CN1655208B - Automatic recognition method for motor vehicle model - Google Patents
Automatic recognition method for motor vehicle model Download PDFInfo
- Publication number
- CN1655208B CN1655208B CN 200410091822 CN200410091822A CN1655208B CN 1655208 B CN1655208 B CN 1655208B CN 200410091822 CN200410091822 CN 200410091822 CN 200410091822 A CN200410091822 A CN 200410091822A CN 1655208 B CN1655208 B CN 1655208B
- Authority
- CN
- China
- Prior art keywords
- parameter
- vehicle
- information
- information parameter
- motor vehicle
- 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.)
- Expired - Fee Related
Links
Images
Abstract
This invention relates to a method for automatically identifying types of automobiles, which detects information parameters by a detection device then transfers them to a process system to identify types of the automobile based on the detection information including the following steps: 1, sample collection 2, information merge, 3, picking up information property parameters, 4, identifying types of automobiles.
Description
Technical field
The present invention relates to transport information control field, particularly relate to a kind of motor vehicle model method of identification automatically.
Background technology
Traffic data collection is the basic work of serving traffic programme, design, maintenance and management.Motor vehicle model identification is the important component part that urban road, highway and common road carry out traffic data collection.In fields such as vehicle toll, traffic controls, it also is unusual important problem that the vehicle of motor vehicle is discerned." the highway communication condition survey system of statistical report " issued in September, 2004 according to Ministry of Communications, when highway and common road are carried out traffic study, motor vehicle should be divided into following 8 classes: jubilee wagon, medium truck, high capacity waggon, station wagon, motorbus, towed vehicle, baby tractor, large and middle tractor.The research of relevant motor vehicle model recognition technology, many colleges and universities and enterprise are all carrying out.Prior art is carried out the following several mode that mainly contains of vehicle identification to motor vehicle:
1, video identification method
The equipment that utilizes Video Detection and image processing techniques is to discern according to carrying out vehicle with the vehicle ' s contour, and advantage is comparatively directly perceived, but has the problem that can't distinguish station wagon and jubilee wagon, medium truck and motorbus.In addition, video detecting method be difficult to overcome mist snow day, insufficient light situation descend can't operate as normal problem.
2, piezoelectricity method of identification
The equipment and the system that utilize piezoelectric detector to carry out vehicle classification serve as according to carrying out vehicle identification with vehicle wheelbase, the number of axle.Because the installation of piezoelectric device can destroy the road surface largely, and axletree can not unique differentiation different automobile types, can't distinguish such as little visitor, little goods, so such detecting device can't carry out vehicle classification exactly according to the relevant requirements of Ministry of Communications.In addition, piezoelectric detector ubiquity short characteristics in serviceable life.
3, radar method of identification
Radar detector is with the work of Doppler effect principle, and have a fatal shortcoming: doppler system can lose efficacy when vehicle slowly travels.In addition, utilize radar system to carry out vehicle identification and have bigger technical difficulty.
According to the electromagnetic theory of classics as can be known: when metal object during by coil, the inductance value of coil can change, and then the frequency of LC oscillatory circuit also can change.Sample with certain sample frequency, the frequency departure value (departing from reference value) of sampling instant is depicted, the frequency variation curve that produces in the time of can obtaining vehicle through coil.The frequency variation curve quite stable that same car produces during through same coil checker with same speed, same class car are during through same coil checker, and some characteristic parameter is stable; And therefore different types of motor vehicle can produce different frequency variation curves because the difference of size, structure, material is also different to the influence that coil checker produces, and this is to carry out the vehicle base of recognition; The frequency variation curve that the LC oscillatory circuit is produced is sent in the computing machine, and computing machine is discerned vehicle according to specific recognition methods again.
