CN107886727A - A kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor - Google Patents
A kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor Download PDFInfo
- Publication number
- CN107886727A CN107886727A CN201711133396.7A CN201711133396A CN107886727A CN 107886727 A CN107886727 A CN 107886727A CN 201711133396 A CN201711133396 A CN 201711133396A CN 107886727 A CN107886727 A CN 107886727A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- waveform data
- frequency
- original waveform
- svm classifier
- 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
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
- G08G1/015—Detecting movement of traffic to be counted or controlled with provision for distinguishing between two or more types of vehicles, e.g. between motor-cars and cycles
-
- 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
- G08G1/042—Detecting movement of traffic to be counted or controlled using inductive or magnetic detectors
Abstract
The application is related to a kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor.The separation vehicle method based on geomagnetic sensor includes：Step a：Gather the first original waveform data and the second original waveform data of automobile respectively by two geomagnetic sensors；Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；Step c：Extract the temporal signatures value and frequency domain character value of first original waveform data and the second original waveform data；Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, separation vehicle is carried out by the svm classifier model.The vehicle classification that the application is more refined using svm classifier model according to the characteristic value of extraction, relative to prior art, vehicle classification is more, and classification accuracy rate is higher.
Description
Technical field
The application is related to technical field of intelligent traffic, more particularly to a kind of separation vehicle method based on geomagnetic sensor,
System and electronic equipment.
Background technology
Traffic study equipment abbreviation intermodulation equipment in intelligent transportation system, according to project study institute of Department of Transportation
What August in 2011 was put into effect on the 30th《" fixed intermodulation equipment and technology condition " and " fixed intermodulation equipment and data service center
Communications protocol "》(《On strengthening the instruction of highway communication condition survey management of equipment technology》(plan word [2007] 52 in the Room
Number) revised draft of text) file, intermodulation equipment needs road pavement automobile to be classified, can be divided mainly into motorcycle, medium and small visitor
Car, buggy, middle lorry, motor bus, truck and the class of super large lorry seven, shown in table specific as follows：
Table 1：Separation vehicle table
| Car category | Vehicle commander's range L (m) |
| Motorcycle | 0<L≤3 |
| Middle minibus | 3<L≤6 |
| Buggy | 3<L≤6 |
| Motor bus | 6<L≤12 |
| Middle lorry | 6<L≤12 |
| Truck | 6<L≤12 |
| Super large lorry | L>12 |
Geomagnetic sensor is the novel pavement wagon detector occurred in recent years, and vehicle classification is wagon detector
One critical function.Geomagnetic sensor mainly determines whether that car passes through by detecting the change in magnetic field.Sensed according to earth magnetism
The performance characteristics of device, Jia Ning etc. propose a kind of vehicle classification algorithm based on earth induction.The algorithm is first from initial data
Middle extraction feature, and then characteristic is clustered to determine the optimal classification ability of detector, finally with " offline instruction
The strategy of white silk, online classification ", real-time vehicle classification is carried out using neutral net, do not increasing the situation of hardware burden excessively
Under obtain relatively good effect.The algorithm is favorably improved the performance of earth magnetism wagon detector, so as to be traffic control system
More rich basic data is provided.
What will is strong etc. to propose a kind of vehicle detecting algorithm based on geomagnetic sensor, and the algorithm can be from motion tracking base
Line value simultaneously extracts signals of vehicles feature.Real road detection test data shows that the algorithm is easy and effective, and accuracy rate is reachable
More than 98.5%.The system can apply to road traffic detection, average speed statistics etc., will be sent out in intelligent transportation field
Wave important function.But the algorithm only have studied four kinds of vehicle classifications of car, SUV, bus and truck, not grind
Study carefully project study institutes of Department of Transportation requirement motorcycle, middle minibus, buggy, motor bus, middle lorry, truck and
Especially big seven kinds of vehicle classifications of lorry.
2007, Sing-Yiu Cheung existed《Traffic surveillance by wireless sensor
networks:Final report》It has studied in article and carry out vehicle disaggregated classification using geomagnetic sensor, use
Average-Bar and Hill-Pattern methods are to signal extraction feature, but its general classification accuracy only has 65%.And
And the requirement of its classification is the vehicle classification requirement according to highway office of the United States Federal, the requirement does not add motor vehicle length
Boundary.
2017, Chang Xu existed《Vehicle classification under different feature sets
with a single anisotropic magnetoresistive sensor》, use Hill-Pattern, Peak-
Peak, Mean-Std and energy eigenvalue, k nearest neighbor (K-nearest neighbor), SVMs (SVM) and BP nerve nets
Network grader carries out separation vehicle.Although accuracy only carries out 4 kinds of classification up to 83.62% to automobile.
In summary, in existing separation vehicle algorithm, classified just for part vehicle, not for traffic
Seven class vehicles (motorcycle, middle minibus, buggy, motor bus, middle lorry, the bulk production of project study institute of Department of Transportation intermodulation equipment
Car and especially big lorry) to be classified, nicety of grading is not fine enough；Meanwhile required according to the separation vehicle of intermodulation equipment, to vapour
The calculating of vehicle commander's degree is very important, and the existing separation vehicle algorithm based on geomagnetic sensor does not make full use of vapour
Car length characteristic carries out vehicle classification, and classification accuracy rate also have it is to be hoisted.
The content of the invention
This application provides a kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor, it is intended at least
Solves one of above-mentioned technical problem of the prior art to a certain extent.
In order to solve the above problems, this application provides following technical scheme：
A kind of separation vehicle method based on geomagnetic sensor, including：
Step a：Gather the first original waveform data and the second original waveform of automobile respectively by two geomagnetic sensors
Data；
Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；
Step c：The temporal signatures value and frequency domain for extracting first original waveform data and the second original waveform data are special
Value indicative；
Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, by described
Svm classifier model carries out separation vehicle.
The technical scheme that the embodiment of the present application is taken also includes：In the step a, described two geomagnetic sensors bury
It is put on trap for automobile, or is positioned over beside trap for automobile, and two geomagnetic sensors is spaced.
The technical scheme that the embodiment of the present application is taken also includes：It is described original according to described first in the step b
Wave data and the second original waveform data calculate motor vehicle length and specifically included：
Step b1：Cross-correlation calculation is done to first original waveform data and the second original waveform data, obtains automobile
Car speed；
Step b2：Interpolation is done using the car speed and normalizes the first original waveform data, obtains time domain waveform；
Step b3：Using high pass and low pass filter processing time domain waveform, the stable frequency in vehicular waveform frequency spectrum is obtained
Section；
Step b4：According to stable frequency section, according to the portion waveshape for blocking area percentage and blocking automobile head and the tailstock
Area, and motor vehicle length is calculated according to remaining waveform area.
