CN107481519B - Automatic traffic accident identification method - Google Patents
Automatic traffic accident identification method Download PDFInfo
- Publication number
- CN107481519B CN107481519B CN201710595902.8A CN201710595902A CN107481519B CN 107481519 B CN107481519 B CN 107481519B CN 201710595902 A CN201710595902 A CN 201710595902A CN 107481519 B CN107481519 B CN 107481519B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- vehicles
- time
- main vehicle
- sensing range
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/065—Traffic control systems for road vehicles by counting the vehicles in a section of the road or in a parking area, i.e. comparing incoming count with outgoing count
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0133—Traffic data processing for classifying traffic situation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses an automatic traffic accident recognition method, which can automatically distinguish vehicles running in the same lane and opposite lanes by skillfully utilizing a simple set thought on the premise of not knowing information such as speed, acceleration, position and the like. And then, by combining the change characteristics of local traffic flow before and after the accident, the traffic accident which influences the normal traffic flow can be automatically identified without a large amount of historical data.
Description
Technical Field
The invention relates to the technical field of intelligent traffic, in particular to an automatic traffic accident identification method.
Background
With the rapid development of the transportation industry and the continuous increase of the automobile holding capacity, the automobile holding capacity in China reaches 1.94 hundred million by 2016 years according to the statistics of the administration of public security of China. The mileage of the Chinese highway is also increasing continuously, and the total mileage of the Chinese highway breaks through 13 kilometers by 2016. Therefore, the number of highway traffic accidents is increasing, and meanwhile, serious traffic jam and life and property loss are brought, and the traffic efficiency and the driving safety of roads are reduced. The world health organization predicts that traffic accidents will become the fifth leading cause of death in the near future.
A survey by Los Angeles shows that traffic congestion can last four minutes or even more for every one minute of delay in clearing a traffic accident. However, in many cases, if an accident is detected early and an alarm message is issued, the driver at the back has sufficient time to slow down or reselect the lane in advance, so that the accident is prevented from being involved, and further congestion of the road and secondary accidents are avoided. Therefore, the highway management system which is efficient, can provide reliable service for drivers and can prevent and identify traffic accidents as early as possible is significant.
Under the background, scholars and experts at home and abroad continuously search for a more reliable and faster traffic accident identification method.
As shown in fig. 1, a bayesian network-based trunk road traffic accident identification algorithm is proposed for k.zhang, m.a.p Taylor, and the algorithm utilizes the bayesian network to quantitatively simulate cause-effect dependency between events and variables, and updates the possibility of occurrence of the events by bidirectional reasoning in the bayesian network at each detection time interval based on real-time traffic data. If this estimated probability is greater than a predetermined decision threshold, an accident alarm is generated. However, the average detection time of the scheme exceeds 120s, and the overlong detection time is not favorable for real-time detection.
As shown in fig. 2, a video-based real-time traffic accident recognition method is proposed for Zuhui, xieyahua and the like, and the method utilizes a gaussian mixture model to extract a foreground and a background from a video screenshot to distinguish a vehicle, then utilizes a mean shift algorithm to track the vehicle, thereby obtaining three traffic parameters of a vehicle position, an acceleration and a course angle, comprehensively analyzes the three parameters, and judges whether an accident occurs or not. However, the method needs to collect video data, is easily affected by light of the use environment, has low reliability, needs to install video collection equipment along the way, and has high cost.
Disclosure of Invention
The invention aims to provide an automatic traffic accident recognition method, which only needs to acquire identification information of each vehicle under the condition of not acquiring traditional information such as position, speed, acceleration and the like, so that accidents influencing normal traffic of traffic flow can be quickly and accurately recognized.
The purpose of the invention is realized by the following technical scheme:
an automatic traffic accident recognition method includes:
arranging a short-distance communication module on each vehicle to enable each vehicle to read identification information of other vehicles in the sensing range of the vehicle;
the main vehicle periodically detects vehicles in the sensing range of the main vehicle;
determining a same-direction vehicle set and an opposite-direction vehicle set which leave the sensing range within a period of time according to the vehicle set which leaves the sensing range within the period of time;
and judging whether the ratio of the equidirectional vehicle set to the opposite vehicle set exceeds a threshold value, if so, judging that the traffic accident happens to the main vehicle or the equidirectional vehicle in front of the main vehicle if the ratio of the equidirectional vehicle set to the opposite vehicle set exceeds the threshold value and the traffic flow density detected by the main vehicle also belongs to an increasing state.
Let T be the detection period of the main vehicle, r be the radius of the induction range, and T be the time when the main vehicle starts to enter the working state0The set of vehicles detected at the moment is A0,t0The set of vehicles detected at + nT is AnAnd during normal driving satisfiesWherein V is the speed of the main vehicle during normal running,is the average speed of the co-current traffic,the average speed of the opposite traffic flow; then the set of vehicles traveling in the same direction during the kT time period is denoted as Sn=An-k∩AnWherein the vehicle set is An-kIs t0+ (n-k) the set of vehicles detected at time T.
Suppose from t0From time + nT to t0Time (n + q) T, the set of vehicles leaving the main vehicle sensing range during this time being
Then the set of co-directional vehicles that left the sensing range at this time is:
Ls(qT)=Ln(qT)∩Sn;
the set of oncoming vehicles that are out of the sensing range at this time period is:
wherein the content of the first and second substances,is not in the set An+qThe elements (A) and (B) in (B),is not in the set LsElements in (qT).
If L iss(qT) and Lr(qT) if the ratio is greater than the threshold, preliminarily judging that the main vehicle or the same-direction vehicle in front of the main vehicle has a traffic accident;
continuously judging t0Set A of vehicles detected at time n + q Tn+qAnd t0Set of vehicles A detected at time T +(n-k)n-kIf the ratio is larger than the set value, the traffic accident of the main vehicle or the same-direction vehicle in front of the main vehicle is finally judged.
According to the technical scheme provided by the invention, the vehicles running in the same lane and the opposite lane are automatically distinguished by skillfully utilizing a simple set thought on the premise of not knowing information such as speed, acceleration, position and the like. And then, by combining the change characteristics of local traffic flow before and after the accident, the traffic accident which influences the normal traffic flow can be automatically identified without a large amount of historical data.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a schematic diagram of a bayesian network-based main road traffic accident identification algorithm according to the background art of the present invention;
FIG. 2 is a flow chart of a video-based real-time traffic accident recognition method provided in the background of the present invention;
fig. 3 is an overall framework diagram of an automatic traffic accident recognition method according to an embodiment of the present invention;
FIG. 4 is an illustration set S provided by the embodiment of the present inventionnA schematic diagram of (a);
FIG. 5 is an illustration set L provided by an embodiment of the present inventions(qT) and LrSchematic of (qT).
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention provides an automatic traffic accident recognition method which is mainly based on an aggregation idea and judges which vehicles run on the same lane and which vehicles run on opposite lanes by an aggregation method on the premise of not needing to know information such as speed, acceleration, position and the like. Further judging the departing same-direction vehicle and the opposite-direction vehicle. After the accident happens, the speed of the accident vehicle is 0, and the traffic of the same-direction lane is influenced by the accident. The number Ls of vehicles leaving the detection area in unit time increases (the vehicles on the same-direction lane and the same-direction vehicles in the set are different concepts, and the same-direction vehicles in the set refer to the vehicles continuously detected by the host vehicle, so after the accident, because the relative speed of the vehicles in the opposite lane and the host vehicle is small, the vehicles in the opposite lane can be detected twice continuously and are regarded as the same-direction vehicles), and the number Lr of the opposite vehicles leaving the detection area in unit time decreases, so that the accident can obviously increase eta l/Lr; and judging whether the accident occurs or not by combining the fact that the traffic flow density is increased due to the fact that the traffic jam occurs in front of the accident car after the accident occurs.
Fig. 3 is a general frame diagram of the automatic traffic accident recognition method, which is specifically as follows:
the method comprises the following steps that a short-distance communication module is arranged on each vehicle, so that each vehicle can read identification information of other vehicles in a sensing range of the vehicle; the short-range communication module may be a WIFI or RFID module.
A vehicle is selected as a research object (called a main vehicle), the main vehicle periodically detects the vehicles in the sensing range of the main vehicle, and the detection period of the main vehicle is assumed to be T, and the radius of the sensing range is assumed to be r.
T at which the master starts to enter the operating state0The set of vehicles detected at the moment is A0,t0The set of vehicles detected at + nT is AnAnd during normal driving satisfiesWherein V is the speed of the main vehicle during normal running,is the average speed of the co-current traffic,is the average speed of the oncoming traffic. Then the set of vehicles traveling in the same direction during the kT time period is denoted as Sn=An-k∩AnWherein the vehicle set is An-kIs t0+ (n-k) the set of vehicles detected at time T.
The above principle is as follows: a formulaWherein 2r represents a distance, which represents that in normal running, if twice the sensing range is smaller than the relative movement distance between the oncoming vehicle and the host vehicle in the kT time and larger than the relative movement distance between the oncoming vehicle and the host vehicle in the kT time, then at t0+ nT and t0The cars detected at both times of + (n + k) T must be co-directional, and the oncoming car must not be likely to be at T0+ nT and t0Both times of + (n + k) T are detected, so the set S can be usedn=An-k∩AnIndicating a co-directional vehicle. After an accident, as the speed of the main vehicle is reduced, if the relative movement distance between the opposite vehicle and the main vehicle in the kT time is less than twice of the induction radius, the opposite vehicle can possibly move at t0+ nT and t0All of + (n + k) T are detected by the host, the result beingThose vehicles traveling in the opposite lane are included in the set S as vehicles traveling in the same directionnIn addition, the vehicles increase the number of vehicles in the same direction which leave the induction range within the qT time, wherein eta is Ls/LrThe numerator in (1) is larger and the denominator is smaller, which is consistent with the judgment basis.
Determining a set of vehicles in the same direction and a set of vehicles in opposite directions which leave the sensing range within a period of time according to the set of vehicles which leave the sensing range within the period of time: suppose from t0From time + nT to t0Time + n + q) T, the set of vehicles that leave the host vehicle sensing range during this time is:
wherein the content of the first and second substances,is not in the set An+qElements of (1), so set Ln(qT) means thatnIn and not in set An+qI.e. t0+ nT to t0Vehicles leaving the primary sensing range within time period of + (n + q) T.
Then the set of co-directional vehicles that left the sensing range at this time is:
Ls(qT)=Ln(qT)∩Sn;
the set of oncoming vehicles that are out of the sensing range at this time period is:
Judging whether the ratio of the equidirectional vehicle set to the opposite vehicle set exceeds a threshold value, if so, increasing the traffic flow density detected by the main vehicleJudging that the main vehicle or the vehicles in the same direction in front of the main vehicle have traffic accidents in the state; specifically, the method comprises the following steps: if L iss(qT) and Lr(qT) if the ratio is greater than the threshold, preliminarily judging that the main vehicle or the same-direction vehicle in front of the main vehicle has a traffic accident; continuously judging t0Set A of vehicles detected at time n + q Tn+qAnd t0Set of vehicles A detected at time T +(n-k)n-kIf the ratio beta is larger than the set value, the traffic accident of the main vehicle or the same-direction vehicle in front of the main vehicle is finally judged.
Illustratively, as shown in FIG. 4, the description set SnA schematic diagram of (a); as shown in fig. 5, is an explanatory set Ls(qT) and LrSchematic of (qT). FIG. 5 illustrates L as an example onlys(qT) and Lr(qT) two sets, FIG. 5 shows an example of a normal case where L is the number of accidents occurrings(qT) increase, Lr(qT) decreases, resulting in Ls(qT)/Lr(qT) increases.
The principle of the above-described scheme is explained below.
The method assumes that each vehicle traveling on the highway is equipped with a short-range communication module, such as WIFI or RFID. Each module can read in any case the identification information of other vehicles within its sensing range (including when an accident occurs). And assume that an accident will only affect the traffic in the direction in which the accident vehicle is located.
Under normal conditions, the average speed of the same-direction traffic flow is set asThe density of the equidirectional traffic (the sum of the number of vehicles evenly distributed in all lanes in one direction within a unit distance) is set as ρAll in one(ii) a The average speed of the oncoming traffic is set asThe density of the oncoming traffic is ρTo pairThe normal running speed of the host vehicle (a vehicle set as a subject of study) is V, and the speed becomes 0 in a very short time after an accident occurs in the host vehicle.
For the same-direction lane, the number of vehicles leaving the sensing range in unit time before the accident occurs isIn particular, if the host vehicle velocity and the velocity of the co-current flow are the same, L s0. The number of vehicles leaving the sensing range per unit time after an accident isWhereinρAll in one' the speed and density of the traffic flow in front of the accident point of the same lane after the accident occurs are respectively shown.
The number of vehicles leaving the sensing range of the host vehicle is related to the relative speed of the host vehicle, and the larger the relative speed is, the more vehicles leave in the same time. Before an accident, the main vehicle and the vehicles in the opposite lane are opposite to each other, so that the relative speed of the vehicles in the opposite lane is added to the speed of the main vehicle, and after the accident, the relative speed is reduced because the speed of the main vehicle is reduced, so that the vehicles leaving the induction range of the main vehicle in the same time are reduced, and a part of the vehicles in the opposite lane is regarded as the same-direction vehicles and contained in the set SnAnd therefore, fewer vehicles leave the main vehicle sensing range in the same time.
Similarly, the number of vehicles in the same direction leaving the sensing range of the host vehicle is also related to the relative speed of the host vehicle. Before an accident, the speed of the vehicle on the same lane is very close to that of the main vehicle, so the relative speed is very low; after the accident, since the host vehicle speed is 0, the vehicle speed on the same lane is small, and therefore the relative speed is also small. The number of vehicles leaving before and after an accident is thus clearly not obvious, but takes into account two factors:
1. the density of vehicles near the host vehicle on the co-directional lane becomes large, so the number of vehicles leaving is easily made large.
2. Some vehicles on the opposite lane are included in the set S as co-directional vehiclesnInside, the vehicles can leaveThe vehicle leaves as a co-directional vehicle.
In summary, the number of vehicles in the same direction that leave after an accident increases.
If the host vehicle velocity and the velocity of the equidirectional traffic flow are close, the above equation becomesThat is, the number L of vehicles leaving the sensing range on the same lane per unit times'=Ls+ΔAll in one。
For the opposite lane, the number of vehicles leaving the sensing range in unit time before the accidentSince the speed of the host vehicle becomes 0 after the occurrence of an accident, the number of vehicles departing the oncoming traffic lane per unit timeNamely, the vehicle number reduction amount of the opposite lane departure is:
that is, the number of vehicles L leaving the sensing range of the same lane per unit timer'=Lr-ΔTo pair。
after an accident, Ls' increase, Lr' decrease, so eta will increase obviously, therefore it can primarily identify the accident according to eta, and then judge whether the accident happens or not by combining the fact that the jam happens before the accident car after the accident happens and the density of the traffic flow increases.
According to the scheme of the embodiment of the invention, the vehicles running in the same lane and the opposite lane are automatically distinguished by skillfully utilizing a simple set thought on the premise of not knowing information such as speed, acceleration, position and the like. And then, by combining the change characteristics of local traffic flow before and after the accident, the traffic accident which influences the normal traffic flow can be automatically identified without a large amount of historical data.
Through the above description of the embodiments, it is clear to those skilled in the art that the above embodiments can be implemented by software, and can also be implemented by software plus a necessary general hardware platform. With this understanding, the technical solutions of the embodiments can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods according to the embodiments of the present invention.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (2)
1. An automatic traffic accident recognition method is characterized by comprising the following steps:
arranging a short-distance communication module on each vehicle to enable each vehicle to read identification information of other vehicles in the sensing range of the vehicle;
the main vehicle periodically detects vehicles in the sensing range of the main vehicle;
determining a same-direction vehicle set and an opposite-direction vehicle set which leave the sensing range within a period of time according to the vehicle set which leaves the sensing range within the period of time;
judging whether the ratio of the equidirectional vehicle set to the opposite vehicle set exceeds a threshold value, if so, judging that the traffic accident happens to the main vehicle or the equidirectional vehicle in front of the main vehicle if the ratio of the equidirectional vehicle set to the opposite vehicle set exceeds the threshold value and the traffic flow density detected by the main vehicle also belongs to an increasing state;
let T be the detection period of the main vehicle, r be the radius of the induction range, and T be the time when the main vehicle starts to enter the working state0The set of vehicles detected at the moment is A0,t0The set of vehicles detected at + nT is AnAnd during normal driving satisfiesWherein V is the speed of the main vehicle during normal running,is the average speed of the co-current traffic,the average speed of the opposite traffic flow; then the set of vehicles traveling in the same direction during the kT time period is denoted as Sn=An-k∩AnWherein the vehicle set is An-kIs t0+ (n-k) set of vehicles detected at time T;
suppose from t0From time + nT to t0Time (n + q) T, the set of vehicles leaving the main vehicle sensing range during this time being
Then the set of co-directional vehicles that left the sensing range at this time is:
Ls(qT)=Ln(qT)∩Sn;
the set of oncoming vehicles that are out of the sensing range at this time period is:
2. The automatic traffic accident recognition method according to claim 1,
if L iss(qT) and Lr(qT) if the ratio is greater than the threshold, preliminarily judging that the main vehicle or the same-direction vehicle in front of the main vehicle has a traffic accident;
continuously judging t0Set A of vehicles detected at time n + q Tn+qAnd t0Set of vehicles A detected at time T +(n-k)n-kIf the ratio is larger than the set value, the traffic accident of the main vehicle or the same-direction vehicle in front of the main vehicle is finally judged.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710595902.8A CN107481519B (en) | 2017-07-20 | 2017-07-20 | Automatic traffic accident identification method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710595902.8A CN107481519B (en) | 2017-07-20 | 2017-07-20 | Automatic traffic accident identification method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107481519A CN107481519A (en) | 2017-12-15 |
CN107481519B true CN107481519B (en) | 2020-10-23 |
Family
ID=60595108
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710595902.8A Active CN107481519B (en) | 2017-07-20 | 2017-07-20 | Automatic traffic accident identification method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107481519B (en) |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4461977B2 (en) * | 2004-09-21 | 2010-05-12 | 株式会社デンソー | Road congestion degree prediction system and road congestion degree prediction apparatus |
CN1766232A (en) * | 2005-11-21 | 2006-05-03 | 钱红波 | Novel traffic layout and control method for city road net and crossing |
CN101407199A (en) * | 2008-10-28 | 2009-04-15 | 湖北汽车工业学院 | Vehicle collision prewarning method and apparatus |
CN104200667B (en) * | 2014-09-19 | 2016-07-27 | 上海美慧软件有限公司 | A kind of traffic congestion hierarchical detection method based on mobile phone signal data |
CN105303852B (en) * | 2015-09-07 | 2018-09-28 | 首都师范大学 | Traffic accident message distributing method based on In-vehicle networking |
CN105448115B (en) * | 2016-01-20 | 2018-09-28 | 南通大学 | Passive cluster inter-vehicular communication method in high-speed transit network |
CN106803359A (en) * | 2016-09-27 | 2017-06-06 | 蔚来汽车有限公司 | Emergency method for early warning and system based on front truck driving information in the same direction |
-
2017
- 2017-07-20 CN CN201710595902.8A patent/CN107481519B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN107481519A (en) | 2017-12-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11024165B2 (en) | Driver behavior monitoring | |
US20220262239A1 (en) | Determining causation of traffic events and encouraging good driving behavior | |
CN102945603B (en) | Method for detecting traffic event and electronic police device | |
CN110136447A (en) | Lane change of driving a vehicle detects and method for distinguishing is known in illegal lane change | |
CN104809887B (en) | A kind of retrograde detection method of vehicle on expressway and autoalarm | |
CN109285349B (en) | Method for detecting highway traffic incident under cooperative vehicle and road environment and early warning system | |
CN104575043B (en) | Automatic prompt system and method during passing of motor vehicle through pedestrian crosswalk | |
CN103021176A (en) | Discriminating method based on section detector for urban traffic state | |
WO2017123665A1 (en) | Driver behavior monitoring | |
CN105118320A (en) | Urban road section traffic guidance method and device based on vehicle-road cooperation | |
CN103914688A (en) | Urban road obstacle recognition system | |
CN108871357B (en) | Method for displaying accident lane of congested road section on electronic map | |
CN103514742A (en) | Intelligent expressway traffic monitoring system based on GPS and GPRS | |
US20220383738A1 (en) | Method for short-term traffic risk prediction of road sections using roadside observation data | |
CN104021683B (en) | The self-adapting signal timing method that a kind of crossing amber light is captured | |
CN104574979B (en) | Motor line does not give way the camera system and method for behavior in accordance with the law through crossing | |
WO2022213542A1 (en) | Method and system for clearing information-controlled intersection on basis of lidar and trajectory prediction | |
JP2020154958A (en) | Abnormal traffic flow detection device, abnormal traffic flow detection method, and abnormal traffic flow detection program | |
CN105279968A (en) | Discrimination system and method for the illegal emergency lane occupancy behavior of highway motor vehicle | |
Zheng et al. | Critical gaps and follow-up headways at congested roundabouts | |
CN201882025U (en) | Anti-collision device for automobile | |
CN115440023A (en) | Method and device for judging traffic violation behaviors | |
CN107481519B (en) | Automatic traffic accident identification method | |
CN107886726A (en) | Road occupying/parking behavior detection method and device | |
CN104015657B (en) | The illegal method and system overtaken other vehicles of a kind of key preventive haulage vehicle |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |