Task scheduling method and system, electronic device, and storage medium

Abstract

This application provides a task scheduling method and system, electronic device, and storage medium, belonging to the field of task scheduling technology. The method includes: acquiring historical site data for all sites within a scheduling area and historical video recording data for all cameras within each site; determining the pre-allocated task volume for each site under different sampling types based on current scheduling resources, historical site task volume, and historical site event volume; determining the first correlation degree between the corresponding camera and the occurring event based on the historical video task volume and historical event volume; selecting multiple target camera devices from all camera devices within the site based on the first correlation degree of each camera device; and generating target scheduling data for each target camera device in the site based on historical time period data, the pre-allocated task volume corresponding to different sampling types, the multiple target camera devices, and the tasks corresponding to the site. This application can improve the efficiency of task scheduling.

Description

Technical Field

[0001] This application relates to the field of task scheduling technology, and in particular to a task scheduling method and system, electronic device and storage medium. Background Technology

[0002] With the development of IoT technology, camera devices can now be configured with different functions to perform specific detection tasks, meeting the detection needs of various scenarios. For example, in the logistics field, camera devices can be configured to detect illegal throwing, crowd gathering, or unauthorized entry onto conveyor belts, thereby achieving real-time monitoring of the cargo transportation process. Since a large number of camera devices are typically deployed in different scenarios, each device has diverse task types and can employ different sampling methods, such as real-time or near-real-time sampling for crowd gathering detection. Therefore, it is necessary to schedule the tasks of the camera devices to enable the system to allocate resources rationally and ensure that critical tasks are executed with priority.

[0003] Currently, related technologies typically employ edge computing nodes to acquire tasks from all camera devices in different scenarios in real time, and allocate sufficient resources to all tasks of different sampling types during execution, thereby achieving task scheduling for camera devices. However, this approach consumes a large amount of resources, leading to reduced efficiency in task scheduling. Summary of the Invention

[0004] The main objective of this application is to provide a task scheduling method and system, electronic device and storage medium, in order to improve the efficiency of task scheduling.

[0005] To achieve the above objectives, a first aspect of this application proposes a task scheduling method, the method comprising:

[0006] Acquire historical site data for all sites within the scheduling area and historical video data for all cameras within each site. The historical site data includes the historical site task volume allocated to the corresponding site and the historical site event volume obtained based on different sampling types. The historical video data includes the historical video task volume allocated to the corresponding camera device and the historical video event volume obtained based on different sampling types. The occurrence of an event is determined when the task is in a triggered state.

[0007] Obtain the current scheduling resources, and determine the pre-allocated task volume for each site under different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume.

[0008] For each camera device in each venue, the first degree of correlation between the corresponding camera device and the event is determined based on the historical task volume and the historical event volume. Based on the first degree of correlation of each camera device, multiple target camera devices are selected from all camera devices in the venue.

[0009] For each venue, historical time period data of the task in the venue when it is in the triggered state is obtained, and target scheduling data of each target camera in the venue is generated based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices and the task corresponding to the venue.

[0010] In some embodiments, determining the pre-allocated task volume for each site corresponding to different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume includes:

[0011] The total number of pre-allocated tasks for the scheduling region is determined based on the current scheduling resources.

[0012] Obtain the number of camera devices in each venue, and determine the second degree of correlation between the corresponding venue and the event based on the number of devices, the historical task volume of the venue, and the historical event volume of the venue;

[0013] The pre-allocated task quantity for each site is determined based on the total pre-allocated task quantity and the second correlation degree of each site for the different sampling types.

[0014] In some embodiments, determining the second degree of correlation between the corresponding site and the event based on the number of devices, the historical task volume of the site, and the historical event volume of the site includes:

[0015] Obtain the first weighting coefficient corresponding to the number of devices, the second weighting coefficient corresponding to the historical task volume of the site, and the third weighting coefficient corresponding to the historical event volume of the site. The first weighting coefficient and the third weighting coefficient are positive numbers, and the second weighting coefficient is negative number.

[0016] The site score for each site is obtained by weighting the first weighting coefficient, the second weighting coefficient, the third weighting coefficient, the number of devices, the historical task volume of the site, and the historical event volume of the site.

[0017] The site score for each site is determined as the second degree of association for that site.

[0018] In some embodiments, determining the pre-allocated task volume for each site corresponding to the different sampling types based on the total pre-allocated task volume and the second correlation degree of each site includes:

[0019] Obtain the sampling ratio of the scheduling area corresponding to different sampling types;

[0020] The total number of tasks for each site is determined based on the total number of pre-allocated tasks and the second degree of association for each site.

[0021] The pre-allocated task volume for each site is determined based on the total site task volume and the sampling ratio for each site, corresponding to the different sampling types.

[0022] In some embodiments, the step of selecting multiple target camera devices from all camera devices in the site based on a first degree of association for each camera device includes:

[0023] Based on a first preset screening ratio and a first degree of correlation for each camera device, a first set of cameras with the highest correlation is selected from the multiple camera devices. A second set of cameras is then determined based on the first set of cameras and all camera devices corresponding to the venue.

[0024] Select a preset number of camera devices from the second set of devices;

[0025] The target camera device is determined by selecting all camera devices included in the first set of devices and the camera devices selected from the second set of devices.

[0026] In some embodiments, generating target scheduling data for each of the target cameras in the venue based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the venue includes:

[0027] Obtain the task time period data and the working time period data of the site for performing the corresponding task;

[0028] The historical time period data is divided into multiple sub-historical time period data based on the task time period data;

[0029] The number of sub-site historical events corresponding to each of the sub-historical time periods is determined based on the multiple sub-historical time period data and the number of site historical events.

[0030] Based on the second preset filtering ratio and the number of historical events corresponding to each sub-historical time period, the data of multiple sub-historical time periods are filtered to obtain the data of multiple sub-historical time periods with the most sub-site historical events as the first time period set.

[0031] The work period data is divided into multiple sub-work period data based on the task period data;

[0032] Based on the data of the multiple sub-working periods and the historical task volume of the site, determine the historical task volume of the sub-site corresponding to each of the sub-working periods;

[0033] Based on the third preset filtering ratio and the historical task volume of the sub-site corresponding to each sub-work period data, the data of the multiple sub-work periods are filtered by time period to obtain the data of the multiple sub-work periods with the highest historical task volume of the sub-site as the second time period set.

[0034] Multiple target time period data are determined based on the first time period set and the second time period set, and target scheduling data for each target camera in the venue is generated based on the multiple target time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the venue.

[0035] In some embodiments, generating target scheduling data for each target camera in the venue based on the multiple target time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the venue includes:

[0036] The task type of the corresponding target camera device is determined based on the task corresponding to the site and the multiple target camera devices;

[0037] For each target camera device within the site, the target sampling type corresponding to the target camera device is determined according to the task type;

[0038] Based on the multiple target time period data, the target sampling type, and the task type, candidate scheduling data for the corresponding target camera equipment is determined;

[0039] The candidate allocation task quantity corresponding to the target sampling type is determined based on the target time period data of all target camera devices in the site.

[0040] When the number of candidate assigned tasks is less than or equal to the number of pre-assigned tasks for the corresponding sampling type, the candidate scheduling data is determined as the target scheduling data for the corresponding target camera device.

[0041] In some embodiments, after generating target scheduling data for each target camera in the venue based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the venue, the method further includes:

[0042] The corresponding target camera devices are scheduled according to the target scheduling data, and the scheduling time is obtained;

[0043] Obtain the data update cycle, and determine the update time based on the data update cycle and the scheduling time;

[0044] When it is determined that the current time has reached the update time, the target site historical data of the site before the current time is obtained, and the target camera historical data of the target camera device before the current time is obtained;

[0045] The historical data of the target site is updated based on the historical data of the target site, and the historical data of the camera is updated based on the historical data of the target camera.

[0046] In some embodiments, determining the first degree of association between the corresponding camera device and the occurrence of the event based on the camera history task volume and the camera history event volume includes:

[0047] Obtain the device parameters of all camera devices in the venue, including the frame rate and detection frequency of the corresponding camera device;

[0048] The camera score for each camera device in the venue is determined based on the frame rate, the detection frequency, the number of historical camera tasks, and the number of historical camera events.

[0049] The camera score for each camera device is determined as the first degree of association for that camera device.

[0050] To achieve the above objectives, a second aspect of this application provides a task scheduling system, the system comprising:

[0051] The data acquisition module is used to acquire the historical data of all venues within the scheduling area and the historical video data of all camera devices in each venue. The historical data of the venues includes the historical task volume allocated to the corresponding venue and the historical event volume of the venues obtained based on different sampling types. The historical video data includes the historical task volume allocated to the corresponding camera device and the historical event volume of the camera device obtained based on different sampling types. The occurrence of an event is determined when the task is in the triggered state.

[0052] The task volume module is used to obtain the current scheduling resources and determine the pre-allocated task volume for each site in the different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume.

[0053] The camera equipment module is used to determine the first correlation degree between the corresponding camera equipment and the occurrence of the event based on the historical task volume and the historical event volume of the camera equipment in each venue, and to select multiple target camera equipment from all camera equipment in the venue based on the first correlation degree of each camera equipment.

[0054] The scheduling module is used to acquire historical time period data of the tasks in the triggered state for each site, and generate target scheduling data for each target camera in the site based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices and the tasks corresponding to the site.

[0055] To achieve the above objectives, a third aspect of the present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the task scheduling method of the first aspect.

[0056] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the task scheduling method of the first aspect.

[0057] The task scheduling method provided in this application determines the pre-allocated task volume for each site under different sampling types by using historical site data and current scheduling resources for all sites within the scheduling area. It then filters out multiple target camera devices based on historical video data of all camera devices within each site. Finally, it generates target scheduling data for each target camera device in the site based on historical time period data of the task in its triggered state and the corresponding task. Thus, this application embodiment can allocate tasks, corresponding time periods, and sampling types to camera devices based on target scheduling data, avoiding the allocation of large amounts of resources to sites and camera devices with low historical event volumes and reducing the risk of ineffective resource occupation during time periods when tasks are not triggered. In other words, this application embodiment can improve the efficiency of task scheduling. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating an existing task scheduling method;

[0059] Figure 2 This is a flowchart illustrating a task scheduling method provided in an embodiment of this application;

[0060] Figure 3 yes Figure 2 A flowchart illustrating step S202;

[0061] Figure 4 yes Figure 3 A flowchart illustrating step S302;

[0062] Figure 5 yes Figure 3 A flowchart illustrating step S303;

[0063] Figure 6 This is a flowchart illustrating the site selection algorithm provided in an embodiment of this application;

[0064] Figure 7 yes Figure 1 A flowchart of step S203;

[0065] Figure 8 yes Figure 1 Another flowchart of step S203;

[0066] Figure 9 This is a flowchart illustrating the camera selection algorithm provided in an embodiment of this application;

[0067] Figure 10 yes Figure 1 A flowchart of step S204;

[0068] Figure 11 This is a flowchart illustrating the time-period selection algorithm provided in an embodiment of this application;

[0069] Figure 12 yes Figure 1 Another flowchart of step S204;

[0070] Figure 13 yes Figure 1 A flowchart illustrating the steps following step S204;

[0071] Figure 14 This is another flowchart of the task scheduling method provided in the embodiments of this application;

[0072] Figure 15 yes Figure 1 Another flowchart illustrating the steps following step S204;

[0073] Figure 16 This is a schematic diagram of the structure of a task scheduling system provided in an embodiment of this application;

[0074] Figure 17 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0075] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only suitable for explaining the present invention, and should not be construed as limiting the present invention.

[0076] It should be noted that although functional modules are divided in the system diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0077] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0078] First, let's analyze some of the terms used in this application:

[0079] Real-time spot checks refer to the immediate monitoring and inspection of the detection tasks currently being performed by camera equipment. For example, in the logistics field, the detection task of detecting illegal throwing requires real-time monitoring to promptly detect and handle violations.

[0080] Near real-time spot checks refer to monitoring and inspecting the detection tasks performed by camera equipment within a certain time delay. For example, in the security field, for the detection task of stranger intrusion, the system can review and analyze within seconds to minutes after the event occurs to confirm the alarm and take appropriate measures.

[0081] Historical spot checks refer to analyzing data recorded by camera equipment over a past period to determine the camera equipment and time periods required for future tasks. For example, in the field of logistics warehouse monitoring, for the task of detecting goods stacking, the system can analyze data on goods stacking events captured by various camera equipment over the past week, identify the time periods and camera locations where goods stacking occurs frequently, and then conduct focused spot checks on these specific cameras in the corresponding time periods in the future.

[0082] Camera equipment refers to devices that can convert actual images or videos into electronic data for storage, transmission, or processing. For example, a camera can capture images or videos on-site and convert them into electronic data.

[0083] Edge computing nodes are computer devices with data processing and analysis capabilities. In task scheduling, edge computing nodes are typically deployed in specific areas or locations to process and analyze image or video data captured by camera devices within that area, enabling task scheduling for the camera devices.

[0084] The widespread application of task scheduling methods has improved the utilization rate of system resources. For example, in logistics or security fields, task scheduling methods can optimize the task execution order and resource allocation of camera equipment, ensuring that critical detection tasks are prioritized. However, existing task scheduling methods still suffer from low efficiency. For instance, ... Figure 1 As shown, the relevant technology typically deploys a T4 server (i.e., an edge computing node) in area A. The T4 server continuously acquires the detection capabilities (e.g., detection of illegal littering, crowd gathering, etc.) of all cameras in areas B, C, and D of area A, and determines the corresponding sampling type (e.g., real-time, near-real-time, or historical) and the camera device to perform the task based on the tasks within the area. The T4 server then transmits the relevant data to the cloud server S1 for task scheduling. The cloud server S1 can perform various functions such as resource management, camera device management, configuration management, area management, task management, event management, site management, and site operation management. For example, the cloud server S1 can perform large-scale model detection on the data transmitted from the T4 server to determine whether an event has occurred during the task performed by the camera, thereby completing event management.

[0085] However, the methods employed by related technologies involve analyzing and processing data from all camera devices at edge computing nodes and transmitting it to cloud servers for further processing, thus consuming significant computing resources. Furthermore, when camera devices are performing tasks within a site, events may only occur during a few key time periods, or only some camera devices may experience events, leading to a significant amount of resources being inefficiently allocated. When an event occurs, the resources are already over-consumed, hindering timely responses and reducing task scheduling efficiency. Therefore, the main objective of this application is to propose a task scheduling method and system, electronic device, and storage medium to improve task scheduling efficiency.

[0086] The task scheduling method provided in this application relates to the field of task scheduling technology. The task scheduling method provided in this application can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the task scheduling method, but is not limited to the above forms.

[0087] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0088] See Figure 2 This application provides a task scheduling method, which may include, but is not limited to, steps S201 to S204:

[0089] Step S201: Obtain historical site data for all sites within the scheduling area and historical video data for all camera devices within each site;

[0090] Step S202: Obtain the current scheduling resources, and determine the pre-allocated task volume for each site in different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume.

[0091] Step S203: Determine the first degree of correlation between the corresponding camera device and the event based on the number of historical camera tasks and the number of historical camera events; and select multiple target camera devices from all camera devices in the venue based on the first degree of correlation of each camera device.

[0092] Step S204: Obtain historical time period data of tasks in the site when they are in the triggered state, and generate target scheduling data for each target camera in the site based on the historical time period data, the pre-allocated task volume corresponding to different sampling types, multiple target camera devices and the tasks corresponding to the site.

[0093] Steps S201 to S204 of this embodiment determine the pre-allocated task volume for each site under different sampling types by using historical site data and current scheduling resources within the scheduling area. Multiple target cameras are then selected based on historical video data of all cameras within each site. Finally, target scheduling data for each target camera in the site is generated based on historical time period data of the task in the triggered state and the corresponding task for the site. Thus, this embodiment can allocate tasks, corresponding time periods, and sampling types to camera devices based on target scheduling data, avoiding the allocation of large amounts of resources to sites and camera devices with low historical event volumes and reducing the risk of ineffective resource occupation during time periods when tasks are not triggered. In other words, this embodiment improves the efficiency of task scheduling.

[0094] In step S201 of some embodiments, the scheduling area may refer to the area where task scheduling needs to be performed. The scheduling area includes multiple sites, and each site will deploy multiple camera devices. For example, in the logistics field, the scheduling area may refer to a logistics warehouse, the sites may be storage areas, loading and unloading areas, or sorting areas within the logistics warehouse, and the camera devices may refer to cameras deployed in each site. It is understood that the number of sites included in the area and the number of camera devices included in each site can be adjusted according to actual needs.

[0095] Historical site data includes the historical task volume allocated to each site and the historical event volume obtained based on different sampling types. The historical task volume can refer to the total duration of tasks performed at the corresponding site over a past period. For example, in the storage area, the historical task volume could be the total duration of tasks performed in the storage area over the past month; or, it could be the total duration of tasks performed in the storage area over the past week. The historical event volume can refer to the total duration of events that occurred at the corresponding site during task performance under different sampling types over a past period. For example, if the historical task volume is the total duration of illegal throwing detection tasks performed in the storage area over the past week, and these tasks are performed based on three sampling types: real-time sampling, near-real-time sampling, and historical sampling, and the cumulative duration of illegal throwing events detected through real-time sampling is 3 hours, the cumulative duration of events detected through near-real-time sampling is 1 hour, and the cumulative duration of events detected through historical sampling is 1 hour, then the total historical event volume for the storage area is 5 hours. It should be noted that the type of sampling used in the execution of the task and the time range of the historical data of the site can be freely adjusted according to actual needs.

[0096] Historical video data includes the historical task volume assigned to each camera device and the historical event volume obtained based on different sampling types. The historical task volume can refer to the total duration of tasks performed by the corresponding camera device over a past period. For example, it could be the total duration of tasks performed by the corresponding camera device over the past month. The historical event volume can refer to the total duration of events that occurred when the corresponding camera device performed tasks under different sampling types over a past period. For example, if the historical task volume is the total duration of illegal throwing detection tasks performed by a certain camera device over the past week, and these tasks are performed based on three sampling types: real-time sampling, near-real-time sampling, and historical sampling, and the cumulative duration of illegal throwing events detected through real-time sampling is 2 hours, the cumulative duration of events detected through near-real-time sampling is 1 hour, and the cumulative duration of events detected through historical sampling is 1 hour, then the total historical event volume for this camera device is 4 hours. It should be noted that the sampling type used by the camera device to perform tasks and the time range for collecting historical video data must be consistent with the corresponding site.

[0097] It should be noted that an event is determined to have occurred when the task is in the triggered state. For example, if a camera takes 0.5 hours to perform a task to detect illegal throwing, and the camera detects illegal throwing during a specific time period (such as 9:00 AM to 9:30 AM), the task enters the triggered state, thus determining that an event has occurred, and this 0.5 hours is part of the historical event volume corresponding to that camera.

[0098] In step S202 of some embodiments, the currently scheduled resource may refer to the computing resources allocated within the scheduling region for executing the current task. For example, the currently scheduled resource may be a graphics processing unit (GPU) resource allocated within the scheduling region; or, the currently scheduled resource may also be a central processing unit (CPU) resource allocated within the scheduling region, without specific limitations.

[0099] It is understood that the embodiments of this application can use the computing resource management module of the cloud server to calculate and adjust the total scheduling resources in the scheduling area in real time, thereby determining the current scheduling resources required to execute the current task in the scheduling area.

[0100] Pre-allocated task volume refers to the task volume corresponding to different sampling types for a given site, pre-determined based on current scheduling resources, historical site task volume, and historical site event volume. For example, if a scheduling area includes sites B, C, and D, the number of GPUs or CPUs allocated to execute the current task within the scheduling area can be determined first using current scheduling resources. Then, the number of GPUs or CPUs needed for sites B, C, and D can be estimated based on their historical site task volume. Finally, the task volume for different sampling types for sites B, C, and D can be estimated based on their historical site event volume. Thus, the pre-allocated task volume for each site under different sampling types can be determined. It should be noted that the sampling type used to execute the current task must be consistent with the sampling type recorded in the site's historical data.

[0101] Please see Figure 3 In some embodiments, step S202 may include, but is not limited to, steps S301 to S303:

[0102] Step S301: Determine the total number of pre-allocated tasks in the scheduling region based on the current scheduling resources;

[0103] Step S302: Obtain the number of camera devices in each site, and determine the second degree of correlation between the corresponding site and the event based on the number of devices, the historical task volume of the site, and the historical event volume of the site.

[0104] Step S303: Determine the pre-allocated task quantity for each site under different sampling types based on the total pre-allocated task quantity and the second correlation degree of each site.

[0105] In step S301 of some embodiments, the total pre-allocated task volume may refer to the total number of tasks used to execute all tasks within the scheduling region, determined based on the current scheduling resources. For example, if the current scheduling resources include two GPUs and a quad-core CPU, the total pre-allocated task volume is calculated to be 200 hours based on a preset resource evaluation model or resource evaluation algorithm.

[0106] It is understood that the embodiments of this application determine the total number of pre-allocated tasks in the scheduling area by using the current scheduling resources, which can reasonably allocate computing resources to different tasks and venues, avoid the computing resources of a certain GPU or CPU being used entirely in a certain venue, and thus improve the utilization rate of computing resources.

[0107] In step S302 of some embodiments, the number of devices may refer to the number of camera devices contained in each site. The second degree of correlation may refer to the degree of correlation between a site and the occurrence of an event, determined based on the number of devices in the site, the historical task volume of the site, and the historical event volume of the site. For example, if a site has a large number of camera devices, a large number of historical events, and a small number of historical tasks, then the site can be considered to have a high degree of correlation with the event. Alternatively, if a site has a large number of camera devices, a small number of historical events, and a large number of historical tasks, then the site can be considered to have a low degree of correlation with the event. It is understood that when comprehensively evaluating the second degree of correlation of a corresponding site based on the number of devices, the historical task volume, and the historical event volume of the site, the weights of these factors in the evaluation can be adjusted according to actual needs.

[0108] In step S303 of some embodiments, the pre-allocated task volume for each site under different sampling types can be determined based on the total pre-allocated task volume and the second correlation degree of each site. For example, if the scheduling area includes two sites, B and C, and site B has a higher second correlation degree while site C has a lower second correlation degree, then most of the total pre-allocated task volume can be allocated to site B, and the remaining portion can be allocated to site C. Then, the task volume for B and C under different sampling types can be estimated based on the historical event volume corresponding to each site, thus determining the pre-allocated task volume for each site under different sampling types.

[0109] It is understood that this application embodiment determines the second degree of correlation between the corresponding site and the event by the number of camera devices included in each site, the historical task volume of the site, and the historical event volume of the site. In this way, this application embodiment can allocate a larger share of the pre-allocated total task volume to sites with high correlation based on the second degree of correlation between each site and the event, avoiding the allocation of resources to sites where tasks will not be triggered, thereby improving resource utilization. Furthermore, this application embodiment also considers the differences in the number of camera devices in different sites, avoiding the problem of allocating fewer resources to sites with a large number of camera devices, thereby improving the monitoring coverage of the camera devices.

[0110] Please see Figure 4 In some embodiments, step S302 may include, but is not limited to, steps S401 to S403:

[0111] Step S401: Obtain the first weight coefficient corresponding to the number of devices, the second weight coefficient corresponding to the historical task volume of the site, and the third weight coefficient corresponding to the historical event volume of the site.

[0112] Step S402: The site score for each site is obtained by weighting the first weighting coefficient, the second weighting coefficient, the third weighting coefficient, the number of equipment, the historical task volume of the site, and the historical event volume of the site.

[0113] Step S403: Determine the site score of each site as the second degree of association of the corresponding site.

[0114] In step S401 of some embodiments, the first weighting coefficient may refer to the weighting coefficient corresponding to the number of devices, used to measure the impact of the number of devices on the correlation between the site and the occurrence of an event. The second weighting coefficient may refer to the weighting coefficient corresponding to the historical task volume of the site, used to measure the impact of the historical task volume of the site on the correlation between the site and the occurrence of an event. The third weighting coefficient may refer to the weighting coefficient corresponding to the historical event volume of the site, used to measure the impact of the historical event volume of the site on the correlation between the site and the occurrence of an event. The first and third weighting coefficients are positive numbers, while the second weighting coefficient is a negative number.

[0115] It is understood that, by assigning positive weighting coefficients to the number of devices and the amount of historical events at the site, and negative weighting coefficients to the amount of historical tasks at the site, the embodiments of this application can reduce the impact of the amount of historical tasks at the site on resource allocation, while increasing the impact of the number of devices and the amount of historical events at the site on resource allocation. This allows sites with a larger number of devices or a larger amount of historical events to be allocated more resources, thereby improving resource utilization. For example, the first weighting coefficient can be set to 0.4, the second weighting coefficient can be set to -0.3, and the third weighting coefficient can be set to 0.6, thus enhancing the impact of the number of devices and the amount of historical events at the site on resource allocation, while relatively weakening the impact of the amount of historical tasks at the site. It should be noted that the values ​​of the first, second, and third weighting coefficients can be freely adjusted according to actual needs, but the first and third weighting coefficients must be positive, and the second weighting coefficient must be negative.

[0116] In steps S402 to S403 of some embodiments, the site score for each site can refer to the second degree of correlation of the corresponding site. The site score for each site can be obtained by weighting the first weight coefficient, the second weight coefficient, the third weight coefficient, the number of devices, the site's historical task volume, and the site's historical event volume. Before calculation, the number of devices, the site's historical task volume, and the site's historical event volume can be normalized so that their values ​​are all between 0 and 1, thereby eliminating the difference in dimensions between different indicators, making the data comparable, and ensuring the accuracy of the weighted calculation. It should be noted that the normalization method is not limited in the embodiments of this application. For example, if the maximum value of the number of devices is 100, the maximum value of the site's historical task volume is 500 hours, and the maximum value of the site's historical event volume is 100 hours. For site B, if the number of devices is 50, the site's historical task volume is 250 hours, and the site's historical event volume is 50 hours, then the normalized number of devices is 50 / 100 = 0.5, the normalized historical task volume is 250 / 500 = 0.5, and the normalized historical event volume is 50 / 100 = 0.5. Assuming the first weighting coefficient is 0.4, the second weighting coefficient is -0.3, and the third weighting coefficient is 0.6, then the site score for site B is: 0.4 × 0.5 + (-0.3) × 0.5 + 0.6 × 0.5 = 0.2 - 0.15 + 0.3 = 0.35. It should be noted that a higher site score indicates a higher correlation between the site and the events that occurred.

[0117] It is understood that this application embodiment, by quantifying the number of devices, the historical task volume of the site, and the historical event volume of the site, and combining them with corresponding weighting coefficients for weighted calculation, can transform the abstract secondary correlation between the site and the occurrence of events into a specific numerical score. Thus, when determining the pre-allocated task volume for each site under different sampling types, this application embodiment can more accurately allocate resources based on the site score of each site, thereby improving resource utilization.

[0118] Please see Figure 5 In some embodiments, step S303 may include, but is not limited to, steps S501 to S503:

[0119] Step S501: Obtain the sampling ratio of the scheduling area corresponding to different sampling types;

[0120] Step S502: Determine the total number of tasks for each site based on the pre-allocated total number of tasks and the second degree of association for each site.

[0121] Step S503: Determine the pre-allocated task volume for each site under different sampling types based on the total site task volume and sampling ratio for each site.

[0122] In step S501 of some embodiments, the sampling ratio may refer to the proportion of different sampling types in the total number of pre-allocated tasks within the scheduling area. For example, if the total number of pre-allocated tasks is 200 hours, real-time sampling accounts for 40%, near-real-time sampling accounts for 30%, and historical sampling accounts for 30%. It is understood that the sampling ratio can be freely adjusted according to actual needs.

[0123] In steps S502 to S503 of some embodiments, the total site task volume can refer to the total task volume of the corresponding site determined based on the pre-allocated total task volume and the second correlation degree. For example, if the pre-allocated total task volume is 200 hours, site B has a high second correlation degree, site C has a moderate second correlation degree, and site D has a low second correlation degree, then the total site task volume for site B can be allocated to 100 hours, the total site task volume for site C can be allocated to 75 hours, and the total site task volume for site D can be allocated to 25 hours. The pre-allocated task volume for each site corresponding to different sampling types can be determined based on the total site task volume and sampling ratio of each site. For example, if the real-time sampling ratio in the scheduling area is 40%, the near-real-time sampling ratio is 30%, and the historical sampling ratio is 30%, and the total site task volume for site B in the scheduling area is allocated to 100 hours, then the real-time sampling task volume for site B is 40 hours, the near-real-time sampling task volume is 30 hours, and the historical sampling task volume is 30 hours.

[0124] It is understood that the embodiments of this application directly determine the pre-allocated task volume for each site under different sampling types by the total number of site tasks and the sampling ratio of each site, avoiding the need to repeatedly evaluate the task volume that needs to be allocated to the corresponding site under different sampling types based on the site's historical data, thereby simplifying the task scheduling process and improving the efficiency of task scheduling.

[0125] Please see Figure 6 In one specific embodiment, the pre-allocated task volume for each site in the scheduling area corresponding to different sampling types can be determined by a site selection algorithm. Figure 6 This is a flowchart of the site selection algorithm provided in this application embodiment, specifically including: First, obtaining the number of GPUs in the scheduling area and the historical task volume, historical event volume, number of devices, and site priority for each site in the scheduling area. Then, calculating the total number of pre-allocated tasks that can be generated by executing this task in the scheduling area based on the number of GPUs. Next, obtaining the sampling ratio of the scheduling area and determining the task volume corresponding to different sampling types in the scheduling area based on the sampling ratio.

[0126] Furthermore, a site score is calculated based on the site's historical task volume, historical event volume, number of devices, and site priority. For example, the site score calculation formula can be shown in Formula 1 below:

[0127] ∑ i＝0 W i *P i (Formula 1)

[0128] Where i is the attribute number, ranging from 0 to 3. W0 represents the historical task volume of the site, and P0 represents the weight coefficient corresponding to the historical task volume of the site. W1 represents the historical event volume of the site, and P1 represents the weight coefficient corresponding to the historical event volume of the site. W2 represents the number of devices, and P2 represents the weight coefficient corresponding to the number of devices. W3 represents the priority of the site, and P3 represents the weight coefficient corresponding to the priority of the site. It can be understood that P0, P1, P2, and P3 can be freely adjusted according to actual needs, but P1, P2, and P3 must be positive numbers, and P0 must be a negative number.

[0129] It should be noted that the priority of a site is pre-set. The priority of a site refers to its importance in the scheduling area. The higher the priority of a site, the more important it is, and vice versa.

[0130] Finally, after calculating the site score for each site within the scheduling area, the pre-allocated task volume for each site under different sampling types can be determined based on the sampling ratio and the task volume corresponding to different sampling types. For example, if the total pre-allocated task volume for the scheduling area is 200 hours, the proportion of near-real-time sampling tasks to the scheduling calculation nodes is 40%, the proportion of historical sampling tasks to the scheduling calculation nodes is 60%, and there are two sites, B and C, then the task volume for near-real-time sampling in the scheduling area is 80 hours, and the task volume for historical sampling is 120 hours. If the site score for B is calculated to be 30 points and the site score for C is calculated to be 50 points using Formula 1, then the two site scores are first normalized. After normalization, the weight of site B is 30 / (30+50) = 0.375, and the weight of site C is 50 / (30+50) = 0.625. Ultimately, based on the weights of sites B and C, as well as the workload of near-real-time and historical spot checks, the workload for near-real-time spot checks at site B is determined to be 80 × 0.375 = 30 hours, and the workload for historical spot checks at site B is 120 × 0.375 = 45 hours. Similarly, the workload for near-real-time spot checks at site C is 50 hours, and the workload for historical spot checks at site C is 75 hours.

[0131] In step S203 of some embodiments, the first degree of association may refer to the degree of association between the camera device and the event, determined based on the number of historical video events and the number of historical video tasks. For example, if a camera device has a large number of historical video events but a small number of historical video tasks, it can be considered that the camera device has a high degree of association with the event; or, if a camera device has a small number of historical video events but a large number of historical video tasks, it can be considered that the camera device has a low degree of association with the event. It is understood that this application will determine the first degree of association for each camera device in the venue, and when comprehensively evaluating the first degree of association of the corresponding camera device based on the number of historical video events and the number of historical video tasks, the weight of these factors in the evaluation can be adjusted according to actual needs. The target camera device may refer to the camera device with a high first degree of association with the event among all camera devices in the venue. It is understood that the number of target camera devices selected can be freely adjusted according to actual needs. For example, if venue B contains five camera devices, and three of them have a high degree of association with the event, then these three camera devices are all identified as target camera devices.

[0132] Please see Figure 7 In some embodiments, step S203 may include, but is not limited to, steps S701 to S703:

[0133] Step S701: Obtain the device parameters of all camera devices in the venue;

[0134] Step S702: Determine the camera score for each camera device in the venue based on frame rate, detection frequency, number of historical camera tasks, and number of historical camera events;

[0135] Step S703: Determine the video score of each camera device as the first degree of correlation of the corresponding camera device.

[0136] In step S701 of some embodiments, the device parameters may refer to indicators used to evaluate the performance of the camera device. The device parameters include the frame rate and detection frequency of the corresponding camera device. The frame rate may refer to the number of image frames captured per second by the camera device. The detection frequency may refer to the frequency at which the camera device performs detection tasks.

[0137] In steps S702 to S703 of some embodiments, the camera score of each camera device can refer to the first degree of correlation between the corresponding camera device and the event. A higher camera score indicates a higher correlation between the corresponding camera device and the event. A lower camera score indicates a lower correlation between the corresponding camera device and the event. The camera score of each camera device can be determined based on frame rate, detection frequency, historical camera task volume, and historical camera event volume.

[0138] It should be noted that the calculation principle for the camera score of each camera device is the same as the calculation principle for the site score of each venue. The camera score of each camera device can also be calculated using Formula 1, which pre-sets a weighting coefficient for frame rate, detection frequency, historical camera task volume, and historical camera event volume. Each attribute (frame rate, detection frequency, historical camera task volume, and historical camera event volume) is weighted with its corresponding weighting coefficient, and the weighted results are summed. The weighting coefficients for detection frequency and historical camera task volume are negative, while the weighting coefficients for historical camera event volume and frame rate are positive.

[0139] It is understood that, in determining the video recording score of the camera device, the embodiments of this application can set negative weighting coefficients for the detection frequency and historical task volume of the camera device, and simultaneously set positive weighting coefficients for the historical event volume and frame rate of the camera device. This improves the correlation between high-performance camera devices and events, and reduces the correlation between high-usage camera devices and events. This allows for the allocation of more tasks to high-performance camera devices with lower usage frequency when allocating tasks, thereby optimizing resource utilization.

[0140] Please see Figure 8 In some embodiments, step S203 may include, but is not limited to, steps S801 to S803:

[0141] Step S801: Select the multiple camera devices with the highest correlation from the multiple camera devices according to the first preset screening ratio and the first correlation degree of each camera device as the first device set, and determine the second device set according to the first device set and all camera devices corresponding to the site.

[0142] Step S802: Select a preset number of camera devices from the second device set;

[0143] Step S803: All camera devices included in the first device set and camera devices selected from the second device set are identified as target camera devices.

[0144] In step S801 of some embodiments, the first preset screening ratio may refer to the ratio used to determine the camera devices with the highest relevance from multiple camera devices. The first device set may refer to the set of multiple camera devices with the highest relevance selected from multiple camera devices according to the first preset screening ratio and the first relevance of each camera device. For example, if the first preset screening ratio is 80%, then the first device set may be the set of 80% of the camera devices with the highest relevance selected from all camera devices. The second device set may refer to the set of camera devices remaining in the site, excluding the first device set. For example, if the first device set is the set of 80% of the camera devices with the highest relevance selected from all camera devices, then the second device set is the remaining 20% ​​of the camera devices. It should be noted that the first preset screening ratio can be freely adjusted according to actual needs.

[0145] In steps S802 to S803 of some embodiments, the preset screening quantity may refer to the number of camera devices selected from the camera devices included in the second device set. For example, the preset screening quantity may be 5 or 6. It should be noted that the preset screening quantity can be freely adjusted according to actual needs. The target camera device may refer to the camera device set composed of all camera devices in the first device set and the camera devices selected from the second device set. For example, if the preset screening quantity is 5, then 5 camera devices are selected from the second device set, which, together with all camera devices in the first device set, constitute the final multiple target camera devices.

[0146] It is understood that, by dividing all the camera devices in the venue into a first set and a second set, and determining all the camera devices in the first set and the camera devices randomly selected from the second set in a preset number as target camera devices, the workload can be prioritized for all camera devices in the first set with high correlation to the event, and the workload can also be allocated to some camera devices in the second set with low correlation to the event, thereby avoiding monitoring blind spots and improving the overall monitoring coverage of the venue.

[0147] Please see Figure 9 In one specific embodiment, multiple target camera devices can be determined by a camera selection algorithm. Figure 9This is a flowchart of the camera selection algorithm provided in this application embodiment, specifically including: first, obtaining the historical task volume, historical event volume, device parameters (including frame rate and detection frequency), and priority of each camera device in the venue. Then, obtaining the preset weight coefficients corresponding to the historical task volume, historical event volume, frame rate, detection frequency, and camera priority, and calculating the camera score corresponding to each camera device using Formula 1. It should be noted that when calculating the camera score, the preset weight coefficients corresponding to the camera priority, frame rate, and historical event volume are positive numbers. The preset weight coefficients corresponding to the detection frequency and historical task volume are negative numbers.

[0148] It should be noted that the priority of the camera equipment is pre-set. The priority of the camera equipment refers to its importance in the venue. The higher the priority of the camera equipment, the higher its importance, and vice versa.

[0149] Furthermore, the camera devices are sorted according to their respective camera ratings, and the top 80% of the cameras with the highest ratings are selected as the first set of devices based on the distribution of the ranking scores. The remaining 20% ​​of the cameras are selected as the second set of devices. Finally, multiple target camera devices are determined based on all the camera devices included in the first set and a specified number of camera devices randomly selected from the second set of devices.

[0150] In step S204 of some embodiments, historical time period data can refer to the time period during which the task was in a triggered state within a past period. For example, historical time period data can include the time period during which illegal throwing events occurred when illegal throwing detection tasks were performed within the past month. Target scheduling data can refer to data generated based on historical time period data, the pre-allocated task volume corresponding to different sampling types, multiple target camera devices, and tasks corresponding to the site, used to guide the corresponding target camera device to perform a specific task within a specific time period. For example, target scheduling data can specify that a certain target camera device performs illegal throwing detection tasks through real-time sampling from 9:00 AM to 11:00 AM every day, with each task requiring a duration of 30 minutes. It is understood that embodiments of this application will generate corresponding target scheduling data for each target camera device in the site.

[0151] Please see Figure 10 In some embodiments, step S204 may include, but is not limited to, steps S1001 to S1008:

[0152] Step S1001: Obtain the task time period data and the work time period data of the site for performing the corresponding task;

[0153] Step S1002: Divide the historical time period data into time periods based on the task time period data to obtain multiple sub-historical time period data;

[0154] Step S1003: Determine the number of sub-site historical events corresponding to each sub-historical time period based on multiple sub-historical time period data and the number of site historical events;

[0155] Step S1004: Based on the second preset filtering ratio and the number of historical events corresponding to each sub-historical time period, the data of multiple sub-historical time periods are filtered to obtain the data of multiple sub-historical time periods with the most historical events as the first time period set.

[0156] Step S1005: Divide the work period data into time periods based on the task period data to obtain multiple sub-work period data;

[0157] Step S1006: Determine the sub-site historical task volume corresponding to each sub-work period data based on multiple sub-work period data and site historical task volume.

[0158] Step S1007: Based on the third preset filtering ratio and the historical task volume of each sub-work period data, the data of multiple sub-work periods are filtered to obtain the data of multiple sub-work periods with the highest historical task volume of each sub-work period as the second time period set.

[0159] Step S1008: Determine multiple target time period data based on the first time period set and the second time period set; generate target scheduling data for each target camera device in the site based on the multiple target time period data, the pre-allocated task volume corresponding to different sampling types, and the tasks corresponding to multiple target camera devices and sites.

[0160] In step S1001 of some embodiments, the task time period data can refer to the duration required for task execution. For example, if the illegal throwing detection task requires 2 hours, then the task time period data is 2 hours. The work time period data can refer to the total time period used by the site to execute tasks. For example, if a logistics warehouse can execute tasks for 8 hours per day, then the work time period data is 8 hours. It should be noted that the tasks to be executed by each site within the scheduling area and the corresponding task time period data can be predetermined.

[0161] In steps S1002 to S1003 of some embodiments, sub-historical time period data can refer to sub-historical time periods with the same duration as the task time period data, obtained by dividing the historical time period data into time periods based on the task time period data. For example, if the task time period data is 2 hours, then all consecutive 2-hour time periods are extracted from the historical time period data as sub-historical time period data. Sub-site historical event quantity can refer to the total duration of events occurring within the corresponding sub-historical time period. For example, if the cumulative duration of illegal throwing events is 1.5 hours within a certain sub-historical time period (such as from 9:00 AM to 11:00 AM on a certain day), then the sub-site historical event quantity corresponding to that sub-historical time period data is 1.5 hours.

[0162] In step S1004 of some embodiments, the second preset filtering ratio may refer to the ratio used for filtering multiple sub-historical time period data. The first time period set may refer to the multiple sub-historical time period data with the highest number of sub-site historical events obtained after filtering multiple sub-historical time period data according to the second preset filtering ratio and the number of sub-site historical events corresponding to each sub-historical time period data. For example, if the second preset filtering ratio is 80%, then the 80% of sub-historical time period data with the highest number of sub-site historical events are selected from all sub-historical time period data as the first time period set. It is understood that the second preset filtering ratio can be freely adjusted according to actual needs.

[0163] In steps S1005 to S1006 of some embodiments, sub-work period data can refer to sub-work period segments with the same duration as the task period data, obtained by dividing the work period data into time segments based on the task period data. For example, if the task period data is 2 hours and the work period data is 8 hours, then all consecutive 2-hour time segments are extracted from the 8-hour work period as sub-work period data. Sub-site historical task volume can refer to the amount of tasks performed within these sub-work period segments. For example, if the duration of the illegal throwing detection task performed within a certain sub-work period segment (such as from 2 PM to 4 PM on a certain day) is 1 hour, then the sub-site historical task volume corresponding to that sub-work period data is 1 hour.

[0164] In step S1007 of some embodiments, the third preset screening ratio may refer to the ratio used to screen multiple sub-work period data. The second time period set may refer to the multiple sub-work period data with the highest historical task volume in the sub-site after screening multiple sub-work period data according to the third preset screening ratio and the historical task volume of the sub-site corresponding to each sub-work period data. For example, if the third preset screening ratio is 20%, then the 20% of sub-work period data with the highest historical task volume in the sub-site is selected from all sub-work period data as the second time period set. It is understood that the third preset screening ratio can be freely adjusted according to actual needs.

[0165] In step S1008 of some embodiments, the target time period data may refer to the time period used to perform the task, determined by combining the first time period set and the second time period set. For example, all time period data in the first time period set and all time period data in the second time period set can be used as target time period data to obtain multiple target time period data; or, all time period data in the first time period set can be used as target time period data, and a portion of time period data is randomly selected from all time period data in the second time period set as target time period data to obtain multiple target time period data. It is understood that the number of target time period data can be freely adjusted according to actual needs. The target scheduling data for each target camera device in the site may refer to data generated based on multiple target time period data, the pre-allocated task volume corresponding to different sampling types, multiple target camera devices, and the tasks corresponding to the site, used to guide the corresponding target camera device to perform a specific task within a specific time period. For example, the target scheduling data may specify that a certain target camera device performs a non-real-time sampling inspection to detect illegal throwing from 8:00 AM to 10:00 AM, and the task execution time is 2 hours.

[0166] It is understood that, according to the embodiments of this application, multiple historical time periods with the most historical events in a sub-site can be selected as a first time period set based on a second preset filtering ratio, and multiple working time periods with the most historical tasks in a sub-site can be selected as a second time period set based on a third preset filtering ratio. Then, the target time period data for all target camera devices in the site can be determined based on the first time period set and the second time period set. In this way, the embodiments of this application can allocate time periods with a high probability of event occurrence to target camera devices to execute tasks, and cover time periods with a large task volume, effectively reducing the risk of invalid resource occupation when camera devices are allocated to time periods where tasks will not be triggered, improving the coverage of task execution time periods and resource utilization, thereby improving the efficiency of task scheduling.

[0167] It should be noted that, in this embodiment of the application, a predetermined key time range of the site can also be used to replace the site's working hours data to determine the second time period set. The calculation principle for determining the second time period set based on the key time range of the site is the same as that based on the site's working hours data.

[0168] Please see Figure 11 In one specific embodiment, multiple target time period data can be determined by a job time period selection algorithm. Figure 11This is a flowchart of the work period selection algorithm provided in this application embodiment, specifically including: First, acquiring task period data, work period data, historical period data, site historical event volume, and site historical task volume. Then, dividing the work period data into multiple sub-work period data based on the task period data corresponding to the task being executed; and dividing the historical period data into multiple sub-historical period data based on the task period data. Further, determining the sub-site historical event volume corresponding to each sub-historical period data based on the site historical event volume, and sorting them from largest to smallest based on the size of the sub-site historical event volume, then selecting the top 80% of the sub-site historical event volume from the sorted sub-historical period data as the first period set. Further, determining the sub-site historical task volume corresponding to each sub-work period data based on the site historical task volume, and sorting them from largest to smallest based on the size of the sub-site historical task volume, then selecting the top 20% from the sorted sub-work period data as the second period set. Finally, using all period data contained in the first period set and all period data contained in the second period set as target period data, thereby determining multiple target period data.

[0169] Please see Figure 12 In some embodiments, step S204 may include, but is not limited to, steps S1201 to S1205:

[0170] Step S1201: Determine the task type of the target camera device based on the task corresponding to the site and the multiple target camera devices;

[0171] Step S1202: For each target camera device in the venue, determine the target sampling type corresponding to the target camera device according to the task type;

[0172] Step S1203: Determine the candidate scheduling data for the corresponding target camera equipment based on multiple target time period data, target sampling type, and task type;

[0173] Step S1204: Determine the candidate allocation task quantity corresponding to the target sampling type based on the target time period data corresponding to all target camera devices in the site;

[0174] Step S1205: When the number of candidate assigned tasks is less than or equal to the number of pre-assigned tasks for the corresponding sampling type, the candidate scheduling data is determined as the target scheduling data for the corresponding target camera device.

[0175] In step S1201 of some embodiments, the task type may refer to the detection function that the target camera device needs to perform based on the task corresponding to the site. For example, if the task corresponding to the site is illegal throwing, then the task type is illegal throwing detection.

[0176] In step S1202 of some embodiments, the target sampling type may refer to the sampling type corresponding to the execution required by the target camera device based on the task type. For example, if the task type is illegal throwing detection, and the site uses real-time sampling when conducting illegal throwing detection, then the target sampling type is real-time sampling.

[0177] In steps S1203 to S1205 of some embodiments, the candidate scheduling data may refer to the preliminary scheduling data corresponding to the target camera device generated based on the target time period data, the target sampling type, and the task type. The candidate allocated task quantity corresponding to the target sampling type may refer to the total task quantity for the target sampling type calculated based on the target time period data corresponding to all target camera devices. The target scheduling data may refer to the candidate scheduling data corresponding to the target camera device when the candidate allocated task quantity is less than or equal to the pre-allocated task quantity for the corresponding sampling type. For example, if the target sampling type is a real-time sampling, the candidate allocated task quantity for the target sampling type is 4 hours, and the pre-allocated task quantity corresponding to the real-time sampling in the venue is 5 hours, then the candidate scheduling data can be determined to be the target scheduling data required for the target camera device to perform the task.

[0178] It should be noted that when the number of candidate assigned tasks exceeds the pre-assigned task volume for the corresponding sampling type, the embodiments of this application can adjust the number of candidate assigned tasks. For example, the embodiments of this application can re-screen the target time period data, reducing the proportion of the sub-work period data with the highest historical task volume selected from multiple sub-work period data, thereby reducing the target time period data to ensure that the number of candidate assigned tasks meets the pre-assigned task volume limit for the corresponding sampling type.

[0179] It is understood that this application embodiment generates candidate scheduling data by using multiple target time period data, the task type of the target camera equipment, and the target sampling type. When the candidate allocation task quantity matches the pre-allocated task quantity corresponding to the target sampling type, the candidate scheduling data is determined as the target scheduling data. In this way, this application embodiment can effectively prevent resource allocation from exceeding the preset limit in task scheduling, reduce the need for rescheduling tasks, and thus improve the efficiency of task scheduling.

[0180] Please see Figure 13 In some embodiments, after step S204, steps S1301 to S1304 may also be included, but are not limited to:

[0181] Step S1301: Schedule the corresponding target camera equipment according to the target scheduling data and obtain the scheduling time;

[0182] Step S1302: Obtain the data update cycle and determine the update time based on the data update cycle and scheduling time;

[0183] Step S1303: When it is determined that the current time has reached the update time, the historical data of the target site before the current time is obtained, and the historical data of the target camera device before the current time is obtained.

[0184] Step S1304: Update the historical data of the site based on the historical data of the target site, and update the historical data of the camera based on the historical data of the target camera.

[0185] In step S1301 of some embodiments, the target camera device is scheduled to perform a specified detection task according to the corresponding target scheduling data. The scheduling time can refer to the actual time when the camera device starts executing the task.

[0186] In step S1302 of some embodiments, the data update cycle can refer to a preset time interval. For example, the data update cycle can be one week or one month. It is understood that the data update cycle can be freely adjusted according to actual needs. The update time can refer to the time point obtained by adding the data update cycle to the scheduling time. For example, if the scheduling time is 9:00 AM on May 3rd and the data update cycle is one week, the update time is 9:00 AM on May 10th.

[0187] In steps S1303 to S1304 of some embodiments, the target site historical data may refer to statistical data on task execution and event occurrence related to the site collected before the current time. The target camera historical data may refer to statistical data on task execution and event occurrence related to the target camera device collected before the current time. Updating the site historical data may refer to replacing or appending newly collected target site historical data to the existing site historical data. For example, if the current time is determined to be 9:00 AM on May 10th, and the data update cycle is one week, target site historical data from May 3rd to May 10th can be collected, and then this target site historical data can be appended to the previous site historical data, thereby updating the site historical task volume and site historical event volume in the site historical data.

[0188] It should be noted that updating camera history data can refer to replacing or appending newly collected target camera history data to the existing camera history data. The specific principle of updating camera history data is the same as that of updating site history data.

[0189] It is understood that when the next task scheduling time arrives, this embodiment of the application can collect historical data of the target site and historical data of the target camera prior to the current time, and update the site historical data and camera historical data used in the previous task scheduling accordingly. In this way, this embodiment of the application can continuously iterate and update the historical task volume and historical event volume of the site in the site historical data, and the historical task volume and historical event volume of the camera in the camera historical data, thereby ensuring that resources are always allocated to sites and camera equipment with a high degree of correlation to the occurrence of events, avoiding the allocation of resources to camera equipment whose tasks will not be triggered, and reducing the allocation of resources to sites with low correlation to events, thereby improving the efficiency of task scheduling.

[0190] The task scheduling method proposed in this application can achieve automated, periodic, and efficient task generation for different sampling types by utilizing historical site data of all sites within the scheduling area, historical video recording data of all cameras within each site, and historical time period data. This application also allocates resources for different sampling types by matching the current scheduling resources within the scheduling area with the sampling ratios of different sampling types, and can allocate resources for different sampling types to sites and cameras with a higher probability of event occurrence. Furthermore, when generating tasks for the current camera device, this application ensures that the sampling type and task type corresponding to the camera device's task are consistent with the site's historical data and the historical video recording data of all cameras within each site. This means that unified and standardized task generation is used for camera devices with different sampling types, thereby achieving a simplified design of the camera device task scheduling architecture in the Internet of Things (IoT) field and standardized and unified construction under the camera device governance dimension.

[0191] Please see Figure 14In a specific embodiment, the task scheduling method provided in this application includes acquiring basic task information, executing a scheduling strategy, configuring alarms, and configuring notifications. Acquiring basic task information refers to acquiring pre-defined basic information within the scheduling area used to generate camera detection tasks. For example, basic task information may include the sampling type corresponding to the task, the task's time period data, the sampling ratio, and the detection functions configured on the camera. Executing the scheduling strategy refers to generating target scheduling data for each target camera within the site based on the acquired basic task information and the method described in any of the above embodiments. Configuring alarms refers to configuring early warning notifications (such as insufficient resources, task execution timeout, task not being executed, or task abnormality) for different task states during task execution when the target camera executes the task according to the corresponding target scheduling data. Configuring notifications refers to sending the early warning information in the alarm configuration to the corresponding operation object through a pre-configured method. It is understood that, after the target scheduling data corresponding to each target camera device is determined, the embodiments of this application can set alarm configuration and notification configuration for each target camera device, so as to issue alarms and notifications to target camera devices with abnormal task status in a timely manner, thereby improving the overall task scheduling efficiency.

[0192] Please see Figure 15In one specific embodiment, after determining the target scheduling data for each target camera device in each location within the scheduling area, the task scheduling method provided in this application includes: firstly, generating scheduling tasks based on the target scheduling data corresponding to all target camera devices within the scheduling area, thereby enabling the corresponding target camera devices to execute tasks. Then, the video streams or images collected by the target camera devices during task execution are sent to edge computing nodes. The edge computing nodes perform edge detection on the video streams or images to determine whether an event has occurred during the target camera device's task execution, and report the operation period of the event, the event volume and task volume of the target camera device, and the location of the target camera device to the cloud server. Alternatively, if the edge computing nodes have high resource consumption, they can directly report the collected video streams or images to the cloud server. The cloud server can perform large-scale model detection on the video or images to determine whether an event has occurred during the target camera device's task execution, and directly calculate the operation period of the event, the event volume and task volume of the target camera device, and the location of the target camera device. Finally, the cloud server outputs records of the tasks performed by the target camera devices within the site. During the current task scheduling process, it statistically analyzes the event volume and task volume of all sites within the scheduling area to determine the target site's historical data, and statistically analyzes the event volume and task volume of all camera devices within the site to determine the target camera's historical data. When task scheduling is required next time, the historical site data and target camera data obtained from this statistical analysis can be used to update the historical site data and camera data needed for the next task scheduling process. It is understood that, after completing the current task scheduling, this embodiment of the application will statistically analyze the assigned task volume, the number of events that occurred, and the time periods of the events in the current task scheduling, and update the data used for the next task scheduling based on the statistical data, thereby gradually improving the efficiency of subsequent task scheduling.

[0193] Please see Figure 16 This application also provides a task scheduling system, which includes:

[0194] The data acquisition module 1601 is used to acquire the historical data of all venues within the scheduling area and the historical video data of all camera devices in each venue. The historical data of the venues includes the historical task volume allocated to the corresponding venues and the historical event volume of the venues obtained based on different sampling types. The historical video data includes the historical task volume allocated to the corresponding camera devices and the historical event volume of the camera devices obtained based on different sampling types. The occurrence of an event is determined when the task is in the triggered state.

[0195] The task quantity module 1602 is used to obtain the current scheduling resources and determine the pre-allocated task quantity for each site under different sampling types based on the current scheduling resources, the site's historical task quantity, and the site's historical event quantity.

[0196] The camera equipment module 1603 is used to determine the first degree of correlation between the corresponding camera equipment and the event based on the number of historical camera tasks and the number of historical camera events for each camera equipment in each venue, and to select multiple target camera equipment from all camera equipment in the venue based on the first degree of correlation of each camera equipment.

[0197] The scheduling module 1604 is used to acquire historical time period data of tasks in the trigger state for each site, and generate target scheduling data for each target camera in the site based on the historical time period data, the pre-allocated task volume corresponding to different sampling types, multiple target camera devices and the tasks corresponding to the site.

[0198] It is evident that the content of the above-described task scheduling method embodiments is applicable to the embodiments of this task scheduling system. The specific functions implemented by the embodiments of this task scheduling system are the same as those of the above-described task scheduling method embodiments, and the beneficial effects achieved are also the same as those achieved by the above-described task scheduling method embodiments.

[0199] Reference Figure 17 , Figure 17 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0200] The processor 1701 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0201] The memory 1702 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1702 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1702 and is called and executed by the processor 1701 using the task scheduling method of the embodiments of this application.

[0202] The input / output interface 1703 is used to implement information input and output;

[0203] The communication interface 1704 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0204] Bus 1705 transmits information between various components of the device (e.g., processor 1701, memory 1702, input / output interface 1703, and communication interface 1704);

[0205] The processor 1701, memory 1702, input / output interface 1703 and communication interface 1704 are connected to each other within the device via bus 1705.

[0206] This application also provides a computer program product, which includes a computer program. The processor of a computer device reads and executes the computer program, causing the computer device to perform the task scheduling method described above.

[0207] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described task scheduling method.

[0208] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0209] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0210] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0211] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0212] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0213] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0214] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0215] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0216] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0217] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0218] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A task scheduling method, characterized by, The method includes: Acquire historical site data for all sites within the scheduling area and historical video data for all cameras within each site. The historical site data includes the historical site task volume allocated to the corresponding site and the historical site event volume obtained based on different sampling types. The historical video data includes the historical video task volume allocated to the corresponding camera device and the historical video event volume obtained based on different sampling types. The occurrence of an event is determined when the task is in a triggered state. Obtain the current scheduling resources, and determine the pre-allocated task volume for each site under different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume. For each camera device in each venue, a first degree of association between the corresponding camera device and the occurrence of the event is determined based on the number of historical camera tasks and the number of historical camera events. Based on the first degree of association of each camera device, multiple target camera devices are selected from all camera devices in the venue. For each venue, historical time period data of the task in the venue when it is in the triggered state is obtained, and target scheduling data of each target camera in the venue is generated based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices and the task corresponding to the venue. The step of determining the first degree of correlation between the corresponding camera device and the occurrence of the event based on the camera history task volume and the camera history event volume includes: Obtain the device parameters of all camera devices in the venue, including the frame rate and detection frequency of the corresponding camera device; The camera score for each camera device in the venue is determined based on the frame rate, the detection frequency, the number of historical camera tasks, and the number of historical camera events. The camera score for each camera device is determined as the first degree of association for that camera device.

2. The method of claim 1, wherein, The step of determining the pre-allocated task volume for each site under different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume includes: The total number of pre-allocated tasks for the scheduling region is determined based on the current scheduling resources. Obtain the number of camera devices in each venue, and determine the second degree of correlation between the corresponding venue and the event based on the number of devices, the historical task volume of the venue, and the historical event volume of the venue; The pre-allocated task quantity for each site is determined based on the total pre-allocated task quantity and the second correlation degree of each site for the different sampling types.

3. The method of claim 2, wherein, The step of determining the second degree of correlation between the corresponding site and the event based on the number of devices, the historical task volume of the site, and the historical event volume of the site includes: Obtain the first weighting coefficient corresponding to the number of devices, the second weighting coefficient corresponding to the historical task volume of the site, and the third weighting coefficient corresponding to the historical event volume of the site. The first weighting coefficient and the third weighting coefficient are positive numbers, and the second weighting coefficient is negative number. The site score for each site is obtained by weighting the first weighting coefficient, the second weighting coefficient, the third weighting coefficient, the number of devices, the historical task volume of the site, and the historical event volume of the site. The site score for each site is determined as the second degree of association for that site.

4. The method of claim 2, wherein, The step of determining the pre-allocated task volume for each site corresponding to the different sampling types based on the total pre-allocated task volume and the second correlation degree of each site includes: Obtain the sampling ratio of the scheduling area corresponding to different sampling types; The total number of tasks for each site is determined based on the total number of pre-allocated tasks and the second degree of association for each site. The pre-allocated task volume for each site is determined based on the total site task volume and the sampling ratio for each site, corresponding to the different sampling types.

5. The method of claim 1, wherein, The step of selecting multiple target camera devices from all camera devices in the site based on a first correlation degree for each camera device includes: Based on a first preset screening ratio and a first degree of correlation for each camera device, a first set of cameras with the highest correlation is selected from the multiple camera devices. A second set of cameras is then determined based on the first set of cameras and all camera devices corresponding to the venue. Select a preset number of camera devices from the second set of devices; The target camera device is determined by selecting all camera devices included in the first set of devices and the camera devices selected from the second set of devices.

6. The method according to claim 1, characterized in that, The step of generating target scheduling data for each target camera in the site based on the historical time period data, the pre-allocated task volume corresponding to different sampling types, the multiple target camera devices, and the tasks corresponding to the site includes: Obtain the task time period data and the working time period data of the site for performing the corresponding task; The historical time period data is divided into multiple sub-historical time period data based on the task time period data; The number of sub-site historical events corresponding to each of the sub-historical time periods is determined based on the multiple sub-historical time period data and the number of site historical events. Based on the second preset filtering ratio and the number of historical events corresponding to each sub-historical time period, the data of multiple sub-historical time periods are filtered to obtain the data of multiple sub-historical time periods with the most sub-site historical events as the first time period set. The work period data is divided into multiple sub-work period data based on the task period data; Based on the data of the multiple sub-working periods and the historical task volume of the site, determine the historical task volume of the sub-site corresponding to each of the sub-working periods; Based on the third preset filtering ratio and the historical task volume of the sub-site corresponding to each sub-work period data, the data of the multiple sub-work periods are filtered by time period to obtain the data of the multiple sub-work periods with the highest historical task volume of the sub-site as the second time period set. Multiple target time period data are determined based on the first time period set and the second time period set, and target scheduling data for each target camera device in the venue is generated based on the multiple target time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the venue.

7. The method according to claim 6, characterized in that, The step of generating target scheduling data for each target camera in the venue based on the multiple target time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the venue includes: The task type of the target camera device is determined based on the task corresponding to the site and the multiple target camera devices; For each target camera device within the site, the target sampling type corresponding to the target camera device is determined according to the task type; Based on the multiple target time period data, the target sampling type, and the task type, candidate scheduling data for the corresponding target camera equipment is determined; The candidate allocation task quantity corresponding to the target sampling type is determined based on the target time period data of all target camera devices in the site. When the number of candidate assigned tasks is less than or equal to the number of pre-assigned tasks for the corresponding sampling type, the candidate scheduling data is determined as the target scheduling data for the corresponding target camera device.

8. The method according to claim 1, characterized in that, After generating target scheduling data for each target camera in the site based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices, and the tasks corresponding to the site, the method further includes: The corresponding target camera devices are scheduled according to the target scheduling data, and the scheduling time is obtained; Obtain the data update cycle, and determine the update time based on the data update cycle and the scheduling time; When it is determined that the current time has reached the update time, the target site historical data of the site before the current time is obtained, and the target camera historical data of the target camera device before the current time is obtained; The historical data of the target site is updated based on the historical data of the target site, and the historical data of the camera is updated based on the historical data of the target camera.

9. A task scheduling system, characterized in that, The system includes: The data acquisition module is used to acquire the historical data of all venues within the scheduling area and the historical video data of all camera devices in each venue. The historical data of the venues includes the historical task volume allocated to the corresponding venue and the historical event volume of the venues obtained based on different sampling types. The historical video data includes the historical task volume allocated to the corresponding camera device and the historical event volume of the camera device obtained based on different sampling types. The occurrence of an event is determined when the task is in the triggered state. The task volume module is used to obtain the current scheduling resources and determine the pre-allocated task volume for each site in the different sampling types based on the current scheduling resources, the site's historical task volume, and the site's historical event volume. The camera equipment module is used to determine the first correlation degree between the corresponding camera equipment and the occurrence of the event based on the historical task volume and the historical event volume of the camera equipment in each venue, and to select multiple target camera equipment from all camera equipment in the venue based on the first correlation degree of each camera equipment. The scheduling module is used to acquire historical time period data of the tasks in the trigger state in each venue, and generate target scheduling data for each target camera in the venue based on the historical time period data, the pre-allocated task volume corresponding to the different sampling types, the multiple target camera devices and the tasks corresponding to the venue. The step of determining the first degree of correlation between the corresponding camera device and the occurrence of the event based on the camera history task volume and the camera history event volume includes: Obtain the device parameters of all camera devices in the venue, including the frame rate and detection frequency of the corresponding camera device; The camera score for each camera device in the venue is determined based on the frame rate, the detection frequency, the number of historical camera tasks, and the number of historical camera events. The camera score for each camera device is determined as the first degree of association for that camera device.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the task scheduling method according to any one of claims 1-8.

11. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the task scheduling method according to any one of claims 1-8.

Images