CN208367913U - A kind of roadside unit - Google Patents
A kind of roadside unit Download PDFInfo
- Publication number
- CN208367913U CN208367913U CN201820444400.5U CN201820444400U CN208367913U CN 208367913 U CN208367913 U CN 208367913U CN 201820444400 U CN201820444400 U CN 201820444400U CN 208367913 U CN208367913 U CN 208367913U
- Authority
- CN
- China
- Prior art keywords
- information
- vehicle
- vehicles
- unit
- sent
- 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.)
- Active
Links
Landscapes
- Traffic Control Systems (AREA)
Abstract
A kind of roadside unit, the roadside unit obtains the information of vehicles for driving into the vehicle in lane in ETC system based on video image processing, and it is matched with information of vehicles bound in the on board unit of the vehicle, and then whether the information of vehicles for judging to drive into lane in ETC system is consistent with information of vehicles bound in the on board unit of the vehicle.By the device, in the case that the actual vehicle information of the information of vehicles and vehicle that can identify in ETC system is not inconsistent, warning information is issued, vehicle fee evasion is prevented.
Description
Technical field
This application involves intelligent transportation fields, and in particular to a kind of roadside unit.
Background technique
Intelligent transportation (Intelligent Transportation System, abbreviation ITS) can be traffic participant
Multifarious service is provided, is the developing direction of traffic system at this stage, electric non-stop toll is applied under this trend
The rapid growth of (Electronic Toll Collection, abbreviation ETC) market scale, 2012-2016 rise to from 500,000,000 yuan
2200000000 yuan, annual compound growth rate is 45% (data source: Shenzhen intelligent transportation employer's organization).According to National Information Center's information
Resources Development Division is up to 2.5 hundred million, it is expected that the year two thousand twenty China car ownership is estimated, and according to " " 13 " are modern comprehensive
Communications and transportation system development plan ", the year two thousand twenty highway passenger vehicle ETC utilization rate is up to 50%, therefore the year two thousand twenty China ETC user
Number is estimated to be expected to reach 1.25 hundred million.
ETC system in the prior art is mainly believed according to the vehicle obtained in OBU in the information of vehicles for determining vehicle
What breath was determined, and in practical application, there are the information of vehicles of information of vehicles and actual travel vehicle in OBU is not corresponding
Situation, for example, the OBU of carriage type is mounted in large-scale vechicle by car owner for fee evasion.In this case, then it can make ETC system vehicle
Information identifies mistake, the phenomenon that vehicle fee evasion occurs.
Utility model content
The application provides a kind of roadside unit.The roadside unit is vehicle when driving into lane in ETC system based on acquisition
Video image obtains the information of vehicles for driving into vehicle according to video image, the vehicle with the on board unit binding driven on vehicle
Information is compared, and judgement is matching.And then judge to enter the vehicle of the on board unit binding in ETC system on the vehicle in lane
Whether information is consistent with the information of vehicles of the vehicle driven into ETC system lane.Solve ETC system vehicle letter in the prior art
, there is the problem of vehicle fee evasion in breath identification mistake.A kind of roadside unit specific embodiment disclosed in the present application is as follows:
According in a first aspect, provide a kind of roadside unit in a kind of embodiment, including the processing of wireless communication unit, video is single
Member and information of vehicles matching unit, wherein
The wireless communication unit, for carrying out data interaction with the on board unit on vehicle, from the on board unit
It obtains the information of vehicles of the vehicle and is sent to the information of vehicles matching unit；
The video processing unit, for acquiring the video image of vehicle, and to the video image of the vehicle using deep
The method of study is spent, to identify the information of vehicles of the vehicle and be sent to the information of vehicles matching unit；
The information of vehicles matching unit is sent for receiving the wireless communication unit and the video processing unit
Information of vehicles, and judge the information of vehicles that the wireless communication unit is sent and the information of vehicles that the video processing unit is sent
Whether match, if mismatching, issues warning information.
Further, which further includes millimetre-wave radar unit, for receiving and emitting MMW RADAR SIGNAL USING, and
The information of vehicles that the vehicle is obtained according to the echo-signal of the MMW RADAR SIGNAL USING received is sent to the vehicle
Information matching unit.
Further, what the millimetre-wave radar unit that the information of vehicles matching unit is also used to judge to receive was sent
Whether information of vehicles matches with the information of vehicles that the wireless communication unit is sent, if mismatching, issues warning information；
And/or
The vehicle letter that the millimetre-wave radar unit that the information of vehicles matching unit is also used to judge to receive is sent
It ceases whether the information of vehicles sent with the video processing unit matches, if mismatching, issues warning information；
And/or
The vehicle letter that the millimetre-wave radar unit that the information of vehicles matching unit is also used to judge to receive is sent
Cease whether the information of vehicles that the information of vehicles sent with the video processing unit and the wireless communication unit are sent matches,
If mismatching, warning information is issued.
Further, the roadside unit further includes perception tracking cell, for using depth to the video image of acquisition
The method of habit, to perceive quantity, flow, and/or the density of the vehicle at high speed crossing；And/or
Method for using deep learning to the video image of acquisition carries out dynamically track to the vehicle to realize.
Further, the information of vehicles of the vehicle include license plate, vehicle, vehicle money, annual test mark, color, sunshading board and/
Or driver.
According to a kind of roadside unit of above-described embodiment, the video image of the vehicle in the lane ETC is driven by acquiring, to vehicle
Video image handled using the method for deep learning, identify the information of vehicles of the vehicle, and with vehicle-mounted list on the vehicle
Information of vehicles bound in member, or obtain information of vehicles with from based on millimetre-wave radar and be compared, judge whether it matches.
And then the phenomenon that improving the accuracy of vehicle identification in ETC system, preventing vehicle fee evasion.
Detailed description of the invention
Fig. 1 is a kind of functional block diagram of roadside unit；
Fig. 2 is a kind of functional block diagram of the video processing unit of embodiment；
Fig. 3 is a kind of functional block diagram of the millimetre-wave radar unit of embodiment；
After Fig. 4 is a kind of millimetre-wave radar echo-signal Fourier's time-frequency conversion after zero intermediate frequency is converted of embodiment
Spectrogram；
Fig. 5 is a kind of LFM video stretching figure of embodiment different automobile types；
Fig. 6 is the functional block diagram of the roadside unit receiving system of another embodiment；
Fig. 7 is the functional block diagram of the roadside unit control system of another embodiment.
Specific embodiment
The application is described in further detail below by specific embodiment combination attached drawing.Wherein different embodiments
Middle similar component uses associated similar element numbers.In the following embodiments, many datail descriptions be in order to
The application is better understood.However, those skilled in the art can recognize without lifting an eyebrow, part of feature
It is dispensed, or can be substituted by other elements, material, method in varied situations.In some cases, this Shen
Please it is relevant it is some operation there is no in the description show or describe, this is the core in order to avoid the application by mistake
More descriptions are flooded, and to those skilled in the art, these relevant operations, which are described in detail, not to be necessary, they
Relevant operation can be completely understood according to the general technology knowledge of description and this field in specification.
It is formed respectively in addition, feature described in this description, operation or feature can combine in any suitable way
Kind embodiment.Meanwhile each step in method description or movement can also can be aobvious and easy according to those skilled in the art institute
The mode carry out sequence exchange or adjustment seen.Therefore, the various sequences in the description and the appended drawings are intended merely to clearly describe a certain
A embodiment is not meant to be necessary sequence, and wherein some sequentially must comply with unless otherwise indicated.
It is herein component institute serialization number itself, such as " first ", " second " etc., is only used for distinguishing described object,
Without any sequence or art-recognized meanings.And " connection ", " connection " described in the application, unless otherwise instructed, include directly and
It is indirectly connected with (connection).
In the embodiment of the present application, it is based on video image analysis technology, the view of the vehicle in lane in ECT system is driven into acquisition
Frequency image, analysis and processing by application deep learning method to video image, to obtain license plate, vehicle, the vehicle money of vehicle
And the information of vehicles such as mark between year, and with the information of vehicles that is obtained from the on board unit of the vehicle, or with from based on millimeter wave
Radar obtains information of vehicles and is compared, and judges whether information of vehicles matches, and then executes corresponding operation.
Embodiment one:
As shown in Figure 1, being a kind of functional block diagram of roadside unit.Roadside unit includes wireless communication unit 120, view
Frequency processing unit 110 and information of vehicles matching unit 140.Wherein, wireless communication unit 120 includes antenna 121,5.8G channel radio
Interrogate module 122 and DSRC protocol resolution module 123.5.8G wireless communication module 122 is the outdoor roadside unit mould of tradition ETC
Block, for interacting the information of vehicles for obtaining vehicle binding with the OBU on vehicle；5.8G wireless communication module 122 includes 5.8G
RF mixer, filter and amplifying circuit etc..5.8G wireless communication module 122 is sailed by the reception of antenna 121 through ETC system vehicle
The frequency microwave signal that OBU on the vehicle in road is returned, then sent out after the signal received is mixed, is filtered, and/or is amplified
DSRC protocol resolution module 123 is given, DSRC protocol resolution module 123 parses the OBU information of the vehicle, and then obtains and sail pathway
The information of vehicles of the vehicle of side unit.The information of vehicles of the vehicle is sent to information of vehicles matching again by wireless communication unit 120
Module 140；
The video processing unit 110 of roadside unit is used to acquire the video image of vehicle, and to collected video image
Using the method for deep learning, to identify the information of vehicles of the vehicle and be sent to information of vehicles matching unit 140.Specially
Video processing unit 110 obtains the video image for sailing the vehicle through roadside unit, then extracts from video image containing vehicle
The target critical hardwood of information image is wiped out by background and extracts target area to be identified, and handled video image,
Obtain the image information of the vehicle.Before this, video processing unit 110 constructs preliminary nerve net using deep learning method
Network, then collected magnanimity vehicle big data is trained, obtain trained neural network.By the image information of the vehicle
Trained neural network is inputted, the information of vehicles that identifies the information of vehicles of the vehicle, and will identify that is sent to vehicle letter
Cease matching module 140.Wherein, information of vehicles includes license plate, vehicle, vehicle money, annual test mark, color, sunshading board, and/or driving
People.
As shown in Fig. 2, being the functional block diagram of video processing unit.Video processing unit include video camera 201,
CPU202, video processing board 203, light compensating lamp 205 and network interface 204.Specifically, light compensating lamp 205 is used for when acquisition video
When information, if there is the needs of illumination, corresponding light filling function is carried out.Video camera 201 obtains the video figure for having target vehicle
Picture is sent to CPU202 by network interface 204 and video handles board 203, and video processing board 203 is extracted from video
Target critical hardwood is wiped out by background and extracts target area to be identified, and handled video image, and the vehicle is obtained
Image information.Certainly, video processing unit can be include video camera 201, CPU202, light compensating lamp 205 and network interface 204
Video processing unit.Wherein, CPU202 is used to wipe out and be extracted wait know by background to Target key frames are extracted from video
Other target area, and video image is handled, obtain the image information of the vehicle.
The identification of the information of vehicles such as Car license recognition and vehicle cab recognition at present is the key problem of video processing, main difficulty
It is that different illumination conditions influence recognition effect；For example, license plate in fine day and the overcast and rainy, weather that snows be visually it is different,
Fine day license plate picture is generally all relatively clear, but strong illumination can bring the reflective too strong problem in part, rainy weather picture
It is dim fuzzy, it snows, it is possible that some regions of covering license plate.In addition even if can on the same day by the variation of sunlight color
Influence the image quality of camera.It is all the direct factor for influencing Car license recognition or vehicle cab recognition effect, traditional image above
Treatment mechanism can not efficiently solve problem above, but deep learning algorithm is applied the identification in information of vehicles, to its video
The recognition capability of processing system effectively big improvement and promotion.Deep learning is that one of machine learning is based on carrying out data
The method of representative learning.The benefit of deep learning is efficient with having supervised or unsupervised feature learning and layered characteristic to extract
Algorithm obtains feature to substitute by hand.Its motivation is to establish, simulate the neural network of human brain progress analytic learning, its imitation people
The mechanism of brain explains data, removes one artificial neural network containing multilayer hidden layer of building using certain methods, network
The number that each layer all corresponds to initial data different levels is abstract, to achieve the purpose that describe image with feature vector.
Deep learning method divides supervised learning and unsupervised learning, and the learning model established under different learning frameworks is not
Together, for example, convolutional neural networks (Convolutional neural networks, abbreviation CNN) are exactly a kind of supervision of depth
Machine learning model under study, and depth confidence net (Deep Belief Nets, abbreviation DBN) is exactly a kind of unsupervised learning
Under machine learning model.
CNN is mainly used to the X-Y scheme of identification displacement, scaling and other forms distortion invariance.Due to the feature of CNN
Detection layers are learnt by training data, so the feature extraction of display is avoided when using CNN, and implicitly from instruction
Practice and is learnt in data；Furthermore since the neuron weight on same Feature Mapping face is identical, so network can be learned parallel
It practises, this is also convolutional network is connected with each other a big advantage of network relative to neuron.Convolutional neural networks are with its local weight
Shared special construction has unique superiority in terms of speech recognition and image procossing, is laid out closer to actual life
Object neural network, the shared complexity for reducing network of weight, the especially image of multidimensional input vector can directly input net
This feature of network avoids the complexity of data reconstruction in feature extraction and assorting process.
Further, information of vehicles matching unit 140 is sent out for receiving video processing unit 110 and wireless communication unit 120
That send drives into the information of vehicles of ETC system vehicle, and the information of vehicles obtained to video processing unit is obtained with wireless communication unit
The information of vehicles taken is compared, if mismatching, issues warning information.
Roadside unit further includes millimetre-wave radar unit 130 in the present embodiment.Millimetre-wave radar unit includes radar antenna
131, millimeter wave radar module 132 and signal processing module 133；Millimeter wave radar module 132 by radar antenna 131 receive and
Emit MMW RADAR SIGNAL USING, and received millimeter wave echo-signal is sent to signal processing module 133, signal processing module
133 obtain the information of vehicles for sailing the vehicle through roadside unit according to the echo-signal of the MMW RADAR SIGNAL USING received, concurrently
Give information of vehicles matching unit 140.
As shown in figure 3, being the functional block diagram of millimetre-wave radar unit.Millimetre-wave radar unit includes radar antenna
131, milli wave radar module 132 and signal processing module 133 may also include thresholding and reset trigger module 134, wherein millimeter wave
Radar module 132 includes radar radio-frequency front-end 1321.Radar antenna 131 uses array antenna, and array antenna is by multiple levels
The submatrix unit of the antenna of direction and vertical direction forms, and the subelement of array antenna presses array arrangement.Radar antenna 131 receives
The reflected millimetre-wave radar echo-signal of target in MMW RADAR SIGNAL USING radiation scope also passes through radar antenna simultaneously
The MMW RADAR SIGNAL USING of 131 transmitting modulation continuous wave LFM (linear frequency modulation).Radar radio-frequency front-end 1321 is by radar antenna 131
It receives and is mixed and is amplified by the echo-signal that target returns, then be converted into zero intermediate frequency signals, while by base band tune
The triangle wave modulation continuous wave signal made is converted to MMW RADAR SIGNAL USING after amplification and mixing and launches.
Further, the millimeter wave radar module 132 of the present embodiment can be millimetre-wave radar sensor, which passes
Sensor is fabricated using planar microstrip technology, and small in size, integration degree is high, is incuded sensitive.This millimetre-wave radar sensing
The radar radio-frequency front-end 1321 of device by voltage controlled oscillator VCO, reception/transmitting antenna, low noise RF preamplifier, coupler and
Frequency mixer composition.
In preferably embodiment, millimeter wave radar module 132 may also include radar signal conditioning module 1322, for connecing
The zero intermediate frequency signals that radar radio-frequency front-end 1321 exports are received, and zero intermediate frequency signals is amplified and filters etc. and handled, are believed
It makes an uproar and exports than higher zero intermediate frequency signals to signal processing module 133.
Signal processing module 133 comes from radar signal conditioning module 1322 treated that signal-to-noise ratio is higher for that will receive
Zero intermediate frequency signals carry out time-frequency conversion to generate range value curve.Range value curve can be range value with the curve of distance change
Figure, i.e. X-axis or Y-axis indicate that amplitude, Y-axis or X-axis indicate distance.Range value curve can specifically use the side of SDIF histogram
Formula, can also be by the way of line chart.Wherein, range value curve includes threshold curve and target amplitude value curve.Thresholding is bent
Line is when not having vehicle to sail in the radiation scope through MMW RADAR SIGNAL USING, and signal processing module 133 will convert into zero intermediate frequency letter
Number echo-signal carry out time-frequency conversion range value curve generated.Target amplitude value curve is then to have vehicle to sail through millimeter wave
When in the radiation scope of radar signal, the echo-signal that signal processing module 133 will convert into zero intermediate frequency signals carries out time-frequency change
Change range value curve generated.In the present embodiment, signal processing module 133 will be handled from radar signal conditioning module 1322
Zero intermediate frequency signals afterwards carry out FFT (Fast Fourier transform, Fast Fourier Transform (FFT)) operation, and then are built into
The target amplitude value curve of histogram is compared by SDIF histogram with threshold curve, is sailed according to comparison result acquisition through institute
State the characteristic information of vehicle in the radiation scope of MMW RADAR SIGNAL USING.Wherein, the characteristic information of vehicle include vehicle length,
Height, and/or profile.
It is illustrated so that the range value of range value curve is with the histogram of distance change as an example below.
Detection threshold curve first.Some range points, range points are first defined in the radiation scope of MMW RADAR SIGNAL USING
It is corresponding with the pre-determined distance point in the effective identification range of vehicle identifier, on the range value curve that signal processing module obtains
The abscissa of each point is corresponding with pre-determined distance point；The distance of adjacent point-to-point transmission corresponds to two neighboring on range value curvilinear abscissa
Actual range between pre-determined distance point is known as the resolution ratio of range value curve；The ordinate of each point is the distance on range value curve
The range value of the echo-signal of point.
In detection threshold curve, determination does not have vehicle to sail in the radiation scope through MMW RADAR SIGNAL USING, setting signal
The range value curve that processing module 133 obtains is threshold curve.
Further, signal processing module 133 generates threshold curve specifically: according to the arrangement of array antenna 131, acquisition milli
In the radiation scope of metre wave radar signal, corresponding multiecho signal in each range points, and repeatedly returned according to collected
Wave signal acquires corresponding multiple amplitude Value Data in each range points；It will be described each apart from corresponding multiple amplitude Value Data
Ascending order arrangement is carried out, is obtained each apart from corresponding effective breadth value array and averaged；Using distance as abscissa, connection
It is each apart from corresponding average value, generate threshold curve.
Target amplitude value curve is compared by signal processing module 133 with threshold curve specifically: by target amplitude value
Curve is compared with threshold curve, or directly that the range value of each point on target amplitude value curve is corresponding on threshold curve
The range value of each point asks poor, such as within the scope of a certain distance, corresponds to the strong of each point on target amplitude value curve and threshold curve
The difference of angle value is positive, i.e., when the range value for having continuously multiple points on target amplitude value curve is greater than the corresponding company of threshold curve
The range value of continuous multiple points, then judgement has vehicle to drive into.The curve that the range value of these continuity points is formed is known as significant figure
According to section, and the vehicle commander of the vehicle can be calculated according to this distance.It can also be according to target amplitude value curve and threshold curve width
The difference of angle value obtains the overall height or vehicle's contour of vehicle.
It will be illustrated by taking the acquisition of the profile of the height of the length of vehicle, vehicle and vehicle as an example below.
For signal processing module 133 in order to reduce influence of the Doppler effect to range accuracy, this programme uses symmetric triangular
Wave modulation.Because triangle arm is difference frequency signal, frequency spectrum be it is discrete, so corresponding distance value be also it is discrete, because
There are a set a distance errors when this correspondence ranging.According to system requirements, the range accuracy of target can be set as 0.3m, i.e., away from
It is 0.3m from resolving power.According to range accuracy formula:
Work as fmWhen=500MHz, R is calculated to obtainminFor 0.3m, distance resolution≤0.3m can satisfy.Wherein, fmFor modulation
Frequency deviation, RminFor range resolution, c0For the light velocity, by formula it is found that range accuracy depends on frequency modulation fm。
In the present embodiment, the length of vehicle can be calculated by the distance resolution of millimetre-wave radar.Specifically, can incite somebody to action
The poor calculating that is multiplied with the distance resolution of the range value curve of abscissa value corresponding to two endpoints of valid data section
Obtain the vehicle commander of target vehicle.
Radar equation are as follows:
Wherein, S/N is the backward energy of radar；For the peak power of transmitting；G is antenna gain；It is dissipated for the radar of target
Penetrate sectional area；λ is wavelength, and R is one way distance of the radar to target.By dissipating for radar return energy known to radar equation and target
Directly proportional and distance the biquadratic for penetrating sectional area is inversely proportional, therefore the size of backward energy can reflect the scattering center of vehicle
At a distance from radar, the overall height of vehicle can reflect out indirectly.Therefore, the vehicle in MMW RADAR SIGNAL USING radiation scope is driven into
, the height of vehicle can be obtained by backward energy value maximum on continuous frequency spectrum.Specifically, can will be each in valid data section
Maximum value in the difference of the range value of dotted or gate limit curve corresponding points brings radar equation into and obtains the height of target vehicle.
The vehicle of radar signal radiation scope is driven into, the profile information of vehicle can be by judging spectral peak continuous frequency nearby
The variation of energy value obtains in spectrum.Specifically, the range value of each point in valid data section and threshold curve corresponding points can be asked poor,
It brings each difference into radar equation and obtains the height value that each point in the valid data section corresponds to the point on the vehicle's contour,
And then obtain the profile of the vehicle.
In the present embodiment, the acquisition of the profile of the length of vehicle, the height of vehicle and vehicle need to be bent by target amplitude value
Wired-AND gate limit curve is compared, and will be illustrated below to target amplitude value curve and threshold curve.
It is exemplified below, as shown in figure 4, carrying out Fu for the millimetre-wave radar echo-signal converted by zero intermediate frequency
In spectrogram after leaf time-frequency conversion.Wherein, block curve 1 is threshold curve in figure, and dashed curve 2 is that have vehicle by millimeter
Target amplitude value curve when wave radar signal radiation areas, dashed curve 2 is higher than the curved section of block curve 1 in circle in figure
For valid data section.
Because radar system is to find target by the rescattering power of target, that is to say, that radar system transmitting
The target that encounters in the air of electromagnetic wave, part energy radiated by target absorption, another part by target again.Target
Again radiation is exactly so-called secondary scattered or reflection.The operation wavelength (or frequency) of the ability of target reflection electromagnetic wave and system,
It is the geometry and size of target, related to absorbability of the target to electromagnetic wave to the incident visual angle of target.ETC system vehicle
The power of background return can cause undesirable influence to the detection performance of millimetre-wave radar on road, therefore effective ambient noise is estimated
It calculates, the target echo signal for obtaining high s/n ratio is very important.This programme is inhibited using Ordered Statistic average background
Method set threshold curve, basic step is as follows: in MMW RADAR SIGNAL USING effective range, by with vehicle identifier
The default unit distance point of distance acquires the millimetre-wave radar echo-signal of predetermined unit range points, and is returned according to millimetre-wave radar
Wave signal obtains the amplitude Value Data on each unit distance point.
The amplitude Value Data on the repeatedly unit distance point is acquired on each unit distance point.
Multiple amplitude Value Datas on the unit distance point are arranged by amplitude ascending order, form an array sequence, this
Array sequence is noise spectrum array sequence.
Give up the low side data in the noise spectrum array sequence, low side data refer to that numerical value is opposite in noise spectrum array sequence
Lesser data.Purpose is in practical applications, to eliminate the interference that part background power rises and falls.
The noise spectrum array sequence of removal low side data seek it is average as thresholding, that is, the default unit away from
Threshold value from point.Briefly be exactly do not have vehicle sail through when, the width of the echo-signal of n times measurement in default unit distance point
The average value of angle value, wherein N is more than or equal to 2.Further be summarised as, signal processing module more than 120 times measurement default unit away from
Range value from point, and the range value for being greater than threshold value in the range value for presetting the unit distance point repeatedly measured is averaging,
Obtain the predetermined amplitude value of default unit distance point.As shown in Figure 4, wherein range value curve resolution rate is decimetre.By effective
Data segment can be represented is detecting target vehicle at 20 meters of radar, and the width that valid data section corresponds to abscissa is vehicle
Length, i.e., vehicle commander be 4 meters, maximum reflection field intensity value be 2.2, can get the height of the vehicle by radar equation.It will
The intensity value of each point and threshold curve corresponding points in valid data section asks poor, each difference can be calculated by radar equation
Vehicle is driven into out and corresponds to the height value that each is put on effective data segment, and then obtains the profile information for driving into vehicle.Its principle
It is that the echo-signal according to radar signal in different distance can generate different spectral lines on baseband frequency spectrum, drives into radar signal
The acquisition of the vehicle of radiation scope, Vehicle length information can be by judging that spectrum width is identified；For example, set distance
Resolution ratio is 0.6 meter, that is to say, that 4.8 meters of long automobiles are possible to generate 8 spectral lines, and 10 meters of vehicle will necessarily also produce
Raw 17 spectral lines.Therefore vehicle commander can be judged by judging the width of continuous frequency spectrum.As shown in figure 5, being two cars, vehicle in figure
Long is respectively LFM (linear frequency modulation) video stretching figure of 4.8 meters and 9.6 meters of vehicle, and previous vehicle commander is as can be seen from Figure 5
4.8 meters and latter vehicle commander is 9.6 meters.And then the length of vehicle can be judged by spectrum width.
Further, which may also include thresholding and resets trigger module 134, can be used for setting signal
Processing module 133 resets thresholding every a unit time, and the unit time can be half an hour, one hour or two hours
Deng.It may include temperature monitoring module, humidity monitoring module, and/or weather monitoring module that thresholding, which resets trigger module 134, wherein
Temperature monitoring module is used to monitor the temperature change in ETC system lane, and when the temperature is changed, thresholding resets the touching of trigger module 134
Signalling processing module 133 resets threshold curve, for example, thresholding is reseted when temperature change value is 5 degree positive and negative greater than preset value
134 trigger signal processing module 133 of trigger module resets threshold curve；Humidity monitoring module is for monitoring ETC system vehicle
The humidity in road, when humidity value variation, thresholding resets 140 trigger signal processing module 133 of trigger module and resets thresholding song
Line；Weather monitoring module is used to monitor the Changes in weather in ETC system lane, and when Changes in weather, thresholding resets trigger module touching
134 signalling processing modules 133 reset threshold curve, for example, weather monitoring module monitors are rained suddenly to weather, thresholding
It resets trigger module 134 and threshold curve is reset with regard to trigger signal processing module 133.
Further, the present embodiment can be used clustering algorithm and believe each feature such as the length, height, and/or profile of vehicle
Breath optimizes processing.
Select the range information away from radar as initial cluster center from collected radar data first；And for institute
The radar data of multi collect is as other objects, then similar to these cluster centres according to their range information below
Degree, the i.e. difference with the range information of initial radar data, assign these to (the cluster centre institute generation most like with it respectively
Table) cluster.
Then each cluster centre newly clustered that obtains is calculated again (to carry out average value processing to the data after cluster, seek
New cluster centre)；Constantly repeat this process becomes only until having been calculated, when the number of cluster centre is more than a certain thresholding
When, then searched out it is optimal as a result, i.e. target preliminary characteristic information.
For the multiple radar echo signal of a vehicle, K- is carried out carrying out vehicle commander, ranging information and amplitude information
Means clustering processing.The algorithm receives multiple input quantities；Then multiple data objects are divided into k cluster to make
The cluster of acquisition meets: the object similarity in same cluster is higher；And the object similarity in different clusters is smaller.Cluster phase
It is to obtain one " center object " (center of attraction) using the mean value of object in each cluster come what is calculated like degree.
Such as the length of vehicle is obtained by the processing of radar echo signal, specific calculating process is as follows:
The long one of object of data Object Selection of radar truck N number of first is as initial cluster center；And for remaining
Other objects, then according to the similarity (i.e. vehicle commander is similar) of they and these cluster centres, assign these to respectively and its
Most like (representated by cluster centre) cluster.
Then each cluster centre (mean values of all objects in the cluster) for obtaining and newly clustering is calculated again；Constantly repeat this
One process becomes only until having been calculated, and when the number of cluster centre is more than a certain thresholding, has then searched out optimal as a result, i.e.
The length of target vehicle.
Information of vehicles matching unit 140 also receive millimetre-wave radar unit 130 transmission sail the vehicle through roadside unit vehicle
Information, is also used to compare the information of vehicles obtained by wireless communication unit 120, the vehicle that millimetre-wave radar unit 130 obtains
Whether the information of vehicles that information and video processing unit 110 obtain matches, if mismatching, issues warning information.Information of vehicles
Matching unit 140 can also individually compare the information of vehicles of the acquisition of wireless communication unit 120 and millimetre-wave radar unit 130 obtains
Information of vehicles, if mismatch, issue warning information.Information of vehicles matching unit 140 can also individually be compared by channel radio
Whether the information of vehicles that the information of vehicles and video processing unit 110 that news unit 120 obtains obtain matches, if mismatching, sends out
Warning information out.Information of vehicles matching unit 140 can also individually compare millimetre-wave radar unit 130 acquisition information of vehicles and
Whether the information of vehicles that video processing unit 110 obtains matches, if mismatching, issues warning information.Information of vehicles matching is single
Member 140 can also compare information of vehicles, the millimetre-wave radar unit 130 for comparing and being obtained by wireless communication unit 120 simultaneously and obtain
Whether the information of vehicles that the information of vehicles and video processing unit 110 obtained obtains matches, if mismatching, issues warning information.
Further, the roadside unit of the present embodiment may also include perception tracking cell 150, perceive 150 pairs of tracking cell views
The video image that frequency processing unit 110 acquires by deep learning method, with perceive the quantity of the vehicle at high speed crossing, flow,
And/or density；And/or perception tracking cell 150 passes through deep learning side to the video image that video processing unit 110 acquires
Method carries out dynamically track to vehicle to realize, and then can prejudge to the pre- travel route of the vehicle for driving into high speed crossing, and subtract
The interference that few information of vehicles to other vehicles obtains.
It can be realized using the present apparatus and mark, vehicle, vehicle body face are examined to the license plate, logo, vehicle of roadside unit vehicle of passing through
Color, sunshading board, and/or the identification of driver, and can be also traffic control department by the training and excavation to history mass data
It formulates comprehensive traffic management command predetermined plan and strong foundation is provided, and then improve traffic efficiency.Further, uniform spaces can be established
Benchmark and data sharing relate to car data center, provide big data concurrent processing and service, so support relate to vehicle management, operation and
Operation service.
Big data of the application based on video image acquisition passes through the training of mass data, energy using deep learning algorithm
The recognition success rate of vehicle detection is increased to 98%, phenomena such as effective solution " fee evasion ", can greatly be retrieved therefore
And the economic loss generated.Simultaneously ETC system is driven into or is driven out to using the alternative traditional ground induction coil identification vehicle of the application
The technology in lane greatly reduces construction and human cost.
Embodiment two:
Fig. 6 and Fig. 7 are please referred to, for the functional block diagram of the roadside unit receiving system of another embodiment and another
A kind of functional block diagram of the roadside unit control system of embodiment；Wherein roadside unit receiving system and roadside unit
Control system is connected by the connector in Fig. 6 with the socket in Fig. 7.Wherein, 16 core connectors can be used in connector, and socket can be used 16
Core socket.
Receiving system includes 5.8GHz transceiver module, locating module, radar module, video camera, Beidou module, power supply
Module and control module.Receiving system uses network RS485 interface and PHY signaling interface for data input/output interface,
For the data exchange between internal system and outside.In addition, control module may include programmable logic circuit FPGA.
5.8GHz transceiver module be tradition ETC outdoor roadside unit module, including 5.8G RF mixer, filter and
Amplifying circuit etc., the frequency microwave signal returned for receiving the OBU to pass through on roadside unit vehicle, then to the signal received
DSRC protocol resolution module is sent to after being mixed, filtered, and/or being amplified, DSRC protocol resolution module parses the vehicle
OBU information, and then obtain the information of the vehicle, and this information of vehicles is sent to control module.
Phased array wireless locating module can be used in locating module, and the location information of the vehicle for will acquire is sent to control
Module.The OBU localization method that phased array wireless locating module provides can refer to that application No. is " 201410593911.X ", the applyings date
For " on October 29th, 2014 ", the Chinese patent application of entitled " a kind of OBU localization method, RSU and ETC system " is complete
Text.
Radar module can be millimeter wave radar module, for receiving and emitting MMW RADAR SIGNAL USING, and according to receiving
The echo-signal of MMW RADAR SIGNAL USING obtain the information of vehicles of vehicle, and information of vehicles is sent to control module.
Beidou module, it may include GNSS (Global Navigation Satellite System, global navigation satellite system
System) unit, for receiving the satellite time of GNSS unit transmission, being mutually synchronized between realization roadside unit is not necessarily to synchronizer
With the equal outer hardware of synchronous cable, the local synchronization system formed in this way simplifies the ad-hoc network of multilane ETC
Scale reduces the integrated difficulty of ETC system, solves not passing through between roadside unit caused by due to constructional difficulties etc.
It connects synchronizer or synchronous cable carries out the synchronous problem of system.
Video camera acquires video image, and light compensating lamp carries out corresponding light filling function according to demand when video acquisition to light.
Video camera carries out data exchange by PHY external network interface and controller part.Receiving system is connect by RS485 network
Mouth carries out data exchange with control system, and the video information that will acquire is sent to video processing board and CPU.
Control system includes power-supply system, CPU, video processing board, PSAM card identifying system, OLED display and input
Outputting communication interface etc..Input and output communication interface includes RS485 interface, RS232 interface, PHY network interface and USB interface
Deng.Power-supply system provides electric power for each system in entire control section.
The received information of control module include 5.8GHz transceiver module obtain information of vehicles, locating module obtain vehicle on
The satellite standard time that the information of vehicles and Beidou module that location information, the radar module of OBU obtains obtain.Roadside unit control
The control module of system carries out received signal to be mutually matched comparison, sails whether the vehicle through roadside unit meets with confirmation
ETC system makes respective handling by requiring, for example, allowing or forbidding vehicle to be driven out to the lane ETC or issue warning information
Deng.
It will be understood by those skilled in the art that this application involves to image procossing, information of vehicles matching etc. need computer to hold
Capable program is all using the prior art.All or part of function of various methods can pass through hardware in above embodiment
Mode is realized, can also be realized by way of computer program.When function all or part of in above embodiment passes through meter
When the mode of calculation machine program is realized, which be can be stored in a computer readable storage medium, and storage medium may include:
Read-only memory, random access memory, disk, CD, hard disk etc. execute the program by computer to realize above-mentioned function.Example
Such as, program is stored in the memory of equipment, when executing program in memory by processor, can be realized it is above-mentioned whole or
Partial function.In addition, when function all or part of in above embodiment is realized by way of computer program, the program
Also it can store in the storage mediums such as server, another computer, disk, CD, flash disk or mobile hard disk, pass through downloading
Or copying and saving carries out version updating into the memory of local device, or to the system of local device, holds when by processor
When program in line storage, all or part of function in above embodiment can be realized.
Use above specific case is illustrated the present invention, is merely used to help understand the application, not to limit
The application processed.For those skilled in the art of the present invention, according to the thought of the application, can also make several
It is simple to deduce, deform or replace.
Claims (5)
1. a kind of roadside unit, which is characterized in that matched including wireless communication unit, video processing unit and information of vehicles single
Member, wherein
The wireless communication unit is obtained from the on board unit for carrying out data interaction with the on board unit on vehicle
The information of vehicles of the vehicle is simultaneously sent to the information of vehicles matching unit；
The video processing unit uses depth for acquiring the video image of vehicle, and to the video image of the vehicle
The method of habit, to identify the information of vehicles of the vehicle and be sent to the information of vehicles matching unit；
The information of vehicles matching unit, the vehicle sent for receiving the wireless communication unit and the video processing unit
Information, and judge whether are the information of vehicles that the wireless communication unit is sent and the information of vehicles that the video processing unit is sent
Matching issues warning information if mismatching.
2. roadside unit as described in claim 1, which is characterized in that further include,
Millimetre-wave radar unit, for receiving and emitting MMW RADAR SIGNAL USING, and according to the millimetre-wave radar received
The echo-signal of signal obtains the information of vehicles of the vehicle, is sent to the information of vehicles matching unit.
3. roadside unit as claimed in claim 2, which is characterized in that
The information of vehicles matching unit be also used to judge information of vehicles that the millimetre-wave radar unit that receives is sent with
Whether the information of vehicles that the wireless communication unit is sent matches, if mismatching, issues warning information；And/or the vehicle
The information of vehicles and the video that the millimetre-wave radar unit that information matching unit is also used to judge to receive is sent are handled
Whether the information of vehicles that unit is sent matches, if mismatching, issues warning information；And/or the information of vehicles matching unit
The vehicle that the information of vehicles and the video processing unit that the millimetre-wave radar unit for being also used to judge to receive is sent are sent
Whether the information of vehicles that information and the wireless communication unit are sent matches, if mismatching, issues warning information.
4. roadside unit as claimed in claim 3, which is characterized in that further include:
Tracking cell is perceived, the method for using deep learning to the video image of acquisition, to perceive the vehicle at high speed crossing
Quantity, flow, and/or density；And/or
Method for using deep learning to the video image of acquisition carries out dynamically track to the vehicle to realize.
5. roadside unit as described in any one of claims 1-3, which is characterized in that the information of vehicles of the vehicle includes vehicle
Board, vehicle, vehicle money, annual test mark, color, sunshading board, and/or driver.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201820444400.5U CN208367913U (en) | 2018-03-30 | 2018-03-30 | A kind of roadside unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201820444400.5U CN208367913U (en) | 2018-03-30 | 2018-03-30 | A kind of roadside unit |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN208367913U true CN208367913U (en) | 2019-01-11 |
Family
ID=64922293
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201820444400.5U Active CN208367913U (en) | 2018-03-30 | 2018-03-30 | A kind of roadside unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN208367913U (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111582174A (en) * | 2020-05-09 | 2020-08-25 | 广西信路威科技发展有限公司 | RSU and multi-target radar detection result matching method based on image recognition |
| CN113256826A (en) * | 2021-03-29 | 2021-08-13 | 北京云星宇交通科技股份有限公司 | ETC intelligence trackside unit |
-
2018
- 2018-03-30 CN CN201820444400.5U patent/CN208367913U/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111582174A (en) * | 2020-05-09 | 2020-08-25 | 广西信路威科技发展有限公司 | RSU and multi-target radar detection result matching method based on image recognition |
| CN113256826A (en) * | 2021-03-29 | 2021-08-13 | 北京云星宇交通科技股份有限公司 | ETC intelligence trackside unit |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108510734A (en) | A kind of information of vehicles matching process of roadside unit and a kind of roadside unit | |
| WO2021068470A1 (en) | Radar signal-based identity and gesture recognition method | |
| CN106373427B (en) | Parking management method and parking management system | |
| CN108765937B (en) | Vehicle identification device, roadside unit and method for ETC system | |
| CN111427031B (en) | Identity and gesture recognition method based on radar signals | |
| US11313950B2 (en) | Machine learning based highway radar vehicle classification across multiple lanes and speeds | |
| CN109583436A (en) | A kind of gesture recognition system based on millimetre-wave radar | |
| CN208367913U (en) | A kind of roadside unit | |
| CN104931977A (en) | Obstacle identification method for smart vehicles | |
| CN107134143A (en) | A kind of vehicle flowrate based on continuous wave radar sentences method for distinguishing with vehicle | |
| CN102968640B (en) | Decompose and the Classification of Polarimetric SAR Image method of data distribution characteristics based on Freeman | |
| CN109444872B (en) | Pedestrian and vehicle distinguishing method and device, computer equipment and storage medium | |
| CN112053584B (en) | Road tooth parking space state prediction management system based on geomagnetism, radar and camera shooting and management method thereof | |
| Cho et al. | A support vector machine approach to CMOS-based radar signal processing for vehicle classification and speed estimation | |
| CN103679202A (en) | Method and device suitable for classifying vegetation through optical remote sensing satellite images | |
| Wang et al. | Multi-classification of UWB signal propagation channels based on one-dimensional wavelet packet analysis and CNN | |
| CN113192337A (en) | Parking space detection method based on millimeter wave radar | |
| CN102360496B (en) | SAR (synthetic aperture radar) image segmentation method based on GMM (gaussian mixture model) parameter transferring and clustering | |
| AU2017287973A1 (en) | System, apparatus and method for non-invasive avian egg fertility detection | |
| Yang et al. | Urban traffic congestion alleviation system based on millimeter wave radar and improved probabilistic neural network | |
| CN106408050B (en) | Electronic tag communication method for ETC system, road side unit and ETC system | |
| CN207020727U (en) | A kind of roadside unit based on DSRC transaction with Radar Technology | |
| CN205541433U (en) | Road vehicle system of testing speed based on communication function radio frequency license plate has | |
| CN112926526B (en) | Parking detection method and system based on millimeter wave radar | |
| CN114884597A (en) | Marine radio wave propagation prediction method based on neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |