CN108830159A - A kind of front vehicles monocular vision range-measurement system and method - Google Patents
A kind of front vehicles monocular vision range-measurement system and method Download PDFInfo
- Publication number
- CN108830159A CN108830159A CN201810475739.6A CN201810475739A CN108830159A CN 108830159 A CN108830159 A CN 108830159A CN 201810475739 A CN201810475739 A CN 201810475739A CN 108830159 A CN108830159 A CN 108830159A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- front vehicles
- information
- module
- relay station
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/582—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of traffic signs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/155—Ground-based stations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Computer Graphics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of front vehicles monocular vision range-measurement system and methods, by constructing the front vehicles monocular vision range-measurement system based on the center of calculating, relay station and vehicle end, after the monocular vision video camera of vehicle loading acquires direction of advance image, lane line information and road signs information the auxiliary monocular vision front vehicles ranging using calculating center identification and extracted front vehicles information, and relay station is combined to provide;Present system structure is simple, and method takes full advantage of the information exchange advantage of car networking, on the basis of monocular vision ranging, using front vehicles trailer information, front vehicles ranging is realized under lane line and road signs information auxiliary, range accuracy is high, calculation amount is small, good reliability.
Description
Technical field
The invention belongs to car steering fields, and in particular to a kind of front vehicles monocular vision range-measurement system and method.
Background technique
Safe distance between vehicles refers to that front vehicle in order to avoid accident collision occurs with front vehicles, is protected with front truck under steam
The necessary spacing distance held.Currently, driver usually passes through the distance that observation judges same front truck indirectly.With computer technology,
Electronic technology continues to develop, and the sensors such as vehicle drive assist system integrated application visual sensor, radar, laser radar can
Independently to find and be vigilant the unsafe condition of road.Wherein, visual sensor information perception, target identification ability are stronger, exist
The functions such as pedestrian detection, road traffic sign detection, lane departure warning are realized in the ADAS of part.However, compared to millimeter wave thunder
It reaches, visual sensor is not yet applied in front vehicles ranging, and main cause is binocular distance measurement distant object ranging essence
Degree is not high, and Stereo Matching Algorithm operand is big, and real-time is not strong;Monocular vision in no space without under geometry dimensional constraints,
It can only determine the orientation of target.
In recent years, front vehicles visual token has obtained related fields scientific research personnel's extensive concern, has carried out based on machine
Front vehicles detection and the ranging of vision are studied.For example, Tong Zhuoyuan (academic dissertation《Front vehicles detection based on machine vision
With Ranging System》The Harbin [D]:Harbin Institute of Technology, 2015.) utilize front vehicles shade width, lane line point
The outer geometrical constraints such as parameter and intrinsic parameter of boundary line, camera, calculate front vehicles distance.However, the prior art relies primarily on vehicle
The environment information of itself perception, and the calculation processing ability of car-mounted computer is limited, and the speed and precision of identification is difficult to
Reach the requirement of high speed moving vehicle ranging.
With using car networking, cloud computing as the development of the intelligent transportation system technology of main feature, make monocular vision application
It is possibly realized in front vehicles ranging:Intelligent identification technology is relied on, vehicle can accurately identify front vehicles vehicle and road road sign
Will type;Car networking technology is relied on, vehicle is expected to obtain road sign geometric scale letter abundant from traffic data center in real time
Breath;Based on image vision processing method, vehicle location is accurately positioned according to tailstock feature in the picture, is mentioned in road ROI region
Pick-up diatom.
Summary of the invention
The purpose of the present invention is to provide a kind of range accuracy is higher, calculation amount is lower, the higher front vehicles of reliability
Monocular vision range-measurement system and method.
The object of the present invention is achieved like this, a kind of front vehicles monocular vision range-measurement system, including:Calculating center,
Relay station, vehicle end;The state of motion of vehicle for calculating the forwarding of central collection relay station sends service response to relay station;Relaying
It stands and vehicle end and calculating center to center communications;Vehicle end sends traffic information, vehicle navigation information to relay station.Communication mode includes
Internet, wireless network, mobile communications network, satellite communication;
Calculating center includes three module layers:Cloud layer, communication layers and client tier;Cloud layer includes that infrastructure services mould
Block, platform, that is, service module, software, that is, service module and data center;Wherein, software, that is, service module includes vehicle cab recognition mould
Block, Traffic Sign Recognition module and supplemental functionality;Communication layers include:The Internet module, wireless network module, mobile communication
Module, satellite communication module;Client tier includes:Relay station management module, relay station respond module, car-mounted terminal management module
With car-mounted terminal respond module;
Relay station includes three module layers:Local cloud layer, local communication layer and native client end layer;Local cloud layer includes this
Ground data center and local software service module;Local software service module includes vehicle cab recognition module, Traffic Sign Recognition mould
Block and supplemental functionality;Local communication layer includes:The Internet module, wireless network module, mobile communication module, satellite communication
Module;Native client end layer includes:Car-mounted terminal management module and car-mounted terminal respond module;
Vehicle end includes:Vehicle communication layer and vehicular client layer;Vehicle communication layer includes wireless network module, moves and lead to
Believe module, satellite communication module;Vehicular client layer includes that human-computer interaction module, vehicle software service module and vehicle intelligent are set
It is standby;Vehicle software service module includes:Vehicle tracking module, lane line extraction module, traffic sign extraction module, vehicle extraction
Module, monocular vision resolve module and ADAS supplementary module;Vehicle intelligent equipment includes vehicle-mounted computer terminal, smart phone, vehicle
Carry global position system and camera.
The realization of the object of the invention further includes a kind of method based on front vehicles monocular vision range-measurement system, including such as
Lower step:
Step 1:Initialization, vehicle end establish the communication with calculating center, relay station respectively;
Calculating task is distributed by calculating center, wherein the center of calculating executes running environment signature analysis, vehicle target detection;
Relay station executes lane line, Traffic Sign Recognition function;Vehicle end executes Lane detection and apart from resolving function;
Step 2:Vehicle advances, and vehicle-mounted camera shoots vehicle forward direction picture, and picture is uploaded to by vehicle communication layer
Calculating center, relay station;Calculating center and relay station determine vision measurement plan by the running environment feature in analysis picture
Slightly, dimensional information is returned;
Step 3:Vehicle end runs vision algorithm, extracts running environment characteristic coordinates, in conjunction with dimensional information, completes front
Vehicle monocular vision ranging;
Further, vision measurement strategy is determined described in step 2 and is the step of returning to dimensional information:
Calculating center determines visual token available information by the running environment feature in analysis picture:
If only vehicle tail is clear in the picture visual field, executes and resolved according to the distance of vehicle tail information:By calculating center
The identification for completing front vehicles sends the tail portion dimensional information of front vehicles to vehicle end by data center;
If vehicle tail is clear in the picture visual field and vehicle traveling direction lane line sharpness of border, execute according to vehicle tail
The distance of information and lane line information resolves:The identification that front vehicles are completed by calculating center exports front vehicle by data center
Tail portion dimensional information to vehicle end;Lane linear content letter of the relay station according to traffic section locating for vehicle location output vehicle
Breath, is sent to vehicle end by local data center;
If vehicle tail is clear in the picture visual field, traffic mark occurs in vehicle traveling direction lane line sharpness of border and the visual field
Will is executed and is resolved according to the distance of vehicle tail information, lane line information and road signs information:Front is completed by calculating center
The identification of vehicle, by the tail portion dimensional information of data center's output front vehicles to vehicle end;Relay station is defeated according to vehicle location
Lane line dimensional information, the road signs information of traffic section locating for vehicle out, is exported by local data center to vehicle end;
Further, vehicle end runs vision algorithm in step 3, specially:
It is resolved if executing according to the distance of vehicle tail information, the calculation expression of front vehicles distance D is:
In formula, S is the front vehicles tail portion dimensional information that calculating center is returned to vehicle end;S is the coordinate of front truck target
The area under image coordinate system;kx,kyIt is preparatory calibration camera parameter;
It is resolved if executing according to the distance of vehicle tail information and lane line information, the calculation method of front vehicles distance D
For:Image coordinate (the u of vehicle end acquisition lane linei,vi) (i=1,2,3 ..., 8), obtain vehicle tail plane people having a common goal
Endpoint (the u of road plane boundary line10,v10)、(u20,v20);The front vehicles tail width information L returned in conjunction with the center of calculating0With
World coordinate system origin is arranged in the lane line dimensional information that relay station returns, and calculates lane line coordinates (Xi,Yi,Zi) (i=1,2,
3 ..., 8), obtain perspective parameter matrix m ';According to perspective projection relationship, the overdetermined equation of fusion vehicle width information is established,
By least square method, extreme coordinates X=[X of the vehicle tail plane with road plane boundary line is solvedw1,Yw1,Xw2,Yw2]T,
Then front vehicles distance is:
Resolved if executing according to the distance of vehicle tail information, lane line information and road signs information, front vehicles away from
Calculation method from D is:
Wherein, S1To calculate the front vehicles tail portion dimensional information that center is returned to vehicle end, S2It is relay station to vehicle end
The traffic sign dimensional information of return;s1、s2Respectively front vehicles tail portion and the traffic sign image coordinate system in the picture visual field
Under area;
Further, vehicle end installs millimetre-wave radar, for detecting the distance between vehicle and front vehicles;
If the distance between vehicle and front vehicles meet 2m < D≤10m, the calculation expression of front vehicles distance D is:
In formula, S is front vehicles tail area information, and the data center by calculating center provides;S is the seat of front truck target
It is marked on area under image coordinate system;kx,kyIt is preparatory calibration camera parameter;
If the distance between vehicle and front vehicles meet 10m < D≤30m, the calculation method of front vehicles distance D is:
Image coordinate (the u of vehicle end acquisition lane linei,vi) (i=1,2,3 ..., 8), it is flat with road to obtain vehicle tail plane
Endpoint (the u of face boundary line10,v10)、(u20,v20);The front vehicles tail width information L returned in conjunction with the center of calculating0And relaying
Stand return lane line dimensional information, be arranged world coordinate system origin, calculate lane line coordinates (Xi,Yi,Zi) (i=1,2,
3 ..., 8), obtain perspective parameter matrix m ';According to perspective projection relationship, the overdetermined equation of fusion vehicle width information is established,
By least square method, extreme coordinates X=[X of the vehicle tail plane with road plane boundary line is solvedw1,Yw1,Xw2,Yw2]T,
Then front vehicles distance is:
If the distance between vehicle and front vehicles meet D > 30m, the calculation method of front vehicles distance D is:
Wherein, S1To calculate the front vehicles tail portion dimensional information that center is returned to vehicle end, S2It is relay station to vehicle end
The traffic sign dimensional information of return;s1、s2Respectively front vehicles tail portion and the traffic sign image coordinate system in the picture visual field
Under area.
The present invention has the advantages that:
1, present system establishes vehicle data center to the information of vehicles in road environment using calculating center, is calculating
The vehicle identification with a wide range of versatility is completed at center, and vehicle image is identified that this needs occupies the portion of a large amount of calculation amounts
Point, it is executed at the center of calculating, recognition efficiency is high;
2, the data center at relay station stores the lane line and road signs information in locating section, according to monocular vision
Effect decision is transferred to vehicle end auxiliary information, effectively controls the data interaction amount of relay station;
3, vehicle end executes leading vehicle distance according to the auxiliary information under different information conditions and resolves, and calculation amount is small, calculates real
When property is good, under the auxiliary of lane line and traffic sign, can be realized and accurately calculates to leading vehicle distance.
Detailed description of the invention
Fig. 1 is present system overall framework figure.
Fig. 2 is the calculating central frame figure of present system.
Fig. 3 is the relay station frame diagram of present system.
Fig. 4 is the vehicle end frame diagram of present system.
Fig. 5 is the basic flow chart of the method for the present invention.
Fig. 6 is the ranging information source schematic diagram of the method for the present invention.
Fig. 7 is vision measurement when meeting 2m < D≤10m at a distance from vehicle is between front vehicles of the method for the present invention
Tactful schematic diagram.
Fig. 8 is that vision when meeting 10m < D≤20m at a distance from vehicle is between front vehicles of the method for the present invention is surveyed
Measure tactful schematic diagram.
Fig. 9 is that vision when meeting 20m < D≤30m at a distance from vehicle is between front vehicles of the method for the present invention is surveyed
Measure tactful schematic diagram.
Figure 10 is vision measurement plan when meeting D > 30m at a distance from vehicle is between front vehicles of the method for the present invention
Slightly schematic diagram.
Figure 11 is that emulation when meeting 2m < D≤10m at a distance from vehicle is between front vehicles of the method for the present invention is real
Test schematic diagram.
Figure 12 is that emulation when meeting 10m < D≤20m at a distance from vehicle is between front vehicles of the method for the present invention is real
Test schematic diagram.
Figure 13 is schematic representation when meeting 20m < D≤30m at a distance from vehicle is between front vehicles of the method for the present invention
It is intended to.
Figure 14 is signal schematic diagram when meeting D > 30m at a distance from vehicle is between front vehicles of the method for the present invention.
Specific embodiment
The present invention is described in further detail below in conjunction with the accompanying drawings.
As shown in Figure 1, a kind of front vehicles monocular vision range-measurement system, including:Calculating center, relay station, vehicle end;Meter
The state of motion of vehicle for calculating the forwarding of central collection relay station sends service response to relay station;Relay station and vehicle end and calculating
Center to center communications;Vehicle end sends traffic information, vehicle navigation information to relay station;
Communication mode includes internet, wireless network, mobile communications network and satellite communication;
As shown in Fig. 2, the center of calculating includes three module layers:Cloud layer, communication layers and client tier;Cloud layer is set including basis
Apply i.e. service module, platform i.e. service module, software i.e. service module and data center;Wherein, software, that is, service module includes
Vehicle cab recognition module, Traffic Sign Recognition module and supplemental functionality.Communication layers include:The Internet module, wireless network mould
Block, mobile communication module, satellite communication module.Client tier includes:It is relay station management module, relay station respond module, vehicle-mounted
Termination management module and car-mounted terminal respond module.
As shown in figure 3, relay station includes three module layers:Local cloud layer, local communication layer and native client end layer;It is local
Cloud layer includes local data center and local software service module.Local software service module includes vehicle cab recognition module, traffic
Landmark identification module and supplemental functionality.Local communication layer includes:The Internet module, wireless network module, mobile communication mould
Block, satellite communication module.Native client end layer includes:Car-mounted terminal management module and car-mounted terminal respond module.
As shown in figure 4, vehicle end includes:Vehicle communication layer and vehicular client layer;Vehicle communication layer includes wireless network
Module, mobile communication module, satellite communication module.Vehicular client layer includes human-computer interaction module, vehicle software service module
And vehicle intelligent equipment.Vehicle software service module includes:Vehicle tracking module, lane line extraction module, cooperation mark extract
Module, vehicle extraction module, monocular vision resolve module and ADAS supplementary module.Vehicle intelligent equipment includes vehicle-mounted computer end
End, smart phone, Vehicular satellite positioning system and monocular cam.
A method of based on front vehicles monocular vision range-measurement system, include the following steps:
Step 1:Initialization, vehicle end establish the communication with calculating center, relay station respectively;
Calculating task is distributed by calculating center, wherein the center of calculating executes running environment signature analysis, vehicle target detection;
Relay station executes lane line, Traffic Sign Recognition function;Vehicle end executes Lane detection and apart from resolving function;
Step 2:Vehicle advances, and vehicle-mounted camera shoots vehicle forward direction picture, and picture is uploaded to by vehicle communication layer
Calculating center, relay station;Calculating center and relay station determine vision measurement plan by the running environment feature in analysis picture
Slightly, dimensional information is returned;
Calculating center, which utilizes, is based on fast area convolutional neural networks (Fast R-CNN), to vehicle mesh in scene image
The targets such as mark, traffic sign are detected and are classified.
When detecting traffic target, visual task is defined first, using target to be detected training network, obtains sample image
Candidate region coordinate is inputted the study of depth convolutional neural networks together with visual task example image, obtains depth by candidate region
Spend convolution feature;Then, it is based on network structure, different complete connect by proper normalization and is inputted by area-of-interest pond layer
Branch is met, parallel return calculates tagsort and detection block coordinate value;Finally, obtaining and specifying by successive ignition training
The target detection model of visual task strong correlation.
When classification traffic target:The data set with mark is input in network first, convolution characteristic pattern is generated, will wait
Favored area is mapped in shared characteristic pattern, obtains corresponding characteristic information;Pondization operation is carried out to this feature, obtains region spy
Chi Huatu is levied, the feature vector of the final feature comprising each candidate region is obtained by full articulamentum, by this feature vector point
It is not input in classifier, using non-maxima suppression, judges target category and the position of candidate region;Finally utilize judgement
Difference between value and practical mark value obtains loss function, is joined by back-propagation algorithm and stochastic gradient descent method to network
Number optimizes, and finally obtains output network.Related experiment shows to choose 15 kinds of traffic signs, the classification success rate of this method
Close to 90%.
To extract lane line, a variety of visual processing methods need to be integrated:Image horizon is extracted first, below horizon
For image ROI region, binary conversion treatment is carried out to road gray level image based on Otsu method, chooses Canny operator to binary map
As carrying out edge extracting;Then, inverse perspective mapping is carried out to image, image is projected to three-dimensional space from two-dimensional space, is obtained
Point in original image space on straight line is converted to the curve in parameter space through Hough transform by road overhead view image, on
Curve intersection is stated in forming peak point on some point, to obtain the straight line in original image, is joined by adjusting the threshold value of Hough transform
Several and straight line angle, finishing screen select lane line;Finally, image same edge image of the registration containing lane line, obtains lane boundary
Line profile and part angle point.
Vision measurement strategy is determined described in step 2 and is the step of returning to dimensional information:
As shown in fig. 6, the center of calculating carries out visual token available information by the running environment feature in analysis picture
Determine.
If only vehicle tail is clear in the picture visual field, executes and resolved according to the distance of vehicle tail information:By calculating center
The identification for completing front vehicles sends the tail portion dimensional information of front vehicles to vehicle end by data center;
If vehicle tail is clear in the picture visual field and vehicle traveling direction lane line sharpness of border, execute according to vehicle tail
The distance of information and lane line information resolves:The identification that front vehicles are completed by calculating center exports front vehicle by data center
Tail portion dimensional information to vehicle end;Lane linear content letter of the relay station according to traffic section locating for vehicle location output vehicle
Breath, is sent to vehicle end by local data center;
If vehicle tail is clear in the picture visual field, traffic mark occurs in vehicle traveling direction lane line sharpness of border and the visual field
Will is executed and is resolved according to the distance of vehicle tail information, lane line information and road signs information:Front is completed by calculating center
The identification of vehicle, by the tail portion dimensional information of data center's output front vehicles to vehicle end;Relay station is defeated according to vehicle location
Lane line dimensional information, the road signs information of traffic section locating for vehicle out, is exported by local data center to vehicle end.
Further, vehicle end runs vision algorithm in step 3, specially:
It is resolved if executing according to the distance of vehicle tail information, the calculation expression of front vehicles distance D is:
In formula, S is the front vehicles tail portion dimensional information that calculating center is returned to vehicle end;S is the coordinate of front truck target
The area under image coordinate system;kx,kyIt is preparatory calibration camera parameter;
In urban environment, lane line continuously occurs on road, and city expressway line of demarcation is to draw the interval 400cm
600cm, other highway boundary lines draw the interval 2m 4m;Traffic sign often appears near crossing, typical rectangular traffic sign ruler
Very little width is 4m, a height of 2m.
It is resolved if executing according to the distance of vehicle tail information and lane line information, the calculation method of front vehicles distance D
For:Image coordinate (the u of vehicle end acquisition lane linei,vi) (i=1,2,3 ..., 8), obtain vehicle tail plane people having a common goal
Endpoint (the u of road plane boundary line10,v10)、(u20,v20);The front vehicles tail width information L returned in conjunction with the center of calculating0With
World coordinate system origin is arranged in the lane line dimensional information that relay station returns, and calculates lane line coordinates (Xi,Yi,Zi) (i=1,2,
3 ..., 8), obtain perspective parameter matrix m ';According to perspective projection relationship, the overdetermined equation of fusion vehicle width information is established,
By least square method, extreme coordinates X=[X of the vehicle tail plane with road plane boundary line is solvedw1,Yw1,Xw2,Yw2]T,
Then front vehicles distance is:
Resolved if executing according to the distance of vehicle tail information, lane line information and road signs information, front vehicles away from
Calculation method from D is:
Wherein, S1To calculate the front vehicles tail portion dimensional information that center is returned to vehicle end, S2It is relay station to vehicle end
The traffic sign dimensional information of return;s1、s2Respectively front vehicles tail portion and the traffic sign image coordinate system in the picture visual field
Under area;
Step 3:Vehicle end runs vision algorithm, extracts running environment characteristic coordinates, in conjunction with dimensional information, completes front
Vehicle monocular vision ranging.
For promoted monocular vision ranging scheme clock availability, can visual token simultaneously, with reference to vehicle-mounted millimeter wave thunder
Up to ranging information, leading vehicle distance range is obtained, to adaptively choose lane line or traffic mark in camera view
Will, the auxiliary positioning mark as monocular vision ranging.Vehicle end installs millimetre-wave radar, for detecting vehicle and front vehicles
The distance between;
As shown in fig. 7, if the distance between vehicle and front vehicles meet 2m < D≤10m, the meter of front vehicles distance D
Operator expression formula is:
In formula, S is front vehicles tail area information, and the data center by calculating center provides;S is the seat of front truck target
It is marked on area under image coordinate system;kx,kyIt is preparatory calibration camera parameter;
As shown in figure 8, leading vehicle distance meets 10m < D≤20m, front truck tail portion, lane two sides boundary line in camera view
And its edge clear is as it can be seen that can be used as cooperation mark auxiliary visual token.Vehicle cab recognition is carried out by calculating center, returns to front
Vehicle tail width information L0, the image coordinate (u of vehicle end acquisition lane linei,vi) (i=1,2,3 ..., 8), obtain
Endpoint (u of the vehicle tail plane with road plane boundary line10,v10)、(u20,v20);Vehicle end is by requesting the area to relay station
Domain lane line, traffic sign dimensional information, and world coordinate system origin is rationally set, calculate cooperation marker coordinates (Xi,Yi,
Zi) (i=1,2,3 ..., 8), obtain perspective parameter matrix m ';According to perspective projection relationship, fusion vehicle width information is established
Overdetermined equation extreme coordinates X=[X of the vehicle tail plane with road plane boundary line is solved by least square methodw1,
Yw1,Xw2,Yw2]T, then front vehicles distance be:
As shown in figure 9, leading vehicle distance meets 20m < D≤30m, front truck tail portion, lane two sides boundary line in camera view
It is high-visible, it can be used as cooperation mark auxiliary visual token.Under the distance, vehicle tail as it can be seen that lane line away from camera shooting
Farther away angle point cannot be extracted accurately, therefore using lane line two sides endpoint as geometrical constraint, then front vehicles distance is:
As shown in Figure 10, leading vehicle distance meets D > 30m, and distant place lane line is unintelligible in camera view, the preceding tailstock
Portion, traffic sign can be used as cooperation mark auxiliary visual token.Vehicle cab recognition is carried out by calculating center, before returning to vehicle end
Square vehicle tail, traffic sign area information S1、S2, the coordinate of vehicle end acquisition front truck target, area is under image coordinate system
s1、s2;Vehicle is D with nearest traffic sign distance1, then front vehicles distance be:
Using the powerful calculating of present system, storage, communication capacity, front truck mesh can be disposed in the center of calculating, relay station
Mark, typical traffic Mark Detection classification feature;It is extracted in vehicle end deployment lane line and apart from resolving function;It is converged through data
Collection is chosen with strategy, final to resolve front vehicles distance.
In practical applications, behind vehicle replacement lane, image is uploaded to calculating center, relay station, calculating center passes through
Environmental characteristic is analyzed, determines that vision measurement strategy, the type of vehicle that identification front travels return to vehicle scale information;Relaying
It stands to detect and can use lane line and traffic sign in the visual field, and return to its dimensional information;Vehicle end runs vision algorithm, extracts traveling
Environmental characteristic, front truck target, lane line, typical traffic sign image coordinate;The monocular view returned according to the center of calculating, relay station
Feel that measurement strategies and dimensional information, vehicle end complete the ranging of front vehicles monocular vision.
Specific embodiment
To verify monocular vision measuring principle correctness, 4 kinds of typical environment are measured for monocular vision, pass through AutoCAD
Software establishes urban road three-dimensional simulation model.In model, lane width 4m, lane line is to draw 2m, lane line width
0.2m;Rectangle traffic sign size width is 4m, a height of 2m;Simulating vehicle-mounted camera installation site is (0m, 0m, 2m), focal length
For 30mm, image resolution ratio is 1024 × 768, intrinsic parameter kx、kyProduct isFront lorry height is set as
4.5m, width are about 3.4m.
As shown in figure 11, in front of adjustment vehicle location to camera within 10m, using truck tail as measured target,
In world coordinate system, width, height, area are respectively 3.400m, 4.492m, 15.274m2, wide under image coordinate system
Degree, height, area are respectively 315pixel (pixel), 418pixel, 131670pixel2, measured value 9.166m, true value
For 9.364m, visual token error is no more than 3%.
As shown in figure 12, it is no more than 20m, using truck tail as quilt more than 10m in front of adjustment vehicle location to camera
It surveys target and 8 lanes edge angle point is chosen, under world coordinate system using lane line length and its interval as geometrical constraint
Coordinate is:(-4,8,0),(-3.8,8,0),(0,8,0),(0.2,8,0),(-4,12,0),(-3.8,12,0),(0,12,0),
(0.2,12,0), unit m;Coordinate under image coordinate system is:(76,610),(98,610),(512,610),(532,
610), (223,535), (236,535), (512,535), (525,535), unit pixel.M ' matrix is calculated, front is obtained
Vehicle measured value is 16.394m, and true value 17.337m, visual token error is close to 6%, if selecting vertical direction cooperation mark
Will, front vehicles measured value are 17.076m, and range error is close to 2%.
As shown in figure 13, it is no more than 30m more than 20m in front of adjustment vehicle location to camera, chooses 8 lanes edge
Angle point, the coordinate in world coordinate system are:(-4,12,0),(-4,14,0),(-4,18,0),(-4,20,0),(0.2,12,
0), (0.2,14,0), (0.2,18,0), (0.2,20,0), unit m;Coordinate in image coordinate system is:(223,535),
(263,514), (319,484), (336,475), (525,535), (524,514), (520,484), (520,475), unit are
pixel.Measured value is 21.257m, true value 23.341m, and visual token error is close to 9%.
As shown in figure 14, other than adjustment vehicle location to 30m in front of camera, using vehicle tail, traffic sign as quilt
Target is surveyed, in world coordinate system, truck tail width, height, area are respectively 3.400m, 4.492m, 15.274m2, traffic
Label width, height, area are respectively 4.000m, 2.000m, 8.000m2, under image coordinate system, truck tail width, height
Degree, area are respectively 69pxiel, 92pxiel, 6348pixel2, traffic sign width, height, area be respectively 96pixel,
48pixel, 4608pixel2, measured value 41.9303m, true value 41.860m.Under ecotopia, when front, spacing is greater than
30m, the vehicle target visual token result based on traffic sign are 41.9294m, and range error is close to 1%;If vehicle in image
Width extracts error and is less than 6pixel, and visual token error is close to 8%.
Based on above-mentioned test data, showing system and method for the present invention range accuracy with higher, data calculation amount is small,
Under the auxiliary of traffic sign and lane line information, the range performance of method is stablized, and has stronger robustness.
Claims (3)
1. a kind of front vehicles monocular vision range-measurement system, which is characterized in that including:
Calculating center, relay station, vehicle end;The state of motion of vehicle for calculating the forwarding of central collection relay station is sent to relay station
Service response;Relay station and vehicle end and calculating center to center communications;Vehicle end sends traffic information and automobile navigation letter to relay station
Breath, communication mode includes internet, wireless network, mobile communications network, satellite communication;
The calculating center includes three module layers:Cloud layer, communication layers and client tier;The cloud layer includes infrastructure
That is service module, platform, that is, service module, software, that is, service module and data center;Software, that is, the service module includes vehicle
Type identification module, Traffic Sign Recognition module and supplemental functionality;The communication layers include the Internet module, wireless network
Module, mobile communication module and satellite communication module;The client tier includes relay station management module, relay station response mould
Block, car-mounted terminal management module and car-mounted terminal respond module;
The relay station includes three module layers:Local cloud layer, local communication layer and native client end layer;The local cloud
Layer includes local data center and local software service module;The local software service module include vehicle cab recognition module,
Traffic Sign Recognition module and supplemental functionality;The local communication layer includes the Internet module, wireless network module, shifting
Dynamic communication module and satellite communication module;The native client end layer includes car-mounted terminal management module and car-mounted terminal response
Module;
The vehicle end includes vehicle communication layer and vehicular client layer;The vehicle communication layer includes wireless network mould
Block, mobile communication module and satellite communication module;The vehicular client layer includes human-computer interaction module, vehicle software service
Module and vehicle intelligent equipment, the vehicle software service module include vehicle tracking module, lane line extraction module, traffic
Indicate that extraction module, vehicle extraction module, monocular vision resolve module and ADAS supplementary module, the vehicle intelligent equipment packet
Include vehicle-mounted computer terminal, smart phone, Vehicular satellite positioning system and camera.
2. a kind of distance measuring method as described in claim 1 based on front vehicles monocular vision range-measurement system, which is characterized in that
Include the following steps
Step 1:Initialization, vehicle end establish the communication with calculating center, relay station respectively;Appointed by calculating center distribution and calculating
Business, wherein the center of calculating executes running environment signature analysis, vehicle target detection;Relay station executes lane line, traffic sign is known
Other function;Vehicle end executes Lane detection and apart from resolving function;
Step 2:Vehicle advances, and vehicle-mounted camera shoots vehicle forward direction picture, and picture is uploaded to calculating by vehicle communication layer
Center and relay station;Calculating center and relay station determine vision measurement strategy simultaneously by the running environment feature in analysis picture
Return to dimensional information;
It wherein calculates center and passes through the running environment feature in analysis picture according to the following steps, visual token available information is carried out
Determine:
If only vehicle tail is clear in the picture visual field, executes and resolved according to the distance of vehicle tail information:It is completed by calculating center
The identification of front vehicles sends the tail portion dimensional information of front vehicles to vehicle end by data center;
If vehicle tail is clear in the picture visual field and vehicle traveling direction lane line sharpness of border, execute according to vehicle tail information
It is resolved with the distance of lane line information:The identification that front vehicles are completed by calculating center, by data center's output front vehicles
Tail portion dimensional information is to vehicle end;Lane line dimensional information of the relay station according to traffic section locating for vehicle location output vehicle, warp
Vehicle end is sent to by local data center;
If vehicle tail is clear in the picture visual field, traffic sign occurs in vehicle traveling direction lane line sharpness of border and the visual field,
It executes and is resolved according to the distance of vehicle tail information, lane line information and road signs information:Front vehicle is completed by calculating center
Identification, by the tail portion dimensional information of data center's output front vehicles to vehicle end;Relay station is exported according to vehicle location
Lane line dimensional information, the road signs information of traffic section locating for vehicle, export via local data center to vehicle end
Step 3:Vehicle end runs vision algorithm, extracts running environment characteristic coordinates, in conjunction with dimensional information, completes front vehicles
Monocular vision ranging;
The vehicle end runs vision algorithm:
It is resolved if executing according to the distance of vehicle tail information, the calculation expression of front vehicles distance D is:
In formula, S is the front vehicles tail portion dimensional information that calculating center is returned to vehicle end;S is that the coordinate of front truck target is being schemed
As area under coordinate system;kx,kyIt is preparatory calibration camera parameter;
It is resolved if executing according to the distance of vehicle tail information and lane line information, the calculation method of front vehicles distance D is:Vehicle
End obtain lane line image coordinate (ui,vi) (i=1,2,3 ..., 8), obtain the same road plane of vehicle tail plane
Endpoint (the u of boundary line10,v10)、(u20,v20);The front vehicles tail width information L returned in conjunction with the center of calculating0And relay station
World coordinate system origin is arranged in the lane line dimensional information of return, calculates lane line coordinates (Xi,Yi,Zi) (i=1,2,3 ...,
8) perspective parameter matrix m ', is obtained;According to perspective projection relationship, the overdetermined equation of fusion vehicle width information is established, by most
Small square law solves extreme coordinates X=[X of the vehicle tail plane with road plane boundary linew1,Yw1,Xw2,Yw2]T, then front
Vehicle distances are:
It is resolved if executing according to the distance of vehicle tail information, lane line information and road signs information, front vehicles distance D's
Calculation method is:
In formula, S1To calculate the front vehicles tail portion dimensional information that center is returned to vehicle end, S2It is returned for relay station to vehicle end
Traffic sign dimensional information;s1、s2Respectively front vehicles tail portion and traffic sign are in the picture visual field under image coordinate system
Area.
3. a kind of front vehicles monocular vision distance measuring method according to claim 2, which is characterized in that the vehicle end
Millimetre-wave radar is installed, for detecting the distance between vehicle and front vehicles;
If the distance between vehicle and front vehicles meet 2m < D≤10m, the calculation expression of front vehicles distance D is:
In formula, S is front vehicles tail area information, is provided by data center;S is the coordinate of front truck target in image coordinate system
Lower area;kx,kyIt is preparatory calibration camera parameter;
If the distance between vehicle and front vehicles meet 10m < D≤30m, the calculation method of front vehicles distance D is:Vehicle
End obtains the image coordinate (u of lane linei,vi) (i=1,2,3 ..., 8) obtain vehicle tail plane and hand over road plane
Endpoint (the u in boundary line10,v10)、(u20,v20);The front vehicles tail width information L returned in conjunction with the center of calculating0It is returned with relay station
World coordinate system origin is arranged in the lane line dimensional information returned, calculates lane line coordinates (Xi,Yi,Zi) (i=1,2,3 ...,
8) perspective parameter matrix m ', is obtained;According to perspective projection relationship, the overdetermined equation of fusion vehicle width information is established, by most
Small square law solves extreme coordinates X=[X of the vehicle tail plane with road plane boundary linew1,Yw1,Xw2,Yw2]T, then front
Vehicle distances are:
If the distance between vehicle and front vehicles meet D > 30m, the calculation method of front vehicles distance D is:
In formula, S1To calculate the front vehicles tail portion dimensional information that center is returned to vehicle end, S2It is returned for relay station to vehicle end
Traffic sign dimensional information;s1、s2Respectively front vehicles tail portion and traffic sign are in the picture visual field under image coordinate system
Area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810475739.6A CN108830159A (en) | 2018-05-17 | 2018-05-17 | A kind of front vehicles monocular vision range-measurement system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810475739.6A CN108830159A (en) | 2018-05-17 | 2018-05-17 | A kind of front vehicles monocular vision range-measurement system and method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108830159A true CN108830159A (en) | 2018-11-16 |
Family
ID=64148214
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810475739.6A Pending CN108830159A (en) | 2018-05-17 | 2018-05-17 | A kind of front vehicles monocular vision range-measurement system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108830159A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110031829A (en) * | 2019-04-18 | 2019-07-19 | 北京联合大学 | A kind of targeting accuracy distance measuring method based on monocular vision |
CN110422176A (en) * | 2019-07-04 | 2019-11-08 | 苏州车萝卜汽车电子科技有限公司 | Intelligent transportation system, automobile based on V2X |
CN111323767A (en) * | 2020-03-12 | 2020-06-23 | 武汉理工大学 | Night unmanned vehicle obstacle detection system and method |
CN111750878A (en) * | 2019-03-28 | 2020-10-09 | 北京初速度科技有限公司 | Vehicle pose correction method and device |
CN112346039A (en) * | 2020-09-15 | 2021-02-09 | 深圳市点创科技有限公司 | Monocular camera-based vehicle distance early warning method, electronic device and storage medium |
CN112428989A (en) * | 2020-10-30 | 2021-03-02 | 惠州华阳通用电子有限公司 | Vehicle control method |
US11024055B2 (en) | 2018-11-29 | 2021-06-01 | Industrial Technology Research Institute | Vehicle, vehicle positioning system, and vehicle positioning method |
CN113111707A (en) * | 2021-03-07 | 2021-07-13 | 上海赛可出行科技服务有限公司 | Preceding vehicle detection and distance measurement method based on convolutional neural network |
CN113232663A (en) * | 2021-06-29 | 2021-08-10 | 西安电子科技大学芜湖研究院 | Control system applied to advanced assistant driving |
TWI754808B (en) * | 2018-11-29 | 2022-02-11 | 財團法人工業技術研究院 | Vehicle, vehicle positioning system, and vehicle positioning method |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101750049A (en) * | 2008-12-05 | 2010-06-23 | 南京理工大学 | Monocular vision vehicle distance measuring method based on road and vehicle information |
CN102490673A (en) * | 2011-12-13 | 2012-06-13 | 中科院微电子研究所昆山分所 | Vehicle active safety control system based on internet of vehicles and control method of vehicle active safety control system |
CN103871243A (en) * | 2014-04-16 | 2014-06-18 | 武汉欧普威科技有限公司 | Wireless vehicle management system and method based on active safety platform |
CN104899554A (en) * | 2015-05-07 | 2015-09-09 | 东北大学 | Vehicle ranging method based on monocular vision |
CN105574552A (en) * | 2014-10-09 | 2016-05-11 | 东北大学 | Vehicle ranging and collision early warning method based on monocular vision |
CN105891811A (en) * | 2015-10-16 | 2016-08-24 | 乐卡汽车智能科技(北京)有限公司 | Method and device for determining distance between vehicles in Internet of Vehicles |
CN106679633A (en) * | 2016-12-07 | 2017-05-17 | 东华大学 | Vehicle-mounted distance measuring system and vehicle-mounted distance measuring method |
JP3212633U (en) * | 2017-07-11 | 2017-09-21 | 新昇汽車科技股▲分▼有限公司 | IoV application service travel monitoring system |
CN107203134A (en) * | 2017-06-02 | 2017-09-26 | 浙江零跑科技有限公司 | A kind of front truck follower method based on depth convolutional neural networks |
AU2015390329A1 (en) * | 2015-04-10 | 2017-11-02 | Huawei Technologies Co., Ltd. | Data sending method and device |
CN107390205A (en) * | 2017-07-20 | 2017-11-24 | 清华大学 | A kind of monocular vision vehicle odometry method that front truck feature is obtained using car networking |
CN107622500A (en) * | 2017-09-29 | 2018-01-23 | 吉林大学 | A kind of the heavy goods vehicles rearward collision method for early warning and its device of view-based access control model sensing |
CN107957583A (en) * | 2017-11-29 | 2018-04-24 | 江苏若博机器人科技有限公司 | A kind of round-the-clock quick unmanned vehicle detection obstacle avoidance system of Multi-sensor Fusion |
-
2018
- 2018-05-17 CN CN201810475739.6A patent/CN108830159A/en active Pending
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101750049A (en) * | 2008-12-05 | 2010-06-23 | 南京理工大学 | Monocular vision vehicle distance measuring method based on road and vehicle information |
CN102490673A (en) * | 2011-12-13 | 2012-06-13 | 中科院微电子研究所昆山分所 | Vehicle active safety control system based on internet of vehicles and control method of vehicle active safety control system |
CN103871243A (en) * | 2014-04-16 | 2014-06-18 | 武汉欧普威科技有限公司 | Wireless vehicle management system and method based on active safety platform |
CN105574552A (en) * | 2014-10-09 | 2016-05-11 | 东北大学 | Vehicle ranging and collision early warning method based on monocular vision |
AU2015390329A1 (en) * | 2015-04-10 | 2017-11-02 | Huawei Technologies Co., Ltd. | Data sending method and device |
CN104899554A (en) * | 2015-05-07 | 2015-09-09 | 东北大学 | Vehicle ranging method based on monocular vision |
CN105891811A (en) * | 2015-10-16 | 2016-08-24 | 乐卡汽车智能科技(北京)有限公司 | Method and device for determining distance between vehicles in Internet of Vehicles |
CN106679633A (en) * | 2016-12-07 | 2017-05-17 | 东华大学 | Vehicle-mounted distance measuring system and vehicle-mounted distance measuring method |
CN107203134A (en) * | 2017-06-02 | 2017-09-26 | 浙江零跑科技有限公司 | A kind of front truck follower method based on depth convolutional neural networks |
JP3212633U (en) * | 2017-07-11 | 2017-09-21 | 新昇汽車科技股▲分▼有限公司 | IoV application service travel monitoring system |
CN107390205A (en) * | 2017-07-20 | 2017-11-24 | 清华大学 | A kind of monocular vision vehicle odometry method that front truck feature is obtained using car networking |
CN107622500A (en) * | 2017-09-29 | 2018-01-23 | 吉林大学 | A kind of the heavy goods vehicles rearward collision method for early warning and its device of view-based access control model sensing |
CN107957583A (en) * | 2017-11-29 | 2018-04-24 | 江苏若博机器人科技有限公司 | A kind of round-the-clock quick unmanned vehicle detection obstacle avoidance system of Multi-sensor Fusion |
Non-Patent Citations (9)
Title |
---|
CHAI L等: ""A Simulation Scheme for Testing the Effects of Key Indicators of the Internet of Vehicles on Vehicle Safety"", 《AUTOMOTIVE ENGINEERING》 * |
LANG W等: ""Research on Preceding Vehicle Distance Measurement with Monocular Vision Based on Lane Plane Geometric Model"", 《AN INTERNATIONAL JOURNAL OF RESEARCH & SURVEYS》 * |
WANG W等: ""A rough vehicle distance measurement method using monocular vision and license plate"", 《IEEE INTERNATIONAL CONFERENCE ON CYBER TECHNOLOGY IN AUTOMATION》 * |
姚春莲等: ""利用车牌面积进行车距测量的研究"", 《系统仿真学报》 * |
孟新育等: ""基于车联网的汽车主动防撞预警系统设计"", 《现代交通技术》 * |
杨炜等: ""融合道路图像关键信息的车距视觉测量方法"", 《计算机工程与应用》 * |
王富罗: ""远程云中心协助的车联网防碰撞系统设计"", 《贵阳学院学报(自然科学版)》 * |
袁帅: ""毫米波雷达在智能网联汽车中的应用"", 《信息技术与标准化》 * |
覃华: ""大数据驱动的车联网信息物理融合系统的分析与设计方法"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11024055B2 (en) | 2018-11-29 | 2021-06-01 | Industrial Technology Research Institute | Vehicle, vehicle positioning system, and vehicle positioning method |
TWI754808B (en) * | 2018-11-29 | 2022-02-11 | 財團法人工業技術研究院 | Vehicle, vehicle positioning system, and vehicle positioning method |
CN111750878A (en) * | 2019-03-28 | 2020-10-09 | 北京初速度科技有限公司 | Vehicle pose correction method and device |
CN110031829A (en) * | 2019-04-18 | 2019-07-19 | 北京联合大学 | A kind of targeting accuracy distance measuring method based on monocular vision |
CN110031829B (en) * | 2019-04-18 | 2021-07-09 | 北京联合大学 | Target accurate distance measurement method based on monocular vision |
CN110422176A (en) * | 2019-07-04 | 2019-11-08 | 苏州车萝卜汽车电子科技有限公司 | Intelligent transportation system, automobile based on V2X |
CN111323767A (en) * | 2020-03-12 | 2020-06-23 | 武汉理工大学 | Night unmanned vehicle obstacle detection system and method |
CN111323767B (en) * | 2020-03-12 | 2023-08-08 | 武汉理工大学 | System and method for detecting obstacle of unmanned vehicle at night |
CN112346039A (en) * | 2020-09-15 | 2021-02-09 | 深圳市点创科技有限公司 | Monocular camera-based vehicle distance early warning method, electronic device and storage medium |
CN112428989A (en) * | 2020-10-30 | 2021-03-02 | 惠州华阳通用电子有限公司 | Vehicle control method |
CN112428989B (en) * | 2020-10-30 | 2022-03-11 | 惠州华阳通用电子有限公司 | Vehicle control method |
CN113111707A (en) * | 2021-03-07 | 2021-07-13 | 上海赛可出行科技服务有限公司 | Preceding vehicle detection and distance measurement method based on convolutional neural network |
CN113111707B (en) * | 2021-03-07 | 2024-05-17 | 上海赛可出行科技服务有限公司 | Front car detection and ranging method based on convolutional neural network |
CN113232663A (en) * | 2021-06-29 | 2021-08-10 | 西安电子科技大学芜湖研究院 | Control system applied to advanced assistant driving |
CN113232663B (en) * | 2021-06-29 | 2022-10-04 | 西安电子科技大学芜湖研究院 | Control system applied to advanced assistant driving |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108830159A (en) | A kind of front vehicles monocular vision range-measurement system and method | |
US11593950B2 (en) | System and method for movement detection | |
CN112184818B (en) | Vision-based vehicle positioning method and parking lot management system applying same | |
CN108572663B (en) | Target tracking | |
CN107862287A (en) | A kind of front zonule object identification and vehicle early warning method | |
US20200041284A1 (en) | Map road marking and road quality collecting apparatus and method based on adas system | |
CN102248947B (en) | Object and vehicle detecting and tracking using a 3-D laser rangefinder | |
WO2022141910A1 (en) | Vehicle-road laser radar point cloud dynamic segmentation and fusion method based on driving safety risk field | |
CN102222236B (en) | Image processing system and position measuring system | |
US20200082614A1 (en) | Intelligent capturing of a dynamic physical environment | |
CN107422730A (en) | The AGV transportation systems of view-based access control model guiding and its driving control method | |
CN108693876A (en) | Target following | |
WO2021046716A1 (en) | Method, system and device for detecting target object and storage medium | |
CN107392103A (en) | The detection method and device of road surface lane line, electronic equipment | |
CN111622145B (en) | Unmanned truck-collecting positioning method and system based on reflective spikes | |
CN110119698A (en) | For determining the method, apparatus, equipment and storage medium of Obj State | |
JP2018092501A (en) | On-vehicle image processing apparatus | |
CN110197173B (en) | Road edge detection method based on binocular vision | |
CN102208013A (en) | Scene matching reference data generation system and position measurement system | |
JP5286653B2 (en) | Stationary object map generator | |
CN106683530A (en) | Computerized judging system and method based on three-dimensional laser vision and high-precision lane model | |
CN117576652B (en) | Road object identification method and device, storage medium and electronic equipment | |
CN112749584B (en) | Vehicle positioning method based on image detection and vehicle-mounted terminal | |
CN209214563U (en) | Multi-information fusion spectrum of road surface roughness real-time testing system | |
Hara et al. | Vehicle localization based on the detection of line segments from multi-camera images |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181116 |