CN104809874A - Traffic accident detection method and device - Google Patents
Traffic accident detection method and device Download PDFInfo
- Publication number
- CN104809874A CN104809874A CN201510182223.9A CN201510182223A CN104809874A CN 104809874 A CN104809874 A CN 104809874A CN 201510182223 A CN201510182223 A CN 201510182223A CN 104809874 A CN104809874 A CN 104809874A
- Authority
- CN
- China
- Prior art keywords
- model
- vehicle
- track
- car
- pavement structure
- 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
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
Abstract
The invention provides a traffic accident detection method and device. The method comprises the following steps: obtaining a traffic monitoring video image which is acquired at a pre-set position by a camera; establishing a vehicle track model and/or a vehicle surrounding environment model according to the traffic monitoring video image; and detecting a traffic accident according to the vehicle track model and a pre-established road surface structure model, and/or detecting a traffic accident according to the vehicle surrounding environment model and the pre-established road surface structure model, wherein the road surface structure model is a model which is established according to a monitoring video image without the traffic accident on the road surface. The automatic traffic accident detection method adopting a traffic accident detection system saves a lot of manpower and material resources, and the working intensity of monitoring personnel is alleviated. The risks of incorrect report and missed report of the traffic accident, caused by the fact that the effort is not enough, do not occur, so that the accuracy of detecting the traffic accident is improved by the detection method.
Description
Technical field
The present invention relates to road monitoring field, particularly relate to a kind of traffic incidents detection method and apparatus.
Background technology
Preventing road monitoring system plays an important role in public security prevention and control.In order to the traffic of Real-time Obtaining road, be provided with preventing road monitoring system at some Important Sections, by video camera shooting road real time status, then the image of shooting uploaded to road monitoring command centre in real time.
In existing preventing road monitoring system, the monitor staff of road monitoring command centre observes the monitor video image of the road all directions uploaded by operation The Cloud Terrace, artificially judge the generation of traffic events by observing monitor video image.This traffic event monitoring method consumes a large amount of human and material resources resources.And due to the energy of monitor staff limited, be easy to cause the wrong report of traffic events and fail to report.
Summary of the invention
In view of this, the invention provides a kind of traffic incidents detection method and apparatus, to realize the automatic detection to traffic events.
In order to solve the problems of the technologies described above, present invention employs following technical scheme:
A detection method for traffic events, comprising:
Obtain the Traffic Surveillance Video image gathered when video camera is in default presetting bit;
Track of vehicle model and/or all environmental models of car is set up according to described Traffic Surveillance Video image;
According to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance; Described pavement structure model is the model set up according to the monitor video image that road surface does not occur traffic events.
Alternatively, described traffic incidents detection method adopts every road video camera between multiple different presetting bit, carry out the method for automatic taking turn monitoring, multiple different presetting bit residing for the video camera of every road can cover the transversal section on whole road surface, setting video camera is after the monitoring period of a presetting bit reaches Preset Time, and automatic alternate is monitored to next presetting bit.
Alternatively, described according to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance after, also comprise:
After receiving the instruction carrying out manual intervention, cancel and load presetting bit associated profile, from receive carry out manual intervention instruction to after reaching default manual monitoring duration, whether prompting loads presetting bit associated profile, if, control the position of camera switching to the next presetting bit of described default presetting bit, with the traffic on the road that the next presetting bit gathering described default presetting bit is corresponding, if not, extend artificial monitor duration, from receive extend artificial monitor duration instruction to after again reaching default manual monitoring duration, return and perform the step whether described prompting loads presetting bit associated profile,
When receive refusal carry out the instruction of manual intervention after, judge whether video camera reaches Preset Time at the monitor event of current presetting bit, if, control the next presetting bit of camera switching to described default presetting bit, and return the step performing the Traffic Surveillance Video image gathered when described acquisition video camera is in default presetting bit.
Alternatively, describedly set up track of vehicle model according to described Traffic Surveillance Video image and comprise:
Detect the vehicle in described Traffic Surveillance Video image;
The vehicle detected is followed the tracks of, obtains the movement locus of described vehicle;
According to the movement locus of described vehicle, set up track of vehicle model.
Alternatively, described track of vehicle model and described pavement structure model include deflection model and the Velicle motion velocity model of vehicle movement;
Describedly specifically to comprise according to track of vehicle model and the pavement structure model inspection traffic events set up in advance:
Deflection model in more described track of vehicle model and described pavement structure model and/or rate pattern, obtain deflection deviation and/or the velocity deviation of described track of vehicle model and described pavement structure model;
Judge whether described deflection deviation reaches first pre-conditioned, and/or, judge whether described velocity deviation reaches second pre-conditioned, to judge whether road surface traffic events occurs;
When the deviation of the deflection model in the deflection model in described track of vehicle model and described pavement structure model is greater than angle threshold, judge that vehicle occurs to drive in the wrong direction event;
When the deviation of the speed variables in the speed variables in described track of vehicle model and described pavement structure model is greater than First Speed threshold value, judge overspeed of vehicle event occurs;
When the deviation of the speed variables in the speed variables in described pavement structure model and described track of vehicle model is greater than second speed threshold value, judge automobile low-speed event occurs;
When the speed variables in track of vehicle model described in first time period continues to be less than third speed threshold value, judge that vehicle occurs stops event;
When the speed variables in track of vehicle model road surface exceeding a certain proportion of vehicle continues to be less than third speed threshold value, judge vehicle congestion event occurs;
When on road surface, at least the track of vehicle model of two vehicles occurs that the overlapping and time remaining of overlap is more than the second time period, judge vehicle crash event occurs.
Alternatively, set up all environmental models of car according to described Traffic Surveillance Video image specifically to comprise:
Vehicle and surrounding environment part is chosen from described Traffic Surveillance Video image;
Car week environment contour feature figure is obtained according to described vehicle and surrounding environment part;
Set up car week environment DPM model according to described car week environment contour feature figure, described car all environment DPM model comprises root model p
0, the first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4, described model p
0for auto model, described first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4be respectively the environmental model on the four direction all around of vehicle.
Alternatively, described according to described car week environmental model and the pavement structure model inspection traffic events set up in advance specifically comprise:
Vehicle environmental model is mated with described pavement structure model, obtains matching value;
When matching value is greater than predetermined threshold value, from all environment of car, be partitioned into car Zhou Yichang object;
The movement locus of described car Zhou Yichang object is followed the tracks of, sets up the locus model of car Zhou Yichang object;
When the medium velocity of the locus model of car Zhou Yichang object is 0, starts timing, after the residence time of car Zhou Yichang object reaches Preset Time, determining that described car Zhou Yichang object is for shedding thing.
A pick-up unit for traffic events, comprising:
Acquiring unit, for obtaining the Traffic Surveillance Video image gathered when video camera is in default presetting bit;
Unit set up by model, for setting up track of vehicle model and/or all environmental models of car according to described Traffic Surveillance Video image;
Detecting unit, for according to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance; Described pavement structure model is the model set up according to monitor video image when there is not traffic events.
Alternatively, the method that described traffic incidents detection device adopts every road video camera to carry out automatic taking turn monitoring between multiple different presetting bit carries out the detection of traffic events, multiple different presetting bit residing for the video camera of every road can cover the transversal section on whole road surface, setting video camera is after the monitoring period of a presetting bit reaches Preset Time, and automatic alternate is monitored to next presetting bit.
Alternatively, described model is set up unit and is comprised:
Detection sub-unit, for detecting the vehicle in described Traffic Surveillance Video image;
Following the tracks of subelement, for following the tracks of the vehicle detected, obtaining the movement locus of described vehicle;
Subelement set up by track of vehicle model, for the movement locus according to described vehicle, sets up track of vehicle model.
Alternatively, described track of vehicle model and described pavement structure model include deflection model and the Velicle motion velocity model of vehicle movement;
Described detecting unit comprises:
Relatively subelement, for the deflection model in more described track of vehicle model and described pavement structure model and/or rate pattern, obtains deflection deviation and/or the velocity deviation of described track of vehicle model and described pavement structure model;
Judgment sub-unit, for judging whether described deflection deviation reaches first pre-conditioned, and/or whether described velocity deviation reaches second pre-conditioned, to judge whether road surface traffic events occurs;
When the deviation of the deflection model in the deflection model in described track of vehicle model and described pavement structure model is greater than angle threshold, judge that vehicle occurs to drive in the wrong direction event;
When the deviation of the speed variables in the speed variables in described track of vehicle model and described pavement structure model is greater than First Speed threshold value, judge overspeed of vehicle event occurs;
When the deviation of the speed variables in the speed variables in described pavement structure model and described track of vehicle model is greater than second speed threshold value, judge automobile low-speed event occurs;
When the speed variables in track of vehicle model described in first time period continues to be less than third speed threshold value, judge that vehicle occurs stops event;
When the speed variables in track of vehicle model road surface exceeding a certain proportion of vehicle continues to be less than third speed threshold value, judge vehicle congestion event occurs;
When on road surface, at least the track of vehicle model of two vehicles occurs that the overlapping and time remaining of overlap is more than the second time period, judge vehicle crash event occurs.
Alternatively, described model is set up unit and is comprised:
Choose subelement, for choosing vehicle and surrounding environment part from described Traffic Surveillance Video image;
Obtain subelement, for obtaining car week environment contour feature figure according to described vehicle and surrounding environment part;
Car week, subelement set up by environment DPM model, and for setting up car week environment DPM model according to described car week environment contour feature figure, described car all environment DPM model comprises root model p
0, the first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4, described model p
0for auto model, described first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4be respectively the environmental model on the four direction all around of vehicle.
Alternatively, described detecting unit comprises:
Coupling subelement, for being mated with described pavement structure model by vehicle environmental model, obtains matching value;
Segmentation subelement, for when described matching value is greater than predetermined threshold value, is partitioned into car Zhou Yichang object from all environment of car;
Car Zhou Yichang object locus model sets up subelement, for following the tracks of the movement locus of described car Zhou Yichang object, sets up the locus model of car Zhou Yichang object;
Judge to shed thing subelement, when the medium velocity for the locus model when car Zhou Yichang object is 0, starts timing, after the residence time of car Zhou Yichang object reaches Preset Time, determining that described car Zhou Yichang object is for shedding thing.
Compared to prior art, the present invention has following beneficial effect:
In the detection method of traffic events provided by the invention, traffic incident detecting system can detect on road surface whether there occurs traffic events automatically according to track of vehicle model and the pavement structure set up in advance model.Therefore, traffic incidents detection method provided by the invention instead of in prior art and observes by staff the method that monitor video image carrys out whether artificial judgment road surface occurs traffic events.This method by traffic incident detecting system automatic transport detection event saves a large amount of human and material resources, reduces the working strength of monitor staff.And can not occur to cause the wrong report of traffic events and the risk failed to report due to deficient in energy, therefore, this detection method improves the degree of accuracy of traffic incidents detection.
Accompanying drawing explanation
In order to be expressly understood technical scheme of the present invention, the accompanying drawing used is done a brief description below when describing the specific embodiment of the present invention.Apparently, these accompanying drawings are only section Example of the present invention, and those of ordinary skill in the art can also obtain other accompanying drawing under the prerequisite not paying creative work.
Fig. 1 is the traffic incidents detection method flow schematic diagram that the embodiment of the present invention one provides;
Fig. 2 is the method for building up schematic flow sheet of the pavement structure model that the embodiment of the present invention one provides;
Fig. 3 is the traffic incidents detection method flow schematic diagram that the embodiment of the present invention two provides;
Fig. 4 is the method flow schematic diagram of the foundation of all environmental models of car that the embodiment of the present invention two provides;
Fig. 5 is the method flow schematic diagram of the utilization intelligence taking turn monitor mode detection traffic events that the embodiment of the present invention three provides;
Fig. 6 is the operating process schematic diagram that intelligent taking turn supervisory system that the embodiment of the present invention three provides is carried out under manual monitoring pattern;
Fig. 7 is the traffic incidents detection apparatus structure schematic diagram that the embodiment of the present invention four provides;
Fig. 8 is the structural representation of first embodiment of the traffic incidents detection device that the embodiment of the present invention four provides;
Fig. 9 is the structural representation of second embodiment of the traffic incidents detection device that the embodiment of the present invention four provides;
Figure 10 is the structural representation of the traffic incidents detection device of the utilization intelligence taking turn supervisory system that the embodiment of the present invention four provides.
Embodiment
For make goal of the invention of the present invention, technical scheme and the technique effect that reaches clearly, complete, below in conjunction with accompanying drawing, the specific embodiment of the present invention is described in detail.
Embodiment one
Fig. 1 is the schematic flow sheet of the traffic incidents detection method that the embodiment of the present invention one provides.As shown in Figure 1, this traffic incidents detection method comprises the following steps:
The Traffic Surveillance Video image that S101, acquisition video camera gather when being in default presetting bit:
Control the video camera be arranged on road to open, make the traffic information on camera acquisition road.Because video camera is arranged on The Cloud Terrace usually, so video camera can be arranged in the default presetting bit of specifying according to demand.After video camera is opened, video camera just can gather traffic information, and the traffic information collected is uploaded to intelligent monitor and control center, and intelligent monitor and control center just can get the Traffic Surveillance Video image gathered when video camera is in default presetting bit.
S102, set up track of vehicle model according to described Traffic Surveillance Video image:
It should be noted that, the usual technological means in this area can be adopted to set up track of vehicle model according to described Traffic Surveillance Video image.
As a specific embodiment of the present invention, describedly set up track of vehicle model according to described Traffic Surveillance Video image and specifically can comprise the following steps:
S1021, the vehicle detected in described Traffic Surveillance Video image:
It should be noted that, the method for the employing background modeling that the method for vehicles in the detection video image that the embodiment of the present invention provides can adopt this area usual.The method of this employing background modeling is by carrying out modeling to the background of image, and present image and background image compare, according to comparative result determination foreground target after setting up by model.But there is following problem in the method:
1) responsive to shade, being easy to shade error detection is moving target;
2) adaptive faculty under low contrast and halation environment is poor;
3) to object without separating capacity, easily produce and identify by mistake, thus have a strong impact on the problem of the statistics of the magnitude of traffic flow and the speed of a motor vehicle.
In order to avoid the problems referred to above, the present invention preferably adopts publication number to be the Chinese patent application of CN101030256A (publication date is on September 5th, 2007).The method of this patented claim can recognition detection target definitely, and compared to traditional method based on background modeling, the method has clear superiority in shadow interference, complex illumination etc.
S1022, the vehicle detected to be followed the tracks of, obtains the movement locus of described vehicle:
The vehicle detected is followed the tracks of, and its movement locus is analyzed, the driving trace of vehicle can be understood definitely.
The wireless vehicle tracking that the present invention adopts can adopt the method based on optical flow tracking or template matches; but usually there is following problem in these methods: 1) occlusion issue; 2) deformation problems; usually can change in size be there is in vehicle in the process travelled; turn simultaneously and also can cause the metamorphosis of vehicle, thus the stability that impact is followed the tracks of.
In order to avoid the problems referred to above, the present invention preferably adopts a kind of new wireless vehicle tracking, the method adopts follows the tracks of and detects the mechanism combined, and the real-time model of target vehicle is set up by on-line study, the result of on-line study reacted on simultaneously and follow the tracks of and detect, final target localization is by following the tracks of and detect collaborative decision.The method effectively can solve the deformation of target in vehicle tracking process, other object block and block rear vehicle due to turn wait generation deformation problem.
S1023, movement locus according to described vehicle, set up track of vehicle model.
S103, according to described track of vehicle model and the pavement structure model inspection traffic events set up in advance:
It should be noted that, the model that monitor video image when pavement structure model is normal according to road traffic situation is set up.
Vehicle movement track time normal by statistics road traffic situation, carries out modeling to track of vehicle, and carries out on-line study, thus automatically identify pavement structure, set up corresponding pavement structure model according to this pavement structure.The normal situation of described road traffic situation refers to the situation that road surface does not occur traffic events.
It should be noted that, the movement locus of vehicle generally includes two factors: travel direction and travel speed.So described pavement structure model comprises the direction of motion angle model of vehicle and the movement velocity model of vehicle, the locus model of same vehicle also comprises the direction of motion angle model of vehicle and the movement velocity model of vehicle.Therefore, the traffic events relevant with travel speed with the travel direction of vehicle can be detected according to the locus model of vehicle with pavement structure model.
In embodiments of the present invention, to be describedly specially according to track of vehicle model and the pavement structure model inspection traffic events set up in advance:
Obtain described track of vehicle model and described pavement structure model:
Deflection model in more described track of vehicle model and described pavement structure model and/or rate pattern, to obtain deflection deviation in described track of vehicle model and described pavement structure model and/or velocity deviation;
Judge whether described deflection deviation reaches first pre-conditioned, and/or, judge whether described velocity deviation reaches second pre-conditioned, to judge whether road surface traffic events occurs.
When the deviation of the deflection model in the deflection model in described track of vehicle model and described pavement structure model is greater than angle threshold, judge that vehicle occurs to drive in the wrong direction event;
When the deviation of the speed variables in the speed variables in described track of vehicle model and described pavement structure model is greater than First Speed threshold value, judge overspeed of vehicle event occurs;
When the deviation of the speed variables in the speed variables in described pavement structure model and described track of vehicle model is greater than second speed threshold value, judge automobile low-speed event occurs;
When the speed variables in track of vehicle model described in first time period continues to be less than third speed threshold value, judge that vehicle occurs stops event;
When the speed variables in track of vehicle model road surface exceeding a certain proportion of vehicle continues to be less than third speed threshold value, judge vehicle congestion event occurs;
When on road surface, at least the track of vehicle model of two vehicles occurs that the overlapping and time remaining of overlap is more than the second time period, judge vehicle crash event occurs.
It should be noted that, in embodiments of the present invention, in order to traffic events be detected more accurately, each two field picture in Traffic Surveillance Video image is become multiple fritter according to identical regular partition, pavement structure model is made up of the pavement structure model in the plurality of fritter.
As a specific embodiment of the present invention, each two field picture in Traffic Surveillance Video image becomes N × M fritter according to identical regular partition, and wherein, N, M are the integer being more than or equal to 1;
In monitor video image, the method for building up of the pavement structure model of each fritter is all identical, is described the method for building up of pavement structure model below for the method for building up of the pavement structure model of any one fritter in N × M the fritter divided.
The default fritter of setting is any one fritter in N × M fritter, if think that N × M fritter forms an array in monitor video image, then presetting fritter position is in the array the i-th row jth row, this default fritter position in monitor video image can be expressed as (i, j), wherein, 1≤i≤N, 1≤j≤M, and i, j are integer.
As a specific embodiment of the present invention, as shown in Figure 2, the method for building up of the pavement structure model of described default fritter comprises the following steps:
S21, from the t two field picture and t-1 two field picture of default fritter, extract unique point for following the tracks of:
Adopt the technological means that this area is usual, from the t two field picture and t-1 two field picture of default fritter, extract the unique point for following the tracks of.
In embodiments of the present invention, t >=2, and t is integer.
S22, light stream estimation is carried out to the movement locus for the unique point of following the tracks of in t two field picture and t-1 two field picture, obtain the initialization model of the pavement structure in described default fritter.
S23, utilize the t+1 two field picture in described default fritter to be optimized described initialization model, obtain the pavement structure model after the renewal in default fritter.
S24, judge whether t is less than default training frame number, if not, perform step S25.If so, step S26 is performed.
S25, model filtering is carried out to the pavement structure model after described renewal, finally obtains the pavement structure model of described default fritter:
Carrying out one of object of model filtering to the pavement structure model after renewal is the movement locus in order to eliminate pedestrian on road and the track of vehicle not in accordance with traffic rules traveling, the information of road surface under making the pavement structure model obtained meet normal traffic conditions.
S26, t two field picture is updated to t+1 two field picture, returns and perform described step S21.
It should be noted that, step S21 to step S26 is the process of establishing of the pavement structure model in any one fritter, also can be regarded as the training process of the pavement structure model in any one fritter.
Statistics display, the deflection θ of road vehicles motion and the movement velocity v of vehicle meets Gaussian distribution, therefore, sets up the direction of vehicle movement of each fritter and the mixed Gauss model P of speed in the training process of the pavement structure model of each fritter
(i, j)(θ, v).
In embodiments of the present invention, the pavement structure model of vehicle can be regarded as the set of the pavement structure model of fritter shared by it.
In embodiments of the present invention, automatically detecting the various types of traffic events occurred on road to make intelligent monitor system, in intelligent monitor system, being previously provided with various types of traffic events model.After road traffic situation meets the traffic events model of a certain type, then intelligent monitor system judges the traffic events that road surface there occurs the type automatically.
As mentioned above, in embodiments of the present invention, traffic events comprise vehicle drive in the wrong direction event, overspeed of vehicle event, automobile low-speed event, vehicle stop event, vehicle congestion event and vehicle crash event.
Accordingly, each traffic events model above-mentioned is as follows:
Before each traffic events model of description, the physical significance of the parameters under first explaining in model below.(i, j) for the position of fritter in monitor video image be i-th row jth row; N represents the feature point number for following the tracks of in fritter; M represents the fritter number that vehicle is shared in monitor video image.
Wherein, the vehicle event model that drives in the wrong direction is:
Wherein,
represent the average angle of pavement structure model in this fritter (i, j);
represent the angle intermediate value of vehicle in this fritter;
δ is angle threshold.
Overspeed of vehicle event model is:
Wherein,
represent pavement structure model in fritter (i, j) average velocity;
for the speed intermediate value of the vehicle in track of vehicle model in this fritter (i, j);
ε 1 represents First Speed threshold value.
Automobile low-speed event model is:
Wherein,
represent pavement structure model in fritter (i, j) average velocity;
for the speed intermediate value of the vehicle in track of vehicle model in this fritter (i, j);
ε 2 represents second speed threshold value.
Vehicle stops event model in first time period, and the relation of the speed variables in the speed variables in track of vehicle model and pavement structure model meets formula below:
Wherein,
represent pavement structure model in fritter (i, j) average velocity;
for the speed intermediate value of the vehicle in track of vehicle model in this fritter (i, j);
ε 3 represents third speed threshold value, and ε 3 is almost nil.
Stop model can determining a vehicle generation Parking according to vehicle, adopt same decision method, after when the vehicle that vehicle most of on road can be set to exceed certain predetermined ratio all occurs to stop time, now judge road there occurs traffic congestion event.
When the track of vehicle model of at least two vehicles on road occurs overlapping, and this at least two vehicles stopping, then intelligent monitor system judges road there occurs vehicle crash event.
Further, after judgement road there occurs vehicle crash event, vehicle collision type can be judged according to the deflection model in the track of vehicle model of vehicle collided and rate pattern further: side collision, to knock into the back or head-on crash.
S104, send alarm when traffic events being detected.
As the preferred embodiments of the present invention, after detecting road surface there occurs traffic events, outwards can also send alarm, to point out staff, the traffic events detected be processed.
The traffic incidents detection method provided by embodiment one, intelligent monitor system is by comparing deflection model in locus model and pavement structure model and/or travel speed model, then whether meet first pre-conditioned according to the deflection deviation in locus model and pavement structure model, and/or whether travel speed deviation meets second pre-conditioned (namely whether road traffic situation reaches traffic events model) and judges whether road surface there occurs traffic events.Traffic incidents detection method provided by the invention can detect the traffic events on road surface automatically.Therefore, traffic incidents detection method provided by the invention instead of in prior art and observes by staff the method that monitor video image carrys out whether artificial judgment road surface occurs traffic events.This method by traffic incident detecting system automatic transport detection event saves a large amount of human and material resources, reduces the working strength of monitor staff.And can not occur to cause because personnel are deficient in energy to the wrong report of traffic events and the risk failed to report, therefore, this detection method improves the degree of accuracy of traffic incidents detection.
In traffic incidents detection method described in embodiment one, travel direction according to track of vehicle model and vehicle to the direction of traffic angle model in pavement structure model and relatively can detecting of rate pattern, whether road surface there occurs the traffic events relevant with track of vehicle with travel speed, such as: vehicle drives in the wrong direction, hypervelocity, at a slow speed, stopping, the event such as collision.In addition, present invention also offers and a kind ofly can detect the method for shedding thing.Specifically see embodiment two.
Embodiment two
Fig. 3 is the schematic flow sheet of the traffic incidents detection method that the embodiment of the present invention two provides, and as shown in Figure 3, this traffic incidents detection method comprises the following steps:
The Traffic Surveillance Video image that S301, acquisition video camera gather when being in default presetting bit:
This step is identical with the step S101 in embodiment one, for the sake of brevity, is not described in detail at this, specifically see the description of embodiment one.
S302, to set up car week environmental model according to described Traffic Surveillance Video image:
As shown in Figure 4, this step specifically comprises the following steps:
S41, from described Traffic Surveillance Video image, choose vehicle and surrounding environment part:
The usual technological means in this area is adopted to choose vehicle and surrounding environment part from described Traffic Surveillance Video image.
S42, obtain car week environment contour feature figure according to described car week and surrounding environment part:
Particularly, according to the vehicle chosen and surrounding environment part, obtain profile diagram based on sobel and SURF algorithm, then remove lane line interference by Hough algorithm, obtain car week environment contour feature figure.
S43, set up car week environment DPM model according to described car week environment contour feature figure:
From described car week environment contour feature figure, be partitioned into car body, set up auto model p according to car body
0, from described car week environment contour feature figure, be partitioned into the surrounding environment of vehicle, set up the model on the four direction all around of vehicle respectively according to the surrounding environment be partitioned into.
With auto model p
0as root model, using the surrounding environment model of vehicle as partial model, the DPM model of these five all environment of model composition car, wherein, surrounding environment model is the partial model of DPM model, and it is respectively the first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4.DPM (can the manufacture) model being formulated car week environment is as follows:
z＝(p
0,p
1,p
2,p
3,p
4)。
S303, according to described car week environmental model and the pavement structure model inspection set up in advance shed formal matter part:
This step specifically comprises the following steps:
S3031, vehicle environmental model to be mated with pavement structure model, obtains matching value:
It should be noted that, the pavement structure model described in the embodiment of the present invention comprises four partial models of root model and the vehicle's surroundings environment structure be made up of auto model.
Wherein, vehicle environmental model and the pavement structure model formula that carries out mating is as follows:
Wherein, score (p
0, p
1, p
2, p
3, p
4) be matching value;
F
0, F
irepresent the root model in pavement structure model and the i-th partial model respectively.
S3032, when matching value is greater than predetermined threshold value, then determine car Zhou Yichang, from car week environment, be partitioned into car Zhou Yichang object:
S3033, the movement locus of car Zhou Yichang object to be followed the tracks of, sets up the locus model of car Zhou Yichang object:
S3034, when the medium velocity of the locus model of car Zhou Yichang object is 0, determines that car Zhou Yichang object is stationary object, and start timing, after the residence time of car Zhou Yichang object reaches Preset Time, feature extraction and classifying is carried out to car Zhou Yichang object:
In order to determine that car Zhou Yichang object is for shedding thing more exactly, after the residence time of car Zhou Yichang object reaches Preset Time, this car Zhou Yichang object is likely pedestrian, in order to avoid to the erroneous judgement of shedding thing, determining car Zhou Yichang object whether for before shedding thing, preferably feature extraction and classifying is carried out to car Zhou Yichang object.
S3035, according to extract feature determination car Zhou Yichang object whether for shedding thing.
S304, send alarm when detecting and shedding formal matter part.
As a preferred embodiment of the present invention, after detecting and road surface there occurs shedding formal matter part, outwards alarm can also be sent, to remind staff.
Automatically can be detected by the traffic incidents detection method described in embodiment two and road surface sheds formal matter part.
By the traffic incidents detection method described in embodiment one and embodiment two, can not only detect and connect the relevant traffic events of track with car, can also detect according to car week environment and shed formal matter part.The traffic events that embodiment one and embodiment two detect has consisted essentially of all traffic events on road, so the traffic incidents detection method provided by embodiment one and embodiment two, intelligent monitor system can detect the traffic events occurred on road automatically, therefore, the traffic incidents detection method that embodiment one and embodiment two provide instead of the method for manual detection traffic events, improve the accuracy detecting traffic events, reduce the risk occurring undetected or flase drop.
It should be noted that, in embodiment one and the traffic incidents detection method described in embodiment two, video camera can be fixed on a certain presetting bit always, in the whole monitoring course of work, camera supervised visual angle immobilizes, but the monitor mode of this fixed viewpoint, likely there is monitoring blind area, thus cause failing to report of traffic events.In order to avoid there is monitoring blind area, and make full use of the feature of monopod video camera, as a preferred embodiment of the present invention, the method that intelligent taking turn can be adopted to monitor controls every road video camera and rotates between multiple different presetting angle.The monitoring visual field that multiple different presetting angle residing for the video camera of every road reaches can cover the transversal section on whole road surface.The intelligent taking turn monitor mode of this utilization detects the method for traffic events as shown in embodiment three.
It should be noted that, when a certain section being provided with multichannel video camera, the position control mode of this multichannel video camera can adopt coordinated signals mode.When the position of a road video camera changes, the position of other road video camera also changes.In order to avoid monitoring blind area, the position of multichannel video camera residing for synchronization be arranged on same section can ensure that monitoring visual field covers the transversal section on whole road surface.In addition, when multichannel video camera carries out rotation between different presetting bit, the seamless connection of monitoring visual field can be ensured.
Embodiment three
It should be noted that, the mode of operation of the intelligent taking turn supervisory system described in the embodiment of the present invention comprises automatic taking turn monitoring mode and manual monitoring pattern.Under normal circumstances, intelligent taking turn supervisory system is in automatic taking turn monitoring mode, only having under special circumstances as there is traffic events, needing just can be in manual monitoring pattern during manual intervention.
Fig. 5 is that the utilization intelligence taking turn monitor mode that provides of the embodiment of the present invention three is to detect the method flow schematic diagram of traffic events.As shown in Figure 5, the method that this utilization intelligence taking turn monitor mode detects traffic events comprises the following steps:
The shooting angle of S501, adjustment video camera reaches a certain default presetting bit, this default presetting bit is regarded as the current presetting bit of video camera:
The shooting angle of intelligent monitor system adjustment video camera reaches a certain default presetting bit, this default presetting bit is regarded as the current presetting bit of video camera.
S502, load the associated profile of current presetting bit.
S503, detect traffic events according to the associated profile of current presetting bit.
It should be noted that, the associated profile according to current presetting bit described in this step detects traffic events and specifically comprises step S101 described in embodiment one to step S103, and/or the step S301 described in embodiment two is to step S303.
S504, when traffic events being detected, outwards send alarm.
S505, after receiving staff and carrying out the instruction of manual intervention, intelligent taking turn supervisory system switches to manual monitoring pattern from automatic taking turn monitoring mode:
After monitoring staff receives the alarm that intelligent monitor system sends, monitoring staff judges whether the traffic events of reporting to the police is rational traffic events, to determine whether carry out manual intervention further.When intelligent taking turn supervisory system receives after staff carries out the instruction of manual intervention, intelligent taking turn supervisory system switches to manual monitoring pattern from automatic taking turn monitoring mode.
After intelligent taking turn supervisory system switches to manual monitoring pattern from automatic taking turn monitoring mode, as shown in Figure 6, intelligent taking turn supervisory system can perform following operation:
S61, cancellation load presetting bit associated profile.
After S62, default manual monitoring duration to be achieved, whether intelligent taking turn supervisory system prompting loads presetting bit associated profile.
If after S63 obtains the positive reply of monitoring staff, intelligent taking turn supervisory system switches to automatic taking turn monitoring mode from manual monitoring pattern.
If S64 obtain monitor staff negative reply after, intelligent taking turn supervisory system extends artificial monitor duration, and return perform step S62.
S506, when receive staff refusal carry out the instruction of manual intervention after, intelligent taking turn supervisory system judges whether video camera reaches Preset Time at the monitoring period of current presetting bit, if so, execution step S507.
Current presetting bit to next presetting bit, and is updated to this next presetting bit by S507, control camera switching, returns and performs step S502.
The intelligent taking turn method for supervising provided by embodiment three can either make video camera play the advantage of automatic taking turn, greatly can reduce again the working strength of monitor staff, the event that farthest reduces is failed to report and improves reaction velocity, and the monitoring capacity of every platform video camera is greatly improved.
The embodiment of the detection method of the traffic events that embodiment one to embodiment three provides for the embodiment of the present invention, based on the detection method of the traffic events that above-described embodiment provides, present invention also offers the embodiment of the pick-up unit of traffic events, specifically see following examples four.
Embodiment four
Fig. 7 is the structural representation of the traffic incidents detection device that the embodiment of the present invention four provides.As shown in Figure 7, this traffic incidents detection device comprises with lower unit:
Acquiring unit 71, for obtaining the Traffic Surveillance Video image gathered when video camera is in default presetting bit;
Unit 72 set up by model, for setting up track of vehicle model and/or all environmental models of car according to described Traffic Surveillance Video image;
Detecting unit 73, for according to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance; Described pavement structure model is the model set up according to monitor video image when there is not traffic events;
There occurs traffic events to remind on staff road surface, described traffic incidents detection device can also comprise:
Alarm unit 74, when traffic events being detected, sends alarm.
The traffic events on road surface automatically can be detected by the traffic incidents detection device shown in Fig. 7, avoid by manually carrying out the method detected, save a large amount of human and material resources resources, reduce the working strength of monitor staff, and can not occur to cause the wrong report of traffic events and the risk failed to report due to deficient in energy, therefore, this pick-up unit improves the degree of accuracy of traffic incidents detection.
In order to the traffic events relevant to track of vehicle can be detected, as first specific embodiment of the present invention, as shown in Figure 8, the described model in the traffic incidents detection device shown in above-described embodiment is set up unit 72 and is comprised:
Detection sub-unit 721, for detecting the vehicle in described Traffic Surveillance Video image;
Following the tracks of subelement 722, for following the tracks of the vehicle detected, obtaining the movement locus of described vehicle;
Subelement 723 set up by track of vehicle model, for the movement locus according to described vehicle, sets up track of vehicle model.
Further, to the further improvement of above-mentioned first specific embodiment, as shown in Figure 8, described track of vehicle model and described pavement structure model include deflection model and the Velicle motion velocity model of vehicle movement;
Described detecting unit 73 specifically comprises:
Relatively subelement 731, for the deflection model in more described track of vehicle model and described pavement structure model and/or rate pattern, obtains deflection deviation and/or the velocity deviation of described track of vehicle model and described pavement structure model;
Judgment sub-unit 732, for judging whether described deflection deviation reaches first pre-conditioned, and/or whether described velocity deviation reaches second pre-conditioned, to judge whether road surface traffic events occurs;
When the deviation of the deflection model in the deflection model in described track of vehicle model and described pavement structure model is greater than angle threshold, judge that vehicle occurs to drive in the wrong direction event;
When the deviation of the speed variables in the speed variables in described track of vehicle model and described pavement structure model is greater than First Speed threshold value, judge overspeed of vehicle event occurs;
When the deviation of the speed variables in the speed variables in described pavement structure model and described track of vehicle model is greater than second speed threshold value, judge automobile low-speed event occurs;
When the speed variables in track of vehicle model described in first time period continues to be less than third speed threshold value, judge that vehicle occurs stops event;
When the speed variables in track of vehicle model road surface exceeding a certain proportion of vehicle continues to be less than third speed threshold value, judge vehicle congestion event occurs;
When on road surface, at least the track of vehicle model of two vehicles occurs that the overlapping and time remaining of overlap is more than the second time period, judge vehicle crash event occurs.
By above-mentioned first specific embodiment, traffic incidents detection device can detect the traffic events relevant to track of vehicle, these traffic events comprise that vehicle drives in the wrong direction, hypervelocity, at a slow speed, stop, blocking up, collision accident.
Traffic incidents detection device provided by the invention, except can detecting the traffic events relevant to track of vehicle, can also detect and shed formal matter part.
Shed formal matter part to detect, as second specific embodiment of the present invention, as shown in Figure 9, the described model in the traffic incidents detection device shown in above-mentioned Fig. 8 is set up unit 72 and can also be comprised:
Choose subelement 721 ', for choosing vehicle and surrounding environment part from described Traffic Surveillance Video image;
Obtain subelement 722 ', for obtaining car week environment contour feature figure according to described vehicle and surrounding environment part;
Car week, subelement 723 ' set up by environment DPM model, and for setting up car week environment DPM model according to described car week environment contour feature figure, described car all environment DPM model comprises root model p
0, the first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4, described model p
0for auto model, described first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4be respectively the environmental model on the four direction all around of vehicle.
Further, described detecting unit 73 can also comprise:
Coupling subelement 731 ', for being mated with described pavement structure model by vehicle environmental model, obtains matching value;
Segmentation subelement 732 ', for when described matching value is greater than predetermined threshold value, is partitioned into car Zhou Yichang object from all environment of car;
Car Zhou Yichang object locus model sets up subelement 733 ', for following the tracks of the movement locus of described car Zhou Yichang object, sets up the locus model of car Zhou Yichang object;
Judge to shed thing subelement 734 ', when the medium velocity for the locus model when car Zhou Yichang object is 0, starts timing, after the residence time of car Zhou Yichang object reaches Preset Time, determining that described car Zhou Yichang object is for shedding thing.
It should be noted that, traffic incidents detection device shown in Fig. 9 is the improvement carried out on the basis of the traffic incidents detection device shown in Fig. 8, in fact, as another embodiment of the present invention, can also improve on the basis of the traffic incidents detection device shown in Fig. 7 for detecting the traffic incidents detection device shedding formal matter part.Information disclosed in Fig. 9, those of ordinary skill in the art can obtain this embodiment under the prerequisite not paying creative work, are not described in detail at this.
In above-mentioned first specific embodiment and second specific embodiment, video camera can be fixed on a certain presetting bit always, in the whole monitoring course of work, camera supervised visual angle immobilizes, but the monitor mode of this fixed viewpoint, likely there is monitoring blind area, thus cause failing to report of traffic events.In order to avoid there is monitoring blind area, and make full use of the feature of monopod video camera, as a preferred embodiment of the present invention, the method that intelligent taking turn can be adopted to monitor controls every road video camera and rotates between multiple different presetting angle.The monitoring visual field that multiple different presetting angle residing for the video camera of every road reaches can cover the transversal section on whole road surface.The intelligent taking turn monitor mode of this utilization detects the device of traffic events as shown in the 3rd embodiment.
It should be noted that, when a certain section being provided with multichannel video camera, the position control mode of this multichannel video camera can adopt coordinated signals mode.When the position of a road video camera changes, the position of other road video camera also changes.In order to avoid monitoring blind area, the position of multichannel video camera residing for synchronization be arranged on same section can ensure that monitoring visual field covers the transversal section on whole road surface.In addition, when multichannel video camera carries out rotation between different presetting bit, the seamless connection of monitoring visual field can be ensured.
Further, the method that described traffic incidents detection device adopts every road video camera to carry out automatic taking turn monitoring between multiple different presetting bit carries out the detection of traffic events, multiple different presetting bit residing for the video camera of every road can cover the transversal section on whole road surface, setting video camera is after the monitoring period of a presetting bit reaches Preset Time, and automatic alternate is monitored to next presetting bit.
As the 3rd embodiment of the present invention, as shown in Figure 10, described device, except having the unit of first embodiment or second embodiment, can also comprise with lower unit:
Cancel and load presetting bit associated profile unit 101, for after receiving staff and carrying out the instruction of manual intervention, cancel and load presetting bit associated profile;
Tip element 102, for reach default manual monitoring duration from receiving staff and carry out the instruction of manual intervention after, whether prompting loads presetting bit associated profile, if, control the position of camera switching to the next presetting bit of described default presetting bit, with the traffic on the road that the next presetting bit gathering described default presetting bit is corresponding, if not, extend artificial monitor duration, from receiving after the instruction extending artificial monitor duration plays and again reach default manual monitoring duration, the signal of manual monitoring duration is sent to described Tip element.
Further, described device can further include: judging unit 103, for when receive staff refusal carry out the instruction of manual intervention after, judge whether video camera reaches Preset Time at the monitor event of current presetting bit, if, control the next presetting bit of camera switching to described default presetting bit, and the position signalling residing for video camera is sent to described acquiring unit.
It should be noted that, the traffic incidents detection device shown in Figure 10 is the improvement carried out on the basis of the embodiment shown in first specific embodiment of the present invention and Fig. 8.In fact, the traffic incidents detection device that can carry out intelligent taking turn can improve on the basis of above-mentioned any embodiment.
Be more than the preferred embodiments of the present invention.It should be pointed out that for the person of ordinary skill of the art, under the prerequisite not departing from inventive concept of the present invention, can also make some improvements and modifications, these improvements and modifications also should at the row of protection scope of the present invention.
Claims (13)
1. a detection method for traffic events, is characterized in that, comprising:
Obtain the Traffic Surveillance Video image gathered when video camera is in default presetting bit;
Track of vehicle model and/or all environmental models of car is set up according to described Traffic Surveillance Video image;
According to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance; Described pavement structure model is the model set up according to the monitor video image that road surface does not occur traffic events.
2. method according to claim 1, it is characterized in that, described traffic incidents detection method adopts every road video camera between multiple different presetting bit, carry out the method for automatic taking turn monitoring, multiple different presetting bit residing for the video camera of every road can cover the transversal section on whole road surface, setting video camera is after the monitoring period of a presetting bit reaches Preset Time, and automatic alternate is monitored to next presetting bit.
3. method according to claim 2, it is characterized in that, described according to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance after, also comprise:
After receiving the instruction carrying out manual intervention, cancel and load presetting bit associated profile, from receive carry out manual intervention instruction to after reaching default manual monitoring duration, whether prompting loads presetting bit associated profile, if, control the position of camera switching to the next presetting bit of described default presetting bit, with the traffic on the road that the next presetting bit gathering described default presetting bit is corresponding, if not, extend artificial monitor duration, from receive extend artificial monitor duration instruction to after again reaching default manual monitoring duration, return and perform the step whether described prompting loads presetting bit associated profile,
When receive refusal carry out the instruction of manual intervention after, judge whether video camera reaches Preset Time at the monitor event of current presetting bit, if, control the next presetting bit of camera switching to described default presetting bit, and return the step performing the Traffic Surveillance Video image gathered when described acquisition video camera is in default presetting bit.
4. the method according to any one of claim 1-3, is characterized in that, describedly sets up track of vehicle model according to described Traffic Surveillance Video image and comprises:
Detect the vehicle in described Traffic Surveillance Video image;
The vehicle detected is followed the tracks of, obtains the movement locus of described vehicle;
According to the movement locus of described vehicle, set up track of vehicle model.
5. the method according to any one of claim 1-3, is characterized in that, described track of vehicle model and described pavement structure model include deflection model and the Velicle motion velocity model of vehicle movement;
Describedly specifically to comprise according to track of vehicle model and the pavement structure model inspection traffic events set up in advance:
Deflection model in more described track of vehicle model and described pavement structure model and/or rate pattern, obtain deflection deviation and/or the velocity deviation of described track of vehicle model and described pavement structure model;
Judge whether described deflection deviation reaches first pre-conditioned, and/or, judge whether described velocity deviation reaches second pre-conditioned, to judge whether road surface traffic events occurs;
When the deviation of the deflection model in the deflection model in described track of vehicle model and described pavement structure model is greater than angle threshold, judge that vehicle occurs to drive in the wrong direction event;
When the deviation of the speed variables in the speed variables in described track of vehicle model and described pavement structure model is greater than First Speed threshold value, judge overspeed of vehicle event occurs;
When the deviation of the speed variables in the speed variables in described pavement structure model and described track of vehicle model is greater than second speed threshold value, judge automobile low-speed event occurs;
When the speed variables in track of vehicle model described in first time period continues to be less than third speed threshold value, judge that vehicle occurs stops event;
When the speed variables in track of vehicle model road surface exceeding a certain proportion of vehicle continues to be less than third speed threshold value, judge vehicle congestion event occurs;
When on road surface, at least the track of vehicle model of two vehicles occurs that the overlapping and time remaining of overlap is more than the second time period, judge vehicle crash event occurs.
6. the method according to any one of claim 1-3, is characterized in that, sets up all environmental models of car specifically comprise according to described Traffic Surveillance Video image:
Vehicle and surrounding environment part is chosen from described Traffic Surveillance Video image;
Car week environment contour feature figure is obtained according to described vehicle and surrounding environment part;
Set up car week environment DPM model according to described car week environment contour feature figure, described car all environment DPM model comprises root model p
0, the first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4, described model p
0for auto model, described first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4be respectively the environmental model on the four direction all around of vehicle.
7. method according to claim 6, is characterized in that, described according to described car week environmental model and the pavement structure model inspection traffic events set up in advance specifically comprise:
Vehicle environmental model is mated with described pavement structure model, obtains matching value;
When matching value is greater than predetermined threshold value, from all environment of car, be partitioned into car Zhou Yichang object;
The movement locus of described car Zhou Yichang object is followed the tracks of, sets up the locus model of car Zhou Yichang object;
When the medium velocity of the locus model of car Zhou Yichang object is 0, starts timing, after the residence time of car Zhou Yichang object reaches Preset Time, determining that described car Zhou Yichang object is for shedding thing.
8. a pick-up unit for traffic events, is characterized in that, comprising:
Acquiring unit, for obtaining the Traffic Surveillance Video image gathered when video camera is in default presetting bit;
Unit set up by model, for setting up track of vehicle model and/or all environmental models of car according to described Traffic Surveillance Video image;
Detecting unit, for according to described track of vehicle model and the pavement structure model inspection traffic events set up in advance, and/or, according to described car week environmental model and the pavement structure model inspection traffic events set up in advance; Described pavement structure model is the model set up according to monitor video image when there is not traffic events.
9. device according to claim 8, it is characterized in that, the method that described traffic incidents detection device adopts every road video camera to carry out automatic taking turn monitoring between multiple different presetting bit carries out the detection of traffic events, multiple different presetting bit residing for the video camera of every road can cover the transversal section on whole road surface, setting video camera is after the monitoring period of a presetting bit reaches Preset Time, and automatic alternate is monitored to next presetting bit.
10. device according to claim 8 or claim 9, it is characterized in that, described model is set up unit and is comprised:
Detection sub-unit, for detecting the vehicle in described Traffic Surveillance Video image;
Following the tracks of subelement, for following the tracks of the vehicle detected, obtaining the movement locus of described vehicle;
Subelement set up by track of vehicle model, for the movement locus according to described vehicle, sets up track of vehicle model.
11. devices according to claim 8 or claim 9, it is characterized in that, described track of vehicle model and described pavement structure model include deflection model and the Velicle motion velocity model of vehicle movement;
Described detecting unit comprises:
Relatively subelement, for the deflection model in more described track of vehicle model and described pavement structure model and/or rate pattern, obtains deflection deviation and/or the velocity deviation of described track of vehicle model and described pavement structure model;
Judgment sub-unit, for judging whether described deflection deviation reaches first pre-conditioned, and/or whether described velocity deviation reaches second pre-conditioned, to judge whether road surface traffic events occurs;
When the deviation of the deflection model in the deflection model in described track of vehicle model and described pavement structure model is greater than angle threshold, judge that vehicle occurs to drive in the wrong direction event;
When the deviation of the speed variables in the speed variables in described track of vehicle model and described pavement structure model is greater than First Speed threshold value, judge overspeed of vehicle event occurs;
When the deviation of the speed variables in the speed variables in described pavement structure model and described track of vehicle model is greater than second speed threshold value, judge automobile low-speed event occurs;
When the speed variables in track of vehicle model described in first time period continues to be less than third speed threshold value, judge that vehicle occurs stops event;
When the speed variables in track of vehicle model road surface exceeding a certain proportion of vehicle continues to be less than third speed threshold value, judge vehicle congestion event occurs;
When on road surface, at least the track of vehicle model of two vehicles occurs that the overlapping and time remaining of overlap is more than the second time period, judge vehicle crash event occurs.
12. devices according to claim 8 or claim 9, it is characterized in that, described model is set up unit and is comprised:
Choose subelement, for choosing vehicle and surrounding environment part from described Traffic Surveillance Video image;
Obtain subelement, for obtaining car week environment contour feature figure according to described vehicle and surrounding environment part;
Car week, subelement set up by environment DPM model, and for setting up car week environment DPM model according to described car week environment contour feature figure, described car all environment DPM model comprises root model p
0, the first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4, described model p
0for auto model, described first partial model p
1, the second partial model p
2, the 3rd partial model p
3with the 4th partial model p
4be respectively the environmental model on the four direction all around of vehicle.
13. devices according to claim 8 or claim 9, it is characterized in that, described detecting unit comprises:
Coupling subelement, for being mated with described pavement structure model by vehicle environmental model, obtains matching value;
Segmentation subelement, for when described matching value is greater than predetermined threshold value, is partitioned into car Zhou Yichang object from all environment of car;
Car Zhou Yichang object locus model sets up subelement, for following the tracks of the movement locus of described car Zhou Yichang object, sets up the locus model of car Zhou Yichang object;
Judge to shed thing subelement, when the medium velocity for the locus model when car Zhou Yichang object is 0, starts timing, after the residence time of car Zhou Yichang object reaches Preset Time, determining that described car Zhou Yichang object is for shedding thing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510182223.9A CN104809874B (en) | 2015-04-15 | 2015-04-15 | A kind of traffic incidents detection method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510182223.9A CN104809874B (en) | 2015-04-15 | 2015-04-15 | A kind of traffic incidents detection method and apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104809874A true CN104809874A (en) | 2015-07-29 |
| CN104809874B CN104809874B (en) | 2017-03-08 |
Family
ID=53694670
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510182223.9A Active CN104809874B (en) | 2015-04-15 | 2015-04-15 | A kind of traffic incidents detection method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104809874B (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105206055A (en) * | 2015-09-24 | 2015-12-30 | 深圳市哈工大交通电子技术有限公司 | Accident detection method for recognizing vehicle collision through traffic monitoring video |
| CN105355039A (en) * | 2015-10-23 | 2016-02-24 | 张力 | Road condition information processing method and equipment |
| 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 |
| CN106101641A (en) * | 2016-07-18 | 2016-11-09 | 中国科学院地理科学与资源研究所 | Video frequency monitoring system and monitoring method thereof |
| CN106340205A (en) * | 2016-09-30 | 2017-01-18 | 广东中星微电子有限公司 | Traffic monitoring method and traffic monitoring apparatus |
| CN106448162A (en) * | 2016-09-30 | 2017-02-22 | 广东中星微电子有限公司 | Road monitoring method and road monitoring device |
| CN107010086A (en) * | 2017-03-29 | 2017-08-04 | 宁夏凯速德科技有限公司 | High ferro line security control method and system |
| CN107274699A (en) * | 2017-06-30 | 2017-10-20 | 重庆交通大学 | Based on what position was shared passing method is cooperateed with without signal-control crossing vehicle certainly |
| CN107705595A (en) * | 2017-09-25 | 2018-02-16 | 航天科技控股集团股份有限公司 | It is a kind of that car-mounted terminal triggering warning device and triggering alarm method are automatically controlled based on recorder management platform |
| CN107742425A (en) * | 2017-09-29 | 2018-02-27 | 公安部交通管理科学研究所 | Abnormal road conditions rapid discovery method based on traffic surveillance videos |
| CN108028012A (en) * | 2015-09-17 | 2018-05-11 | 大众汽车有限公司 | For providing equipment, the method and computer program of the information on congestion end by vehicle-to-vehicle interface |
| CN108307315A (en) * | 2016-09-07 | 2018-07-20 | 北京嘀嘀无限科技发展有限公司 | A kind of processing method of traffic accident, server and mobile terminal |
| CN109035776A (en) * | 2018-08-23 | 2018-12-18 | 深圳大学 | A kind of accident detection method and system |
| CN109934075A (en) * | 2017-12-19 | 2019-06-25 | 杭州海康威视数字技术股份有限公司 | Accident detection method, apparatus, system and electronic equipment |
| CN110113579A (en) * | 2019-05-30 | 2019-08-09 | 浙江大华技术股份有限公司 | A kind of method and device tracking target object |
| CN111145378A (en) * | 2019-12-30 | 2020-05-12 | 中设设计集团股份有限公司 | Automatic video inspection system and method for road traffic incident |
| CN111402612A (en) * | 2019-01-03 | 2020-07-10 | 北京嘀嘀无限科技发展有限公司 | Traffic incident notification method and device |
| CN111640309A (en) * | 2020-05-22 | 2020-09-08 | 浙江工贸职业技术学院 | Swift vehicle detecting system |
| CN111753634A (en) * | 2020-03-30 | 2020-10-09 | 上海高德威智能交通系统有限公司 | Traffic incident detection method and device |
| CN112367475A (en) * | 2021-01-15 | 2021-02-12 | 上海闪马智能科技有限公司 | Traffic incident detection method and system and electronic equipment |
| CN112435470A (en) * | 2020-11-11 | 2021-03-02 | 宁波职业技术学院 | Traffic incident video detection system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090189979A1 (en) * | 2008-01-29 | 2009-07-30 | Navteq North America, Llc | Method for Providing Images of Traffic Incidents |
| CN102945603A (en) * | 2012-10-26 | 2013-02-27 | 青岛海信网络科技股份有限公司 | Method for detecting traffic event and electronic police device |
| JP2013161281A (en) * | 2012-02-06 | 2013-08-19 | Sumitomo Electric Ind Ltd | Traffic image acquisition device, traffic image acquisition method, and traffic image acquisition program |
| CN103366569A (en) * | 2013-06-26 | 2013-10-23 | 东南大学 | Method and system for snapshotting traffic violation vehicle in real time |
| CN103971521A (en) * | 2014-05-19 | 2014-08-06 | 清华大学 | Method and device for detecting road traffic abnormal events in real time |
-
2015
- 2015-04-15 CN CN201510182223.9A patent/CN104809874B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090189979A1 (en) * | 2008-01-29 | 2009-07-30 | Navteq North America, Llc | Method for Providing Images of Traffic Incidents |
| JP2013161281A (en) * | 2012-02-06 | 2013-08-19 | Sumitomo Electric Ind Ltd | Traffic image acquisition device, traffic image acquisition method, and traffic image acquisition program |
| CN102945603A (en) * | 2012-10-26 | 2013-02-27 | 青岛海信网络科技股份有限公司 | Method for detecting traffic event and electronic police device |
| CN103366569A (en) * | 2013-06-26 | 2013-10-23 | 东南大学 | Method and system for snapshotting traffic violation vehicle in real time |
| CN103971521A (en) * | 2014-05-19 | 2014-08-06 | 清华大学 | Method and device for detecting road traffic abnormal events in real time |
Non-Patent Citations (2)
| Title |
|---|
| 李国锋等: "带云台摄像机异常交通事件视频识别技术研究", 《公路交通科技(应用技术版)》 * |
| 王宁: "基于视频的车辆跟踪与交通事件检测", 《中国优秀博硕士学位论文全文数据库 (硕士) 信息科技辑》 * |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108028012A (en) * | 2015-09-17 | 2018-05-11 | 大众汽车有限公司 | For providing equipment, the method and computer program of the information on congestion end by vehicle-to-vehicle interface |
| CN108028012B (en) * | 2015-09-17 | 2021-04-23 | 大众汽车有限公司 | Device, method and data carrier for providing information about congestion ending via a vehicle-to-vehicle interface |
| CN105206055A (en) * | 2015-09-24 | 2015-12-30 | 深圳市哈工大交通电子技术有限公司 | Accident detection method for recognizing vehicle collision through traffic monitoring video |
| CN105206055B (en) * | 2015-09-24 | 2018-09-21 | 深圳市哈工大交通电子技术有限公司 | A kind of accident detection method of Traffic Surveillance Video identification vehicle collision |
| CN105355039A (en) * | 2015-10-23 | 2016-02-24 | 张力 | Road condition information processing method and equipment |
| 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 |
| CN106101641A (en) * | 2016-07-18 | 2016-11-09 | 中国科学院地理科学与资源研究所 | Video frequency monitoring system and monitoring method thereof |
| CN106101641B (en) * | 2016-07-18 | 2022-06-14 | 中国科学院地理科学与资源研究所 | Video monitoring system and monitoring method thereof |
| CN108307315B (en) * | 2016-09-07 | 2020-08-25 | 北京嘀嘀无限科技发展有限公司 | Traffic accident processing method, server and mobile terminal |
| CN108307315A (en) * | 2016-09-07 | 2018-07-20 | 北京嘀嘀无限科技发展有限公司 | A kind of processing method of traffic accident, server and mobile terminal |
| CN106340205A (en) * | 2016-09-30 | 2017-01-18 | 广东中星微电子有限公司 | Traffic monitoring method and traffic monitoring apparatus |
| CN106448162A (en) * | 2016-09-30 | 2017-02-22 | 广东中星微电子有限公司 | Road monitoring method and road monitoring device |
| CN107010086A (en) * | 2017-03-29 | 2017-08-04 | 宁夏凯速德科技有限公司 | High ferro line security control method and system |
| CN107274699A (en) * | 2017-06-30 | 2017-10-20 | 重庆交通大学 | Based on what position was shared passing method is cooperateed with without signal-control crossing vehicle certainly |
| CN107705595A (en) * | 2017-09-25 | 2018-02-16 | 航天科技控股集团股份有限公司 | It is a kind of that car-mounted terminal triggering warning device and triggering alarm method are automatically controlled based on recorder management platform |
| CN107742425A (en) * | 2017-09-29 | 2018-02-27 | 公安部交通管理科学研究所 | Abnormal road conditions rapid discovery method based on traffic surveillance videos |
| CN107742425B (en) * | 2017-09-29 | 2021-02-12 | 公安部交通管理科学研究所 | Method for rapidly discovering abnormal road conditions based on road monitoring video |
| CN109934075A (en) * | 2017-12-19 | 2019-06-25 | 杭州海康威视数字技术股份有限公司 | Accident detection method, apparatus, system and electronic equipment |
| CN109035776A (en) * | 2018-08-23 | 2018-12-18 | 深圳大学 | A kind of accident detection method and system |
| CN111402612A (en) * | 2019-01-03 | 2020-07-10 | 北京嘀嘀无限科技发展有限公司 | Traffic incident notification method and device |
| CN110113579B (en) * | 2019-05-30 | 2021-04-16 | 浙江大华技术股份有限公司 | Method and device for tracking target object |
| CN110113579A (en) * | 2019-05-30 | 2019-08-09 | 浙江大华技术股份有限公司 | A kind of method and device tracking target object |
| CN111145378A (en) * | 2019-12-30 | 2020-05-12 | 中设设计集团股份有限公司 | Automatic video inspection system and method for road traffic incident |
| CN111145378B (en) * | 2019-12-30 | 2021-06-25 | 华设设计集团股份有限公司 | Automatic video inspection system and method for road traffic incident |
| CN111753634A (en) * | 2020-03-30 | 2020-10-09 | 上海高德威智能交通系统有限公司 | Traffic incident detection method and device |
| CN111640309A (en) * | 2020-05-22 | 2020-09-08 | 浙江工贸职业技术学院 | Swift vehicle detecting system |
| CN112435470A (en) * | 2020-11-11 | 2021-03-02 | 宁波职业技术学院 | Traffic incident video detection system |
| CN112367475A (en) * | 2021-01-15 | 2021-02-12 | 上海闪马智能科技有限公司 | Traffic incident detection method and system and electronic equipment |
| CN112367475B (en) * | 2021-01-15 | 2021-03-30 | 上海闪马智能科技有限公司 | Traffic incident detection method and system and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104809874B (en) | 2017-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104809874A (en) | Traffic accident detection method and device | |
| CN103465906B (en) | A kind of parking area automatic train stop implementation method based on telepresenc | |
| CN103473951B (en) | A kind of Automatic Carpark Control System based on telepresenc | |
| US20230236602A1 (en) | Systems and Methods for Controlling an Autonomous Vehicle with Occluded Sensor Zones | |
| CN102332209B (en) | Automobile violation video monitoring method | |
| CN103770780B (en) | A kind of active safety systems of vehicles alarm shield device | |
| CN106327863A (en) | Method, device and system for detection and early warning of vehicle congestion | |
| CN108536149A (en) | A kind of automatic driving vehicle avoidance obstacle device and control method based on the paths Dubins | |
| CN202563344U (en) | Magnetic navigation and routing inspection robot | |
| CN112373485A (en) | Decision planning method for automatic driving vehicle considering interactive game | |
| CN104809887B (en) | A kind of retrograde detection method of vehicle on expressway and autoalarm | |
| CN105151043A (en) | Emergency avoidance system and method for unmanned automobile | |
| CN102541064A (en) | Magnetic navigation routing inspection robot | |
| CN104442566A (en) | Vehicle inside passenger dangerous state alarming device and alarming method | |
| CN105425757A (en) | Intelligent coke oven car equipment control system and method | |
| DE102014110206B4 (en) | Vehicle external environment detection device | |
| CN103313039B (en) | A kind of freeway tunnel entrance security prompt device on daytime and reminding method | |
| CN115603466A (en) | Ship shore power system based on artificial intelligence visual recognition and control method thereof | |
| CN202345598U (en) | Track change danger early-warning device for vehicles | |
| CN104157143A (en) | Illegal parking detection system and detection method thereof | |
| CN105574502A (en) | Automatic detection method for violation behaviors of self-service card sender | |
| CN103136514A (en) | Parking event detecting method based on double tracking system | |
| WO2019162447A1 (en) | Method for operating a door system, and door system | |
| CN103692974A (en) | Vehicle driving safety pre-warning method and system based on environmental monitoring | |
| CN115465293A (en) | Multi-sensor safety self-cognition and safety processing device and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20171120 Address after: 518057 the South District of Nanshan District high tech Industrial Park, Shenzhen City, Guangdong Province, R1-B Patentee after: Shenzhen Neusoft Software Co., Ltd. Address before: Hunnan rookie street Shenyang city Liaoning province 110179 No. 2 Patentee before: Neusoft Group Corporation |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20181213 Address after: 110179 2 Hsin street, Hunnan New District, Shenyang, Liaoning Patentee after: Neusoft Group Corporation Address before: 518057 Block R1-B, South District of Nanshan High-tech Industrial Park, Shenzhen City, Guangdong Province Patentee before: Shenzhen Neusoft Software Co., Ltd. |