CN106204653B - A kind of monitoring tracking and device - Google Patents
A kind of monitoring tracking and device Download PDFInfo
- Publication number
- CN106204653B CN106204653B CN201610556686.1A CN201610556686A CN106204653B CN 106204653 B CN106204653 B CN 106204653B CN 201610556686 A CN201610556686 A CN 201610556686A CN 106204653 B CN106204653 B CN 106204653B
- Authority
- CN
- China
- Prior art keywords
- frame
- tracking box
- video frame
- monitoring video
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20224—Image subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
Landscapes
- Closed-Circuit Television Systems (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of monitoring tracking and device, which comprises carries out foreground extraction to the monitor video got;The first tracking box is determined according to the prospect extracted;First tracking box is handled using default track algorithm, to obtain stronger second tracking box of stability;Target following frame is determined according to first tracking box and second tracking box;Monitored object is tracked according to the target following frame.The automatic monitoring tracking of monitored object may be implemented using the embodiment of the present invention, and guarantee the accuracy for being monitored tracking to monitored object.
Description
Technical field
The present invention relates to technical field of video monitoring more particularly to a kind of monitoring trackings and device.
Background technique
With the development of Video Supervision Technique, Video Supervision Technique applications in various fields is more more and more universal, and by
Gradually become an important component of safety and protection system.It, can be effectively in monitoring scene by Video Supervision Technique
The various actions of monitored object be monitored, and handled in time when found the abnormal situation.
However practice discovery, in existing Video Supervision Technique, the various actions monitoring to monitored object is mainly still
Pass through the monitoring of the artificial real time inspection monitoring scene of monitoring personnel (including but not limited to market, bank security, prison prison guard etc.)
The mode of video is realized, needs to expend a large amount of manpower, and is easy to cause to found the abnormal situation in time due to human negligence,
And then lead to adverse events.
Summary of the invention
The present invention provides a kind of monitoring tracking and device, with solve it is existing by manually to the behavior of monitored object into
It needs to expend a large amount of manpowers in the scheme of row monitoring, and is easy to lead to the problem of cannot noting abnormalities in time due to artificial carelessness.
According to the first aspect of the invention, a kind of monitoring tracking is provided, video monitoring system is applied to, comprising:
Foreground extraction is carried out to the monitor video got;
The first tracking box is determined according to the prospect extracted;
First tracking box is handled using default track algorithm, to obtain stronger second tracking of stability
Frame;
Target following frame is determined according to first tracking box and second tracking box;
Monitored object is tracked according to the target following frame.
According to the second aspect of the invention, a kind of monitoring tracking device is provided, comprising:
Extraction unit, for carrying out foreground extraction to the monitor video got;
First determination unit, for determining the first tracking box according to the prospect extracted;
Processing unit, for being handled using default track algorithm first tracking box, to obtain stability more
The second strong tracking box;
Second determination unit, for determining target following frame according to first tracking box and second tracking box;
Tracking cell, for being tracked according to the target following frame to monitored object.
Using technical solution disclosed by the invention, by carrying out foreground extraction to the monitor video got, and according to mentioning
The prospect got determines the first tracking box, and then is handled using default track algorithm the first tracking box, to be stablized
Stronger second tracking box of property, and target following frame is determined according to the first tracking box and the second tracking box, and according to the target with
Track frame tracks monitored object, realizes the automatic monitoring tracking of monitored object, and ensure that and supervise to monitored object
Control the accuracy of tracking.
Detailed description of the invention
Fig. 1 is a kind of flow diagram for monitoring tracking provided in an embodiment of the present invention;
Fig. 2 is a kind of structural schematic diagram for monitoring tracking device provided in an embodiment of the present invention;
Fig. 3 is the structural schematic diagram of another monitoring tracking device provided in an embodiment of the present invention;
Fig. 4 is the structural schematic diagram of another monitoring tracking device provided in an embodiment of the present invention;
Fig. 5 is the structural schematic diagram of another monitoring tracking device provided in an embodiment of the present invention.
Specific embodiment
Technical solution in embodiment in order to enable those skilled in the art to better understand the present invention, and make of the invention real
The above objects, features, and advantages for applying example can be more obvious and easy to understand, with reference to the accompanying drawing to technical side in the embodiment of the present invention
Case is described in further detail.