China Patent No. is 98200486.9, provide a name to be called the utility model patent of " motor vehicle model recognition device ", output connection interface control circuit by sensor, the output of interface control circuit connects existing computing machine again and forms, wherein interface control circuit is connect the input end of central processor CPU by signal processing circuit, the mouth line of CPU connects data register, and data register connects the input/output bus of microcomputer; Sensor is the toroidal inductor that is embedded on the highway.This utility model reflects the variation of winding inductance quantity by the variation that oscillatory circuit converts frequency to, each coil frequency value is constantly described out, thereby obtain vehicle through toroidal frequency curve, detect vehicle, vehicle is classified with it.This utility model patent has provided a kind of method of vehicle classification identification theoretically, but but do not provide specific embodiment, because when identical car passes through solenoid with friction speed, the frequency curve of its generation is also inequality, so the emphasis of vehicle identification does not lie in the method for describing, and be how this theory is applied in the practice, particularly, emphasis and difficult point are that how solenoid being responded to variable converts frequency curve to, determine the one-to-one relationship of all kinds of vehicles and frequency curve again, but this utility model patent is not pointed out.
Summary of the invention
The objective of the invention is to overcome the deficiencies in the prior art, a kind of effective motor vehicle model method of identification automatically is provided.
The technical solution used in the present invention is:
A kind of motor vehicle model is the method for identification automatically, and the information parameter by pick-up unit machines motor-car is sent to disposal system to information parameter again, identifies the vehicle of motor vehicle by disposal system according to parameter information, it is characterized in that comprising the steps:
(1), sample collection,
Gather the information parameter of various motor vehicle models;
(2), information categorization
Set up the information parameter storehouse of all kinds of motor vehicle model correspondences;
(3), information extraction characteristic parameter
Extract the specific characteristic of the pairing information parameter of all kinds of vehicles;
(4), identification vehicle
With the detected information parameter contrast of pick-up unit information characteristics parameter, and then draw recognition result.
In the technique scheme, the sample collection of described step (1) realizes by the solenoid of LC oscillatory circuit.
The information parameter that described step (1) is gathered is the frequency departure value of LC oscillatory circuit.
Step of the present invention (1) also comprises the span of determining information parameter, and the span of described information parameter can be set by artificial, or adopts the algorithm of support vector machine to learn voluntarily.
The information parameter storehouse that step (2) is set up is a waveform library.
The specific characteristic of the information parameter that step (3) is extracted comprises the moment and the ripple ratio of the finish time that crest value, crest number, crest occur, and these three are characterized as mandatory parameter.
The specific characteristic of the information parameter that step (3) is extracted also comprises popin average or the speed of a motor vehicle or vehicle commander, and these three are characterized as optional parameter, can select the combination in any in three to cooperate with mandatory parameter according to actual needs.
Described step (4) also can cooperate in the historical data ratio of all kinds of vehicles to discern.
Principle of work of the present invention is:
The frequency variation curve that the LC oscillatory circuit produced when the present invention crossed vehicle by pick-up unit is sent in the computing machine of disposal system;
Carry out sample collection then earlier, the method by the pairing vehicle of artificial selection waveform makes sample have teacher's information; Before carrying out the identification of robotization vehicle, also must determine the span of each sorting parameter of each class car, this process is called parameter configuration.When carrying out parameter configuration, the invention provides two kinds of methods: the one, artificial setting, the one, utilize the algorithm of support vector machine to learn voluntarily by machine.
Again sample is classified, set up the waveform library of all kinds of motor vehicle correspondences;
From waveform library, extract 3 mandatory parameters, be respectively the ratio that crest value, crest number, crest go out now and ripple finish time, from popin average, speed, these 3 optional parameter of vehicle commander, choose several parameters as required arbitrarily again.Because above parameter has only speed, two optional parameter of vehicle commander to need two coil joint-detection just can obtain, even therefore any one in two coils in certain track breaks down, system can both utilize another coil to carry out the volume of traffic separately and detect and vehicle classification, certainly, nicety of grading may descend to some extent in such cases.Mandatory parameter, optional parameter have been formed the vehicle classification parameter jointly;
The standard of vehicle classification is open, so the user can set up the standard of vehicle classification as required on their own.Because the vehicle classification standard that the user is provided with is varied, therefore may exist and to discern some vehicle, the possibility that some other class vehicles can't be distinguished, the present invention adopts following method to overcome this defective: for the some class motor vehicles that can't distinguish vehicle, with the historical data is to divide according to carrying out ratio, for example can't distinguish A, B two class cars, collect totally 100 of the vehicle numbers of A and B, A in the historical data: B is 6: 4, then these 100 vehicles that can't distinguish vehicle have 60 to belong to the A vehicle, and 40 belong to the B vehicle.The present invention is stored in the configuration result of vehicle classification parameter in the format specification file, para.txt for example, sample data with teacher signal is stored in the specific file equally, as sample.txt, if the staff of research station can't be configured the vehicle classification parameter from sample data, the support vector machine self-learning module that system carries also can't intactly be carried out parameter configuration, the sample.txt of storing sample can be sent to processing enter, by the technical professional it is handled and analyzes, after drawing satisfactory parameter configuration scheme it is stored among the para.txt, and beam back the research station replacing original para.txt, thereby reach the purpose of parameter configuration.
The configuration of vehicle classification characteristic parameter of the present invention can be undertaken by artificial setting and machine self study dual mode, but also by mandatory parameter and optional parameter in conjunction with the information extraction characteristic parameter, so flexible configuration is compatible good.In addition, method involved in the present invention also has automatic detection of fault and functions such as shielding, Redundant Control.
Coil of the present invention does not have strict requirement, the most of coil checker that is applied to traffic data collection at present all is suitable for native system, add that coil checker is stable, economy, reliability height, be most widely used traffic detecting device in China, the worldwide, so the present invention have very big actual promotional value.
Description of drawings
Fig. 1 is the signal pickup assembly hardware frame figure of one embodiment of the present of invention;
Fig. 2 is a recognition principle block diagram of the present invention;
Fig. 3 is the parameter configuration block diagram of one embodiment of the present of invention;
Fig. 4 is the LC oscillatory circuit schematic diagram of one embodiment of the present of invention;
Fig. 5 is provided with the interface circuit schematic diagram for the parameter of one embodiment of the present of invention;
Fig. 6 is the single chip circuit schematic diagram of one embodiment of the present of invention;
Fig. 7 is the impulse output circuit schematic diagram of one embodiment of the present of invention;
Fig. 8 is the RS232 interface circuit schematic diagram of one embodiment of the present of invention.
Embodiment
The present invention is described further below in conjunction with accompanying drawing.
Present embodiment is by testing in tunnel, main road, Huangpu, Guangzhou, the hardware block diagram of its signals collecting as shown in Figure 1, main processing terminal is a single-chip microcomputer, solenoid is connected on the LC oscillatory circuit, and LC oscillatory circuit, parameter are provided with interface circuit, impulse output circuit, RS232 interface circuit and are connected with single-chip microcomputer respectively.
LC oscillatory circuit schematic diagram is made up of inductance L 1, capacitor C 10, C11, resistance R 23, R24, R25, R26, R27, R28 and triode P5, P6 as shown in Figure 4, LOOP1 and LOOP2 port among the outside inductive coil two termination circuit figure.In the LC oscillatory circuit, add self-bias on the one hand, bias voltage strengthens in reaching the process of sustained vibration, thereby make transistor produce non-linear action and come the restriction set electrode current, the oscillatory circuit that L, C form is set on feedback circuit on the other hand, make it and the tuning generation sine voltage of the first-harmonic of output current, its part is feedbacked, take out as output simultaneously.Because the mutual inductance effect, when vehicle passed through the coil checker top, the variation of coil oscillation frequency made the oscillation frequency of LC oscillatory circuit also change thereupon, and the sinusoidal signal of generation outputs to single-chip microcomputer the 25th pin (P2.1) by triode P6.
Parameter is provided with the interface circuit schematic diagram as shown in Figure 5, form by toggle switch, pull-up resistor row, the parameter signalization respectively with the 10th, 12,14,16,18,20,24,26 pin of single-chip microcomputer (P1.0～P1.7 and P2.2～P2.6) be connected, read in signalization from single-chip microcomputer, single-chip microcomputer is handled accordingly according to setting.Parameter is provided with the various parameters that selection that interface dials key SWG by switch is provided with the vehicle classification detecting device: induction precision, susceptibility, output mode, delay parameter etc.
Impulse output circuit is made up of triode P1, P2, P3, P4 and relay R ELAYC1, RELAYC2 as shown in Figure 7.Signal derives from the 21st, 23 pin (P2.3, P2.4 pin) of single-chip microcomputer, is output as RELAYOUT1, RELAYOUT2 by relay.Impulse output circuit is isolated by relay R ELAYC1 and RELAYC2, the pulse signal that on behalf of vehicle, output whether pass through.
RS232 interface circuit schematic diagram mainly is made up of the MAX232 chip as shown in Figure 8.The serial communication pin of the RXT of single-chip microcomputer and TXD connects R1OUT and the T1IN mouth of MAX232.Output T1OUT, the R1IN of MAX232 receives output port and is connected with PC or other peripherals.
The single chip circuit schematic diagram as shown in Figure 6, it is the core of vehicle classification detecting device, the input sinusoidal signal is carried out filtering, detected and handle, and read in parameter interface signal is set, the pulse signal that on behalf of vehicle, output whether pass through, and with peripherals RS232 serial communication, output data information.
The present invention is when coming into operation, information classification is finished, so its recognition principle block diagram as shown in Figure 2, by on highway, being embedded with two solenoids of certain intervals, these two solenoids are sent to the parameter of the electromagnetic frequency off-set value that motor vehicle produced that crosses from above in the host computer of disposal system by slave computer by the RS232 network interface, host computer filters to the received signal, make signal parameter meet the requirements of scope, then satisfactory signal parameter is extracted characteristic parameter, the characteristic parameter of present embodiment comprises crest value, the crest number, the moment and the ripple ratio of the finish time that crest occurs, the speed of a motor vehicle, again the characteristic parameter and the information parameter storehouse of extracting contrasted, then can identify vehicle according to " highway communication condition survey system of statistical report " desired vehicle standard.
The schematic diagram of present embodiment parameter configuration as shown in Figure 3 can be by machine learning, manually set the configuration result that two kinds of methods obtain characteristic parameter, and by 2000 samples are carried out vehicle classification, classification accuracy reaches 95.1% as a result.
Claims (4)
1. the motor vehicle model method of identification automatically, the information parameter by pick-up unit machines motor-car is sent to disposal system to information parameter again, identifies the vehicle of motor vehicle by disposal system according to parameter information, it is characterized in that comprising the steps:
(1), sample collection,
Sample collection realizes by the solenoid of LC oscillatory circuit, it gathers the information parameter of the motor vehicle of various vehicles, described information parameter is the frequency departure value of LC oscillatory circuit, sample collection also comprises the span of determining information parameter, and this span adopts the algorithm of support vector machine to learn voluntarily;
(2), information categorization
Set up the information parameter storehouse of all kinds of vehicle correspondences, the information parameter storehouse of being set up is a waveform library;
(3), information extraction characteristic parameter
Extract the specific characteristic of the pairing information parameter of all kinds of vehicles;
(4), identification vehicle
With the detected information parameter contrast of pick-up unit information characteristics parameter, and then draw recognition result;
2. recognition methods according to claim 1 is characterized in that the specific characteristic of the information parameter that described step (3) is extracted comprises the moment and the ripple ratio of the finish time that crest value, crest number, crest occur.
3. recognition methods according to claim 1 and 2 is characterized in that the specific characteristic of the information parameter that described step (3) is extracted also comprises popin average or the speed of a motor vehicle or vehicle commander.
4. recognition methods according to claim 3 is characterized in that described step (4) can cooperate in the historical data ratio of all kinds of vehicles to discern.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410091822 CN1655208B (en) | 2004-12-31 | 2004-12-31 | Automatic recognition method for motor vehicle model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410091822 CN1655208B (en) | 2004-12-31 | 2004-12-31 | Automatic recognition method for motor vehicle model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1655208A CN1655208A (en) | 2005-08-17 |
| CN1655208B true CN1655208B (en) | 2010-08-11 |
Family
ID=34892446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410091822 Expired - Fee Related CN1655208B (en) | 2004-12-31 | 2004-12-31 | Automatic recognition method for motor vehicle model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1655208B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102982684B (en) * | 2012-11-28 | 2015-05-13 | 深圳市迈科龙影像技术有限公司 | Vehicle model recognition method and system |
| CN103745598B (en) * | 2014-01-09 | 2015-12-30 | 中科联合自动化科技无锡有限公司 | Based on the model recognizing method of front face feature |
| CN107085074B (en) * | 2017-04-19 | 2019-07-23 | 中国科学技术大学 | A method of classification monitoring motor-vehicle tail-gas |
| CN110793482A (en) * | 2019-11-13 | 2020-02-14 | 佛山科学技术学院 | Vehicle sample data acquisition system for collecting data conforming to normal distribution |
| CN116645818B (en) * | 2023-07-27 | 2023-10-10 | 山东高速集团有限公司创新研究院 | Vehicle type recognition method based on multidimensional feature extraction |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1052961A (en) * | 1990-12-30 | 1991-07-10 | 天津大学电力及自动化工程系 | Multifunctional vehicle tester |
| CN2343635Y (en) * | 1998-01-22 | 1999-10-13 | 北京师范大学 | Motor-vehicle type identifier |
-
2004
- 2004-12-31 CN CN 200410091822 patent/CN1655208B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1052961A (en) * | 1990-12-30 | 1991-07-10 | 天津大学电力及自动化工程系 | Multifunctional vehicle tester |
| CN2343635Y (en) * | 1998-01-22 | 1999-10-13 | 北京师范大学 | Motor-vehicle type identifier |
Non-Patent Citations (2)
| Title |
|---|
| 朱海涛.一种基于感应线圈的车型识别系统 2003.2003,7-43、52. |
| 朱海涛.一种基于感应线圈的车型识别系统 2003.2003,7-43、52. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1655208A (en) | 2005-08-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102779281B (en) | Vehicle type identification method based on support vector machine and used for earth inductor | |
| CN101833862B (en) | False-detection resistant annular coil vehicle detector | |
| CN106327911B (en) | A kind of intelligent parking control system for parking lot | |
| CN103500504A (en) | Urban road traffic parameter estimation and road situation discrimination method based on RFID (Radio Frequency Identification) data and application system | |
| CN101694746A (en) | Magnetic resistant vehicle quantity and speed detecting node based on Zigbee wireless protocol and use methods thereof | |
| CN101908250A (en) | Full RFID (Radio Frequency Identification Device) license plate based parkinglay-by intelligent monitoring system and method | |
| CN102779413A (en) | Motor vehicle characteristics data acquisition analyzer and analyzing method | |
| CN1655208B (en) | Automatic recognition method for motor vehicle model | |
| CN109034265A (en) | Wiper self-starting device and method | |
| CN107273771B (en) | Vehicle driving direction identification method and system, identification station, memory and background server | |
| US6865518B2 (en) | Method and device for classifying vehicles | |
| CN206249557U (en) | Based on the vehicle detecting system that Magnetic Sensor and ultrasonic sensor are merged | |
| CN103646553A (en) | Investigation system for road traffic flow and realization method thereof | |
| CN101814237A (en) | Road condition identifier based on license plate recognition technology and identification method thereof | |
| CN109325409B (en) | Passing vehicle fake plate inspection platform | |
| CN104200670B (en) | A kind of vehicle model recognizing method based on vehicle earth induction feature | |
| CN208477734U (en) | A kind of traffic data collection terminal based on earth induction | |
| CN108984758A (en) | Car owner's association address analysis method and system based on point identification | |
| CN103050014A (en) | Traffic speed detection system and detection method | |
| CN106710238B (en) | Method for improving accuracy of geomagnetic vehicle detector | |
| CN205405899U (en) | Forbidden candid photograph device violating regulations of freight train | |
| CN201392591Y (en) | Portable automatic vehicle type sorting device | |
| CN202153396U (en) | Vehicle detector based on inductance variation of measuring coil | |
| CN201707792U (en) | Reluctance vehicle number and speed detection node based on Zigbee protocol | |
| CN100501796C (en) | Minisize buried automobile detector |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100811 Termination date: 20141231 |
|
| EXPY | Termination of patent right or utility model |