CN105719486B - Highway zig zag vehicle-passing intelligent caution control system and method - Google Patents
Highway zig zag vehicle-passing intelligent caution control system and method Download PDFInfo
- Publication number
- CN105719486B CN105719486B CN201610292299.1A CN201610292299A CN105719486B CN 105719486 B CN105719486 B CN 105719486B CN 201610292299 A CN201610292299 A CN 201610292299A CN 105719486 B CN105719486 B CN 105719486B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- module
- warning
- pedestrian
- image
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/042—Detecting movement of traffic to be counted or controlled using inductive or magnetic detectors
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/09—Arrangements for giving variable traffic instructions
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09F—DISPLAYING; ADVERTISING; SIGNS; LABELS OR NAME-PLATES; SEALS
- G09F9/00—Indicating arrangements for variable information in which the information is built-up on a support by selection or combination of individual elements
- G09F9/30—Indicating arrangements for variable information in which the information is built-up on a support by selection or combination of individual elements in which the desired character or characters are formed by combining individual elements
- G09F9/33—Indicating arrangements for variable information in which the information is built-up on a support by selection or combination of individual elements in which the desired character or characters are formed by combining individual elements being semiconductor devices, e.g. diodes
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
Taken a sudden turn vehicle-passing intelligent caution control system, including image capture module, image processing and analyzing module, the number of axle for obtaining vehicle, piezoelectric sensing detection module, information of vehicles extraction module and the warning mechanism of weight and velocity information the invention discloses a kind of highway;Warning mechanism includes warning prompting module and module of warning;Module of warning is connected by control circuit with warning prompting module;According to the direction apart from road bend from the close-by examples to those far off, warning mechanism, image capture module and piezoelectric sensing detection module are arranged at intervals in order.The invention also discloses the warning control method using above-mentioned highway zig zag vehicle-passing intelligent caution control system.For highway zig zag road, type of vehicle, velocity information can accurately be informed by the present invention, realize classification alarm, warning.Especially for large-scale, heavy-load automobile, caution control system of the invention reminds opposed vehicle to take care by word, loudspeaker, can effectively lower pernicious traffic accident and occur.
Description
Technical field
The present invention relates to traffic alarm technical field, and in particular to a kind of highway zig zag vehicle-passing intelligent warning control
Method and its system.
Background technology
The possibility that the traffic safety of road means people or thing suffers a loss is acceptable;If this possibility surpasses
Acceptable level has been crossed, it is as dangerous.Road traffic system is as dynamic open system, and its safety is both by internal system
The restriction of factor, and being disturbed by its exterior environment, and it is closely related with the factor such as people, vehicle and road environment.In system
Any factor it is unreliable, uneven, unstable, may all lead to a conflict and contradiction, produce unsafe factor or dangerous shape
State.
With the development and popularization of automobile, road safety traffic accident also accordingly increases, and the lives and properties to people are brought
Loss.Particularly road conditions are constantly improved, speed also more and more higher, in some racing sweeps, especially one
The climb and fall racing sweep of a little sky ways, when two cars that sharp turn goes in the same direction carry out meeting often because driver regards
Angle is limited, often hear have at sharp turn due to avoid it is too late, cause the declaration of an accident of automobile crash, cause greatly safety and
Property loss.And it is existing for these racing sweeps typically using being handled in turning installation reflective mirror, it is but this
Mode darkness or rain wait weather under driver be difficult the enough information warning of acquisition, actual effect is very undesirable.In addition,
If in climb and fall racing sweep sharp turn installation reflective mirror if, because climb and fall has a gradient, therefore install
It is less desirable that reflective mirror is difficult to the driver alert signal to up-hill journey and descent run, its actual effect simultaneously.
Find there are some Patents documents to report by the domestic Searches of Patent Literature, mainly have following some:
1st, Publication No. CN202644425U, entitled " a kind of sharp highway turn vehicle is by prompting loudspeaker " are practical
New patent, belong to engineering goods field.The patent is related to a kind of sharp highway turn vehicle by prompting loudspeaker, utilizes loudspeaker
Auditory tone cues bend the opposing party there is vehicle to drive into bend, on the highway pavement on bend both sides, be respectively equipped with and produce gas
Gas source generator, and steam whistle loudspeaker are set respectively beside the highway on bend both sides, the source of the gas of steam whistle loudspeaker and highway pavement
Generating means is connected, and flowing gas sends tucket as caused by gas source generator.It is possible to prevente effectively from driver is led to
Bend is crossed, does not pay attention to seeing reflective mirror, does not also have the situation on ring loudspeaker prompting opposite, the mesh of mandatory prompting can be reached
's.
2nd, Publication No. CN201785709U, entitled " warning device for up/down slope sharp turn " utility model patent, the patent
It is related to a kind of warning device for up/down slope sharp turn.Belong to electromechanical control field, include and be arranged at climb and fall zig zag side upward slope
Detector and the detector being arranged on climb and fall zig zag side gug, climb and fall zig zag both sides on road
In control the warning sign being connected with the detector on opposite side track per also erectting to set respectively on the track of side.
3rd, Patent No. CN201020219521.3, a kind of entitled " Warning for vehicle for highway zig zag crossing
Device " utility model patent, the patent are related to a kind of vehicle warning device for highway zig zag crossing, belong to Electromechanical Control
Technical field.Including power electricity transition components, electric energy processing component and electronic display, wherein power electricity transition components are installed on public affairs
The porch at road zig zag crossing, electronic display are installed on the exit at highway zig zag crossing.Power electricity transition components will row
The kinetic energy for sailing vehicle is converted to electric energy;Electric energy processing component, which is converted to alternating current caused by power electricity transition components, can continue a little
The direct current of bright electronic display;Electronic display includes some LED lamps in parallel, forms in the word to serve as warning
Hold.When being crossed if vehicle, power electricity transition components electricity can light electronic display, so as to warn opposite to drive into car
Driver, with the generation to cut down traffic accidents.
