Abnormal activity monitoring method and device based on artificial intelligence

Abstract

The present invention relates to the field of artificial intelligence technology, and in particular to an abnormal activity monitoring method and device based on artificial intelligence. The method comprises: generating a first association mapping diagram for projecting a monitoring field corresponding to each monitoring device based on campus building and road planning parameters combined with the setting position of a monitoring device; obtaining abnormal moving targets captured by multiple monitoring devices during night patrols, and updating the position information of the abnormal moving targets in real time based on the first association mapping diagram; sending an early warning execution instruction to a drone so that the drone calling device performs a call attendance task; the application first constructs a monitoring area mapping diagram based on an existing monitoring device to achieve real-time tracking of abnormal targets across cameras, and then dispatches a corresponding calling drone for remote calling, which greatly reduces the number of related security personnel and reduces personnel costs. At the same time, intelligent automatic monitoring can also effectively avoid monitoring omissions caused by human negligence.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an abnormal activity monitoring method and device based on artificial intelligence. Background Art

[0002] Campus is a place where students gather, so it is of great significance to monitor abnormal activities of students on campus. Effective supervision methods can reduce the accident rate of students on campus, especially for monitoring abnormal behaviors during non-normal activity time periods (such as night patrols after bedtime);

[0003] The existing night patrol still uses manpower to manually traverse the campus, such as regular and fixed area inspections by security personnel. Secondly, the security personnel in the monitoring room call the corresponding field patrol personnel to implement the on-site situation after seeing abnormal behavior. However, the above manual methods have the following problems:

[0004] On the one hand, there is a possibility of negligence in human monitoring during the second half of the night, such as security personnel in the monitoring room falling asleep, which may lead to supervision omissions;

[0005] Secondly, the high labor cost. When the campus is large (such as a university campus), the cost of supporting night patrol field staff and monitoring personnel is high.

[0006] However, with the maturity of artificial intelligence algorithms and image recognition algorithms and the popularization of drone applications, it has become possible to realize intelligent monitoring of abnormal activities on campus at night based on the existing surveillance cameras on campus in combination with drones and artificial intelligence. Therefore, how to realize the above-mentioned intelligent integration has become an urgent problem to be solved. Summary of the invention

[0007] The purpose of the present invention is to provide an abnormal activity monitoring method and device based on artificial intelligence to improve the problems of high manual inspection costs and easy supervision omissions in the above-mentioned prior art.

[0008] In order to achieve the above objectives, the present application provides the following technical solutions:

[0009] On the one hand, an embodiment of the present application provides an abnormal activity monitoring method based on artificial intelligence, the method comprising: generating a first association mapping diagram for projecting the monitoring areas corresponding to each monitoring device based on campus building and road planning parameters combined with the setting positions of the monitoring devices; during the night patrol period, obtaining multiple abnormal moving targets captured by the monitoring devices, and updating the position information of the abnormal moving targets in real time based on the first association mapping diagram; based on the real-time position of the abnormal moving object and the campus distribution information of the drone calling device, selecting the optimal drone calling device according to the task priority, and sending an early warning execution instruction to it, so that the drone calling device performs the call attendance task, and establishes a call connection with the security call station after arriving at the scene, so that the security personnel at the security call station can warn the person with abnormal activities at night through the loudspeaker broadcast of the drone calling device.

[0010] Optionally, the generating of a first association map for displaying a monitoring area corresponding to each monitoring device based on campus building and road planning parameters combined with the setting position of the monitoring device includes:

[0011] Obtain a campus planning map, and generate a three-dimensional campus building simulation map according to the campus planning map, wherein the three-dimensional campus building simulation map includes location information of multiple teaching buildings, building height information of each teaching building, green vegetation location information and campus road planning information;

[0012] Encode each teaching building and each campus road to obtain a virtual building number corresponding to each teaching building and a virtual road number corresponding to each campus road;

[0013] According to the distribution positions of the campus surveillance cameras and the orientation of each fixed camera, the virtual building number corresponding to the teaching building monitored by each fixed camera and the virtual road number corresponding to the road are determined, and then at least one fixed camera number corresponding to each virtual building number and at least one fixed camera number corresponding to each virtual road number are obtained, and then a mapping relationship diagram reflecting the fixed cameras and the virtual road numbers and virtual building numbers is obtained, which is recorded as the first association mapping diagram.

[0014] Optionally, during the night patrol period, acquiring a plurality of abnormal moving targets captured by the monitoring device, and updating the position information of the abnormal moving targets in real time based on the first association map, includes:

[0015] During the night patrol period, multiple moving object monitoring video data captured by multiple fixed cameras are obtained and recorded as the first data to be inspected;

[0016] Sequentially peeling off the outline of the moving object in each first data to be inspected based on the boundary segmentation algorithm of the dynamic image, and marking the corresponding first data to be inspected as the second data to be inspected when the outline of the moving object is determined to be the outline of a human body;

[0017] Retrieving the number of the fixed camera corresponding to the second data to be inspected, and combining it with the first association map, obtaining the number of the virtual road where the moving object in the second data to be inspected is located, and recording it as the second virtual road number;

[0018] Based on the Kalman filtering algorithm, the motion vector of the mobile object in the second data to be inspected is calculated and combined with the second virtual road number, the virtual road number and the corresponding fixed camera number corresponding to the next destination of the mobile object are predicted, and the camera corresponding to the corresponding fixed camera number is called to realize continuous tracking of the abnormal moving object and obtain the real-time position of the abnormal moving object.

[0019] Optionally, sending an early warning execution instruction to the drone calling device so that the drone calling device performs a calling attendance task includes:

[0020] The drone calling device formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map, and goes to the real-time position according to the navigation path. After reaching the real-time position, the drone calling device captures the moving target through the smart camera under the drone calling device and gradually approaches the moving object. When the distance detection unit on the smart camera detects that the distance between the moving object and the drone calling device is within the preset calling distance, a communication connection request is fed back to the campus detection platform;

[0021] The communication connection request is accepted and a communication connection is established between the security call station and the drone calling device, so that the security personnel at the security call station can warn the person of abnormal activities at night through the speaker broadcast of the drone calling device.

[0022] In a second aspect, an embodiment of the present application provides an abnormal activity monitoring device based on artificial intelligence, the device comprising:

[0023] A first calculation module, for generating a first association map for displaying a monitoring area corresponding to each monitoring device based on campus building and road planning parameters combined with the setting position of the monitoring device;

[0024] a first identification module, used for acquiring a plurality of abnormal moving targets captured by the monitoring device during a night patrol period, and updating the position information of the abnormal moving targets in real time based on the first association map;

[0025] The second computing module is used to select the best drone calling device according to the task priority based on the real-time position of the abnormal mobile object and the campus distribution information of the drone calling device, and send an early warning execution instruction to it, so that the drone calling device can perform the call attendance task and establish a call connection with the security call station after arriving at the scene, so that the security personnel at the security call station can warn the person with abnormal activities at night through the loudspeaker broadcast of the drone calling device.

[0026] Optionally, the first calculation module includes:

[0027] A first acquisition unit is used to acquire a campus planning map, and generate a three-dimensional campus building simulation map according to the campus planning map, wherein the three-dimensional campus building simulation map includes location information of multiple teaching buildings, building height information of each teaching building, green vegetation location information and campus road planning information;

[0028] The first numbering unit is used to encode each teaching building and each campus road to obtain a virtual building number corresponding to each teaching building and a virtual road number corresponding to each campus road;

[0029] The first mapping unit is used to determine the virtual building number corresponding to the teaching building monitored by each fixed camera and the virtual road number corresponding to the road according to the distribution position of the campus surveillance cameras and the direction of each fixed camera, and then obtain at least one fixed camera number corresponding to each virtual building number and at least one fixed camera number corresponding to each virtual road number, and then obtain a mapping relationship diagram reflecting the fixed cameras and the virtual road numbers and virtual building numbers, which is recorded as the first association mapping diagram.

[0030] Optionally, the first identification module includes:

[0031] The second acquisition unit is used to acquire a plurality of moving object monitoring video data captured by a plurality of fixed cameras during a night patrol period, and record the data as first data to be inspected;

[0032] A first recognition unit is used to sequentially peel off the outline of the moving object in each first to-be-checked data based on a boundary segmentation algorithm of a dynamic image, and when the outline of the moving object is determined to be a human body outline, mark the corresponding first to-be-checked data as second to-be-checked data;

[0033] A first mapping unit is used to retrieve the number of the fixed camera corresponding to the second data to be inspected, and obtain the number of the virtual road where the moving object in the second data to be inspected is located in combination with the first association mapping map, which is recorded as the second virtual road number;

[0034] The first calculation unit is used to calculate the motion vector of the mobile object in the second data to be inspected based on the Kalman filtering algorithm and combine it with the second virtual road number, predict the virtual road number and the corresponding fixed camera number corresponding to the next destination of the mobile object, and call the camera corresponding to the corresponding fixed camera number, so as to realize continuous tracking of the abnormal moving object and obtain the real-time position of the abnormal moving object.

[0035] Optionally, the second calculation module includes:

[0036] The instruction sending unit is used to send an early warning execution instruction to the designated drone, so that the drone calling device formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map, and goes to the real-time position according to the navigation path. After reaching the real-time position, the moving target is captured by the smart camera under the drone calling device, and gradually approaches the moving object. When the distance detection unit on the smart camera detects that the distance between the moving object and the drone calling device is within a preset calling distance, a communication connection request is fed back to the campus detection platform;

[0037] The communication connection unit is used to accept the communication connection request and then establish a communication connection between the security call station and the drone calling device, so that the security personnel located at the security call station can warn the person with abnormal activities at night through the speaker broadcast of the drone calling device.

[0038] In a third aspect, an embodiment of the present application provides an abnormal activity monitoring device based on artificial intelligence, wherein the device includes a memory and a processor.

[0039] The memory is used to store computer programs; the processor is used to implement the steps of the above-mentioned abnormal activity monitoring method based on artificial intelligence when executing the computer program.

[0040] In a fourth aspect, an embodiment of the present application provides a medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned artificial intelligence-based abnormal activity monitoring method are implemented.

[0041] The beneficial effects of the present invention are:

[0042] The present invention discloses an abnormal activity monitoring method, device, equipment and medium based on artificial intelligence. A monitoring area map is first constructed based on the existing monitoring device, so as to facilitate the switching of the monitoring screen of the automatic camera of artificial intelligence, realize the real-time tracking of abnormal targets across cameras, and then dispatch the corresponding calling drone to make a remote call (instead of field staff), thereby realizing the intelligent integration of monitoring, tracking and calling warning of abnormal personnel at night. Security personnel only need to make a remote call, which greatly reduces the number of related security personnel and reduces personnel costs. At the same time, intelligent automatic monitoring can also effectively avoid monitoring omissions due to human negligence.

[0043] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 It is a flow chart of an abnormal activity monitoring method based on artificial intelligence described in an embodiment of the present invention;

[0046] Figure 2 It is a schematic diagram of the structure of an abnormal activity monitoring device based on artificial intelligence described in an embodiment of the present invention;

[0047] Figure 3 It is a schematic diagram of the structure of an abnormal activity monitoring device based on artificial intelligence described in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0049] It should be noted that similar reference numerals or letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0050] Before giving examples, a brief description is needed of the distribution of collectors and power equipment. A fixed number of test specimens are randomly selected from the power equipment in a region, and corresponding collectors are set on the specimens to implement detection of the equipment. Secondly, the type of equipment and the specific type of power equipment to be monitored need to be the same, such as transformers, cables, substations, etc. This embodiment will take high-voltage transformers as the object of explanation, and the remaining power equipment can achieve the same identification of partial discharge types of power equipment according to the relevant principles disclosed in the embodiment, and this specification does not make relevant explanations.

[0051] Example 1

[0052] like Figure 1 As shown, this embodiment provides an abnormal activity monitoring method based on artificial intelligence, and the method includes step S1, step S2 and step S3.

[0053] Step S1, based on campus building and road planning parameters combined with the setting positions of the monitoring devices, generating a first association map for displaying the monitoring areas corresponding to each monitoring device;

[0054] The specific method of generating the first association map is as follows:

[0055] Step S11, obtaining a campus planning map, and generating a three-dimensional campus building simulation map according to the campus planning map, wherein the three-dimensional campus building simulation map includes location information of multiple teaching buildings, building height information of each teaching building, green vegetation location information and campus road planning information;

[0056] Step S12: Encode each teaching building and each campus road to obtain a virtual building number corresponding to each teaching building and a virtual road number corresponding to each campus road;

[0057] Step S13, according to the distribution position of the campus surveillance cameras and the direction of each fixed camera, determine the virtual building number corresponding to the teaching building monitored by each fixed camera and the virtual road number corresponding to the road, and then obtain at least one fixed camera number corresponding to each virtual building number and at least one fixed camera number corresponding to each virtual road number, and then obtain a mapping relationship diagram reflecting the fixed cameras and the virtual road numbers and virtual building numbers, which is recorded as the first association mapping diagram.

[0058] The role of establishing the above-mentioned basic information parameter model is divided into two aspects. On the one hand, the location information of green vegetation and the building height information of each teaching building are used for the subsequent planning of drone routes. For example, the location of green vegetation is a no-fly zone for drones, thereby preventing the crash of drones. Secondly, the building height information of each teaching building is used as the flight height reference information for later drones crossing the teaching buildings. On the other hand, a virtual campus simulation map and the correlation between the camera coverage area of ​​each area are established, thereby forming a linkage of multiple cameras indexed by the ground area. In other words, the artificial intelligence algorithm only needs to know the target area and its corresponding moving vector to call the surveillance camera of the next destination in advance, thereby realizing real-time intelligent monitoring of dynamic cross-cameras indexed by moving routes.

[0059] Step S2: during the night patrol period, obtain a plurality of abnormal moving targets captured by the monitoring device, and update the position information of the abnormal moving targets in real time based on the first association map;

[0060] Among them, the specific methods for real-time follow-up of abnormal moving target location information are:

[0061] Step S21: During the night patrol period, a plurality of moving object monitoring video data captured by a plurality of fixed cameras are obtained, which are recorded as the first data to be inspected. The first data to be inspected includes the image data of vehicle movement and abnormal personnel movement, etc. Therefore, it is necessary to remove the relevant interference images. For the specific stripping method, see step S22;

[0062] Step S22, based on the boundary segmentation algorithm of the dynamic image, the contour of the moving object in each first data to be inspected is sequentially stripped, and when the contour of the moving object is determined to be a human contour, the corresponding first data to be inspected is marked as the second data to be inspected, wherein the method for determining the human contour can be: 1. Obtaining the spectrum of the contour of the moving object, and judging that it is not human movement if there is strong white light during the movement; 2. Presetting multiple human contour parameters at different distances and angles, and assisting other cameras to perform multi-angle contour comprehensive weighted analysis and judgment when a single camera cannot make a judgment;

[0063] Step S23, retrieve the number of the fixed camera corresponding to the second data to be inspected, and combine it with the first association map to obtain the number of the virtual road where the moving object in the second data to be inspected is located, which is recorded as the second virtual road number;

[0064] Step S24, based on the Kalman filter algorithm, the motion vector of the mobile object in the second data to be inspected is calculated and combined with the second virtual road number, the virtual road number and the corresponding fixed camera number corresponding to the next destination of the mobile object are predicted, and the camera corresponding to the corresponding fixed camera number is called up to realize continuous tracking of the abnormal moving object and obtain the real-time position of the abnormal moving object.

[0065] Step S3: Based on the real-time location of the abnormal moving object and the campus distribution information of the drone calling device, an optimal drone calling device is selected according to the task priority, and an early warning execution instruction is sent to the drone calling device, so that the drone calling device performs the call attendance task and establishes a call connection with the security calling station after arriving at the scene, so that the security personnel at the security calling station can warn the person with abnormal activities at night through the loudspeaker broadcast of the drone calling device;

[0066] Among them, after receiving the early warning execution instruction, the drone first formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map. The formulated navigation path includes the flight altitude and route, and should avoid entering the high vegetation airspace and flying at an altitude 20% higher than the building height when crossing the building. Then, it goes to the real-time position according to the navigation path, and after reaching the real-time position, the moving target is captured by the smart camera under the drone calling device, and gradually approaches the moving object. When the distance detection unit on the smart camera detects that the distance between the moving object and the drone calling device is within the preset calling distance, a communication connection request is fed back to the campus detection platform;

[0067] The control center accepts the communication connection request and establishes a communication connection between the security call center and the drone calling device, so that the security personnel at the security call center can warn the person of abnormal activities at night through the speaker broadcast of the drone calling device.

[0068] This invention discloses an abnormal activity monitoring method based on artificial intelligence. It first constructs a monitoring area map based on the existing monitoring device, and then facilitates the switching of the artificial intelligence automatic camera monitoring screen, realizes the real-time tracking of abnormal targets across cameras, and then dispatches the corresponding calling drone to make a remote call (instead of field staff), thereby realizing the intelligent integration of monitoring, tracking and calling warning of abnormal personnel at night. Security personnel only need to make remote calls, which greatly reduces the number of relevant security personnel and reduces personnel costs. At the same time, intelligent automatic monitoring can also effectively avoid monitoring omissions due to human negligence.

[0069] Secondly, in step S3, based on the real-time location of the abnormal moving object and the campus distribution information of the drone calling device, the optimal drone calling device is selected according to the task priority, and an early warning execution instruction is sent to it. The specific operation method is as follows:

[0070] Step S31, obtaining the real-time positions of the plurality of mobile objects, counting their number, and sending online detection to each drone calling device in turn, so that each drone calling device feeds back the local power and idle status to the patrol station server;

[0071] Step S31: If it is detected that the number of drone call devices in the current idle state is less than the number of the mobile objects, the emergency call tasks corresponding to the multiple mobile objects are graded according to the real-time positions of the mobile objects, and the highest-level call tasks are screened out and recorded as the first call tasks, wherein the closer the person is to the predetermined specific position, the higher the corresponding call task level;

[0072] Step S31: according to the position of each idle drone calling device, the theoretical power consumption and redundant power of each idle drone calling device in the round trip to the destination corresponding to the first calling task are calculated, and the actual power consumption of each idle drone calling device in performing the first calling task is obtained, wherein the redundant power is the power consumption calculated according to the temperature and the round trip distance, thereby ensuring that the drone calling device has sufficient power to complete the task;

[0073] Step S31, sequentially detecting the current remaining power and the corresponding actual power consumption of each idle drone calling device, and recording the drone calling device whose current remaining power is greater than the corresponding actual power consumption as the first drone calling device;

[0074] Step S31: select a first drone calling device that is closest to the destination corresponding to the first calling task, record it as the optimal drone calling device, and send the early warning execution instruction corresponding to the first calling task to it.

[0075] Example 2

[0076] like Figure 2 As shown, this embodiment provides an abnormal activity monitoring device based on artificial intelligence, and the device includes:

[0077] A first calculation module 71, for generating a first association map for displaying a monitoring area corresponding to each monitoring device based on campus building and road planning parameters combined with the setting position of the monitoring device;

[0078] A first identification module 72 is used to obtain a plurality of abnormal moving targets captured by the monitoring device during a night patrol period, and update the position information of the abnormal moving targets in real time based on the first association map;

[0079] The second calculation module 73 is used to select the best drone calling device according to the task priority based on the real-time position of the abnormal mobile object and the campus distribution information of the drone calling device, and send an early warning execution instruction to it, so that the drone calling device can perform the call attendance task and establish a call connection with the security call station after arriving at the scene, so that the security personnel at the security call station can warn the person with abnormal activities at night through the loudspeaker broadcast of the drone calling device.

[0080] Optionally, the first calculation module 71 includes:

[0081] The first acquisition unit 711 is used to acquire a campus planning map, and generate a three-dimensional campus building simulation map according to the campus planning map, wherein the three-dimensional campus building simulation map includes location information of multiple teaching buildings, building height information of each teaching building, green vegetation location information and campus road planning information;

[0082] The first numbering unit 712 is used to encode each teaching building and each campus road to obtain a virtual building number corresponding to each teaching building and a virtual road number corresponding to each campus road;

[0083] The first mapping unit 713 is used to determine the virtual building number corresponding to the teaching building monitored by each fixed camera and the virtual road number corresponding to the road according to the distribution position of the campus surveillance cameras and the direction of each fixed camera, and then obtain at least one fixed camera number corresponding to each virtual building number and at least one fixed camera number corresponding to each virtual road number, and then obtain a mapping relationship diagram reflecting the fixed cameras and the virtual road numbers and virtual building numbers, which is recorded as the first association mapping diagram.

[0084] Optionally, the first identification module 72 includes:

[0085] The second acquisition unit 721 is used to acquire a plurality of moving object monitoring video data captured by a plurality of fixed cameras during the night patrol period, and record the data as the first data to be inspected;

[0086] The first identification unit 722 is used to sequentially peel off the outline of the moving object in each first to-be-checked data based on the boundary segmentation algorithm of the dynamic image, and when the outline of the moving object is determined to be a human body outline, mark the corresponding first to-be-checked data as the second to-be-checked data;

[0087] The first mapping unit 723 is used to retrieve the number of the fixed camera corresponding to the second data to be inspected, and obtain the number of the virtual road where the moving object in the second data to be inspected is located in combination with the first association mapping diagram, which is recorded as the second virtual road number;

[0088] The first calculation unit 724 is used to calculate the motion vector of the mobile object in the second data to be inspected based on the Kalman filtering algorithm and combine it with the second virtual road number, predict the virtual road number and the corresponding fixed camera number corresponding to the next destination of the mobile object, and call the camera corresponding to the corresponding fixed camera number, so as to realize continuous tracking of the abnormal moving object and obtain the real-time position of the abnormal moving object.

[0089] Optionally, the second calculation module 73 includes:

[0090] The instruction sending unit 731 is used to send an early warning execution instruction to the designated drone, so that the drone calling device formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map, and goes to the real-time position according to the navigation path. After reaching the real-time position, the moving target is captured by the smart camera under the drone calling device, and gradually approaches the moving object. When the distance detection unit on the smart camera detects that the distance between the moving object and the drone calling device is within the preset calling distance, a communication connection request is fed back to the campus detection platform;

[0091] The communication connection unit 732 is used to accept the communication connection request and then establish a communication connection between the security call station and the drone calling device, so that the security personnel located at the security call station can warn the person with abnormal activities at night through the speaker broadcast of the drone calling device.

[0092] It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0093] Example 3

[0094] Corresponding to the above method embodiment, the embodiment of the present disclosure also provides an abnormal activity monitoring device based on artificial intelligence. The abnormal activity monitoring device based on artificial intelligence described below and the abnormal activity monitoring method based on artificial intelligence described above can refer to each other.

[0095] Figure 3 FIG. 8 is a block diagram of an abnormal activity monitoring device 800 based on artificial intelligence according to an exemplary embodiment. Figure 3 As shown, the electronic device 800 may include: a processor 801 and a memory 802. The electronic device 800 may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0096] The processor 801 is used to control the overall operation of the electronic device 800 to complete all or part of the steps in the above-mentioned abnormal activity monitoring method based on artificial intelligence. The memory 802 is used to store various types of data to support the operation of the electronic device 800, which may include, for example, instructions for any application or method used to operate on the electronic device 800, and application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read-Only Memory, referred to as EPROM), programmable read-only memory (Programmable Read-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone, which is used to receive external audio signals. The received audio signal may be further stored in the memory 802 or sent through the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules may be keyboards, mice, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the electronic device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 may include: Wi-Fi module, Bluetooth module, NFC module.

[0097] In an exemplary embodiment, the electronic device 800 can be implemented by one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned artificial intelligence-based abnormal activity monitoring method.

[0098] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, and when the program instructions are executed by a processor, the steps of the abnormal activity monitoring method based on artificial intelligence are implemented. For example, the computer-readable storage medium can be the memory 802 including the program instructions, and the program instructions can be executed by the processor 801 of the electronic device 800 to complete the abnormal activity monitoring method based on artificial intelligence.

[0099] Example 4

[0100] Corresponding to the above method embodiment, the embodiment of the present disclosure further provides a readable storage medium. The readable storage medium described below and the abnormal activity monitoring method based on artificial intelligence described above can refer to each other.

[0101] A readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the abnormal activity monitoring method based on artificial intelligence of the above method embodiment.

[0102] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An abnormal activity monitoring method based on artificial intelligence, characterized in that: The method comprises: Based on the campus building and road planning parameters combined with the setting positions of the monitoring devices, a first association map for displaying the monitoring areas corresponding to the monitoring devices is generated; During the night patrol period, a plurality of abnormal moving targets captured by the monitoring device are acquired, and the position information of the abnormal moving targets is updated in real time based on the first association map; Based on the real-time location of the abnormal moving object and the campus distribution information of the drone calling device, the best drone calling device is selected according to the task priority, and an early warning execution instruction is sent to it, so that the drone calling device performs the call attendance task, and establishes a call connection with the security calling station after arriving at the scene, so that the security personnel at the security calling station can warn the person with abnormal activities at night through the loudspeaker broadcast of the drone calling device. Secondly, after receiving the early warning execution instruction, the drone first formulates a corresponding navigation path according to the real-time location of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map. The formulated navigation path includes the flight altitude and route, and should avoid entering the high vegetation airspace and flying at an altitude 20% higher than the building height when crossing the building; Secondly, based on the real-time location of the abnormal moving object and the campus distribution information of the drone calling device, the optimal drone calling device is selected according to the task priority, and an early warning execution instruction is sent to it. The specific operation method is as follows: Obtain the real-time positions of the plurality of mobile objects, count their number, and send online detection to each drone calling device in turn, so that each drone calling device can feedback the local power and idle status to the patrol station server; If it is detected that the number of drone call devices currently in idle state is lower than the number of the mobile objects, the emergency call tasks corresponding to the multiple mobile objects are graded according to the real-time positions of the mobile objects, and the highest-level call tasks are screened out and recorded as the first call tasks, wherein the closer the person is to the predetermined specific position, the higher the corresponding call task level; According to the position of each idle drone calling device, the theoretical power consumption and redundant power of each idle drone calling device in the round trip to the destination corresponding to the first calling task are calculated, and the actual power consumption of each idle drone calling device in performing the first calling task is obtained, wherein the redundant power is the power consumption calculated according to the temperature and the round trip distance, thereby ensuring that the drone calling device can have sufficient power to complete this task; Detect the current remaining power and the corresponding actual power consumption of each idle drone calling device in turn, and record the drone calling device whose current remaining power is greater than the corresponding actual power consumption as the first drone calling device; Selecting a first drone calling device that is closest to the destination corresponding to the first calling task as the best drone calling device, and sending the warning execution instruction corresponding to the first calling task to the drone calling device; The method of generating a first association map for displaying the monitoring area corresponding to each monitoring device based on campus building and road planning parameters combined with the setting position of the monitoring device includes: Obtain a campus planning map, and generate a three-dimensional campus building simulation map according to the campus planning map, wherein the three-dimensional campus building simulation map includes location information of multiple teaching buildings, building height information of each teaching building, green vegetation location information and campus road planning information; Encode each teaching building and each campus road to obtain a virtual building number corresponding to each teaching building and a virtual road number corresponding to each campus road; According to the distribution positions of the campus surveillance cameras and the orientation of each fixed camera, the virtual building number corresponding to the teaching building monitored by each fixed camera and the virtual road number corresponding to the road are determined, and then at least one fixed camera number corresponding to each virtual building number and at least one fixed camera number corresponding to each virtual road number are obtained, and then a mapping relationship diagram reflecting the fixed cameras and the virtual road numbers and virtual building numbers is obtained, which is recorded as a first association mapping diagram; The method of acquiring a plurality of abnormal moving targets captured by the monitoring device during the night patrol period and updating the position information of the abnormal moving targets in real time based on the first association map includes: During the night patrol period, multiple moving object monitoring video data captured by multiple fixed cameras are obtained and recorded as the first data to be inspected; The boundary segmentation algorithm based on dynamic images sequentially strips the contour of the moving object in each first data to be inspected, and when the contour of the moving object is determined to be a human contour, the corresponding first data to be inspected is marked as the second data to be inspected, wherein the method for determining the human contour is: obtaining the spectrum of the contour of the moving object, and determining that it is not a human movement if there is strong white light during the movement; presetting multiple human contour parameters at different distances and angles, and assisting other cameras in performing a comprehensive weighted analysis and determination of multi-angle contours when a single camera cannot make a determination; Retrieving the number of the fixed camera corresponding to the second data to be inspected, and combining it with the first association map, obtaining the number of the virtual road where the moving object in the second data to be inspected is located, and recording it as the second virtual road number; Based on the Kalman filter algorithm, the motion vector of the mobile object in the second to-be-inspected data is calculated and combined with the second virtual road number, the virtual road number and the corresponding fixed camera number corresponding to the next destination of the mobile object are predicted, and the camera corresponding to the corresponding fixed camera number is retrieved, so as to realize continuous tracking of the abnormal mobile object and obtain the real-time position of the abnormal mobile object; Among them, sending an early warning execution instruction to the drone calling device so that the drone calling device performs the calling attendance task includes: The drone calling device formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map, and goes to the real-time position according to the navigation path. After reaching the real-time position, the drone calling device captures the moving target through the smart camera under the drone calling device and gradually approaches the moving object. When the distance detection unit on the smart camera detects that the distance between the moving object and the drone calling device is within the preset calling distance, a communication connection request is fed back to the campus detection platform; The communication connection request is accepted and a communication connection is established between the security call station and the drone calling device, so that the security personnel at the security call station can warn the person of abnormal activities at night through the speaker broadcast of the drone calling device.

2. An abnormal activity monitoring device based on artificial intelligence, characterized in that: The device comprises: A first calculation module, for generating a first association map for displaying a monitoring area corresponding to each monitoring device based on campus building and road planning parameters combined with the setting position of the monitoring device; a first identification module, used for acquiring a plurality of abnormal moving targets captured by the monitoring device during a night patrol period, and updating the position information of the abnormal moving targets in real time based on the first association map; The second calculation module is used to select the best drone calling device according to the task priority based on the real-time position of the abnormal moving object and the campus distribution information of the drone calling device, and send an early warning execution instruction to it, so that the drone calling device performs the call attendance task, and establishes a call connection with the security calling station after arriving at the scene, so that the security personnel at the security calling station can warn the person with abnormal activities at night through the loudspeaker broadcast of the drone calling device. Secondly, after receiving the early warning execution instruction, the drone first formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map, wherein the formulated navigation path includes the flight altitude and route, and should avoid entering the high vegetation airspace and flying at an altitude 20% higher than the building height when crossing the building; Secondly, based on the real-time location of the abnormal moving object and the campus distribution information of the drone calling device, the optimal drone calling device is selected according to the task priority, and an early warning execution instruction is sent to it. The specific operation method is as follows: Obtain the real-time positions of the plurality of mobile objects, count their number, and send online detection to each drone calling device in turn, so that each drone calling device can feedback the local power and idle status to the patrol station server; If it is detected that the number of drone call devices currently in idle state is lower than the number of the mobile objects, the emergency call tasks corresponding to the multiple mobile objects are graded according to the real-time positions of the mobile objects, and the highest-level call tasks are screened out and recorded as the first call tasks, wherein the closer the person is to the predetermined specific position, the higher the corresponding call task level; According to the position of each idle drone calling device, the theoretical power consumption and redundant power of each idle drone calling device in the round trip to the destination corresponding to the first calling task are calculated, and the actual power consumption of each idle drone calling device in performing the first calling task is obtained, wherein the redundant power is the power consumption calculated according to the temperature and the round trip distance, thereby ensuring that the drone calling device can have sufficient power to complete this task; Detect the current remaining power and the corresponding actual power consumption of each idle drone calling device in turn, and record the drone calling device whose current remaining power is greater than the corresponding actual power consumption as the first drone calling device; Selecting a first drone calling device that is closest to the destination corresponding to the first calling task as the best drone calling device, and sending the warning execution instruction corresponding to the first calling task to the drone calling device; Wherein, the first calculation module includes: A first acquisition unit is used to acquire a campus planning map, and generate a three-dimensional campus building simulation map according to the campus planning map, wherein the three-dimensional campus building simulation map includes location information of multiple teaching buildings, building height information of each teaching building, green vegetation location information and campus road planning information; The first numbering unit is used to encode each teaching building and each campus road to obtain a virtual building number corresponding to each teaching building and a virtual road number corresponding to each campus road; A first mapping unit is used to determine, according to the distribution positions of the campus surveillance cameras and the orientation of each fixed camera, the virtual building number corresponding to the teaching building monitored by each fixed camera and the virtual road number corresponding to the road, and then obtain at least one fixed camera number corresponding to each virtual building number and at least one fixed camera number corresponding to each virtual road number, and then obtain a mapping relationship diagram reflecting the fixed cameras and the virtual road numbers and virtual building numbers, which is recorded as a first association mapping diagram; Wherein, the first identification module includes: The second acquisition unit is used to acquire a plurality of moving object monitoring video data captured by a plurality of fixed cameras during a night patrol period, and record the data as first data to be inspected; The first recognition unit is used to sequentially peel off the contour of the moving object in each first inspection data based on the boundary segmentation algorithm of the dynamic image, and when the contour of the moving object is determined to be a human contour, mark the corresponding first inspection data as the second inspection data, wherein the method for determining the human contour is: obtaining the spectrum of the contour of the moving object, and determining that the movement is not human if there is strong white light during the movement; presetting multiple human contour parameters at different distances and angles, and assisting other cameras in performing multi-angle contour comprehensive weighted analysis and determination when a single camera cannot make a determination; A first mapping unit is used to retrieve the number of the fixed camera corresponding to the second data to be inspected, and obtain the number of the virtual road where the moving object in the second data to be inspected is located in combination with the first association mapping map, which is recorded as the second virtual road number; A first calculation unit is used to calculate the motion vector of the mobile object in the second to-be-detected data based on a Kalman filter algorithm and combine it with the second virtual road number, predict the virtual road number and the corresponding fixed camera number corresponding to the next destination of the mobile object, and call the camera corresponding to the corresponding fixed camera number, so as to realize continuous tracking of the abnormal mobile object and obtain the real-time position of the abnormal mobile object; Wherein, the second calculation module includes: The instruction sending unit is used to send an early warning execution instruction to the designated drone, so that the drone calling device formulates a corresponding navigation path according to the real-time position of the abnormal moving object and the campus road planning information in the three-dimensional campus building simulation map, and goes to the real-time position according to the navigation path. After reaching the real-time position, the moving target is captured by the smart camera under the drone calling device, and gradually approaches the moving object. When the distance detection unit on the smart camera detects that the distance between the moving object and the drone calling device is within a preset calling distance, a communication connection request is fed back to the campus detection platform; The communication connection unit is used to accept the communication connection request and then establish a communication connection between the security call station and the drone calling device, so that the security personnel located at the security call station can warn the person with abnormal activities at night through the speaker broadcast of the drone calling device.

Images