CN107920223A - A kind of object behavior detection method and device - Google Patents
A kind of object behavior detection method and device Download PDFInfo
- Publication number
- CN107920223A CN107920223A CN201610875498.5A CN201610875498A CN107920223A CN 107920223 A CN107920223 A CN 107920223A CN 201610875498 A CN201610875498 A CN 201610875498A CN 107920223 A CN107920223 A CN 107920223A
- Authority
- CN
- China
- Prior art keywords
- destination object
- behavior
- image
- time period
- preset time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
Abstract
An embodiment of the present invention provides a kind of object behavior detection method and device.The described method includes：The first image is obtained, detects in the predeterminable area of the first image and whether there is destination object；The current state of destination object is determined according to testing result；According to the laststate and current state of destination object, judge whether the state of destination object changes；If it is, obtaining the first collection moment of described first image, and laststate is updated to current state；According to the relation between the first collection moment and preset time period, the behavior of destination object is determined.The embodiment of the present invention can improve the detection efficiency to object behavior.
Description
Technical field
This application involves intelligent Video Surveillance Technology field, more particularly to a kind of object behavior detection method and device.
Background technology
With the development of Video Supervision Technique, some important places often can all install camera, to obtain these
The monitor video in place.These monitor videos can be provided the information in monitoring place by people.At present, can be regarded according to monitoring
Late behavior of the frequency to the personnel in workplace is monitored.
In order to which the personnel for realizing to workplace are late the detection of behavior, in the prior art, usually using manually checking prison
The mode of video is controlled, testing staff whether there is late behavior.But, it is necessary to manual identified people by the way of manually checking
Member arrives the time of seat, and determines that personnel whether there is late behavior.The detection method of this late behavior is less efficient.
The content of the invention
The purpose of the embodiment of the present application is the provision of a kind of object behavior detection method and device, to improve to object row
For detection efficiency.
In order to achieve the above object, the invention discloses a kind of object behavior detection method, including：
Obtain the first image；
Detect in the predeterminable area of described first image and whether there is destination object；
The current state of the destination object is determined according to testing result；
According to the laststate of the destination object and the current state, judge whether the state of the destination object is sent out
Changing；
If it is, obtaining the first collection moment of described first image, and the laststate is updated to described work as
Preceding state；
According to the relation between the described first collection moment and preset time period, the behavior of the destination object is determined.
Optionally, it whether there is destination object in the predeterminable area of the detection described first image, including：
Detect in the predeterminable area of described first image and whether there is doubtful human face region；
If in the presence of judging the doubtful human face region, whether the feature of the destination object with prestoring matches；
If matching, judge that there are destination object in the predeterminable area of described first image.
Optionally, when the current state shows that the destination object occurs, the laststate shows the target pair
During as not occurring, the relation according between the described first collection moment and preset time period, determines the destination object
Behavior, including：
When described first gather moment ∈ (a, b] when, and the destination object is first in the preset time period
In the case of secondary appearance, determine that the destination object has late behavior；Wherein, a is the starting of the preset time period
Moment, the b are the end time of the preset time period.
Optionally, when the current state shows that the destination object occurs, the laststate shows the target pair
During as not occurring, the relation according between the described first collection moment and preset time period, determines the destination object
Behavior, including：
When described first gather moment ∈ (a, b] when, and the destination object has gone out in the preset time period
In the case of now crossing, determine that the destination object has behavior of leaving the table；Wherein, when a is the starting of the preset time period
Carve, the b is the end time of the preset time period.
Optionally, the method further includes：
When destination object is not present in the predeterminable area for detect described first image, the first collection moment is equal to institute
When stating b, and in the case that the destination object had occurred in the preset time period, determine that the destination object is deposited
In the behavior of leaving early.
Optionally, the method further includes：
When destination object is not present in the predeterminable area for detect described first image, the first collection moment is equal to institute
When stating b, and in the case that the destination object did not occur in the preset time period, determine that the destination object is deposited
In absent behavior.
Optionally, the method further includes：
When there are during the destination object, mould is detected according to the dressing previously generated in the predeterminable area of described first image
Whether type, the dressing for detecting the destination object meet the requirements；
According to testing result, the dressing behavior of the destination object is determined.
Optionally, it is described according to testing result, determine the dressing behavior of the destination object, including：
When being equal to the b at the described first collection moment, the satisfactory image of dressing of the destination object is counted
First quantity, counts the second quantity of the undesirable image of dressing of the destination object；
According to first quantity and second quantity, the dressing behavior of the destination object is determined.
In order to achieve the above object, the invention also discloses a kind of object behavior detection device, including：
Image obtains module, for obtaining the first image；
Obj ect detection module, whether there is destination object in the predeterminable area for detecting described first image；
State determining module, for determining the current state of the destination object according to testing result；
Condition judgment module, for the laststate according to the destination object and the current state, judges the mesh
Whether the state of mark object changes；
Moment acquisition module, for when the state of the destination object changes, obtaining the of described first image
One collection moment, and the laststate is updated to the current state；
Behavior determining module, for according to the relation between the described first collection moment and preset time period, determining described
The behavior of destination object.
Optionally, the obj ect detection module, including：
Detection sub-module, whether there is doubtful human face region in the predeterminable area for detecting described first image；
Judging submodule, for when, there are during doubtful human face region, sentencing in the predeterminable area for detecting described first image
Breaking, whether the feature of the destination object with prestoring matches the doubtful human face region；
Decision sub-module, judges the doubtful human face region and the feature phase of the destination object prestored for working as
Timing, judges that there are destination object in the predeterminable area of described first image.
Optionally, the behavior determining module, is specifically used for：
When the current state shows that the destination object occurs, the laststate shows that the destination object does not go out
Now, when described first gather moment ∈ (a, b] when, and the destination object be in the preset time period for the first time go out
In the case of existing, determine that the destination object has late behavior；Wherein, a is the initial time of the preset time period,
The b is the end time of the preset time period.
Optionally, the behavior determining module, is specifically used for：
When the current state shows that the destination object occurs, the laststate shows that the destination object does not go out
Now, when described first gather moment ∈ (a, b] when, and the destination object had occurred in the preset time period
In the case of, determine that the destination object has behavior of leaving the table；Wherein, a be the preset time period initial time, institute
State the end time that b is the preset time period.
Optionally, described device further includes behavior determining module of leaving early；
The behavior determining module of leaving early, detects in the predeterminable area of described first image for working as and target pair is not present
As, first collection moment is when being equal to the b, and the destination object had occurred in the preset time period
In the case of, determine that the destination object has behavior of leaving early.
Optionally, described device further includes absent behavior determining module；
The absence behavior determining module, detects in the predeterminable area of described first image for working as and target pair is not present
As, first collection moment is when being equal to the b, and the destination object did not occur in the preset time period
In the case of, determine that the destination object has absent behavior.
Optionally, described device further includes dressing detection module；
The dressing detection module, for when in the predeterminable area of described first image there are during the destination object, root
According to the dressing detection model previously generated, whether the dressing for detecting the destination object meets the requirements；
The behavior determining module, specifically for according to testing result, determining the dressing behavior of the destination object.
Optionally, the behavior determining module, including：
Statistic submodule, for when being equal to the b at the described first collection moment, counting the dressing symbol of the destination object
The first quantity of desired image is closed, counts the second quantity of the undesirable image of dressing of the destination object；
Determination sub-module, for according to first quantity and second quantity, determining the dressing of the destination object
Behavior.
As seen from the above technical solution, in the embodiment of the present application, in the predeterminable area of the first image for detecting acquisition first
With the presence or absence of destination object, the current state of the destination object is determined according to testing result, according to a upper shape for destination object
State and current state, judge whether the state of destination object changes, if it is, when obtaining the first collection of the first image
Carve, and laststate is updated to current state；According to the relation between the first collection moment and preset time period, determine described
The behavior of destination object.
That is, when the embodiment of the present application changes according to the laststate and current state of destination object, first
The relation between moment and preset time period is gathered, determines the behavior of destination object, it is not necessary to the behavior of artificial detection object, because
This can improve the detection efficiency to object behavior.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is attached drawing needed in technology description to be briefly described.It should be evident that drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is a kind of flow diagram of object behavior detection method provided by the embodiments of the present application；
Fig. 2 is several behaviors and the schematic diagram of time relationship of destination object；
Fig. 3 is a kind of structure diagram of object behavior detecting system provided by the embodiments of the present application；
Fig. 4 is a kind of structure diagram of object behavior detection device provided by the embodiments of the present application.
Embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Whole description.Obviously, described embodiment is only the part of the embodiment of the present invention, instead of all the embodiments.Base
Embodiment in the present invention, those of ordinary skill in the art are obtained all on the premise of creative work is not made
Other embodiment, belongs to the scope of protection of the invention.
The embodiment of the present application provides a kind of object behavior detection method and device, is imitated with the detection improved to object behavior
Rate.Object behavior detection method and device provided by the embodiments of the present application can be applied to terminal or server, can also apply
In other electronic equipments, the application is not specifically limited this.
Below by specific embodiment, the application is described in detail.
Fig. 1 is a kind of flow diagram of object behavior detection method provided by the embodiments of the present application, the described method includes
Following steps：
Step S101：Obtain the first image.
Specifically, obtain the first image, can be the image that is gathered in real time from monitor video collecting device or from
The image obtained in the monitor video prerecorded.It is of course also possible to it is the image obtained otherwise, the application is to this
It is not specifically limited.
The object that the present embodiment is detected has a certain range of fixed position in the picture.Therefore, in monitor video
Scene, can be office space, and the present embodiment can examine the behavior of the staff in office space in this case
Survey；Can be conference meeting-place, the present embodiment can be detected the behavior of the speechmaker in meeting-place in this case.This Shen
Please the social role of the personnel to being detected is not specifically limited.
Step S102：Detect in the predeterminable area of described first image and whether there is destination object.
Specifically, predeterminable area is pre-set according to fixed position of the object in the first image.Predeterminable area
Any position that can be located in the first image.Predeterminable area can be the region for having in image certain regular shape, in advance
Predeterminable area is first set, that is, pre-sets the coordinate range in region in monitored picture.
Above-mentioned destination object can be people, naturally it is also possible to be animal, e.g., pet dog etc., the application, which does not do this, to be had
Body limits.Mainly the people in monitoring scene is monitored in being monitored due to actual video, as a kind of embodiment,
Above-mentioned destination object is behaved.
In the present embodiment, step S102, that is, detect and whether there is destination object in the predeterminable area of the first image, can wrap
Include numerous embodiments：
First, it whether there is human face region in the predeterminable area of the first image of detection, if it is, determining the first image
There are destination object in predeterminable area.
Due under normal circumstances, the destination object in predeterminable area is that comparison is fixed, its in addition to destination object
He will not frequently appear in the predeterminable area by destination object.Therefore, need to only detect in this embodiment in predeterminable area is
It is no there are face, it is whether identical without distinguishing the face detected in each image, it is possible to increase detection efficiency.
Second, it whether there is doubtful human face region in detecting the predeterminable area of described first image；If in the presence of judging institute
Stating doubtful human face region, whether the feature of the destination object with prestoring matches；If matching, judges described first image
Predeterminable area in there are destination object.
In this embodiment, the feature of destination object is stored in terminal, server or other electronic equipments in advance,
And the doubtful human face region detected is matched with this feature, i.e. judge doubtful human face region whether all with above-mentioned
Feature, if it is, determining that doubtful human face region matches with this feature.
Wherein, the feature of the destination object prestored, can be the facial image according to the destination object prestored
Obtain.
For example, in workplace, in order to detect the behavior of employee A (i.e. A is destination object), it can gather A's in advance
Facial image, obtains the feature of the facial image of A, by this feature storage in the electronic device., can be with when detecting the behavior of A
It whether there is doubtful human face region first in detection image in predeterminable area, if it is present obtaining the spy of doubtful human face region
Sign, and judges that the feature of doubtful human face region is matched with the feature of the facial image of the A prestored, if match into
Work(, it is determined that employee A is detected in the predeterminable area of image.Employee B enters the operating position of employee A, then can not be in image
Predeterminable area in detect that there are destination object.Therefore, the above embodiment can improve the accuracy of detection.
It should be noted that detection human face region belongs to the prior art in the picture, details are not described herein again for its detailed process.
Step S103：The current state of the destination object is determined according to testing result.
Above-mentioned predeterminable area whether is appeared according to destination object, the state of destination object can be divided into appearance and do not had
There are two states.
When there are the current state for during destination object, determining destination object being in the predeterminable area for detecting the first image
It is existing.When destination object is not present in the predeterminable area for detecting the first image, the current state for determining destination object is not have
Occur.
Since monitor video is made of great amount of images, when being detected to each image, one can be determined
The state of destination object, the state or to occur, or not occur.For the first image of current detection, definite target
The state of object is current state, and a upper image for the first image for current detection, the shape of definite destination object
State is laststate.
Step S104：According to the laststate of the destination object and the current state, the destination object is judged
Whether state changes, if it is, performing step S105.Otherwise, disregard.
In the present embodiment, the laststate of the destination object and the current state, judge the shape of the destination object
Whether state changes, and can specifically include numerous embodiments：
When the laststate of destination object is occurs, current state is when occurring, to judge that the state of destination object is not sent out
Changing；
When the laststate of destination object not occur, current state is that the shape of destination object is judged when not occurring
State does not change；
When the laststate of destination object is occurs, current state is that the state hair of destination object is judged when not occurring
Changing；
When the laststate of destination object not occur, current state is that the state hair of destination object is judged when occurring
Changing.
Step S105：The first collection moment of the first image is obtained, and the laststate is updated to the current shape
State.
As the embodiment of the present embodiment, the first collection moment of the first image is obtained, can be included：According to opening
At the time of beginning recorded video, the position of frame per second and the first image in whole video, determine the first image first collection the moment.
For example, 0 separately beginning recorded video when 8, code check is 25 frames/s, and the first image is the 10000th frame of video, then the first image
00 6 divides 40 seconds when dividing+10000 frames/(25 frames/s)=8 when first collection moment was 8.
Can also for each image of collection, store the collection moment of the image, according to the collection moment of storage,
Directly acquire the first collection moment of the first image.
Step S106：According to the relation between the first collection moment and preset time period, the row of the destination object is determined
For.
It is understood that preset time period can be understood as under normal circumstances, destination object should be in predeterminable area
In period, preset time period determines by initial time and end time.
Preset time period can be represented with the absolute moment.For example, for workplace, work hours 8：00, when coming off duty
Between be 16：00, then preset time period could be provided as 8：00—16：00.Wherein, 8：00 be initial time, 16：00 is
End time.
Preset time period can also be represented with relative instant.For example, the work hours are 0, the quitting time 8, then default
Period could be provided as [0,8].Wherein, 0 is initial time, and 8 be end time.
Relation between first collection moment t and preset time period [a, b] can include：T ＜ a, t=a, a ＜ t ＜ b, t
=b, t ＞ b, and the combination of the above situation.Wherein, a is the initial time of preset time period, and b is the termination of preset time period
Moment.
Different, the behavior to the destination object that should determine that according to the relation between the first collection moment and preset time period
Difference, the behavior of destination object can include：Normally take seat, be late, leave the table.
In addition, in the embodiment of the present embodiment, can also include：Record the first image, the first collection moment
And corresponding behavior.In more specifically embodiment, can also obtain the forward default quantity image of the first image and
The default quantity image of first image backward, according to the forward default quantity image of the first image and the first image backward
Default quantity image, generates the first video, the first video of record, the first collection moment and corresponding behavior.
It is understood that the evidence of respective behavior can occur as destination object for the above- mentioned information of record.
As shown in the above, in the present embodiment, whether there is in the predeterminable area for the first image for detecting acquisition first
Destination object, the current state of the destination object is determined according to testing result, according to the laststate of destination object and currently
State, judges whether the state of destination object changes, if it is, the first collection moment of the first image is obtained, and will
Laststate is updated to current state；According to the relation between the first collection moment and preset time period, the target pair is determined
The behavior of elephant.That is, when the embodiment of the present application changes according to the laststate and current state of destination object, first
The relation between moment and preset time period is gathered, determines the behavior of destination object, it is not necessary to the behavior of artificial detection object, because
This can improve the detection efficiency to object behavior.
In the embodiment shown in fig. 1, according to the laststate of destination object and the different content of current state, it may be determined that
There are different behaviors for destination object.In order to more specifically determine the different behaviors of destination object, embodiment illustrated in fig. 1 can wrap
Include different embodiments.Introduce separately below.
In another embodiment herein, in the embodiment shown in fig. 1, step S106, according to the described first collection
Relation between moment and preset time period, determines the behavior of the destination object, can include：
When the current state shows that the destination object occurs, the laststate shows that the destination object does not go out
Now, when described first gather moment ∈ (a, b] when, and the destination object be in the preset time period for the first time go out
In the case of existing, determine that the destination object has late behavior.
Wherein, the ∈ is belongs to symbol, and a is the initial time of the preset time period, and the b is described default
The end time of period, (a, b] represent not including a points, include b points.
It should be noted that late behavior refers to that the time that destination object occurs for the first time in preset time period has been later than
Begin the moment.
Specifically, after the current state of destination object is determined, can record and store the current state of destination object with
And the number that each state occurs, include the occurrence number of destination object, destination object is without the number occurred.Wherein, target
Object can be understood as the number that destination object leaves predeterminable area without the number occurred.
In the specific implementation of the present embodiment, target can be determined according to the occurrence number of the destination object of storage
Whether object is to occur for the first time in preset time period.
If it is understood that the first collection moment t=a, then it is assumed that destination object is normally taken seat, and there is no late row
For.
In another embodiment herein, in the embodiment shown in fig. 1, step S106, according to the described first collection
Relation between moment and preset time period, determines the behavior of the destination object, can include：
When the current state shows that the destination object occurs, the laststate shows that the destination object does not go out
Now, when described first gather moment ∈ (a, b] when, and the destination object had occurred in the preset time period
In the case of, determine that the destination object has behavior of leaving the table.
It should be noted that the behavior of leaving the table refers to that destination object once occurred in preset time period, but midway from
Seat has been opened, also, seat is returned again before end time.
Specifically, it can determine that destination object is in preset time period according to the occurrence number of the destination object of storage
It is no to have occurred.
If it is understood that there is the behavior of leaving the table in preset time period in destination object, then it is considered that target
Object first has to occur in preset time period, and then, destination object is made leave seat successively, to return to seat etc. dynamic
Make.When destination object returns to seat again, and it is no more than end time at the time of return seat, can just thinks that destination object is deposited
In the behavior of leaving the table.
It is worth noting that in preset time period, behavior of repeatedly leaving the table may occur for destination object.Therefore, this implementation
In the embodiment of example, the number for the behavior of leaving the table can also be recorded.
In another embodiment herein, it can also include in embodiment illustrated in fig. 1：
When destination object is not present in the predeterminable area for detect described first image, the first collection moment is equal to institute
When stating b, and in the case that the destination object had occurred in the preset time period, determine that the destination object is deposited
In the behavior of leaving early.
It should be noted that the behavior of leaving early refers to that destination object once occurred in preset time period, but midway from
Seat has been opened, and until end time, destination object all do not return again to seat.Destination object in preset time period
In the case of occurring, only destination object is not detected in the predeterminable area of corresponding first images of end time b,
It just can determine that destination object has behavior of leaving early.
In another embodiment herein, embodiment illustrated in fig. 1 can also include：
When destination object is not present in the predeterminable area for detect described first image, the first collection moment is equal to institute
When stating b, and in the case that the destination object did not occur in the preset time period, determine that the destination object is deposited
In absent behavior.
It should be noted that absent behavior refers to that destination object did not all occur from start to finish in preset time period.
Destination object is not only detected in the predeterminable area of corresponding first images of end time b, and destination object is pre-
If in the case of never being occurred in the period, it just can determine that destination object has absent behavior.
In addition, whether meeting the requirements for the dressing of detected target object, in another embodiment herein, scheming
It can also include in 1 illustrated embodiment：
Step 1：When in the predeterminable area of described first image there are during the destination object, according to the dressing previously generated
Whether detection model, the dressing for detecting the destination object meet the requirements.
Specifically, the image for including satisfactory dressing can be gathered in advance, and to the dressing in the image into rower
Note.Then, default machine learning model is trained using above-mentioned image, obtains dressing detection model.
Step 2：According to testing result, the dressing behavior of the destination object is determined.
Specifically, when testing result shows that destination object dressing meets the requirements, the dressing behavior of destination object is being determined just
Often.When testing result shows that destination object dressing is undesirable, the dressing abnormal behavior of destination object is determined.
In a kind of embodiment of this step, step 2 can include：
Step 2A：When being equal to the b at the first collection moment, the satisfactory image of dressing of the destination object is counted
The first quantity, count the second quantity of the undesirable image of dressing of the destination object.
Step 2B：According to first quantity and second quantity, the dressing behavior of the destination object is determined.
Specifically, the normal quantitative proportion of destination object dressing behavior can be obtained according to the first quantity and the second quantity,
As the first quantity/(first the+the second quantity of quantity).Judge whether the quantitative proportion is more than preset ratio threshold value, if so,
Then determine that destination object dressing behavior is normal, otherwise, it determines destination object dressing abnormal behavior.
Smaller value, example are could be provided as in view of there are a variety of disturbing factors, preset ratio threshold value during actually detected
Such as 10% or 30%.
Elaborate again to the application with reference to instantiation.
Fig. 2 be late, leave the table, leaving early, the schematic diagram of absent behavior and time relationship.Wherein, a is rising for preset time period
Begin the moment, b is the end time of preset time period.Destination object is represented with A.
Represent that destination object is the state of appearance in the predeterminable area of image in Fig. 2 with the square frame of different the fill styles,
Again without the state of appearance.As shown in (1) in Fig. 2, when the image of current detection is between a~t1, preset areas is determined
A is not present in domain, the laststate and current state of A show that A does not occur, i.e. the state of A does not change.
When the image of current detection is located at t1, i.e. t1 was the first collection moment, determined that A's is upper there are A in predeterminable area
One state shows that A does not occur, and current state shows that A occurs, and judges that the state of A changes at this time, meanwhile, determine t1
Between ab, therefore it can determine that A has late behavior.
When the image of current detection is located at t1~t2, determine in predeterminable area there are A, but the laststate of A and work as
Preceding state shows that A occurs, i.e. the state of A does not change.
When the image of current detection is located at t2, i.e. t2 was the first collection moment, determined that A is not present in predeterminable area, A's
Laststate shows that A occurs, and current state shows that A does not occur, judges that the state of A changes at this time, till now also not
It can determine that A leaves the table or leaves early.
When the image of current detection is located at t2~t3, determine that A is not present in predeterminable area, the laststate of A and current
State shows that A does not occur, i.e. the state of A does not change.
When the image of current detection is located at t3, determine in predeterminable area that there are A, the laststate of A to show that A does not go out
Existing, current state shows that A occurs, and judges that the state of A changes at this time.Simultaneously as A once occurred between ab,
Therefore it can determine that A has behavior of leaving the table at this time.
When the image of current detection is located at t3~t4, determine in predeterminable area there are A, but the laststate of A and work as
Preceding state shows that A occurs, i.e. the state of A does not change.
When the image of current detection is located at t4, i.e. t4 was the first collection moment, determined that A is not present in predeterminable area, A's
Laststate shows that A occurs, and current state shows that A does not occur, judges that the state of A changes at this time, but cannot be true
Determine A and there is behavior of leaving the table still to leave early behavior.
When the image of current detection is located at t4~b, determine that A, the laststate of A and current shape are not present in predeterminable area
State shows that A does not occur, i.e. the state of A does not change.
If the image of current detection is located at b, i.e. b was the first collection moment, determined that A is not present in predeterminable area, then such as
Fruit A within the period from a to b had occurred, then can determine that A has behavior of leaving early.
As shown in (2) in Fig. 2, if the image of current detection is located at b, i.e. b was the first collection moment, determined predeterminable area
In A is not present, then if A did not occur within the period from a to b, can determine that A has absent behavior.
The application can also provide a kind of object behavior detecting system, including video acquisition unit 301, video analysis unit
302nd, data associating unit 303 and violation display unit 304.Fig. 3 is the structure diagram of the system, shown in the system and Fig. 1
Embodiment of the method is corresponding.
Wherein, video acquisition unit 301, the first image of predeterminable area is included for gathering, and image is sent to regarding
Frequency analysis unit 302.
Specific video acquisition unit 301 can be realized by video camera.Different, the video camera according to the situation of different scenes
It can select wall or lifting.
According to practical application request, video acquisition unit 301 can also include light filling equipment.
Video analysis unit 302, for obtaining the first image of the transmission of video acquisition unit 301；Detect first figure
It whether there is destination object in the predeterminable area of picture；The current state of the destination object is determined according to testing result；According to institute
The laststate of destination object and the current state are stated, judges whether the state of the destination object changes；If so,
The first collection moment of described first image is then obtained, and the laststate is updated to the current state；According to described
Relation between first collection moment and preset time period, determines the behavior of the destination object；By described first image, first
The behavior at collection moment and destination object is sent to data associating unit 303.
Specifically, according to the actual requirements and in video acquisition unit video camera type, can be by video analysis unit
302 are integrated into camera chip, by video analysis unit 301 and 302 integrator of video acquisition unit.Certainly, video point
Analysis unit 302 can also be realized by server or embedded device.
Data associating unit 303, for receive video analysis unit 302 transmission described first image, first collection when
The behavior with destination object is carved, and is stored.
Specifically, data associating unit 303 can store above- mentioned information to database, and by above- mentioned information with it is relevant
Destination object is associated, and the above- mentioned information of all destination objects is updated.
Violation display unit 304, for the described first image of storage, first to be gathered to the behavior of moment and destination object
Counted, and generate destination object behavior report, shown in a manner of visual.
Fig. 4 is a kind of structure diagram of object behavior detection device provided by the embodiments of the present application, with method shown in Fig. 1
Embodiment is corresponding, and described device includes：
Image obtains module 401, for obtaining the first image；
Obj ect detection module 402, whether there is destination object in the predeterminable area for detecting described first image；
State determining module 403, for determining the current state of the destination object according to testing result；
Condition judgment module 404, for the laststate according to the destination object and the current state, described in judgement
Whether the state of destination object changes；
Moment acquisition module 405, for when the state of the destination object changes, obtaining described first image
First collection moment, and the laststate is updated to the current state；
Behavior determining module 406, for according to the relation between the described first collection moment and preset time period, determining institute
State the behavior of destination object.
As another embodiment herein, in the embodiment shown in fig. 4, the obj ect detection module 402, can wrap
Include detection sub-module, judging submodule and decision sub-module；(not shown)
Detection sub-module, whether there is doubtful human face region in the predeterminable area for detecting described first image；
Judging submodule, for when, there are during doubtful human face region, sentencing in the predeterminable area for detecting described first image
Breaking, whether the feature of the destination object with prestoring matches the doubtful human face region；
Decision sub-module, judges the doubtful human face region and the feature phase of the destination object prestored for working as
Timing, judges that there are destination object in the predeterminable area of described first image.
As another embodiment herein, in the embodiment shown in fig. 4, the behavior determining module 406 is specific to use
In：
When the current state shows that the destination object occurs, the laststate shows that the destination object does not go out
Now, when described first gather moment ∈ (a, b] when, and the destination object be in the preset time period for the first time go out
In the case of existing, determine that the destination object has late behavior；Wherein, a is the initial time of the preset time period,
The b is the end time of the preset time period.
As another embodiment herein, in the embodiment shown in fig. 4, the behavior determining module 406 is specific to use
In：
When the current state shows that the destination object occurs, the laststate shows that the destination object does not go out
Now, when described first gather moment ∈ (a, b] when, and the destination object had occurred in the preset time period
In the case of, determine that the destination object has behavior of leaving the table.
As another embodiment herein, embodiment illustrated in fig. 4 can also include leaving early behavior determining module (in figure
It is not shown)；
The behavior determining module of leaving early, detects in the predeterminable area of described first image for working as and target pair is not present
As, first collection moment is when being equal to the b, and the destination object had occurred in the preset time period
In the case of, determine that the destination object has behavior of leaving early.
As another embodiment herein, embodiment illustrated in fig. 4 can also include absent behavior determining module (in figure
It is not shown)；
The absence behavior determining module, detects in the predeterminable area of described first image for working as and target pair is not present
As, first collection moment is when being equal to the b, and the destination object did not occur in the preset time period
In the case of, determine that the destination object has absent behavior.
As another embodiment herein, in the embodiment shown in fig. 4, described device can also include dressing and detect
Module (not shown)；
The dressing detection module, for when in the predeterminable area of described first image there are during the destination object, root
According to the dressing detection model previously generated, whether the dressing for detecting the destination object meets the requirements；
The behavior determining module 406, specifically for according to testing result, determining the dressing behavior of the destination object.
As another embodiment herein, in the embodiment shown in fig. 4, the behavior determining module 406, can wrap
Include statistic submodule and determination sub-module；(not shown)
Wherein, statistic submodule, for when being equal to the b at the described first collection moment, counting the destination object
First quantity of the satisfactory image of dressing, counts the second number of the undesirable image of dressing of the destination object
Amount；
Determination sub-module, for according to first quantity and second quantity, determining the dressing of the destination object
Behavior.
Since above device embodiment and system embodiment are all based on what embodiment of the method obtained, there is phase with this method
Same technique effect, therefore details are not described herein for the technique effect of device embodiment and system embodiment.
For device embodiment and system embodiment, since it is substantially similar to embodiment of the method, so describing
Fairly simple, the relevent part can refer to the partial explaination of embodiments of method.
It should be noted that herein, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any this actual relation or order.Moreover, term " comprising ", "comprising" or any other variant be intended to it is non-
It is exclusive to include, so that process, method, article or equipment including a series of elements not only include those key elements,
But also including other elements that are not explicitly listed, or further include solid by this process, method, article or equipment
Some key elements.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that including
Also there are other identical element in the process of the key element, method, article or equipment.
It will appreciated by the skilled person that all or part of step in the above embodiment is can to pass through journey
Sequence instructs relevant hardware, and come what is completed, the program can be stored in computer read/write memory medium.It is designated herein
Storage medium, refers to ROM/RAM, magnetic disc, CD etc..
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all
Any modification, equivalent substitution, improvement and etc. done within the spirit and principles in the present invention, are all contained in protection scope of the present invention
It is interior.
Claims (16)
- A kind of 1. object behavior detection method, it is characterised in that including：Obtain the first image；Detect in the predeterminable area of described first image and whether there is destination object；The current state of the destination object is determined according to testing result；According to the laststate of the destination object and the current state, judge whether the state of the destination object becomes Change；If it is, obtaining the first collection moment of described first image, and the laststate is updated to the current shape State；According to the relation between the described first collection moment and preset time period, the behavior of the destination object is determined.
- 2. according to the method described in claim 1, it is characterized in that, it is described detection described first image predeterminable area in whether There are destination object, including：Detect in the predeterminable area of described first image and whether there is doubtful human face region；If in the presence of judging the doubtful human face region, whether the feature of the destination object with prestoring matches；If matching, judge that there are destination object in the predeterminable area of described first image.
- 3. according to the method described in claim 1, it is characterized in that, when the current state shows that the destination object occurs, When the laststate shows that the destination object does not occur, it is described according to the described first collection moment and preset time period it Between relation, determine the behavior of the destination object, including：When described first gather moment ∈ (a, b] when, and the destination object be in the preset time period for the first time go out In the case of existing, determine that the destination object has late behavior；Wherein, a is the initial time of the preset time period, The b is the end time of the preset time period.
- 4. according to the method described in claim 1, it is characterized in that, when the current state shows that the destination object occurs, When the laststate shows that the destination object does not occur, it is described according to the described first collection moment and preset time period it Between relation, determine the behavior of the destination object, including：When described first gather moment ∈ (a, b] when, and the destination object had occurred in the preset time period In the case of, determine that the destination object has behavior of leaving the table；Wherein, a be the preset time period initial time, institute State the end time that b is the preset time period.
- 5. according to the described method of any one of claim 1-4, it is characterised in that the method further includes：When destination object is not present in the predeterminable area for detect described first image, the first collection moment is equal to the b When, and in the case that the destination object had occurred in the preset time period, determine that the destination object exists Leave early behavior.
- 6. according to the described method of any one of claim 1-4, it is characterised in that the method further includes：When destination object is not present in the predeterminable area for detect described first image, the first collection moment is equal to the b When, and in the case that the destination object did not occur in the preset time period, determine that the destination object exists Absent behavior.
- 7. according to the described method of any one of claim 1-4, it is characterised in that the method further includes：When in the predeterminable area of described first image there are during the destination object, according to the dressing detection model previously generated, Whether the dressing for detecting the destination object meets the requirements；According to testing result, the dressing behavior of the destination object is determined.
- 8. the method according to the description of claim 7 is characterized in that described according to testing result, the destination object is determined Dressing behavior, including：When being equal to the b at the described first collection moment, the first of the satisfactory image of dressing of the destination object is counted Quantity, counts the second quantity of the undesirable image of dressing of the destination object；According to first quantity and second quantity, the dressing behavior of the destination object is determined.
- A kind of 9. object behavior detection device, it is characterised in that including：Image obtains module, for obtaining the first image；Obj ect detection module, whether there is destination object in the predeterminable area for detecting described first image；State determining module, for determining the current state of the destination object according to testing result；Condition judgment module, for the laststate according to the destination object and the current state, judges the target pair Whether the state of elephant changes；Moment acquisition module, first for when the state of the destination object changes, obtaining described first image adopts Collect the moment, and the laststate is updated to the current state；Behavior determining module, for according to the relation between the described first collection moment and preset time period, determining the target The behavior of object.
- 10. device according to claim 9, it is characterised in that the obj ect detection module, including：Detection sub-module, whether there is doubtful human face region in the predeterminable area for detecting described first image；Judging submodule, for when, there are during doubtful human face region, judging institute in the predeterminable area for detecting described first image Stating doubtful human face region, whether the feature of the destination object with prestoring matches；Decision sub-module, for matching when the feature for judging destination object of the doubtful human face region with prestoring When, judge that there are destination object in the predeterminable area of described first image.
- 11. device according to claim 9, it is characterised in that the behavior determining module, is specifically used for：When the current state shows that the destination object occurs, the laststate shows that the destination object does not occur When, when described first gather moment ∈ (a, b] when, and the destination object be in the preset time period for the first time occur In the case of, determine that the destination object has late behavior；Wherein, a be the preset time period initial time, institute State the end time that b is the preset time period.
- 12. device according to claim 9, it is characterised in that the behavior determining module, is specifically used for：When the current state shows that the destination object occurs, the laststate shows that the destination object does not occur When, when described first gather moment ∈ (a, b] when, and the destination object had occurred in the preset time period In the case of, determine that the destination object has behavior of leaving the table；Wherein, a is the initial time of the preset time period, described B is the end time of the preset time period.
- 13. according to the device any one of claim 9-12, it is characterised in that it is true that described device further includes the behavior of leaving early Cover half block；The behavior determining module of leaving early, detects in the predeterminable area of described first image for working as and destination object is not present, When the first collection moment is equal to the b, and the feelings that the destination object had occurred in the preset time period Under condition, determine that the destination object has behavior of leaving early.
- 14. according to the device any one of claim 9-12, it is characterised in that it is true that described device further includes absent behavior Cover half block；The absence behavior determining module, detects in the predeterminable area of described first image for working as and destination object is not present, When the first collection moment is equal to the b, and the feelings that the destination object did not occur in the preset time period Under condition, determine that the destination object has absent behavior.
- 15. according to the device any one of claim 9-12, it is characterised in that described device further includes dressing detection mould Block；The dressing detection module, for when in the predeterminable area of described first image there are during the destination object, according to pre- Whether the dressing detection model first generated, the dressing for detecting the destination object meet the requirements；The behavior determining module, specifically for according to testing result, determining the dressing behavior of the destination object.
- 16. device according to claim 15, it is characterised in that the behavior determining module, including：Statistic submodule, for when being equal to the b at the described first collection moment, the dressing for counting the destination object to conform to First quantity of the image asked, counts the second quantity of the undesirable image of dressing of the destination object；Determination sub-module, for according to first quantity and second quantity, determining the dressing behavior of the destination object.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610875498.5A CN107920223B (en) | 2016-10-08 | 2016-10-08 | Object behavior detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610875498.5A CN107920223B (en) | 2016-10-08 | 2016-10-08 | Object behavior detection method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107920223A true CN107920223A (en) | 2018-04-17 |
| CN107920223B CN107920223B (en) | 2020-08-28 |
Family
ID=61892184
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610875498.5A Active CN107920223B (en) | 2016-10-08 | 2016-10-08 | Object behavior detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107920223B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109345887A (en) * | 2018-09-29 | 2019-02-15 | 北京猎户星空科技有限公司 | Acquisition methods, device and the electronic equipment of task execution duration |
| CN109993436A (en) * | 2019-03-30 | 2019-07-09 | 王浩 | A kind of intelligent building total management system |
| CN110309768A (en) * | 2019-06-28 | 2019-10-08 | 上海眼控科技股份有限公司 | The staff's detection method and equipment of car test station |
| CN110543830A (en) * | 2019-08-12 | 2019-12-06 | 珠海格力电器股份有限公司 | motion detection method, motion detection device, and storage medium |
| CN111723603A (en) * | 2019-03-19 | 2020-09-29 | 杭州海康威视数字技术股份有限公司 | Material monitoring method, system and device |
| CN112150435A (en) * | 2020-09-23 | 2020-12-29 | 徐州联城住宅工业科技发展有限公司 | Construction management method and device for prefabricated part |
| WO2021164644A1 (en) * | 2020-02-18 | 2021-08-26 | 深圳市商汤科技有限公司 | Violation event detection method and apparatus, electronic device, and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001160161A (en) * | 1999-12-02 | 2001-06-12 | Softbrain Co Ltd | Action management system utilizing the internet |
| CN101577812A (en) * | 2009-03-06 | 2009-11-11 | 北京中星微电子有限公司 | Method and system for post monitoring |
| CN202159375U (en) * | 2011-03-14 | 2012-03-07 | 惠州学院 | Classroom roll-call system |
| CN103810696A (en) * | 2012-11-15 | 2014-05-21 | 浙江大华技术股份有限公司 | Method for detecting image of target object and device thereof |
| CN104408406A (en) * | 2014-11-03 | 2015-03-11 | 安徽中科大国祯信息科技有限责任公司 | Staff off-post detection method based on frame difference method and background subtraction method |
| CN105488490A (en) * | 2015-12-23 | 2016-04-13 | 天津天地伟业数码科技有限公司 | Judge dressing detection method based on video |
-
2016
- 2016-10-08 CN CN201610875498.5A patent/CN107920223B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001160161A (en) * | 1999-12-02 | 2001-06-12 | Softbrain Co Ltd | Action management system utilizing the internet |
| CN101577812A (en) * | 2009-03-06 | 2009-11-11 | 北京中星微电子有限公司 | Method and system for post monitoring |
| CN202159375U (en) * | 2011-03-14 | 2012-03-07 | 惠州学院 | Classroom roll-call system |
| CN103810696A (en) * | 2012-11-15 | 2014-05-21 | 浙江大华技术股份有限公司 | Method for detecting image of target object and device thereof |
| CN104408406A (en) * | 2014-11-03 | 2015-03-11 | 安徽中科大国祯信息科技有限责任公司 | Staff off-post detection method based on frame difference method and background subtraction method |
| CN105488490A (en) * | 2015-12-23 | 2016-04-13 | 天津天地伟业数码科技有限公司 | Judge dressing detection method based on video |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109345887A (en) * | 2018-09-29 | 2019-02-15 | 北京猎户星空科技有限公司 | Acquisition methods, device and the electronic equipment of task execution duration |
| CN111723603A (en) * | 2019-03-19 | 2020-09-29 | 杭州海康威视数字技术股份有限公司 | Material monitoring method, system and device |
| CN109993436A (en) * | 2019-03-30 | 2019-07-09 | 王浩 | A kind of intelligent building total management system |
| CN110309768A (en) * | 2019-06-28 | 2019-10-08 | 上海眼控科技股份有限公司 | The staff's detection method and equipment of car test station |
| CN110543830A (en) * | 2019-08-12 | 2019-12-06 | 珠海格力电器股份有限公司 | motion detection method, motion detection device, and storage medium |
| CN110543830B (en) * | 2019-08-12 | 2022-05-13 | 珠海格力电器股份有限公司 | Motion detection method, motion detection device, and storage medium |
| WO2021164644A1 (en) * | 2020-02-18 | 2021-08-26 | 深圳市商汤科技有限公司 | Violation event detection method and apparatus, electronic device, and storage medium |
| CN112150435A (en) * | 2020-09-23 | 2020-12-29 | 徐州联城住宅工业科技发展有限公司 | Construction management method and device for prefabricated part |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107920223B (en) | 2020-08-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107920223A (en) | A kind of object behavior detection method and device | |
| CN108846365B (en) | Detection method and device for fighting behavior in video, storage medium and processor | |
| CN107004271B (en) | Display method, display apparatus, electronic device, computer program product, and storage medium | |
| CN108174165A (en) | Electric power safety operation and O＆M intelligent monitoring system and method | |
| CN108319926A (en) | A kind of the safety cap wearing detecting system and detection method of building-site | |
| CN111507283A (en) | Student behavior identification method and system based on classroom scene | |
| CN110674790B (en) | Abnormal scene processing method and system in video monitoring | |
| CN103957387B (en) | A kind of security protection channel sundries detection method and detection system | |
| CN113411542A (en) | Intelligent working condition monitoring equipment | |
| CN108027973A (en) | Crowded resolver, crowded analytic method and crowded analysis program | |
| CN102306304A (en) | Face occluder identification method and device | |
| CN110059761A (en) | A kind of human body behavior prediction method and device | |
| CN109886555A (en) | The monitoring method and device of food safety | |
| CN106901686B (en) | Execution method, server, test lead and the system of test of eye movement task | |
| CN109867186A (en) | A kind of elevator malfunction detection method and system based on intelligent video analysis technology | |
| CN111178241A (en) | Intelligent monitoring system and method based on video analysis | |
| CN111223011A (en) | Food safety supervision method and system for catering enterprises based on video analysis | |
| CN114821483B (en) | Monitoring method and system capable of measuring temperature and applied to monitoring video | |
| CN207233038U (en) | Face is called the roll and number system | |
| CN111325119A (en) | Video monitoring method and system for safety production | |
| CN110807117A (en) | User relationship prediction method and device and computer readable storage medium | |
| CN111860457A (en) | Fighting behavior recognition early warning method and recognition early warning system thereof | |
| CN115798047A (en) | Behavior recognition method and apparatus, electronic device, and computer-readable storage medium | |
| CN111385530A (en) | Intelligent camera combined encryption method and system | |
| CN113158724A (en) | Monitoring method and system |
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 |