The technical scheme that the embodiment of the present application is taken also includes：It is described original according to described first in the step b
Wave data and the second original waveform data, which calculate motor vehicle length, also to be included：Vehicle commander's range intervals according to corresponding to motor vehicle length
Carry out vehicle commander's classification.
The technical scheme that the embodiment of the present application is taken also includes：In the step c, the first original waveform of the extraction
The temporal signatures value and frequency domain character value of data and the second original waveform data specifically include：
Step c1：According to the time domain waveform, automobile head and the tailstock are distinguished into cut through partial wave according to threshold values is blocked
Shape, and average, mean square deviation, center of gravity and discreteness are calculated according to residual waveform, obtain 4 characteristic values of time domain；
Step c2：The time domain waveform Fast Fourier Transform (FFT) is obtained into frequency-region signal, and the frequency-region signal is carried out
Amplitude normalizes, and obtains normalizing frequency-region signal；
Step c3：The low frequency waveform for intercepting the normalization frequency-region signal obtains low frequency frequency-region signal；
Step c4：Average, mean square deviation, center of gravity and the discreteness of the low frequency frequency-region signal are calculated, obtains 4 of frequency domain
Characteristic value.
The technical scheme that the embodiment of the present application is taken also includes：It is described to be entered by svm classifier model in the step d
Row separation vehicle is specially：3 svm classifiers are trained using 4 characteristic values of time domain, 4 characteristic values of frequency domain and motor vehicle length
Model, 3 svm classifier models be respectively the first svm classifier model model1, the second svm classifier model model2 and
3rd svm classifier model model3；It is 3 to vehicle commander's range intervals by the first svm classifier model model1<L≤6 it is medium and small
Car and buggy are classified, and are 6 to vehicle commander's range intervals by the second svm classifier model model2<The bus of L≤12
Car is classified with lorry, then by the 3rd svm classifier model model3 to vehicle commander's range intervals be 6<The middle lorry of L≤12
Classified with truck.
Another technical scheme that the embodiment of the present application is taken is：A kind of separation vehicle system based on geomagnetic sensor, bag
Include：
Data acquisition module：For gathering the first original waveform data and the second original waveform data of automobile respectively；
Vehicle commander's computing module：For calculating automobile according to first original waveform data and the second original waveform data
Length；
Time-domain and frequency-domain characteristic extracting module：For extracting first original waveform data and the second original waveform data
Temporal signatures value and frequency domain character value；
Vehicle classification module：For the motor vehicle length, temporal signatures value and frequency domain character value to be inputted into svm classifier mould
Type, separation vehicle is carried out by the svm classifier model.
The technical scheme that the embodiment of the present application is taken also includes：The data acquisition module is two geomagnetic sensors, institute
State two geomagnetic sensors and bury and be put on trap for automobile, or be positioned over beside trap for automobile, and two geomagnetic sensors are mutual
Every.
The technical scheme that the embodiment of the present application is taken also includes：Vehicle commander's computing module includes：
Speed computing unit：Based on doing cross-correlation to first original waveform data and the second original waveform data
Calculate, obtain the car speed of automobile；
Time domain waveform computing unit：The first original waveform data is normalized for doing interpolation using the car speed, is obtained
To time domain waveform；
Filter computing unit：For using high pass and low pass filter processing time domain waveform, obtaining in vehicular waveform frequency spectrum
Stable frequency section；
Vehicle commander's computing unit：For according to stable frequency section, automobile head and the tailstock to be blocked according to area percentage is blocked
Portion waveshape area, and motor vehicle length is calculated according to remaining waveform area.
The technical scheme that the embodiment of the present application is taken also includes：Vehicle commander's computing module also includes vehicle commander's taxon,
Vehicle commander's taxon is used for vehicle commander's range intervals according to corresponding to motor vehicle length and carries out vehicle commander's classification.
The technical scheme that the embodiment of the present application is taken also includes：The time-domain and frequency-domain characteristic extracting module includes：
Temporal signatures extraction unit：For according to the time domain waveform, automobile head and the tailstock to be divided according to threshold values is blocked
Other truncation part waveform, and average, mean square deviation, center of gravity and discreteness are calculated according to residual waveform, obtain 4 features of time domain
Value；
Normalization unit：For the time domain waveform Fast Fourier Transform (FFT) to be obtained into frequency-region signal, and by the frequency domain
Signal carries out amplitude normalization, obtains normalizing frequency-region signal；
Low frequency signal interception unit：Low frequency waveform for intercepting the normalization frequency-region signal obtains low frequency frequency domain letter
Number；
Frequency domain character extraction unit：For calculating average, mean square deviation, center of gravity and the discreteness of the low frequency frequency-region signal,
Obtain 4 characteristic values of frequency domain.
The technical scheme that the embodiment of the present application is taken also includes：The vehicle classification module is carried out by svm classifier model
Separation vehicle is specially：3 svm classifier moulds are trained using 4 characteristic values of time domain, 4 characteristic values of frequency domain and motor vehicle length
Type, 3 svm classifier models are respectively the first svm classifier model model1, the second svm classifier model model2 and
Three svm classifier model model3；It is 3 to vehicle commander's range intervals by the first svm classifier model model1<The medium and small visitor of L≤6
Car and buggy are classified, and are 6 to vehicle commander's range intervals by the 2nd SVM disaggregated models model2<The motor bus of L≤12
Classified with lorry, then by the 3rd svm classifier model model3 to vehicle commander's range intervals be 6<The middle lorry of L≤12 with
Truck is classified.
The another technical scheme that the embodiment of the present application is taken is：A kind of electronic equipment, including：
At least one processor；And
The memory being connected with least one processor communication；Wherein,
The memory storage has can be by the instruction of one computing device, and the instruction is by described at least one
Computing device, so that at least one processor is able to carry out the above-mentioned separation vehicle method based on geomagnetic sensor
Following operation：
Step a：Gather the first original waveform data and the second original waveform of automobile respectively by two geomagnetic sensors
Data；
Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；
Step c：The temporal signatures value and frequency domain for extracting first original waveform data and the second original waveform data are special
Value indicative；
Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into SVM disaggregated models, by described
Svm classifier model carries out separation vehicle.
Relative to prior art, beneficial effect caused by the embodiment of the present application is：The embodiment of the present application based on earth magnetism
Separation vehicle method, system and the electronic equipment of sensor obtain automobile length first by stable energy frequency spectrum method is found
Degree and length classification, then obtain geomagnetic sensor signal time domain, frequency-domain waveform geometric properties using normalization spacetime geometry method
Value, the vehicle classification finally more refined according to the characteristic value of extraction using svm classifier model, relative to prior art,
Vehicle classification is more, and classification accuracy rate is higher.
Brief description of the drawings
Fig. 1 is the flow chart of the separation vehicle method based on geomagnetic sensor of the embodiment of the present application；
Fig. 2 is the flow chart of vehicle commander's sorting technique of the embodiment of the present application；
Fig. 3 is the flow chart of the normalization spacetime geometry algorithm of the embodiment of the present application；
Fig. 4 is the structure chart of the svm classifier model of the embodiment of the present application；
Fig. 5 is the structural representation of the separation vehicle system based on geomagnetic sensor of the embodiment of the present application；
Fig. 6 is that the hardware device structure for the separation vehicle method based on geomagnetic sensor that the embodiment of the present application provides is shown
It is intended to.
Embodiment
In order that the object, technical solution and advantage of the application are more clearly understood, below in conjunction with drawings and Examples,
The application is further elaborated.It should be appreciated that specific embodiment described herein is only to explain the application,
It is not used to limit the application.
Referring to Fig. 1, it is the flow chart of the separation vehicle method based on geomagnetic sensor of the embodiment of the present application.This Shen
Please the separation vehicle method based on geomagnetic sensor of embodiment comprise the following steps：
Step 100：By two spaced geomagnetic sensors gather respectively automobile the first original waveform data and
Second original waveform data；
In step 100, two geomagnetic sensors, which can bury, is placed on trap for automobile, can also be placed on by trap for automobile
Side.The acquisition mode of first original waveform data and the second original waveform data is：When automobile is by two geomagnetic sensors,
Pass through ADC (Analog-to-Digital Converter, A/D converter or the modulus of two geomagnetic sensors respectively
Converter) collector samples to magnitude of voltage caused by geomagnetic sensor, sample frequency 1kHz, obtain two it is substantially similar
The first original waveform data and the second original waveform data, and by the first original waveform data and the second original waveform data
It is set to x1And x2.In the embodiment of the present application, the spaced distance of two geomagnetic sensors is 1 meter, specifically can be according to reality
Border operation is set.
Step 200：According to the first original waveform data and the second original waveform data of collection, stable using searching
Energy frequency spectrum method calculates motor vehicle length, and vehicle commander's range intervals carry out vehicle commander's classification to automobile according to corresponding to motor vehicle length；
In order to clearly illustrate step 200, referring to Fig. 2, being the flow of vehicle commander's sorting technique of the embodiment of the present application
Figure.Vehicle commander's sorting technique of the embodiment of the present application comprises the following steps：
Step 201：To the first original waveform data x1With the second original waveform data x2Cross-correlation calculation is done, obtains the vapour
Car speed corresponding to car；
Step 202：Interpolation, which is done, using car speed normalizes the first original waveform data x1, obtain time domain waveform x3；
In step 202, because earth magnetism voltage signal sampling rate is constant, and automobile by the speed of geomagnetic sensor not
Together, therefore, it is necessary to vehicular waveform data are normalized with each car speed, it is the later stage so that release rate influences
Computational length and classification are prepared.
Step 203：Use high pass and low pass filter processing time domain waveform x3, obtain the stabilization in vehicular waveform frequency spectrum
Frequency band x4；
Step 204：According to stable frequency section x4, automobile head and the tailstock are blocked according to the area percentage that blocks of setting
Portion waveshape area, and motor vehicle length L is calculated according to remaining waveform area；
In step 204, blocking the concrete numerical value of area percentage can be set according to practical operation, in the application
In embodiment, block area percentage and be set as such as 4%.
Step 205：Four class vehicle commander range intervals carry out vehicle commander's classification according to corresponding to motor vehicle length L；
In step 205, four class vehicle commander range intervals progress vehicle commander's classification is specially according to corresponding to motor vehicle length L：Will
Motorcycle, middle minibus, buggy, motor bus, middle lorry, truck, the class vehicle of especially big lorry seven respectively with A, B, C, D, E,
F, G is represented, four length range sections 0<L≤3、3<L≤6、6<L≤12 and L>12 are represented with T1, T2, T3 and T4 respectively,
Car category and its length range interval table are as shown in table 2 below：
Table 2：Car category and its length range interval table
0 is directed to by using the method for finding stable frequency spectrum<L≤3、3<L≤6、 6<L≤12 and L>12 4 length models
Enclose section and carry out vehicle commander's classification, confusion matrix accuracy is more than 95%, shown in vehicle commander's classification confusion matrix table 3 specific as follows：
Table 3：Vehicle commander's classification confusion matrix
| T1 | T2 | T3 | T4 | |
| T1 | 15 | 0 | 0 | 0 |
| T2 | 0 | 806 | 35 | 0 |
| T3 | 0 | 16 | 255 | 2 |
| T4 | 0 | 0 | 1 | 23 |
Step 206：Vehicle commander's classification accuracy rate is counted, completes an iteration；
In step 206, through statistics, four length range sections 0<L≤3、 3<L≤6、6<L≤12 and L>Corresponding to 12
Vehicle commander's classification accuracy rate is respectively 100%, 95.84%, 93.41% and 95.83%, shown in table 4 specific as follows：
Table 4：Vehicle commander's classification accuracy rate
| Vehicle commander's type | T1 | T2 | T3 | T4 |
| Vehicle commander's classification accuracy rate | 100% | 95.84% | 93.41% | 95.83% |
Step 207：Change low pass, high pass cut off frequency and block area percentage, and re-execute step 203, directly
Vehicle commander's classification accuracy rate highest filter coefficient and area percentage is blocked to choosing.
Step 300：Time domain waveform x is calculated using normalization spacetime geometry algorithm3Temporal signatures value and frequency domain character
Value；
In step 300, it is the flow of the normalization spacetime geometry algorithm of the embodiment of the present application also referring to Fig. 3
Figure.The embodiment of the present application calculates time domain waveform x using normalization spacetime geometry algorithm3Temporal signatures value and frequency domain character
The mode of value comprises the following steps：
Step 301：According to time domain waveform x3, automobile head and the tailstock are blocked less than the default portion for blocking threshold values respectively
Partial wave shape, and average, mean square deviation, center of gravity and discreteness are calculated according to residual waveform, obtain 4 characteristic values of time domain；
In step 301, if time domain waveform x3Average beMeansquaredeviationσt, time domain waveform x3Length is n, then time domain
Waveform x3Center of gravity calculation formula it is as follows：
Time domain waveform x3Discreteness calculation formula it is as follows：
Step 302：By time domain waveform x3Fast Fourier Transform (FFT) obtains frequency domain signal X3, and by frequency domain signal X3Carry out width
Value normalization, obtain normalizing frequency domain signal X4；
Step 303：Interception normalization frequency domain signal X4Part low frequency (such as 1 to 40Hz) waveform obtain low frequency frequency domain letter
Number X5；
Step 304：Calculate low frequency frequency domain signal X5Average, mean square deviation, center of gravity and discreteness, obtain 4 spies of frequency domain
Value indicative；
In step 304, if low frequency frequency domain signal X5Average beMeansquaredeviationσf, low frequency frequency domain signal X5Length is
N, then low frequency frequency domain signal X5Center of gravity calculation formula it is as follows：
Low frequency frequency domain signal X5Discreteness calculation formula it is as follows：
Step 400：Svm classifier model is built according to the motor vehicle length of extraction, temporal signatures value and frequency domain character value, led to
Cross the separation vehicle that svm classifier model carries out seven kinds of vehicles；
In step 400, the structure of svm classifier model is as shown in Figure 4.The building mode of svm classifier model is specially：
Using 4 characteristic values of time domain, 4 characteristic values of frequency domain and motor vehicle length train 3 svm classifier models, respectively first
Svm classifier model model1, the second svm classifier model model2 and the 3rd svm classifier model model3；Pass through the first SVM
Disaggregated model model1 is 3 to vehicle commander's range intervals<The middle minibus and buggy of L≤6 are classified, and pass through the 2nd SVM
Disaggregated model model2 is 6 to vehicle commander's range intervals<The motor bus of L≤12 is classified with lorry (middle goods, bulk production), then
Further classified with truck by the 3rd svm classifier model model3 centerings lorry, so as to obtain seven kinds of vehicles
Separation vehicle result.By using svm classifier model to motorcycle, middle minibus, buggy, motor bus, middle lorry, big
Lorry and especially big lorry carry out vehicle classification, confusion matrix accuracy about 90%.Specific separation vehicle confusion matrix such as following table
Shown in 5：
Table 5：Separation vehicle confusion matrix
Motorcycle, middle minibus, buggy, motor bus, middle lorry, truck and especially big lorry use svm classifier model
The separation vehicle accuracy classified is respectively 100%, 98.3%, 93.8%, 86.8%, 77.6%, 72.2% and
95.8%, shown in table 6 specific as follows：
Table 6：Separation vehicle accuracy
| Car category | A | B | C | D | E | F | G |
| Accuracy | 100% | 98.3% | 93.8% | 86.8% | 77.6% | 72.2% | 95.8% |
Referring to Fig. 5, it is the structural representation of the separation vehicle system based on geomagnetic sensor of the embodiment of the present application.
The separation vehicle system based on geomagnetic sensor of the embodiment of the present application includes data acquisition module, vehicle commander's computing module, time domain
Frequency domain character extraction module and vehicle classification module.
Data acquisition module：For gathering the first original waveform data and the second original waveform data of automobile；Wherein,
Data acquisition module is two spaced geomagnetic sensors, and two geomagnetic sensors, which can bury, to be placed on trap for automobile,
It can be placed on beside trap for automobile.The original waveform data of data collecting module collected first and the second original waveform data are adopted
Mode set is：When automobile is by two geomagnetic sensors, pass through the ADC (Analog-to- of two geomagnetic sensors respectively
Digital Converter, A/D converter or analog-digital converter) collector enters to magnitude of voltage caused by geomagnetic sensor
Row sampling, sample frequency 1kHz, obtains two substantially similar the first original waveform datas and the second original waveform data, and
First original waveform data and the second original waveform data are set to x1And x2.In the embodiment of the present application, two earth magnetism pass
The spaced distance of sensor is 1 meter, can specifically be set according to practical operation.
Vehicle commander's computing module：For the first original waveform data and the second original waveform data according to collection, using seeking
Stable energy frequency spectrum method is looked for calculate motor vehicle length, and vehicle commander's range intervals are carried out to automobile according to corresponding to motor vehicle length
Vehicle commander classifies；Specifically, vehicle commander's computing module includes：
Speed computing unit：For to the first original waveform data x1With the second original waveform data x2Make cross-correlation meter
Calculate, obtain car speed corresponding to the automobile；
Time domain waveform computing unit：The first original waveform data x is normalized for doing interpolation using car speed1, obtain
Time domain waveform x3；Wherein, because earth magnetism voltage signal sampling rate is constant, and automobile is different by the speed of geomagnetic sensor, because
This is, it is necessary to vehicular waveform data are normalized with each car speed, so that release rate influences, for later stage calculating
Length and classification are prepared.
Filter computing unit：For using high pass and low pass filter processing time domain waveform x3, obtain vehicular waveform frequency spectrum
In stable frequency section x4；
Vehicle commander's computing unit：For according to stable frequency section x4, Automobile is blocked according to the area percentage that blocks of setting
The portion waveshape area of head and the tailstock, and motor vehicle length L is calculated according to remaining waveform area；Wherein, area hundred is blocked
The concrete numerical value of point ratio can be set according to practical operation, in the embodiment of the present application, block area percentage be set as
4%.
Vehicle commander's taxon：Vehicle commander's classification is carried out for the four class vehicle commander range intervals according to corresponding to motor vehicle length L；Its
In, the four class vehicle commander range intervals according to corresponding to motor vehicle length L carry out vehicle commander's classification and are specially：By motorcycle, middle minibus,
Buggy, motor bus, middle lorry, truck, the class vehicle of especially big lorry seven are represented with A, B, C, D, E, F, G respectively, four long
Spend range intervals 0<L≤3、3<L≤6、6<L≤12 and L>12 are represented with T1, T2, T3 and T4 respectively, car category and its length
Range intervals table is as shown in table 2 below：
Table 2：Car category and its length range interval table
| Car category | Car category code name | Vehicle commander interval range L (m) | Vehicle commander's type |
| Motorcycle | A | 0<L≤3 | T1 |
| Middle minibus | B | 3<L≤6 | T2 |
| Buggy | C | 3<L≤6 | T2 |
| Motor bus | D | 6<L≤12 | T3 |
| Middle lorry | E | 6<L≤12 | T3 |
| Truck | F | 6<L≤12 | T3 |
| Super large lorry | G | L>12 | T4 |
0 is directed to by using the method for finding stable frequency spectrum<L≤3、3<L≤6、 6<L≤12 and L>12 4 length models
Enclose section and carry out vehicle commander's classification, confusion matrix accuracy is more than 95%, shown in vehicle commander's classification confusion matrix table 3 specific as follows：
Table 3：Vehicle commander's classification confusion matrix
| T1 | T2 | T3 | T4 | |
| T1 | 15 | 0 | 0 | 0 |
| T2 | 0 | 806 | 35 | 0 |
| T3 | 0 | 16 | 255 | 2 |
| T4 | 0 | 0 | 1 | 23 |
Accuracy statistic unit：For being counted to vehicle commander's classification accuracy rate, an iteration is completed；Wherein, through system
Meter, four length range sections 0<L≤3、3<L≤6、 6<L≤12 and L>Vehicle commander's classification accuracy rate is respectively corresponding to 12
100%th, 95.84%, 93.41% and 95.83%, shown in table 4 specific as follows：
Table 4：Vehicle commander's classification accuracy rate
| Vehicle commander's type | T1 | T2 | T3 | T4 |
| Vehicle commander's classification accuracy rate | 100% | 95.84% | 93.41% | 95.83% |
Iteration unit：For changing low pass, high pass cut off frequency and blocking area percentage, and pass through filter unit
Again to time domain waveform x3Handled, until choosing vehicle commander's classification accuracy rate highest filter coefficient and blocking area
Percentage.
Time-domain and frequency-domain characteristic extracting module：For calculating time domain waveform x using normalization spacetime geometry algorithm3Time domain
Characteristic value and frequency domain character value；Specifically, time-domain and frequency-domain characteristic extracting module includes：
Temporal signatures extraction unit：For according to time domain waveform x3, automobile head and the tailstock are blocked less than default respectively
The portion waveshape of threshold values is blocked, and average, mean square deviation, center of gravity and discreteness are calculated according to residual waveform, obtains 4 of time domain
Characteristic value；Wherein, if time domain waveform x3Average beMeansquaredeviationσt, time domain waveform x3Length is n, then time domain waveform x3's
Center of gravity calculation formula is as follows：
Time domain waveform x3Discreteness calculation formula it is as follows：
Normalization unit：For by time domain waveform x3Fast Fourier Transform (FFT) obtains frequency domain signal X3, and by frequency-region signal
X3Amplitude normalization is carried out, obtains normalizing frequency domain signal X4；
Low frequency signal interception unit：Frequency domain signal X is normalized for intercepting4Part low frequency (such as 1 to 40Hz) waveform obtain
To low frequency frequency domain signal X5；
Frequency domain character extraction unit：For calculating low frequency frequency domain signal X5Average, mean square deviation, center of gravity and discreteness, obtain
Obtain 4 characteristic values of frequency domain；Wherein, if low frequency frequency domain signal X5Average beMeansquaredeviationσf, low frequency frequency domain signal X5Length
For N, then low frequency frequency domain signal X5Center of gravity calculation formula it is as follows：
Low frequency frequency domain signal X5Discreteness calculation formula it is as follows：
Vehicle classification module：Divide for the motor vehicle length according to extraction, temporal signatures value and frequency domain character value structure SVM
Class model, the separation vehicle of seven kinds of vehicles is carried out by svm classifier model；The building mode of svm classifier model is specially：Profit
3 svm classifier models, respectively the first SVM are trained with 4 characteristic values of time domain, 4 characteristic values of frequency domain and motor vehicle length
Disaggregated model model1, the second svm classifier model model2 and the 3rd svm classifier model model3；Pass through the first svm classifier
Model model1 is 3 to vehicle commander's range intervals<The middle minibus and buggy of L≤6 are classified, and pass through the second svm classifier
Model model2 is 6 to vehicle commander's range intervals<The motor bus of L≤12 is classified with lorry (middle goods, bulk production), then is passed through
3rd svm classifier model model3 centerings lorry is further classified with truck, so as to obtain the automobile of seven kinds of vehicles
Classification results.By using svm classifier model to motorcycle, middle minibus, buggy, motor bus, middle lorry, truck and
Especially big lorry carries out vehicle classification, confusion matrix accuracy about 90%.Specific separation vehicle confusion matrix is as shown in table 5 below：
Table 5：Separation vehicle confusion matrix
| A | B | C | D | E | F | G | |
| A | 15 | 0 | 0 | 0 | 0 | 0 | 0 |
| B | 0 | 499 | 12 | 12 | 8 | 1 | 0 |
| C | 0 | 8 | 287 | 5 | 3 | 6 | 0 |
| D | 0 | 2 | 2 | 66 | 3 | 3 | 0 |
| E | 0 | 2 | 5 | 9 | 125 | 19 | 1 |
| F | 0 | 0 | 5 | 2 | 2 | 26 | 1 |
| G | 0 | 0 | 0 | 0 | 0 | 1 | 23 |
Motorcycle, middle minibus, buggy, motor bus, middle lorry, truck and especially big lorry use svm classifier model
The separation vehicle accuracy classified is respectively 100%, 98.3%, 93.8%, 86.8%, 77.6%, 72.2% and
95.8%, shown in table 6 specific as follows：
Table 6：Separation vehicle accuracy
| Car category | A | B | C | D | E | F | G |
| Accuracy | 100% | 98.3% | 93.8% | 86.8% | 77.6% | 72.2% | 95.8% |
The application is trained by using the Wave data of 1000 automobiles to svm classifier model, and uses 1153
Car is tested, and the results show the application can be very good to search out the stable wavelength coverage of geomagnetic sensor, steady based on this
Determine wavelength coverage, vehicle commander is calculated using the method for blocking area from beginning to end, and four class vehicle commanders classification is carried out according to vehicle commander, according to vehicle commander
The four class vehicle classification accuracy carried out use depot, temporal signatures value and frequency domain character value to carry out vehicle more than 95%
The final overall classification accuracy rate about 90% of the svm classifier model of classification, relative to prior art, vehicle classification is more, classification
Accuracy is higher.
Fig. 6 is that the hardware device structure for the separation vehicle method based on geomagnetic sensor that the embodiment of the present application provides is shown
It is intended to.As shown in fig. 6, the equipment includes one or more processors and memory.By taking a processor as an example, the equipment
It can also include：Input system and output system.
Processor, memory, input system and output system can be connected by bus or other modes, in Fig. 6 with
Exemplified by being connected by bus.
Memory as a kind of non-transient computer readable storage medium storing program for executing, available for store non-transient software program, it is non-temporarily
State computer executable program and module.Processor is by running storage non-transient software program in memory, instruction
And module, so as to perform the various function application of electronic equipment and data processing, that is, realize the place of above method embodiment
Reason method.
Memory can include storing program area and storage data field, wherein, storing program area can storage program area,
Application program required at least one function；Storage data field can data storage etc..In addition, memory can be included at a high speed
Random access memory, can also include non-transient memory, a for example, at least disk memory, flush memory device or its
His non-transient solid-state memory.In certain embodiments, memory is optional including relative to the remotely located storage of processor
Device, these remote memories can pass through network connection to processing system.The example of above-mentioned network includes but is not limited to interconnect
Net, intranet, LAN, mobile radio communication and combinations thereof.
Input system can receive the numeral or character information of input, and produce signal input.Output system may include to show
The display devices such as display screen.
One or more of modules are stored in the memory, are held when by one or more of processors
During row, the following operation of any of the above-described embodiment of the method is performed：
Step a：Gather the first original waveform data and the second original waveform of automobile respectively by two geomagnetic sensors
Data；
Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；
Step c：The temporal signatures value and frequency domain for extracting first original waveform data and the second original waveform data are special
Value indicative；
Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, by described
Svm classifier model carries out separation vehicle.
The said goods can perform the method that is provided of the embodiment of the present application, possess the corresponding functional module of execution method and
Beneficial effect.Not ins and outs of detailed description in the present embodiment, reference can be made to the method that the embodiment of the present application provides.
The embodiment of the present application provides a kind of non-transient (non-volatile) computer-readable storage medium, the computer storage
Media storage has computer executable instructions, the executable following operation of the computer executable instructions：
Step a：Gather the first original waveform data and the second original waveform of automobile respectively by two geomagnetic sensors
Data；
Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；
Step c：The temporal signatures value and frequency domain for extracting first original waveform data and the second original waveform data are special
Value indicative；
Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, by described
Svm classifier model carries out separation vehicle.
The embodiment of the present application provides a kind of computer program product, and the computer program product is non-including being stored in
Computer program in transitory computer readable storage medium, the computer program include programmed instruction, when described program refers to
When order is computer-executed, the computer is set to perform following operate：
Step a：Gather the first original waveform data and the second original waveform of automobile respectively by two geomagnetic sensors
Data；
Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；
Step c：The temporal signatures value and frequency domain for extracting first original waveform data and the second original waveform data are special
Value indicative；
Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, by described
Svm classifier model carries out separation vehicle.
Separation vehicle method, system and the electronic equipment based on geomagnetic sensor of the embodiment of the present application is first by seeking
Look for stable energy frequency spectrum method to obtain motor vehicle length and length classification, then obtain earth magnetism using normalization spacetime geometry method
Sensor signal time domain, frequency-domain waveform geometrical characteristic, finally carried out more according to the characteristic value of extraction using svm classifier model
The vehicle classification of refinement, relative to prior art, vehicle classification is more, and classification accuracy rate is higher.
The foregoing description of the disclosed embodiments, professional and technical personnel in the field are enable to realize or using the application.
A variety of modifications to these embodiments will be apparent for those skilled in the art, defined herein
General Principle can be realized in other embodiments in the case where not departing from spirit herein or scope.Therefore, originally
Application is not intended to be limited to the embodiments shown herein, and is to fit to special with principles disclosed herein and novelty
The consistent most wide scope of point.
Claims (13)
- A kind of 1. separation vehicle method based on geomagnetic sensor, it is characterised in that including：Step a：Gather the first original waveform data and the second original waveform data of automobile respectively by two geomagnetic sensors；Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；Step c：Extract the temporal signatures value and frequency domain character value of first original waveform data and the second original waveform data；Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, pass through the SVM points Class model carries out separation vehicle.
- 2. the separation vehicle method according to claim 1 based on geomagnetic sensor, it is characterised in that in the step a In, described two geomagnetic sensors, which bury, to be put on trap for automobile, or is positioned over beside trap for automobile, and two geomagnetic sensor phases Mutually interval.
- 3. the separation vehicle method according to claim 2 based on geomagnetic sensor, it is characterised in that in the step b In, it is described to be specifically included according to first original waveform data and the second original waveform data calculating motor vehicle length：Step b1：Cross-correlation calculation is done to first original waveform data and the second original waveform data, obtains the vapour of automobile Vehicle speed；Step b2：Interpolation is done using the car speed and normalizes the first original waveform data, obtains time domain waveform；Step b3：Using high pass and low pass filter processing time domain waveform, the stable frequency section in vehicular waveform frequency spectrum is obtained；Step b4：According to stable frequency section, according to the portion waveshape area for blocking area percentage and blocking automobile head and the tailstock, And motor vehicle length is calculated according to remaining waveform area.
- 4. the separation vehicle method according to claim 3 based on geomagnetic sensor, it is characterised in that in the step b In, it is described also to be included according to first original waveform data and the second original waveform data calculating motor vehicle length：According to automobile Vehicle commander's range intervals corresponding to length carry out vehicle commander's classification.
- 5. the separation vehicle method according to claim 3 based on geomagnetic sensor, it is characterised in that in the step c In, the temporal signatures value and frequency domain character value of the first original waveform data of the extraction and the second original waveform data are specifically wrapped Include：Step c1：According to the time domain waveform, automobile head and the tailstock are distinguished into truncation part waveform, and root according to threshold values is blocked Average, mean square deviation, center of gravity and discreteness are calculated according to residual waveform, obtains 4 characteristic values of time domain；Step c2：The time domain waveform Fast Fourier Transform (FFT) is obtained into frequency-region signal, and the frequency-region signal is subjected to amplitude Normalization, obtain normalizing frequency-region signal；Step c3：The low frequency waveform for intercepting the normalization frequency-region signal obtains low frequency frequency-region signal；Step c4：Average, mean square deviation, center of gravity and the discreteness of the low frequency frequency-region signal are calculated, obtains 4 features of frequency domain Value.
- 6. the separation vehicle method according to claim 5 based on geomagnetic sensor, it is characterised in that in the step d In, it is described to be specially by svm classifier model progress separation vehicle：Utilize 4 characteristic values of time domain, 4 characteristic values of frequency domain 3 svm classifier models are trained with motor vehicle length, 3 svm classifier models are respectively the first svm classifier model model1, the Two svm classifier model model2 and the 3rd svm classifier model model3；By the first svm classifier model model1 to vehicle commander's model Section is enclosed for 3<The middle minibus and buggy of L≤6 are classified, by the second svm classifier model model2 to vehicle commander's scope Section is 6<The motor bus of L≤12 is classified with lorry, then by the 3rd svm classifier model model3 to vehicle commander's range intervals For 6<Middle lorry and the truck of L≤12 are classified.
- A kind of 7. separation vehicle system based on geomagnetic sensor, it is characterised in that including：Data acquisition module：For gathering the first original waveform data and the second original waveform data of automobile respectively；Vehicle commander's computing module：For calculating motor vehicle length according to first original waveform data and the second original waveform data；Time-domain and frequency-domain characteristic extracting module：For extracting the time domain of first original waveform data and the second original waveform data Characteristic value and frequency domain character value；Vehicle classification module：For the motor vehicle length, temporal signatures value and frequency domain character value to be inputted into svm classifier model, lead to Cross the svm classifier model and carry out separation vehicle.
- 8. the separation vehicle system according to claim 7 based on geomagnetic sensor, it is characterised in that the data acquisition Module is two geomagnetic sensors, and described two geomagnetic sensors, which bury, to be put on trap for automobile, or is positioned over beside trap for automobile, And two geomagnetic sensors are spaced.
- 9. the separation vehicle system according to claim 8 based on geomagnetic sensor, it is characterised in that the vehicle commander calculates Module includes：Speed computing unit：For doing cross-correlation calculation to first original waveform data and the second original waveform data, obtain Obtain the car speed of automobile；Time domain waveform computing unit：The first original waveform data is normalized for doing interpolation using the car speed, when obtaining Domain waveform；Filter computing unit：For using high pass and low pass filter processing time domain waveform, obtaining steady in vehicular waveform frequency spectrum Determine frequency band；Vehicle commander's computing unit：For according to stable frequency section, according to the portion for blocking area percentage and blocking automobile head and the tailstock Divide waveform area, and motor vehicle length is calculated according to remaining waveform area.
- 10. the separation vehicle system according to claim 9 based on geomagnetic sensor, it is characterised in that vehicle commander's meter Calculating module also includes vehicle commander's taxon, and vehicle commander's taxon is entered for vehicle commander's range intervals according to corresponding to motor vehicle length The long classification of driving.
- 11. the separation vehicle system according to claim 9 based on geomagnetic sensor, it is characterised in that the time domain frequency Characteristic of field extraction module includes：Temporal signatures extraction unit：For according to the time domain waveform, automobile head and the tailstock to be cut respectively according to threshold values is blocked Disconnected portion waveshape, and average, mean square deviation, center of gravity and discreteness are calculated according to residual waveform, obtain 4 characteristic values of time domain；Normalization unit：For the time domain waveform Fast Fourier Transform (FFT) to be obtained into frequency-region signal, and by the frequency-region signal Amplitude normalization is carried out, obtains normalizing frequency-region signal；Low frequency signal interception unit：Low frequency waveform for intercepting the normalization frequency-region signal obtains low frequency frequency-region signal；Frequency domain character extraction unit：For calculating average, mean square deviation, center of gravity and the discreteness of the low frequency frequency-region signal, obtain 4 characteristic values of frequency domain.
- 12. the separation vehicle system according to claim 11 based on geomagnetic sensor, it is characterised in that the vehicle point Generic module carries out separation vehicle by svm classifier model：Using 4 characteristic values of time domain, 4 characteristic values of frequency domain and Motor vehicle length trains 3 svm classifier models, and 3 svm classifier models are respectively the first svm classifier model model1, second Svm classifier model model2 and the 3rd svm classifier model model3；By the first svm classifier model model1 to vehicle commander's scope Section is 3<The middle minibus and buggy of L≤6 are classified, by the second svm classifier model model2 to vehicle commander's scope area Between be 6<The motor bus of L≤12 is classified with lorry, then is to vehicle commander's range intervals by the 3rd svm classifier model model3 6<Middle lorry and the truck of L≤12 are classified.
- 13. a kind of electronic equipment, including：At least one processor；AndThe memory being connected with least one processor communication；Wherein,The memory storage has can be by the instruction of one computing device, and the instruction is by least one processor Perform, so that at least one processor is able to carry out the automobile based on geomagnetic sensor point described in above-mentioned 1 to 6 any one The following operation of class method：Step a：Gather the first original waveform data and the second original waveform data of automobile respectively by two geomagnetic sensors；Step b：Motor vehicle length is calculated according to first original waveform data and the second original waveform data；Step c：Extract the temporal signatures value and frequency domain character value of first original waveform data and the second original waveform data；Step d：The motor vehicle length, temporal signatures value and frequency domain character value are inputted into svm classifier model, pass through the SVM points Class model carries out separation vehicle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711133396.7A CN107886727B (en) | 2017-11-13 | 2017-11-13 | Geomagnetic sensor-based automobile classification method and system and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711133396.7A CN107886727B (en) | 2017-11-13 | 2017-11-13 | Geomagnetic sensor-based automobile classification method and system and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107886727A true CN107886727A (en) | 2018-04-06 |
| CN107886727B CN107886727B (en) | 2020-04-14 |
Family
ID=61776812
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711133396.7A Active CN107886727B (en) | 2017-11-13 | 2017-11-13 | Geomagnetic sensor-based automobile classification method and system and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107886727B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816992A (en) * | 2018-12-13 | 2019-05-28 | 杭州电子科技大学 | A kind of model recognizing method based on Geomagnetic signal |
| CN110491141A (en) * | 2019-08-22 | 2019-11-22 | 长安大学 | A kind of information of vehicles identifying system and recognition methods |
| CN111524365A (en) * | 2020-05-11 | 2020-08-11 | 西安电子科技大学 | Method for classifying vehicle types by using multiple geomagnetic sensors |
| CN112611313A (en) * | 2020-12-11 | 2021-04-06 | 交控科技股份有限公司 | Rail vehicle length measuring device, rail vehicle length measuring method and traffic facility |
| CN114202927A (en) * | 2021-12-27 | 2022-03-18 | 招商局重庆公路工程检测中心有限公司 | Vehicle type detection method based on multi-sensor fusion |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11271441A (en) * | 1998-03-23 | 1999-10-08 | Hino Motors Ltd | On-vehicle radar apparatus |
| US6014447A (en) * | 1997-03-20 | 2000-01-11 | Raytheon Company | Passive vehicle classification using low frequency electro-magnetic emanations |
| CN101110161A (en) * | 2007-08-31 | 2008-01-23 | 北京科技大学 | System for automatic cab model recognition and automatic vehicle flowrate detection and method thereof |
| CN101266717A (en) * | 2008-04-25 | 2008-09-17 | 北京科技大学 | A car detection recognition system and method based on MEMS sensor |
| CN101923781A (en) * | 2010-07-28 | 2010-12-22 | 北京交通大学 | Vehicle type recognizing method based on geomagnetic sensing technology |
| 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 |
| CN103345841A (en) * | 2013-06-19 | 2013-10-09 | 北京交通大学 | Frequency domain spectrum energy vehicle type classifying method based on geomagnetic sensor |
| CN104112359A (en) * | 2013-04-22 | 2014-10-22 | 徐州工程学院 | Vehicle type and speed detection device based on bias magnetic field and working method |
| CN104200670A (en) * | 2014-08-26 | 2014-12-10 | 长沙理工大学 | Vehicle model identifying method based on vehicle earth induction characteristics |
| 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 |
| CN105761497A (en) * | 2013-04-22 | 2016-07-13 | 徐州工程学院 | Vehicle solid line pressing detection method and solid line pressing snapshooting device |
-
2017
- 2017-11-13 CN CN201711133396.7A patent/CN107886727B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6014447A (en) * | 1997-03-20 | 2000-01-11 | Raytheon Company | Passive vehicle classification using low frequency electro-magnetic emanations |
| JPH11271441A (en) * | 1998-03-23 | 1999-10-08 | Hino Motors Ltd | On-vehicle radar apparatus |
| CN101110161A (en) * | 2007-08-31 | 2008-01-23 | 北京科技大学 | System for automatic cab model recognition and automatic vehicle flowrate detection and method thereof |
| CN101266717A (en) * | 2008-04-25 | 2008-09-17 | 北京科技大学 | A car detection recognition system and method based on MEMS sensor |
| CN101923781A (en) * | 2010-07-28 | 2010-12-22 | 北京交通大学 | Vehicle type recognizing method based on geomagnetic sensing technology |
| 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 |
| CN104112359A (en) * | 2013-04-22 | 2014-10-22 | 徐州工程学院 | Vehicle type and speed detection device based on bias magnetic field and working method |
| CN105761497A (en) * | 2013-04-22 | 2016-07-13 | 徐州工程学院 | Vehicle solid line pressing detection method and solid line pressing snapshooting device |
| CN103345841A (en) * | 2013-06-19 | 2013-10-09 | 北京交通大学 | Frequency domain spectrum energy vehicle type classifying method based on geomagnetic sensor |
| CN103345842A (en) * | 2013-07-25 | 2013-10-09 | 重庆邮电大学 | Road vehicle classification system and method |
| CN104200670A (en) * | 2014-08-26 | 2014-12-10 | 长沙理工大学 | Vehicle model identifying method based on vehicle earth induction characteristics |
| 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 |
Non-Patent Citations (3)
| Title |
|---|
| JINHUI LAN: "Vehicle detection and classification by measuring and processing magnetic signal", 《MEASUREMENT》 * |
| 李盼: "基于地磁信息的车型识别技术的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 赵济民: "基于SVM的地磁车辆检测器车型分类方法研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109816992A (en) * | 2018-12-13 | 2019-05-28 | 杭州电子科技大学 | A kind of model recognizing method based on Geomagnetic signal |
| CN109816992B (en) * | 2018-12-13 | 2020-08-18 | 杭州电子科技大学 | Vehicle type identification method based on geomagnetic signals |
| CN110491141A (en) * | 2019-08-22 | 2019-11-22 | 长安大学 | A kind of information of vehicles identifying system and recognition methods |
| CN111524365A (en) * | 2020-05-11 | 2020-08-11 | 西安电子科技大学 | Method for classifying vehicle types by using multiple geomagnetic sensors |
| CN111524365B (en) * | 2020-05-11 | 2021-12-17 | 西安电子科技大学 | Method for classifying vehicle types by using multiple geomagnetic sensors |
| CN112611313A (en) * | 2020-12-11 | 2021-04-06 | 交控科技股份有限公司 | Rail vehicle length measuring device, rail vehicle length measuring method and traffic facility |
| CN114202927A (en) * | 2021-12-27 | 2022-03-18 | 招商局重庆公路工程检测中心有限公司 | Vehicle type detection method based on multi-sensor fusion |
| CN114202927B (en) * | 2021-12-27 | 2022-11-15 | 招商局重庆公路工程检测中心有限公司 | Vehicle type detection method based on multi-sensor fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107886727B (en) | 2020-04-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107886727A (en) | A kind of separation vehicle method, system and electronic equipment based on geomagnetic sensor | |
| Won | Intelligent traffic monitoring systems for vehicle classification: A survey | |
| CN102779281B (en) | Vehicle type identification method based on support vector machine and used for earth inductor | |
| CN108345822B (en) | Point cloud data processing method and device | |
| CN107665353A (en) | Model recognizing method, device, equipment and computer-readable recording medium based on convolutional neural networks | |
| CN108513676B (en) | Road condition identification method, device and equipment | |
| CN102509291B (en) | Pavement disease detecting and recognizing method based on wireless online video sensor | |
| CN104680795B (en) | A kind of model recognizing method and device based on local features | |
| CN103743980A (en) | Power quality disturbance recognition and classification method based on PSO (Particle Swarm Optimization) for SVM (Support Vector Machine) | |
| CN106250838A (en) | vehicle identification method and system | |
| CN107529659B (en) | Seatbelt wearing detection method, device and electronic equipment | |
| US20170336215A1 (en) | Classifying entities in digital maps using discrete non-trace positioning data | |
| CN106203437B (en) | Individual driving behavior recognition methods and device | |
| CN103646553A (en) | Investigation system for road traffic flow and realization method thereof | |
| CN104794184A (en) | Illegal car identifying method utilizing Bayesian classification algorithm based on massive data | |
| CN110517500A (en) | A kind of people's vehicle association process method and device | |
| CN108510396A (en) | It insures method, apparatus, computer equipment and the storage medium of verification | |
| CN113283331A (en) | Multi-class target identification method and system for unattended sensor system | |
| CN115346183A (en) | Lane line detection method, terminal and storage medium | |
| CN110263836B (en) | Bad driving state identification method based on multi-feature convolutional neural network | |
| CN106650801A (en) | GPS data-based method for classifying multiple types of vehicles | |
| Tobe et al. | vcity map: Crowdsensing towards visible cities | |
| CN106095915A (en) | The processing method and processing device of user identity | |
| Kubo et al. | Design of ultra low power vehicle detector utilizing discrete wavelet transform | |
| CN104616495A (en) | Bus GPS (Global Positioning System) probe vehicle based urban traffic state recognition method |
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 |