CN108848348A - A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane - Google Patents
A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane Download PDFInfo
- Publication number
- CN108848348A CN108848348A CN201810762340.6A CN201810762340A CN108848348A CN 108848348 A CN108848348 A CN 108848348A CN 201810762340 A CN201810762340 A CN 201810762340A CN 108848348 A CN108848348 A CN 108848348A
- Authority
- CN
- China
- Prior art keywords
- crowd
- unmanned plane
- abnormal behaviour
- wireless communication
- infrared
- 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
-
- 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
- H04N7/183—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a single remote source
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Alarm Systems (AREA)
Abstract
A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention, including unmanned plane body, flight control system, infrared vision system, earth station system and hand-held remote controller；Infrared vision system and flight control system are built in unmanned plane body center, hand-held remote controller wirelessly communicates flight control system, earth station system wirelessly communicates infrared vision system, infrared vision system communications flight control system solves the problems, such as that the manpower consumption of previous crowd's abnormal behaviour monitoring is high.
Description
Technical field
The present invention relates to crowd's abnormal behaviours of unmanned plane to monitor field, and especially a kind of crowd based on unmanned plane is abnormal
Behavior monitoring device and method.
Background technique
With the raising that people realize public safety, crowd's unusual checking is received more and more attention, so that
The research of crowd's abnormal behaviour becomes an academic hot spot in computer vision field.The monitoring of China public domain at present
System is that crowd's unusual checking based on visible light, but based on visible light is affected by environment larger, and faces mostly
When property large-scale assembly place proposes higher challenge to fixed video monitoring system.
Currently, the monitoring of crowd's abnormal behaviour has become a vital research direction in field of video monitoring, but
It is only to increase human cost by complete artificial monitoring crowd behaviour, the features such as unmanned plane is with its light weight, at low cost, mobility strong
It receives significant attention, thus by military and civilian field is largely put into, in military surveillance, target search, information search and safety
The application fields such as protection have important research significance.It is small for group's sexuality of burst when monitoring crowd's abnormal behaviour
The advantages of type unmanned plane is with its response quickly and mobility strong, the developing state of energy real-time tracking event, facilitates command centre
Implement uninterrupted commander's processing.After installing embedded image processor additional, unmanned plane can also carry out people to monitoring area in real time
Group's abnormal behaviour monitoring, and alarm abnormal behaviour, so that security personnel can timely and effectively take counter-measure.Cause
This, develop it is a kind of based on unmanned plane crowd's abnormal behaviour monitoring system seem very necessary.
Summary of the invention
To solve problems of the prior art, the present invention provides a kind of, and crowd's abnormal behaviour based on unmanned plane is supervised
Device and method is surveyed, solves the problems, such as that the manpower consumption of previous crowd's abnormal behaviour monitoring is high.
The technical solution adopted by the present invention is that a kind of crowd's abnormal behaviour monitoring device based on unmanned plane, including, nobody
Machine body, flight control system, infrared vision system, earth station system and hand-held remote controller；The infrared vision system and fly
Row control system is built in unmanned plane body center, and hand-held remote controller wirelessly communicates flight control system, earth station system without
Line communicates infrared vision system.
A kind of crowd's abnormal behaviour monitoring method based on unmanned plane, includes the following steps：
Step S1：The monitoring region of monitoring unmanned plane is set by earth station；
Step S2：By the monitoring for monitoring unmanned plane height be set as 20m, concurrently set thermal infrared imager direction with it is vertical
Angular separation α is 45 °；
Step S3：Spiral search is carried out according to monitoring height, the angle of setting thermal infrared imager and specified region, is led to
It crosses infrared vision system and carries out crowd density estimation；
Step S4：Judge whether crowd's number of infrared vision system detection is less than threshold value, if so, S5 is entered step,
Otherwise, S6 is entered step；
Step S5：Continue spiral search, return step S4；
Step S6：The fixed current visual angle of unmanned plane is without spiral search；
Step S7：Infrared video information is obtained by thermal infrared imager FLIR TAU2, and passes through embedded platform Jetson
TX1 estimates crowd density using multitask concatenated convolutional neural network；
Step S8：Judge whether variable density and average movement velocity variation are abnormal, if so, S9 is entered step,
Otherwise, system does not alarm and enters step S6.
Step S9：System alarm terminates program.
The present invention is based on crowd's abnormal behaviour monitoring device of unmanned plane and having the beneficial effect that for method：
The Full-automatic monitoring to crowd behaviour is realized, there is crowd behaviour real-time monitoring, the alarm of crowd's abnormal behaviour, number
According to functions such as storage, the inquiry of historical data, the monitoring of crowd's abnormal behaviour in time, accurately, improves security protection efficiency, reduces manpower
Cost has good practicability and scalability.
Detailed description of the invention
Fig. 1 is a kind of overall structure figure of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention.
Fig. 2 is a kind of infrared vision system knot of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention
Structure schematic diagram.
Fig. 3 is a kind of flight control system knot of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention
Structure schematic diagram.
Fig. 4 is a kind of crowd's abnormal behaviour prison of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention
Survey method key step schematic diagram.
Fig. 5 is that a kind of infrared crowd density of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention is estimated
Meter method key step schematic diagram.
Fig. 6 is that a kind of infrared crowd density of crowd's abnormal behaviour monitoring device and method based on unmanned plane of the present invention is estimated
Count convolutional neural networks structural schematic diagram.
Specific embodiment
A specific embodiment of the invention is described below, in order to facilitate understanding by those skilled in the art this hair
It is bright, it should be apparent that the present invention is not limited to the ranges of specific embodiment, for those skilled in the art,
As long as various change is in the spirit and scope of the present invention that the attached claims limit and determine, these variations are aobvious and easy
See, all are using the innovation and creation of present inventive concept in the column of protection.
As shown in Figure 1, a kind of crowd's abnormal behaviour monitoring device based on unmanned plane, including, unmanned plane body, flight control
System, infrared vision system, earth station system and hand-held remote controller processed；The infrared vision system and flight control system are built
In in unmanned plane body center, hand-held remote controller wirelessly communicates flight control system, and the earth station system wireless communication is infrared
Vision system, infrared vision system serial communication flight control system.
As shown in Fig. 2, unmanned plane body includes X-shaped rack, propeller and battery, the propeller is set to X-shaped rack
Four angles top, the battery is set to machine frame inside.
Infrared vision system include embedded platform Jetson TX1, thermal infrared imager FLIR TAU2, video frequency collection card and
Image transmission system；The thermal infrared imager FLIR TAU2 is separately connected video frequency collection card and image transmission system；The image transmission system communication
Earth station system, the video frequency collection card connect embedded platform Jetson TX1, and the embedded platform Jetson TX1 connects
Connect flight controller.
As shown in figure 3, flight control system includes microcontroller STM32F427, brushless motor, electron speed regulator, gyro
Instrument, electronic compass, accelerometer, barometer, GPS module and the first wireless communication module, microcontroller STM32F427 connect respectively
Connect brushless motor, electron speed regulator, gyroscope, electronic compass, accelerometer, barometer, GPS module and the first radio communication mold
Block, the main control chip of first wireless communication module are nRF24L01.
Hand-held remote controller includes microcontroller STM32F103VET6 and the second wireless communication module, the microcontroller
STM32F103VET6 the second wireless communication module of connection, the main control chip of second wireless communication module are nRF24L01, hand
Remote controler is held for manually controlling unmanned plane.
As shown in figure 4, a kind of crowd's abnormal behaviour monitoring method based on unmanned plane, includes the following steps：
Step S1：The monitoring region of monitoring unmanned plane is set by earth station,
Step S2：By the monitoring for monitoring unmanned plane height be set as 20m, concurrently set thermal infrared imager direction with it is vertical
Angular separation α is 45 °；
Step S3：Spiral search is carried out according to monitoring height, the angle of setting thermal infrared imager and specified region, is led to
It crosses infrared vision system and carries out crowd density estimation；
Step S4：Judge whether crowd's number of infrared vision system detection is less than threshold value, if so, S5 is entered step,
Otherwise, S6 is entered step；
Step S5：Continue spiral search, return step S4；
Step S6：The fixed current visual angle of unmanned plane is without spiral search；
Step S7：Infrared video information is obtained by thermal infrared imager FLIR TAU2, and passes through embedded platform Jetson
TX1 estimates crowd density using multitask concatenated convolutional neural network；
Step S8：Judge whether variable density and average movement velocity variation are abnormal, if so, S9 is entered step,
Otherwise, system does not alarm and enters step S6.
Step S9：System alarm terminates program.
As shown in fig. 6, step S6's uses multitask concatenated convolutional neural network by embedded platform Jetson TX1
Method estimates that crowd density, multitask concatenated convolutional neural network method include the following steps：
Step A1：Infrared image is inputted, image is inputted into shared network phase；
Step A2：According to the image of shared network phase, output image is inputted into the advanced priori knowledge stage；
Step A3：The 10 class crowds by the output in advanced priori knowledge stage are inputted into the crowd density estimation stage；
Step A4：After the crowd density estimation stage, crowd density is exported, terminates program.
The infrared vision system of step S2 carries out crowd density estimation method and includes the following steps：
As shown in figure 5, step B1：Design crowd density estimation CNN network model；
Step B2：According to CNN network model, training sample is demarcated, then the ideal density figure D of training samplei；
In formula, σ is the scale parameter of two-dimensional Gaussian kernel, and S is the set of all the points where the number of people；
Step B3：According to training sample, training network is established,
The cross entropy in training network advanced priori knowledge stage damages function：
In formula, N is number of training, and Θ is a set of network parameters, XiIt is i-th of training sample, Fc(Xi, Θ) and it is classification
Output, yiIt is true value classification, M is classification sum；
The loss function in density estimation stage is：
Wherein Fd(Xi, CiIt Θ) is estimation density map, DiIt is real density figure, CiIt is last from the advanced priori knowledge stage
The characteristic pattern that one convolutional layer obtains.
Entirely the loss function of cascade network is：
L=λ Lc+Ld
Wherein λ is weight factor, and λ takes 0.00001；
Step B4：It is tested using trained multi-cascade convolutional neural networks.
The velocity magnitude of angle point is calculated as：
Wherein, fps is video frame rate, and the Optic flow information of motion corner point is (vx, vy)T；
Crowd's average movement velocity size is calculated as in xth frame image：
Wherein, n indicates to detect the quantity of motion corner point, viIndicate the movement velocity size of i-th of angle point.
The present embodiment is when implementing, a kind of crowd's abnormal behaviour monitoring system based on unmanned plane, including unmanned plane machine
Body, flight control system, infrared vision system, earth station system and hand-held remote controller；Unmanned plane body includes rack, spiral
Paddle, battery；Infrared vision system and flight control system are equipped in unmanned plane body center, wherein thermal infrared imager FLIR
TAU2 is located at below unmanned plane body, and is furnished with the dedicated holder of unmanned plane；Earth station system obtains Jetson by image transmission system
The collected video information of TX1；
Hand-held remote controller includes microcontroller STM32F103VET6 and the second wireless communication module nRF24L01, is used for hand
Dynamic control unmanned plane, and pass through the status information of the second wireless communication module nRF24L01 acquisition crowd, including crowd density, people
Group mean movement velocity and crowd state.
Infrared vision system include embedded platform Jetson TX1, thermal infrared imager FLIR TAU2, video frequency collection card and
Image transmission system, thermal infrared imager FLIR TAU2 are used for infrared image acquisition, send video information to ground by image transmission system
It stands, embedded platform Jetson TX1 realizes crowd's unusual checking, and crowd's information is sent to flight control by serial ports
System processed.
Flight control system includes microcontroller STM32F427, brushless motor, electron speed regulator, gyroscope, electronics sieve
Disk, accelerometer, barometer, GPS module and the first wireless communication module nRF24L01, for control unmanned plane during flying posture and
Position, flight control system send crowd's information to hand-held remote controller, convenient for peace by the first wireless communication module nRF24L01
Anti- personnel timely and effectively take counter-measure.
A kind of embedded platform Jetson TX1 of crowd's abnormal behaviour monitoring system based on unmanned plane is red by extracting
Shi-Tomasi angle point and pyramid LK optical flow method in outer image track angle point, and the crowd calculated between video two continuous frames is average
Movement velocity.It is (v by the Optic flow information that the above method obtains motion corner point for two continuous frames imagex, vy)T, as a result, may be used
It is calculated as with calculating the velocity magnitude of angle point：
Wherein, fps is video frame rate.
Crowd's average movement velocity is defined as in X frame image：
Wherein, n indicates to detect the quantity of motion corner point, viIndicate the movement velocity size of i-th of angle point.
Crowd density estimation uses multitask concatenated convolutional neural fusion, and cascade neural network is divided into three phases,
Respectively shared network phase, advanced priori knowledge stage and crowd density estimation stage, the input of concatenated convolutional neural network are
Infrared image exports as crowd's density map and crowd's quantity.
Shared network in multitask concatenated convolutional neural network includes 4 layers of convolutional layer, there is one after every layer of convolutional layer
The convolution kernel size of a linear excitation function PReLU of the amendment with parameter, first layer convolution are 9*9, export 16 characteristic patterns, the
The convolution kernel size of two layers of convolutional layer is 7*7, exports 32 characteristic patterns.
The advanced priori knowledge stage in multitask concatenated convolutional neural network includes 4 layers of convolutional layer, 1 pond SPP layer
With 3 full articulamentums, each convolutional layer has the linear excitation function PReLU of the amendment with parameter, level 1 volume lamination later
Convolution kernel size be 9*9, the convolution kernel size of level 2 volume lamination is 7*7, comprising maximum pond layer after the first two convolutional layer,
Pond layer step-length is that the convolution kernel size of the 2, the 3rd layer of convolutional layer is 7*7, and the convolution kernel size of the 4th layer of convolutional layer is 7*7, convolution
Pass through the pond spatial pyramid SPP layer after layer, passes through 3 full articulamentums after the layer of pond, first full articulamentum there are 512
Neuron, second full articulamentum have 256 neurons, the last one full articulamentum has 10 neurons, indicate crowd's number
Amount is divided into 10 classes.
The crowd density estimation stage in multitask concatenated convolutional neural network includes 7 layers of convolutional layer and 2 layers of warp lamination,
There is a PReLU excitation function after preceding 6 layers of convolutional layer, the convolution kernel size of level 1 volume lamination is 7*7, the 2nd, 3 and 4 layer
The convolution kernel size of convolutional layer is all 7*7, there is a maximum pond layer, the step-length of each pond layer after preceding level 2 volume lamination
It is that characteristic pattern of the output characteristic pattern after 4 layers of convolution in advanced priori knowledge stage combines 2,4 layers of convolution later, connects the most
The input of two layers of the convolutional layer to get off, the convolution kernel size of the 5th and 6 layer of convolutional layer are all 3*3, then pass through two layers of deconvolution
Layer finally passes through the 7th layer of convolutional layer, which is 1*1, exports as crowd's density map and crowd's quantity.
Infrared crowd density estimation method, includes the following steps：
Step 1：Design crowd density estimation CNN network model
Using multitask concatenated convolutional neural network, inputs as infrared image, export as crowd's density map and crowd's quantity.
Step 2：Demarcate training sample
Ideal density figure D corresponding to i-th of training sampleiIt is calculated by two-dimensional Gaussian kernel summation, with everyone
Position xgCentered on dimensional Gaussian kernel, be defined as follows：
Wherein σ is the scale parameter of two-dimensional Gaussian kernel, and S is the set of all the points where the number of people.
Step 3：Training network
The cross entropy loss function in advanced priori knowledge stage is defined as follows：
Wherein N is number of training, and Θ is a set of network parameters, XiIt is i-th of training sample, Fc(Xi, Θ) and it is that classification is defeated
Out, yiIt is true value classification, M is classification sum.
The loss function in density estimation stage is defined as：
Wherein Fd(Xi, CiIt Θ) is estimation density map, DiIt is real density figure, CiFrom last of advanced priori knowledge stage
The characteristic pattern that a convolutional layer obtains.
Entire cascade network is trained using following loss function：
L=λ Lc+Ld
Wherein λ is weight factor, and λ takes 0.00001.
Step 4：Test network
Using trained multitask concatenated convolutional neural network, the crowd density in actual scene is tested.
A kind of crowd's abnormal behaviour monitoring method based on unmanned plane, includes the following steps：
Step 1：Monitoring personnel reaches monitoring region center, and monitoring personnel sets monitoring unmanned plane by earth station
Monitoring region, by the monitoring for monitoring unmanned plane height be set as 20m, concurrently set thermal infrared imager direction and vertical direction
Angle α is 45 °；
Step 2：After monitoring unmanned plane arrival 20m height, monitoring unmanned plane independently carries out spiral according to specified region
Formula search carries out crowd density estimation by airborne infrared vision system.If crowd's number is less than threshold value, continue
Spiral search, otherwise unmanned plane fixes current visual angle, without spiral search；
Step 3：After the fixed position of unmanned plane and visual angle, infrared video is obtained by thermal infrared imager FLIR TAU2
Information estimates crowd density using multitask concatenated convolutional neural network by embedded platform Jetson TX1, if people
Group's density is higher than threshold value, then continues to calculate the crowd's average movement velocity and crowd density between video two continuous frames, if crowd
Average speed increases or crowd density increases, and system alarm, otherwise system is not alarmed.
Claims (8)
1. a kind of crowd's abnormal behaviour monitoring device based on unmanned plane, which is characterized in that including unmanned plane body, flight control
System, infrared vision system, earth station system and hand-held remote controller；The infrared vision system and flight control system are built
In unmanned plane body center, hand-held remote controller and flight control system wireless communication are connected, the earth station system and infrared view
Feel Radio Communication connection, infrared vision system and flight control system communicate to connect.
2. crowd's abnormal behaviour monitoring device according to claim 1 based on unmanned plane, which is characterized in that it is described nobody
Machine body includes X-shaped rack, propeller and battery, and the propeller is set to the top at four angles of X-shaped rack, the battery
It is set to X-shaped machine frame inside.
3. crowd's abnormal behaviour monitoring device according to claim 1 based on unmanned plane, which is characterized in that described infrared
Vision system includes embedded platform Jetson TX1, thermal infrared imager FLIR TAU2, video frequency collection card and image transmission system；Institute
It states thermal infrared imager FLIR TAU2 and is separately connected video frequency collection card and image transmission system；The image transmission system and earth station system are logical
Letter connection, the video frequency collection card connect embedded platform Jetson TX1, and the embedded platform Jetson TX1 connection flies
Line control unit.
4. crowd's abnormal behaviour monitoring device according to claim 1 based on unmanned plane, which is characterized in that the flight
Control system includes microcontroller STM32F427, brushless motor, electron speed regulator, gyroscope, electronic compass, accelerometer, gas
Pressure meter, GPS module and the first wireless communication module, microcontroller STM32F427 be separately connected brushless motor, electron speed regulator,
Gyroscope, electronic compass, accelerometer, barometer, GPS module and the first wireless communication module, first radio communication mold
The main control chip of block is nRF24L01.
5. crowd's abnormal behaviour monitoring device according to claim 4 based on unmanned plane, which is characterized in that described hand-held
Remote controler includes microcontroller STM32F103VET6 and the second wireless communication module, and the microcontroller STM32F103VET6 connects
The second wireless communication module is connect, the main control chip of second wireless communication module is nRF24L01, second wireless communication
Module and the first wireless communication module connect by wireless communication, and hand-held remote controller is for manually controlling unmanned plane.
6. a kind of crowd's abnormal behaviour monitoring method based on unmanned plane according to claim 1, which is characterized in that including
Following steps：
Step S1：The monitoring region of monitoring unmanned plane is set by earth station；
Step S2：The monitoring for monitoring unmanned plane height is set as 20m, concurrently sets the direction and vertical direction of thermal infrared imager
Angle α is 45 °；
Step S3：Spiral search is carried out according to monitoring height, the angle of setting thermal infrared imager and specified region, by red
Outer vision system carries out crowd density estimation；
Step S4：Judge whether crowd's number of infrared vision system detection is less than threshold value, if so, S5 is entered step, otherwise,
Enter step S6；
Step S5：Continue spiral search, return step S4；
Step S6：The fixed current visual angle of unmanned plane is without spiral search；
Step S7：Infrared video information is obtained by thermal infrared imager FLIR TAU2, and passes through embedded platform Jetson TX1
Crowd density is estimated using multitask concatenated convolutional neural network；
Step S8：Judge whether variable density and average movement velocity variation are abnormal, if so, S9 is entered step, otherwise,
System does not alarm and enters step S6.
Step S9：System alarm terminates program.
7. a kind of crowd's abnormal behaviour monitoring method based on unmanned plane according to claim 6, which is characterized in that described
Step S7's estimates crowd density using multitask concatenated convolutional neural network method by embedded platform Jetson TX1,
The multitask concatenated convolutional neural network method includes the following steps：
Step A1：Infrared image is inputted, image is inputted into shared network phase；
Step A2：According to the image of shared network phase, output image is inputted into the advanced priori knowledge stage；
Step A3：The 10 class crowds by the output in advanced priori knowledge stage are inputted into the crowd density estimation stage；
Step A4：After the crowd density estimation stage, crowd density is exported, terminates program.
8. a kind of crowd's abnormal behaviour monitoring method based on unmanned plane according to claim 6, which is characterized in that described
The infrared vision system of step S3 carries out crowd density estimation method and includes the following steps：
Step B1：Design crowd density estimation CNN network model；
Step B2：According to CNN network model, training sample is demarcated, then the ideal density figure D of training samplei；
In formula, σ is the scale parameter of two-dimensional Gaussian kernel, and S is the set of all the points where the number of people；
Step B3：According to training sample, training network is established,
The cross entropy in training network advanced priori knowledge stage damages function LcFor：
In formula, N is number of training, and Θ is a set of network parameters, XiIt is i-th of training sample, Fc(Xi, Θ) and it is classification output,
yiIt is true value classification, M is classification sum；
The loss function L in density estimation stagedFor：
Wherein Fd(Xi,CiIt Θ) is estimation density map, DiIt is real density figure, CiBe from the advanced priori knowledge stage the last one
The characteristic pattern that convolutional layer obtains；
Entirely the loss function of cascade network is：
L=λ Lc+Ld
Wherein λ is weight factor, and λ takes 0.00001；
Step B4：It is tested using trained multi-cascade convolutional neural networks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810762340.6A CN108848348A (en) | 2018-07-12 | 2018-07-12 | A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810762340.6A CN108848348A (en) | 2018-07-12 | 2018-07-12 | A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108848348A true CN108848348A (en) | 2018-11-20 |
Family
ID=64196992
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810762340.6A Pending CN108848348A (en) | 2018-07-12 | 2018-07-12 | A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108848348A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109934148A (en) * | 2019-03-06 | 2019-06-25 | 华瑞新智科技（北京）有限公司 | A kind of real-time people counting method, device and unmanned plane based on unmanned plane |
| CN110611877A (en) * | 2019-04-24 | 2019-12-24 | 西南科技大学 | Violence abnormal behavior monitoring system and method based on unmanned aerial vehicle |
| CN110991375A (en) * | 2019-12-10 | 2020-04-10 | 北京航空航天大学 | Group behavior analysis method and device |
| CN111291597A (en) * | 2018-12-07 | 2020-06-16 | 杭州海康威视数字技术股份有限公司 | Image-based crowd situation analysis method, device, equipment and system |
| CN111797739A (en) * | 2020-06-23 | 2020-10-20 | 中国平安人寿保险股份有限公司 | Reminding information sending method and device based on double scanning and computer equipment |
| CN118366065A (en) * | 2024-04-26 | 2024-07-19 | 武汉大学 | Unmanned aerial vehicle image vehicle detection method and system based on height information |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102682303A (en) * | 2012-03-13 | 2012-09-19 | 上海交通大学 | Crowd exceptional event detection method based on LBP (Local Binary Pattern) weighted social force model |
| US20160042529A1 (en) * | 2014-08-11 | 2016-02-11 | Nongjian Tao | Systems and Methods for Non-Contact Tracking and Analysis of Physical Activity |
| CN205098466U (en) * | 2015-10-08 | 2016-03-23 | 安徽理工大学 | Population density monitoring devices based on many rotors |
| CN205499368U (en) * | 2016-04-06 | 2016-08-24 | 成都积格科技有限公司 | Crowd gathers early warning unmanned aerial vehicle |
| CN106203331A (en) * | 2016-07-08 | 2016-12-07 | 苏州平江历史街区保护整治有限责任公司 | A kind of crowd density evaluation method based on convolutional neural networks |
| WO2017041303A1 (en) * | 2015-09-11 | 2017-03-16 | SZ DJI Technology Co., Ltd. | Systems and methods for detecting and tracking movable objects |
| CN106533594A (en) * | 2017-01-04 | 2017-03-22 | 上海量明科技发展有限公司 | Broadcast and unmanned aerial vehicle combined system and implementation method thereof |
| CN107330372A (en) * | 2017-06-05 | 2017-11-07 | 四川大学 | A kind of crowd density based on video and the analysis method of unusual checking system |
| CN107352032A (en) * | 2017-07-14 | 2017-11-17 | 广东工业大学 | A kind of monitoring method and unmanned plane of flow of the people data |
| US20170341746A1 (en) * | 2016-05-27 | 2017-11-30 | International Business Machines Corporation | Unmanned aerial vehicle for crowd control amelioration |
| CN206968975U (en) * | 2017-07-14 | 2018-02-06 | 广东工业大学 | A kind of unmanned plane |
| CN108052859A (en) * | 2017-10-31 | 2018-05-18 | 深圳大学 | A kind of anomaly detection method, system and device based on cluster Optical-flow Feature |
| US20180189557A1 (en) * | 2016-05-23 | 2018-07-05 | Intel Corporation | Human detection in high density crowds |
-
2018
- 2018-07-12 CN CN201810762340.6A patent/CN108848348A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102682303A (en) * | 2012-03-13 | 2012-09-19 | 上海交通大学 | Crowd exceptional event detection method based on LBP (Local Binary Pattern) weighted social force model |
| US20160042529A1 (en) * | 2014-08-11 | 2016-02-11 | Nongjian Tao | Systems and Methods for Non-Contact Tracking and Analysis of Physical Activity |
| WO2017041303A1 (en) * | 2015-09-11 | 2017-03-16 | SZ DJI Technology Co., Ltd. | Systems and methods for detecting and tracking movable objects |
| CN205098466U (en) * | 2015-10-08 | 2016-03-23 | 安徽理工大学 | Population density monitoring devices based on many rotors |
| CN205499368U (en) * | 2016-04-06 | 2016-08-24 | 成都积格科技有限公司 | Crowd gathers early warning unmanned aerial vehicle |
| US20180189557A1 (en) * | 2016-05-23 | 2018-07-05 | Intel Corporation | Human detection in high density crowds |
| US20170341746A1 (en) * | 2016-05-27 | 2017-11-30 | International Business Machines Corporation | Unmanned aerial vehicle for crowd control amelioration |
| CN106203331A (en) * | 2016-07-08 | 2016-12-07 | 苏州平江历史街区保护整治有限责任公司 | A kind of crowd density evaluation method based on convolutional neural networks |
| CN106533594A (en) * | 2017-01-04 | 2017-03-22 | 上海量明科技发展有限公司 | Broadcast and unmanned aerial vehicle combined system and implementation method thereof |
| CN107330372A (en) * | 2017-06-05 | 2017-11-07 | 四川大学 | A kind of crowd density based on video and the analysis method of unusual checking system |
| CN107352032A (en) * | 2017-07-14 | 2017-11-17 | 广东工业大学 | A kind of monitoring method and unmanned plane of flow of the people data |
| CN206968975U (en) * | 2017-07-14 | 2018-02-06 | 广东工业大学 | A kind of unmanned plane |
| CN108052859A (en) * | 2017-10-31 | 2018-05-18 | 深圳大学 | A kind of anomaly detection method, system and device based on cluster Optical-flow Feature |
Non-Patent Citations (2)
| Title |
|---|
| VISHWANATH A. SINDAGI、VISHAL M. PATEL: "CNN-based Cascaded Multi-task Learning of High-level Prior and Density Estimation for Crowd Counting", 《2017 14TH IEEE INTERNATIONAL CONFERENCE ON ADVANCED VIDEO AND SIGNAL BASED SURVEILLANCE (AVSS)》 * |
| 陈禹: "基于视频的人群数量统计及异常检测方法研究", 《中国优秀硕士学位论文全文数据库—信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111291597A (en) * | 2018-12-07 | 2020-06-16 | 杭州海康威视数字技术股份有限公司 | Image-based crowd situation analysis method, device, equipment and system |
| CN111291597B (en) * | 2018-12-07 | 2023-10-13 | 杭州海康威视数字技术股份有限公司 | Crowd situation analysis method, device, equipment and system based on image |
| CN109934148A (en) * | 2019-03-06 | 2019-06-25 | 华瑞新智科技（北京）有限公司 | A kind of real-time people counting method, device and unmanned plane based on unmanned plane |
| CN110611877A (en) * | 2019-04-24 | 2019-12-24 | 西南科技大学 | Violence abnormal behavior monitoring system and method based on unmanned aerial vehicle |
| CN110611877B (en) * | 2019-04-24 | 2020-09-29 | 西南科技大学 | Violence abnormal behavior monitoring system and method based on unmanned aerial vehicle |
| CN110991375A (en) * | 2019-12-10 | 2020-04-10 | 北京航空航天大学 | Group behavior analysis method and device |
| CN111797739A (en) * | 2020-06-23 | 2020-10-20 | 中国平安人寿保险股份有限公司 | Reminding information sending method and device based on double scanning and computer equipment |
| CN111797739B (en) * | 2020-06-23 | 2023-09-08 | 中国平安人寿保险股份有限公司 | Dual-scanning-based reminding information sending method and device and computer equipment |
| CN118366065A (en) * | 2024-04-26 | 2024-07-19 | 武汉大学 | Unmanned aerial vehicle image vehicle detection method and system based on height information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108848348A (en) | A kind of crowd's abnormal behaviour monitoring device and method based on unmanned plane | |
| CN105759834B (en) | A kind of system and method actively capturing low latitude small-sized unmanned aircraft | |
| Jian et al. | Combining unmanned aerial vehicles with artificial-intelligence technology for traffic-congestion recognition: electronic eyes in the skies to spot clogged roads | |
| CN107817820A (en) | A kind of unmanned plane autonomous flight control method and system based on deep learning | |
| CN104236548B (en) | Autonomous navigation method in a kind of MAV room | |
| CN109255286A (en) | A kind of quick detection recognition method of unmanned plane optics based on YOLO deep learning network frame | |
| CN102190081B (en) | Vision-based fixed point robust control method for airship | |
| CN107850902A (en) | Camera configuration in loose impediment | |
| Karim et al. | Image processing based proposed drone for detecting and controlling street crimes | |
| CN106339006A (en) | Object tracking method of aircraft and apparatus thereof | |
| CN111679695B (en) | Unmanned aerial vehicle cruising and tracking system and method based on deep learning technology | |
| CN107209514A (en) | The selectivity processing of sensing data | |
| CN105334347B (en) | A kind of particle image velocimetry detection system and method based on unmanned plane | |
| CN106155082B (en) | A kind of unmanned plane bionic intelligence barrier-avoiding method based on light stream | |
| CN108500992A (en) | A kind of multi-functional mobile security robot | |
| CN109035665A (en) | A kind of novel forest fire early-warning system and fire alarm method | |
| CN110109469A (en) | It is a kind of with color, identification, positioning, following function quadrotor drone control system | |
| CN104816829A (en) | Skyeye aircraft applicable to investigation | |
| CN109801484A (en) | A kind of emergency communication UAV system and emergency communication system | |
| CN115035470A (en) | Low, small and slow target identification and positioning method and system based on mixed vision | |
| CN109960278B (en) | LGMD-based bionic obstacle avoidance control system and method for unmanned aerial vehicle | |
| CN108614572A (en) | A kind of target identification method for tracing, equipment and storage device based on aircraft | |
| CN109709975A (en) | A kind of quadrotor indoor security system and method for view-based access control model SLAM | |
| CN109649654A (en) | A kind of low altitude coverage localization method | |
| CN107168362A (en) | A kind of monitoring method, system and flying robot |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181120 |
|
| RJ01 | Rejection of invention patent application after publication |