Data early warning method fusing multiple devices and multiple terminals

Abstract

The invention belongs to the technical field of city management, and provides a multi-device and multi-terminal integrated data early warning method, which comprises the following steps: receiving first video data shot by a first camera of a city monitoring system, and analyzing whether preset abnormal data exists in the first video data to obtain a first abnormal analysis probability; determining a first number and a second number according to the first anomaly analysis probability, and screening the first number of second cameras belonging to the city monitoring system and the second number of third cameras not belonging to the city monitoring system in a preset area where a single camera shooting the first video data is located; based on the first video data, second video data shot by each second camera and third video data shot by each third camera, carrying out secondary analysis on preset abnormal data to obtain a second abnormal analysis probability; and if the second anomaly analysis probability is higher than an anomaly threshold value, outputting alarm information. According to the invention, more accurate anomaly analysis is realized.

Description

Technical Field

[0001] The present invention relates to the field of urban management technology, and in particular to a data early warning method integrating multiple devices and multiple terminals. Background Art

[0002] In order to achieve smarter and more efficient urban management, many cities have deployed data early warning systems based on surveillance videos. By analyzing the behavior of objects in the surveillance videos and judging anomalies, alarms are issued when anomalies occur, so that relevant personnel can deal with them in a timely manner.

[0003] However, the existing data early warning system is only based on the urban monitoring system deployed by the city management department, but the urban monitoring system often cannot fully cover the city, resulting in many abnormal objects and abnormal behaviors not being monitored. Moreover, the existing data early warning system generally only analyzes abnormal objects and abnormal behaviors based on a single surveillance video source, resulting in insufficient accuracy of the analysis results and prone to false alarms.

[0004] It can be seen that how to improve the comprehensiveness and accuracy of the data early warning system is a technical problem that needs to be solved urgently. Summary of the invention

[0005] In this regard, the present invention provides a data warning method, electronic device, computer storage medium and computer program product integrating multiple devices and multiple terminals to solve at least one of the above-mentioned technical problems.

[0006] The present invention provides a data early warning method integrating multiple devices and multiple terminals, comprising the following method steps: receiving first video data shot by a first camera of a city monitoring system, analyzing whether there is preset abnormal data in the first video data, and obtaining a first abnormal analysis probability; wherein the abnormal data refers to abnormal objects and abnormal behaviors; determining a first number and a second number according to the first abnormal analysis probability, screening the first number of second cameras belonging to the city monitoring system and screening the second number of third cameras not belonging to the city monitoring system in a preset area where a single camera shooting the first video data is located; based on the first video data, the second video data shot by each of the second cameras, and the third video data shot by each of the third cameras, performing a secondary analysis on the preset abnormal data to obtain a second abnormal analysis probability; if the second abnormal analysis probability is higher than an abnormal threshold, outputting an alarm message.

[0007] As an example, the receiving of the first video data captured by the first camera of the city monitoring system includes: determining the selection information of the management personnel on a certain monitoring area, and determining a number of cameras belonging to the city monitoring system located in the monitoring area according to the selection information; determining a number of monitoring sub-areas according to the monitoring type corresponding to the selection information, and determining the camera with the largest coverage angle located in each of the monitoring sub-areas as the first camera; and receiving the first video data captured by each of the first cameras.

[0008] As an example, the analysis of whether there is preset abnormal data in the first video data to obtain a first abnormal analysis probability includes: filtering out the corresponding abnormal object analysis model and abnormal behavior analysis model from a database according to the monitoring type, and inputting the first video data into the abnormal object analysis model and the abnormal behavior analysis model, respectively, to obtain a first abnormal object probability and a first abnormal behavior probability, respectively; the first abnormal object probability and the first abnormal behavior probability constitute the first abnormal analysis probability.

[0009] As an example, the first number and the second number determined according to the first abnormality analysis probability include: calculating the third number of cameras used for collaborative analysis according to the first abnormality analysis probability; and also extracting the motion intensity of preset abnormal data from the first video data, and determining an adjustment value according to the motion intensity, wherein the adjustment value is greater than or equal to 1; calculating a fourth number according to the third number and the adjustment value; obtaining the overall monitoring coverage of each camera belonging to the urban monitoring system in the monitoring area, determining a first percentage according to the overall monitoring coverage, and further calculating a second percentage = 1-first percentage; multiplying the fourth number by the first percentage to obtain the first number, and multiplying the fourth number by the second percentage to obtain the second number.

[0010] As an example, calculating the third number of cameras for collaborative analysis according to the first abnormality analysis probability includes: ; In the formula, is the third number of cameras used for collaborative analysis, is the first abnormal analysis probability, is the average value of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data, The median value of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data; is a preset number related to the number of cameras used for collaborative analysis; is the weight value, .

[0011] As an example, the method also includes: obtaining the carrier information of each of the third cameras that are screened out and do not belong to the city monitoring system, performing a handshake connection with the corresponding carrier according to the carrier information, and after the handshake connection is successful, retrieving the third video data taken by the corresponding third camera.

[0012] As an example, the abnormal object analysis model and the abnormal behavior analysis model are constructed based on a generative adversarial network.

[0013] The present invention also provides a computer storage medium storing a computer program executable by a processor to implement the method steps implemented by the processing device as described in any of the above items.

[0014] The present invention also provides a computer program product, which includes a computer program executable by a processor to implement the method steps implemented by the processing device as described in any one of the above items.

[0015] The beneficial effect of the present invention is that the scheme of the present invention can achieve more accurate anomaly analysis effect by fusing and analyzing video surveillance data from multiple devices and multiple terminals, compared with the background technology of performing anomaly analysis based only on video data captured by a single camera of the urban monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be 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 paying creative work.

[0017] Figure 1 It is a flow chart of a data early warning method integrating multiple devices and multiple terminals disclosed in an embodiment of the present invention.

[0018] Figure 2 It is a schematic diagram of a process of receiving first video data shot by a first camera of a city monitoring system disclosed in an embodiment of the present invention.

[0019] Figure 3 It is a flowchart of determining a first number and a second number according to a first abnormality analysis probability disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0021] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0022] The existing data early warning system is only based on the urban monitoring system deployed by the city management department. However, the urban monitoring system often cannot fully cover the city, resulting in many abnormal objects and abnormal behaviors not being monitored. Moreover, the existing data early warning system generally only analyzes abnormal objects and abnormal behaviors based on a single surveillance video source, resulting in insufficient accuracy of the analysis results and prone to false alarms. It can be seen that how to improve the comprehensiveness and accuracy of the data early warning system is a technical problem that needs to be solved urgently.

[0023] like Figure 1 As shown, an embodiment of the present invention discloses a data early warning method integrating multiple devices and multiple terminals, including the following method steps: receiving first video data captured by a first camera of a city monitoring system, analyzing whether there is preset abnormal data in the first video data, and obtaining a first abnormal analysis probability; wherein the abnormal data refers to abnormal objects and abnormal behaviors; determining a first number and a second number according to the first abnormal analysis probability, screening the first number of second cameras belonging to the city monitoring system and screening the second number of third cameras not belonging to the city monitoring system in a preset area where a single camera that captured the first video data is located; based on the first video data, the second video data captured by each of the second cameras, and the third video data captured by each of the third cameras, performing a secondary analysis on the preset abnormal data to obtain a second abnormal analysis probability; if the second abnormal analysis probability is higher than the abnormal threshold, outputting an alarm message.

[0024] Compared with the prior art, the present invention determines whether there are abnormal objects and abnormal behaviors in the corresponding area based on the video data captured by the camera of the urban monitoring system and the video data captured by the camera of the non-urban monitoring system, that is, based on the fusion analysis of the monitoring data of multiple devices and multiple terminals to achieve the purpose of accurate monitoring of abnormalities. Specifically as follows: First, receive the first video data captured by the first camera of the urban monitoring system. According to the management requirements of the city, the urban management department will distribute many cameras in the city. These cameras constitute the urban monitoring system. The cameras are, for example, ball cameras, gun cameras, etc. According to the preset analysis method, analyze whether there is preset abnormal data in the first video data to obtain the first abnormal analysis probability. Among them, abnormal data refers to abnormal objects and abnormal behaviors, that is, the presence of abnormal objects and abnormal behaviors in the first video data is detected and analyzed, and the corresponding probability is output. Abnormal objects are, for example, wanted persons, monitored persons, etc., and abnormal behaviors are, for example, illegal stalls, graffiti, etc.

[0025] Then, the first number and the second number are determined based on the first abnormal analysis probability, and thus, some cameras within the preset area (e.g., 300 meters) where the single camera that shoots the first video data is located are screened out, including the first number of second cameras belonging to the city monitoring system and the second number of third cameras that do not belong to the city monitoring system. Among them, the third cameras that do not belong to the city monitoring system refer to cameras installed in street shops and shopping malls, and vehicle-mounted cameras installed in various vehicles with network transmission functions.

[0026] Finally, based on the aforementioned first video data and the video data captured by the second camera and the third camera, the preset abnormal data obtained by the preliminary determination is analyzed again to obtain the second abnormal analysis probability. If the second abnormal analysis probability is higher than the abnormal threshold, it means that there is indeed an abnormality. At this time, an alarm message can be output to the relevant personnel or terminal device to remind them to deal with it.

[0027] It can be seen that the solution of the present invention can achieve a more accurate anomaly analysis effect by fusing and analyzing video surveillance data from multiple devices and multiple terminals, compared to the background technology of performing anomaly analysis based only on video data captured by a single camera of the urban monitoring system.

[0028] As an example, Figure 2As shown, the receiving of the first video data captured by the first camera of the city monitoring system includes: determining the selection information of the management personnel on a certain monitoring area, and determining a number of cameras belonging to the city monitoring system located in the monitoring area according to the selection information; determining a number of monitoring sub-areas according to the monitoring type corresponding to the selection information, and determining the camera with the largest coverage angle located in each of the monitoring sub-areas as the first camera; and receiving the first video data captured by each of the first cameras.

[0029] In this embodiment, the manager of the monitoring center can select an area, i.e., the monitoring area, on the monitoring map based on the actual monitoring needs. In response to the selection of the monitoring area, several cameras belonging to the city monitoring system located in the monitoring area can be determined. At the same time, the manager also selects or inputs the monitoring type of this time before or after selecting the monitoring area, such as fugitive tracking, anti-theft monitoring, city appearance management, etc., and several different monitoring sub-areas can be determined according to different monitoring types. For example, when the monitoring type is fugitive tracking, the road area can be used as the monitoring sub-area, when the monitoring type is anti-theft monitoring, the street shop area can be used as the monitoring sub-area, and when the monitoring type is city appearance management, the pedestrian area between the road and the building, the alley, etc. can be used as the monitoring sub-area. Finally, the camera with the largest coverage angle in each monitoring sub-area is determined as the first camera. These first cameras can achieve the maximum coverage of the monitoring sub-area, which is conducive to reducing the probability of missing abnormal objects and abnormal behaviors.

[0030] As an example, the analysis of whether there is preset abnormal data in the first video data to obtain a first abnormal analysis probability includes: filtering out the corresponding abnormal object analysis model and abnormal behavior analysis model from a database according to the monitoring type, and inputting the first video data into the abnormal object analysis model and the abnormal behavior analysis model, respectively, to obtain a first abnormal object probability and a first abnormal behavior probability, respectively; the first abnormal object probability and the first abnormal behavior probability constitute the first abnormal analysis probability.

[0031] In this embodiment, the database is preset with a variety of abnormal object analysis models and abnormal behavior analysis models, which are pre-built and trained for different types of monitoring needs, and can achieve targeted analysis of abnormal objects and abnormal behaviors in fugitive tracking, anti-theft monitoring, city appearance management, etc. Compared with the traditional composite model, the classification preset model in the present invention has a smaller model size and a faster analysis rate, which can meet the timeliness requirements in urban monitoring management.

[0032] It is worth pointing out that the abnormal object analysis model and the abnormal behavior analysis model are preferably constructed based on generative adversarial networks. Generative Adversarial Networks (GANs) is a deep learning model that has been widely used in the field of artificial intelligence since it was proposed by Ian Goodfellow and others in 2014. It contains two neural networks: the generator and the discriminator, which compete with each other to achieve the goal of generating realistic data through adversarial training. The goal of the generator is to generate artificial samples that are as close to the real data distribution as possible, while the goal of the discriminator is to determine whether the input data is real or generated by the generator.

[0033] The core of GANs is that its generator can learn the distribution of real data and generate realistic samples. This feature enables GANs to generate samples that are highly similar to normal data in the anomaly detection of the present invention, thereby helping to more accurately identify abnormal objects and behaviors that deviate from normal data. Moreover, since GANs can use unsupervised learning in anomaly detection without the need for labeled data, this allows GANs to still perform effective anomaly detection in scenarios where there is a lack of labeled data, greatly reducing the difficulty of model construction.

[0034] It should be noted that the method adopted for the secondary analysis is the same as that in this embodiment. The main difference is that the input of the abnormal object analysis model and the abnormal behavior analysis model becomes the first video data, the second video data and the third video data, which will not be described in detail.

[0035] As an example, Figure 3 As shown, the first number and the second number determined according to the first abnormality analysis probability include: calculating the third number of cameras used for collaborative analysis according to the first abnormality analysis probability; and also extracting the motion intensity of the preset abnormal data from the first video data, and determining an adjustment value according to the motion intensity, wherein the adjustment value is greater than or equal to 1; calculating a fourth number according to the third number and the adjustment value; obtaining the overall monitoring coverage of each camera belonging to the urban monitoring system in the monitoring area, determining a first percentage according to the overall monitoring coverage, and further calculating a second percentage = 1-first percentage; multiplying the fourth number by the first percentage to obtain the first number, and multiplying the fourth number by the second percentage to obtain the second number.

[0036] In this embodiment, the present invention performs secondary abnormal analysis based on the cameras of the city monitoring system and the cameras of the non-city monitoring system, and at this time, it is necessary to determine the number of the two cameras. Specifically, first, the total number of cameras used for collaborative analysis, i.e., the third number, is calculated based on the first abnormal analysis probability. Here, the third number is negatively correlated with the first abnormal analysis probability, that is, the higher the first abnormal analysis probability, the higher the abnormality of the abnormal object and abnormal behavior (for example, a person is setting up a stall on the roadside, and some buyers are surrounding the stall, and the behavior has lasted for an hour). At this time, it is not necessary to obtain more video data for secondary analysis, and the third number is set to a smaller value accordingly; and when the first abnormal analysis probability is lower, it indicates that the abnormality of the abnormal object and abnormal behavior is lower (for example, the walking posture of a certain object has a certain similarity with the fugitive being tracked, but the facial information of the object cannot be collected because the object wears a hat, a mask, etc.), at this time, it is necessary to obtain more video data for secondary analysis, and the third number is set to a larger value accordingly.

[0037] At the same time, the motion intensity of the preset abnormal data is also extracted from the first video data. The motion intensity refers to the motion speed involved in the abnormal object and abnormal behavior, such as the abnormal object running fast (escaping), drunk driving vehicles driving fast, driving irregularly in place, etc. The motion intensity here can be evaluated based on preset rules or video semantic understanding models (such as BERT models), and the details are not repeated here. Then, the adjustment value can be determined based on the obtained motion intensity. The adjustment value is a value greater than or equal to 1, which can appropriately increase the third number to the fourth number. Among them, when the determined motion intensity is higher, it indicates that the object involved in the abnormal behavior is less likely to capture key information, or the behavior is more complex. At this time, the adjustment value is set to be larger, otherwise the adjustment value is set to be smaller.

[0038] Next, the overall monitoring coverage of each camera belonging to the city monitoring system in the monitoring area is obtained. The overall monitoring coverage here refers to the ratio of the actual coverage area of ​​the cameras of each city monitoring system in the monitoring area to the area of ​​the monitoring area. Obviously, the higher the overall monitoring coverage, the fewer monitoring blind spots in the monitoring area, and vice versa. Therefore, the present invention is set to determine the first percentage according to the overall monitoring coverage, and multiply the fourth number by the first percentage to obtain the first number of cameras in the city monitoring system that needs to be called; correspondingly, the fourth number is multiplied by the second percentage, that is, (1-first percentage), to obtain the second number of cameras in the non-city monitoring system that needs to be called.

[0039] Therefore, the above scheme of the present invention determines the proportion of the second camera and the third camera to be dispatched according to the size of the overall monitoring coverage of the urban monitoring system in the monitoring area.

[0040] It should be noted that, since the difficulty of accessing cameras in the city monitoring system is significantly lower than that of cameras in shops, shopping malls, and car cameras (there are situations such as refusal to access, vehicles driving out of the monitoring area, etc.), when the overall monitoring coverage is low, more third cameras are prioritized, and this can reserve more time to communicate with the carrier of the third camera, improving the communication success rate; otherwise, more second cameras are prioritized. The comparison example of the overall monitoring coverage and the first percentage and the second percentage is shown in the following table:

[0041] It should be noted that the above comparison example table is only for illustrative purposes and is not intended to limit the comparison relationship between the overall monitoring coverage and the first percentage and the second percentage in the present invention to only the above-mentioned mode.

[0042] As an example, calculating the third number of cameras for collaborative analysis according to the first abnormality analysis probability includes: ; In the formula, is the third number of cameras used for collaborative analysis, is the first abnormal analysis probability, is the average value of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data, The median value of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data; is a preset number related to the number of cameras used for collaborative analysis; is the weight value, .

[0043] In this embodiment, based on the above calculation formula, when calculating the total number of cameras used for collaborative analysis, i.e., the third number, the present invention comprehensively considers the first abnormal analysis probability and the general situation of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data. The general situation is represented by the sum of the average value and the median value of each first abnormal analysis probability, which can basically be regarded as a fixed value. Therefore, the third number is negatively correlated with the first abnormal analysis probability. Since the above content has been analyzed in detail, it will not be repeated here.

[0044] As an example, the method also includes: obtaining the carrier information of each of the third cameras that are screened out and do not belong to the city monitoring system, performing a handshake connection with the corresponding carrier according to the carrier information, and after the handshake connection is successful, retrieving the third video data taken by the corresponding third camera.

[0045] In this embodiment, after determining the third cameras that do not belong to the city monitoring system, the carrier information of these cameras is further obtained. The carrier is, for example, a vehicle, a business, a shopping mall, etc. The corresponding carrier information is the coding information of the vehicle, the business, and the shopping mall. Based on this carrier information, a connection can be established with the corresponding carrier, specifically, a handshake signal is sent to these carriers, and these carriers respond to the handshake signal, thereby achieving a communication connection with these carriers, and then retrieving the third video data shot by the corresponding third camera.

[0046] In addition, for the connection with each second camera belonging to the city monitoring system, if the second camera and the management personnel belong to the same system, they can be directly dispatched; if the second camera and the management personnel belong to different systems, the second video data can also be retrieved in the same way as described above for connecting with the third camera. The details will not be repeated here.

[0047] The present invention also discloses an electronic device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein when the computer program is executed by the processor, the method steps implemented by the processing device described in any of the preceding items are implemented.

[0048] The present invention also discloses a computer storage medium storing a computer program executable by a processor to implement the method steps implemented by the processing device described in any of the above items.

[0049] The present invention also discloses a computer program product, which includes a computer program that can be executed by a processor to implement the method steps implemented by the processing device described in any of the above items.

[0050] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0051] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A data early warning method integrating multiple devices and multiple terminals, characterized in that: The method comprises the following steps: receiving first video data shot by a first camera of a city monitoring system, analyzing whether there is preset abnormal data in the first video data, and obtaining a first abnormal analysis probability; wherein the abnormal data refers to abnormal objects and abnormal behaviors; determining a first number and a second number according to the first abnormal analysis probability, screening the first number of second cameras belonging to the city monitoring system and screening the second number of third cameras not belonging to the city monitoring system in a preset area where a single camera shooting the first video data is located; performing a secondary analysis on the preset abnormal data based on the first video data, the second video data shot by each of the second cameras, and the third video data shot by each of the third cameras, and obtaining a second abnormal analysis probability; if the second abnormal analysis probability is higher than an abnormal threshold, outputting an alarm message.

2. According to claim 1, a data early warning method integrating multiple devices and multiple terminals is characterized in that: Receiving first video data captured by a first camera of a city monitoring system, including: determining selection information of a management personnel on a certain monitoring area, and determining a number of cameras belonging to the city monitoring system and located in the monitoring area according to the selection information; determining a number of monitoring sub-areas according to the monitoring type corresponding to the selection information, and determining the camera with the largest coverage angle located in each of the monitoring sub-areas as the first camera; and receiving the first video data captured by each of the first cameras.

3. According to claim 2, a data early warning method integrating multiple devices and multiple terminals is characterized in that: Analyze whether there is preset abnormal data in the first video data to obtain a first abnormal analysis probability, including: screening out corresponding abnormal object analysis models and abnormal behavior analysis models from a database according to the monitoring type, inputting the first video data into the abnormal object analysis model and the abnormal behavior analysis model respectively, and obtaining a first abnormal object probability and a first abnormal behavior probability respectively; the first abnormal object probability and the first abnormal behavior probability constitute the first abnormal analysis probability.

4. The data early warning method integrating multiple devices and multiple terminals according to claim 3 is characterized in that: Determining a first number and a second number based on the first abnormality analysis probability includes: calculating a third number of cameras for collaborative analysis based on the first abnormality analysis probability; and extracting motion intensity of preset abnormal data from the first video data, and determining an adjustment value based on the motion intensity, wherein the adjustment value is greater than or equal to 1; calculating a fourth number based on the third number and the adjustment value; obtaining the overall monitoring coverage of each camera belonging to the urban monitoring system within the monitoring area, determining a first percentage based on the overall monitoring coverage, and further calculating a second percentage = 1-first percentage; multiplying the fourth number by the first percentage to obtain the first number, and multiplying the fourth number by the second percentage to obtain the second number.

5. The data early warning method integrating multiple devices and multiple terminals according to claim 4 is characterized in that: Calculating a third number of cameras for collaborative analysis according to the first abnormality analysis probability includes: ; In the formula, is the third number of cameras used for collaborative analysis, is the first abnormal analysis probability, is the average value of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data, The median value of each first abnormal analysis probability in the monitoring type corresponding to the preset abnormal data; is a preset number related to the number of cameras used for collaborative analysis; is the weight value, .

6. A data early warning method integrating multiple devices and multiple terminals according to claim 5, characterized in that: The method also includes: obtaining the carrier information of each of the third cameras that are screened out and do not belong to the city monitoring system, performing a handshake connection with the corresponding carrier according to the carrier information, and after the handshake connection is successful, retrieving the third video data shot by the corresponding third camera.

7. The data early warning method integrating multiple devices and multiple terminals according to claim 5 is characterized in that: The abnormal object analysis model and the abnormal behavior analysis model are constructed based on a generative adversarial network.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein when the computer program is executed by the processor, the method steps implemented by the processing device as described in any one of claims 1 to 6 are implemented.

9. A computer storage medium, characterized in that: The computer storage medium stores a computer program that can be executed by a processor to implement the method steps implemented by the processing device as claimed in any one of claims 1 to 6.

10. A computer program product, characterized in that: The computer program product comprises a computer program executable by a processor to implement the method steps implemented by the processing device as claimed in any one of claims 1 to 6.

Images