CN108389351A - Tag match, fence border control method, apparatus and storage medium - Google Patents
Tag match, fence border control method, apparatus and storage medium Download PDFInfo
- Publication number
- CN108389351A CN108389351A CN201810373250.8A CN201810373250A CN108389351A CN 108389351 A CN108389351 A CN 108389351A CN 201810373250 A CN201810373250 A CN 201810373250A CN 108389351 A CN108389351 A CN 108389351A
- Authority
- CN
- China
- Prior art keywords
- target
- monitored
- label
- coordinate
- fence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/22—Electrical actuation
- G08B13/24—Electrical actuation by interference with electromagnetic field distribution
- G08B13/2402—Electronic Article Surveillance [EAS], i.e. systems using tags for detecting removal of a tagged item from a secure area, e.g. tags for detecting shoplifting
- G08B13/2451—Specific applications combined with EAS
- G08B13/2462—Asset location systems combined with EAS
Abstract
The present invention relates to electronic security technical field, especially a kind of tag match, fence border control method, apparatus and storage medium.It includes location information to target to be monitored information to combine monocular cam or both to be integrated through the fusion realization of vision positioning technology and wireless distance finding technology using a single base station, the accurate acquisition of identity information and authority information into fence, it can also detect that do not carry the case where label card is put into fence boundary or region and a target carries the case where blocking, convenient unattended monitoring and management for the personnel such as large scale industry, building yard, mining region, prison, equipment etc. more.The system function is comprehensive, and system deployment is simple, facilitates easy to operate, realizes better management effect with less resource, is also greatly reduced cost needed for system.
Description
Technical field
The present invention relates to electronic security technical field, especially a kind of tag match, fence border control method, dress
It sets and storage medium.
Background technology
It needs to carry out management and control to region for requirements such as safety management, secret management and control in many occasions.Area is realized at present
The means of domain management and control are broadly divided into two kinds of physics fence and fence.Physics fence needs a large amount of constructions, and using ineffective
It is living, it can not be changed, generally be can be only applied in application that is fixed and will not changing according to demand.So virtual electronic
Fence obtains more application.
Existing virtual electronic fence usually has following several：First, the infrared emitting using central apparatus area surrounding fills
Infrared ray of the transmitting through ovennodulation is set, an infrared receiving device is placed in position in contrast, converts the signal into electric signal,
Its system administration manager is sent to after processing appropriate.Under normal operating mode, multiple fixed electronics stakes can opposite hair
Go out and receive infrared light, infrared light beam constitutes blockade line together to hair with to receipts, and multiple infrared light beams are just formed
One closed electronics regions, when someone passes through or object stops infrared ray, infrared signal is interrupted, system administration
Machine will give audible alarm and light alarm immediately.Thus achieving the purpose that prevent invasion.
Second is that by determining fence region on the electronic map in advance, then one is laid in corresponding actual place
The locating base station of fixed number amount forms virtual fence region.It is communicated wirelessly by locating base station and terminal to be positioned,
The location information for getting terminal to be positioned determines location information after being compared by location information and fence relative position
Whether it is located in fence.In existing technology, such as selection is using the UWB electronic enclosure system generally uses positioned
The localization method of range difference, theoretical minimum are solved based on TDOA (Time Difference of Arrival reaching time-differences)
It needs to lay 4 locating base stations and just can determine that the two-dimensional position of terminal to be positioned, but in fact, because of protected from environmental, one
A region usually requires more ultra wide band base stations, using more redundant datas, could obtain better locating effect.
The third is that the electronic enclosure system circumference that for example Haikang prestige regards that common camera is monitored monitoring area is anti-
Model system, dependence is that image frame changes and then judge to whether there is in scene mobile object, and is advised according to the alarm of setting
Then alarm, can detect the warning face of passing through, region invasion, into region, leave zone issue.
To sum up, the fence of infrared tube needs artificial arranging electronic stake, deploy troops on garrison duty, needs of withdrawing a garrison it is a large amount of artificial, make
With inconvenience, missing inspection and rate of false alarm are high, and do not have rights management function.Virtual electronic fence system based on base station location needs
The base station number of laying is more, thus cost needs are very high.Another aspect system function is single, is only merely to be positioned according to judgement
Whether card is located in fence area to decide whether triggering alarm, can not solve the personnel area management for not carrying positioning card
Problem can not just alarm without card electronic enclosure system, then have the case where can not exactly obtaining visual information, cards more for band
It can not handle.Maximum disadvantage existing for electronic enclosure system based on common camera is exactly can only target in detection zone
Activity, cannot achieve the differentiation to target identity information, and can not judgment object classification, often occur judging by accident caused by animal is invaded
The case where.
Invention content
The technical problem to be solved by the present invention is to：In view of the above problems, tag match, fence side are provided
Boundary's management method, equipment or storage medium.Using the vision positioning technology of monocular cam and the wireless distance finding of single base station or determine
The fusion of position technology is realized:To prison, large scale industry, the target to be monitored in the specified supervision region such as mining industry, construction site
The identification of position tracking, identity and access entitlements, and whether there is or not carry label card, whether carry the differentiation more blocked.Specifically：
1) by matching target to be monitored and label card, obtaining the corresponding identity of target to be monitored and whether having into certain
The authority information in fence region.
2) according to the unique mark in the position and matched label card of the change in location of target to be monitored or label
Information carries out position tracking to target to be monitored and enters the rights management of fence, realizes and haves no right to enter electronics to target
Fence and the accurate monitoring of illegal invasion；
3) compared with the existing technology in virtual electronic fence, the present invention it is full-featured, can not be taken detecting target
It alarms in the case of tape label and the multiple labels of carrying.
The technical solution adopted by the present invention is as follows：
A kind of tag match method includes：
The target data of target to be monitored based on monocular cam acquisition, obtain target to be monitored and monocular cam it
Between the coordinate of the target range change curve and target to be monitored that change over time in map coordinates system；
Meanwhile according to the range information of single base station and label, obtain the label that is changed over time between label and base station away from
From change curve；
Then, according to the target range change curve and the tag distances change curve, matching label with it is corresponding
Target to be monitored；
The wherein described monocular cam is mounted on same position with single base station.
Further include when label is with corresponding target to be monitored successful match, in conjunction with described based on the tag match method
Coordinate of the target to be monitored in map coordinates system obtains coordinate of the label in map.
A kind of tag match method includes：
The target data of target to be monitored based on monocular cam acquisition obtains target to be monitored in map coordinates system
Coordinate；
Meanwhile coordinate of the label in map is obtained to tag location based on single base station；
Then, the coordinate of coordinate and label in map according to the target to be monitored in map matches label
With corresponding target to be monitored；
The wherein described monocular cam is mounted on same position with single base station.
Further, further include working as label and corresponding target to be monitored successful match according to the tag match method
When, the unique mark in conjunction with possessed by the label including tag ID and the authority information for entering fence obtains the mesh that is monitored
Target identity and the permission for entering fence.
Fence border control method based on the tag match method further includes：
According to coordinate of the label in map, the relative position of label and fence boundary or region is calculated；
Unique mark possessed by combination tag, including tag ID and the authority information for entering fence, judge to carry
Whether the target to be monitored of the label enters fence boundary, makes corresponding alarm response；
When coordinate of the multiple labels of acquisition in map is identical, that is, there are two or more tag match and arrive
The same target to be monitored then judges that the monitoring objective carries more cards, makes corresponding alarm immediately.Based on the label
The fence border control method of matching process further includes：
According to coordinate of the target to be monitored in map, the opposite of target to be monitored and fence boundary or region is calculated
Position；
Whether identity in conjunction with the target to be monitored got and the permission into fence, judge target to be monitored
It is responded into fence boundary or region, and by alarm unit correspondence；
When target to be monitored is without matching label, i.e., the identity and the corresponding electronics that enters that can not get target to be monitored are enclosed
The permission on column boundary, target to be monitored make corresponding alarm immediately once entering fence.
Further, the target range change curve changed over time between the target to be monitored and monocular cam is logical
Cross the acquisition of one of following two methods：
Method 1：
The target data of target to be monitored based on monocular cam acquisition obtains seat of the target to be monitored on map
Mark；According to the coordinate of monocular cam, by the coordinate of the target to be monitored changed over time be converted to target to be monitored with
The target range change curve changed over time between monocular cam；
Method 2：The target data of target to be monitored based on monocular cam acquisition obtains the depth letter of target to be monitored
Breath, the target range changed over time between target to be monitored and monocular cam is fitted according to the depth information of different moments
Change curve.
Further, the coordinate that the acquisition target to be monitored is fastened in map reference is to be based on following two vision positionings
One of method is realized：
Method one：
Step 1：Target detection：
The picture of monocular cam transmission is handled in real time and obtains position (x, y) of the target in picture, is taken the photograph in conjunction with monocular
As the focal length f of head, vectorial (x, y, f) is obtained, determines monitoring objective relative to monocular cam according to the direction of vectorial (x, y, f)
The angle of optical center point；
Step 2：Estimation of Depth：
The coordinate (x, y) of picture and target to be monitored that monocular cam is taken is obtained by depth estimation method
The distance between target to be monitored and monocular cam optical center point d；
Step 3：Target to be monitored is obtained in conjunction with the range information in (x, the y, f) and step 2 in step 1 to sit in camera
Coordinate in mark system；
Step 4：It (spin matrix and is translated towards according to the transformational relation between monocular cam coordinate system and map coordinates system
Amount), the coordinate of target to be monitored in the camera coordinate system is converted into coordinate of the target to be monitored in map coordinates system.
Method two：
In target detection in the method one, height h of the people in camera picture can also be provided；According to similar
Triangle Principle sets the average height of target to be monitored as H, then can calculate target to be monitored to monocular cam optical center
Distance d=f × H/h of point；It executes with the step 3 and step 4 described in method one, obtains target to be monitored in map coordinates system
Coordinate.
Further, the matching label and corresponding target to be monitored detailed process：
Under the same coordinate system, the target range change curve and the tag distances change curve are adopted
Sample counts the range information of a period of time, obtains corresponding vectorial (di0, di1 ... din) and (Dj0, Dj1 ... Djn)；In calculating
State the cosine degree of correlation between two vectors；If the cosine degree of correlation is more than threshold value, then it is assumed that target to be monitored i and j, label
Match, the two is same target；Wherein i, j are greater than the positive integer equal to 1；N is natural number；Din expressions are inscribed i-th in tn
The distance between target to be monitored and camera optical center point；Djn indicate to inscribe in tn between j-th of label and single base station away from
From.
Further, the matching label is with corresponding target to be monitored detailed process：
The distance between label and target to be monitored information are calculated, when distance is less than or equal to threshold value, then judges the label
Match for the target to be monitored, i.e., the label is entrained by the target to be monitored；The wherein described threshold value is actual test threshold value.
Tag match device based on the tag match method includes：
Monocular cam, the target data for acquiring target to be monitored；And target data is uploaded to server；
Single base station：For being communicated wirelessly between label, the range information of label and single base station is obtained, also by data
It is uploaded to server；
Several labels with unique mark, for being communicated wirelessly between single base station；
Server：According to the target data of target to be monitored, obtain between target to be monitored and monocular cam at any time
Coordinate of the target range change curve and target to be monitored of variation in map coordinates system；Meanwhile according to single base station and label
Range information obtains the tag distances change curve changed over time between label and base station；Changed according to the target range
Curve and the tag distances change curve, match label and corresponding target to be monitored；
Wherein monocular cam is mounted on same position with single base station.
Tag match device according to the tag match method includes：
Monocular cam, the target data for acquiring target to be monitored；And target data is uploaded to server；
Single base station：For being communicated wirelessly between label, position coordinates of the label in map coordinates system are obtained, also
Data are uploaded to server；
Several labels with unique mark, for being communicated wirelessly between single base station；
Server：According to the target data of target to be monitored, coordinate of the acquisition target to be monitored in map coordinates system；Together
When, according to the position coordinates of label, match label and corresponding target to be monitored；
Wherein monocular cam is mounted on same position with single base station.
Fence border control device based on the tag match device further includes；
Label and corresponding target to be monitored of the server based on successful match, obtain identity and the entrance of target to be monitored
The permission of fence and the coordinate for matching label calculate the phase of label or target to be monitored and fence boundary or region
To position, judge whether the target to be monitored for carrying label enters fence boundary or region, is done by alarm unit correspondence
Go out response.
Fence border control device based on the tag match device further includes：
When server can not carry out tag match to target to be monitored, i.e., it can not get the identity of target to be monitored and right
The permission into fence boundary answered, target to be monitored once entering fence, make corresponding immediately by alarm unit
Alarm.
Fence border control device based on the tag match device further includes:
When coordinate of the multiple labels that server obtains in map is identical, that is, there are two or more labels
It is fitted on the same target to be monitored, server judges that the monitoring objective carries more cards, made immediately by alarm unit corresponding
Alarm.
Further, based on alarm described in the fence border control device be following two methods it
One：
Reminding method one：Alarm unit is connect with server, and server directly issues control instruction extremely according to judging result
Alarm unit, alarm unit make corresponding alarm according to the instruction；
Reminding method two：Alarm unit is connect with single base station, and server issues control instruction to single base according to judging result
It stands, corresponding alarm is made by single base station triggers alarm unit.
A kind of computer storage media, the computer storage media are stored with computer program, the computer program
It is realized such as the step of above-mentioned any one the method when being executed by server.
In conclusion by adopting the above-described technical solution, the beneficial effects of the invention are as follows：
The present invention is integrated so simple only with a base station and a monocular cam or both
Device acquires the data of label and target to be monitored, is supervised with the i.e. matching of merging of wireless distance finding or positioning by vision positioning
Control target includes identity information to target to be monitored information in realization with label and enters the standard of the authority information of fence
It also achieves label is accurately positioned while really acquisition, and then realize for large scale industry, building yard, mining region,
The monitoring and management of the illegal invasion of the specific regions such as prison.In addition, the present invention can also realize the personnel's to not carrying label
Intrusion detection and the case where carrying more card to same target, are detected.Entire technical solution can be realized with less resource
Better management effect, system deployment is simple, facilitates easy to operate, is also greatly reduced cost needed for system.
Description of the drawings
Examples of the present invention will be described by way of reference to the accompanying drawings, wherein：
Fig. 1 is a kind of structure chart of the present invention.
Fig. 2 is another structure chart of the present invention.
Fig. 3 label ranging flow diagrams.
Specific implementation mode
All features disclosed in this specification or disclosed all methods or in the process the step of, in addition to mutually exclusive
Feature and/or step other than, can combine in any way.
Any feature disclosed in this specification unless specifically stated can be equivalent or with similar purpose by other
Alternative features are replaced.That is, unless specifically stated, each feature is an example in a series of equivalent or similar characteristics
.
Structure：One embodiment of the present invention structure as shown in Figure 1, including a base station, a monocular cam,
Several labels with unique mark, server and alarm unit, the unique mark include tag ID and enter fence
Authority information, base station are mounted on the same position either integrated structure, alarm unit and server with monocular cam
Connection.Base station and several labels interact formula wireless distance finding or positioning, and resolve between base station and several labels away from
From or label coordinate, monocular cam is used to acquire the target data (image) of monitoring area, and base station uploads range information
, can be by wired or wirelessly transmit to server, the target data of acquisition is uploaded to server by camera, clothes
Business device is handled in real time according to the target data received, extracts the distance between target to be monitored and camera optical center point, and
And calculate position coordinates of the target to be monitored in map coordinates system；Server also matches label and target to be monitored, for
The target to be monitored of label on successful match, target to be monitored then have unique ID number and enter fence boundary or region
Authority information, and the ID and authority information are corresponding with the unique mark entrained by matching label.Server calculates monitored
The relative position relation of target and fence boundary or region according to the relative position relation and combines target to be monitored
ID and enter fence permission, judge whether target to be monitored enters and fence boundary or region and whether have the right to enter,
Server triggers alarm unit according to judging result and makes corresponding alarm response；When target to be monitored is without matching label, clothes
Business device equally makes corresponding alarm response with the relative position relation in fence boundary or region according to target to be monitored, i.e.,
When target to be monitored, which does not carry label, is put into fence boundary or region, server triggers alarm unit alarm,
Otherwise, alarm unit is not responding to alarm；When there is multiple labels to be matched on the same target to be monitored, server judgement detects
The phenomenon that a people with more cards, alarm unit alarm is triggered immediately.
Another embodiment of the invention structure as shown in Figure 2, by alarm unit in structure shown in FIG. 1 and service
Device connection replaces with alarm unit and is connect with base station, and server will determine that result is exported to single base station, is controlled and is alarmed by single base station
Whether unit triggers alarm.Other content and connection type and the triggering that alarm unit and server are removed in the first embodiment
The content described except type of alarm is identical.
Tag match method according to the present invention, fence border control method are as described below：
One, two kinds of matching process：
(1), the first tag match method includes：
The target data of target to be monitored based on monocular cam acquisition, obtain target to be monitored and monocular cam it
Between the target range change curve that changes over time；
Meanwhile according to the range information of single base station and label, obtain the label that is changed over time between label and base station away from
From change curve；
Then, according to the target range change curve and the tag distances change curve, matching label with it is corresponding
Target to be monitored.
(2), second of tag match method includes：
The target data of target to be monitored based on monocular cam acquisition obtains target to be monitored in map coordinates system
Coordinate；
Meanwhile coordinate of the label in map is obtained to tag location based on single base station；
Then, the coordinate of coordinate and label in map according to the target to be monitored in map matches label
With corresponding target to be monitored.
Based on the first tag match method, the label of target to be monitored on successful match can be in conjunction with described monitored
Coordinate of the target in map coordinates system obtains coordinate of the label in map；Unique mark possessed by combination tag obtains
It carries the identity of the target to be monitored of the label and enters the permission of fence.
Based on second of tag match method, the target to be monitored of label, can obtain and take monitored mesh on successful match
Target identity and the authority information for entering fence boundary or region.
Two, two kinds of fence border control methods：
The first fence border control method：On the basis of two kinds of tag match methods, further include：
According to coordinate of the label in map, the relative position of label and fence boundary or region is calculated；
Server obtains the identity information of corresponding target to be monitored by label and the target enters fence boundary
Permission；Position relationship of the combination tag relative to fence boundary or region judges that the target to be monitored for carrying label is
It is no to enter fence boundary or region, respond that (target to be monitored of lack of competence enters electronics by alarm unit correspondence
Fence, server controls alarm unit carry out alarm, otherwise do not alarm).
For example, when target to be monitored moves closer to fence boundary outside fence boundary and has trend to want to cross over fence
When boundary, server can judge whether to need to do according to the authority information into fence boundary of the monitoring objective got
Go out response.If monitoring, the target to be monitored into fence does not have the permission entered, and monitoring server is to there is abnormal phenomenon
Occur, server controls alarm unit carries out alarm, and staff scene is prompted to observe conditions while prompting to be not belonging to this
The staff in region leaves the fence area in time, to avoid unnecessary safety accident.Such as prompt " note that inspection
Measure the target to be monitored entrance for not allowing access into the region ".It is on the contrary then do not alarm.
When server get coordinate of multiple labels in map it is identical when, that is, there are two or more labels
It is matched to the same target to be monitored, then judges that the monitoring objective carries more cards, control alarm unit makes corresponding report immediately
Alert prompt.
Second of fence border control method：On the basis of two kinds of tag match methods, further include：
According to coordinate of the target to be monitored in map, the opposite of target to be monitored and fence boundary or region is calculated
Position；Identity in conjunction with the target to be monitored got and the permission into fence judge to carry the monitored of label
Whether target enters fence boundary or region, and responds the (target to be monitored of lack of competence by alarm unit correspondence
Into fence, alarm carries out alarm and does not otherwise alarm).
Assuming that the target to be monitored of monitoring area has the case where not carrying label and entering fence, such case can
Can be originally belong to this working region personnel forget wear it could also be possible that the target to be monitored belongs to external inoperative people
Member, regardless of situation, as long as not carrying label, then server can only get the data under vision positioning, can not
With label, in this case, server triggers alarm unit prompt staff checks while prompting not carry label in scene
Target leaves the region, such as prompt " note that and detect that the personnel for not carrying label enter the region " in time.
It is that related content in above-mentioned one to two is further explained below：
1, it obtains the aim curve changed over time between target to be monitored and monocular cam and passes through following two methods
One of：
11) method 1：
The target data of target to be monitored based on monocular cam acquisition obtains seat of the target to be monitored on map
Mark；According to the coordinate of monocular cam, by the coordinate of the target to be monitored changed over time be converted to target to be monitored with
The target range change curve changed over time between monocular cam；
Specifically, obtaining coordinate of the target to be monitored on map may be used the following two kinds vision positioning method：
111) vision positioning method one：
Step 11：The picture of monocular cam transmission is handled to obtain position of the target in picture, and is supervised
Target is controlled relative to the angle of monocular cam (with projection point coordinates of the monocular cam optical center point on monocular cam picture
For (0,0), it is assumed that plane coordinates of the target to be monitored in monocular cam picture is (x, y), and monocular cam focal length is f,
Then the direction of vectorial (x, y, f) can determine angle of the target to be monitored relative to monocular cam optical center point)；Especially by
Algorithm of target detection such as HOG pedestrian detections algorithm, Faster RCNN etc. obtain position of the target to be monitored in picture.
Wherein, HOG pedestrian detections algorithm is in Dalal N, meters of the Triggs B. in IEEE Computer Association tissues in 2005
What calculation machine vision and pattern-recognition meeting were delivered《Histograms of oriented gradients for human
Pedestrian detections of the detection based on histograms of oriented gradients》It is described in meeting paper.
Faster RCNN algorithm of target detection was believed in 2015 about nerve in Ren S, He K, Girshick R et al.
It is delivered in the meeting of the progress of breath processing system《Faster r-cnn:Towards real-time object
Detection with region proposal networks propose the real-time target inspection of convolutional neural networks based on region
It surveys》There is discussion in paper.
Step 12：The picture that monocular cam takes is passed through into depth with coordinate (x, y) of the target to be monitored in picture
Spend method of estimation, obtain target to be monitored (such as：Staff or illegal invasion personnel) and monocular cam optical center point (0,
The distance between 0) d.
Wherein, depth estimation method is《The non-limiting rings of Single-image depth perception in the wild
Single photograph frame estimation of Depth under border》Method described in paper, the Authors of Science Articles are Weifeng Chen, Zhao Fu, Dawei
Yang, Jia Deng are published in Advances in neural information processing systems (NIPS).
Step 13：It calculates along the direction (x, y, f), the distance with optical center point (0,0,0) is the target to be monitored of d in camera
Coordinate (Xw, Yw, Zw) in coordinate system；
Because the distance between 2 points (x1, y1, z1) and (x2, y2, z2) mathematically D calculation formula are：
D2=(x1-x2)2+(y1-y2)2+(z1-z2)2
Therefore, the distance between (x, y, f) and (0,0,0) d0 is calculated according to above-mentioned principle (to represent in camera picture
Point to optical center point distance),
The distance between the controlled object and camera optical center point that are then extracted according to estimation of Depth d, obtains the mesh that is monitored
The coordinate (Xw, Yw, Zw) of mark in the camera coordinate system is (x × d/d0, y × d/d0, f × d/d0).
Step 14:Further according to transformational relation (the spin matrix peace between monocular cam coordinate system and map coordinates system
The amount of shifting to)：Wherein R is 3*3 matrixes；T is D translation vector；O=(0,0,0).
Calculate coordinate of the target to be monitored in map coordinates system.
112) vision positioning method 2：
Step 110：The picture of monocular cam transmission is handled to obtain position of the target to be monitored in picture, and
Height h of the people in camera picture is provided, determines that target to be monitored (is taken the photograph relative to the angle of monocular cam with monocular
As projection point coordinates of the head optical center point on monocular cam picture is (0,0), it is assumed that target to be monitored is drawn in monocular cam
Plane coordinates in face is (x, y), and monocular cam focal length is f, then the direction of vectorial (x, y, f) can determine target to be monitored
Angle relative to monocular cam optical center point)；Especially by targets to be monitored such as HOG pedestrian detections algorithm, FasterRCNN
Detection algorithm etc. obtains position of the target to be monitored in picture.
Step 120：According to similar triangle theory, the average height of target to be monitored is set as H, then is calculated monitored
Distance d=f × H/h of the target to monocular cam optical center point；
Step 130：It calculates along the direction (x, y, f), the coordinate for the point that the distance with optical center point (0,0,0) is d.With step 13
Described acquisition methods.
Step 140：:According to transformational relation (the spin matrix peace between monocular cam coordinate system and map coordinates system
The amount of shifting to), calculate coordinate of the target to be monitored in map coordinates system.It is identical that method is described with step 14.
12) method 2：
The target data of target to be monitored based on monocular cam acquisition, obtains the depth information of target to be monitored, root
The target range variation changed over time between target to be monitored and monocular cam is fitted according to the depth information of different moments
Curve；
It is described obtain depth information process with step 11 in method 1 and 12 or step 110 and 120 description it is identical.
2, the range information acquisition modes of single base station and label are one of following three kinds：21) on the label
Into resolving of the row label at a distance from base station and obtain the range information.
Detailed process is：Label is resolved according to related ranging information, into row label at a distance from base station.
Such as ranging flow according to Fig.3,：
Label card timing sends a distance measurement request to single base station, by taking a certain label card as an example, it is assumed that when label card is sent
Between stamp be TspSingle base station receives the distance measurement request that label card is sent, and receiving time stamp is Trp, in the ranging for receiving label card
After request, a response signal can be sent under a random delay, if the timestamp that base station sends response signal is Tsr, sending out
In the response signal sent, T can be takenrpAnd Tsr, the response signal of the single base station of label card reception, receiving time stamp is Trr。
Using formulaCalculate the distance between label card and single base station.
22) in single base station, enterprising row label resolves at a distance from base station and obtains the range information.Detailed process is：Label
Related ranging information is back to single base station by base station UWB channels, single base station is resolved into row label at a distance from base station.
23) except label with base station the enterprising row label of third party device and base station at a distance from resolve and obtain the distance
Information.
Such as third party refers to server：Specifically solution process is：By the correlative measurement of label or respective base station away from letter
Breath is uploaded to server, is resolved at a distance from base station into row label.
It wherein uploads mode and is not limited in following manner：
231) the related ranging information of label or respective base station is back to base station by base station UWB channels, base station will
The information uploads；
232) by the related ranging information of label either respective base station by the way that Ethernet or WIFI be wired in this way or nothing
The transmission mode of line uploads.
3, the process that single base station obtains tag location on coordinate of the label in map is：
Single base station obtains the angle of label and single base station based on phase discriminator, at this point, the knot of single base station described in Fig. 1 and Fig. 2
Include phase discriminator in structure, and phase discriminator is connected with two radio-frequency antennas, two antennas are respectively the first radio-frequency antenna and
Two radio-frequency antennas, and the distance between the first radio-frequency antenna and the second radio-frequency antenna can be measured as known quantity；Single base station with
Label, which carries out wireless interactive ranging, can obtain the distance between single base station and label, and base station is based on angle information and apart from letter
Breath just can obtain the coordinate of label.Specific localization method is as described below：
According to formulaCalculate angle of the label relative to single base station；
Wherein α is to measure the signal that the label obtained is sent by the phase discriminator to reach first radio-frequency antenna
With the phase difference of the second radio-frequency antenna；
D is the distance between described two radio-frequency antennas；
λ is the wavelength of the ultra-broadband signal；
By wireless distance finding module based on above-mentioned range formulaIt calculates
The distance between label and single base station D；
According to the distance between label and single base station angle of D and the label relative to single base station, the mark is determined
The coordinate of label.
4, it can also be fence area that fence, which can be fence boundary, if fence boundary, then only need monitoring
A fence boundary line is drawn on the electronic map of display end.If fence area, then only need to draw one on the electronic map
A fence closed area.The unlimited form in the closed area.All fence areas total management system in the server
Electronic map display interface is shown.
5, requiring to enter specifies each target of fence boundary or region to need to carry a label, and every label card is all recorded
Enter corresponding target and enters the permission on fence boundary or region and corresponding ID number.
6, there are two types of the connection types of alarm unit：Alarm unit includes loud speaker and alarm.
61) alarm unit is connect with server, when server monitoring to target to be monitored without enter fence permission, or
When person does not carry label or carries multiple labels, then control instruction is sent to alarm unit, alarm unit is according to the instruction
Make corresponding alarm；
62) alarm unit is connect with single base station, when server server monitors target to be monitored without into fence
Permission then sends instructions to alarm unit either without carrying label or when carrying multiple labels by single base station, and alarm is single
Member makes corresponding alarm according to the instruction.
7, related description of the present invention：Single base station refers to communicating wirelessly with label, obtain label and the list base station away from
From or to tag location base station；The list base station can carry out data interaction with server simultaneously.
The invention is not limited in specific implementation modes above-mentioned.The present invention, which expands to, any in the present specification to be disclosed
New feature or any new combination, and disclose any new method or process the step of or any new combination.
Claims (17)
1. a kind of tag match method, it is characterised in that including：
The target data of target to be monitored based on monocular cam acquisition, obtain between target to be monitored and monocular cam with
Coordinate of the target range change curve and target to be monitored of time change in map coordinates system；
Meanwhile according to the range information of single base station and label, obtains the tag distances changed over time between label and base station and become
Change curve；
Then, according to the target range change curve and the tag distances change curve, matching label is supervised with corresponding
Control target；
The wherein described monocular cam is mounted on same position with single base station.
2. based on tag match method described in claim 1, it is characterised in that further include working as label and corresponding target to be monitored
When successful match, in conjunction with coordinate of the target to be monitored in map coordinates system, coordinate of the label in map is obtained.
3. a kind of tag match method, it is characterised in that including：
The target data of target to be monitored based on monocular cam acquisition obtains seat of the target to be monitored in map coordinates system
Mark；
Meanwhile coordinate of the label in map is obtained to tag location based on single base station；
Then, the coordinate of coordinate and label in map according to the target to be monitored in map matches label and phase
The target to be monitored answered；
The wherein described monocular cam is mounted on same position with single base station.
4. being based on the tag match method of claim 1 or 3, it is characterised in that further include when label and corresponding monitored mesh
When marking successful match, the unique mark in conjunction with possessed by the label including tag ID and the authority information for entering fence obtains
It takes the identity of target to be monitored and enters the permission of fence.
5. the fence border control method based on tag match method according to claim 2 or 3, it is characterised in that also wrap
It includes：
According to coordinate of the label in map, the relative position of label and fence boundary or region is calculated；
Unique mark possessed by combination tag, including tag ID and the authority information for entering fence, judge to carry the mark
Whether the target to be monitored of label enters fence boundary, makes corresponding alarm response；
When coordinate of the multiple labels of acquisition in map is identical, that is, there are two or more tag match to same
A target to be monitored then judges that the monitoring objective carries more cards, makes corresponding alarm immediately.
6. the fence border control method based on the tag match method described in claim 4, it is characterised in that further include：
According to coordinate of the target to be monitored in map, target to be monitored and fence boundary or the opposite position in region are calculated
It sets；
Identity in conjunction with the target to be monitored got and the permission into fence, judge whether target to be monitored enters
Fence boundary or region, and responded by alarm unit correspondence；
When target to be monitored is without matching label, i.e., it can not get the identity of target to be monitored and corresponding entrance fence side
The permission on boundary, target to be monitored make corresponding alarm immediately once entering fence.
7. tag match method according to claim 1 is it is characterized in that between the target to be monitored and monocular cam
The target range change curve changed over time is obtained by one of following two methods：
Method 1：
The target data of target to be monitored based on monocular cam acquisition obtains coordinate of the target to be monitored on map；Root
According to the coordinate of monocular cam, the coordinate of the target to be monitored changed over time is converted into target to be monitored and is taken the photograph with monocular
As the target range change curve changed over time between head；
Method 2：The target data of target to be monitored based on monocular cam acquisition, obtains the depth information of target to be monitored,
The target range changed over time between target to be monitored and monocular cam is fitted according to the depth information of different moments to become
Change curve.
8. tag match method according to claim 1,2 or 3, it is characterised in that the acquisition target to be monitored is in map
Coordinate on coordinate system is realized based on one of following two vision positioning methods：
Method one：
Step 1：Target detection：
The picture of monocular cam transmission is handled in real time and obtains position (x, y) of the target in picture, in conjunction with monocular cam
Focal length f, obtain vectorial (x, y, f), determine monitoring objective relative to monocular cam optical center according to the direction of vectorial (x, y, f)
The angle of point；
Step 2：Estimation of Depth：
The coordinate (x, y) of picture and target to be monitored that monocular cam is taken obtains being supervised by depth estimation method
Control the distance between target and monocular cam optical center point d；
Step 3：Target to be monitored is obtained in camera coordinate system in conjunction with the range information in (x, the y, f) and step 2 in step 1
In coordinate；
Step 4：According to the transformational relation (spin matrix and translation vector) between monocular cam coordinate system and map coordinates system,
The coordinate of target to be monitored in the camera coordinate system is converted into coordinate of the target to be monitored in map coordinates system.
Method two：
In target detection in the method one, height h of the people in camera picture can also be provided；According to similar triangle
Shape principle sets the average height of target to be monitored as H, then can calculate target to be monitored to monocular cam optical center point
Distance d=f × H/h；It executes with the step 3 and step 4 described in method one, obtains target to be monitored and sat in map coordinates system
Mark.
9. according to the method described in claim 1, it is characterized in that the matching label and the specific mistake of corresponding target to be monitored
Journey：
Under the same coordinate system, the target range change curve and the tag distances change curve are sampled,
The range information for counting a period of time obtains corresponding vectorial (di0, di1 ... din) and (Dj0, Dj1 ... Djn)；It calculates above-mentioned
The cosine degree of correlation between two vectors；If the cosine degree of correlation is more than threshold value, then it is assumed that target to be monitored i is matched with label j,
The two is same target；Wherein i, j are greater than the positive integer equal to 1；N is natural number；I-th of quilt is inscribed in din expressions in tn
The distance between monitoring objective and camera optical center point；Djn indicate to inscribe in tn between j-th of label and single base station away from
From.
10. tag match method according to claim 3, it is characterised in that matching label has with corresponding target to be monitored
Body process is：
The distance between label and target to be monitored information are calculated, when distance is less than or equal to threshold value, then judges the label to be somebody's turn to do
Target to be monitored matches, i.e., the label is entrained by the target to be monitored；The wherein described threshold value is actual test threshold value.
11. the tag match device based on tag match method described in claim 1, it is characterised in that including：
Monocular cam, the target data for acquiring target to be monitored；And target data is uploaded to server；
Single base station：For being communicated wirelessly between label, the range information of label and single base station is obtained, also uploads data
To server；
Several labels with unique mark, for being communicated wirelessly between single base station；
Server：According to the target data of target to be monitored, obtains and changed over time between target to be monitored and monocular cam
Target range change curve and target to be monitored map coordinates system coordinate；Meanwhile according to single base station at a distance from label
Information obtains the tag distances change curve changed over time between label and base station；According to the target range change curve
With the tag distances change curve, label and corresponding target to be monitored are matched；
Wherein monocular cam is mounted on same position with single base station.
12. the tag match device of tag match method according to claim 3, it is characterised in that including：
Monocular cam, the target data for acquiring target to be monitored；And target data is uploaded to server；
Single base station：For being communicated wirelessly between label, position coordinates of the label in map coordinates system are obtained, will also be counted
According to being uploaded to server；
Several labels with unique mark, for being communicated wirelessly between single base station；
Server：According to the target data of target to be monitored, coordinate of the acquisition target to be monitored in map coordinates system；Meanwhile root
According to the position coordinates of label, label and corresponding target to be monitored are matched；
Wherein monocular cam is mounted on same position with single base station.
13. the fence border control device based on tag match device described in claim 11, it is characterised in that further include；
Label and corresponding target to be monitored of the server based on successful match obtain the identity of target to be monitored and enter electronics
The permission of fence and the coordinate for matching label calculate label or target to be monitored and fence boundary or the opposite position in region
It sets, judges whether the target to be monitored for carrying label enters fence boundary or region, sound is made by alarm unit correspondence
It answers.
14. the fence border control device based on tag match device described in claim 11, it is characterised in that further include：
When server can not carry out tag match to target to be monitored, i.e., it can not get the identity of target to be monitored and corresponding
Into the permission on fence boundary, for target to be monitored once entering fence, alarm unit makes corresponding alarm immediately
Prompt.
15. based on the fence border control device of one of claim 11 to 14 tag match device, feature exists
In further including:
When coordinate of the multiple labels that server obtains in map is identical, that is, there are two or more tag match and arrive
The same target to be monitored, server judge that the monitoring objective carries more cards, make corresponding alarm immediately by alarm unit
Prompt.
16. based on the fence border control device described in claim 15, it is characterised in that the alarm is following
One of two methods：
Reminding method one：Alarm unit is connect with server, and server directly issues control instruction according to judging result to alarm
Unit, alarm unit make corresponding alarm according to the instruction；
Reminding method two：Alarm unit is connect with single base station, and server issues control instruction to single base station according to judging result, by
Single base station triggers alarm unit makes corresponding alarm.
17. a kind of computer storage media, it is characterised in that the computer storage media is stored with computer program, the meter
It is realized such as the step of claims 1 to 10 any one the method when calculation machine program is executed by server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810373250.8A CN108389351A (en) | 2018-04-24 | 2018-04-24 | Tag match, fence border control method, apparatus and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810373250.8A CN108389351A (en) | 2018-04-24 | 2018-04-24 | Tag match, fence border control method, apparatus and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108389351A true CN108389351A (en) | 2018-08-10 |
Family
ID=63064584
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810373250.8A Pending CN108389351A (en) | 2018-04-24 | 2018-04-24 | Tag match, fence border control method, apparatus and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108389351A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109246691A (en) * | 2018-11-19 | 2019-01-18 | 南京沃旭通讯科技有限公司 | A kind of intrusion detection method based on channel impulse response |
| CN109493550A (en) * | 2018-11-08 | 2019-03-19 | 北京中盈安信技术服务股份有限公司 | Site safety management method, system and electronic equipment based on fence |
| CN109597027A (en) * | 2018-12-06 | 2019-04-09 | 清华大学 | A kind of positioning system and method based on single base station |
| CN109618287A (en) * | 2018-12-29 | 2019-04-12 | 杭州平普智能科技有限公司 | A kind of convict's management method and device based on UWB positioning |
| CN109996182A (en) * | 2019-04-19 | 2019-07-09 | 无锡艾森汇智科技有限公司 | A kind of localization method, apparatus and system combined based on UWB positioning with monitoring |
| CN110557569A (en) * | 2019-09-09 | 2019-12-10 | 郑州联睿电子科技有限公司 | Intelligent monitoring system and method integrating ultra-wideband AoA positioning and video tracking |
| CN110825037A (en) * | 2018-08-14 | 2020-02-21 | 臧博琦 | Intelligent active management and control system for operation site |
| CN111564016A (en) * | 2020-06-12 | 2020-08-21 | 夸克云智科技(深圳)有限公司 | Regional intrusion early warning method based on computer vision |
| CN112098985A (en) * | 2020-09-09 | 2020-12-18 | 杭州中芯微电子有限公司 | UWB positioning method based on millimeter wave detection |
| CN112802302A (en) * | 2020-12-31 | 2021-05-14 | 国网浙江省电力有限公司双创中心 | Electronic fence method and system based on multi-source algorithm |
| CN114049710A (en) * | 2021-10-08 | 2022-02-15 | 江苏禹治流域管理技术研究院有限公司 | Intelligent supervision system for sand production in river channel |
| CN114120620A (en) * | 2021-11-18 | 2022-03-01 | 海南大学 | DIKW-based regional perception and traffic prompting method |
| CN115150802A (en) * | 2022-09-05 | 2022-10-04 | 广东新邦智联科技有限公司 | User monitoring method, device and system based on Bluetooth detection and storage medium |
| WO2023273149A1 (en) * | 2021-06-29 | 2023-01-05 | 北京市商汤科技开发有限公司 | Method and apparatus for managing electronic fence, method and apparatus for configuring electronic fence, and device, medium and product |
| CN116052223A (en) * | 2023-04-03 | 2023-05-02 | 浪潮通用软件有限公司 | Method, system, equipment and medium for identifying people in operation area based on machine vision |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102667518A (en) * | 2009-12-08 | 2012-09-12 | 真实定位公司 | Multi-sensor location and identification |
| US20130117266A1 (en) * | 2011-11-09 | 2013-05-09 | Microsoft Corporation | Geo-fence based on geo-tagged media |
-
2018
- 2018-04-24 CN CN201810373250.8A patent/CN108389351A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102667518A (en) * | 2009-12-08 | 2012-09-12 | 真实定位公司 | Multi-sensor location and identification |
| US20130117266A1 (en) * | 2011-11-09 | 2013-05-09 | Microsoft Corporation | Geo-fence based on geo-tagged media |
Non-Patent Citations (1)
| Title |
|---|
| 沈慧杰: ""基于单目视觉的摄像机定位方法的研究"" * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110825037A (en) * | 2018-08-14 | 2020-02-21 | 臧博琦 | Intelligent active management and control system for operation site |
| CN109493550A (en) * | 2018-11-08 | 2019-03-19 | 北京中盈安信技术服务股份有限公司 | Site safety management method, system and electronic equipment based on fence |
| CN109246691B (en) * | 2018-11-19 | 2021-10-22 | 南京沃旭通讯科技有限公司 | Intrusion detection method based on channel impulse response |
| CN109246691A (en) * | 2018-11-19 | 2019-01-18 | 南京沃旭通讯科技有限公司 | A kind of intrusion detection method based on channel impulse response |
| CN109597027A (en) * | 2018-12-06 | 2019-04-09 | 清华大学 | A kind of positioning system and method based on single base station |
| CN109618287A (en) * | 2018-12-29 | 2019-04-12 | 杭州平普智能科技有限公司 | A kind of convict's management method and device based on UWB positioning |
| CN109996182A (en) * | 2019-04-19 | 2019-07-09 | 无锡艾森汇智科技有限公司 | A kind of localization method, apparatus and system combined based on UWB positioning with monitoring |
| CN110557569A (en) * | 2019-09-09 | 2019-12-10 | 郑州联睿电子科技有限公司 | Intelligent monitoring system and method integrating ultra-wideband AoA positioning and video tracking |
| CN111564016A (en) * | 2020-06-12 | 2020-08-21 | 夸克云智科技(深圳)有限公司 | Regional intrusion early warning method based on computer vision |
| CN112098985A (en) * | 2020-09-09 | 2020-12-18 | 杭州中芯微电子有限公司 | UWB positioning method based on millimeter wave detection |
| CN112098985B (en) * | 2020-09-09 | 2024-04-12 | 杭州中芯微电子有限公司 | UWB positioning method based on millimeter wave detection |
| CN112802302A (en) * | 2020-12-31 | 2021-05-14 | 国网浙江省电力有限公司双创中心 | Electronic fence method and system based on multi-source algorithm |
| CN112802302B (en) * | 2020-12-31 | 2022-11-18 | 国网浙江省电力有限公司双创中心 | Electronic fence method and system based on multi-source algorithm |
| WO2023273149A1 (en) * | 2021-06-29 | 2023-01-05 | 北京市商汤科技开发有限公司 | Method and apparatus for managing electronic fence, method and apparatus for configuring electronic fence, and device, medium and product |
| CN114049710A (en) * | 2021-10-08 | 2022-02-15 | 江苏禹治流域管理技术研究院有限公司 | Intelligent supervision system for sand production in river channel |
| CN114049710B (en) * | 2021-10-08 | 2023-10-20 | 江苏禹治流域管理技术研究院有限公司 | River course adopts sand intelligent supervisory systems |
| CN114120620A (en) * | 2021-11-18 | 2022-03-01 | 海南大学 | DIKW-based regional perception and traffic prompting method |
| CN115150802A (en) * | 2022-09-05 | 2022-10-04 | 广东新邦智联科技有限公司 | User monitoring method, device and system based on Bluetooth detection and storage medium |
| CN115150802B (en) * | 2022-09-05 | 2022-12-13 | 广东新邦智联科技有限公司 | User monitoring method, device and system based on Bluetooth detection and storage medium |
| CN116052223A (en) * | 2023-04-03 | 2023-05-02 | 浪潮通用软件有限公司 | Method, system, equipment and medium for identifying people in operation area based on machine vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108389351A (en) | Tag match, fence border control method, apparatus and storage medium | |
| CN108550234A (en) | Tag match, fence boundary management method, device and the storage medium of Dual base stations | |
| CN208689702U (en) | A kind of monitoring management system based on fence | |
| CN208675549U (en) | A kind of fence management system | |
| KR101797231B1 (en) | BLE beacon-based system and method for managing entrance and exit | |
| CN109076191B (en) | Monitoring system, method, non-transitory computer-readable medium, and control unit | |
| US7535353B2 (en) | Surveillance system and surveillance method | |
| CN101496074A (en) | Device and method for detecting suspicious activity, program, and recording medium | |
| US11398121B2 (en) | Method for provisioning a device with an information element allowing to identify unauthorized users in a restricted area | |
| JP2006311111A (en) | Position sensing system and method for sensing position | |
| CN111083642A (en) | Method for positioning personnel in treasury | |
| CN108650485A (en) | Personnel monitoring's method and system based on image detection and wireless probe | |
| CN104777522B (en) | Foreign matter detection system and detection method between a kind of Train door and platform screen door based on compound detection | |
| KR101823147B1 (en) | BLE Beacon based safety assitance service system | |
| US7295106B1 (en) | Systems and methods for classifying objects within a monitored zone using multiple surveillance devices | |
| CN107066945B (en) | A kind of quick identity checking method of big flow clearance and system | |
| KR100815920B1 (en) | Security system using location data of gps and image data of ccd camera and invader detection method using the same | |
| KR101305371B1 (en) | Real time location monitoring system and method for transmitting of image data | |
| KR101946317B1 (en) | User Identification and Tracking Method Using CCTV System | |
| CN208737505U (en) | Personnel safety monitoring and Time Attendance Device based on wireless sensor network | |
| JPH09130781A (en) | Broad area supervisory equipment | |
| CN209312052U (en) | A kind of carry-on police system of recognition of face | |
| KR101829274B1 (en) | System and apparatus for image security using beacon | |
| JP2004030022A (en) | Invasion detection system | |
| KR20170127960A (en) | Enforcement system of illegal parking |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190418 Address after: Room 904, 905, 906, 1005, 1006 and 1007, No. 898 Baicao Road, Chengdu High-tech Zone, Sichuan Province, 610000 Applicant after: Chengdu Sixiangzhi New Technology Co., Ltd. Address before: 610000 No. 88 Tianchen Road, Chengdu High-tech Zone, Sichuan Province Applicant before: CHENGDU EVERHIGH TECHNOLOGY CO., LTD. |
|
| TA01 | Transfer of patent application right |