CN106530825A - Method for detecting traffic conflict between motor-assisted bicycle and automobile based on ST-MRF model - Google Patents
Method for detecting traffic conflict between motor-assisted bicycle and automobile based on ST-MRF model Download PDFInfo
- Publication number
- CN106530825A CN106530825A CN201611030277.4A CN201611030277A CN106530825A CN 106530825 A CN106530825 A CN 106530825A CN 201611030277 A CN201611030277 A CN 201611030277A CN 106530825 A CN106530825 A CN 106530825A
- Authority
- CN
- China
- Prior art keywords
- electric bicycle
- motion vector
- conflict
- block
- pixels
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/166—Anti-collision systems for active traffic, e.g. moving vehicles, pedestrians, bikes
Abstract
The invention discloses a method for detecting traffic conflict between a motor-assisted bicycle and an automobile based on an ST-MRF model. The method includes the steps of a traffic camera taking three-dimensional video images, tracking single motor-assisted bicycle and automobile by using the ST-MRF model to obtain the traffic information of the single motor-assisted bicycle and automobile, converting the three-dimensional video images taken by the traffic camera into two-dimensional coordinate data through a coordinate conversion algorithm, predicting the parking points of the motor-assisted bicycle and the automobile in the next frame image, and establishing driver conflict determination criteria according to the distance between the parking points. A method for determining traffic between a red-light running motor-assisted bicycle and a normally driving automobile by tracking moving tracks of vehicles, and real-time and accurate traffic information for accident identification can be provided to a traffic management department.
Description
Technical field
The present invention relates to a kind of electric bicycle based on ST-MRF models and mechanical transport collision detection method.
Background technology
As city video equipment quantity, coverage rate are increased rapidly, the lifting of Computer Image Processing level, by video
Image processing techniques becomes a kind of trend for City ITS.Traffic transport system engineering and information, 2013 the 13rd
Rolled up for the 3rd phase:65-70 page. disclose a kind of adaptive vehicle track algorithm research (Zhou Jun, Cheng Lin) based on ST-MRF, its base
In ST-MRF models adaptive vehicle track algorithm can the volume of traffic than it is larger, and vehicle occur mutually block in the case of,
Accurate acquisition vehicle tracking data, provide important data basis for later traffic incidents detection.
In some small and medium-sized cities, as population is more intensive, economy is in developing stage, electric bicycle go on a journey quantity compared with
It is many, go on a journey commonplace, but electric bicycle driver causes the behavior of making a dash across the red light to occur repeatedly due to lacking awareness of safety, causes
Traffic accident takes place frequently, and causes substantial amounts of personnel, property loss.By substantial amounts of accident statistics, accident area occurred frequently is found exactly
City intersection, so outstanding to traffic conflict detection is carried out between vehicle and the electric bicycle for making a dash across the red light in crossing intersection part
For important.
It is very few with the traffic conflict of car detection research with regard to electric bicycle in prior art, herein in current city
Crossing intersection part for electric bicycle traffic conflict detection research it is weaker on the basis of, study and propose it is a set of be adapted in
Traffic conflict detection method between small city crossing intersection part vehicle and the electric bicycle that makes a dash across the red light, and driver is conflicted
Judgment criterion carries out the judgement of conflict spectrum in being used in Computer Vision.
The content of the invention
For the problems referred to above, the present invention is provided a kind of electric bicycle based on ST-MRF models and is conflicted with mechanical transport inspection
A kind of survey method, by vehicle movement track following, there is provided between electric bicycle for making a dash across the red light and the car of normally travel
Traffic conflict detection method, can for vehicle supervision department real-time, accurate transport information be provided for accident identify.
To realize above-mentioned technical purpose, above-mentioned technique effect is reached, the present invention is achieved through the following technical solutions：
Electric bicycle based on ST-MRF models and mechanical transport collision detection method, it is characterised in that including as follows
Step：
Step 1, traffic camera shoot 3 d video images, track single electrical salf-walking simultaneously using ST-MRF models
Car and automobile, obtain the transport information of single electric bicycle and automobile, such as vehicle, the volume of traffic, speed and coordinate value；
Step 2, the 3 d video images that traffic camera shoots are converted into by two-dimensional coordinate number by coordinates transformation method
According to；
Step 3, the prediction stop of electric bicycle and automobile in next two field picture；
Step 4, set up according to the distance between stop driver conflict judgment criterion.
It is preferred that, in step 1, the energy function U of ST-MRF modelsstmrfIt is：
In formula：Part I a (Nyk-μNy)2Represent in target map, the neighbouring relations of label, Part II b (Mxyk-
μMxy)2Represent the hiding relation of the label in successive objective map, Part IIIRepresent the company relevant with motion vector
Texture relation in continuous image, Part IVRepresent the fortune in motion vector map
The neighbouring relations of dynamic vector；
Nyk：Refer to that the adjacent pixel blocks and the block of pixels of a block of pixels have the number of identical label；
Nxk：Represent the number of the adjacent pixel blocks of a block of pixels；
Dxyk：Texture correlation between the image of the image and t that represent the t-1 moment, when blocking, respectively
Calculating belongs to the probability of each car；
Mxyk：The number of pixels of shield portions in two block of pixels of partial occlusion；
μNy：Neighborhood group, if using 8- neighborhood groups, μ Ny=8 is maximum；
Ck：Current pixel block；
Bk：Adjacent pixel blocks；
In the difference of the motion vector of (t-1) moment, current pixel block and adjacent pixel blocks；
A, b, c, f and μMxyFor the parameter of setting.
It is preferred that, experimental calibration parameter index：A=1/2, b=1/256, c=32/1000000, f=1/4, μMxy=0.
It is preferred that, the target map, motion vector and present image according to previous moment, while considering the fortune in adjacent block
The similitude of the texture relation in dynamic vector and consecutive image come minimize current time target map and motion vector most
Little energy, specifically includes following steps：
01) all pieces of motion vector is obtained by block matching method, the original state V (t- of motion vector map is determined
1；T)=V0, V (t-1；T) moment t-1 to t, the motion vector of each block are represented；
02) each piece of candidate's label is set to the initial of target map by the original state according to motion vector map
State X (t)=y0；
03) estimate that target map and motion vector map are carrying out X (t)=yi and V (t-1；T) it is total after=Vi iteration
Energy, i are iterations；
04) target map X (t)=yi and motion vector map V (t-1 of the random conversion simultaneously under current state；T)=
Target map X (t+1)=yi+1 and motion vector map V (t of the Vi to NextState；T+1)=Vi+1；
05) in step 03) and 04) between iterative cycles operation, until X (t) and V (t-1；T) converge to energy function value
UstmrfTo minimum.
It is preferred that, in step 3, braking distance S of the electric bicycle with automobile in next two field picture is：
In formula, ξ is speed；It is attachment coefficient；ψ is road longitudinal grade degree, and % goes up a slope as just, descending is negative.
It is preferred that, in step 4, driver's conflict judgment criterion is divided into 4 grades of conflicts：1 grade, behavior of making a dash across the red light；It is 2 grades, slight to rush
It is prominent；3 grades, dangerous conflict；4 grades, Serious conflicts or there is accident, criterion difference is as follows：
When intersection is amber light, and the two-dimensional coordinate of electric bicycle is fallen in the coordinate range of intersection, then sentence
It is set to 1 grade of behavior of making a dash across the red light；
When intersection is red light, first, by obtaining in vehicle with car tracing to the electric bicycle in image
Heart X, Y-coordinate predict the braking distance of electric bicycle and car with formula 2；Then, estimate two cars in next two field picture
In stop, and obtain the distance between two cars stop；Finally, the stop of two cars is marked at into two-dimensional coordinate axle
On, and be mapped on real road：
If the extended line of the stop of two cars is non-intersect, it is judged to 2 grades of slight conflict；
If the distance between stop of two cars is not in safe range and when the two cars being marked on coordinate
When movement locus intersects, then it is judged to 3 grades of dangerous conflict；
If the distance between stop of two cars is less than or equal to 0, it is judged to 4 grades of Serious conflicts or thing occurs
Therefore.
The invention has the beneficial effects as follows：
(1) a kind of ST-MRF models based on optimization are proposed while tracking the algorithm of electric bicycle and automobile.The party
Method can not only optimize the spatial distribution of image, moreover it is possible to optimize distribution of the image sequence along time shaft, improve former vehicle with
The shortcoming of track algorithm poor robustness in the case of the intersection vehicles serious shielding of traffic disturbance.
(2) traffic conflict detection method of the invention is applied to the intersection of various geometries, is rotated by coordinate
3-D view is converted into 2-D data by algorithm, and (electric bicycle is little for the vehicle in calculating crossing intersection part different motion direction
Automobile) movement locus and two-dimensional coordinate, and experimental data is compared with commercialization data measured by Autoscope softwares,
Show that the error of the traffic data obtained using the technology is less, precision comparison is high.
(3) according to driver's conflict judgment criterion, set up the levels of conflict of 4 grades：1 grade, behavior of making a dash across the red light；2 grades, slightly
Conflict；3 grades, dangerous conflict；4 grades, Serious conflicts (accident), and level Four conflict is detected by Success in Experiment.
(4) when the electric bicycle for making a dash across the red light and car occur 3 grades to conflict, this method energy success prediction is potentially handed over
Interpreter's event, can provide real-time, accurate, reliable transport information for vehicle supervision department, it is to avoid traffic accident occurs.
Description of the drawings
Fig. 1 is label flow chart of the present invention；
Fig. 2 is coordinate rotation schematic diagram；
Fig. 3 is vehicle label analogous diagram of the present invention；
Fig. 4 is two grades of conflict analogous diagrams in present invention test；
Fig. 5 is three-level conflict analogous diagram in present invention test；
Fig. 6 is level Four conflict analogous diagram in present invention test.
Specific embodiment
Technical solution of the present invention is described in further detail with specific embodiment below in conjunction with the accompanying drawings, so that ability
The technical staff in domain can be better understood from the present invention and can be practiced, but illustrated embodiment is not as the limit to the present invention
It is fixed.
Electric bicycle based on ST-MRF models and mechanical transport collision detection method, comprise the steps：
Step 1, traffic camera shoot 3 d video images, track single electrical salf-walking simultaneously using ST-MRF models
Car and automobile, obtain the transport information of single electric bicycle and automobile, such as vehicle, the volume of traffic, speed and coordinate value etc..
In vehicle tracking algorithm, block of pixels and ST-MRF (Spatial Temporal Markov Random Field)
In scene be one-to-one, i.e., each block of pixels and a vehicle communication are got up, that is, give to each block of pixels
The label of one corresponding vehicle.When using vehicle tracking algorithm, an initial label distribution, Ran Houtong are determined first
The energy function distribution label crossed in ST-MRF is refined and optimization processing.In extractive process, algorithm considers block in the time
Domain and the contact of space (two-dimentional x-y), that is, contact and the connection between adjacent block of the block between consecutive image
System, by ST-MRF models, is assigned to block of pixels label.One scene of block correspondence in ST-MR models, block are numbered as mesh
The part in mark region is different with the texture of background image, and these blocks flock together and are referred to as target map, on each block
There is the motion vector for representing block, ST-MRF models are exactly the target that the next moment is estimated according to the target map of previous moment
Map, symbolically are as follows：
G (t-1)=g, G (t)=h：Gray values of the image G in moment t-1 is g, and in moment t, gray value is h.It is right
In each pixel, can be expressed as：G(t-1；α, β)=g (α, β), G (t；α, β)=h (α, β), in formula, α, β are that pixel is sat
Mark, G (t-1；α, β)=g (α, β) is meant that：In image G, coordinate is α, and the gray value of the pixel of β in moment t-1 is equal to figure
As coordinate is α in g, the corresponding gray value of pixel of β, G (t；α, β)=h (α, β) is meant that：In image G, coordinate is α, β's
Value of the pixel in moment t is α equal to coordinate in image h, the corresponding value of pixel of β.
X (t-1)=x, X (t)=y：Target map X is detected label in moment t-1 and is distributed as x, the quilt in moment t
The label for detecting is distributed as y.For each block of pixels, can be expressed as：Xk(t-1)=xk,Xk(t)=yk, in formula, k is
The numbering of block, Xk(t-1)=xkIt is meant that：In target map X, label distribution of k-th block of pixels in moment t-1 is x
In k-th block of pixels label；Xk(t)=ykIt is meant that：Label of k-th block of pixels in moment t point in target map X
Cloth is the label of k-th block of pixels in y.
V(t-1；T)=v：Moment t-1 to t, the motion vector of all pixels block are estimated as v.In this case, often
The motion vector of individual block can be expressed as：Vk=(t-1；T)=vk：Moment t-1 to t, the motion vector of k-th block of pixels is vk。
By ST-MRF models, previous moment image G (t-1)=g, present image G (t)=h, and previous moment are given
Target map X (t-1)=x, can determine target map X (t)=y and motion vector map V (t- simultaneously with maximum a posteriori probability
1；T)=V.Therefore, the problem of optimization aim map and motion vector map has reformed into the energy function of minimum formula (1)：
UstmrfIt is the energy function of ST-MRF models, can be minimized by relaxed algorithm, in formula：Part I a (Nyk-
μNy)2Represent in target map, the neighbouring relations of label, Part II b (Mxyk-μMxy)2Represent in successive objective map
The hiding relation of label, Part IIITexture relation in the expression consecutive image relevant with motion vector, Part IVRepresent the neighbouring relations of the motion vector in motion vector map.
Nyk：Refer to that the adjacent pixel blocks and the block of pixels of a block of pixels have the number of identical label；
Nxk：Represent the number of the adjacent pixel blocks of a block of pixels；
Dxyk：Texture correlation between the image of the image and t that represent the t-1 moment, when blocking, respectively
Calculating belongs to the probability of each car, make block of pixels time t coordinate value be [x (t), y (t)], motion vector be [u (t), v
(t)], in the coordinate value of time t-1 be：[x (t-1), y (t-1)]=[x (t)-u (t), y (t)-v (t)], when the figure of former frame
The vehicle label O as inMMotion vector beWhen：
G(t；α,β)；Gray-scale intensity of the pixel (α, β) in time t, each block of pixels is made up of 8 × 8 pixels, so 0
≤dα≤8,0≤dβ≤8；
Mxyk：The number of pixels of shield portions in two block of pixels of partial occlusion；
μNy：Neighborhood group, if using 8- neighborhood groups, μ Ny=8 is maximum；
Ck：Current pixel block；
Bk：Adjacent pixel blocks；
In the difference of the motion vector of (t-1) moment, current pixel block and adjacent pixel blocks；
A, b, c, f and μMxyFor the parameter of setting.It is preferred that, experimental calibration parameter index：A=1/2, b=1/256, c=32/
1000000, f=1/4, μMxy=0.
ST-MRF models can be while segmentation object border and motion vector.The process of its optimization is according to previous moment
Target map, motion vector and present image, while the texture in considering the motion vector and consecutive image in adjacent block is closed
The similitude of system specifically includes following steps minimizing the least energy of the target map and motion vector at current time：
01) motion vector of all pixels block is obtained by simple block matching method, it is then determined that motion vector map
Original state V (t-1；T)=V0, V (t-1；T) moment t-1 to t, the motion vector of all pixels block are represented；
02) candidate's label of each block of pixels is set to target map by the original state according to motion vector map
Original state X (t)=y0.In this process, in single target area, single label of being selected is named to block of pixels, so
And more than one is selected label and gives block of pixels name around the Ouluding boundary, there is no candidate's label to block of pixels in other places
Name.
03) estimate that target map and motion vector map are carrying out X (t)=yi and V (t-1；T) it is total after=Vi iteration
Energy, i are iterations；
04) for searching target map and the optimal result of motion vector map, random conversion simultaneously is under current state
Target map X (t)=yi and motion vector map V (t-1；T) target map X (the t+1)=yi+1s of=Vi to NextState
With motion vector map V (t；T+1)=Vi+1；
05) in step 03) and 04) between iterative cycles operation, until X (t) and V (t-1；T) converge to energy function value
UstmrfTo minimum.
Finally, by the belongingness of the minimum decision block of energy function, energy function is less, CkBlock is more likely to belong to
Corresponding vehicle, according to the differentiation electric bicycle of different sizes and car of image block.
In vehicle label procedure, compare the average gray and threshold size of block on detection rectangle frame, if the gray scale of block of pixels
More than threshold value, illustrate with the presence of vehicle, and give vehicle label, in order to verify vehicle tracking algorithm, vehicle label starts to tire out from 1
Meter.After vehicle enters detection rectangle frame, need to be tracked vehicle.Track algorithm based on ST-MRF models is along time sequence
Row automatically update the shape of vehicle.For this renewal, algorithm estimates motion vector in the block of vehicle region.In each picture
On plain block, motion vector of each target pixel block in consecutive frame is estimated with the matching process of block.
Detection calculating and picture frame of the block of pixels detected on rectangle frame to vehicle are can be seen that by Fig. 1 label flow charts
Between the calculating of motion vector As time goes on run.That is, when detection vehicle presence on t time detecting rectangle frames
Afterwards, the initial label of vehicle is obtained, and then tracking module begins to block of pixels be calculated corresponding to consecutive frame with block matching method
Motion vector, and to detect rectangle frame on initial vehicle label shift, obtain by detect rectangle frame after sequence chart
The vehicle label of picture, as shown in Figure 3.
Image block-matching technique is by the initial vehicle label obtained in detection module and the car for coming into detection zone
Label is moved, so as to moving vehicle image sequence is converted into vehicle label sequence.The label sequence is whole video
Important Floor layer Technology in event detection technology, the solution to block a difficult problem provide technology platform, while label sequence is traffic
The extraction of parameter provides data source.
Step 2, the 3 d video images that traffic camera shoots are converted into by two-dimensional coordinate number by coordinates transformation method
According to.
As the intersection obtained from video camera is indirect angle, it is difficult to obtain the speed in addition to the volume of traffic, vehicle
The aspect such as two-dimensional coordinate information, so as to the more difficult reaction braking calculated between the electric bicycle and car that make a dash across the red light away from
From, it is impossible to correct judgement is given to the order of severity of traffic conflict.In order to obtain actual road speed, coordinate value, need by
3-D view is converted into 2-D data coordinate, as shown in Fig. 2 specifically including following steps：
01) set up global coordinates system (XG,YG), come from video camera and the reference axis (X based on imageM,YM)；
02) calculate by intersection polygon detecting region (x1,y1), (x2,y2), (x3,y3) and (x4,y4) extended line
(p1,q1) and (p2,q2)；
03) derive rotational alignment W1, with (x3,y3) for round dot, from (p2,q2) rotation-θ angles obtain W1；
04) straight line (p1,q1)→(x1,y1) extended line hand over W1For (x5,y5), point (x5,y5) and point (x3,y3) between appoint look for
A bit (xn,yn), then calculate point (xn,yn) to point (p1,q1) linear equation；
05) calculate straight line (p2,q2)→(x3,y3), (p2,q2)→(x4,y4) intersection point (p2,q2)；
06) in the same manner 03), derive rotational alignment W2, with (x3,y3) for round dot, from (p1,q1) rotation θ angles obtain W2；
07) in straight line W2Take up an official post and look for a bit (xm,ym), then calculate point (xm,ym) to point (p2,q2) linear equation；
08) calculate straight line (p1,q1)→(x1,y1), (p1,q1)→(x3,y3) intersection point (p1,q1), straight line (p2,q2)→
(x3,y3), (p2,q2)→(x4,y4) intersection point (p2,q2)。
Step 3, the prediction stop of electric bicycle and automobile in next two field picture.
The research of most of conventional traffic conflicts is to carry out conflict set recognition by employing observer, and observer verifies in person and works as
Front direction have electric bicycle make a dash across the red light phenomenon when, whether the car of other direction normally travel has avlidance behavior, then right
Conflict after checking is counted.However, this method depends on the judgement of people, it is impossible to take into full account accident with conflict
The order of severity.Additionally, former judgement method for confliction detection only counts the electric bicycle for making a dash across the red light, and do not consider it is other because
Reaction braking distance between element, such as vehicle (electric bicycle and car), so as to reduce event detection precision.Cause
This, conflict spectrum is divided into 4 etc. according to reaction braking distance in order to set up more accurately driver's conflict judgment criterion by the present invention
Level：1 grade, behavior of making a dash across the red light；It is 2 grades, slight to conflict；3 grades, dangerous conflict；4 grades, Serious conflicts or there is accident, this result of study
Real-time, accurate, reliable transport information can be provided for vehicle supervision department, be easy to process event in time.Carry out below
It is discussed in detail.
Single electric bicycle and automobile are tracked simultaneously using ST-MRF models, single electric bicycle and vapour is obtained
After the speed and coordinate value of car, the braking distance of each car can be computed.The calculating parameter of braking distance includes：Calculate
The top rake of speed, attachment coefficient and road.Wherein, the top rake of road can be by extending former frame (t-1) and present frame
T the slope of () is obtained；Coefficient of road adhesion is referred to《Highway technical standard》, vehicle next frame can be estimated according to formula (2)
(t+1) braking distance S：
In formula, ξ is speed；It is attachment coefficient, good bituminous paving value is 1, good cement concrete pavement value
For 0.7, wet road surface value is 0.5；ψ is road longitudinal grade degree, and % goes up a slope as just, descending is negative.
Step 4, set up according to the distance between stop driver conflict judgment criterion.
Driver's conflict judgment criterion is divided into by 4 grades of conflicts according to the braking distance of vehicle：1 grade, behavior of making a dash across the red light；2 grades,
It is slight to conflict；3 grades, dangerous conflict；4 grades, Serious conflicts or there is accident, criterion difference is as follows：
1 grade：Make a dash across the red light behavior
The conflict of intersection is solved using traditional method, when intersection or amber light, electric bicycle drives
Member does not observe traffic rules and regulations harum-scarum into intersection, and this kind of behavior is considered as the behavior of making a dash across the red light and is counted.Namely work as crossroad
When mouth is amber light, and the two-dimensional coordinate of electric bicycle is fallen in the coordinate range of intersection, then be judged to 1 grade of the row that makes a dash across the red light
For.
Can be (relevant with time and signal) next using the two-dimensional coordinate of intersection image tracking method acquisition each car
The number of times that makes a dash across the red light of statistics, the number of times of each car turnover intersection with signal time compared with, then statistics make a dash across the red light it is secondary
Number.
2 grades：It is slight to conflict
The electric bicycle of such as east-west direction traveling makes a dash across the red light into after intersection, the car of North and South direction traveling
Into intersection (now North and South direction is green light, and east-west direction is red light), conflict now can be electronic certainly by estimation
Driving is judged with the distance between the stop of car.First, by obtaining in vehicle to the vehicle tracking in image
Heart X, Y-coordinate predict the braking distance of electric bicycle and car with formula 2；Then, estimate two cars in next two field picture
In stop, and obtain the distance between two cars stop；Finally, the stop of two cars is marked at into two-dimensional coordinate axle
On, and be reflected on real road, to recognize stop of the two cars on two-dimensional coordinate.If the stop of two cars
Extended line is non-intersect, it is possible to be judged as 2 grades of slight conflict.That is, the stop of two cars mutually disjoints, it is this kind of
Conflict is considered as slight conflict；If the extended line of the stop of two cars intersects, then two cars must be clashed, punching
Prominent degree of danger can be given and be evaluated according to 3 grades of conflicts and 4 grades of conflict criterions.
3 grades：Dangerous conflict
3 grades of conflicts be the distance between stop when two cars less than safe distance and the two cars that estimate with
The situation of track intersection of locus.In order to the conflict judgment criterion studied under 3 grades of conflicts need to provide the judgement standard of quantization to slight conflict
Then, if whether the distance between two car stops are in safe range.If the distance between two car stops not in safe range and
It is considered as 3 grades of dangerous conflict when the movement locus of the two cars being marked on coordinate is intersecting.
4 grades：Serious conflicts
On the basis of tracking vehicle movement track obtains the coordinate of two cars, when the rectangular area part for representing two cars
During overlap, it is traffic accident to be considered as this conflict.Specifically, if the distance between stop of two cars is less than or equal to 0,
Then it is judged to 4 grades of Serious conflicts or accident occurs.
Experimental result and analysis：
The video that this experiment shoots is the morning 9 on November 14th, 2015：30 are handed over Liberation Road at Huai'an Huai-Hai South Road
One section of video of prong northing mouth, can obtain the traffic parameters such as the volume of traffic and speed by the vehicle tracking technology of the present invention,
Data of the data of collection with the Autoscope softwares measurement of commercialization are compared, as shown in table 1.As can be seen from the table
Difference between the volume of traffic measured by the volume of traffic, speed and Autoscope softwares that obtains under video environment, speed less,
Illustrate that the error of traffic data measured by the vehicle tracking technology is less, precision comparison is high.Therefore, vehicle movement track skill
The traffic conflict detection and accident that art can be used between the electric bicycle for making a dash across the red light intersection and car judges.
1 video data of table and Autoscope Data Comparisons
Traffic conflict test experience：On the basis of tracking vehicle movement track obtains coordinate, with Huai-Hai South Road and liberation
As a example by the intersection of road, application carries out traffic conflict detection to captured video, when the electric bicycle for making a dash across the red light and normal row
When intersection experiences traffic conflict, system passes through the braking distance for predicting electric bicycle and car to the car sailed,
Determine stop of the two cars in next frame, and obtain the distance between stop, driver's conflict is set up according to this distance and is sentenced
Disconnected criterion determines conflict grade, and as a result as shown in table 2, levels of conflict is that space representative does not have electric bicycle to make a dash across the red light.
2 traffic conflict testing result of table
| Time | Vehicle ID number | Vehicle center X-coordinate | Vehicle center Y-coordinate | Levels of conflict | Speed (km/h) |
| 10:15 | 012 | 167.0 | 282.4 | 1 | 38.3 |
| 10:16 | 013 | 155.8 | 59.7 | 43.5 | |
| 10:17 | 014 | 227.2 | 97.0 | 3 | 32.5 |
| 10:18 | 016 | 163.3 | 288.4 | 1 | 38.3 |
| 10:19 | 017 | 145.6 | 60.4 | 43.4 | |
| 10:20 | 020 | 220.2 | 96.6 | 2 | 32.5 |
| 10:21 | 024 | 159.6 | 294.6 | 38.4 | |
| 10:22 | 028 | 134.5 | 60.4 | 3 | 43.4 |
| 10:23 | 030 | 211.3 | 95.0 | 1 | 32.6 |
| 10:24 | 031 | 155.3 | 301.4 | 2 | 38.5 |
| 10:25 | 032 | 122.8 | 60.9 | 43.5 | |
| 10:26 | 036 | 201.3 | 95.0 | 1 | 32.5 |
| 10:27 | 038 | 152.2 | 307.0 | 2 | 38.3 |
| 10:28 | 040 | 110.7 | 61.3 | 4 | 43.4 |
| 10:29 | 044 | 190.9 | 93.9 | 32.5 |
Reached a conclusion by experiment：In one-level consequences of hostilities, record has 10 electric bicycles to make a dash across the red light, in 10 cars,
6 entrance, two grades of conflicts, 3 records in three-level conflict finally have 1 car to reach level Four levels of conflict, traffic will occur
Accident.Fig. 4,5,6 show the traffic conflict situation that tracking system is detected, wherein, Fig. 4 shows 2 grades of conflict situations, and No. ID is
But 27 electric bicycle violates traffic signals does not have other vehicles；Fig. 5 shows 3 grades of conflict situations, and this conflict occurs
No. ID for 27 electric bicycle and No. ID for 28 car between, this kind of situation, No. ID for 27 electric bicycle violate
Traffic signals, the stop of the two cars being marked on coordinate intersects, but the rectangular area for representing two cars is not sent out
It is raw to overlap；Finally, Fig. 6 show level Four conflict situations, this conflict occur No. ID for 11 electric bicycle with No. ID be 9 it is little
Between automobile, the rectangular area of this interval scale two cars partly overlaps, it is believed that this conflict will occur traffic accident.
The invention has the beneficial effects as follows：
(1) a kind of ST-MRF models based on optimization are proposed while tracking the algorithm of electric bicycle and automobile.The party
Method can not only optimize the spatial distribution of image, moreover it is possible to optimize distribution of the image sequence along time shaft, improve former vehicle with
The shortcoming of track algorithm poor robustness in the case of the intersection vehicles serious shielding of traffic disturbance.
(2) traffic conflict detection method of the invention is applied to the intersection of various geometries, is rotated by coordinate
3-D view is converted into 2-D data by algorithm, and (electric bicycle is little for the vehicle in calculating crossing intersection part different motion direction
Automobile) movement locus and two-dimensional coordinate, and experimental data is compared with commercialization data measured by Autoscope softwares,
Show that the error of the traffic data obtained using the technology is less, precision comparison is high.
(3) according to driver's conflict judgment criterion, set up the levels of conflict of 4 grades：1 grade, behavior of making a dash across the red light；2 grades, slightly
Conflict；3 grades, dangerous conflict；4 grades, Serious conflicts (accident), and level Four conflict is detected by Success in Experiment.
(4) when the electric bicycle for making a dash across the red light and car occur 3 grades to conflict, this method energy success prediction is potentially handed over
Interpreter's event, can provide real-time, accurate, reliable transport information for vehicle supervision department, it is to avoid traffic accident occurs.
The preferred embodiments of the present invention are these are only, the scope of the claims of the present invention is not thereby limited, it is every using this
Equivalent structure or equivalent flow conversion that bright specification and accompanying drawing content are made, or directly or indirectly it is used in other correlations
Technical field, be included within the scope of the present invention.
Claims (6)
1. the electric bicycle based on ST-MRF models and mechanical transport collision detection method, it is characterised in that including following step
Suddenly：
Step 1, traffic camera shoot 3 d video images, using ST-MRF models simultaneously track single electric bicycle with
Automobile, obtains the transport information of single electric bicycle and automobile；
Step 2, the 3 d video images that traffic camera shoots are converted into by two-dimensional coordinate data by coordinates transformation method；
Step 3, the prediction stop of electric bicycle and automobile in next two field picture；
Step 4, set up according to the distance between stop driver conflict judgment criterion.
2. the electric bicycle based on ST-MRF models according to claim 1 and mechanical transport collision detection method, its
It is characterised by, in step 1, the energy function U of ST-MRF modelsstmrfIt is：
In formula：Part I a (Nyk-μNy)2Represent in target map, the neighbouring relations of label, Part II b (Mxyk-μMxy)2
Represent the hiding relation of the label in successive objective map, Part IIIRepresent the sequential chart relevant with motion vector
Texture relation as in, Part IVRepresent the motion arrow in motion vector map
The neighbouring relations of amount；
Nyk：Refer to that the adjacent pixel blocks and the block of pixels of a block of pixels have the number of identical label；
Nxk：Represent the number of the adjacent pixel blocks of a block of pixels；
Dxyk：Texture correlation between the image of the image and t that represent the t-1 moment, when blocking, is calculated respectively
Belong to the probability of each car；
Mxyk：The number of pixels of shield portions in two block of pixels of partial occlusion；
μNy：Neighborhood group, if using 8- neighborhood groups, μ Ny=8 is maximum；
Ck：Current pixel block；
Bk：Adjacent pixel blocks；
In the difference of the motion vector of (t-1) moment, current pixel block and adjacent pixel blocks；
A, b, c, f and μMxyFor the parameter of setting.
3. the electric bicycle based on ST-MRF models according to claim 2 and mechanical transport collision detection method, its
It is characterised by, the target map, motion vector and present image according to previous moment, while considering the fortune in adjacent pixel blocks
The similitude of the texture relation in dynamic vector and consecutive image come minimize current time target map and motion vector most
Little energy, specifically includes following steps：
01) motion vector of all pixels block is obtained by block matching method, the original state V (t- of motion vector map is determined
1；T)=V0, V (t-1；T) moment t-1 to t, the motion vector of each block of pixels are represented；
02) candidate's label of each block of pixels is set to the initial of target map by the original state according to motion vector map
State X (t)=y0；
03) estimate that target map and motion vector map are carrying out X (t)=yi and V (t-1；T) total energy after=Vi iteration
Amount, i is iterations；
04) target map X (t)=yi and motion vector map V (t-1 of the random conversion simultaneously under current state；T)=Vi is arrived
Target map X (the t+1)=yi+1 and motion vector map V (t of NextState；T+1)=Vi+1；
05) in step 03) and 04) between iterative cycles operation, until X (t) and V (t-1；T) converge to energy function value Ustmrf
To minimum.
4. the electric bicycle based on ST-MRF models according to claim 1 and mechanical transport collision detection method, its
It is characterised by, in step 3, braking distance S of the electric bicycle with automobile in next two field picture is：
In formula, ξ is speed；It is attachment coefficient；ψ is road longitudinal grade degree, and % goes up a slope as just, descending is negative.
5. the electric bicycle based on ST-MRF models according to claim 1 and mechanical transport collision detection method, its
It is characterised by, in step 4, driver's conflict judgment criterion is divided into 4 grades of conflicts：1 grade, behavior of making a dash across the red light；It is 2 grades, slight to conflict；3
Level, dangerous conflict；4 grades, Serious conflicts or there is accident, criterion difference is as follows：
When intersection is amber light, and the two-dimensional coordinate of electric bicycle is fallen in the coordinate range of intersection, then be judged to 1
The behavior of making a dash across the red light of level；
When intersection is red light, first, by obtaining vehicle center X, Y with car tracing to the electric bicycle in image
Coordinate, predicts the braking distance of electric bicycle and car with formula 2；Then, estimate two cars in next two field picture
Stop, and obtain the distance between two cars stop；Finally, the stop of two cars is marked on two-dimensional coordinate axle,
And be mapped on real road：
If the extended line of the stop of two cars is non-intersect, it is judged to 2 grades of slight conflict；
If the distance between stop of two cars is not in safe range and when the motion of the two cars being marked on coordinate
During intersection of locus, then it is judged to 3 grades of dangerous conflict；
If the distance between stop of two cars is less than or equal to 0, it is judged to 4 grades of Serious conflicts or accident occurs.
6. the electric bicycle based on ST-MRF models according to claim 2 and mechanical transport collision detection method, its
It is characterised by, a=1/2, b=1/256, c=32/1000000, f=1/4, μMxy=0.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611030277.4A CN106530825B (en) | 2016-11-16 | 2016-11-16 | Electric bicycle and mechanical transport collision detection method based on ST-MRF model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611030277.4A CN106530825B (en) | 2016-11-16 | 2016-11-16 | Electric bicycle and mechanical transport collision detection method based on ST-MRF model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106530825A true CN106530825A (en) | 2017-03-22 |
| CN106530825B CN106530825B (en) | 2019-01-22 |
Family
ID=58356105
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611030277.4A Active CN106530825B (en) | 2016-11-16 | 2016-11-16 | Electric bicycle and mechanical transport collision detection method based on ST-MRF model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106530825B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107590462A (en) * | 2017-09-13 | 2018-01-16 | 淮阴工学院 | The traffic flow parameter extracting method being combined based on pattern-recognition with ST MRF |
| CN107808549A (en) * | 2017-11-08 | 2018-03-16 | 华为技术有限公司 | The methods, devices and systems of mobile control |
| CN110809792A (en) * | 2017-06-29 | 2020-02-18 | 株式会社电装 | Collision estimation device and collision estimation method |
| CN111243274A (en) * | 2020-01-20 | 2020-06-05 | 陈俊言 | Road collision early warning system and method for non-internet traffic individuals |
| CN112368757A (en) * | 2018-07-04 | 2021-02-12 | 日产自动车株式会社 | Driving assistance method and driving assistance device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008070955A (en) * | 2006-09-12 | 2008-03-27 | Sumitomo Electric Ind Ltd | Display system for displaying traveling object, on-vehicle apparatus, image transmitter and display method |
| CN103942359A (en) * | 2014-03-13 | 2014-07-23 | 华南理工大学 | Modeling method for simulation model of bicycle conflict avoidance behaviors at non-signalized intersection |
| CN104050837A (en) * | 2014-06-10 | 2014-09-17 | 奇瑞汽车股份有限公司 | Driving safety prompting method and device |
| US20150003670A1 (en) * | 2013-06-28 | 2015-01-01 | Bendix Commercial Vehicle Systems Llc | Measuring the range to an object, in an image, using size categorization |
| CN104376725A (en) * | 2014-11-28 | 2015-02-25 | 东南大学 | Signalized intersection non-motor vehicle lane channelizing control method under expansion effect |
| CN105761547A (en) * | 2016-03-28 | 2016-07-13 | 安徽云森物联网科技有限公司 | Traffic collision pre-warning technique and system based on images |
-
2016
- 2016-11-16 CN CN201611030277.4A patent/CN106530825B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008070955A (en) * | 2006-09-12 | 2008-03-27 | Sumitomo Electric Ind Ltd | Display system for displaying traveling object, on-vehicle apparatus, image transmitter and display method |
| US20150003670A1 (en) * | 2013-06-28 | 2015-01-01 | Bendix Commercial Vehicle Systems Llc | Measuring the range to an object, in an image, using size categorization |
| CN103942359A (en) * | 2014-03-13 | 2014-07-23 | 华南理工大学 | Modeling method for simulation model of bicycle conflict avoidance behaviors at non-signalized intersection |
| CN104050837A (en) * | 2014-06-10 | 2014-09-17 | 奇瑞汽车股份有限公司 | Driving safety prompting method and device |
| CN104376725A (en) * | 2014-11-28 | 2015-02-25 | 东南大学 | Signalized intersection non-motor vehicle lane channelizing control method under expansion effect |
| CN105761547A (en) * | 2016-03-28 | 2016-07-13 | 安徽云森物联网科技有限公司 | Traffic collision pre-warning technique and system based on images |
Non-Patent Citations (4)
| Title |
|---|
| 周君等: "基于ST-MRF的自适应车辆跟踪算法研究", 《交通运输系统工程与信息》 * |
| 周君等: "基于反向ST-MRF模型的车辆遮挡分割算法", 《公路交通科技》 * |
| 汽车驾驶员新登记标准教材委员会: "《高级汽车驾驶员培训教材上册》", 31 October 1994 * |
| 王小凤: "电动自行车交通冲突分析与交通安全评价", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110809792A (en) * | 2017-06-29 | 2020-02-18 | 株式会社电装 | Collision estimation device and collision estimation method |
| CN107590462A (en) * | 2017-09-13 | 2018-01-16 | 淮阴工学院 | The traffic flow parameter extracting method being combined based on pattern-recognition with ST MRF |
| CN107590462B (en) * | 2017-09-13 | 2020-11-20 | 淮阴工学院 | Traffic flow parameter extraction method based on combination of pattern recognition and ST-MRF |
| CN107808549A (en) * | 2017-11-08 | 2018-03-16 | 华为技术有限公司 | The methods, devices and systems of mobile control |
| CN107808549B (en) * | 2017-11-08 | 2020-06-16 | 华为技术有限公司 | Method, device and system for mobile control |
| CN112368757A (en) * | 2018-07-04 | 2021-02-12 | 日产自动车株式会社 | Driving assistance method and driving assistance device |
| CN112368757B (en) * | 2018-07-04 | 2022-11-01 | 日产自动车株式会社 | Driving assistance method and driving assistance device |
| CN111243274A (en) * | 2020-01-20 | 2020-06-05 | 陈俊言 | Road collision early warning system and method for non-internet traffic individuals |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106530825B (en) | 2019-01-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106530825B (en) | Electric bicycle and mechanical transport collision detection method based on ST-MRF model | |
| CN105512720B (en) | A kind of public transit vehicle passenger flow statistics method and system | |
| KR100201739B1 (en) | Method for observing an object, apparatus for observing an object using said method, apparatus for measuring traffic flow and apparatus for observing a parking lot | |
| US9292750B2 (en) | Method and apparatus for detecting traffic monitoring video | |
| CN103593678B (en) | A kind of long-span bridge vehicle dynamic load distribution detection method | |
| CN108596129A (en) | A kind of vehicle based on intelligent video analysis technology gets over line detecting method | |
| WO2022206942A1 (en) | Laser radar point cloud dynamic segmentation and fusion method based on driving safety risk field | |
| Zhangyu et al. | A camera and LiDAR data fusion method for railway object detection | |
| CN106951879A (en) | Multiple features fusion vehicle checking method based on camera and millimetre-wave radar | |
| Goyat et al. | Vehicle trajectories evaluation by static video sensors | |
| CN101105893A (en) | Automobile video frequency discrimination speed-testing method | |
| CN102592454A (en) | Intersection vehicle movement parameter measuring method based on detection of vehicle side face and road intersection line | |
| CN105260719A (en) | Railway platform line-crossing detection method | |
| CN103325255A (en) | Regional traffic condition detection method based on photogrammetric technology | |
| CN107909012B (en) | Real-time vehicle tracking detection method and device based on disparity map | |
| Feng et al. | Mixed road user trajectory extraction from moving aerial videos based on convolution neural network detection | |
| CN110379168A (en) | A kind of vehicular traffic information acquisition method based on Mask R-CNN | |
| CN110298300A (en) | A method of detection vehicle violation crimping | |
| CN104183142A (en) | Traffic flow statistics method based on image visual processing technology | |
| CN109791607A (en) | It is detected from a series of images of video camera by homography matrix and identifying object | |
| CN103794050A (en) | Real-time transport vehicle detecting and tracking method | |
| CN113759391A (en) | Passable area detection method based on laser radar | |
| Jwa et al. | Kalman filter based railway tracking from mobile Lidar data | |
| CN111160132B (en) | Method and device for determining lane where obstacle is located, electronic equipment and storage medium | |
| CN106951820B (en) | Passenger flow statistical method based on annular template and ellipse fitting |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 223005 8th Floor, Andong Building, No. 10 Haian Road, Lianshui County, Huaian City, Jiangsu Province Applicant after: Huaijin Polytechnical College Address before: 223005 Jiangsu Huaian Higher Education Park 1 East Road Applicant before: Huaijin Polytechnical College |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200831 Address after: No.12, Keji Road, higher education park, Huaian Economic and Technological Development Zone, Jiangsu Province Patentee after: JIANGSU TS TRAFFIC DESIGN & RESEARCH INSTITUTE Co.,Ltd. Address before: 223005 8th Floor, Andong Building, No. 10 Haian Road, Lianshui County, Huaian City, Jiangsu Province Patentee before: HUAIYIN INSTITUTE OF TECHNOLOGY |
|
| TR01 | Transfer of patent right | ||
| CP02 | Change in the address of a patent holder |
Address after: 210019 No. 8 East Street, Fuchunjiang, Jianye District, Nanjing City, Jiangsu Province Patentee after: JIANGSU TS TRAFFIC DESIGN & RESEARCH INSTITUTE Co.,Ltd. Address before: 223243 No.12, Keji Road, higher education park, Huaian Economic and Technological Development Zone, Jiangsu Province Patentee before: JIANGSU TS TRAFFIC DESIGN & RESEARCH INSTITUTE Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |