CN109859240A - Video object tracking method and apparatus, vehicle - Google Patents
Video object tracking method and apparatus, vehicle Download PDFInfo
- Publication number
- CN109859240A CN109859240A CN201711243752.0A CN201711243752A CN109859240A CN 109859240 A CN109859240 A CN 109859240A CN 201711243752 A CN201711243752 A CN 201711243752A CN 109859240 A CN109859240 A CN 109859240A
- Authority
- CN
- China
- Prior art keywords
- identification
- frame image
- tracking
- registration
- video
- 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
Abstract
The invention discloses a kind of Video object tracking method and apparatus, vehicle, the described method comprises the following steps: obtaining the video information of current scene in real time, and carry out Object identifying to each frame image in video information to obtain recognition result；The registration between the identification object of adjacent two frame is calculated according to the corresponding recognition result of previous frame image and the corresponding recognition result of current frame image；Registration is judged, and pursuit gain is updated according to judging result；Pursuit gain is judged；When pursuit gain is greater than the first tracking threshold value, judge that target vehicle tracking is effective.Tracking of the invention can adapt to Object identifying and the tracking of the very big detection scene of range by comparing the registration of the identified object of two continuous frames video.
Description
Technical field
The present invention relates to object tracing technique field, in particular to a kind of Video object tracking method, a kind of the video object
Tracking device and a kind of vehicle with the Video object tracking device.
Background technique
Object identifying is the important component of video image processing, is realized on this basis to image tracing.To image tracing
Result can feed back to Object identifying, improve the confidence level of Object identifying.
In the related technology, method for tracing object is: each moving target is in previous frame in detection motion target tracking database
Area in region and Moving Objects reference database where in video image where each Moving Objects of current frame video image
Domain whether there is intersection；Where detecting the moving target in motion target tracking database in previous frame video image
Region and Moving Objects reference database in current frame video image Moving Objects where there are intersections in region
In the case of, calculate the area of intersection；Judge whether area is greater than the first preset threshold；Judging that it is default that area is greater than first
In the case where threshold value, the tracking particle information of moving target is updated；Movement is calculated according to updated tracking particle information
Region of the target where in current frame video image, to realize the tracking of the moving target.
But whether the foundation judged in the above method is greater than the first preset threshold as area, this is for monitoring scene
In movement detection be it is feasible, since range is little in monitoring scene, so threshold value is also easier to determine.But for vehicle
Detection scene for be it is inadequate, range detected is from less than 1m to nearly 100m range, and target object (vehicle) is in video
Shared area change is very big in image.Although the object in sequential frame image has the part of overlapping, the area of lap
Variation range it is very big.Occupied area is big on hand for the object for including in successive frame, and occupied area is small at a distance, thus same knowledge
The judgment threshold of the overlapping area of other object on hand is not suitable for the tracking of distant object, and threshold value at a distance is not suitable for yet
The nearby tracking of object.
Summary of the invention
The present invention is directed at least solve one of the technical problem in above-mentioned technology to a certain extent.For this purpose, of the invention
First purpose is to propose a kind of Video object tracking method, by comparing the coincidence of the identified object of two continuous frames video
Degree, can adapt to Object identifying and the tracking of the very big detection scene of range.
Second object of the present invention is to propose a kind of non-transitorycomputer readable storage medium.
Third object of the present invention is to propose a kind of Video object tracking device.
Fourth object of the present invention is to propose a kind of vehicle.
In order to achieve the above objectives, a kind of Video object tracking method that first aspect present invention embodiment proposes, including with
Lower step: the video information of current scene is obtained in real time, and Object identifying is carried out to each frame image in the video information
To obtain recognition result；It is calculated according to the corresponding recognition result of previous frame image and the corresponding recognition result of current frame image adjacent
Registration between the identification object of two frames, and pursuit gain is updated according to the registration；The pursuit gain is judged；When
When the pursuit gain is greater than the first tracking threshold value, judge that the identification is effective to image tracing.
Video object tracking method according to an embodiment of the present invention obtains the video information of current scene in real time, and to view
Each frame image in frequency information carries out Object identifying to obtain recognition result, according to the corresponding recognition result of previous frame image and
The corresponding recognition result of current frame image calculate adjacent two frame identification object between registration, and according to registration update with
Track value, and pursuit gain is judged, when pursuit gain is greater than the first tracking threshold value, judge that identification is effective to image tracing.The party
Method can adapt to the object knowledge of the very big detection scene of range by comparing the registration of the identified object of two continuous frames video
Not and track.
In addition, the Video object tracking method proposed according to that above embodiment of the present invention can also have following additional skill
Art feature:
According to one embodiment of present invention, corresponding also according to previous frame image after the identification object track is effective
Recognition result and the corresponding recognition result of current frame image obtain the direction of motion and motion profile of the identification object, and root
The direction of extinction of the identification object is prejudged according to the direction of motion and motion profile of the identification object, and is being prejudged
It sets up and when the pursuit gain is less than the second tracking threshold value, judge identification object disappearance.
According to one embodiment of present invention, the recognition result is indicated with rectangle frame.
According to one embodiment of present invention, the registration calculates acquisition according to the following formula:
Registration=α × S3/S1+(1-α)×S3/S2,
Wherein, S1 is the area of the corresponding rectangle frame of the previous frame image, and S2 is the corresponding square of the current frame image
The area of shape frame, S3 are the corresponding rectangle frame of previous frame image rectangle frame intersection corresponding with the current frame image
Area, α is constant, and 0≤α≤1.
According to one embodiment of present invention, pursuit gain is updated according to the registration, comprising: carry out to the registration
Judgement, and search in preset tracking object list with the presence or absence of the identification object；If the registration is greater than registration
The threshold value and identification object is in tracking object list, then add up on the basis of initial pursuit gain, until it is described with
Track value reaches pre-set limit；If the registration is less than or equal to registration threshold value and the identification object is in tracking object column
In table, then successively decrease on the basis of initial pursuit gain, until being deleted in the tracking object list when pursuit gain is equal to zero
Except the corresponding entry of the identification object；If the registration is greater than registration threshold value and the identification object is not in tracking
In list object, then entry is increased newly in the tracking object list, and the pursuit gain of the newly-increased entry is set one.
According to one embodiment of present invention, corresponding according to the corresponding recognition result of previous frame image and current frame image
Recognition result obtains the direction of motion and motion profile of the identification object, comprising: by the corresponding rectangle of the previous frame image
The center of the center of frame rectangle frame corresponding with current frame image carries out line, to obtain the motion profile of the identification object；
When the corresponding rectangle frame of previous frame image rectangle frame corresponding with current frame image the line of centres and video collector just
Angle between apparent direction is greater than the first predetermined angle and the corresponding rectangle frame of the previous frame image is corresponding with current frame image
Rectangle frame or so distribution when, the direction of motion of the identification object is laterally；When the corresponding rectangle frame of the previous frame image
Angle between the line of centres and the positive apparent direction of video collector of rectangle frame corresponding with current frame image is default less than second
The direction of motion of angle, the identification object is longitudinal, wherein second predetermined angle is less than first predetermined angle.
According to one embodiment of present invention, according to the direction of motion of the target vehicle and motion profile to the identification
The direction of extinction of object is prejudged, comprising: when the direction of motion of the identification object is lateral, if the identification object
Motion profile have moved into preset Video Edge region, then prejudge establishment；When the direction of motion of the identification object is longitudinal
When, if the motion profile of the identification object has moved into default vanishing line and the corresponding rectangle frame size of the current frame image
For preset minimum identification frame, then establishment is prejudged；When the direction of motion of the identification object is longitudinal, if the present frame
The corresponding rectangle frame of image then prejudges establishment having a size of preset maximum identification frame.
In order to achieve the above objectives, second aspect of the present invention embodiment proposes a kind of non-transitory computer-readable storage medium
Matter is stored thereon with computer program, which realizes above-mentioned Video object tracking method when being executed by processor.
The non-transitorycomputer readable storage medium of the embodiment of the present invention, by executing the computer program of its storage,
By comparing the registration of the identified object of two continuous frames video, the Object identifying of the very big detection scene of range can adapt to
And tracking.
In order to achieve the above objectives, third aspect present invention embodiment proposes a kind of Video object tracking device, including deposits
Reservoir, processor and it is stored in the Video object tracking program that can be run on the memory and on the processor, it is described
The control program of Video object tracking device realizes the step of above-mentioned Video object tracking method when being executed by the processor.
The Video object tracking device of the embodiment of the present invention, by comparing the coincidence of the identified object of two continuous frames video
Degree, can adapt to Object identifying and the tracking of the very big detection scene of range.
In order to achieve the above objectives, fourth aspect present invention embodiment proposes a kind of vehicle comprising above-mentioned video pair
Image tracing device.
The vehicle of the embodiment of the present invention, by above-mentioned Video object tracking device, by comparing two continuous frames video
The registration of identified object can adapt to Object identifying and the tracking of the very big detection scene of range.
Detailed description of the invention
Fig. 1 is the flow chart of Video object tracking method according to an embodiment of the present invention；
Fig. 2 is first frame rectangle frame when closely identifying in Video object tracking method according to an embodiment of the invention
Big logotype；
Fig. 3 is the second frame rectangle frame when closely identifying in Video object tracking method according to an embodiment of the invention
Big logotype；
N-th frame rectangle frame when Fig. 4 is Video object tracking method medium and long distance according to an embodiment of the invention identification
Big logotype；
(n+1)th frame rectangle when Fig. 5 is Video object tracking method medium and long distance according to an embodiment of the invention identification
The big logotype of frame.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to is used to explain the present invention, and is not considered as limiting the invention.
The Video object tracking method proposed according to embodiments of the present invention, Video object tracking described with reference to the accompanying drawing
Device and vehicle with the Video object tracking device.
Fig. 1 is the flow chart of Video object tracking method according to an embodiment of the present invention.
As shown in Figure 1, the Video object tracking method of the embodiment of the present invention can comprise the following steps that
S1 obtains the video information of current scene in real time, and carries out Object identifying to each frame image in video information
To obtain recognition result.Wherein, recognition result is indicated with rectangle frame.
Specifically, the process of Object identifying is carried out to each frame image in video information can include: image preprocessing, spy
Sign extraction, Object identifying.Wherein, due to weather, illumination, ambient enviroment, the image of shooting can have noise, need elder generation
The image of target vehicle (identification object) is pre-processed (including image grayscale, image enhancement, median filtering, histogram are equal
Weighing apparatusization etc.), to eliminate most noise jamming.Then carry out feature extraction, extract on vehicle the apparent key point of feature or
Key lines, to obtain recognition result (for example, identification of vehicle).Identification side in the prior art can be used by specifically how identifying
Method is not detailed here.
In practical applications, due to the principal outline shape invariance of rigid objects, for the ease of knowing to identification object
Not, horizontal linear is done to the key point in the vehicle's contour of extraction, obtains the rectangle block diagram of identification object, as Figure 2-Figure 5,
Wherein, Fig. 2-Fig. 5 is by taking tracking object is 2 as an example.
S2 calculates adjacent two frame according to the corresponding recognition result of previous frame image and the corresponding recognition result of current frame image
Identification object between registration, and according to registration obtain pursuit gain.
S3 judges pursuit gain.
S4 judges that identification is effective to image tracing when pursuit gain is greater than the first tracking threshold value.Wherein, the first tracking threshold value
It can be demarcated according to the actual situation.
According to one embodiment of present invention, registration can be calculated according to following formula (1) and be obtained:
Registration=α × S3/S1+(1-α)×S3/S2 (1)
Wherein, S1 is the area of the corresponding rectangle frame of previous frame image, and S2 is the face of the corresponding rectangle frame of current frame image
Product, S3 are the area of the corresponding rectangle frame of previous frame image rectangle frame intersection corresponding with current frame image, and α is constant,
And 0≤α≤1.
Further, according to one embodiment of present invention, according to registration update pursuit gain, comprising: to registration into
Row judgement, and search in preset tracking object list with the presence or absence of identification object, if registration be greater than registration threshold value and
Identification object is in tracking object list, then is added up on the basis of initial pursuit gain, until pursuit gain reaches pre- and limits
Value；If registration is less than or equal to registration threshold value and identification object is in tracking object list, initial pursuit gain basis
On successively decrease, until deleting the corresponding entry of identification object in tracking object list when pursuit gain is equal to zero；If be overlapped
Degree is greater than registration threshold value and identification object is not in tracking object list, then increases entry newly in tracking object list, and
The pursuit gain of newly-increased entry is set one.Wherein, registration threshold value and pre-set limit can be demarcated according to the actual situation.
Specifically, when starting to carry out to image tracing, first to tracking object list, former frame list object, present frame pair
As list is initialized, Object identifying then is carried out to obtain recognition result to current frame image, and recognition result is recorded
It is calculated into present frame list object, and according to the corresponding recognition result of previous frame image and the corresponding identification of current frame image
Registration between the identification object of newest adjacent two frame, wherein registration is the area of two rectangle intersections in plane
With the weighted average of the ratio between the two rectangular areas, it can specifically be calculated and be obtained according to above-mentioned formula (1).
After calculating registration, registration is judged, if the registration of two rectangle frames is greater than registration threshold
Value, the recognition result that two rectangle frames are included is considered as same target, and this object be it is traceable, for example, target carriage
When (identification object) is closer relative to this vehicle, as shown in Figures 2 and 3, the corresponding rectangle frame of first frame image is S101,
The corresponding rectangle frame of second frame image is S201, and registration is very big, can be used as tracking foundation；The corresponding rectangle frame of previous frame image
For S102, the corresponding rectangle frame of current frame image is S202, and registration is very big, can be used as tracking foundation.For another example, target vehicle
(identification object) relative to this vehicle distance farther out when, as shown in Figure 4 and Figure 5, the corresponding rectangle frame of n-th frame image is S301, n-th
The corresponding rectangle frame of+1 frame image is S401, although small compared to the overlapping area of S101 and S201 or S102 and S202 very
It is more, but the registration of adjacent two frame (n-th frame and the (n+1)th frame) is very big, can be used as tracking foundation.
It searches with the presence or absence of identification object in preset tracking object list, if the identification object is in tracking object column
In table, then add up pursuit gain, if pursuit gain is more than pre-set limit, setting pursuit gain is pre-set limit；If the identification object
It is not in tracking object list, then increases entry newly in tracking object list, and the pursuit gain of newly-increased entry is set 1.If
The registration of two rectangle frames is less than or equal to registration threshold value, and identifies that object is in tracking object list, then tracking of successively decreasing
Value deletes the entry if pursuit gain is equal to 0 in tracking object list, for example, target vehicle (identification object) is blocked
Afterwards, the target vehicle being blocked on video image does not identify frame, does not have associated in newest two frame video image
With as a result, therefore, the corresponding pursuit gain of identification object in tracking list is subtracted 1, is decremented to 0 by one frame video image of every acquisition
When delete the entry.
After obtaining pursuit gain according to registration, pursuit gain is judged, wherein when pursuit gain is greater than the first tracking
When threshold value, shows to identify Object identifying effectively and tracking is effective.It can adapt to not by carrying out the comparison of opposite registration as a result,
With the identified object of size, it can adapt to Object identifying and the tracking of the very big detection scene of range, effectively prevent carrying out
Close objects and remote object must be taken into account when the comparison of overlapping area absolute value, so that judging when tracking close objects
Threshold value is too small, judgment threshold excessive problem when tracking remote object.
In order to improve the accuracy identified to image tracing, according to one embodiment of present invention, have in identification to image tracing
After effect, identification object is obtained also according to the corresponding recognition result of previous frame image and the corresponding recognition result of current frame image
The direction of motion and motion profile, and the direction of extinction of identification object is carried out according to the direction of motion of identification object and motion profile
Anticipation, and when anticipation is set up and pursuit gain is less than the second tracking threshold value, judge to identify that object disappears.Wherein, the second tracking
Threshold value can be demarcated according to the actual situation.
According to one embodiment of present invention, corresponding according to the corresponding recognition result of previous frame image and current frame image
Recognition result obtains the direction of motion and motion profile of identification object, comprising: by the center of the corresponding rectangle frame of previous frame image
The center of rectangle frame corresponding with current frame image carries out line expression, to obtain identification object motion trajectory.When former frame figure
As corresponding rectangle frame rectangle frame corresponding with current frame image the line of centres and the positive apparent direction of video collector between folder
Angle is distributed greater than the first predetermined angle and the corresponding rectangle frame of previous frame image rectangle frame corresponding with current frame image or so
When, identify the direction of motion of object for laterally；When the corresponding rectangle frame of previous frame image rectangle frame corresponding with current frame image
The line of centres and the positive apparent direction of video collector between angle less than the second predetermined angle when, identify the direction of motion of object
For longitudinal direction.Wherein, for the second predetermined angle less than the first predetermined angle, the first predetermined angle and the second predetermined angle can be according to reality
Situation is demarcated.
Further, according to one embodiment of present invention, according to the direction of motion of identification object and motion profile to knowledge
The direction of extinction of other object is prejudged, comprising: when the direction of motion for identifying object is lateral, if the movement of identification object
Track has moved into preset Video Edge region, then prejudges establishment；When the direction of motion for identifying object is longitudinal, if identification
The motion profile of object has moved into default vanishing line and the corresponding rectangle frame of current frame image is having a size of preset minimum identification frame,
Then prejudge establishment；When the direction of motion for identifying object is longitudinal, if the corresponding rectangle frame of current frame image is having a size of default
Maximum identification frame, then prejudge establishment.
Specifically, after judgement identification is effective to image tracing, the traffic direction and running track of identification object can be also obtained,
And the direction of extinction of identification object is prejudged according to the direction of motion and running track, to increase the confidence level to image tracing.
As Figure 2-Figure 5, by taking first frame image and the second frame image, n-th frame and the (n+1)th frame image as an example.By first
The center of the center of the corresponding rectangle frame of frame image rectangle frame corresponding with the second frame image carries out line, to obtain identification object
Running track.When the line of centres of two rectangle frames and the angle of the positive apparent direction of video collector are bigger (for example, angle is greater than
First predetermined angle), and (i.e. the corresponding rectangle frame of first frame image is in a left side, the second frame image pair for two rectangle frames or so distribution
The rectangle frame answered is on the right side or the corresponding rectangle frame of first frame image on the right side, and the corresponding rectangle frame of the second frame image is on a left side) when,
The direction of motion for identifying object is laterally.
When the line of centres of two rectangle frames and the angle of the positive apparent direction of video collector are smaller (for example, angle is less than
Second predetermined angle) when, the direction of motion of object is identified to be longitudinal, for example, as shown in Figures 2 and 3, first frame image is corresponding
Rectangle frame is S101, and the corresponding rectangle frame of the second frame image is S201 or the corresponding rectangle frame of first frame image is S102, the
The corresponding rectangle frame of two frame images is S202.For another example, as shown in Figure 4 and Figure 5, the corresponding rectangle frame of first frame image is S301,
The corresponding rectangle frame of second frame image is S401 or the corresponding rectangle frame of first frame image is S302, and the second frame image is corresponding
Rectangle frame be S402.
After obtaining the direction of motion for having tracked vehicle (identification object) and motion profile, these information are recorded in tracking
(relative to this vehicle) in list object, to improve the accuracy of tracking, to be prejudged to the direction of extinction for having tracked vehicle
Foundation is provided.When the direction of motion for identifying object is lateral, if the motion profile of identification object has moved into preset video
Fringe region then prejudges establishment, for example, identification object may turn right or turn left or turn around.When the direction of motion of identification object is
When longitudinal, if the motion profile of identification object has moved into default vanishing line and the corresponding rectangle frame of current frame image having a size of pre-
If minimum identification frame, then establishments is prejudged, for example, target vehicle (identification object) improves speed, close to presetting vanishing line；If
The corresponding rectangle frame of present frame then prejudges establishment having a size of preset maximum identification frame, for example, this vehicle improves speed, it is more than mesh
Mark vehicle (identification object).
When prejudging establishment, pursuit gain is judged, if pursuit gain judges identification pair less than the second tracking threshold value
As disappearing, when identifying that object disappears, corresponding entry in tracking object list is deleted.It is blocked and identifies object, when its pursuit gain
When being decremented to 0, judge that object disappears.As a result, by identification object motion profile and the direction of motion recorded (relative to
The positive apparent direction of video collector), the motion conditions of identification object are prejudged according to motion profile and the direction of motion, increase pair
The confidence level of image tracing.
Therefore, the Video object tracking method of the embodiment of the present invention, by reasonably selecting different distance to the threshold of image tracing
Value, comes Object identifying and the tracking of the very big detection scene of adaptation range, while to the motion profile and fortune of different tracking objects
Dynamic direction is recorded, and the accuracy of tracking is improved.
In conclusion Video object tracking method according to an embodiment of the present invention, obtains the video letter of current scene in real time
Breath, and to each frame image progress Object identifying in video information to obtain recognition result, it is corresponding according to previous frame image
Recognition result and the corresponding recognition result of current frame image calculate the registration between the identification object of adjacent two frame, and according to weight
Right update pursuit gain, and pursuit gain is judged, when pursuit gain is greater than the first tracking threshold value, judge identification to image tracing
Effectively.This method can adapt to the very big detecting field of range by comparing the registration of the identified object of two continuous frames video
The Object identifying of scape and tracking.
Corresponding above-described embodiment, the invention also provides a kind of non-transitorycomputer readable storage mediums, store thereon
There is computer program, which realizes above-mentioned Video object tracking method when being executed by processor.
The non-transitorycomputer readable storage medium of the embodiment of the present invention, by executing the computer program of its storage,
By comparing the registration of the identified object of two continuous frames video, the Object identifying of the very big detection scene of range can adapt to
And tracking.
In addition, the invention also provides a kind of Video object tracking device, including memory, processor and it is stored in storage
On device and the Video object tracking program that can run on a processor, the control program of Video object tracking device are held by processor
The step of above-mentioned Video object tracking method is realized when row.
The Video object tracking device of the embodiment of the present invention, by comparing the coincidence of the identified object of two continuous frames video
Degree, can adapt to Object identifying and the tracking of the very big detection scene of range.
Corresponding above-described embodiment, the invention also provides a kind of vehicles comprising above-mentioned Video object tracking device.
The vehicle of the embodiment of the present invention, by above-mentioned Video object tracking device, by comparing two continuous frames video
The registration of identified object can adapt to Object identifying and the tracking of the very big detection scene of range.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present invention, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing custom logic function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discussed suitable
Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, Lai Zhihang function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system, including the system of processor or other can be held from instruction
The instruction fetch of row system, device or equipment and the system executed instruction) it uses, or combine these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium ", which can be, any may include, stores, communicates, propagates or pass
Defeated program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment
It sets.The more specific example (non-exhaustive list) of computer-readable medium include the following: there is the electricity of one or more wirings
Interconnecting piece (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other are suitable
Medium, because can then be edited, be interpreted or when necessary with it for example by carrying out optical scanner to paper or other media
His suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of the invention can be realized with hardware, software, firmware or their combination.Above-mentioned
In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage
Or firmware is realized.Such as, if realized with hardware in another embodiment, following skill well known in the art can be used
Any one of art or their combination are realized: have for data-signal is realized the logic gates of logic function from
Logic circuit is dissipated, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), scene can compile
Journey gate array (FPGA) etc..
Those skilled in the art are understood that realize all or part of step that above-described embodiment method carries
It suddenly is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer-readable storage medium
In matter, which when being executed, includes the steps that one or a combination set of embodiment of the method.
It, can also be in addition, each functional unit in each embodiment of the present invention can integrate in a processing module
It is that each unit physically exists alone, can also be integrated in two or more units in a module.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized and when sold or used as an independent product in the form of software function module, also can store in a computer
In read/write memory medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..Although having been shown and retouching above
The embodiment of the present invention is stated, it is to be understood that above-described embodiment is exemplary, and should not be understood as to limit of the invention
System, those skilled in the art can be changed above-described embodiment, modify, replace and become within the scope of the invention
Type.
In the description of the present invention, it is to be understood that, term " center ", " longitudinal direction ", " transverse direction ", " length ", " width ",
" thickness ", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outside", " up time
The orientation or positional relationship of the instructions such as needle ", " counterclockwise ", " axial direction ", " radial direction ", " circumferential direction " be orientation based on the figure or
Positional relationship is merely for convenience of description of the present invention and simplification of the description, rather than the device or element of indication or suggestion meaning must
There must be specific orientation, be constructed and operated in a specific orientation, therefore be not considered as limiting the invention.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.
Term shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or integral；It can be mechanical connect
It connects, is also possible to be electrically connected；It can be directly connected, can also can be in two elements indirectly connected through an intermediary
The interaction relationship of the connection in portion or two elements.It for the ordinary skill in the art, can be according to specific feelings
Condition understands the concrete meaning of above-mentioned term in the present invention.
In the present invention unless specifically defined or limited otherwise, fisrt feature in the second feature " on " or " down " can be with
It is that the first and second features directly contact or the first and second features pass through intermediary mediate contact.Moreover, fisrt feature exists
Second feature " on ", " top " and " above " but fisrt feature be directly above or diagonally above the second feature, or be merely representative of
First feature horizontal height is higher than second feature.Fisrt feature can be under the second feature " below ", " below " and " below "
One feature is directly under or diagonally below the second feature, or is merely representative of first feature horizontal height less than second feature.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, modifies, replacement and variant.
Claims (10)
1. a kind of Video object tracking method, which comprises the following steps:
The video information of current scene is obtained in real time, and Object identifying is carried out to obtain to each frame image in the video information
Obtain recognition result；
The identification of adjacent two frame is calculated according to the corresponding recognition result of previous frame image and the corresponding recognition result of current frame image
Registration between object, and pursuit gain is updated according to the registration；
The pursuit gain is judged；
When the pursuit gain is greater than the first tracking threshold value, judge that the identification is effective to image tracing.
2. Video object tracking method as described in claim 1, which is characterized in that it is described identification to image tracing effectively it
Afterwards, the identification object is obtained also according to the corresponding recognition result of previous frame image and the corresponding recognition result of current frame image
The direction of motion and motion profile, and the disappearance according to the direction of motion of the identification object and motion profile to the identification object
Direction is prejudged, and when anticipation is set up and the pursuit gain is less than the second tracking threshold value, judges that the identification object disappears
It loses.
3. Video object tracking method as claimed in claim 1 or 2, which is characterized in that the recognition result with rectangle frame into
Row indicates.
4. Video object tracking method as claimed in claim 3, which is characterized in that the registration calculates according to the following formula
It obtains:
Registration=α × S3/S1+(1-α)×S3/S2,
Wherein, S1 is the area of the corresponding rectangle frame of the previous frame image, and S2 is the corresponding rectangle frame of the current frame image
Area, S3 be the corresponding rectangle frame of previous frame image rectangle frame intersection corresponding with the current frame image face
Product, α are constant, and 0≤α≤1.
5. Video object tracking method as described in claim 1, which is characterized in that pursuit gain is updated according to the registration,
Include:
The registration is judged, and is searched in preset tracking object list with the presence or absence of the identification object；
If the registration is greater than registration threshold value and the identification object is in tracking object list, initially tracking
It adds up on the basis of value, until the pursuit gain reaches pre-set limit；
If the registration is less than or equal to registration threshold value and the identification object and is in tracking object list, initially with
Successively decrease on the basis of track value, until deleting the identification pair in the tracking object list when pursuit gain is equal to zero
As corresponding entry；
If the registration is greater than registration threshold value and the identification object and is not in tracking object list, it is described with
It increases entry in track list object newly, and the pursuit gain of the newly-increased entry is set one.
6. Video object tracking method as claimed in claim 3, which is characterized in that tied according to the corresponding identification of previous frame image
Fruit and the corresponding recognition result of current frame image obtain the direction of motion and motion profile of the identification object, comprising:
The center of the center of the corresponding rectangle frame of previous frame image rectangle frame corresponding with current frame image is subjected to line,
To obtain the motion profile of the identification object；
When the line of centres and video acquisition of the corresponding rectangle frame of previous frame image rectangle frame corresponding with current frame image
Angle between the positive apparent direction of device is greater than the first predetermined angle and the corresponding rectangle frame of the previous frame image and current frame image
When corresponding rectangle frame or so is distributed, the direction of motion of the identification object is laterally；
When the line of centres and video acquisition of the corresponding rectangle frame of previous frame image rectangle frame corresponding with current frame image
Less than the second predetermined angle, the direction of motion of the identification object is longitudinal for angle between the positive apparent direction of device, wherein described the
Two predetermined angles are less than first predetermined angle.
7. Video object tracking method as claimed in claim 6, which is characterized in that according to the direction of motion of the identification object
It is prejudged with direction of extinction of the motion profile to the identification object, comprising:
When the direction of motion of the identification object is lateral, if the motion profile of the identification object has moved into preset view
Frequency fringe region then prejudges establishment；
When the direction of motion of the identification object is longitudinal, if the motion profile of the identification object has moved into default disappearance
Line and the corresponding rectangle frame of the current frame image then prejudge establishment having a size of preset minimum identification frame；
When the direction of motion of the identification object is longitudinal, if the corresponding rectangle frame of the current frame image is having a size of default
Maximum identification frame, then prejudge establishment.
8. a kind of non-transitorycomputer readable storage medium, is stored thereon with computer program, which is characterized in that the program quilt
Such as Video object tracking method of any of claims 1-7 is realized when processor executes.
9. a kind of Video object tracking device, which is characterized in that including memory, processor and be stored on the memory simultaneously
The Video object tracking program that can be run on the processor, the control program of the Video object tracking device is by the place
It manages when device executes and realizes such as the step of Video object tracking method of any of claims 1-7.
10. a kind of vehicle, which is characterized in that including Video object tracking device as claimed in claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711243752.0A CN109859240B (en) | 2017-11-30 | 2017-11-30 | Video object tracking method and device and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711243752.0A CN109859240B (en) | 2017-11-30 | 2017-11-30 | Video object tracking method and device and vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109859240A true CN109859240A (en) | 2019-06-07 |
| CN109859240B CN109859240B (en) | 2021-06-18 |
Family
ID=66888813
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711243752.0A Active CN109859240B (en) | 2017-11-30 | 2017-11-30 | Video object tracking method and device and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109859240B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348460A (en) * | 2019-07-04 | 2019-10-18 | 成都旷视金智科技有限公司 | A kind of target detection training method, object detection method and device based on angle |
| CN110969647A (en) * | 2019-12-04 | 2020-04-07 | 苏州智加科技有限公司 | Method for integrating vehicle identification tracking and vehicle lamp detection |
| CN111046820A (en) * | 2019-12-17 | 2020-04-21 | 上海船舶研究设计院（中国船舶工业集团公司第六0四研究院） | Statistical method and device for vehicles in automobile roll-on-roll-off ship and intelligent terminal |
| CN111079675A (en) * | 2019-12-23 | 2020-04-28 | 武汉唯理科技有限公司 | Driving behavior analysis method based on target detection and target tracking |
| CN111259754A (en) * | 2020-01-10 | 2020-06-09 | 中国海洋大学 | End-to-end plankton database construction system and method |
| CN112017319A (en) * | 2020-08-21 | 2020-12-01 | 中建二局第一建筑工程有限公司 | Intelligent patrol security method, device and system and storage medium |
| CN112037257A (en) * | 2020-08-20 | 2020-12-04 | 浙江大华技术股份有限公司 | Target tracking method, terminal and computer readable storage medium thereof |
| CN114582140A (en) * | 2022-01-17 | 2022-06-03 | 浙江银江智慧交通工程技术研究院有限公司 | Method, system, device and medium for identifying traffic flow of urban road intersection |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130129157A1 (en) * | 2008-06-27 | 2013-05-23 | Canon Kabushiki Kaisha | Image processing apparatus for retrieving object from moving image and method thereof |
| CN103259962A (en) * | 2013-04-17 | 2013-08-21 | 深圳市捷顺科技实业股份有限公司 | Target tracking method and related device |
| US20140003725A1 (en) * | 2012-06-15 | 2014-01-02 | Canon Kabushiki Kaisha | Video processing apparatus, video processing method, and recording medium |
| CN104217417A (en) * | 2013-05-31 | 2014-12-17 | 张伟伟 | A video multiple-target tracking method and device |
| CN105551063A (en) * | 2016-01-29 | 2016-05-04 | 中国农业大学 | Method and device for tracking moving object in video |
| US20170061229A1 (en) * | 2015-09-01 | 2017-03-02 | Sony Corporation | Method and system for object tracking |
| CN106682619A (en) * | 2016-12-28 | 2017-05-17 | 上海木爷机器人技术有限公司 | Object tracking method and device |
| CN106846374A (en) * | 2016-12-21 | 2017-06-13 | 大连海事大学 | The track calculating method of vehicle under multi-cam scene |
| CN107292297A (en) * | 2017-08-09 | 2017-10-24 | 电子科技大学 | A kind of video car flow quantity measuring method tracked based on deep learning and Duplication |
| CN107316322A (en) * | 2017-06-27 | 2017-11-03 | 上海智臻智能网络科技股份有限公司 | Video tracing method and device and object identifying method and device |
| CN107358621A (en) * | 2016-05-10 | 2017-11-17 | 腾讯科技（深圳）有限公司 | Method for tracing object and device |
-
2017
- 2017-11-30 CN CN201711243752.0A patent/CN109859240B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130129157A1 (en) * | 2008-06-27 | 2013-05-23 | Canon Kabushiki Kaisha | Image processing apparatus for retrieving object from moving image and method thereof |
| US20140003725A1 (en) * | 2012-06-15 | 2014-01-02 | Canon Kabushiki Kaisha | Video processing apparatus, video processing method, and recording medium |
| CN103259962A (en) * | 2013-04-17 | 2013-08-21 | 深圳市捷顺科技实业股份有限公司 | Target tracking method and related device |
| CN104217417A (en) * | 2013-05-31 | 2014-12-17 | 张伟伟 | A video multiple-target tracking method and device |
| US20170061229A1 (en) * | 2015-09-01 | 2017-03-02 | Sony Corporation | Method and system for object tracking |
| CN105551063A (en) * | 2016-01-29 | 2016-05-04 | 中国农业大学 | Method and device for tracking moving object in video |
| CN107358621A (en) * | 2016-05-10 | 2017-11-17 | 腾讯科技（深圳）有限公司 | Method for tracing object and device |
| CN106846374A (en) * | 2016-12-21 | 2017-06-13 | 大连海事大学 | The track calculating method of vehicle under multi-cam scene |
| CN106682619A (en) * | 2016-12-28 | 2017-05-17 | 上海木爷机器人技术有限公司 | Object tracking method and device |
| CN107316322A (en) * | 2017-06-27 | 2017-11-03 | 上海智臻智能网络科技股份有限公司 | Video tracing method and device and object identifying method and device |
| CN107292297A (en) * | 2017-08-09 | 2017-10-24 | 电子科技大学 | A kind of video car flow quantity measuring method tracked based on deep learning and Duplication |
Non-Patent Citations (5)
| Title |
|---|
| RAN ESHEL等: "Homography based multiple camera detection and tracking of people in a dense crowd", 《2008 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION》 * |
| ROHIT C. PHILIP等: "A comparison of tracking algorithm performance for objects in wide area imagery", 《2014 SOUTHWEST SYMPOSIUM ON IMAGE ANALYSIS AND INTERPRETATION》 * |
| 付苗: "基于在线学习的非特定目标长时间跟踪技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 甘玲等: "一种应用于交通环境中的运动车辆跟踪方法", 《重庆邮电大学学报》 * |
| 黄德宝: "智能交通监控系统中运动目标的检测与预警技术", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348460A (en) * | 2019-07-04 | 2019-10-18 | 成都旷视金智科技有限公司 | A kind of target detection training method, object detection method and device based on angle |
| CN110348460B (en) * | 2019-07-04 | 2021-10-22 | 成都旷视金智科技有限公司 | Angle-based target detection training method, target detection method and device |
| CN110969647A (en) * | 2019-12-04 | 2020-04-07 | 苏州智加科技有限公司 | Method for integrating vehicle identification tracking and vehicle lamp detection |
| CN111046820A (en) * | 2019-12-17 | 2020-04-21 | 上海船舶研究设计院（中国船舶工业集团公司第六0四研究院） | Statistical method and device for vehicles in automobile roll-on-roll-off ship and intelligent terminal |
| CN111079675A (en) * | 2019-12-23 | 2020-04-28 | 武汉唯理科技有限公司 | Driving behavior analysis method based on target detection and target tracking |
| CN111259754A (en) * | 2020-01-10 | 2020-06-09 | 中国海洋大学 | End-to-end plankton database construction system and method |
| CN112037257A (en) * | 2020-08-20 | 2020-12-04 | 浙江大华技术股份有限公司 | Target tracking method, terminal and computer readable storage medium thereof |
| CN112037257B (en) * | 2020-08-20 | 2023-09-29 | 浙江大华技术股份有限公司 | Target tracking method, terminal and computer readable storage medium thereof |
| CN112017319A (en) * | 2020-08-21 | 2020-12-01 | 中建二局第一建筑工程有限公司 | Intelligent patrol security method, device and system and storage medium |
| CN114582140A (en) * | 2022-01-17 | 2022-06-03 | 浙江银江智慧交通工程技术研究院有限公司 | Method, system, device and medium for identifying traffic flow of urban road intersection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109859240B (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109859240A (en) | Video object tracking method and apparatus, vehicle | |
| CN109145931B (en) | Object detecting method, device and storage medium | |
| US10417503B2 (en) | Image processing apparatus and image processing method | |
| AU2008210192B2 (en) | Method of moving target tracking and number accounting | |
| JP3781370B2 (en) | Mobile device | |
| Lee | A machine vision system for lane-departure detection | |
| CN105426864B (en) | One kind being based on the matched Multi-lane Lines Detection method of equidistant marginal point | |
| TWI459312B (en) | Face tracking method | |
| CN105260713B (en) | A kind of method for detecting lane lines and device | |
| JP3880759B2 (en) | Moving object detection method | |
| CN109147341B (en) | Violation vehicle detection method and device | |
| CN101681424B (en) | Sequential image alignment | |
| Lee et al. | A lane-departure identification based on LBPE, Hough transform, and linear regression | |
| US20090037039A1 (en) | Method for locomotive navigation and track identification using video | |
| CN102799857B (en) | Video multi-vehicle outline detection method | |
| TW201120807A (en) | Apparatus and method for moving object detection | |
| Cheung et al. | Efficient object-based video inpainting | |
| CN103140876B (en) | Information processing device, information processing method, program for information processing device, and recording medium | |
| JP6226368B2 (en) | Vehicle monitoring apparatus and vehicle monitoring method | |
| RU2395787C2 (en) | Method of detecting objects | |
| KR101552600B1 (en) | Method for tracking multiple walking object simultaneously using sectioned image | |
| CN107705577B (en) | Real-time detection method and system for calibrating illegal lane change of vehicle based on lane line | |
| KR102226845B1 (en) | System and method for object recognition using local binarization | |
| Janda et al. | Road boundary detection for run-off road prevention based on the fusion of video and radar | |
| CN114998317B (en) | Lens occlusion detection method and device, camera device and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |