A method and system for real-time video tracking of abnormal events

Abstract

The present invention discloses a method and system for real-time video tracking of abnormal events. The method includes: acquiring a real-time video and determining abnormal events in the real-time video; if it is determined that there are abnormal events, associating the special locations where the abnormal events occur in the real-time video with locations, and performing real-time tracking by using special persons and / or special key objects as tag classes; when there is at least one special person or special key object, taking the special locations where the abnormal events occur in the real-time video as tracing points, and performing tracing by using special persons and / or special key objects as screening classes. The present invention can timely identify abnormal events in real-time video monitoring, accurately and timely track according to the abnormal elements involved in the abnormal events, improve the processing efficiency of abnormal events, and at the same time perform multiple tracing and tracking in combination with the tracing rules of the relevant locations and abnormal elements of the abnormal events to ensure that there are no other potential hazards, and at the same time reduce the workload and improve the processing efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic tracking, and particularly to a method and system for real-time video tracking of abnormal events. Background Art

[0002] Currently, monitoring systems are becoming more and more popular. In most large-scale scenarios, there are a large number of monitoring devices, and unexpected situations are inevitable in various places. The existing monitoring modes mostly directly transmit or record and store several videos in real time, and someone needs to watch, query, track, and trace multiple videos online in real time, which requires a relatively high comprehensive quality of people and also requires a large amount of data and computing power that the system can carry, and is also cumbersome.

[0003] After a sudden and abnormal event occurs, the targets involved will move, and the direction of their movement is uncertain. If not tracked in time, their real-time whereabouts cannot be obtained in time, resulting in more manpower and material resources being consumed in later investigations. In addition, the monitoring personnel need to perform a large number of switching operations to lock the event. When the event occurs, they cannot accurately track the movement of the target object. After the event, due to the inability to accurately determine the time of the event, a large number of manual retrievals and reviews of video recordings are required, which is time-consuming and laborious and has low efficiency. Therefore, a method that can accurately and quickly perform real-time tracking and reporting based on real-time video frame information and a method for locking abnormal events from a large number of monitoring recordings at the same time are needed. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed. Therefore, the present invention provides a method for real-time video tracking of abnormal events to solve the problems that when an abnormal event occurs, the movement of the target object cannot be accurately tracked, and at the same time, tracing cannot be carried out. After the event, due to the inability to accurately determine the time of the event, a large number of manual retrievals and reviews of video recordings are required, resulting in low efficiency.

[0005] To solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides a method for real-time video tracking of abnormal events, including:

[0007] Obtain real-time video and judge abnormal events in the real-time video;

[0008] If it is judged that there are abnormal events, associate the special positions where the abnormal events appear in the real-time video with locations, and perform real-time tracking by using special persons and / or special key objects as tags;

[0009] When there is at least one special person or special key object, use the special position where the abnormal event appears in the real-time video as the tracing point, and perform tracing by using special persons and / or special key objects as screening criteria.

[0010] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: the judgment includes identifying or receiving abnormal elements, and the abnormal elements include special persons, special key objects or special data.

[0011] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: the judgment further includes setting different recognition algorithms for each real-time video according to the location and running time of its camera device.

[0012] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: the location association for special locations includes

[0013] invoking the first set of real-time videos of relevant locations connected to the special location, and screening with special persons and / or special key objects as tag categories;

[0014] If there are special persons and / or special key objects in the first set of real-time videos, lock the location where the corresponding real-time video is located as the first relevant location.

[0015] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: the location association for special locations further includes

[0016] invoking the first set of real-time videos of relevant locations connected to the special location, and screening with special persons and / or special key objects as tag categories;

[0017] If there are no special persons and / or special key objects in the first set of real-time videos, take the relevant location as the association point to perform the association of the second relevant location, form the second set of real-time videos, and screen with special persons and / or special key objects as tag categories.

[0018] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: if there are no special persons and / or special key objects in the second set of real-time videos, loop to associate relevant locations until the location where the special persons and / or special key objects are located is screened out.

[0019] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: when there is at least one special person or special key object, take the special location where the abnormal event appears in the real-time video as the tracing point, and perform tracing with special persons and / or special key objects as the screening categories, including

[0020] For the special location, with special persons and / or special key objects as the screening categories, trace the preset time range from the time point of the abnormal event, and combine with the intelligent judgment of the relevant locations of the special location to lock the target node of the abnormal event.

[0021] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: if there are other special persons and / or other special key objects during the tracing process, then use the other special persons and / or other special key objects as label classes for real-time tracking in the real-time video.

[0022] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: it further includes, while using the other special persons and / or other special key objects as label classes for real-time tracking in the real-time video, continuing to perform video tracing of relevant positions of special locations to lock the target node of the abnormal event.

[0023] As a preferred solution of the method for real-time video tracking of abnormal events according to the present invention, wherein: the target node of the abnormal event includes the node where the special person and / or special key object first appears.

[0024] In a second aspect, the present invention provides a system for real-time video tracking of abnormal events, including:

[0025] An acquisition module, configured to acquire real-time video and judge abnormal events in the real-time video;

[0026] A tracking module, configured to, if it is judged that an abnormal event exists, perform location association on the special location where the abnormal event appears in the real-time video, and perform real-time tracking by combining special persons and / or special key objects as label classes;

[0027] A tracing module, configured to, when there is at least one special person or special key object, use the special location where the abnormal event appears in the real-time video as a tracing point, and perform tracing by combining special persons and / or special key objects as screening classes.

[0028] In a third aspect, the present invention provides a computing device, including:

[0029] A memory and a processor;

[0030] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, and when the computer-executable instructions are executed by the processor, the steps of the method for real-time video tracking of abnormal events are implemented.

[0031] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the method for real-time video tracking of abnormal events are implemented.

[0032] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention can accurately identify abnormal events in real-time video monitoring in a timely manner, accurately and timely track according to the abnormal elements involved in the abnormal events, thereby improving the processing efficiency of abnormal events. At the same time, it can combine the relevant retrieval and traceability rules of the relevant positions and abnormal elements of the abnormal events for multiple traceability and tracking to ensure that there are no other hidden dangers in the abnormal events, while reducing the workload of staff and related systems and improving the processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings. Among them:

[0034] Figure 1 It is a schematic flowchart of the method for real-time video tracking of abnormal events according to the first embodiment of the present invention;

[0035] Figure 2 It is a schematic flowchart of the implementation method of the first scenario in the method for real-time video tracking of abnormal events according to the second embodiment of the present invention;

[0036] Figure 3 It is a schematic flowchart of the implementation method of the second scenario in the method for real-time video tracking of abnormal events according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0038] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar promotions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0039] Second, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not all refer to the same embodiment, nor is it an individual or alternative embodiment that is mutually exclusive with other embodiments.

[0040] The present invention will be described in detail in conjunction with the schematic diagrams. When describing the embodiments of the present invention in detail, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally in a non-general proportion, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0041] At the same time, in the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper, lower, inner, and outer" is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation of the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0042] Unless otherwise clearly defined and limited in the present invention, the terms "installed, connected, and connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can also be a mechanical connection, an electrical connection, or a direct connection, and can also be indirectly connected through an intermediate medium, or can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0043] Embodiment 1

[0044] Refer to Figures 1-3 , which is an embodiment of the present invention, and provides a method for real-time video tracking of abnormal events, including:

[0045] Step 101: Obtain a real-time video and judge the abnormal events in the real-time video;

[0046] It should be noted that the real-time video acquisition method can be obtained from all camera devices in the abnormal event occurrence scenario and the camera devices in other associated scenarios.

[0047] In an implementable manner, the judgment includes identifying or receiving abnormal elements, and the abnormal elements include special persons, special key objects, or special data.

[0048] Among them, the special persons can be some people with iconic clothing or iconic accessories such as doctors, students, teachers, etc., or the persons marked when abnormal events occur in the camera device; the special key objects can be ambulances, stretchers or harmful items; the special data can be abnormal data obtained by object sensors, physiological sensors and other types of sensors.

[0049] In an implementable manner, the judgment further includes setting different recognition algorithms for each real-time video according to the position and running time of the camera device where it is located.

[0050] It should be noted that different recognition algorithms can be set or retrieved in each camera device according to the running time of the relevant activities involved in the location, and the algorithms are run according to different time periods, so that the system operation amount can be reduced and resource waste can be reduced.

[0051] Among them, the recognition algorithms include, through human action recognition, such as: falling to the ground, standing or gathering, etc.; special key object recognition; color recognition and so on.

[0052] The judgment can be, for example: in the school scenario, during the class time period, the corridor camera device sets or retrieves the algorithms for gathering and falling to the ground. If it is recognized that there is such a situation, the next step and early warning are carried out.

[0053] Step 102: If it is judged that an abnormal event exists, the special location where the abnormal event appears in the real-time video is associated with the location, and the special persons and / or special key objects are used as tag classes for real-time tracking;

[0054] In an implementable manner, the location association of the special location includes

[0055] Calling the first set of real-time videos of the relevant locations connected to the special location, and screening with the special persons and / or special key objects as tag classes;

[0056] If there are special persons and / or special key objects in the first set of real-time videos, lock the location where the corresponding real-time video is located as the first relevant location.

[0057] It should be noted that the first set of real-time videos is the set of each real-time video of the location where the abnormal event occurs and the relevant locations; screening with the special persons and / or special key objects as tag classes to screen out all the line videos with special persons and / or special key objects.

[0058] Screening. For example, in a school scenario, during class time, after an abnormal event occurs and a special person moves, the first real-time video set of the corridor and its connected relevant locations is called, and screening is carried out with the marked special person as the label. If the special person appears in several videos in the first real-time video set, the location where the video camera device is located is locked, that is: the first relevant location, and the first relevant location is at least one location.

[0059] In an implementable manner, the location association of special locations further includes

[0060] Calling the first real-time video set of the relevant locations connected to the special location, and screening with the special person and / or special key object as the label class;

[0061] If there is no special person and / or special key object in the first real-time video set, the relevant location is used as the association point to perform the association of the second relevant location, form the second real-time video set, and screen with the special person and / or special key object as the label class.

[0062] It should be noted that the second real-time video set is the collection of each real-time video of the relevant locations of the first relevant location of the location where the abnormal event occurs; screening is carried out with the special person and / or special key object as the label class to screen out all the line videos with the special person and / or special key object.

[0063] In an implementable manner, if there is no special person and / or special key object in the second real-time video set, the relevant locations are cyclically associated until the location where the special person and / or special key object is located is screened out.

[0064] Step 103: When there is at least one special person or special key object, using the special location where the abnormal event appears in the real-time video as the tracing point, and tracing in combination with the special person and / or special key object as the screening category.

[0065] In an implementable manner, for the special location, with the special person and / or special key object as the screening category, tracing the preset time range based on the time point of the abnormal event, and combining with the intelligent judgment of the relevant locations of the special location to lock the target node of the abnormal event.

[0066] In an implementable manner, the intelligent judgment further includes

[0067] If there is an abnormal event in the relevant location of the special location, associate the relevant location of the special location judged in the previous link and perform the next round of judgment.

[0068] For example: when there is only one abnormal event that the user needs to trace, if there is such an abnormal event in the relevant location of the special location, associate the video recording of the relevant location of the special location, and thus perform the next round of judgment.

[0069] Specifically, the relevant positions associated with this special position can be expressed as {position N - 1, position N - 2,..., position N - M, (M < N)}, where M is the number of relevant positions.

[0070] When the abnormal events that the user needs to trace are multiple simultaneous events, if there are abnormal events in the relevant positions of the special positions involved during the same period, then associate the videos of the relevant positions of the special positions involved, so as to conduct the next round of research and judgment.

[0071] If there is an abnormal event, trace the preset time range based on the time point of the abnormal event, and use special persons and / or special key items as the screening categories. Combine the relevant positions of the special positions in the previous link of intelligent research and judgment to lock the video information of the target node of the abnormal event.

[0072] Among them, the preset time range can be reset multiple times.

[0073] For example, when presetting the time for a special time point, when no special persons and / or special key items are recognized in the videos within this preset time, reset the time, and the value of the newly preset time is greater than the previous preset time and less than the time span from this abnormal event time point to the initial position of the video.

[0074] In an implementable manner, if there are other special persons and / or other special key items during the tracing process, then use other special persons and / or other special key items as the tagging categories to conduct real-time tracking in the real-time video.

[0075] It should be noted that the tracking method for other special persons and / or other special key items is the same as the above real-time video tracking mode.

[0076] In an implementable manner, while conducting real-time tracking of other special persons and / or other special key items as the tagging categories in the real-time video, continue to trace the videos of the relevant positions of the special positions to lock the target node of the abnormal event.

[0077] In an implementable manner, the target node of the abnormal event includes the node where the special person and / or the special key item first appear.

[0078] In an implementable manner, the videos during the tracing process can be frame-extracted, and the corresponding frame-extraction operation can be performed according to user requirements or other requirements.

[0079] For example, extract the key frames among the twenty-five frames in one second, and combine the extracted key frames into a new video file every twenty-five frames in sequence to achieve lossless speed-up playback of the new video.

[0080] In an implementable manner, during the process of retrospective or real-time video tracking, manual judgment is also included, which is carried out after each round of AI judgment or after an abnormal element is detected in the real-time video and a warning is issued. Among them, after manual judgment confirms the existence of an abnormal event, screenshots are saved, and the AI extracts the video event data; manual annotation and review opinions are provided.

[0081] The above is a schematic solution of a method for real-time video tracking of abnormal events in this embodiment. It should be noted that the technical solution of the system for real-time video tracking of abnormal events belongs to the same concept as the technical solution of the above-mentioned method for real-time video tracking of abnormal events. For the details not described in detail in the technical solution of the system for real-time video tracking of abnormal events in this embodiment, reference can be made to the description of the technical solution of the above-mentioned method for real-time video tracking of abnormal events.

[0082] The system for real-time video tracking of abnormal events in this embodiment includes:

[0083] An acquisition module, configured to acquire real-time video and judge abnormal events in the real-time video;

[0084] A tracking module, configured to, if an abnormal event is judged to exist, perform location association on the special location where the abnormal event appears in the real-time video, and perform real-time tracking with special persons and / or special key objects as tag classes;

[0085] A retrospective module, configured to, when there is at least one special person or special key object, use the special location where the abnormal event appears in the real-time video as a retrospective point, and perform retrospective with special persons and / or special key objects as screening classes.

[0086] This embodiment also provides a computing device, applicable to the situation of real-time video tracking of abnormal events, including:

[0087] A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for real-time video tracking of abnormal events as proposed in the above embodiment.

[0088] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the method for real-time video tracking of abnormal events as proposed in the above embodiment.

[0089] The storage medium proposed in this embodiment and the method for real-time video tracking of abnormal events proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0090] From the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general-purpose hardware. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk, or optical disc of a computer, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0091] Embodiment 2

[0092] Referring to Figures 1-3 , an embodiment of the present invention provides a method for real-time video tracking of abnormal events, taking a factory scenario as an example.

[0093] First scenario: When a physical conflict occurs among workers in the ninth workshop during the working hours in the factory, the recognition algorithm in the camera device of the ninth workshop detects actions such as someone falling to the ground or gathering during this time period, determines that an abnormal event exists, marks the special person and the special key items that may be carried at the time of the conflict, and associates the relevant locations in the ninth workshop. Real-time tracking is performed using the special person and / or special key items as tag classes.

[0094] The relevant locations in the ninth workshop are the corridor at the entrance of the ninth workshop, the staircase, etc. The real-time videos of these locations are called, that is, the first video set; screening is performed using the special person and / or special key items as tag classes to determine the real-time locations of the special person and / or special key items, that is, the first relevant locations.

[0095] If there are no special person and / or special key items in the first real-time video set, then the corridor at the entrance of the ninth workshop and the staircase are used as the associated points to perform the association of the second relevant locations, forming the second real-time video set, and continue to screen using the special person and / or special key items as tag classes. If there are no special person and / or special key items in the second real-time video set, then the relevant locations are cyclically associated until the locations where the special person and / or special key items are located are screened out.

[0096] Second scenario: During normal operation, if the recognition algorithm in the camera device at the factory gate detects the presence of Person A wearing a factory uniform during this time period, while performing the above tracking, traceability is carried out with this special person A as the screening category. Set half an hour as the traceability time range, and combine AI to analyze relevant positions of the gate, such as aisles, roads, parking lots, etc., until the target node of the abnormal event is locked, that is, the node where the special person A first appeared.

[0097] If, during the tracing process, video analysis finds that this special person A has other associated special person B and / or other special key item C, then real-time tracking is carried out in the real-time video with other special person B and / or other special key item C as the tagging category. While carrying out real-time tracking in the real-time video with other special person B and / or other special key item C as the tagging category, continue to carry out video tracing of relevant positions of special locations, and lock the target node of the abnormal event, that is, the node where the special person A, other special person B and / or other special key item C first appeared.

[0098] During the real-time tracking process, an alarm can be issued when an abnormal event is discovered, and manual analysis is carried out. During the tracing process, screenshots and frame extraction compression can be performed on relevant videos, and manual analysis is carried out after intelligent analysis to further confirm the results.

[0099] It should be noted that this method can also be applied to large-scale places such as hospitals, factories, nursing homes, detention houses, detention centers, drug rehabilitation centers, prisons, grain depots, oil depots, and storage yards.

[0100] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for real-time video tracking of abnormal events, characterized in that, Including: Obtain a live video and judge abnormal events in the live video; If it is judged that there are abnormal events, associate the special location where the abnormal events occur in the live video with a location, and conduct real-time tracking with special persons and / or special key objects as tag classes; The location association of the special location includes: Call the first set of live videos of related locations connected to the special location, and screen with special persons and / or special key objects as tag classes; If there are special persons and / or special key objects in the first set of live videos, lock the location where the corresponding live video is located as the first related location; The location association of the special location also includes: Call the first set of live videos of related locations connected to the special location, and screen with special persons and / or special key objects as tag classes; If there are no special persons and / or special key objects in the first set of live videos, use the related location as an association point to conduct the association of the second related location, form the second set of live videos, and screen with special persons and / or special key objects as tag classes; When there is at least one special person or special key object, use the special location where the abnormal event occurs in the live video as a tracing point, and conduct tracing with special persons and / or special key objects as screening classes; If there are other special persons and / or other special key objects during the tracing process, conduct real-time tracking in the live video with other special persons and / or other special key objects as tag classes; While conducting real-time tracking in the live video with other special persons and / or other special key objects as tag classes, continue to conduct video tracing of related locations of the special location to lock the abnormal event target node; The judgment also includes that different recognition algorithms are set for each live video according to the location and running time of its camera device.

2. The method for real-time video tracking of abnormal events according to claim 1, wherein The judgment includes identifying or receiving abnormal elements, and the abnormal elements include special persons, special key objects or special data.

3. The method for real-time video tracking abnormal events according to claim 2, characterized in that, If there are no special persons and / or special key objects in the second set of live videos, circularly associate related locations until the location where the special persons and / or special key objects are located is screened out.

4. The method for real-time video tracking abnormal events according to claim 3, characterized in that, When there is at least one special person or special key object, use the special location where the abnormal event occurs in the live video as a tracing point, and conduct tracing with special persons and / or special key objects as screening classes, including: For the special location, use special persons and / or special key objects as screening classes, trace the preset time range based on the time point of the abnormal event, and combine intelligent judgment of related locations of the special location to lock the abnormal event target node.

5. The method for real-time video tracking abnormal events according to claim 1 or 4, characterized in that, The abnormal event target node includes the node where the special person and / or special key object first appears.

6. A real-time video tracking abnormal event system, applying the method as described in claim 1, characterized in that, Including: An acquisition module for obtaining a live video and judging abnormal events in the live video; A tracking module for, if it is judged that there are abnormal events, associating the special location where the abnormal events occur in the live video with a location, and conducting real-time tracking with special persons and / or special key objects as tag classes; A tracing module, configured to, when there is at least one special person or special key item, use the special position where an abnormal event occurs in the live video as a tracing point, and combine the special person and / or special key item to perform tracing for a screening category.

7. An electronic device, comprising: a memory and a processor; The memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method for tracking abnormal events in a live video according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for tracking abnormal events in a live video according to any one of claims 1 to 5.

Images