4th, Patent No. CN200420073346.6, entitled " intersection acousto-optic warning fills zig zag on-highway motor vehicle safely
Put " utility model patent, the patent is related to acousto-optic warning device during for sharp highway turn automobile meeting.Provided with pickup
Device/microphone, sound control circuit, acousto-optic warning circuit, the input of the output termination sound control circuit of sound pick-up/microphone, sound control circuit
Output end connect acousto-optic warning circuit through connecting wire.In use, sound pick-up/microphone is arranged in front of sharp highway turn about
At 50m, and set into safety line etc. and mark, acousto-optic warning circuit can be appropriate located at sharp highway turn opposite side according to road conditions
Position, when vehicle enters safety line or/and sees mark, driver's ring loudspeaker, acousto-optic warning circuit is simultaneously emitted by warning acousto-optic
Signal, driver is reminded to enter sharp turn front reduction gear, giving precedence to.
Although above-mentioned patent proposes technical scheme during sharp highway turn automobile meeting, but many drivers at present
Still there is certain idea of leaving things to chance in member, it not is to take much count of that acousto-optic, which is reminded,.Sharp highway turn, full-sized car and car
It is very heavy Deng the improper caused serious accident of meeting.If can further lift technique scheme, pass through alarming device
Accurate to inform type of vehicle, velocity information, classification alarm, warning, such effect can more targetedly, and effect can be more preferable.Enter one
Step, above-mentioned patent more concern in the road vehicles intersection safety, for the pedestrian participant in road not it is considered that
Enough warnings are not given to pedestrian.
The content of the invention
It is an object of the invention to take a sudden turn vehicle pass-through just for vehicle detection and the letter of detection for current highway
A kind of present situations such as breath amount deficiency, highway zig zag vehicle-passing intelligent caution control system of proposition, can detect vehicle and row
The details of people, avoid sharp turn that traffic accident occurs by different information warnings.
To achieve the above object, highway of the invention zig zag vehicle-passing intelligent caution control system includes IMAQ
Module, image processing and analyzing module, the number of axle for obtaining vehicle, piezoelectric sensing detection module, the car of weight and velocity information
Information extraction modules and warning mechanism;Warning mechanism includes warning prompting module and module of warning;The module of warning passes through
Control circuit is connected with the warning prompting module;
Piezoelectric sensing detection module includes the ground induction coil being disposed adjacent along the length direction of road and piezoelectric membrane senses
Device, for obtaining the number of axle, weight and velocity information of vehicle;
Image capture module includes multiple pedestrians shooting of the vehicle camera and alignment pedestrian area of alignment vehicle region
Head, vehicle camera are used for collection vehicle image, and pedestrian's camera is used to gather pedestrian image;
Image processing and analyzing module includes background separation module and convolutional neural networks;Background separation module is used in vehicle
Vehicle's contour shape is extracted in the image of camera collection;Convolutional neural networks are used to carry out vehicle according to vehicle's contour shape
Classify and calculate the type of corresponding vehicle and the positional information of vehicle;Image processing and analyzing module is provided with two sets, respectively car
Image processing and analyzing module and pedestrian image Treatment Analysis module;
Information of vehicles extraction module includes radial base neural net grader, for piezoelectric sensing detection module and image
The data of Treatment Analysis module transmission are merged and the type of vehicle of image processing and analyzing module transmission are modified;
It is integrated circuit or PLC or single-chip microcomputer to warn prompting module, for receiving the information of information of vehicles extraction module
And module of warning is controlled to warn;
Module of warning includes LED display and loudspeaker, and LED display is used to show warning image and caveat, raised one's voice
Device is used to report warning sound;
Image capture module and piezoelectric sensing detection module are connected with image processing and analyzing module respectively, image procossing point
Analysis module is connected with information of vehicles extraction module;Information of vehicles extraction module is connected with warning prompting module, and warning is reminded
Module is connected with module of warning;
Wide-angle lens is provided with road sharp turn;
The road of the warning mechanism, described image acquisition module and the piezoelectric sensing detection module in turning both sides
Two sets are provided with, and according to the direction apart from road bend from the close-by examples to those far off, are arranged at intervals the warning mechanism, described in order
Image capture module and the piezoelectric sensing detection module.
The vehicle camera is provided with two or more relative to each track of vehicle region.
Two camera installation directions in pedestrian's camera mounted in pairs and each pair pedestrian's camera are opposite.
Length direction of the piezoelectric film sensor along road is interval with two.
It is respectively positioned in the middle part of the middle part of the ground induction coil and piezoelectric film sensor at the divisional line in two tracks.
The warning mechanism includes warning prompting module and module of warning;Warning prompting module and module control of warning connect
Connect;Module of warning includes LED display and loudspeaker.
It is a kind of using above-mentioned highway zig zag vehicle-passing intelligent caution control system the present invention also aims to provide
Warning control method.
To realize the purpose, the invention provides one kind to warn control method, and the highway zig zag car is driven into vehicle
Behind the setting area of passing intelligent caution control system, carry out according to the following steps successively:
(1) number of axle, weight and the velocity information of vehicle are obtained by piezoelectric sensing detection module;
(2) vehicle image of current system institute viewed area is gathered by the vehicle camera of described image acquisition module, led to
Cross the pedestrian image of pedestrian's camera collection current system institute viewed area;
(3) the preliminary identification of type of vehicle and the extraction of pedestrian information are completed by described image Treatment Analysis module:
The vehicle region picture gathered for the vehicle camera of image capture module, background separation module pass through background point
The foreground area for including vehicle sections is obtained from operation, further according to Edge Gradient Feature vehicle's contour shape, and utilizes the volume
Product neutral net is classified to vehicle, and calculates the type of corresponding vehicle and its positional information of vehicle;
By setting multiple pedestrian's cameras to gather pedestrian area picture, prevent blind area being present;The figure of each pedestrian area
Piece all obtains the foreground area for including pedestrian part using background separation module by background separation operation, further according to edge feature
Pedestrian contour shape is extracted, and the pedestrian in road is picked out using the convolutional neural networks trained;
(4) fusion that vehicle multi-source information is completed by the information of vehicles extraction module is extracted:
By the vehicle number of axle acquired in piezoelectric sensing detection module, weight, velocity information and image processing and analyzing module institute
The type of vehicle of acquisition carries out data fusion, using the radial base neural net grader that off-line training is crossed to the(3)In step
The type of vehicle of identification is modified;
(5) vehicle acquired in aforementioned vehicle information extraction modules, pedestrian information are sent to described warning and remind mould
Block, the warning prompting module are acted accordingly according to following situations:
A. certain when track on have pedestrian in the case of, control warn module send signal remind to being kept away to track vehicle deceleration
Allow;
B. certain when vehicle road occupying traveling and following two condition in either condition meet in the case of, control is warned
Module sends signal and reminds Current vehicle by regulation slow down and to drive to correct track, while reminds to track vehicle
Parking avoids;
1. item condition is that the travel speed of the vehicle is more than or equal to 30 kilometers per hour to the of two condition;Described two
2. item condition is that the vehicle is heavily loaded oversize vehicle to the of part;
C. certain when vehicle be the oversize vehicle of heavy duty in the case of, control module of warning sends signal and reminded to track
Vehicle deceleration and avoidance of going slowly;
D. certain when vehicle be normal speed it is current and according to lane in the case of, control module of warning sends signal
Remind to track Slow Down;
E. to track without road participant in the case of, control module of warning sends signal bidirectional reminding and driven with caution.
The present invention advantage and its have the beneficial effect that:For highway zig zag road, pass through highway racing of the invention
Curved vehicle-passing intelligent caution control system and method, type of vehicle, velocity information are accurately informed, realize classification alarm, warning.
Especially for large-scale, heavy-load automobile, caution control system of the invention passes through(It is red)Word, loudspeaker remind opposed vehicle note
Meaning safety, it can effectively lower pernicious traffic accident and occur.Further, this patent is additionally contemplates that the peace of the pedestrian participant in road
Entirely, enough current security warnings are given to pedestrian by loudspeaker, avoid the road participant of road sharp turn because not shifting to an earlier date
Know turning opposite side road vehicle and pedestrian's situation and traffic accident occurs.
Brief description of the drawings
Fig. 1 is the principle schematic of the warning control method of the present invention;
Fig. 2 is that warning mechanism in the present invention and image capture module, sensing module are illustrated in the arrangement of road bend
Figure;
Reference is meant that in Fig. 2:101st, ground induction coil;102nd, piezoelectric film sensor, 103, IMAQ mould
Block;104th, warn module;301st, wide-angle lens.
Fig. 3 is the scheme of installation for the pedestrian's camera being arranged in pairs;
Reference is meant that in Fig. 3:111st, pedestrian's camera;112nd, angle adjustable screw;213rd, detachably focus
Camera lens.Fig. 4 is the flow chart that vehicle image Treatment Analysis module carries out type of vehicle identification;
Fig. 5 is the convolutional neural networks structural representation that vehicle image Treatment Analysis module carries out type of vehicle and pedestrian's identification
Figure;
Fig. 6 is the flow chart that pedestrian image Treatment Analysis module carries out pedestrian's identification;
Fig. 7 is the pedestrian area identification template schematic diagram of pedestrian image Treatment Analysis module;
Fig. 8 is the flow chart that information of vehicles extraction module carries out type of vehicle multi-source fusion detection.
Embodiment
As shown in Figures 1 to 8, the invention discloses a kind of highway zig zag vehicle-passing intelligent caution control system, bag
The piezoelectricity for including image capture module 103, image processing and analyzing module, the number of axle for obtaining vehicle, weight and velocity information passes
Feel detection module, information of vehicles extraction module and warning mechanism;Warning mechanism includes warning prompting module and module 104 of warning;
The core of system preferably uses distributed embedded processing systems, each module can complete independently oneself task, mutually it
Between communicated according to the flow shown in Fig. 1, complete all Data Detections, logic judgment, graphical analysis and warning control.
Piezoelectric sensing detection module includes the ground induction coil 101 being disposed adjacent along the length direction of road and piezoelectric membrane passes
Sensor 102, for obtaining the number of axle, weight and velocity information of vehicle;
Multiple pedestrians of vehicle camera and alignment pedestrian area that image capture module 103 includes alignment vehicle region take the photograph
As first 111, vehicle camera is used for collection vehicle image, and pedestrian's camera 111 is used to gather pedestrian image;
In order to lift the coverage of pedestrian detection and its precision, patent of the present invention is carried out to pedestrian's camera 111
Adjustment, its structural representation are as shown in Figure 3.Can be to the pedestrian for gathering pedestrian area image by angle adjustable screw 112
The shooting angle of camera 111 is finely adjusted, and reduces field erected difficulty, and the effective image sampling of maximum magnitude can be achieved.Enter
One step, can be to the image sampling of pedestrian key area by detachable tight shot 213, and the number of mounting rod is laid at the scene of reduction
Amount.In addition, in order to reduce the blind area of pedestrian's detecting, the preferably 180 ° of positive and negative installations of pedestrian's camera 111.
Image processing and analyzing module includes background separation module(Background separation module is this area routine techniques, and its is specific
Composition is no longer described in detail)With the convolutional neural networks trained;Background separation module is used in the image of vehicle camera collection
Extract vehicle's contour shape;Convolutional neural networks are used to carry out preliminary classification to vehicle according to vehicle's contour shape and calculated
The type of corresponding vehicle and the positional information of vehicle;Image processing and analyzing module is provided with two sets, and respectively vehicle image is handled
Analysis module and pedestrian image Treatment Analysis module, two sets of image processing and analyzing modules are combined into one use in Fig. 2 for simplicity
One square frame represents.
Information of vehicles extraction module includes radial base neural net(SVM)Grader, for piezoelectric sensing detection module
Merged with the data of image processing and analyzing module transmission and the type of vehicle of image processing and analyzing module transmission is repaiied
Just;
It is integrated circuit or PLC or single-chip microcomputer to warn prompting module, for receiving the information of information of vehicles extraction module
And accordingly module 104 of warning is controlled to warn;Either single-chip microcomputer or the specified function of PLC programming realizations are integrated design circuit
The conventional technical ability of those skilled in the art, warns the concrete structure of prompting module to be no longer described in detail.Warn prompting module also normal for this area
Rule technology, its concrete composition are no longer described in detail.
Module of warning 104 includes LED display and loudspeaker, and LED display is used to show warning image and caveat, raised
Sound device is used to report warning sound;
Module annexation:Image capture module 103 and piezoelectric sensing detection module respectively with image processing and analyzing module
It is connected, image processing and analyzing module is connected with information of vehicles extraction module;Information of vehicles extraction module reminds mould with warning
Block is connected;The module 104 of warning is connected by control circuit with the warning prompting module;
Wide-angle lens 301 is provided with road sharp turn(That is convex mirror, so as to make passing driver or pedestrian see in advance
The vehicles or pedestrians at turning rear);
Warning mechanism and image capture module 103, the ground location relation of sensing module:The warning mechanism, the figure
As acquisition module 103 and the piezoelectric sensing detection module are provided with two sets on the road of turning both sides, and according to apart from road
The direction of road turning from the close-by examples to those far off, the warning mechanism, described image acquisition module 103 and the pressure are arranged at intervals in order
Fax sense detection module.
The vehicle camera is provided with two or more relative to each track of vehicle region, blind so as to prevent from existing
Area.
Two camera installation direction phases in the mounted in pairs of pedestrian's camera 111 and each pair pedestrian camera 111
Instead(I.e. two cameras are in the positive and negative installation in 180 degree angle), prevent blind area being present.
Length direction of the piezoelectric film sensor 102 along road is interval with two.By setting two piezoelectricity thin
Film sensors 102, the time difference that calculating vehicle passes through can accurately obtain the velocity information of vehicle.
The middle part of the ground induction coil 101 and the middle part of piezoelectric film sensor 102 are respectively positioned on the track point in two tracks
At line.So only need the information of vehicles disposed on i.e. detectable two tracks of a set of piezoelectric sensing detection module.Certainly, if
Vehicle is not in the center traveling in two tracks, and ground induction coil is pressed in due to only having vehicle to close on the tire of divisional line side
101 and piezoelectric film sensor 102 on, therefore the vehicle weight measured needs to put and be twice.
The warning mechanism includes warning prompting module and module 104 of warning;Warning prompting module is controlled with module 104 of warning
System connection;Module of warning 104 includes LED display and loudspeaker;
The invention also discloses the warning controlling party using above-mentioned highway zig zag vehicle-passing intelligent caution control system
Method, after vehicle drives into the setting area of the highway zig zag vehicle-passing intelligent caution control system, successively by following step
It is rapid to carry out:
(1) number of axle, weight and the velocity information of vehicle are obtained by piezoelectric sensing detection module;
Piezoelectric sensing detection module is responsible for obtaining the number of axle, weight, the velocity information of vehicle.In order to reduce cost, for two
The highway situation in track, ground induction coil 101 and piezoelectric film sensor 102 are deployed in the middle position in two tracks, so such as
Fruit vehicle is travelled in two lane centers, then the vehicle weight of measurement needs to put to be twice.By setting two piezoelectricity
Thin film sensor 102, the velocity information of vehicle can accurately be obtained by calculating the time difference passed through.The installation and deployment of its module are as schemed
Shown in 2.
Because temperature and speed have a great influence to the detection accuracy of piezoelectric film sensor 102, thus need to establish one
Individual fuzzy inference system, appropriate amendment is carried out to the vehicle weight after detection, and modification method is the weight and reasoning obtained
Multiplication factor carries out product calculation, and its dependent blur inference rule is as shown in table 1:
Inference rule table is pasted in the vehicle detection weight amendment of table 1
(2) vehicle image of current system institute viewed area is gathered by the vehicle camera of described image acquisition module 103,
The pedestrian image of current system institute viewed area is gathered by pedestrian's camera 111;
(3) the preliminary identification of type of vehicle and the extraction of pedestrian information are completed by described image Treatment Analysis module.
The vehicle region picture gathered for the vehicle camera of image capture module 103, background separation module pass through the back of the body
Scape lock out operation(Background separation operation is this area routine techniques)The foreground area for including vehicle sections is obtained, further according to edge
Feature extraction vehicle's contour shape, and preliminary classification is carried out to vehicle using the convolutional neural networks trained, and count
Calculate the type of corresponding vehicle and its positional information of single unit vehicle;
By setting multiple pedestrian's cameras 111 to gather pedestrian area picture, prevent blind area being present;Each pedestrian area
Picture all obtains the foreground area for including pedestrian part using background separation module by background separation operation, special further according to edge
Sign extraction pedestrian contour shape, and utilize the convolutional neural networks trained(Effectively)Pick out the pedestrian in road.
Specifically, as shown in figure 4, vehicle image Treatment Analysis module is carried out according to the following steps at work:
1. the foreground pixel part of vehicle is obtained by present image and background image subtraction.Background image is no car
Picture captured by camera when current, the background picture is to maintain changeless in certain period of time(Positive reason
Under condition, the road conditions of road sharp turn will not often change).The system can accurately detect car by ground induction coil 101
Vehicle pass-through situation on road, thus accurate reference signal input can be provided for this module, upgrade in time background picture.Pass through
Two field pictures are subtracted each other, the pixel set differed.For this one part of pixel set, obtain corresponding in vehicle region picture
Pixel portion and carry out gray processing, that is, obtain foreground part.
2. extract edge feature.For the foreground part obtained, vehicle image Treatment Analysis module needs further
Handled, obtain the edge feature of vehicle.Edge feature is completed by the hybrid operation of Fourier's operator+Canny operators, so as to
The bianry image of a secondary black and white is obtained, represents the principal outline information of vehicle.
3. convolutional neural networks are classified.The system is by off-line training convolutional neural networks, and sample is different distances, no
The vehicle two-value profile picture of same type, exports the type for vehicle, so as to realize the preliminary judgement to type of vehicle, its convolution
Neural network structure is as shown in figure 5, the flow chart of data processing of convolutional neural networks is as follows:
A. the input layer of convolutional neural networks is then whole image, and as shown in Figure 5, picture size can be set to 29*29
Pixel.Image is deployed by row, forms 841 nodes;And the node of first layer is forward without any tie line.
B. three feature extraction figures are produced after convolution, then every group of four pixels are summed again in feature extraction figure,
Weighted value, biasing are put, and three Feature Mapping figures are obtained by a Sigmoid function.
C. convolution is further carried out again to caused three Feature Mapping figures, producing three quadratic characters after convolution carries
Figure is taken, then every group in Further Feature Extraction figure of four pixels are summed again, weighted value, biasing is put, again by one
Individual Sigmoid functions obtain three quadratic character mapping graphs.
D. above-mentioned caused quadratic character mapping graph is rasterized, and connect into a vector be input to it is traditional
Neutral net, obtain output result.
In the system, convolutional neural networks are inherently a kind of mapping for being input to output, and it can learn largely
Mapping relations between input and output, without the accurate mathematic(al) representation between any input and output, as long as with
Known pattern is trained to convolutional network, and network just has the mapping ability between inputoutput pair.Convolutional network performs
Be to have tutor's training, so its sample set be by shaped like:(Vehicle's contour two-value input picture vector, type of vehicle export to
Amount)Vector to composition.All these vectors are right, should all be derived from the actual " RUN " for the system that network will simulate
As a result, they are to gather to come from actual motion system.Before training is started, all power all with some it is different it is small at random
Number is initialized.
In highway zig zag section, if the vehicle of current driving occupies to being danger close to track, its consequence, especially
It is the situation that oversize vehicle, heavy-duty vehicle intersect with dilly.Therefore, image capture module 103 can also complete Current vehicle
The preliminary judgement in place track, its basis for estimation use fuzzy inference system.By taking two tracks as an example, its dependent blur inference rule
As shown in table 2:
Track fuzzy inference rule table shared by the vehicle of table 2
Pedestrian image Treatment Analysis module operation principle as shown in fig. 6, pedestrian image Treatment Analysis module to all panoramas,
The pedestrian image of key area is all analyzed and processed, collected one by one.Obtained by background separation operation comprising pedestrian part
Foreground area, pedestrian is carried out further according to Edge Gradient Feature pedestrian contour shape, and using the convolutional neural networks trained
Identification, and calculate the positional information in corresponding pedestrian place track.Mainly include the following steps that:
1. the foreground pixel part of pedestrian is obtained by present image and background image subtraction.Background image is no row
Picture captured by camera when people is current.The background picture be to maintain in certain period of time it is fixed, in order to avoid car
Interference, the sampling instant of the background is identical with vehicle context sampling instant.
2. extract edge feature.For the foreground part obtained, pedestrian image Treatment Analysis module needs further
Handled, obtain the edge feature of pedestrian.Edge feature is completed by the hybrid operation of Fourier's operator+Canny operators.So as to
The bianry image of a secondary black and white is obtained, represents the principal outline information of pedestrian.
3. convolutional neural networks are classified.By off-line training convolutional neural networks, sample is different distances, different statures
Pedestrian's black and white two-value profile picture, exports the probability for pedestrian, so as to realize that the identification to pedestrian judges that it uses convolutional Neural
Network structure is identical with type of vehicle preliminary classification, also shown in FIG. 5.
In addition, in order to recognize the region of pedestrian place road, traffic safety is lifted, its pedestrian area identification Prototype drawing is as schemed
Shown in 7.By pedestrian contour where judging position in a template, you can separate whether in safety zone.Pass through in Fig. 7
Different colours identify safe class, and the configuring condition of the security template in Fig. 7 can be carried out according to the camera installation site at scene
It is specific to set.Red area is high risk zone in Fig. 7, and yellow area is the region with danger, and green area is safety zone.
(4) fusion that vehicle multi-source information is completed by the information of vehicles extraction module is extracted:
By the vehicle number of axle acquired in piezoelectric sensing detection module, weight, velocity information and image processing and analyzing module institute
The type of vehicle of acquisition carries out data fusion, the radial base neural net crossed using off-line training(SVM)Grader is to(3)Step
The type of vehicle of identification is modified in rapid, and extracts the information in vehicle place track, is provided reliably for warning prompting module
Basic data.The operation principle of information of vehicles extraction module is as shown in Figure 8.
(5) vehicle acquired in information of vehicles extraction module in (4) step, pedestrian information are sent to described police
Show prompting module, the warning prompting module is acted accordingly according to following situations:
A. certain when track on have pedestrian in the case of, control warn module 104 send signal remind to subtracting to track vehicle
Speed avoids;
B. certain when vehicle road occupying traveling and following two condition in either condition meet in the case of, control is warned
Module 104 sends signal and reminds Current vehicle by regulation slow down and to drive to correct track, while reminds to track
Vehicle parking avoids;
1. item condition is that the travel speed of the vehicle is more than or equal to 30 kilometers per hour to the of two condition;Described two
2. item condition is that the vehicle is heavily loaded oversize vehicle to the of part;
C. certain when vehicle be heavy duty oversize vehicle in the case of, control warn module 104 send signal remind to
Track vehicle slows down with caution and avoidance of going slowly;
D. certain when vehicle be normal speed it is current and according to lane in the case of, control module 104 of warning is sent
Signal is reminded to travel to being given precedence to track Slow Down;
E. to track without road participant(I.e. not only without vehicle but also without pedestrian)In the case of, control module 104 of warning to send out
Go out signal bidirectional reminding to drive with caution;
Above-mentioned(5)Traffic rules in are all suitable for passing in and out two-way vehicle.
The module 104 of warning shows related caveat by LED display, and red, yellow can be used according to grade(It is red
It is more urgent that grade compares yellowness ratings)It has been shown that, to prompt the associated vehicle human pilot of meeting.Further, carry for convenience
Wake up road on pedestrian, it is described warn module 104 also warning prompting module control under can be informed by audible alarm to
Road carrys out car situation, as according to the car category travelled on present road, the sound of blowing a whistle of automobile corresponding to automatic imitation;Again
Such as when the situation is critical, direct voice reminder pedestrian pays attention to avoiding.
General principle, the main features and advantages of the present invention have been shown and described above.The technical staff of the industry should
Understand, the present invention is not limited to the above embodiments, the original for simply illustrating the present invention described in above-described embodiment and specification
Reason, without departing from the spirit and scope of the present invention, various changes and modifications of the present invention are possible, and these are without departing from this hair
Bright substantive changes and improvements all fall within the protetion scope of the claimed invention.The claimed scope of the invention is by appended power
Sharp claim and its equivalent thereof.
Claims (6)
- The vehicle-passing intelligent caution control system 1. highway takes a sudden turn, it is characterised in that:Including image capture module, image procossing Analysis module, the number of axle for obtaining vehicle, piezoelectric sensing detection module, the information of vehicles extraction module of weight and velocity information And warning mechanism;Warning mechanism includes warning prompting module and module of warning;It is described warn module by control circuit with it is described Warning prompting module is connected;Piezoelectric sensing detection module includes the ground induction coil and piezoelectric film sensor being disposed adjacent along the length direction of road, uses In the number of axle, weight and the velocity information that obtain vehicle;Image capture module includes the vehicle camera of alignment vehicle region and multiple pedestrian's cameras of alignment pedestrian area, car Camera is used for collection vehicle image, and pedestrian's camera is used to gather pedestrian image;Image processing and analyzing module includes background separation module and convolutional neural networks;Image processing and analyzing module is provided with two Set, respectively vehicle image Treatment Analysis module and pedestrian image Treatment Analysis module;The back of the body of vehicle image Treatment Analysis module Scape separation module is used to extract vehicle's contour shape in the image of vehicle camera collection;Vehicle image Treatment Analysis module Convolutional neural networks are used for type and the position of vehicle for being classified and being calculated corresponding vehicle to vehicle according to vehicle's contour shape Confidence ceases;Information of vehicles extraction module includes radial base neural net grader, for piezoelectric sensing detection module and image procossing The data of analysis module transmission are merged and the type of vehicle of image processing and analyzing module transmission are modified;It is integrated circuit or PLC or single-chip microcomputer to warn prompting module, for receiving information and the control of information of vehicles extraction module Module of warning is made to warn;Module of warning includes LED display and loudspeaker, and LED display is used to show warning image and caveat, and loudspeaker is used Sound is warned in reporting;The vehicle camera and piezoelectric sensing detection module of image capture module are connected with vehicle image Treatment Analysis module respectively Connect, image processing and analyzing module is connected with information of vehicles extraction module;Information of vehicles extraction module is with warning prompting module phase Connection, warning prompting module are connected with module of warning;Wide-angle lens is provided with road sharp turn;The warning mechanism, described image acquisition module and the piezoelectric sensing detection module are set on the road of turning both sides There are two sets, and according to the direction apart from road bend from the close-by examples to those far off, be arranged at intervals the warning mechanism, described image in order Acquisition module and the piezoelectric sensing detection module.
- The vehicle-passing intelligent caution control system 2. highway according to claim 1 takes a sudden turn, it is characterised in that:The car Camera is provided with two or more relative to each track of vehicle region.
- The vehicle-passing intelligent caution control system 3. highway according to claim 1 takes a sudden turn, it is characterised in that:The row Two camera installation directions in people's camera mounted in pairs and each pair pedestrian's camera are opposite.
- The vehicle-passing intelligent caution control system 4. highway according to any one of claim 1 to 3 takes a sudden turn, its feature It is:Length direction of the piezoelectric film sensor along road is interval with two.
- The vehicle-passing intelligent caution control system 5. highway according to any one of claim 1 to 3 takes a sudden turn, its feature It is:It is respectively positioned in the middle part of the middle part of the ground induction coil and piezoelectric film sensor at the divisional line in two tracks.
- 6. the warning control method of highway zig zag vehicle-passing intelligent caution control system described in usage right requirement 5, its Be characterised by after vehicle drives into the setting area of highway zig zag vehicle-passing intelligent caution control system, press successively with Lower step is carried out:(1) number of axle, weight and the velocity information of vehicle are obtained by piezoelectric sensing detection module;(2) vehicle image of current system institute viewed area is gathered by the vehicle camera of described image acquisition module, passes through row The pedestrian image of people's camera collection current system institute viewed area;(3) the preliminary identification of type of vehicle and the extraction of pedestrian information are completed by described image Treatment Analysis module:The vehicle region picture gathered for the vehicle camera of image capture module, background separation module are grasped by background separation Make to obtain the foreground area for including vehicle sections, further according to Edge Gradient Feature vehicle's contour shape, and utilize convolution god Vehicle is classified through network, and calculates the type of corresponding vehicle and its positional information of vehicle;By setting multiple pedestrian's cameras to gather pedestrian area picture, prevent blind area being present;The picture of each pedestrian area The foreground area for including pedestrian part is obtained by background separation operation using background separation module, further according to Edge Gradient Feature Pedestrian contour shape, and pick out using the convolutional neural networks trained the pedestrian in road;(4) fusion that vehicle multi-source information is completed by the information of vehicles extraction module is extracted:By acquired in the vehicle number of axle acquired in piezoelectric sensing detection module, weight, velocity information and image processing and analyzing module Type of vehicle carry out data fusion, using the radial base neural net grader that off-line training is crossed to the(3)Recognized in step Type of vehicle be modified;(5) vehicle acquired in aforementioned vehicle information extraction modules, pedestrian information are sent to described warning prompting module, The warning prompting module is acted accordingly according to following situations:A. certain when track on have pedestrian in the case of, control warn module send signal remind to track vehicle deceleration avoid;B. certain when vehicle road occupying traveling and following two condition in either condition meet in the case of, control is warned module Sending signal reminds Current vehicle by regulation slow down and to drive to correct track, while reminds to track vehicle parking Avoid;1. item condition is that the travel speed of the vehicle is more than or equal to 30 kilometers per hour to the of two condition;Two condition 2. item condition be that the vehicle is heavily loaded oversize vehicle;C. certain when vehicle be the oversize vehicle of heavy duty in the case of, control module of warning sends signal and reminded to track car Slow down and avoidance of going slowly;D. certain when vehicle be normal speed it is current and according to lane in the case of, control module of warning sends signal and carried Wake up to track Slow Down;E. to track without road participant in the case of, control module of warning sends signal bidirectional reminding and driven with caution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610292299.1A CN105719486B (en) | 2016-05-05 | 2016-05-05 | Highway zig zag vehicle-passing intelligent caution control system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610292299.1A CN105719486B (en) | 2016-05-05 | 2016-05-05 | Highway zig zag vehicle-passing intelligent caution control system and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105719486A CN105719486A (en) | 2016-06-29 |
CN105719486B true CN105719486B (en) | 2017-12-01 |
Family
ID=56161973
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610292299.1A Active CN105719486B (en) | 2016-05-05 | 2016-05-05 | Highway zig zag vehicle-passing intelligent caution control system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105719486B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106205163B (en) * | 2016-07-20 | 2019-06-07 | 长安大学 | Mountain-area road-curve sight blind area meeting early warning system based on panoramic shooting technology |
CN106571046B (en) * | 2016-11-11 | 2021-07-16 | 上海市政工程设计研究总院(集团)有限公司 | Vehicle-road cooperative driving assisting method based on road surface grid system |
CN106530835A (en) * | 2017-01-09 | 2017-03-22 | 山东轻工职业学院 | Road-sharp-corner-crossing smooth-traffic intelligence dispersion apparatus and method thereof |
CN106875751A (en) * | 2017-03-27 | 2017-06-20 | 华南理工大学 | Meeting prompt system and method at a kind of bicycle road country road sharp turn section |
CN107292997A (en) * | 2017-06-21 | 2017-10-24 | 成都锐奕信息技术有限公司 | Safety detecting system based on high capacity waggon |
CN107230365A (en) * | 2017-06-21 | 2017-10-03 | 成都锐奕信息技术有限公司 | A kind of detection method for being used to monitor the lorry of traveling |
CN107610527A (en) * | 2017-09-21 | 2018-01-19 | 临沂市华夏高科信息有限公司 | One kind zig zag vehicle avoids caution system |
CN109410589B (en) * | 2018-08-30 | 2021-07-27 | 广州大学 | Vehicle violation detection system based on piezoelectric composite board |
CN109147326A (en) * | 2018-09-06 | 2019-01-04 | 北京理工大学 | A kind of Campus transport safety warning system |
CN111815974A (en) * | 2019-04-12 | 2020-10-23 | 天津大学 | Curve speed measuring and calculating indicator based on piezoelectric induction technology |
CN112458945B (en) * | 2020-12-11 | 2021-09-24 | 郑州轻工业大学 | Safety control method for vehicle passing through sharp curve in mountain area |
CN113022443A (en) * | 2021-03-10 | 2021-06-25 | 白田梅荣 | Traffic safety reminding device for bus |
CN114438919A (en) * | 2022-03-17 | 2022-05-06 | 湘潭大学 | Device and system applied to traffic safety guarantee of sharp-bend and steep-slope |
CN114648892A (en) * | 2022-03-25 | 2022-06-21 | 北京中交兴路信息科技有限公司 | Vehicle driving safety early warning method and device, vehicle, equipment and medium |
CN117012020A (en) * | 2023-01-10 | 2023-11-07 | 深圳市睿星旗航科技有限公司 | Individualized warning system of wisdom zebra stripes pedestrian |
CN116876381B (en) * | 2023-07-13 | 2024-03-29 | 云南八瑞科技有限公司 | Car yard streaming media reflector based on artificial intelligence technology |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005025528A (en) * | 2003-07-02 | 2005-01-27 | Sony Corp | Vehicle detection method, vehicle detection system, vehicle detection device, information transmission device, information reception device, navigation system, information notification device, automobile and vehicle detecting program |
JP2006236029A (en) * | 2005-02-25 | 2006-09-07 | Alpine Electronics Inc | Warning device and warning method |
KR20100053318A (en) * | 2008-11-12 | 2010-05-20 | (주) 서돌 전자통신 | A sensing system of a lane offense cars using the corner data and sensing method using the same |
CN201489668U (en) * | 2009-09-03 | 2010-05-26 | 中山市海辰科技发展有限公司 | Pedestrian safety pre-warning system |
CN101763720A (en) * | 2009-08-12 | 2010-06-30 | 江苏大学 | Slope turn traffic command system based on wireless sensor network |
CN202887476U (en) * | 2012-09-06 | 2013-04-17 | 上海市政工程设计研究总院(集团)有限公司 | Early warning system for preventing right-turning large-scale vehicle traffic accident |
CN203038485U (en) * | 2012-11-23 | 2013-07-03 | 长安大学 | Device for monitoring vehicle at sharp curves of mountain roads |
CN203300053U (en) * | 2013-06-25 | 2013-11-20 | 天津云视科技发展有限公司 | Safety warning device for vehicle to pass pavement in right-hand bend mode |
CN204808588U (en) * | 2015-07-29 | 2015-11-25 | 王云 | Vehicle type identification system |
-
2016
- 2016-05-05 CN CN201610292299.1A patent/CN105719486B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005025528A (en) * | 2003-07-02 | 2005-01-27 | Sony Corp | Vehicle detection method, vehicle detection system, vehicle detection device, information transmission device, information reception device, navigation system, information notification device, automobile and vehicle detecting program |
JP2006236029A (en) * | 2005-02-25 | 2006-09-07 | Alpine Electronics Inc | Warning device and warning method |
KR20100053318A (en) * | 2008-11-12 | 2010-05-20 | (주) 서돌 전자통신 | A sensing system of a lane offense cars using the corner data and sensing method using the same |
CN101763720A (en) * | 2009-08-12 | 2010-06-30 | 江苏大学 | Slope turn traffic command system based on wireless sensor network |
CN201489668U (en) * | 2009-09-03 | 2010-05-26 | 中山市海辰科技发展有限公司 | Pedestrian safety pre-warning system |
CN202887476U (en) * | 2012-09-06 | 2013-04-17 | 上海市政工程设计研究总院(集团)有限公司 | Early warning system for preventing right-turning large-scale vehicle traffic accident |
CN203038485U (en) * | 2012-11-23 | 2013-07-03 | 长安大学 | Device for monitoring vehicle at sharp curves of mountain roads |
CN203300053U (en) * | 2013-06-25 | 2013-11-20 | 天津云视科技发展有限公司 | Safety warning device for vehicle to pass pavement in right-hand bend mode |
CN204808588U (en) * | 2015-07-29 | 2015-11-25 | 王云 | Vehicle type identification system |
Non-Patent Citations (2)
Title |
---|
一种基于单片机的交通警示装置的设计;胡勇 等;《仪表技术》;20101231(第12期);正文第9-11,15页 * |
盘山公路转弯处车辆监测***;贺桂芳,张佰力;《仪表技术》;20111231(第5期);正文第37-39页 * |
Also Published As
Publication number | Publication date |
---|---|
CN105719486A (en) | 2016-06-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105719486B (en) | Highway zig zag vehicle-passing intelligent caution control system and method | |
CN106251701B (en) | Vehicle rearview monitor and alarm system and method based on rotation zoom multi-cam | |
CN105015548B (en) | A kind of longitudinal collision avoidance is reminded and follows integrated system and method automatically | |
KR102133913B1 (en) | Digital display board system using object recognition based on deep learning for providing customized information of object and method for controlling thereof | |
US11912286B2 (en) | Driving risk identification model calibration method and system | |
CN101763720B (en) | Slope turn traffic command system based on wireless sensor network | |
CN108859961A (en) | Circular camera system for automatic Pilot | |
CN107953830A (en) | A kind of motorbus blind area detection alarm system and its detection alarm method | |
CN106828304A (en) | Car backing warning method using structure light sensing obstacle | |
CN110085042A (en) | A kind of vehicle drive early warning system and method based on information fusion | |
CN108346316A (en) | A kind of prevention and control system violating the regulations of the intelligence based on lane detection | |
CN107146423A (en) | A kind of crossing inlet track vehicle guiding signal control device and method | |
CN116030662B (en) | Intelligent safety detection system and method based on big data | |
CN109348179A (en) | A kind of road monitoring detection system and method based on artificial intelligence | |
CN112466159B (en) | Right-turning safety early warning system for large vehicle | |
CN110533959A (en) | No-shell culture conflict Vehicle intelligent collision warning method and system based on bus or train route collaboration | |
KR102107791B1 (en) | Safety speed control system for children's protection zone | |
CN108010324A (en) | Highway and Emergency Vehicle Lane traffic alarm system | |
CN102024336A (en) | Intelligent derivation device of slip-road | |
CN207216868U (en) | A kind of crossing inlet track vehicle guiding signal control device | |
CN113178081B (en) | Vehicle immission early warning method and device and electronic equipment | |
CN115762174A (en) | Intelligent transportation system based on image processing | |
CN212847147U (en) | Pedestrian perception road side system with cooperative warning signals | |
CN114299715A (en) | Expressway information detection system based on videos, laser radar and DSRC | |
KR102187441B1 (en) | Distinguishable drone for abnormal driving |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |