CN109815933A - A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method - Google Patents
A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method Download PDFInfo
- Publication number
- CN109815933A CN109815933A CN201910110065.4A CN201910110065A CN109815933A CN 109815933 A CN109815933 A CN 109815933A CN 201910110065 A CN201910110065 A CN 201910110065A CN 109815933 A CN109815933 A CN 109815933A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- automatic identification
- evidence
- obtaining
- axle
- 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.)
- Pending
Links
Landscapes
- Traffic Control Systems (AREA)
Abstract
Present disclose provides a kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method, wherein, described device includes the formal dress snap unit and side dress snap unit of front-end server and linkage setting, the formal dress snap unit, it is configured as capturing vehicle frontal, and picture is captured according to front, automatic identification is carried out to vehicle license plate and vehicle；The side fills snap unit, is configured as with formal dress snap unit synchronization action for capturing to vehicular sideview, and captures picture according to side and carry out automatic identification to the vehicle number of axle；The front-end server, it is configured as when axle number reaches setting value, it is found according to the serial number that record is captured in side and corresponds to front candid photograph picture for illegal synthesis, and according to the synthesis and upload of progress picture after the information and illegal activities of evidence obtaining requirement superposition vehicle.The disclosure is utilized to the vehicle number of axle automatic identification in high clear video image, is realized to five axis, six axis and the above vehicle automatic screening in truck.
Description
Technical field
This disclosure relates to intelligent traffic monitoring technical field more particularly to a kind of multiaxle trucks automatic identification apparatus for obtaining evidence, be
System and method.
Background technique
During Widening of Freeway, to guarantee that Expressway Road is unimpeded, prevents road traffic accident with all strength, need to prohibit
Only five axis and with Truck highway traveling.But with regard to known to inventor, original electric police grasp shoot system can not be done
To the automatic identification to Vehicle Axles number, even if being screened to a certain bayonet according to yellow card spoke part by vehicle administrating system
It is still very big that car data amount is crossed afterwards, and it is very low to carry out secondary review efficiency to the data after screening by manually.
Therefore, illegal activities are more effectively accurately hit in order to realize, high speed traffic police passes through the people on duty that arrange an order according to class and grade daily
It is alert that scene evidence taking is carried out to forbidden truck illegal activities to high speed roadside.Although can solve illegal evidence obtaining in this way
Difficult problem, but round-the-clock real-time, traffic police personal safety the problems such as be unable to get and be effectively ensured again.
The formation of " AI+ traffic " mode has led the new route of future transportation, intelligent security guard development, video monitoring level
From initial digitlization, networking, high Qinghua to intelligent fast transition.With traffic video access scale it is swift and violent increase and
Requirement of the video monitoring to high definition, intelligence, networking is higher and higher, and the traditional algorithm of shallow hierarchy analysis identification has been unable to satisfy prison
Control demand.
Summary of the invention
In order to solve the deficiencies in the prior art, present disclose provides a kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and
Method, by the association of side camera and front camera, carries out linkage candid photograph to vehicle in lane using video interlink technology,
Using deep learning algorithm, to the vehicle number of axle automatic identification captured in picture, realize in truck five axis, six axis and with
Upper vehicle automatic screening, realization rush forbidden behavior to illegal vehicle and carry out evidence obtaining synthesis.
To achieve the goals above, the technical solution of the disclosure is as follows:
A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, including front-end server and linkage setting formal dress snap unit and
Side fills snap unit, wherein
The formal dress snap unit, is configured as capturing vehicle frontal, and captures picture to vehicle according to front
License plate and vehicle carry out automatic identification；
The side fills snap unit, is configured as with formal dress snap unit synchronization action for grabbing to vehicular sideview
It claps, and picture is captured according to side, automatic identification is carried out to the vehicle number of axle；
The front-end server is configured as when axle number reaches setting value, and the serial number of record is captured according to side
It finds corresponding front and captures picture, and according to the synthesis for carrying out picture after the information and illegal activities of evidence obtaining requirement superposition vehicle
And it uploads.
Further, the formal dress snap unit is by Recognition Algorithm of License Plate and vehicle targets to vehicle frontal picture
License plate and vehicle cab recognition are carried out, and vehicle frontal picture and recognition result are sent to front-end server.
Further, the side dress snap unit, which isolates to vehicle by deep learning algorithm, identifies, and when vehicle is isolated
Vehicular sideview picture and recognition result are sent to front-end server when reaching setting value.
Further, the deep learning algorithm specifically includes:
According to the feature of a variety of models under varying environment, vehicle sample training library is established；
Establish depth convolutional neural networks model, and by vehicle sample training library input depth convolutional neural networks model into
Row training；
Number of axle detection identification is carried out to real-time vehicle pictures using the depth convolutional Neural model after training.
Further, the vehicle feature includes vehicle's contour, vehicle dimension and the vehicle number of axle.
Further, the apparatus for obtaining evidence further includes light filling unit, and the light filling unit includes formal dress environment light-supplementing lamp, just
Quick-fried sudden strain of a muscle light compensating lamp and side dress light compensating lamp are filled, the light filling unit is effectively promoted and obtained for ensuring night apparatus for obtaining evidence blur-free imaging
The accuracy for taking license plate number and the vehicle number of axle to detect.
Further, the apparatus for obtaining evidence is connected with a trigger device, and the trigger device is installed on to be set on highway
It captures at trigger position.
A kind of multiaxle trucks automatic identification evidence-obtaining system, including multiaxle trucks automatic identification apparatus for obtaining evidence as described above.
Further, the system also includes back-end central platform, the multiaxle trucks automatic identification apparatus for obtaining evidence passes through net
Network exchange and transmission unit are connected with back-end central platform.
A kind of multiaxle trucks automatic identification evidence collecting method, including multiaxle trucks automatic identification apparatus for obtaining evidence as described above, specifically
Include:
It has detected whether that vehicle passes through, vehicle side image has been acquired if having, and obtained vehicle frontal image；
The vehicle number of axle, vehicle license plate and vehicle vehicle are identified respectively, and judge whether the vehicle number of axle is more than setting
Value；
When the vehicle number of axle is more than or equal to setting value, corresponding front is found according to the serial number that record is captured in side and is captured
Image carries out the synthesis of vehicle violation image, and carries out on picture after requiring the information and illegal information that are superimposed vehicle according to evidence obtaining
It passes.
Compared with prior art, the beneficial effect of the disclosure is:
1) processing of multiaxle trucks intelligent recognition is sunk to front-end A I smart machine by the disclosure, and edge calculations theory is introduced and is handed over
Logical illegal activities identification, solves the problems, such as that traditional ARM+DSP and centric computing model recognition efficiency are low.
2) disclosure is obtained using the intelligent candid machine for having deep learning algorithm by the analysis to video capture picture
Truck axle number, compared to tradition by identification methods such as laser, coils, with construction cost is low, the construction period is short, Yi Wei
The advantages such as shield.
3) disclosure is by video interlink technology, using the association of side camera and front camera, to vehicle in lane into
Row linkage is captured, and by deep learning algorithm to five axis and the above truck automatic screening, solves traditional violation snap-shooting system
Forbidden behavior automatic evidence-collecting problem can not be rushed to five axis and with Truck.
4) disclosure solves highway law enforcement agency to forbidden truck by the application of artificial intelligence science and technology
Manage difficult problem.
5) disclosure system uses damascene structures, and safety and stability, black box is few, multiple functional, debugs convenient for construction,
With good generalization, can be carried out for the other highways of the whole province under five axis and the control of the above vehicle and divided lane driving mode
It manages forbidden vehicle and effective solution is provided.
Detailed description of the invention
The accompanying drawings constituting a part of this application is used to provide further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation are not constituted an undue limitation on the present application for explaining the application.
Fig. 1 is the structural schematic diagram of a certain embodiment of disclosure multiaxle trucks automatic identification apparatus for obtaining evidence；
Fig. 2 is the structural schematic diagram of a certain embodiment of disclosure multiaxle trucks automatic identification evidence-obtaining system；
Fig. 3 is the deep learning algorithm model figure of a certain embodiment of the disclosure；
Fig. 4 is that the convolutional neural networks of a certain embodiment of the disclosure carry out the building schematic diagram of axle identification model；
Fig. 5 is the training process schematic diagram of a certain embodiment of the disclosure；
Fig. 6 is the back-end central platform composed structure schematic diagram of a certain embodiment of the disclosure；
Fig. 7 is the multiaxle trucks automatic identification evidence-obtaining system flow chart of a certain embodiment of the disclosure；
In figure: 1, the quick-fried sudden strain of a muscle light compensating lamp of formal dress；2, formal dress snap unit；3, formal dress environment light-supplementing lamp；4, side fills light compensating lamp；5,
Side fills snap unit；6, front-end server；7, back-end central platform；8, network exchange and transmission unit.
Specific embodiment
The disclosure is described further with specific embodiment with reference to the accompanying drawing.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms used herein has usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
In the disclosure, term for example "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", " side ",
The orientation or positional relationship of the instructions such as "bottom" is to be based on the orientation or positional relationship shown in the drawings, only to facilitate describing this public affairs
The relative for opening each component or component structure relationship and determination, not refers in particular to either component or element in the disclosure, cannot understand
For the limitation to the disclosure.
In the disclosure, term such as " affixed ", " connected ", " connection " be shall be understood in a broad sense, and indicate may be a fixed connection,
It is also possible to be integrally connected or is detachably connected；It can be directly connected, it can also be indirectly connected through an intermediary.For
The related scientific research of this field or technical staff can determine the concrete meaning of above-mentioned term in the disclosure as the case may be,
It should not be understood as the limitation to the disclosure.
As one or more embodiments, as shown in Figure 1, a kind of multiaxle trucks automatic identification apparatus for obtaining evidence, including front end take
The formal dress snap unit 2 and side of business device 6 and linkage setting fill snap unit 5, wherein
The formal dress snap unit 2, is configured as capturing vehicle frontal, and captures picture to vehicle according to front
License plate and vehicle carry out automatic identification；
The side fills snap unit 5, is configured as capturing vehicular sideview, and captures picture to vehicle according to side
The number of axle carries out automatic identification；
The front-end server 6 is configured as when axle number reaches setting value, and the serial number of record is captured according to side
It finds corresponding front and captures picture for illegal synthesis, and require the information of superposition vehicle and illegal activities laggard according to evidence obtaining
The synthesis and upload of row picture.
The formal dress snap unit 2 carries out license plate to vehicle frontal picture by Recognition Algorithm of License Plate and vehicle targets
And vehicle cab recognition, and vehicle frontal picture and recognition result are sent to front-end server 6.
The side dress snap unit 5, which isolates to vehicle by deep learning algorithm, to be identified, and reaches setting when vehicle is isolated
Vehicular sideview picture and recognition result are sent to front-end server 6 when value.
The present embodiment use lane it is positive, laterally a set of AI intelligent identification equipment (snap unit) is respectively installed, pass through two
The identification methods such as the synchronous interaction acquisition, multidimensional image fitting and characteristic fusion of equipment are covered, vehicle is effectively increased
Imaging definition, discrimination and accuracy.
Formal dress snap unit 2 and side dress snap unit 5 are captured by control signal linkage in the present embodiment, to guarantee two
The vehicle scene of video capture is consistent.
In the present embodiment formal dress snap unit 2 and side dress snap unit 5 be 3,200,000 pixel bayonet capture machines, 2 3,200,000
Pixel bayonet capture machine covers 1 lane (3.75 meters of national standard width), guarantees all standing of field range.It is separated in addition to that can capture
Outside truck, it was also equipped with the function of vehicle record.
The deep learning algorithm specifically includes:
As shown in figure 3, establishing vehicle sample training library according to the feature of a variety of models under varying environment；
Establish depth convolutional neural networks model, and by vehicle sample training library input depth convolutional neural networks model into
Row training；
Number of axle detection identification is carried out to real-time vehicle pictures using the depth convolutional Neural model after training.
Convolutional neural networks have good resistivity to translation, scaling equiaffine deformation, can effectively overcome vehicle
Appearance is changeable, the influence of video camera distance, height and the factors such as angle and ambient lighting.
The vehicle feature includes vehicle's contour, vehicle dimension and the vehicle number of axle.
It is many in sample database specifically, vehicle data sample database high definition bayonet vehicle image library on highway
High definition bayonet of the image from each section, shooting environmental, time, angle are all different, so the data in the present embodiment
Library can simulate the identification situation under actual scene well, can preferably test the performance of recognizer.
Depth convolutional neural networks characteristic model (Convolutional Neural Networks, CNN) training process packet
Include successively building CNN and training two parts CNN.Using five layer networks, the building process of network such as Fig. 4 in the present embodiment
It is shown.First layer is the input of network, and the second layer is the convolutional layer of network, convolutional layer by filter with can biasing set and rolled up
Product, obtains Ni characteristic pattern.Third layer is the down-sampling layer of network, and all down-sampling layers are all to pass sequentially through to handle as follows
It arrives: the four pixel summations of every neighborhood, weight (convolution nuclear element) weighting, in addition biasing, small finally by an influence function core
Sigmoid activation primitive.Then with onesize filter convolution third layer, the 4th layer of characteristic pattern is obtained, is equally passed through
Down-sampling obtains layer 5, and the characteristic pattern of layer 5 is finally arranged as a column vector, obtains final feature vector.
The training part is as shown in figure 5, include the following two stage:
A) the forward-propagating stage.Training sample is inputted into CNN network, obtains reality output；
B) back-propagation phase.The error for calculating reality output and ideal outlet chamber, is successively passed by back-propagation algorithm
Error is broadcast, and updates each layer weight.
After obtaining feature vector, using classifier, classify, exports the vehicle number of axle.
Edge calculations theory is introduced intelligent transportation field, integrated deep learning algorithm to front-end camera etc. by the present embodiment
Edge device carries out the dedicated chip design that 5-axle car captures detection, is intelligently calculated by high-performance calculation chip and image recognition
Method is energized edge device, detection, extraction, modeling and the parsing of video image target is realized at edge, a large amount of of image analysis
It calculates pressure uniformly to share in the large-scale edge calculations resource of little particle, only the structuring valid data of refining be uploaded public
Pacify traffic and integrate maneuvering platform processing, effectively reduce the transmission and carrying cost of video flowing, shares center calculation and storage pressure
Power realizes that efficiency maximizes, while terminal being made to realize low delay and fast-response, and 5~10 times are improved in disposed of in its entirety speed,
Real-time property, accuracy are more guaranteed.
In view of the situation of night side dress capture machine identification scene brightness deficiency, the apparatus for obtaining evidence further includes light filling list
Member, the light filling unit are double quick-fried sudden strain of a muscle+bis- strobe modes, specifically include quick-fried 1 and of sudden strain of a muscle light compensating lamp of formal dress environment light-supplementing lamp 3, formal dress
Side fills light compensating lamp 4.The light filling unit can ensure that night apparatus for obtaining evidence blur-free imaging, is effectively promoted and obtains license plate number and vehicle
The accuracy of number of axle detection.
The apparatus for obtaining evidence is connected with a trigger device, and the trigger device is installed on the candid photograph trigger bit set on highway
Set place.
Project is in implementation process, by taking 5-axle car as an example, according to " national highway engineering technology criterion ", " over-limited transport vehicle
Traveling highway administration regulation " etc. relevant regulations, the outer dimension of 5-axle car is generally in 5 meters of bodywork height, 2.5 meters of vehicle width
Interior, vehicle body integrally grows the actual characteristic in 18 meters, repeatedly to the upright bar height of intelligent acquisition equipment, imaging angle, two equipment rooms
Away from etc. parameters proved, tested, final to determine that upright bar height is 6-7 meter, 4-6 meters of equipment spacing, imaging angle is at 22 meter
Triggering.
Specifically, because of the imaged viewing angle factor of intelligent acquisition equipment, if including vehicle whole profile when imaging, and can be clear
It can be seen that equipment should keep certain distance with vehicle when axle.Therefore through scene, experiment is tested repeatedly, when vehicle exists in the present embodiment
When being imaged at about 22 meters before forward direction candid photograph equipment, image, which had both been avoided, can not acquire vehicle whole profile because distance is close,
Evaded again because distance it is remote caused by be repeatedly imaged the problems such as, vehicle characteristics picture that is positive and laterally acquiring is ideal.Therefore this
The trigger position of embodiment is at about 22 meters before positive candid photograph equipment.
When lateral equipment is mounted on perpendicular at vehicle, the vehicle number of axle is most obvious, but equipment installation position has been in high
Except fast road range；It is lateral to be imaged when lateral equipment is closer with forward device spacing, and spacing is less than 4 meters in specific experiment
Axle feature it is unobvious, axle can not accurately identify axle number at adhesion property；When lateral equipment and forward device spacing exist
It include the features such as vehicle whole profile, axle in image after imaging, physical separation is more apparent between each axle when at 4-6 meters.
As shown in Fig. 2, a kind of multiaxle trucks automatic identification evidence-obtaining system, including multiaxle trucks automatic identification as described above evidence obtaining
Device.
The system also includes back-end central platform 7, the multiaxle trucks automatic identification apparatus for obtaining evidence by network exchange and
Transmission unit 8 is connected with back-end central platform 7.
As shown in fig. 6, the back-end central platform includes platform base functional module and business application module.
Platform base functional module is mainly the function services that the specific business of system carries out offer basis, including data are adopted
Collection, data storage, video preview, playing back videos, electronic map, system administration configuration and system O&M etc..
The business application module includes traffic base data management system, traffic condition monitoring system, traffic safety state
Gesture assessment system, vehicle cloud analysis system, traffic offence evidence-obtaining system and five axis truck capturing systems.
As shown in fig. 7, a kind of multiaxle trucks automatic identification evidence collecting method, including multiaxle trucks automatic identification as described above evidence obtaining
Device specifically includes:
It has detected whether that vehicle passes through, vehicle side image has been acquired if having, and obtained vehicle frontal image；
The vehicle number of axle, vehicle license plate and vehicle vehicle are identified respectively, and judge whether the vehicle number of axle is more than setting
Value；
When the vehicle number of axle is more than or equal to setting value, corresponding front is found according to the serial number that record is captured in side and is captured
Image carries out the synthesis of vehicle violation image, and carries out on picture after requiring the information and illegal information that are superimposed vehicle according to evidence obtaining
It passes.
Specifically, by taking 5-axle car or more as an example, 5-axle car automatic evidence-collecting identifying system process mainly include " detection -- grab
Four major parts of bat-identification-evidence obtaining ".When system detection has vehicle to pass through to road, snap unit 5 vehicle in side is grabbed
While bat, vehicle snapshot is carried out by the positive snap unit 2 of control signal linkage, passes through the purpose of control signal linkage candid photograph
It is that the serial number that vehicle is crossed in positive side when guaranteeing illegal synthesis is consistent.Vehicle characteristics and license plate are identified simultaneously, according to vehicle side
Piece determines the axle number of lorry, is just found pair according to serial number that record is captured in side when axle number reaches five axis or more
It answers front bayonet to capture picture for illegal synthesis, and is carried out after requiring the information and illegal activities that are superimposed vehicle according to evidence obtaining
The synthesis of picture uploads.
The construction of multiaxle trucks automatic identification evidence-obtaining system in the disclosure, realizes effective control to forbidden vehicle first,
Improve the professional skill of traffic administration.Next reduces construction cost, by realizing that video is known using existing hardware condition
Not, building time is substantially reduced.
(1) traffic violation people is frightened, enhances driver's traffic law-abiding awareness.In Jinan, Zibo, the Weihe River
Larger impact is formed within the scope of the line highway of the Jinan-Qingdaos north such as mill.On the one hand, illegal activities can be carried out in conjunction with law enforcement process
Effectively punish；On the other hand, make each truck driver that can feel a kind of invisible pressure, force it to driving self
The behavior of sailing is constrained, and numerous drivers is promoted to eliminate idea of leaving things to chance, improves consciousness of abiding by the law, and reduces traffic violation and road
Road traffic accident, to reduce the personal injury of the people, vehicle asset loss, brought social benefit exceeds
The range that money is measured.
(2) traffic safety, maintenance driving order are improved.During reorganization and expansion, road conditions are complicated, current condition ratio
It is poor, jam situation caused by cart is gone slowly for a long time is advantageously reduced by the control to oversize vehicle, to realize vehicle height
Imitate safe passing.
(3) be other other highway extension projects of the whole province when need to carry out vehicle control and divided lane driving mode under
It manages forbidden vehicle and provides effective solution.
(4) construction of 5-axle car automatic identification evidence-obtaining system brings following benefit to traffic administration:
1, front end recognition system is basis and the support of traffic police's backend services application, front-end architecture type is more abundant,
More accurate, front end stability is higher, can just support the more preferable application of backend services, meets data query statistics, studies and judges point
The business demands such as application are expanded in analysis.
3, traffic safety, maintenance driving order are improved.
By the control to 5-axle car or more, reduce because large car is current slowly, manipulates not flexible caused road
The generation of jam situation, to realize safe and orderly driving order.
(5) construction of multiaxle trucks automatic identification evidence-obtaining system realizes effective control to forbidden vehicle, improves traffic
The professional skill of management.Video identification is realized using existing hardware condition simultaneously, is reduced construction cost, is substantially reduced and build
If the time.Since 2 months installed multiaxle trucks evidence taking equipment, capture altogether operation lorry 36855 times in violation of rules and regulations.Five axis and to get in stocks
Vehicle traffic volume is by original daily more than 400 more than 120 times declined till now, and decline nearly 70%, traffic order is shown
Writing improves.
Two, implement front and back comparison
On November 1st, 2017, Jinan-Qingdao north line start five axis of restricted driving and with Trucks.At the beginning of restricted driving, people's police utilize camera people
Work capture in violation of rules and regulations operation five axis and with Truck, although at that time in violation of rules and regulations operation vehicle had 500 or so daily, by police strength,
Time restriction, practical quantity of effectively capturing daily is only more than 40 items, and effective candid photograph rate is less than 10%.2 months in 2018, start to pacify
It fills multiaxis lorry and captures equipment, it, can only be by manually from all Huangs without lorry number of axle automatic identification function because technology is not perfect
Board lorry data screening, it is time-consuming and laborious；By 1000 artificial screening, typing, audit valid data, about 35 hours are needed；
Equipment is captured to multiaxis lorry and is optimized, realizes the automatic identification of the lorry number of axle, discrimination exists in March, 2018
90% or more, forensic data can be directly entered integrated maneuvering platform and carry out audit check and correction, 1000 significant figures of screening audit at present
According to, it is only necessary to 3 hours.
Since 2 month of this year installs multiaxis lorry automatic identification evidence-obtaining system, December is ended, Jinan-Qingdao north line is related to lorry
Traffic accident reduces 234 on year-on-year basis, and decline 33.38%, wherein general procedure accident reduces 10, declines 58.82%, dead people
Number reduces 11 people, decline 64.71%.
The foregoing is merely preferred embodiment of the present application, are not intended to limit this application, for the skill of this field
For art personnel, various changes and changes are possible in this application.Within the spirit and principles of this application, made any to repair
Change, equivalent replacement, improvement etc., should be included within the scope of protection of this application.
Although above-mentioned be described in conjunction with specific embodiment of the attached drawing to the disclosure, model not is protected to the disclosure
The limitation enclosed, those skilled in the art should understand that, on the basis of the technical solution of the disclosure, those skilled in the art are not
Need to make the creative labor the various modifications or changes that can be made still within the protection scope of the disclosure.
Claims (10)
1. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence, it is characterised in that: the formal dress including front-end server and linkage setting
Snap unit and side fill snap unit, wherein
The formal dress snap unit, is configured as capturing vehicle frontal, and captures picture to vehicle license plate according to front
Automatic identification is carried out with vehicle；
The side fills snap unit, is configured as with formal dress snap unit synchronization action for capturing to vehicular sideview, and
Picture is captured according to side, and automatic identification is carried out to the vehicle number of axle；
The front-end server is configured as when axle number reaches setting value, is found according to the serial number that record is captured in side
Corresponding positive picture of capturing is used for illegal synthesis, and carries out figure after requiring the information and illegal activities that are superimposed vehicle according to evidence obtaining
The synthesis and upload of piece.
2. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence as described in claim 1, which is characterized in that the formal dress snap unit
License plate and vehicle cab recognition carried out to vehicle frontal picture by Recognition Algorithm of License Plate and vehicle targets, and by vehicle frontal figure
Piece and recognition result are sent to front-end server.
3. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence as described in claim 1, which is characterized in that the side fills snap unit
It is isolated and is identified to vehicle by deep learning algorithm, and tie vehicular sideview picture and identification when vehicle isolates and reaches setting value
Fruit is sent to front-end server.
4. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence as claimed in claim 3, which is characterized in that the deep learning algorithm
It specifically includes:
According to the feature of a variety of models under varying environment, vehicle sample training library is established；
Depth convolutional neural networks model is established, and vehicle sample training library input depth convolutional neural networks model is instructed
Practice；
Number of axle detection identification is carried out to real-time vehicle pictures using the depth convolutional Neural model after training.
5. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence as claimed in claim 3, which is characterized in that the vehicle feature includes
Vehicle's contour, vehicle dimension and the vehicle number of axle.
6. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence as described in claim 1, which is characterized in that the apparatus for obtaining evidence also wraps
Light filling unit is included, the light filling unit includes formal dress environment light-supplementing lamp, the quick-fried sudden strain of a muscle light compensating lamp of formal dress and side dress light compensating lamp, the light filling
Unit is for ensuring night apparatus for obtaining evidence blur-free imaging.
7. a kind of multiaxle trucks automatic identification apparatus for obtaining evidence as described in claim 1, which is characterized in that the apparatus for obtaining evidence and one
Trigger device is connected, and the trigger device is installed at the candid photograph trigger position set on highway.
8. a kind of multiaxle trucks automatic identification evidence-obtaining system, which is characterized in that including multiaxle trucks as claimed in claim 1
Automatic identification apparatus for obtaining evidence.
9. a kind of multiaxle trucks automatic identification evidence-obtaining system as claimed in claim 8, which is characterized in that the system also includes rear
Central platform is held, the multiaxle trucks automatic identification apparatus for obtaining evidence passes through network exchange and transmission unit and back-end central platform phase
Even.
10. a kind of multiaxle trucks automatic identification evidence collecting method, including multiaxle trucks automatic identification as claimed in claim 1 take
Card device characterized by comprising
It has detected whether that vehicle passes through, vehicle side image has been acquired if having, and obtained vehicle frontal image；
The vehicle number of axle, vehicle license plate and vehicle vehicle are identified respectively, and judge whether the vehicle number of axle is more than setting value；
When the vehicle number of axle is more than or equal to setting value, corresponding front is found according to the serial number that record is captured in side and captures image
The synthesis of vehicle violation image is carried out, and carries out picture upload after requiring the information and illegal information that are superimposed vehicle according to evidence obtaining.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910110065.4A CN109815933A (en) | 2019-02-11 | 2019-02-11 | A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910110065.4A CN109815933A (en) | 2019-02-11 | 2019-02-11 | A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109815933A true CN109815933A (en) | 2019-05-28 |
Family
ID=66606448
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910110065.4A Pending CN109815933A (en) | 2019-02-11 | 2019-02-11 | A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109815933A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110516524A (en) * | 2019-06-26 | 2019-11-29 | 东南大学 | Vehicle number of axle recognition methods based on Mask R-CNN in a kind of traffic scene |
| CN111325146A (en) * | 2020-02-20 | 2020-06-23 | 吉林省吉通信息技术有限公司 | Truck type and axle type identification method and system |
| CN113537238A (en) * | 2021-07-05 | 2021-10-22 | 上海闪马智能科技有限公司 | Information processing method and image recognition device |
| CN118334875A (en) * | 2024-05-16 | 2024-07-12 | 江苏新菲鹤信息科技有限公司 | Motor vehicle side window snapshot system |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129775A (en) * | 2010-12-30 | 2011-07-20 | 上海安防电子有限公司 | Method and system for obtaining evidence by capturing vehicles at traffic crossing under panoramic video detection |
| CN102564546A (en) * | 2010-12-14 | 2012-07-11 | 金基太 | Method and system for detecting overload and unlawful measurement of vehicle |
| CN102768800A (en) * | 2012-07-12 | 2012-11-07 | 复旦大学 | Motor vehicle red light running rule violation evidence obtaining method based on high-definition videos |
| CN103150774A (en) * | 2013-03-07 | 2013-06-12 | 吉林省高速公路管理局 | System and method for identifying vehicles of highway green channel |
| CN103794056A (en) * | 2014-03-06 | 2014-05-14 | 北京卓视智通科技有限责任公司 | Vehicle type accurate classification system and method based on real-time double-line video stream |
| CN105303824A (en) * | 2015-09-07 | 2016-02-03 | 北京信路威科技股份有限公司 | Wireless triggering image collection device and method employing same for obtaining evidence of traffic violation |
| CN105427619A (en) * | 2015-12-24 | 2016-03-23 | 上海新中新猎豹交通科技股份有限公司 | Vehicle following distance automatic recording system and method |
| CN106127107A (en) * | 2016-06-14 | 2016-11-16 | 宁波熵联信息技术有限公司 | The model recognizing method that multi-channel video information based on license board information and vehicle's contour merges |
| CN106530747A (en) * | 2016-11-01 | 2017-03-22 | 公安部交通管理科学研究所 | Novel vehicle-mounted evidence collection system and method |
| CN106982319A (en) * | 2016-01-19 | 2017-07-25 | 杭州羊道科技有限公司 | A kind of violation violation snap-shooting prosecution system violating the regulations |
| CN107481525A (en) * | 2017-08-25 | 2017-12-15 | 安徽实运信息科技有限责任公司 | A kind of vehicle on highway illegal road occupation capturing system |
| CN207097236U (en) * | 2017-08-10 | 2018-03-13 | 江苏网进科技股份有限公司 | A kind of overspeed snapping system for traffic |
| CN108091142A (en) * | 2017-12-12 | 2018-05-29 | 公安部交通管理科学研究所 | For vehicle illegal activities Tracking Recognition under highway large scene and the method captured automatically |
| CN108334892A (en) * | 2017-12-26 | 2018-07-27 | 新智数字科技有限公司 | A kind of model recognizing method, device and equipment based on convolutional neural networks |
| CN108694399A (en) * | 2017-04-07 | 2018-10-23 | 杭州海康威视数字技术股份有限公司 | Licence plate recognition method, apparatus and system |
-
2019
- 2019-02-11 CN CN201910110065.4A patent/CN109815933A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102564546A (en) * | 2010-12-14 | 2012-07-11 | 金基太 | Method and system for detecting overload and unlawful measurement of vehicle |
| CN102129775A (en) * | 2010-12-30 | 2011-07-20 | 上海安防电子有限公司 | Method and system for obtaining evidence by capturing vehicles at traffic crossing under panoramic video detection |
| CN102768800A (en) * | 2012-07-12 | 2012-11-07 | 复旦大学 | Motor vehicle red light running rule violation evidence obtaining method based on high-definition videos |
| CN103150774A (en) * | 2013-03-07 | 2013-06-12 | 吉林省高速公路管理局 | System and method for identifying vehicles of highway green channel |
| CN103794056A (en) * | 2014-03-06 | 2014-05-14 | 北京卓视智通科技有限责任公司 | Vehicle type accurate classification system and method based on real-time double-line video stream |
| CN105303824A (en) * | 2015-09-07 | 2016-02-03 | 北京信路威科技股份有限公司 | Wireless triggering image collection device and method employing same for obtaining evidence of traffic violation |
| CN105427619A (en) * | 2015-12-24 | 2016-03-23 | 上海新中新猎豹交通科技股份有限公司 | Vehicle following distance automatic recording system and method |
| CN106982319A (en) * | 2016-01-19 | 2017-07-25 | 杭州羊道科技有限公司 | A kind of violation violation snap-shooting prosecution system violating the regulations |
| CN106127107A (en) * | 2016-06-14 | 2016-11-16 | 宁波熵联信息技术有限公司 | The model recognizing method that multi-channel video information based on license board information and vehicle's contour merges |
| CN106530747A (en) * | 2016-11-01 | 2017-03-22 | 公安部交通管理科学研究所 | Novel vehicle-mounted evidence collection system and method |
| CN108694399A (en) * | 2017-04-07 | 2018-10-23 | 杭州海康威视数字技术股份有限公司 | Licence plate recognition method, apparatus and system |
| CN207097236U (en) * | 2017-08-10 | 2018-03-13 | 江苏网进科技股份有限公司 | A kind of overspeed snapping system for traffic |
| CN107481525A (en) * | 2017-08-25 | 2017-12-15 | 安徽实运信息科技有限责任公司 | A kind of vehicle on highway illegal road occupation capturing system |
| CN108091142A (en) * | 2017-12-12 | 2018-05-29 | 公安部交通管理科学研究所 | For vehicle illegal activities Tracking Recognition under highway large scene and the method captured automatically |
| CN108334892A (en) * | 2017-12-26 | 2018-07-27 | 新智数字科技有限公司 | A kind of model recognizing method, device and equipment based on convolutional neural networks |
Non-Patent Citations (7)
| Title |
|---|
| ANSHUL GOYAL等: "A Neural Network based Approach for the Vehicle Classification", 《2007 IEEE SYMPOSIUM ON COMPUTATIONAL INTELLIGENCE IN IMAGE AND SIGNAL PROCESSING》 * |
| CARLOS SUN等: "Inductive Classifying Artificial Network for Vehicle Type Categorization", 《2003 COMPUTER–AIDED CIVIL AND INFRASTRUCTURE ENGINEERING》 * |
| MAJURA F. SELEKWA等: "Setting up a Probabilistic Neural Network for Classification of Highway Vehicles", 《INTERNATIONAL JOURNAL OF COMPUTATIONAL INTELLIGENCE AND APPLICATIONS》 * |
| VALERIAN KWIGIZILE等: "Highway Vehicle Classification by Probabilistic Neural Networks", 《FLAIRS CONFERENCE, 2004》 * |
| 丁乐乐: "基于深度学习和强化学习的车辆定位与识别", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王茜: "基于深度神经网络的汽车车型识别", 《现代计算机（普及版）》 * |
| 陈宏彩: "基于卷积神经网络的轿车车型精细识别方法", 《河北科技大学学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110516524A (en) * | 2019-06-26 | 2019-11-29 | 东南大学 | Vehicle number of axle recognition methods based on Mask R-CNN in a kind of traffic scene |
| CN111325146A (en) * | 2020-02-20 | 2020-06-23 | 吉林省吉通信息技术有限公司 | Truck type and axle type identification method and system |
| CN111325146B (en) * | 2020-02-20 | 2021-06-04 | 吉林省吉通信息技术有限公司 | Truck type and axle type identification method and system |
| CN113537238A (en) * | 2021-07-05 | 2021-10-22 | 上海闪马智能科技有限公司 | Information processing method and image recognition device |
| CN113537238B (en) * | 2021-07-05 | 2022-08-05 | 上海闪马智能科技有限公司 | Information processing method and image recognition device |
| CN118334875A (en) * | 2024-05-16 | 2024-07-12 | 江苏新菲鹤信息科技有限公司 | Motor vehicle side window snapshot system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109815933A (en) | A kind of multiaxle trucks automatic identification apparatus for obtaining evidence, system and method | |
| CN107066953B (en) | A kind of vehicle cab recognition towards monitor video, tracking and antidote and device | |
| CN112700470B (en) | Target detection and track extraction method based on traffic video stream | |
| CN103902976B (en) | A kind of pedestrian detection method based on infrared image | |
| CN103971097B (en) | Vehicle license plate recognition method and system based on multiscale stroke models | |
| CN109887283A (en) | A kind of congestion in road prediction technique, system and device based on bayonet data | |
| CN202838681U (en) | Security intelligent checkpoint system | |
| CN110084165A (en) | The intelligent recognition and method for early warning of anomalous event under the open scene of power domain based on edge calculations | |
| CN105118305B (en) | Motor pool outlet vehicle management platform | |
| CN103714363A (en) | Motor vehicle exhaust smoke video identification system | |
| CN109191830A (en) | A kind of congestion in road detection method based on video image processing | |
| CN109033175A (en) | A kind of method and system to scheme to search vehicle | |
| CN109712406A (en) | A kind of pedestrian running red light and motor vehicle do not give precedence to pedestrian and monitor capturing system | |
| CN103903440B (en) | A kind of electric police grasp shoot method and device | |
| CN102819764A (en) | Method for counting pedestrian flow from multiple views under complex scene of traffic junction | |
| CN103164958B (en) | Method and system for vehicle monitoring | |
| CN102867417A (en) | Taxi anti-forgery system and taxi anti-forgery method | |
| TWI613108B (en) | Driving behavior analysis system and method for accident | |
| CN112883936A (en) | Method and system for detecting vehicle violation | |
| CN112084928A (en) | Road traffic accident detection method based on visual attention mechanism and ConvLSTM network | |
| CN106205135A (en) | A kind of detection method of vehicle behavior that turns around violating the regulations, Apparatus and system and a kind of ball machine | |
| CN110189425A (en) | Multilane free-flow vehicle detection method and system based on binocular vision | |
| CN112687103A (en) | Vehicle lane change detection method and system based on Internet of vehicles technology | |
| Ravish et al. | Intelligent traffic violation detection | |
| CN114926984B (en) | Real-time traffic conflict collection and road safety evaluation 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190528 |