Referring to Figure 1, Fig. 1 is a kind of flow diagram for monitoring tracking provided in an embodiment of the present invention, such as Fig. 1 institute
Show, which may include:
It should be noted that step 101~step 105 executing subject can be the background service in video monitoring system
The processor of device or background server, such as CPU (Center Process Unit, central processing unit), for ease of description, with
Under by step 101~step 105 executing subject be server for be illustrated.
Step 101 carries out foreground extraction to the monitor video got.
In the embodiment of the present invention, server can pass through the video acquisition device disposed in monitoring scene, such as IPC
(Internet Protocol Camera, web camera), acquires the monitor video of monitoring scene, and to collected monitoring
Video carries out foreground extraction.
As an alternative embodiment, in embodiments of the present invention, carrying out prospect to the monitor video got and mentioning
It takes, may include:
Mixed Gaussian background modeling is carried out to the monitoring video frame got, extracts the prospect in monitoring video frame, and
Interference filtering is carried out to the prospect extracted.
In this embodiment, after server gets monitor video, the monitoring video frame got can be mixed
Gaussian Background modeling is closed, the prospect in monitoring video frame is extracted.Due to the prospect meeting directly extracted from monitoring video frame
There is interference caused by many noises and light variation etc., therefore, server may be used also after extracting prospect in monitor video
To carry out interference filtering to the prospect extracted, burn into expansion, denoising, size exclusion, region such as are carried out to the prospect extracted
The morphological operations such as connection obtain effective prospect to filter out interference.
It is worth noting that in embodiments of the present invention, server carries out foreground extraction to the monitor video got
Mode can also be other way, such as ViBe (Visual other than the mixed Gaussian background modeling mode of foregoing description
Background Extractor, visual background extract) background modeling algorithm, codebook (code book) background modeling algorithm,
SOBS (Self-organization background subtraction, self-organizing background subtraction) background modeling algorithm etc.,
Specific implementation does not repeat them here herein.
In the embodiment of the present invention, it is contemplated that when monitored object moves suddenly (suddenly switching to move by quiet), server passes through
To monitoring video frame carry out foreground extraction when, in addition to can the practical position of monitored object extract corresponding prospect it
Outside, it is also possible to prospect (prospect is properly termed as " ghost ") can be extracted at the position where before monitored object, in order to keep away
Exempting from " ghost " leads to interference to the behavioral value of monitored object, and server needs after extracting prospect in monitor video
" ghost " is gone to handle the prospect extracted.
Accordingly, as a kind of optional embodiment, in embodiments of the present invention, the monitor video got is carried out
After foreground extraction, can also include:
If extract prospect from monitoring video frame, there are overlapped prospects in the monitoring video frame, then will prison
Control the lesser prospect removal of area in prospect overlapped in video frame.
In this embodiment, it is contemplated that " ghost " extracted from monitoring video frame would generally be corresponding with monitored object
Actual foreground exist overlapping, and " ghost " would generally actual foreground more corresponding than monitored object area it is small, therefore, work as service
When device extracts prospect from monitoring video frame, there are when overlapped prospect in the monitoring video frame, server can be recognized
To include " ghost " in the prospect extracted, in turn, server can be gone the lesser prospect of area in overlapped prospect
It removes, the interference to avoid " ghost " to subsequent processing.
As an example it is assumed that certain monitored object (as supervised indoor criminal) is under sleep state (lying on a bed)
It stands up to leave the position of sleep suddenly, at this point, when server carries out foreground extraction to monitoring video frame, for the monitored object,
Two prospects may be generated, one of prospect is generated since monitored object leaves sleep position, which is " ghost
Shadow ", due to the real prospect of the mobile generation of monitored object, which usually there will be overlapped feelings for another
Condition, and " ghost " prospect can be smaller than actual foreground, thus, when server extracts prospect, and presence from monitoring video frame
When overlapped two prospects, server can be removed the lesser prospect of area in overlapped prospect, to avoid " ghost
Influence of the shadow " to monitoring tracking.
Step 102 determines the first tracking box according to the prospect extracted.
In the embodiment of the present invention, server, can be according to the prospect extracted after extracting prospect in monitoring video frame
The first tracking box (being referred to as weak tracking box) is determined, as server can will be from the prospect extracted in monitoring video frame
First tracking box of the minimum circumscribed rectangle frame as the monitoring video frame.
Wherein, if server is after extracting prospect in the monitoring video frame got, the prospect extracted is carried out
Interference processing or removal " ghost " processing are filtered out, server needs according to treated that prospect is true when determining the first tracking box
It is fixed.
It is worth noting that server is somebody's turn to do after extracting prospect for the first time in monitoring video frame for monitor video
It can determine that first tracking box in each monitoring video frame after monitoring video frame.
Step 103 is handled the first tracking box using default track algorithm, with obtain stability stronger second with
Track frame.
In the embodiment of the present invention, server extracts prospect from video frame for the first time, and has been determined first according to the prospect
When tracking box, server can use default track algorithm and handle first tracking box, to enhance the first tracking box
Stability obtains stronger second tracking box of stability.
Wherein, which can include but is not limited to STC (Fast Tracking via Spatio-
Temporal Context Learning, the tracking of space-time context vision) track algorithm, LK (Lucas-Kanade) tracking calculation
Method etc..
For ease of understanding, below by taking default track algorithm is STC track algorithm as an example.
In the embodiment of the present invention, server extracts prospect from video frame for the first time, and has been determined first according to the prospect
When tracking box, the second tracking box can be obtained by STC track algorithm using first tracking box as the input of STC track algorithm
(being referred to as STC tracking box in this embodiment).
Step 104 determines target following frame according to the first tracking box and the second tracking box.
In the embodiment of the present invention, server, can be according to first after the first tracking box and the second tracking box has been determined
Tracking box and the second tracking box determine target following frame.
As an alternative embodiment, in embodiments of the present invention, it is true according to the first tracking box and the second tracking box
Set the goal tracking box, may include:
For extracting the target monitoring video frame of prospect for the first time, the first tracking box of target monitoring video frame is determined as
The target following frame of target monitoring video frame;
For the monitoring video frame after target monitoring video frame, the second tracking box is determined as to the target of monitoring video frame
Tracking box, and target following frame is repaired according to the first tracking box in each monitoring video frame after target monitoring video frame
Just.
In this embodiment, server extracts prospect for the first time from monitoring video frame (this extracts prospect for the first time
Monitoring video frame is referred to herein as target monitoring video frame), it, can should and when the first tracking box has been determined according to the prospect
First tracking box is as the target following frame in the target monitoring video frame;Meanwhile server can make first tracking box
For the input of STC track algorithm, the second tracking box is obtained by STC track algorithm.
For the monitoring video frame after target monitoring video frame, server can using second tracking box as target with
Track frame, and the first tracking box in each monitoring video frame after the target monitoring video frame repairs target following frame
Just.
Wherein, the first tracking box in each monitoring video frame after target monitoring video frame can by server by pair
The mode that each monitoring video frame carries out foreground extraction determines that specific implementation may refer to the correlation in step 101~step 102
Description, details are not described herein for the embodiment of the present invention.
As an alternative embodiment, in embodiments of the present invention, according to each prison after target monitoring video frame
The first tracking box in control video frame is modified target following frame, comprising:
For any monitoring video frame after target monitoring video frame, if meeting the first preset condition, by target with
Track frame is revised as the first tracking box in the monitoring video frame;
If meeting the second preset condition, delete target tracking box, and according to the first tracking box in the monitoring video frame
Redefine target following frame;
If being unsatisfactory for the first preset condition, and it is unsatisfactory for the second preset condition, then keeps target following frame constant;
Wherein, the first preset condition is the area of target following frame and the area of the first tracking box in the monitoring video frame
Ratio be greater than first threshold, be less than second threshold, and the first tracking box in target following frame and the monitoring video frame exists
Overlapping;Second preset condition is that the ratio of the area and the area of the first tracking box in the monitoring video frame of target following frame is small
In being equal to third threshold value or be more than or equal to the 4th threshold value, alternatively, the in the central point of target following frame and the monitoring video frame
The distance between central point of one tracking box is greater than the half of the width and the sum of height of video image;Second threshold is greater than the first threshold
Value, third threshold value are less than first threshold, and the 4th threshold value is greater than second threshold.
In this embodiment, it is assumed that target monitoring video frame is monitoring video frame t, then server can be by monitor video
The first tracking box in frame t is determined as the target following frame of the monitoring video frame;Each prison is waited for monitoring video frame t+1, t+2 ...
Video frame is controlled, target following frame can be by the first tracking box in each monitoring video frame, and by first in monitoring video frame t
Tracking box determines together as the second tracking box that the input of STC track algorithm obtains.
For example, for monitoring video frame t+1, t+2 ..., server can be first using above-mentioned second tracking box as target following
Frame, and it is modified according to the first tracking box in each monitoring video frame.
For monitoring video frame t+1, server may determine that the first tracking box and current mesh in monitoring video frame t+1
Whether mark tracking box (target following frame at this time is above-mentioned second tracking box) meets the first preset condition or second condition.
If meeting first condition, i.e., the area of the first tracking box in the area of target following frame and monitoring video frame t+1
Ratio be greater than first threshold (can be set according to concrete scene, such as 0.5), be less than second threshold (can be according to concrete scene
Setting, such as 2 and second threshold is greater than first threshold), and there is weight in the first tracking box in target following frame and monitoring video frame
Folded, then target following frame can be modified to the first tracking box in monitoring video frame t+1 by server;
If meeting the second preset condition, i.e. the first tracking box in the area of target following frame and monitoring video frame in t+1
Area ratio be less than or equal to third threshold value (can be set according to concrete scene, such as 0.4, third threshold value be less than first threshold)
Or it is more than or equal to the 4th threshold value (can set according to concrete scene, such as 2.5, the 4th threshold value is greater than second threshold);Alternatively, target
The distance between central point and the central point of the first tracking box in monitoring video frame t+1 of tracking box are greater than the width of video image
With the half of the sum of height, then server can delete current target following frame, and according to the first tracking in monitoring video frame
Frame redefines target following frame, and specific implementation may refer to the associated description in step 101~step 104, and the present invention is real
Applying example, details are not described herein.
If the first tracking box and current target following frame in monitoring video frame t+1 be both unsatisfactory for the first preset condition,
Also it is unsatisfactory for the second preset condition, then the target following frame that server can keep current is constant.
Similarly, the target following frame in the monitoring video frames such as monitoring video frame t+2, t+3 can also according to the method described above really
It is fixed.
Step 105 tracks monitored object according to target following frame.
It, can be according to the target after server has determined the target following frame in monitor video in the embodiment of the present invention
Tracking box tracks monitored object, such as row of the change in location according to target following frame in video image to monitored object
To analyze etc..
Further, as an alternative embodiment, in embodiments of the present invention, according to target following frame to monitoring
After object is tracked, further includes:
When the central point and target following frame of target following frame are moved to mesh out of object detection area in monitor video
When marking outside detection zone, alarming processing is carried out.
In this embodiment, can the specified region in monitoring scene be set as object detection area, correspondingly, should
Corresponding region in the monitor video of monitoring scene is object detection area.Server has determined the target following in monitor video
After frame, can to target following frame in monitor video position movement be monitored;If monitoring server is to target following frame
Central point be moved to outside object detection area out of target area in monitor video, alternatively, monitoring server to target with
Track frame is moved to outside object detection area out of target area in monitor video, and server can consider that monitored object does not exist
In object detection area, currently there is unusual condition, thus, server can carry out alarming processing, such as by the tinkle of bells or its
Its mode notifies administrative staff to carry out follow-up processing.
It for example,, can be with for monitored object is the criminal supervised in room using monitoring scene as the prison room in prison
Object detection area is set by bed region in prison room in advance, and in daily specified time (the prison room such as provided
The time that interior criminal is slept on bed) monitoring tracking in the present invention is enabled to the indoor criminal progress of each prison
Monitoring;When within a specified time, when detecting that monitored object moves suddenly, process generates in monitor video according to the method described above
Target following frame, and work as the central point for detecting target following frame or target following frame from the object detection area in monitor video
It is inside moved to outside object detection area, that is, when having left bed between supervising indoor criminal during the break, it is believed that exception has occurred
Situation, thus, server can be alerted to prison guard, notified prison guard's follow-up to handle, prevented adverse events.
It can be seen that in technical solution provided in an embodiment of the present invention by the above process, by the prison got
It controls video and carries out foreground extraction, and the first tracking box is determined according to the prospect extracted, and then using default track algorithm to the
One tracking box is handled, to obtain stronger second tracking box of stability, and it is true according to the first tracking box and the second tracking box
Set the goal tracking box, and is tracked according to the target following frame to monitored object, realize monitored object it is automatic monitoring with
Track, and ensure that the accuracy that tracking is monitored to monitored object.
Fig. 2 is referred to, for a kind of structural schematic diagram for monitoring tracking device provided in an embodiment of the present invention, wherein the prison
Control tracking device can be applied to video monitoring system, the server being such as applied in above method embodiment, as shown in Fig. 2,
The monitoring tracking device may include:
Extraction unit 210, for carrying out foreground extraction to the monitor video got;
First determination unit 220, for determining the first tracking box according to the prospect extracted;
Processing unit 230, for being handled using default track algorithm first tracking box, to obtain stability
Stronger second tracking box;
Second determination unit 240, for determining target following frame according to first tracking box and second tracking box;
Tracking cell 250, for being tracked according to the target following frame to monitored object.
In an alternative embodiment, the extraction unit 210 can be specifically used for mixing the monitoring video frame got
Gaussian Background modeling is closed, extracts the prospect in the monitoring video frame, and interference filtering is carried out to the prospect extracted;
Correspondingly, first determination unit 220, can be specifically used for according to interfere filtered prospect determine first with
Track frame.
Scheming also referring to Fig. 3 for the structural schematic diagram of another monitoring tracking device provided in an embodiment of the present invention
On the basis of monitoring tracking device shown in 2, the monitoring tracking device shown in Fig. 3 further include:
Removal unit 260, if extract prospect from monitoring video frame for the extraction unit 210, monitoring view
There are overlapped prospects in frequency frame, then remove the lesser prospect of area in prospect overlapped in monitoring video frame;
Correspondingly, first determination unit 220, specifically for being determined according to the prospect not being removed in monitoring video frame
First tracking box.
Referring to Figure 4 together, it is the structural schematic diagram of another monitoring tracking device provided in an embodiment of the present invention, is scheming
On the basis of monitoring tracking device shown in 2, in the monitoring tracking device shown in Fig. 4, second determination unit 240, packet
It includes:
First determines that subelement 241 supervises the target for the target monitoring video frame for extracting prospect for the first time
The first tracking box in control video frame is determined as the target following frame of target monitoring video frame;Wherein, second tracking box is
It is obtained after being handled using the default track algorithm the first tracking box in the target monitoring video frame;
Second determines subelement 242, for for the monitoring video frame after the target monitoring video frame, by described the
Two tracking box are determined as the target following frame of monitoring video frame, and according in each monitoring video frame after target monitoring video frame
The first tracking box the target following frame is modified.
In an alternative embodiment, it is described second determine subelement 242, can be specifically used for for target monitoring video frame it
Any monitoring video frame afterwards, if meeting the first preset condition, the target following frame is revised as in the monitoring video frame
One tracking box;If meeting the second preset condition, delete target tracking box, and according to the first tracking box in the monitoring video frame
Redefine target following frame;If being unsatisfactory for first preset condition, and it is unsatisfactory for second preset condition, then keeps institute
It is constant to state target following frame;
Wherein, the area and the first tracking box in the monitoring video frame that first preset condition is target following frame
The ratio of area is greater than first threshold, is less than second threshold, and the first tracking box in target following frame and the monitoring video frame
There are overlappings;Second preset condition is the area of target following frame and the area of the first tracking box in the monitoring video frame
Ratio be less than or equal to third threshold value or be more than or equal to the 4th threshold value, alternatively, the central point of target following frame and the monitor video
The distance between central point of the first tracking box in frame is greater than the half of the width and the sum of height of video image;The second threshold
Greater than the first threshold, the third threshold value is less than the first threshold, and the 4th threshold value is greater than the second threshold.
Please refer to fig. 5, scheming for the structural schematic diagram of another monitoring tracking device provided in an embodiment of the present invention
On the basis of monitoring tracking device shown in 2, the monitoring tracking device shown in Fig. 5 further include:
Alarm Unit 270, the central point or the target following frame for working as the target following frame are from monitor video
Object detection area in when being moved to outside object detection area, carry out alarming processing.
The function of each unit and the realization process of effect are specifically detailed in the above method and correspond to step in above-mentioned apparatus
Realization process, details are not described herein.
For device embodiment, since it corresponds essentially to embodiment of the method, so related place is referring to method reality
Apply the part explanation of example.The apparatus embodiments described above are merely exemplary, wherein described be used as separation unit
The unit of explanation may or may not be physically separated, and component shown as a unit can be or can also be with
It is not physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to actual
The purpose for needing to select some or all of the modules therein to realize the present invention program.Those of ordinary skill in the art are not paying
Out in the case where creative work, it can understand and implement.
As seen from the above-described embodiment, by carrying out foreground extraction to the monitor video that gets, and according to extracting before
Scape determines the first tracking box, and then is handled using default track algorithm the first tracking box, stronger to obtain stability
Second tracking box, and target following frame is determined according to the first tracking box and the second tracking box, and according to the target following frame to prison
Control object is tracked, and the automatic monitoring tracking of monitored object is realized, and be ensure that and be monitored tracking to monitored object
Accuracy.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to of the invention its
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the invention, these modifications, purposes or
Person's adaptive change follows general principle of the invention and including the undocumented common knowledge in the art of the present invention
Or conventional techniques.The description and examples are only to be considered as illustrative, and true scope and spirit of the invention are by following
Claim is pointed out.
It should be understood that the present invention is not limited to the precise structure already described above and shown in the accompanying drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present invention is limited only by the attached claims.
Claims (10)
1. a kind of monitoring tracking is applied to video monitoring system characterized by comprising
Foreground extraction is carried out to the monitor video got;
The first tracking box is determined according to the prospect extracted;
First tracking box is handled using default track algorithm, to obtain stronger second tracking box of stability;
Target following frame is determined according to first tracking box and second tracking box;
Monitored object is tracked according to the target following frame;
It is wherein, described that target following frame is determined according to first tracking box and the second tracking box, comprising:
For extracting the target monitoring video frame of prospect for the first time, the first tracking box in the target monitoring video frame is determined
For the target following frame of target monitoring video frame;Wherein, second tracking box is using the default track algorithm to described
What the first tracking box in target monitoring video frame obtained after being handled;
For the monitoring video frame after the target monitoring video frame, second tracking box is determined as monitoring video frame
Target following frame, and according to the first tracking box in each monitoring video frame after target monitoring video frame to the target following
Frame is modified.
2. the method according to claim 1, wherein the described pair of monitor video that gets carries out foreground extraction,
Include:
Mixed Gaussian background modeling is carried out to the monitoring video frame got, extracts the prospect in the monitoring video frame, and
Interference filtering is carried out to the prospect extracted;
The prospect that the basis is extracted determines the first tracking box, comprising:
The first tracking box is determined according to the filtered prospect of interference.
3. the method according to claim 1, wherein the method also includes:
If extract prospect from monitoring video frame, there are overlapped prospects in the monitoring video frame, then regard monitoring
The lesser prospect removal of area in overlapped prospect in frequency frame;
The prospect that the basis is extracted determines the first tracking box, comprising:
The first tracking box is determined according to the prospect not being removed in monitoring video frame.
4. the method according to claim 1, wherein each monitoring after the video frame according to target monitoring regards
The first tracking box in frequency frame is modified the target following frame, comprising:
For any monitoring video frame after target monitoring video frame, if meeting the first preset condition, by target following frame
The first tracking box being revised as in the monitoring video frame;
If meeting the second preset condition, delete target tracking box, and again according to the first tracking box in the monitoring video frame
Determine target following frame;
If being unsatisfactory for first preset condition, and it is unsatisfactory for second preset condition, then keeps the target following frame not
Become;
Wherein, first preset condition is the area of target following frame and the area of the first tracking box in the monitoring video frame
Ratio be greater than first threshold, be less than second threshold, and the first tracking box in target following frame and the monitoring video frame exists
Overlapping;Second preset condition is the ratio of the area of target following frame and the area of the first tracking box in the monitoring video frame
Value is less than or equal to third threshold value or is more than or equal to the 4th threshold value, alternatively, in the central point of target following frame and the monitoring video frame
The distance between the central point of the first tracking box be greater than video image width and the sum of height half;The second threshold is greater than
The first threshold, the third threshold value are less than the first threshold, and the 4th threshold value is greater than the second threshold.
5. the method according to claim 1, wherein described carry out monitored object according to the target following frame
After tracking, the method also includes:
When the central point of the target following frame or the target following frame are moved out of object detection area in monitor video
When outside to object detection area, alarming processing is carried out.
6. a kind of monitoring tracking device is applied to video monitoring system characterized by comprising
Extraction unit, for carrying out foreground extraction to the monitor video got;
First determination unit, for determining the first tracking box according to the prospect extracted;
Processing unit, it is stronger to obtain stability for being handled using default track algorithm first tracking box
Second tracking box;
Second determination unit, for determining target following frame according to first tracking box and second tracking box;
Tracking cell, for being tracked according to the target following frame to monitored object;
Wherein, second determination unit, comprising:
First determines subelement, for the target monitoring video frame for extracting prospect for the first time, by the target monitoring video
The first tracking box in frame is determined as the target following frame of target monitoring video frame;Wherein, second tracking box is utilizes institute
It states and obtains after default track algorithm handles the first tracking box in the target monitoring video frame;
Second determines subelement, for for the monitoring video frame after the target monitoring video frame, described second to be tracked
Frame is determined as the target following frame of monitoring video frame, and according to first in each monitoring video frame after target monitoring video frame
Tracking box is modified the target following frame.
7. device according to claim 6, which is characterized in that
The extraction unit extracts described specifically for carrying out mixed Gaussian background modeling to the monitoring video frame got
Prospect in monitoring video frame, and interference filtering is carried out to the prospect extracted;
First determination unit, specifically for determining the first tracking box according to the filtered prospect of interference.
8. device according to claim 6, which is characterized in that described device further include:
Removal unit exists in the monitoring video frame if extract prospect from monitoring video frame for the extraction unit
Overlapped prospect then removes the lesser prospect of area in prospect overlapped in monitoring video frame;
First determination unit, specifically for determining the first tracking box according to the prospect not being removed in monitoring video frame.
9. device according to claim 6, which is characterized in that
Described second determines subelement, specifically for for any monitoring video frame after target monitoring video frame, if meeting
First preset condition, then the first tracking box being revised as target following frame in the monitoring video frame;If meeting the second default item
Part, then delete target tracking box, and target following frame is redefined according to the first tracking box in the monitoring video frame;If discontented
Foot first preset condition, and be unsatisfactory for second preset condition, then keep the target following frame constant;
Wherein, first preset condition is the area of target following frame and the area of the first tracking box in the monitoring video frame
Ratio be greater than first threshold, be less than second threshold, and the first tracking box in target following frame and the monitoring video frame exists
Overlapping;Second preset condition is the ratio of the area of target following frame and the area of the first tracking box in the monitoring video frame
Value is less than or equal to third threshold value or is more than or equal to the 4th threshold value, alternatively, in the central point of target following frame and the monitoring video frame
The distance between the central point of the first tracking box be greater than video image width and the sum of height half;The second threshold is greater than
The first threshold, the third threshold value are less than the first threshold, and the 4th threshold value is greater than the second threshold.
10. device according to claim 6, which is characterized in that described device further include:
Alarm Unit, for when the target following frame central point or the target following frame from monitor video target inspection
When being moved to outside object detection area in survey region, alarming processing is carried out.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610556686.1A CN106204653B (en) | 2016-07-13 | 2016-07-13 | A kind of monitoring tracking and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610556686.1A CN106204653B (en) | 2016-07-13 | 2016-07-13 | A kind of monitoring tracking and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106204653A CN106204653A (en) | 2016-12-07 |
CN106204653B true CN106204653B (en) | 2019-04-30 |
Family
ID=57475843
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610556686.1A Active CN106204653B (en) | 2016-07-13 | 2016-07-13 | A kind of monitoring tracking and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106204653B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108665476B (en) * | 2017-03-31 | 2022-03-11 | 华为技术有限公司 | Pedestrian tracking method and electronic equipment |
CN111052753A (en) * | 2017-08-30 | 2020-04-21 | Vid拓展公司 | Tracking video scaling |
CN109191495B (en) * | 2018-07-17 | 2022-02-15 | 东南大学 | Black smoke vehicle detection method based on self-organizing background difference model and multi-feature fusion |
CN109472226B (en) * | 2018-10-29 | 2021-07-09 | 上海交通大学 | Sleeping behavior detection method based on deep learning |
CN110969097B (en) * | 2019-11-18 | 2023-05-12 | 浙江大华技术股份有限公司 | Method, equipment and storage device for controlling linkage tracking of monitoring target |
CN113766175A (en) * | 2020-06-04 | 2021-12-07 | 杭州萤石软件有限公司 | Target monitoring method, device, equipment and storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101226691A (en) * | 2007-12-21 | 2008-07-23 | 北京中星微电子有限公司 | Vehicle count method based on video image |
CN101739686A (en) * | 2009-02-11 | 2010-06-16 | 北京智安邦科技有限公司 | Moving object tracking method and system thereof |
CN103489199A (en) * | 2012-06-13 | 2014-01-01 | 通号通信信息集团有限公司 | Video image target tracking processing method and system |
-
2016
- 2016-07-13 CN CN201610556686.1A patent/CN106204653B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101226691A (en) * | 2007-12-21 | 2008-07-23 | 北京中星微电子有限公司 | Vehicle count method based on video image |
CN101739686A (en) * | 2009-02-11 | 2010-06-16 | 北京智安邦科技有限公司 | Moving object tracking method and system thereof |
CN103489199A (en) * | 2012-06-13 | 2014-01-01 | 通号通信信息集团有限公司 | Video image target tracking processing method and system |
Also Published As
Publication number | Publication date |
---|---|
CN106204653A (en) | 2016-12-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106204653B (en) | A kind of monitoring tracking and device | |
CN104123544B (en) | Anomaly detection method and system based on video analysis | |
CN109117827A (en) | Work clothes work hat wearing state automatic identifying method and alarm system based on video | |
Nait-Charif et al. | Activity summarisation and fall detection in a supportive home environment | |
CN211293956U (en) | AI-based identification and alarm system for abnormal agent on construction site | |
CN111144291B (en) | Video monitoring area personnel intrusion discrimination method and device based on target detection | |
CN107911653A (en) | The module of intelligent video monitoring in institute, system, method and storage medium | |
Lim et al. | iSurveillance: Intelligent framework for multiple events detection in surveillance videos | |
CN109377697A (en) | Rapid Alarm method of disposal under a kind of intensive camera head environment | |
KR101884611B1 (en) | Method of extracting objects-of-Interest out of CCTV video using metadate filtering of moving objects | |
CN103576660A (en) | Intelligent home monitoring system and method | |
KR20160011523A (en) | Device and method surveiling innormal behavior using 3d image information | |
CN113486777A (en) | Behavior analysis method and device for target object, electronic equipment and storage medium | |
CN103150736A (en) | Camera motion detecting method based on video monitoring | |
CN107688793A (en) | A kind of outside transformer substation fire automatic monitoring method for early warning | |
CN103870824A (en) | Method and device for capturing face in face detecting and tracking process | |
CN113887445A (en) | Method and system for identifying standing and loitering behaviors in video | |
KR101454644B1 (en) | Loitering Detection Using a Pedestrian Tracker | |
CN103049748B (en) | Behavior monitoring method and device | |
KR101572366B1 (en) | Kidnapping event detector for intelligent video surveillance system | |
CN103093197A (en) | Monitoring method and system for recognizing hanging behavior | |
Tao et al. | Real-time detection and tracking of moving object | |
KR101581162B1 (en) | Automatic detection method, apparatus and system of flame, smoke and object movement based on real time images | |
CN113505709A (en) | Method and system for monitoring dangerous behaviors of human body in real time | |
CN112532927A (en) | Intelligent safety management and control system for construction site |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |