CN110059578A - A kind of method and device of vehicle tracking - Google Patents
A kind of method and device of vehicle tracking Download PDFInfo
- Publication number
- CN110059578A CN110059578A CN201910239369.0A CN201910239369A CN110059578A CN 110059578 A CN110059578 A CN 110059578A CN 201910239369 A CN201910239369 A CN 201910239369A CN 110059578 A CN110059578 A CN 110059578A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- block diagram
- target vehicle
- target
- default
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
This application provides a kind of methods of vehicle tracking, which comprises the position that the target vehicle is identified by presetting recognizer obtains the first vehicle block diagram at the target vehicle current time；The position that default track algorithm starts to track the target vehicle is controlled by the default recognizer, obtains the second vehicle block diagram at the target vehicle current time；The repetitive rate of the first vehicle block diagram and the second vehicle block diagram is calculated, if the judgement repetitive rate is greater than or equal to preset threshold, the target vehicle block diagram at current time is determined according to the first vehicle block diagram and the second vehicle block diagram.Using method provided by the present application, continuity when tracking target vehicle can be effectively improved, the loss of target vehicle is avoided.Present invention also provides a kind of devices of vehicle tracking.
Description
Technical field
This application involves technical field of vehicle control more particularly to a kind of method and devices of vehicle tracking.
Background technique
With the continuous quickening of urbanization process, the intelligent of City road traffic system gradually becomes research hotspot, vehicle
An important component of the tracking technique as traffic monitoring system, also become an important research contents, it is existing
Have and generally uses track algorithm when tracking in technology to vehicle.
The track algorithm includes detection module and tracking module, needs to first pass through the detection module detection mesh of front end
Vehicle is marked, it is general to detect whether next frame predicted position is target vehicle by detection module, then reuse new detection knot
Fruit goes to update training set and then updates detection module, and generally chooses target vehicle region in training detection module and be positive
The peripheral region of sample, target vehicle is negative sample, and after detecting the target vehicle, tracking module is started to work, but is worked as
When the target vehicle is blocked, scene shown in FIG. 1 specifically may refer to, target vehicle is by obstacle vehicle in capturing the video
(can also be pedestrian, trees or other building facilities etc.) stops, and leads to that complete mesh cannot be obtained from the shooting video
Vehicle is marked, or may refer to scene shown in Fig. 2, the part vehicle body of target vehicle is blocked by trees in capturing the video at this time,
Cause the target vehicle shot in video to be separated into two parts, referring also to scene shown in Fig. 3, when unmanned plane or
The barriers such as when person installs pinnacled photographic device shooting video, the target vehicle in video counted, height-limiting bar block or
Separate, tracking module tends not to the position for being appropriately determined target vehicle when there is similar above situation, it is easy to target occurs
The loss of vehicle.Specifically the reason is that being mixed with institute in current positive sample after since target vehicle is blocked by barrier at this time
Barrier region is stated, the pollution of sample is caused, so that detection module is using the sample of the barrier as training sample,
Cause the detection module may be in the case where the target vehicle is not blocked completely, really by the barrier mistake
It is set to target vehicle, and then carries out the tracking to the barrier, there is a phenomenon where lose target vehicle.
Summary of the invention
In order to solve above-mentioned technical problem of the existing technology, this application provides a kind of method of vehicle tracking and dresses
It sets, continuity when tracking target vehicle can be effectively improved, avoid the loss of target vehicle.
The embodiment of the present application provides a kind of method of vehicle tracking, described applied to the target vehicle in tracking video
Method includes:
The position that the target vehicle is identified by presetting recognizer, obtains the first of the target vehicle current time
Vehicle block diagram；
The position that default track algorithm starts to track the target vehicle is controlled by the default recognizer, obtains institute
State the second vehicle block diagram at target vehicle current time；
Calculate the repetitive rate of the first vehicle block diagram and the second vehicle block diagram, if the judgement repetitive rate be greater than or
Equal to preset threshold, the target vehicle frame at current time is determined according to the first vehicle block diagram and the second vehicle block diagram
Figure.
Optionally, described to start to track the target vehicle by the default default track algorithm of recognizer control
Position includes:
The default recognizer by using the first vehicle block diagram as the initial value of the default track algorithm to control
The default track algorithm is made to start to track the position of the target vehicle.
Optionally, the repetitive rate is the repetition area and block diagram of the first vehicle block diagram and the second vehicle block diagram
The ratio of the gross area；The sum of the area that the block diagram gross area is equal to the first vehicle block diagram and the second vehicle block diagram subtracts
Remove the repetition area of the first vehicle block diagram and the second vehicle block diagram.
Optionally, if the repetitive rate is less than preset threshold, the method also includes:
Control the position that the default track algorithm stops tracking the target vehicle；
The first vehicle block diagram is determined as to the target vehicle block diagram at the current time.
Optionally, if the default recognizer can not recognize the target vehicle, and the default track algorithm energy
The second vehicle block diagram is enough obtained, or,
When the default track algorithm can not track the target vehicle, determine that the target vehicle is blocked by barrier,
The method also includes:
Start target carriage described in preset number frame video image before being blocked by the barrier according to the target vehicle
Offset information obtain the speed of service and traffic direction of the target vehicle；
According to the appearance position of the speed of service of the target vehicle and traffic direction prediction vehicle.
Optionally, the method also includes:
The default track algorithm is controlled when the target vehicle starts to be blocked by barrier to stop tracking the target
The position of vehicle；
Identify whether the target vehicle reappears by the default recognizer, and the target vehicle again
The position of the target vehicle is identified when appearance to obtain the first vehicle block diagram of the target vehicle.
Optionally, if judging, the target vehicle is reappeared, the method also includes:
The default track algorithm is controlled by the default recognizer start again at track the position of the target vehicle
It sets.
The embodiment of the present application also provides a kind of device of vehicle tracking, described device includes: first acquisition unit, second
Acquiring unit, determination unit；
The first acquisition unit, for identifying the position of the target vehicle by presetting recognizer, described in acquisition
The first vehicle block diagram at target vehicle current time；
The second acquisition unit starts described in tracking for controlling default track algorithm by the default recognizer
The position of target vehicle obtains the second vehicle block diagram at the target vehicle current time；
The determination unit judges for calculating the repetitive rate of the first vehicle block diagram and the second vehicle block diagram
If the repetitive rate is greater than or equal to preset threshold, determined according to the first vehicle block diagram and the second vehicle block diagram current
The target vehicle block diagram at moment.
Optionally, described device further include: first control unit and third acquiring unit；
The first control unit stops tracking the position of the target vehicle for controlling the default track algorithm；
The third acquiring unit, for the first vehicle block diagram to be determined as to the target carriage at the current time
Block diagram.
Optionally, described device further include: the second control unit and third acquiring unit；
Second control unit, for controlling the default tracking when the target vehicle starts to be blocked by barrier
Algorithm stops tracking the position of the target vehicle；
The third acquiring unit, for identifying whether the target vehicle goes out again by the default recognizer
It is existing, and when the target vehicle reappears the position of the target vehicle is identified to obtain described the of the target vehicle
One vehicle block diagram.
Herein described method has the advantage that
This application provides a kind of method of vehicle tracking, the method identifies the target carriage by default recognizer
Position, obtain the first vehicle block diagram at the target vehicle current time；It is default by the default recognizer control
Track algorithm starts to track the position of the target vehicle, obtains the second vehicle block diagram at the target vehicle current time；Meter
The repetitive rate of the first vehicle block diagram and the second vehicle block diagram is calculated, if the judgement repetitive rate is greater than or equal to default threshold
Value, the target vehicle block diagram at current time is determined according to the first vehicle block diagram and the second vehicle block diagram, i.e., is only existed
After the repetitive rate for the target vehicle block diagram that two kinds of algorithms obtain respectively reaches certain value, just show two algorithms for same vehicle
, and then determine the vehicle block diagram of the target vehicle.Using method provided by the present application, avoid merely with recognizer or
The problem of be easy to causeing target to lose when person is merely with track algorithm can effectively improve continuity when tracking target vehicle,
Avoid the loss of target vehicle.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The some embodiments recorded in application, for those of ordinary skill in the art, without creative efforts,
It can also be obtained according to these attached drawings other attached drawings.
Fig. 1 is schematic diagram of a scenario one provided by the present application；
Fig. 2 is schematic diagram of a scenario two provided by the present application；
Fig. 3 is schematic diagram of a scenario three provided by the present application；
Fig. 4 is a kind of flow chart of the method for vehicle tracking that the embodiment of the present application one provides；
Fig. 5 is the schematic diagram of a scenario four that the embodiment of the present application one provides；
Fig. 6 is the schematic diagram of a scenario five that the embodiment of the present application one provides；
Fig. 7 is the flow chart of the method for another vehicle tracking that the embodiment of the present application two provides；
Fig. 8 is a kind of schematic diagram of the device for vehicle tracking that the embodiment of the present application three provides.
Specific embodiment
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only this
Apply for a part of the embodiment, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art exist
Every other embodiment obtained under the premise of creative work is not made, shall fall in the protection scope of this application.
Embodiment one:
The embodiment of the present application one provides a kind of method of vehicle tracking, is specifically described with reference to the accompanying drawing.
Referring to fig. 4, which is a kind of flow chart of the method for vehicle tracking that the embodiment of the present application one provides.
The embodiment of the present application the described method comprises the following steps:
S401: the position of the target vehicle is identified by presetting recognizer, obtains the target vehicle current time
The first vehicle block diagram.
The recognizer, being sent into neural network model in video comprising target vehicle picture will be shot by, which referring to, carries out deeply
Degree study, the neural network model can for convolutional neural networks (Convolutional Neural Networks, CNN),
Recognition with Recurrent Neural Network (Recurrent Neural Network, RNN) etc., by carrying out deep learning to judge vehicle at this time
It whether is target vehicle and the real time position for obtaining the target vehicle.
Target vehicle is chosen from the shooting video, the equipment for obtaining the shooting video can be the bat of Near Ground
Take the photograph equipment, such as video, the shooting video of traffic camera or the shooting picture of roadside monitoring camera of automobile data recorder
Face, can also be the capture apparatus in high altitude operation, such as the unmanned plane with aerial photography function, and the application does not do specific limit to this
Fixed, the working condition of the capture apparatus can be at certain angle for the camera of totally stationary state or capture apparatus
The focal length of the camera of rotation or capture apparatus can be automatically adjusted according at a distance from target vehicle in degree, then or
Capture apparatus described in person is with the automobile data recorder in locomotivity such as vehicle or the unmanned plane with aerial photography function etc., this Shen
Please the working condition of the capture apparatus is not specifically limited.
It should be noted that obtaining the first vehicle block diagram at the target vehicle current time by presetting recognizer
When, the target vehicle block diagram not being blocked completely can be therefrom obtained from several frame video images, the application is for the mesh
The type of mark vehicle is not specifically limited with quantity, for convenience of explanation, below by taking the number of the target vehicle is 1 as an example
It is illustrated.
The position for identifying target vehicle described in shooting video in real time by presetting recognizer, obtains the target vehicle
The first vehicle block diagram at current time, since the capture apparatus is during shooting video, the target vehicle is in fortune
Dynamic state, i.e., the relative distance of the described target vehicle and the capture apparatus are simultaneously not fixed, and cause the target vehicle in institute
It states the morphological feature shown in shooting video not to be fixed and invariable, therefore the area of the first vehicle block diagram of different moments
It is possible that variation, and the first vehicle block diagram is used to identify the position of current goal vehicle, is not the target
The exact outline of vehicle, therefore allow the target vehicle to there is part and be not located in the first vehicle block diagram, while part
Environmental area around target vehicle also may be embodied in the first vehicle block diagram.
The application is not especially limited the default recognizer.
S402: starting to track the position of the target vehicle by the default default track algorithm of recognizer control,
Obtain the second vehicle block diagram at the target vehicle current time.
Control the position that default track algorithm starts to track the target vehicle by presetting recognizer so that it is default with
Track algorithm provides the accurate of target vehicle in start time and default recognizer target having the same, for default track algorithm
Positive sample.
While identifying the position of the target vehicle by default recognizer by default track algorithm in real time with
Track shoots the position of target vehicle described in video, obtains the second vehicle block diagram at the target vehicle current time, similarly not
The area of the second vehicle block diagram in the same time is it is possible that variation, specific reason are same as above, and details are not described herein.
And the second vehicle block diagram is equally used for the position of mark current goal vehicle, is not the target vehicle
Exact outline, therefore allow the target vehicle to there is part and be not located in the second vehicle block diagram, while partial target
The environmental area of vehicle periphery also may be embodied in the second vehicle block diagram.
S403: calculating the repetitive rate of the first vehicle block diagram and the second vehicle block diagram, if the judgement repetitive rate
More than or equal to preset threshold, the target carriage at current time is determined according to the first vehicle block diagram and the second vehicle block diagram
Block diagram.
Since the default recognizer is identical with the target vehicle of the default track algorithm, completely in the state of ideal
The first vehicle block diagram should repeat completely with the second vehicle block diagram, but actual tracking during due to it is possible that
The interference of barrier or there are two types of algorithm itself levels of precision difference two cause error, lead to the first vehicle block diagram
It can not be completely coincident with the second vehicle block diagram, such as may refer to schematic diagram of a scenario shown in fig. 5, the repetitive rate can
To be calculated using following formula:
η in formula (1)1Indicate the repetitive rate, S0Indicate the weight of the first vehicle block diagram Yu the second vehicle block diagram
Coverlay product, S1Indicate the area of the first vehicle block diagram, S2Indicate the area of the second vehicle block diagram.The i.e. described repetitive rate
For the ratio of the repetition area of the first vehicle block diagram and the second vehicle block diagram and the block diagram gross area；The total face of block diagram
The sum of the area that product is equal to the first vehicle block diagram and the second vehicle block diagram subtracts the first vehicle block diagram and described
The repetition area of second vehicle block diagram.
Further, when the first repetitive rate of the first vehicle block diagram and the second vehicle block diagram is less than preset threshold
When, show the default recognizer described at this time and the default track algorithm when carrying out the tracking to the target vehicle at least
The main reason for having one tracking target selection mistake, mistake generation occur is that target vehicle is blocked very by barrier section
To being to block completely, such as may refer to schematic diagram of a scenario shown in fig. 6, when vehicle passes through from tree is lower, track algorithm is preset
There is mistake in tracking target, leads to the practical position block diagram for becoming tree of the second vehicle block diagram, the position with target vehicle
There is obvious deviation, so that first repetitive rate is almost nil.
It should be noted that above steps is intended merely to the convenient explanation to herein described method, do not constitute pair
In the restriction of herein described method.
The embodiment of the present application provides a kind of method of vehicle tracking, and the method passes through described in default recognizer identification
The position of target vehicle obtains the first vehicle block diagram at the target vehicle current time；It is pre- by default recognizer control
If track algorithm tracks the position of the target vehicle, the second vehicle block diagram at current time described in the target vehicle is obtained；
Calculate the first repetitive rate of the first vehicle block diagram and the second vehicle block diagram, if judgement first repetitive rate be greater than or
Equal to preset threshold, the repeating part of the first vehicle block diagram and the second vehicle block diagram is determined as the current time
Target vehicle block diagram, first repetitive rate is that the first vehicle block diagram and the second vehicle block diagram repeat area and institute
State the ratio of the sum of the first vehicle block diagram and the second vehicle block diagram area, i.e., the target only obtained respectively in two kinds of algorithms
After the repetitive rate of vehicle block diagram reaches certain value, just show that for same target vehicle, lap is determined as two algorithms
The vehicle block diagram of the target vehicle, to obtain the position of target vehicle；Track the target vehicle block diagram.Utilize the application reality
The method for applying example offer be easy to cause asking for target loss when avoiding merely with recognizer or merely with track algorithm
Topic can effectively improve continuity when tracking target vehicle, avoid the loss of target vehicle.
Embodiment two:
The embodiment of the present application two additionally provides the method for another vehicle tracking, can be applied to the target vehicle and is hindered
Object is hindered to block completely so that illustrating with reference to the accompanying drawing from the shooting video the case where temporary extinction.
Referring to Fig. 7, which is the flow chart of the method for another vehicle tracking that the embodiment of the present application two provides.
The method also includes following steps:
S701: mesh described in preset number frame video image before being blocked completely according to the target vehicle by the barrier
The offset information of mark vehicle obtains the speed of service and traffic direction of the target vehicle；According to the operation of target vehicle speed
The appearance position of degree and traffic direction prediction vehicle.
If the default recognizer can not recognize the target vehicle, and the default track algorithm can obtain institute
The second vehicle block diagram is stated, shows that target vehicle is partly or entirely blocked by barrier at this time, leads to target in video image vehicle
Presentation it is imperfect, the default recognizer can not recognize the vehicle, and even if stating default track algorithm can track
Target vehicle, the target of actual tracking at this time may be barrier, i.e., default track algorithm at this time tracks target accuracy
It can not ensure.
If the default track algorithm can not track the target vehicle, determine that the target vehicle is complete by barrier
It blocks.It should be noted that such situation is equally applicable to target vehicle when sailing out of video image scope, i.e. target carriage
Although not being blocked but still track algorithm tracking can not be predetermined.
When there is above situation, it can be determined that the target vehicle is blocked by barrier, acquisition can be passed through at this time
Start the offset of target vehicle described in preset number frame video image before being blocked by the barrier according to the target vehicle
The speed of service and traffic direction of target vehicle described in acquisition of information, such as obtain target vehicle and start to be blocked by the barrier
Front cross frame video image, extracts the range ability of target vehicle in two frame video images, and according to two frame video images when
Between interval determine the speed of service and range ability of target vehicle.Then, according to the speed of service and fortune of the target vehicle
The appearance position of line direction prediction vehicle.
S702: the control default track algorithm stops tracking the position of the target vehicle.
Due to the track algorithm, when having generally comprised detection module and tracking module, then having carried out the tracking of target vehicle,
Need to first pass through front end detection module detection target vehicle, generally by detection module detection next frame predicted position whether be
Then target vehicle reuses new testing result and goes to update training set and then update detection module, and in training detection module
When generally to choose target vehicle region be positive sample, the peripheral region of target vehicle is negative sample, when detecting the mesh
After marking vehicle, during tracking module is started to work, and hereafter detection module would not get involved in tracking.Due to the target carriage
Gradually blocked by barrier until disappearing completely from the shooting video is a progressive formation, understand gradually it is current just
The barrier region is mixed in sample, causes the pollution of sample, causes the detection module may be in the mesh
In the case that mark vehicle is not blocked completely, the barrier mistake is determined as target vehicle, it is therefore desirable to described in stopping
Default track algorithm, avoids track algorithm that barrier mistake is determined as target vehicle.
S703: identify whether target vehicle reappears by default recognizer, and when target vehicle reappears
The position of target vehicle is identified to obtain the first vehicle block diagram of target vehicle.
S704: if judging, the target vehicle is reappeared, and controls the default track algorithm by default recognizer
Start again at the position for tracking the target vehicle.
It prompts the track algorithm target vehicle to occur by the default recognizer, is continued with realizing with two kinds of algorithms
Track same target vehicle.
It, can be when target vehicle be obscured by an obstacle completely still through institute using method provided by the embodiments of the present application
The position of default recognizer prediction target vehicle is stated to obtain the target vehicle occurred again in time, and stop described pre-
If track algorithm, avoid the default track algorithm mistake is determined as the target vehicle for the barrier, in target carriage
When occurring again, the default track algorithm can also be controlled by the default recognizer and restarted, it can be effective
Continuity when tracking target vehicle is improved, the loss of target vehicle is further avoided.
Embodiment three:
The method of the vehicle tracking provided based on the above embodiment, the embodiment of the present application three additionally provide a kind of vehicle tracking
Device, illustrate with reference to the accompanying drawing.
Referring to Fig. 8, which is a kind of structure chart of the device for vehicle tracking that the embodiment of the present application three provides.
The embodiment of the present application described device includes: first acquisition unit 801, second acquisition unit 802, determination unit 803
With the first tracking cell 804.
The first acquisition unit 801 obtains institute for identifying the position of the target vehicle by presetting recognizer
State the first vehicle block diagram at target vehicle current time.
The second acquisition unit 802 starts to track for controlling default track algorithm by the default recognizer
The position of the target vehicle obtains the second vehicle block diagram at the target vehicle current time.
The determination unit 803 is sentenced for calculating the repetitive rate of the first vehicle block diagram and the second vehicle block diagram
If the disconnected repetitive rate is greater than or equal to preset threshold, worked as according to the first vehicle block diagram and the second vehicle block diagram determination
The target vehicle block diagram at preceding moment.
In one possible implementation, described device further include: first control unit and third acquiring unit.
The first control unit stops tracking the position of the target vehicle for controlling the default track algorithm.
The third acquiring unit, for the first vehicle block diagram to be determined as to the target carriage at the current time
Block diagram.
In alternatively possible implementation, described device further include: the second control unit and the 4th acquiring unit.
Second control unit, for controlling the default tracking when the target vehicle starts to be blocked by barrier
Algorithm stops tracking the position of the target vehicle.
4th acquiring unit, for identifying whether the target vehicle goes out again by the default recognizer
It is existing, and when the target vehicle reappears the position of the target vehicle is identified to obtain described the of the target vehicle
One vehicle block diagram.
The embodiment of the present application provides a kind of device of vehicle tracking, presets recognizer identification by first acquisition unit
The position of the target vehicle obtains the first vehicle block diagram at the target vehicle current time；It is pre- by second acquisition unit
If recognizer controls the position that default track algorithm tracks the target vehicle, current time described in the target vehicle is obtained
The second vehicle block diagram；Calculate the first vehicle block diagram and the second vehicle block diagram by determination unit first repeats
Rate, if judgement first repetitive rate is greater than or equal to preset threshold, by the first vehicle block diagram and the second vehicle frame
The repeating part of figure is determined as the target vehicle block diagram at the current time, and first repetitive rate is the first vehicle block diagram
The ratio of the sum of area and the first vehicle block diagram and the second vehicle block diagram area is repeated with the second vehicle block diagram,
I.e. only after the repetitive rate for the target vehicle block diagram that two kinds of algorithms obtain respectively reaches certain value, just show that two algorithms are directed to
Lap is determined as the vehicle block diagram of the target vehicle by same target vehicle, to obtain the position of target vehicle；Tracking
The target vehicle block diagram.Using device provided by the embodiments of the present application, avoid merely with recognizer or merely with
The problem of be easy to causeing target to lose when track algorithm can effectively improve continuity when tracking target vehicle, avoid target carriage
Loss.
It should be appreciated that in this application, " at least one (item) " refers to one or more, and " multiple " refer to two or two
More than a."and/or" indicates may exist three kinds of relationships, for example, " A and/or B " for describing the incidence relation of affiliated partner
It can indicate: only exist A, only exist B and exist simultaneously tri- kinds of situations of A and B, wherein A, B can be odd number or plural number.Word
Symbol "/" typicallys represent the relationship that forward-backward correlation object is a kind of "or"." at least one of following (a) " or its similar expression, refers to
Any combination in these, any combination including individual event (a) or complex item (a).At least one of for example, in a, b or c
(a) can indicate: a, b, c, " a and b ", " a and c ", " b and c ", or " a and b and c ", and wherein a, b, c can be individually, can also
To be multiple.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device reality
For applying example, since it is substantially similar to the method embodiment, so describing fairly simple, related place is referring to embodiment of the method
Part explanation.The apparatus embodiments described above are merely exemplary, wherein described be used as separate part description
Unit and module may or may not be physically separated.Furthermore it is also possible to select it according to the actual needs
In some or all of unit and module achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying
In the case where creative work, it can understand and implement.
The above is only the specific embodiment of the application, it is noted that for the ordinary skill people of the art
For member, under the premise of not departing from the application principle, several improvements and modifications can also be made, these improvements and modifications are also answered
It is considered as the protection scope of the application.
Claims (10)
1. a kind of method of vehicle tracking, which is characterized in that applied to the target vehicle in tracking video, which comprises
The position that the target vehicle is identified by presetting recognizer, obtains first vehicle at the target vehicle current time
Block diagram；
The position that default track algorithm starts to track the target vehicle is controlled by the default recognizer, obtains the mesh
Mark the second vehicle block diagram at vehicle current time；
The repetitive rate of the first vehicle block diagram and the second vehicle block diagram is calculated, if the judgement repetitive rate is greater than or equal to
Preset threshold determines the target vehicle block diagram at current time according to the first vehicle block diagram and the second vehicle block diagram.
2. the method for vehicle tracking according to claim 1, which is characterized in that described to pass through the default recognizer control
Making the position that default track algorithm starts to track the target vehicle includes:
The default recognizer by using the first vehicle block diagram as the initial value of the default track algorithm to control
Default track algorithm is stated to start to track the position of the target vehicle.
3. the method for vehicle tracking according to claim 1, which is characterized in that the repetitive rate is the first vehicle frame
The ratio of the repetition area and the block diagram gross area of figure and the second vehicle block diagram；The block diagram gross area is equal to first vehicle
The sum of the area of block diagram and the second vehicle block diagram subtracts the weight of the first vehicle block diagram and the second vehicle block diagram
Coverlay product.
4. the method for vehicle tracking according to claim 1, which is characterized in that if the repetitive rate is less than preset threshold,
The method also includes:
Control the position that the default track algorithm stops tracking the target vehicle；
The first vehicle block diagram is determined as to the target vehicle block diagram at the current time.
5. the method for vehicle tracking according to claim 1, which is characterized in that if the default recognizer can not identify
To the target vehicle, and the default track algorithm can obtain the second vehicle block diagram, or,
When the default track algorithm can not track the target vehicle, determine that the target vehicle is blocked by barrier, it is described
Method further include:
Start target vehicle described in preset number frame video image before being blocked by the barrier according to the target vehicle
Offset information obtains the speed of service and traffic direction of the target vehicle；
According to the appearance position of the speed of service of the target vehicle and traffic direction prediction vehicle.
6. the method for vehicle tracking according to claim 5, which is characterized in that the method also includes:
The default track algorithm is controlled when the target vehicle starts to be blocked by barrier to stop tracking the target vehicle
Position；
It identifies whether the target vehicle reappears by the default recognizer, and is reappeared in the target vehicle
When identify the position of the target vehicle to obtain the first vehicle block diagram of the target vehicle.
7. the method for vehicle tracking according to claim 6, which is characterized in that if judging, the target vehicle goes out again
It is existing, the method also includes:
The default track algorithm is controlled by the default recognizer start again at track the position of the target vehicle.
8. a kind of device of vehicle tracking, which is characterized in that described device include: first acquisition unit, second acquisition unit, really
Order member；
The first acquisition unit obtains the target for identifying the position of the target vehicle by presetting recognizer
The first vehicle block diagram at vehicle current time；
The second acquisition unit starts to track the target for controlling default track algorithm by the default recognizer
The position of vehicle obtains the second vehicle block diagram at the target vehicle current time；
The determination unit, for calculating the repetitive rate of the first vehicle block diagram and the second vehicle block diagram, if judgement institute
Repetitive rate is stated more than or equal to preset threshold, current time is determined according to the first vehicle block diagram and the second vehicle block diagram
Target vehicle block diagram.
9. the device of vehicle tracking according to claim 8, which is characterized in that described device further include: the first control is single
Member and third acquiring unit；
The first control unit stops tracking the position of the target vehicle for controlling the default track algorithm；
The third acquiring unit, for the first vehicle block diagram to be determined as to the target vehicle frame at the current time
Figure.
10. the method for vehicle tracking according to claim 8, which is characterized in that described device further include: the second control is single
Member and the 4th acquiring unit；
Second control unit, for controlling the default track algorithm when the target vehicle starts to be blocked by barrier
Stop tracking the position of the target vehicle；
4th acquiring unit, for identifying whether the target vehicle reappears by the default recognizer, and
The position of the target vehicle is identified when the target vehicle reappears to obtain first vehicle of the target vehicle
Block diagram.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910239369.0A CN110059578A (en) | 2019-03-27 | 2019-03-27 | A kind of method and device of vehicle tracking |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910239369.0A CN110059578A (en) | 2019-03-27 | 2019-03-27 | A kind of method and device of vehicle tracking |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110059578A true CN110059578A (en) | 2019-07-26 |
Family
ID=67317514
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910239369.0A Pending CN110059578A (en) | 2019-03-27 | 2019-03-27 | A kind of method and device of vehicle tracking |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110059578A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111046752A (en) * | 2019-11-26 | 2020-04-21 | 上海兴容信息技术有限公司 | Indoor positioning method and device, computer equipment and storage medium |
| CN112200126A (en) * | 2020-10-26 | 2021-01-08 | 上海盛奕数字科技有限公司 | Method for identifying limb shielding gesture based on artificial intelligence running |
| CN112766216A (en) * | 2021-01-29 | 2021-05-07 | 深圳市捷顺科技实业股份有限公司 | Method and device for detecting vehicle driving into parking space, electronic equipment and storage medium |
| CN112885108A (en) * | 2020-12-23 | 2021-06-01 | 爱泊车美好科技有限公司 | Vehicle change detection method and system on parking space based on deep learning algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101986348A (en) * | 2010-11-09 | 2011-03-16 | 上海电机学院 | Visual target identification and tracking method |
| CN104008371A (en) * | 2014-05-22 | 2014-08-27 | 南京邮电大学 | Regional suspicious target tracking and recognizing method based on multiple cameras |
| CN105869186A (en) * | 2016-04-23 | 2016-08-17 | 上海大学 | FPGA-based mass center correction and status differentiation particle filtering algorithm |
-
2019
- 2019-03-27 CN CN201910239369.0A patent/CN110059578A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101986348A (en) * | 2010-11-09 | 2011-03-16 | 上海电机学院 | Visual target identification and tracking method |
| CN104008371A (en) * | 2014-05-22 | 2014-08-27 | 南京邮电大学 | Regional suspicious target tracking and recognizing method based on multiple cameras |
| CN105869186A (en) * | 2016-04-23 | 2016-08-17 | 上海大学 | FPGA-based mass center correction and status differentiation particle filtering algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 余铎 等: "基于视觉的移动机器人目标跟踪方法", 《仪器仪表学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111046752A (en) * | 2019-11-26 | 2020-04-21 | 上海兴容信息技术有限公司 | Indoor positioning method and device, computer equipment and storage medium |
| CN112200126A (en) * | 2020-10-26 | 2021-01-08 | 上海盛奕数字科技有限公司 | Method for identifying limb shielding gesture based on artificial intelligence running |
| CN112885108A (en) * | 2020-12-23 | 2021-06-01 | 爱泊车美好科技有限公司 | Vehicle change detection method and system on parking space based on deep learning algorithm |
| CN112885108B (en) * | 2020-12-23 | 2022-02-01 | 爱泊车美好科技有限公司 | Vehicle change detection method and system on parking space based on deep learning algorithm |
| CN112766216A (en) * | 2021-01-29 | 2021-05-07 | 深圳市捷顺科技实业股份有限公司 | Method and device for detecting vehicle driving into parking space, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107134145B (en) | Road side parking management device, system and method based on multi-type image acquisition | |
| CN110059578A (en) | A kind of method and device of vehicle tracking | |
| CN109817014B (en) | Roadside inspection parking charging method based on mobile video and high-precision positioning | |
| KR20200064873A (en) | Method for detecting a speed employing difference of distance between an object and a monitoring camera | |
| CN106571039A (en) | Automatic snapshot system for highway traffic offence | |
| CN107305627A (en) | A kind of automobile video frequency monitoring method, server and system | |
| CN111126235A (en) | Method and device for detecting and processing illegal berthing of ship | |
| CN105894817A (en) | Evidence taking method of vehicle illegal parking and device thereof | |
| CN102673979B (en) | Method and device for judging deviation of conveying belt | |
| CN111985365A (en) | Straw burning monitoring method and system based on target detection technology | |
| CN106503622A (en) | A kind of vehicle antitracking method and device | |
| CN103376112A (en) | Elevated road auxiliary judging system, navigation equipment with same and navigation method thereof | |
| KR102061264B1 (en) | Unexpected incident detecting system using vehicle position information based on C-ITS | |
| CN109446926A (en) | A kind of traffic monitoring method and device, electronic equipment and storage medium | |
| CN113763484A (en) | Ship target positioning and speed estimation method based on video image analysis technology | |
| CN115909240A (en) | Road congestion detection method based on lane line and vehicle identification | |
| CN112200092A (en) | Intelligent smoking detection method based on variable-focus movement of dome camera | |
| CN110264729A (en) | A kind of asynchronous capture System and method for of license plate | |
| CN111383248A (en) | Method and device for judging red light running of pedestrian and electronic equipment | |
| CN111429723B (en) | Communication and perception data fusion method based on road side equipment | |
| CN113591643A (en) | Underground vehicle station entering and exiting detection system and method based on computer vision | |
| CN110070724A (en) | A kind of video monitoring method, device, video camera and image information supervisory systems | |
| CN112365740A (en) | Alarm display method and device | |
| CN112406700B (en) | Blind area early warning system based on upper and lower binocular vision analysis range finding | |
| CN112601021B (en) | Method and system for processing monitoring video of network camera |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190726 |
|
| RJ01 | Rejection of invention patent application after publication |