state detection method and device based on earth magnetic field induction |
CN105321354A (en) * | 2015-09-22 | 2016-02-10 | 中国科学院上海微系统与信息技术研究所 | Simple vehicle detection, classification and identification method based on geomagnetic sensor |
CN108509962A (en) * | 2017-02-28 | 2018-09-07 | 优信互联(北京)信息技术有限公司 | A kind of method and its device of identification information of vehicles |
CN107657817A (en) * | 2017-09-29 | 2018-02-02 | 北京航空航天大学 | A kind of vehicle condition detection device based on more geomagnetic sensors |
CN107886727A (en) * | 2017-11-13 | 2018-04-06 | 深圳先进技术研究院 | A kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor |
CN107886727B (en) * | 2017-11-13 | 2020-04-14 | 深圳先进技术研究院 | Geomagnetic sensor-based automobile classification method and system and electronic equipment |
CN111742356A (en) * | 2018-03-27 | 2020-10-02 | 株式会社京三制作所 | Detection system |
CN110310490A (en) * | 2019-04-23 | 2019-10-08 | 深圳市戴升智能科技有限公司 | Vehicle speed estimation method, apparatus, computer equipment and storage medium |
CN110310490B (en) * | 2019-04-23 | 2020-11-17 | 深圳市戴升智能科技有限公司 | Vehicle speed estimation method, vehicle speed estimation device, computer equipment and storage medium |
CN112419744A (en) * | 2020-08-27 | 2021-02-26 | 宁波大榭招商国际码头有限公司 | Container truck detection method based on waveform detection |
Also Published As
Publication number | Publication date |
---|---|
CN101923781B (en) | 2013-11-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101923781B (en) | Vehicle type recognizing method based on geomagnetic sensing technology | |
CN102779281B (en) | Vehicle type identification method based on support vector machine and used for earth inductor | |
CN106529608B (en) | A kind of telemetering motor vehicle tail equipment complex system for arranging gravity points | |
CN103345842B (en) | A kind of road vehicle classification system and method | |
Souza | Asphalt pavement classification using smartphone accelerometer and complexity invariant distance | |
CN104809180B (en) | Illegal vehicle in use recognition methods based on unsupervised intelligence learning algorithm | |
CN102829974B (en) | LMD (local mean decomposition) and PCA (principal component analysis) based rolling bearing state identification method | |
CN106856049B (en) | Key intersection demand aggregation analysis method based on bayonet number plate identification data | |
CN105320966A (en) | Vehicle driving state recognition method and apparatus | |
CN105447504A (en) | Traffic mode behavior recognition method and corresponding recognition model construction method | |
CN102778355A (en) | Rolling bearing state identification method based on empirical mode decomposition (EMD) and principal component analysis (PCA) | |
CN104408927B (en) | Model classification method based on frequency modulation continuous wave radar | |
CN103592587A (en) | Partial discharge diagnosis method based on data mining | |
CN110188797A (en) | A kind of intelligent automobile method for rapidly testing based on Bayes's optimization | |
CN101799968A (en) | Detection method and device for oil well intrusion based on video image intelligent analysis | |
CN111290370B (en) | Automatic driving performance detection method and device | |
CN103310199B (en) | Based on the vehicle model recognition methods of high-definition remote sensing data | |
CN102779410A (en) | Parallel implementation method of multi-source heterogeneous traffic data fusion | |
CN102194132A (en) | Detection and identification system and method for accompanying vehicle | |
CN104299417A (en) | Vehicle identification method based on waveform detection | |
CN103559508A (en) | Video vehicle detection method based on continuous Adaboost | |
CN107274668A (en) | A kind of congestion in road modeling method based on vehicle detection | |
CN102567715A (en) | Human body action hierarchical identification method based on pyroelectric infrared detection | |
CN115311229A (en) | Laser radar-based pavement disease detection and classification method and system and storage medium | |
CN107545244A (en) | Speed(-)limit sign detection method based on image processing techniques |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |