Task-driven video inspection method and device, computing device, computer readable storage medium and program product

Abstract

The embodiment of the specification provides a task-driven video inspection method and device, a computing device, a computer readable storage medium and a program product, the method comprises the following steps: in response to a target account login, a target inspection task corresponding to a current time is acquired, and a plurality of inspection items contained in the target inspection task and respective recommended time are determined; in a dynamic triggering mode, the system automatically calculates and generates an inspection time period according to a real-time flight operation plan and a preset time offset rule. When the time period matches the current time, the corresponding target inspection task is triggered; a target item is selected from the plurality of inspection items, and a video analysis algorithm is used to analyze the monitoring video; the analysis result is compared with the corresponding qualified standard, so as to generate the inspection result of the item; in the case that the corresponding recommended time is reached, the step of selecting the target inspection item from the plurality of inspection items is executed, and in the case that the inspection result contains an inspection problem, the inspection problem is sent to an alarm system.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of computer technology, and in particular to a task-driven video inspection method. Background Technology

[0002] As large public places such as airports and train stations continue to expand, the coverage and number of video surveillance devices are also growing. Currently, there are two main methods for inspecting surveillance videos: one is to review the recordings after an accident or report, which is slow and cannot intervene in time before or during an incident; the other is to manually inspect each monitoring screen in real time, but due to the large number of monitoring devices and the wide area covered, this method suffers from high labor costs, low inspection efficiency, and is prone to missed inspections due to visual fatigue. Furthermore, it is difficult to dynamically adjust the inspection focus according to changes in the on-site operational situation, resulting in insufficient control over the on-site security situation. Summary of the Invention

[0003] In view of this, embodiments of this specification provide a task-driven video inspection method. One or more embodiments of this specification also relate to a task-driven video inspection device, a computing device, a computer-readable storage medium, and a computer program product, to address the technical deficiencies existing in the prior art.

[0004] According to a first aspect of the embodiments of this specification, a task-driven video inspection method is provided, comprising: In response to the login operation of the target account, the system obtains the target inspection task corresponding to the current time under the target account, and determines multiple inspection items corresponding to the target inspection task and the suggested time for each inspection item. The inspection time period corresponding to the target inspection task corresponds to the current time, and the inspection time period is determined based on the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the inspection time period is automatically calculated and generated according to external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. Select a target inspection project from the multiple inspection projects, and analyze the monitoring video corresponding to the target inspection project based on the inspection content and video analysis algorithm to obtain the analysis results; The analysis results are compared with the pass / fail standards in the inspection content, and the inspection results of the target inspection project are generated based on the comparison results. If the recommended time is reached, continue to execute the step of selecting the target inspection project from the multiple inspection projects, and if the inspection results contain inspection problems, send the inspection problems to the alarm system.

[0005] According to a second aspect of the embodiments of this specification, a task-driven video inspection device is provided, comprising: The task acquisition module is configured to respond to the login operation of the target account, acquire the target inspection task corresponding to the current time under the target account, and determine the multiple inspection items corresponding to the target inspection task and the suggested time for each inspection item. The inspection time period corresponding to the target inspection task corresponds to the current time, and the inspection time period is determined based on the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the inspection time period is automatically calculated and generated according to external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. The project inspection module is configured to select a target inspection project from the multiple inspection projects, analyze the monitoring video corresponding to the target inspection project based on the inspection content corresponding to the target inspection project and in combination with video analysis algorithms, and obtain analysis results. The result generation module is configured to compare the analysis results with the pass / fail standards in the inspection content, and generate the inspection results of the target inspection project based on the comparison results. The problem reporting module is configured to continue executing the step of selecting the target inspection project from the multiple inspection projects when the corresponding suggested time is reached, and to send the inspection problem to the alarm system if the inspection result contains an inspection problem.

[0006] According to a third aspect of the embodiments of this specification, a computing device is provided, comprising: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, they implement the steps of the above-described task-driven video inspection method.

[0007] According to a fourth aspect of the embodiments of this specification, a computer-readable storage medium is provided that stores a computer program / instructions that, when executed by a processor, implement the steps of the task-driven video inspection method described above.

[0008] According to a fifth aspect of the embodiments of this specification, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement the steps of the task-driven video inspection method described above.

[0009] This specification provides an embodiment of a task-driven video inspection method. By responding to a login operation of a target account, it obtains the target inspection task corresponding to the current time under the target account, and determines the multiple inspection items included in the target inspection task and their corresponding suggested time. The inspection time period corresponding to the target inspection task is determined based on the task triggering mode. When the task triggering mode is dynamic, the inspection time period can be automatically calculated and generated based on external operation plan data and preset time offset rules. This external operation plan data includes dynamically accessed flight operation plans. Therefore, in airport inspection scenarios, it enables the automatic planning and issuance of inspection tasks before peak operation periods such as concentrated flight departures or arrivals. This transforms the passive mode of traditional solutions, which relies on manual experience to determine inspection timing and is disconnected from the on-site operational situation, into a proactive prediction mode driven by operation plan data. This effectively avoids the problem of inspection tasks running idle during off-peak hours or being absent during peak hours, allowing video inspection resources to be concentrated on the critical periods with the most urgent regulatory needs, significantly improving the matching degree between inspection resources and the actual operational situation. Based on this, from multiple inspection... The project selects target inspection items, analyzes the corresponding monitoring videos based on the inspection content of these target inspection items and video analysis algorithms, and obtains analysis results to improve the processing efficiency of video inspection. The analysis results are then compared with the pass / fail standards in the inspection content, and inspection results for the target inspection items are generated based on the comparison results. This ensures that each inspection operation by inspection personnel has clear and standardized inspection content for reference, avoiding missed inspections and misjudgments caused by visual fatigue or distraction during manual inspections. When each inspection item reaches its suggested time, the project automatically continues to the next inspection item. Time constraints standardize the inspection time of each item, preventing excessive time consumption of a single inspection item from affecting the overall task progress, and also preventing inspections from becoming merely a formality due to insufficient time, ensuring the standardization and consistency of inspection work. If inspection results contain inspection problems, the problems are sent to the alarm system, realizing a complete closed loop from problem discovery to information reporting. This significantly improves the closed-loop management capabilities of security patrols in airport inspection scenarios and the overall control of the on-site operational situation.

[0010] It should be noted that in the field of video surveillance at large transportation hubs such as airports and train stations, enabling the focus of inspections to automatically shift to corresponding key areas in response to changes in operational status, rather than relying on manual, one-by-one, and universally applicable inspections of surveillance footage, has long been a key technical challenge that needs to be addressed, impacting inspection efficiency. Although experienced inspectors can manually retrieve surveillance footage of relevant areas for focused inspections during peak flight arrival times based on their experience, the locations of key nodes such as exits, baggage claim carousels, and security checkpoints vary for different flights, and flight schedules themselves are subject to dynamic changes such as delays, advances, and cancellations. Inspectors relying solely on memory and experience to determine "when to inspect which area" is highly prone to omissions and oversights.

[0011] This embodiment of the specification introduces a dynamic triggering mode and preset time offset rules, enabling the planning and distribution of inspection tasks to be completed automatically before the peak operating period. When the inspection period can be automatically calculated and generated based on dynamically accessed external operation plan data, there is no need for inspection personnel to manually determine the inspection timing. Furthermore, since the inspection period is bound to specific flight operation data, and the key nodes such as exits, security checkpoints, and check-in counters corresponding to each flight are pre-configured through the binding relationship between the project and the monitoring equipment when the task is created, the system can simultaneously determine the monitoring equipment and areas that need to be inspected during the dynamic generation of the inspection period. Combined with video analysis algorithms, it can also significantly improve inspection efficiency, avoid the problems of missed inspections and forgetting caused by manual memory and inspection, and achieve the unexpected effect of "inspection before the peak and precise matching of inspection focus". Attached Figure Description

[0012] Figure 1 This is a flowchart of a task-driven video inspection method provided in one embodiment of this specification; Figure 2 This is a schematic diagram illustrating the processing procedure of a task-driven video inspection method provided in one embodiment of this specification; Figure 3 This is a schematic diagram illustrating an inspection project management system provided in one embodiment of this specification. Figure 4 This is a schematic diagram illustrating the processing of a camera configuration provided in one embodiment of this specification; Figure 5 This is a schematic diagram illustrating the processing of a newly added patrol task provided in one embodiment of this specification; Figure 6 This is a schematic diagram illustrating the task execution process provided in one embodiment of this specification; Figure 7 This is a schematic diagram of the subtask inspection process provided in one embodiment of this specification; Figure 8This is a schematic diagram of the state change of a subtask inspection provided in one embodiment of this specification; Figure 9 This is a schematic diagram of a problem reporting process provided in one embodiment of this specification; Figure 10 This is a schematic diagram of the structure of a task-driven video inspection device provided in one embodiment of this specification; Figure 11 This is a structural block diagram of a computing device provided in one embodiment of this specification. Detailed Implementation

[0013] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0014] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0015] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0016] Furthermore, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0017] This specification provides a task-driven video inspection method, and also relates to a task-driven video inspection device, a computing device, and a computer-readable storage medium, which will be described in detail in the following embodiments.

[0018] See Figure 1 , Figure 1 The flowchart of a task-driven video inspection method provided in one embodiment of this specification is shown, which specifically includes the following steps.

[0019] The task-driven video inspection method provided in the embodiments of this specification is a method for inspecting public places, especially places covered by video surveillance equipment such as airports, train stations, and shopping malls.

[0020] This embodiment takes the application of the video inspection method to an airport inspection scenario as an example to provide a detailed explanation of the method.

[0021] Step 102: In response to the login operation of the target account, obtain the target inspection task corresponding to the current time under the target account, and determine the multiple inspection items corresponding to the target inspection task and the suggested time for each inspection item. The inspection time period corresponding to the target inspection task corresponds to the current time, and the inspection time period is determined based on the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the inspection time period is automatically calculated and generated according to external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans.

[0022] The target account can be understood as the identity identifier of the inspector or role who logs into the inspection system, used to distinguish the identity and permissions of different inspectors; the target inspection task can be understood as a pre-configured inspection work plan assigned to the target account that needs to be executed within a specified time period, and the target inspection task may contain multiple inspection items; the target inspection task can be a task that is automatically released daily or a task that is temporarily released in response to management operations, and there is no limitation here.

[0023] A patrol project can be understood as a video patrol project with independent inspection content, divided according to business type or regulatory area. The patrol project includes the content name (for example, in the airport patrol scenario, the content name of a patrol project can be the pre-security check area), the type (for example, process node supervision), the suggested time (for example, 5 seconds), and the inspection content (for example, the inspection content description for the pre-security check area is ① whether the staff are on duty, whether the service is standardized, and whether there is any refusal of passengers ② whether the number of people queuing for each pre-security check gate in international and regional areas is greater than 8, and whether the number of people queuing for each pre-security check gate in domestic areas is greater than 15 ③ whether the queuing is orderly, whether it obstructs the passage of people and the flow of passengers, and whether the staff maintains the queuing order).

[0024] Different inspection projects correspond to different monitoring equipment information. For example, monitoring equipment information can be understood as the identification, installation location, monitoring coverage, and real-time image access address of the video surveillance equipment (such as cameras) bound to the inspection project, which is used to determine the specific monitoring equipment to be called during the inspection. The suggested time can be understood as the preset duration for completing the inspection project, which is used to standardize the inspection time and ensure the inspection quality, and prevent the time from being too long or too short.

[0025] Specifically, when inspection personnel log into the inspection system using a target account, in response to the login operation of that target account, they obtain the target inspection task pre-assigned to that target account and corresponding to the current time. In practice, this target inspection task is usually configured by the management personnel during the task planning stage. The target inspection task contains multiple inspection items, and each inspection item is bound to monitoring equipment information and a configured suggested time.

[0026] By determining the monitoring equipment information and preset suggested time for each inspection project, a data foundation is provided for subsequent inspections of each inspection project. By associating and matching inspection tasks with the target accounts corresponding to the inspection personnel, the inspection personnel do not need to manually search for or select inspection tasks.

[0027] In practical applications, airport inspection tasks typically have strong time-sensitive requirements. For example, they may require focused inspections of key areas such as boarding gates and baggage carousels before and after flight landings or during peak travel periods. Therefore, this embodiment, based on these time-sensitive characteristics, responds to the login operation of the target account and obtains the target inspection task corresponding to the current time. The inspection period of the target inspection task matches the current time, and this inspection period is determined by the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. For time-sensitive target inspection tasks, a dynamic triggering mode can be used. In this mode, the inspection period corresponding to the task is automatically calculated and generated based on real-time accessed external operation plan data (including dynamically accessed flight operation plans) and preset time offset rules.

[0028] The preset time offset rule is not a fixed time constant, but a dynamically adjustable rule base pre-configured according to the actual business logic of the airport. For example, for the "incoming baggage carousel inspection task", its triggering time is not simply set to be executed immediately after the flight lands, but can be calculated based on multiple factors such as "flight planned landing time + taxiing time offset + baggage conveyor start delay + preset division window".

[0029] By setting time offset rules, the inspection period of the inspection task can be accurately matched with the actual operation status of the airport, avoiding inspection delays or resource waste caused by the disconnect between the fixed schedule and the dynamic operating environment, thereby effectively improving the timeliness of inspection tasks and the efficiency of resource scheduling.

[0030] For example, in a scenario involving airport terminal inspections, the target account can be "Inspector A." The system automatically assigns a "T2 Terminal Morning Peak Flow Node Supervision Task" (i.e., the target inspection task) corresponding to the current time to this account. This task includes three inspection items: "Check-in Area Inspection," "Security Check Area Inspection," and "Waiting Area Inspection." The system determines that the monitoring equipment bound to the "Check-in Area Inspection" item is "Check-in Area Monitoring Equipment 1-10," with a suggested time of 2 minutes; the monitoring equipment bound to the "Security Check Area Inspection" item is "Security Check Lane Monitoring Equipment 1-8," with a suggested time of 3 minutes; and the monitoring equipment bound to the "Waiting Area Inspection" item is "Waiting Hall A Area Monitoring Equipment 1-20," with a suggested time of 2 minutes.

[0031] For the "incoming baggage carousel inspection task", the system can access dynamic flight operation data in real time. For example, when the estimated arrival time of an incoming flight A is 16:00, the estimated trigger time can be determined by automatically obtaining the preset "taxiing time offset" (e.g., 8 minutes for Terminal 1 and 12 minutes for Terminal 2) and "baggage conveyor start delay" (e.g., 5 minutes). Then, combined with the preset division window (e.g., the preset division window is a time window consisting of 10 minutes before the estimated trigger time and half an hour after the estimated trigger time), the inspection period of the inspection task can be calculated.

[0032] For example, if inbound flight A corresponds to Terminal 1, the estimated trigger time is 16:13, and the inspection period is 16:03-16:43. Furthermore, if inbound flight A is delayed, the system will recalculate and dynamically update the inspection period for this task based on the updated estimated arrival time, ensuring that the inspection task remains synchronized with actual operational events.

[0033] The task-driven video inspection method provided in this specification automatically triggers the acquisition of target inspection tasks through login operations, avoiding the tedious process of inspection personnel manually searching for tasks. At the same time, by using pre-bound monitoring equipment information and suggested time, it lays a data foundation for subsequent standardized and efficient video inspections, improving the automation level and execution efficiency of video inspections.

[0034] In one or more embodiments of this specification, the inspection type of the target inspection project is any one of process node supervision, regional supervision, periodic supervision, and spot check; The process node inspection refers to the inspection of process nodes divided according to the flow of traffic; the area inspection refers to the inspection of the area divided according to the area of ​​the supervised site; the periodic inspection refers to the inspection with fixed inspection frequency requirements; and the spot check refers to the inspection without fixed frequency.

[0035] Specifically, process node inspection refers to a patrol method that sequentially checks key nodes or bottlenecks according to personnel flow lines (such as passenger entry and exit flow lines, tourist tour routes), focusing on monitoring traffic order and the operational status of key areas. Regional inspection refers to a patrol method that comprehensively inspects facilities, equipment, environmental sanitation, safety conditions, etc., within a specific area according to the geographical division of the supervised site, focusing on comprehensive inspection of various patrol items within the area.

[0036] Regular inspections refer to a patrol method that involves repetitive checks of high-risk areas or important operational resources at a fixed frequency (e.g., daily), focusing on continuous monitoring of the status of critical facilities. Spot checks refer to an inspection method without a fixed frequency, initiated flexibly by patrol personnel based on on-site operational conditions, focusing on quality control and internal control supervision of operating procedures, construction activities, etc.

[0037] Specifically, the inspection projects are configured as one of the following types: process node inspection, regional inspection, periodic inspection, or spot check. Different types of projects correspond to different organizational logics and execution rules. For example, process node inspection connects different process nodes according to the personnel flow, with each process node being equipped with a fixed camera, which is suitable for monitoring traffic order during peak hours or in key areas. Regional inspection is divided into areas, combining multiple cameras and inspection projects within a region, which is suitable for comprehensive inspections of specific areas.

[0038] For example, in an airport setting, a process node inspection project could require checking each node sequentially along the passenger flow path: "Check-in Area—Security Check Area—Waiting Area—Boarding Gate." Each node would be linked to a corresponding camera and inspection content to monitor passenger flow during peak hours. When inspecting the "Waiting Hall B" area, this area could be covered by multiple inspection projects targeting fire safety facilities, shops, restrooms, and other areas, providing a comprehensive inspection of the area.

[0039] A regular inspection-type patrol project could involve setting up three fixed patrols per day for high-risk areas such as the "power supply room" to continuously monitor the operating status of electrical equipment. A spot-check type patrol project could be initiated randomly by management personnel, such as conducting ad-hoc checks on the "operational procedures for baggage sorting area personnel."

[0040] By dividing inspection tasks into four types—process node supervision, regional supervision, regular supervision, and spot checks—video inspections can adapt to different regulatory scenarios and business needs. Different types of tasks correspond to different organizational methods, execution rules, and frequency requirements, which improves the flexibility and scenario adaptability of video inspections.

[0041] In one or more embodiments of this specification, before obtaining the target inspection task, it is necessary to bind the inspection item with the corresponding monitoring equipment information. This allows, once the inspection item is determined, the corresponding monitoring video to be obtained based on the bound monitoring equipment information, and the inspection of the inspection item to be completed. Specific implementation methods are described below: Before responding to the login operation of the target account and obtaining the target patrol task corresponding to the current time under the target account, the following steps are included: The system acquires monitoring equipment information for the target area and binds this information to patrol projects within the target area. It then determines the monitoring equipment information corresponding to each patrol project. This monitoring equipment information is configured through forward binding or reverse binding. Forward binding involves binding monitoring equipment information to patrol projects, while reverse binding involves binding monitoring equipment information to corresponding patrol projects. The attribute information for each patrol project includes the project name, patrol type, suggested time, and inspection content.

[0042] The target area can be understood as the physical space where video surveillance needs to be implemented, such as airport terminals, train station waiting halls, or shopping mall interiors—places covered by video surveillance equipment. Forward binding can be understood as a configuration method that proactively associates one or more monitoring device information with a specific inspection project; reverse binding can be understood as a configuration method that proactively associates one or more inspection projects with a specific monitoring device.

[0043] The attribute information can be understood as the metadata characteristics describing the inspection project. The project name is the content name in the above embodiment. The inspection type is used to distinguish the business classification of the inspection project. The suggested time is used to standardize the inspection time of the inspection project. The inspection content is used to describe the specific inspection items and qualification standards that need to be verified for the inspection project.

[0044] Specifically, during the configuration phase, monitoring equipment information within the target area is obtained, including equipment number (i.e., identifier), physical installation location, and monitoring coverage. Based on the regulatory characteristics and business needs of the target area, and after identifying and defining multiple patrol items, the association between monitoring equipment information and patrol items is established through forward or reverse binding.

[0045] In practice, forward binding is suitable for scenarios configured according to business processes. That is, when creating or editing a patrol project, a monitoring device suitable for monitoring the inspection content of the project is selected from the monitoring device library and associated with it. Reverse binding is suitable for scenarios managed by device. That is, when viewing or configuring a monitoring device, the monitoring device is associated with one or more patrol projects that need to be inspected by it.

[0046] For example, when an administrator creates a patrol project called "Check-in Area Inspection", they can select multiple monitoring devices covering the check-in area from the monitoring equipment library and set a suggested time of 2 minutes. The inspection content includes "whether the check-in counters are open normally, whether the queuing order is good, and whether the equipment screens are displaying normally".

[0047] Alternatively, when configuring monitoring equipment for a public area on a certain floor, the management personnel can simultaneously bind the monitoring equipment to three inspection items: "fire exit inspection", "sanitation inspection" and "facility integrity inspection", and set different suggested time and inspection content for each item.

[0048] In practical applications, once a patrol project is defined or configured, its corresponding attribute information can be obtained, including but not limited to project name, patrol type, suggested time, and inspection content. Therefore, when a patrol project is bound to corresponding monitoring equipment information, the inspection of the patrol project can be carried out based on the monitoring video provided by the corresponding monitoring equipment and the inspection content corresponding to the patrol project.

[0049] The task-driven video inspection method provided in this specification uses two flexible configuration mechanisms, forward binding and reverse binding, to associate monitoring equipment information with inspection items, thus meeting the configuration requirements of different business scenarios.

[0050] In one or more embodiments of this specification, a target account may contain multiple candidate inspection tasks. The currently processed candidate inspection task is determined as the target inspection task. Given that each candidate inspection task is configured with a corresponding inspection time period, at least one candidate inspection task that can be processed at the current time is determined based on the current time and the inspection time period of each candidate inspection task, and the target inspection task is then determined from among them. Specific implementation methods are as follows: The step of responding to a login operation of the target account and obtaining the target patrol task corresponding to the current time under the target account includes: In response to a login operation of the target account, obtain multiple candidate patrol tasks corresponding to the current time under the target account; Based on the current time and the inspection period of each candidate inspection task, at least one inspection task corresponding to the current time is obtained from the plurality of candidate inspection tasks, and a target inspection task is determined from the at least one inspection task to be executed, wherein the inspection period of each candidate inspection task is determined when the corresponding candidate inspection task is created.

[0051] In this context, candidate inspection tasks can be understood as inspection tasks that have been pre-created and assigned to target accounts but have not yet been executed. Each candidate inspection task has its suggested inspection period determined at the time of its creation. The inspection period can be understood as the time range within which the inspection task is planned to be executed, such as the start and end times.

[0052] The inspection tasks to be performed can be understood as one or more candidate inspection tasks that have arrived or fallen within the inspection period and are available for immediate execution by the inspection personnel; the target inspection task can be understood as the specific inspection task to be performed at present, selected from the inspection tasks to be performed.

[0053] Specifically, when the target account logs into the inspection system, the inspection system responds to the login operation by obtaining all candidate inspection tasks pre-assigned to the account, and based on the current time, it iterates through the inspection time periods of each candidate inspection task, filters out inspection tasks whose inspection time periods cover the current time, and obtains at least one inspection task to be executed.

[0054] If there are multiple inspection tasks to be performed, one of them is selected as the target inspection task according to preset rules. For example, preset rules may include selecting the task that starts earliest in the inspection period, selecting the task with the highest priority, or allowing the inspection personnel to select it manually. This ensures that the inspection personnel perform the correct inspection task at the correct time.

[0055] In practical applications, when at least one inspection task is available, it can be displayed on the system interface of the inspection system. When the inspection personnel interact with the system interface (by clicking or speaking, etc., which is not limited here) to select an inspection task from the at least one inspection task, the selected inspection task will be used as the target inspection task.

[0056] The task-driven video inspection method provided in this specification automatically matches the current time with the inspection time period of candidate inspection tasks, filters out the inspection tasks to be executed, and determines the target inspection task. This avoids the tedious process of inspection personnel manually judging and selecting tasks, and ensures that inspection tasks are executed in an orderly manner according to the planned inspection time period, thereby improving the execution efficiency of task scheduling.

[0057] In one or more embodiments of this specification, before obtaining the target inspection task under the target account, it is necessary to formulate the inspection task according to actual needs. Specifically, when formulating the inspection task, it is necessary to configure the target account, inspection items, and inspection time period corresponding to the inspection task. Detailed implementation methods are as follows: Before responding to the login operation of the target account and obtaining the target patrol task corresponding to the current time under the target account, the method further includes: In response to a request to add an inspection task, a task identifier corresponding to the inspection task is generated, and an inspection task template is obtained; An initial inspection task is generated based on the inspection task parameters and the inspection task template, wherein the inspection task parameters include the target account, at least one inspection item, and the task triggering mode. According to the task triggering mode, the patrol period corresponding to the initial patrol task is determined. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the patrol period is automatically calculated and generated based on external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. Determine whether there is a time conflict between the patrol time period of the initial patrol task and the patrol time period of the existing candidate patrol tasks under the target account. If a time slot conflict exists, update the patrol time slot corresponding to the initial patrol task to ensure that the updated initial patrol task does not conflict with any existing candidate patrol tasks under the target account. Then, designate the updated initial patrol task as a candidate patrol task under the target account. If there is no time conflict, the initial patrol task will be identified as a candidate patrol task under the target account.

[0058] The "Inspection Task Addition Request" can be understood as an instruction to create a new inspection task. The "Task Identifier" can be understood as a unique number generated for each new inspection task, used to distinguish and retrieve different inspection tasks.

[0059] The inspection task template can be understood as a predefined task structure framework, which includes the basic field structure of the task, default configuration items and data format specifications, etc., and is used to guide the standardized generation of inspection tasks.

[0060] Patrol task parameters can be understood as the specific details that need to be configured when creating a patrol task, including the target account, at least one patrol item, and the suggested patrol time period. Time period conflict can be understood as the patrol time period of a newly created initial patrol task overlapping or intersecting with the patrol time period of an existing candidate patrol task under the same target account, resulting in the same account needing to execute two different tasks at the same time.

[0061] Specifically, administrators or personnel with task creation permissions can initiate a request to add a patrol task. The patrol system responds to the request, generates a unique task identifier for the task to be created, and retrieves the patrol task template from the system configuration.

[0062] The patrol task parameters entered by the manager are combined with the patrol task template to generate an initial patrol task. This initial patrol task will be assigned to the target account, and the patrol time period of the initial patrol task will be checked to see if there is any conflict with the patrol time period of the existing candidate patrol tasks under the target account.

[0063] If a time slot conflict exists, the system will automatically or prompt the administrator to update the patrol time slot of the initial patrol task. For example, the patrol time slot may be shifted entirely to an idle period or the task duration may be shortened until there is no time slot conflict with existing candidate patrol tasks. The updated initial patrol task will then be designated as a candidate patrol task for that target account. If there is no time slot conflict, the initial patrol task will be directly designated as a candidate patrol task.

[0064] In practical applications, when creating an inspection task, the inspection task parameters can include the task triggering mode. For example, the task triggering mode includes a fixed triggering mode and a dynamic triggering mode. When the task triggering mode is a dynamic triggering mode, the inspection period can be dynamically determined based on external operation plan data. This external operation plan data includes, but is not limited to, flight operation plans, train operation plans, ship operation plans, or configured peak passenger flow periods. In this embodiment, when applied to an airport inspection scenario, the external operation plan is a flight operation plan. When the task triggering mode is a fixed triggering mode, the inspection period entered by the user is directly used as the inspection period of the initial inspection task.

[0065] In practical implementation, when the task triggering mode is set to dynamic triggering mode, the inspection system no longer requires managers to manually input fixed inspection time periods. Instead, it can automatically query external operation plan data and, in conjunction with preset time offset rules, dynamically calculate suitable inspection time periods for execution. The inspection task will then be automatically sent during the corresponding inspection time period. It should be noted that this dynamic triggering mode can be determined based on preset triggering conditions set by managers. For example, if an airline's hourly flight volume exceeds 10 flights or the total number of international and domestic flights in the security checkpoint exceeds 30 flights, the task triggering mode will be set to dynamic triggering mode. This allows the generated inspection tasks to match actual peak operating times on-site, preventing inspection personnel from viewing empty screens during off-peak hours, thereby improving inspection efficiency and effectiveness. In other words, based on dynamically accessed flight operation plan data, inspection time periods can be automatically calculated and generated according to time offset rules, ensuring that inspection tasks are automatically issued before peak operating times arrive.

[0066] For example, in an airport scenario, an administrator creates a new patrol task for account "Inspector A," generating task ID "TASK-005," and obtains the patrol task template. By configuring the patrol task parameters in the template (target account is Inspector A, patrol item is "B Zone Fire Protection Facilities Inspection," patrol time is 09:00-09:30), an initial patrol task is obtained. If an inspection reveals that Inspector A already has a candidate patrol task with a patrol time of 09:20-09:50, the two tasks overlap. In this case, the system can prompt the administrator to update the initial patrol task's patrol time to 09:50-10:20 to avoid conflict with the existing candidate patrol task. Through administrator confirmation or publication, the updated initial patrol task is then designated as a candidate patrol task.

[0067] In another embodiment, when configuring patrol task parameters, the administrator selects "dynamic trigger mode" for the task triggering mode and sets the triggering condition to "Aviation A's check-in area has more than 10 flights per hour" and the time offset rule to "2 hours before flight departure". The patrol system automatically queries the accessed flight schedule data and finds that the Aviation A's check-in area has 12 flights per hour between 06:00 and 07:00 the next morning, meeting the triggering condition. Based on this, the patrol system dynamically calculates the patrol period to be 04:00 to 05:00 the next day (i.e., 2 hours before flight departure) and binds this patrol period to the patrol item "Aviation A's check-in area passenger queue order inspection", generating an initial patrol task. The patrol system then checks whether the dynamically generated patrol period conflicts with the existing candidate patrol tasks under the "Inspector A" account. If there is no conflict, the initial patrol task is identified as a candidate patrol task for "Inspector A". When the system arrives at 04:00 the next day, it will automatically issue the inspection task to "Inspector A". After logging in, Inspector A can view the camera footage of the check-in area of ​​Flight A during the corresponding time period, so as to carry out targeted inspections during peak flight periods.

[0068] The task-driven video inspection method provided in this specification improves the efficiency and consistency of inspection task creation by automatically generating task identifiers and standardizing and reusing task templates. Furthermore, by detecting time period conflicts, it avoids execution time conflicts of tasks under the same account, ensuring the rationality and feasibility of task scheduling.

[0069] Step 104: Select a target inspection project from the multiple inspection projects, and analyze the monitoring video corresponding to the target inspection project based on the inspection content and video analysis algorithm to obtain the analysis results.

[0070] Among them, the target inspection project can be understood as the inspection project selected from the multiple inspection projects included in the target inspection task, which is currently being inspected. Each inspection project corresponds to a specific inspection object or inspection area in the airport, such as a boarding gate, a check-in counter area, or a baggage carousel. Inspection can be understood as the process of checking the on-site operation status item by item by reviewing the monitoring video and comparing it with the preset inspection content.

[0071] Surveillance video can be understood as real-time video or recorded playback video retrieved through monitoring equipment, reflecting the current or historical on-site conditions of the monitored area. Inspection content can be understood as the pre-configured verification items (i.e., inspection items) and pass / fail standards for each inspection project, used to guide inspection personnel or the system in determining whether the on-site operational status of the area under the inspection project's jurisdiction is normal. Examples include checking whether passenger queues are blocking passageways, whether there is unattended luggage, and whether staff are on duty.

[0072] Video analytics algorithms can be understood as image recognition and processing models built based on computer vision and deep learning technologies. They can analyze surveillance videos in real time and extract preset target information, which includes at least the number of people in the area, people density, people behavior characteristics, or the presence status of specific targets. Video analytics algorithms can adaptively call corresponding algorithm instances according to different patrol projects, including but not limited to people density detection algorithms, target detection algorithms, and behavior recognition algorithms.

[0073] After determining the target inspection task, the system selects the target inspection item to be executed from the multiple inspection items included in the target inspection task, and then proceeds to the specific inspection stage of the target inspection item. When selecting one target inspection item from multiple inspection items, the selection order can be in the order of the preset arrangement of the inspection items, or it can be manually selected by the inspection personnel; there is no restriction here.

[0074] In this embodiment, a combination of preset inspection content and video analysis algorithms can be used to analyze the monitoring video corresponding to the target inspection project.

[0075] Specifically, while retrieving the monitoring video corresponding to the target inspection project, the system can invoke a matching video analysis algorithm based on the project type to perform real-time image recognition and data analysis on the video stream. This allows for the extraction of quantitative feature information related to the inspection content from the monitoring footage, serving as the basis for subsequent analysis to determine whether inspection problems exist. Through this algorithm-driven automated analysis process, reliance on manual labor is effectively reduced, improving inspection efficiency and identification accuracy.

[0076] For example, when the system issues a "gate inspection task" in a dynamically triggered mode, the user selects the inspection item corresponding to a specific gate from the multiple inspection items included in the task as the target inspection item. The system then retrieves the real-time monitoring video of that gate and automatically calls a preset crowd density analysis algorithm to process the video stream. This algorithm, based on a deep learning model, identifies and statistically analyzes the number and distribution of people in the queuing area of ​​the monitoring screen. The statistical results output by the algorithm are the analysis results described in this step, such as "The current number of people in the queuing area is X, and the queue extends to the main flow channel." This analysis result will serve as input data for comparison with the acceptance criteria in subsequent steps to determine whether there is a channel congestion problem.

[0077] For example, taking the "unattended baggage detection" target inspection project as an example, the video analysis algorithm continuously tracks baggage items in the monitoring screen. When it detects that no one approaches or claims a piece of baggage within a preset time threshold, the video analysis algorithm can output analysis results including "suspicious unattended baggage" and "location coordinates". Furthermore, in practical applications, the area where the abnormal baggage is located can be highlighted on the inspection interface, and a pop-up window can remind the user to pay attention. This "human-machine collaboration" working mode also allows users to automatically complete the analysis of the screen content without having to continuously stare at the screen for extended periods. Users only need to manually confirm and make decisions regarding the abnormal results indicated by the algorithm, significantly reducing the human burden while ensuring inspection quality.

[0078] It should be noted that, when the inspection type of the target inspection project is regional supervision, the analysis of the monitoring video corresponding to the target inspection project based on the inspection content and combined with video analysis algorithms to obtain the analysis results includes: At least one regional subtask corresponding to the target inspection project is determined. Based on the inspection content corresponding to each regional subtask and combined with video analysis algorithms, the monitoring videos corresponding to each regional subtask are analyzed to obtain the analysis results corresponding to each regional subtask. Each regional subtask corresponds to a combination of different monitoring devices within the target area where the target inspection project is located. The target inspection project is obtained by connecting and combining each regional subtask in series. The monitoring devices of each regional subtask are configured according to the splitting principle.

[0079] Among them, regional sub-tasks can be understood as multiple sub-units obtained by decomposing a complete regional supervision project (i.e., a regional supervision type of inspection task), with each sub-unit corresponding to a section within the target area.

[0080] The splitting principle can be understood as the rules for breaking down a regional task into multiple sub-tasks. These include specific requirements such as each sub-task containing different cameras, each sub-task covering all patrol items in the target area, and flexibly adjusting camera bindings according to the importance of the patrol items. The sequential combination can be understood as connecting multiple regional sub-tasks in a certain order to form a complete regional supervision task, ensuring full coverage of the target area's patrol items after executing all sub-tasks in sequence.

[0081] Specifically, when the inspection type of the target inspection project is regional supervision, at least one regional sub-task corresponding to the target inspection project can be further determined. During the inspection, the system sequentially enters each regional sub-task, obtains the monitoring video corresponding to each regional sub-task based on the monitoring equipment information bound to each regional sub-task, and analyzes the monitoring video based on the inspection content corresponding to each regional sub-task and combined with video analysis algorithms to realize the inspection of each regional sub-task and obtain the analysis results.

[0082] Since each area subtask contains different combinations of cameras, and all area subtasks together cover the inspection items of the target area, the inspection personnel complete a comprehensive inspection of the entire target area after completing all area subtasks in sequence.

[0083] This split design breaks down large-area tasks that originally required checking all cameras at once into multiple smaller checks, each checking only a portion of the cameras, but cumulatively covering all inspection items.

[0084] Suppose that a regional inspection project A within an airport terminal needs to conduct video surveillance of "T2 Terminal B Area." This area has 16 cameras deployed, and four inspection items need to be checked: the operational status of check-in counters, the queuing order at security checkpoints, the integrity of fire safety facilities, and the operational status of lighting equipment. Each inspection item is covered by four different cameras simultaneously. For example, the operational status of check-in counters is covered by cameras C01 to C04, the queuing order at security checkpoints by cameras C05 to C08, the integrity of fire safety facilities by cameras C09 to C12, and the operational status of lighting equipment by cameras C13 to C16. If all 16 cameras were inspected at once using the traditional method, 16 inspections would be required, which is time-consuming and involves a significant amount of repetitive work.

[0085] To address this issue, this embodiment divides the regional inspection project A into four sub-tasks based on the splitting principle. The first principle requires that each sub-task include different cameras to avoid the same camera being repeatedly inspected in multiple sub-tasks. Accordingly, sub-task 1 is assigned four cameras: C01, C05, C09, and C13; sub-task 2 is assigned C02, C06, C10, and C14; sub-task 3 is assigned C03, C07, C11, and C15; and sub-task 4 is assigned C04, C08, C12, and C16. The cameras in the four sub-tasks do not overlap. The second principle requires that each sub-task include all inspection items within the area; that is, sub-tasks 1 through 4 must cover the four inspection items: check-in counter operation status, security checkpoint queuing order, fire safety facilities condition, and lighting equipment operation status. The third principle allows for flexible adjustment of the binding frequency or number based on the importance of the inspection items.

[0086] In this example, the operational status of check-in counters and the queuing order at security checkpoints are high-priority inspection items, which need to be checked in each subtask. Furthermore, the operational status of check-in counters can be checked in each subtask by binding multiple related cameras to obtain a more comprehensive view. On the other hand, the integrity of fire protection facilities and the operational status of lighting equipment are general-priority inspection items. Provided that they have already been inspected, they can be bound in only some subtasks and do not need to be checked repeatedly in each subtask.

[0087] Based on the above decomposition principles, the final configuration of the four sub-tasks is as follows: Sub-task 1 is bound to all four inspection items, namely, using C01 to check the operational status of check-in counters, C05 to check the queuing order at security checkpoints, C09 to check the integrity of fire protection facilities, and C13 to check the operational status of lighting equipment. Sub-task 2 is bound to only the two most important inspection items, namely, using C02 to check the operational status of check-in counters and C06 to check the queuing order at security checkpoints, without binding fire protection facilities and lighting equipment. Sub-task 3 is also bound to only the two most important inspection items, using C03 and C07 to check the operational status of check-in counters and the queuing order at security checkpoints, respectively. Sub-task 4 is also bound to only the two most important inspection items, using C04 and C08 to check the operational status of check-in counters and the queuing order at security checkpoints, respectively.

[0088] With this configuration, the operational status of critical check-in counters is checked once from different angles by four different cameras in four sub-tasks, achieving sufficient multi-angle coverage; the queuing order of critical security checkpoints is also checked once by four different cameras; while general important fire protection facilities and lighting equipment are only checked once in sub-task 1, which is sufficient to meet basic regulatory requirements.

[0089] After the tasks are broken down, the system sequentially combines these four subtasks into a complete regional inspection task. Patrol personnel then execute subtasks 1, 2, 3, and 4 in sequence. Compared to the traditional method of inspecting all 16 cameras at once, using regional subtasks significantly improves video patrol efficiency.

[0090] The task-driven video inspection method provided in this specification solves the problem of the large number of cameras in large public places and the excessive time required for a single inspection by breaking down regional supervision tasks into multiple regional sub-tasks and executing them in series, thereby improving the execution efficiency of video inspection.

[0091] In the embodiments of this specification, when the inspection type of the target inspection project is a periodic inspection, the target inspection project usually corresponds to a fixed monitoring project in a high-risk area; while when the inspection type of the target inspection project is a spot check, spot check information can be obtained and a spot check work order can be generated. The spot check work order includes at least one of the following: spot check project, spot check content, spot check result, and monitoring device information.

[0092] High-risk areas refer to specific areas within public places that pose a high level of safety risk or have a significant impact on operational safety, such as electrical rooms, power distribution rooms, and charging rooms in airport terminals, and signal equipment rooms in train stations. Fixed monitoring items refer to specific, pre-defined inspection items that require repeated checks in high-risk areas, such as equipment operating temperature, the status of fire protection facilities, and the condition of electrical wiring.

[0093] The inspection information refers to the specific details related to the inspection recorded by the inspection personnel based on the monitoring device screen. This includes the name of the inspected project, the specific inspection items, the handling conclusions of the on-site situation, and relevant information about the inspected monitoring device. The inspection work order refers to the structured data record generated by the system based on the inspection information, used to record and track inspection activities, and can serve as the basis for subsequent quality traceability and management audits.

[0094] Specifically, when the inspection type for a target inspection project is periodic supervision, the inspection targets are typically limited to fixed monitoring items within high-risk areas. These high-risk areas, due to their involvement in electrical safety, fire safety, or the operation of critical equipment, require continuous monitoring and frequent status checks. Therefore, this target inspection project is usually configured with fixed daily inspections, meaning that inspections are conducted on the target inspection items every single day. Fixed inspection frequencies ensure that high-risk areas receive regular and continuous monitoring, preventing safety hazards from going undetected due to excessively long inspection intervals.

[0095] For example, in an airport scenario, the "Regular Inspection of Electrical Rooms" project is classified as a regular inspection, and the corresponding high-risk area is the "Electrical Room in Area B of Terminal 2." Fixed monitoring items include checking whether the distribution cabinet temperature is normal, whether the insulation mats are intact, and whether there are any abnormal noises. The system is configured with three fixed inspection times per day for this "Regular Inspection of Electrical Rooms" project, automatically issuing tasks to the on-duty inspection personnel at 08:00, 14:00, and 20:00 respectively.

[0096] For inspection projects involving random checks, inspectors conduct ad-hoc and flexible checks on the on-site operational status based on the monitoring equipment footage linked to the target inspection project. Inspectors then fill out inspection information based on the results, including confirming which project was inspected, the specific details checked, and whether the on-site condition is normal or requires further attention. After completion, the system generates an inspection work order based on this information.

[0097] As an official record of this inspection project, the inspection work order can be stored in the system for managers to view and review. The monitoring device information in the inspection work order makes it easy to trace the source of the footage during the inspection. If the inspection results are found to be inconsistent with the actual situation, the historical recordings of the corresponding cameras can be played back for verification.

[0098] By regularly inspecting fixed monitoring projects within high-risk areas, continuous and regular monitoring of key safety areas was ensured, guaranteeing timely confirmation of the safety status of important operational resources and sensitive areas, and improving the level of safety management in public places. By generating inspection work orders for random inspection projects, standardized recording and traceable management of inspection activities without fixed frequency were achieved, providing complete data support for subsequent quality audits and accountability, and improving the standardization of inspection work.

[0099] Step 106: Compare the analysis results with the qualification standards in the inspection content, and generate the inspection results of the target inspection project based on the comparison results.

[0100] The pass / fail criteria can be understood as quantitative thresholds or qualitative rules set in advance in the inspection content to measure whether the on-site operation status of the inspected project meets the standards. For example, a queue of more than 20 people is considered a congestion, or luggage left unattended for more than 10 minutes in a specific area is considered abnormal.

[0101] The inspection result can be understood as a judgment conclusion on the current status of the target inspection item after logically comparing the objective data output by the video analysis algorithm with the qualification standard. The inspection result can be "normal / qualified" or "abnormal / unqualified", and can further include structured information such as the specific type of abnormality, the location of occurrence and the severity, to guide subsequent alarm push or manual handling.

[0102] Specifically, objective analysis results of the target inspection project are obtained through video analysis algorithms. Since the analysis results are only objective data extraction from the on-site images (such as the number of people, the status of items, etc.), they are not directly equivalent to the inspection conclusions. Therefore, they need to be automatically compared with pre-configured qualification standards. In practice, the qualification standards corresponding to the target inspection project can be read, and the quantitative features or status indicators in the analysis results can be matched with standard thresholds or rules.

[0103] If the analysis results meet or exceed the qualified standards, the current status of the target inspection item is determined to be normal; if the analysis results exceed or do not meet the qualified standards, it is determined to be abnormal, and an inspection result containing abnormal details is generated.

[0104] Of course, in practical applications, patrol personnel can also check the monitoring videos corresponding to the target patrol items, and mark, describe or report any abnormalities. By combining the abnormalities determined manually with the analysis results obtained from the algorithm, the patrol results can be determined comprehensively.

[0105] For example, in a boarding gate queue detection scenario, the video analytics algorithm outputs the result: "The current queue size is 28 people, and the queue extends into the main flow channel." Given that the pass / fail standard for this boarding gate inspection item is "the number of people in the queue ≤ 20 and the queue does not obstruct the main flow channel," comparing the analysis result with the standard reveals that 28 people > 20, and the queue status does not meet the requirements. Therefore, the comparison result is "non-compliant." Based on this, the system generates the inspection result for this target inspection item as "Abnormal: Boarding gate queue congestion" and automatically triggers an alarm to notify on-duty personnel to manage the queue.

[0106] Once the inspection of a target inspection item is completed and the recommended time for that target inspection item has been reached, the system allows the inspection personnel to select the next target inspection item from the remaining inspection items and continue to perform the inspection until all inspection items are completed.

[0107] In one or more embodiments of this specification, corresponding monitoring video can be obtained based on the monitoring equipment information corresponding to the target inspection item. Based on this monitoring video and the inspection content corresponding to the target inspection item, the target inspection item can be inspected. Therefore, it is necessary to obtain the inspection content corresponding to the target inspection item before performing the inspection, thereby achieving accurate inspection. Specific implementation methods are as follows: Before inspecting the target inspection project based on the inspection content and the corresponding monitoring video of the target inspection project, and obtaining the inspection results of the target inspection project, the method further includes: The inspection content corresponding to the target inspection project is determined according to the preset configuration information, wherein the inspection content includes inspection items for inspecting the on-site conditions and corresponding qualification standards.

[0108] The preset configuration information can be understood as a set of parameters and rules for the inspection task that are pre-set and stored during the system initialization or task issuance phase. It is used to define the business logic that different inspection items need to follow when performing inspections and serves as the data basis for the system to automatically match and load the corresponding inspection rules.

[0109] The inspection content can be understood as pre-set, specific inspection items and compliance standards used to guide patrol personnel in verifying the on-site conditions, such as whether facilities and equipment are operating normally, whether personnel are working in accordance with regulations, and whether fire exits are unobstructed. Inspection items can be understood as specific verification dimensions or observation objects set for the target patrol project, such as "passenger queue length," "passage unobstructedness," "staff on-duty status," or "whether there are any leftover items." Compliance standards are the benchmark conditions corresponding to the inspection items, used to determine whether the on-site conditions are compliant. These can be quantitative thresholds (such as the maximum number of people queuing, the maximum time items can remain in the queue) or qualitative rules (such as whether specific personnel are wearing prescribed uniforms, or whether there are obstacles in a specific area), and are not limited here.

[0110] Specifically, based on the monitoring equipment information associated with the target inspection project, the video stream of the corresponding monitoring equipment is obtained through a video access protocol and displayed on the inspection interface. This allows inspection personnel to verify the on-site conditions item by item based on the monitoring video, enabling them to judge and inspect the actual on-site situation. Of course, when using video analysis algorithms for inspection, it is also necessary to analyze the monitoring video based on the inspection content.

[0111] Before selecting a target inspection item and preparing to conduct the inspection, it is necessary to first clarify "what to check" and "what constitutes a pass / fail result". Since inspection tasks in complex scenarios such as airports usually involve multiple different types of items, this embodiment dynamically and accurately matches specific inspection content to the target inspection item by reading preset configuration information.

[0112] In practice, when the inspection content includes multiple inspection items, for each inspection item, it can be confirmed whether the on-site situation presented by the monitoring video meets the preset qualification standards, and the inspection results (such as normal, abnormal or need to be reported) are recorded in the system.

[0113] When the inspection time of the target inspection item reaches the preset recommended time, the system allows or prompts the inspection personnel to end the inspection of the current target inspection item and return to continue to execute the step of selecting the next target inspection item, that is, to select the next inspection item to be inspected from the remaining inspection items as the target inspection item.

[0114] For example, when the system selects "Security Checkpoint B" as the target inspection item, the system matches the inspection content of that item based on preset configuration information. These inspection items include "Checkpoint Open Status" and "Queue Order." For the "Queue Order" inspection item, the system's acceptable standard is that "the queue must not occupy the main passage, and the number of people in the queue must not exceed 20." The system uses these inspection items and acceptable standards as the direct basis for subsequent target detection and comparison in the video analysis algorithm.

[0115] In fact, in the airport scenario, the target inspection task includes three inspection items: "check-in area inspection", "security check area inspection" and "waiting area inspection", with recommended time of 2 minutes, 3 minutes and 2 minutes respectively.

[0116] Select "Check-in Area Inspection" as the target inspection item. The "Check-in Area Inspection" item is linked to the monitoring equipment information of cameras 1-10 on the check-in island. The system obtains the inspection images from these cameras and determines that the inspection content of the "Check-in Area Inspection" item includes: "Whether the check-in counters are open normally", "Whether the queuing order is good", and "Whether the equipment screens are displaying normally".

[0117] By reviewing the inspection footage from each camera in turn and verifying each item against the inspection criteria, it was confirmed that all check-in counters were open, the queues were orderly, and the screen displays were normal. The system then selected "Normal" for this task. After two minutes of inspection, the system automatically ended the task and selected "Security Check Area Inspection" as the next target. The system then retrieved camera footage from the security checkpoint for inspection. After completing a three-minute inspection of the security check area, the system moved on to "Waiting Area Inspection."

[0118] The task-driven video inspection method provided in this specification automatically acquires monitoring video through monitoring equipment information and verifies each item based on the inspection content determined by preset configuration information. This achieves standardization and normalization of video inspection, ensuring that each inspection has clear visual evidence and uniform inspection content reference, reducing subjective judgment differences, and improving the accuracy and consistency of inspection results.

[0119] Step 108: If the corresponding suggested time is reached, continue to execute the step of selecting the target inspection project from the multiple inspection projects, and if the inspection results contain inspection problems, send the inspection problems to the alarm system.

[0120] Inspection issues can be understood as specific matters discovered by inspection personnel during the inspection of target items, through the inspection screen, that do not meet the preset inspection standards, or involve safety hazards, equipment malfunctions, operational violations, or other abnormal situations. The alarm system is used to receive, record, assign, and track various inspection issues.

[0121] In practical applications, when the target inspection item reaches the corresponding recommended time, the inspection of that target inspection item ends, and the next inspection item is automatically selected in sequence, or the next inspection item can be manually selected for execution. When inspection issues are sent to the alarm system, information such as the issue type, area location information, fault description, reporter information, issue tracking time limit settings, and on-site screenshots can be included to facilitate subsequent processing by the alarm system.

[0122] Specifically, when patrol personnel discover anomalies through monitoring video during the inspection of a target inspection project, they generate an inspection issue by filling in the issue type, area location information, fault description, associated monitoring equipment information, and screenshots or video clips from the scene. The inspection issue is then packaged according to a preset data format and sent to the alarm system via API calls or message queues. After sending, the patrol system can continue to perform the remaining inspections of the target inspection project or switch to the next inspection project.

[0123] In this embodiment, to further improve patrol efficiency and reduce patrol personnel fatigue, the patrol system can integrate intelligent video analytics technology to achieve collaborative patrols between artificial intelligence and human personnel. Specifically, during the patrol personnel's execution of the target patrol project, the intelligent video analytics module can be used to monitor and analyze the currently retrieved surveillance video footage in real time, automatically identifying abnormal situations in the footage, such as items improperly placed on counters, blackout flight display screens, unclaimed luggage, and abnormal queuing.

[0124] When the intelligent video analytics module detects a suspected problem, the patrol system can automatically alert patrol personnel and provide feedback on the abnormal footage and preliminary analysis results. Upon receiving the alert, patrol personnel verify the suspected problem. If the problem is confirmed, they can report it to the alarm system with a single click; if the identification is incorrect, the patrol personnel can ignore the alert. This approach frees patrol personnel from the tedious task of visually searching for anomalies throughout the entire process, allowing them to focus on verification and problem handling, thereby effectively improving patrol efficiency and reducing personnel fatigue.

[0125] After receiving an inspection issue, the alarm system can automatically or manually assign it to the appropriate personnel based on the issue type and urgency, thus initiating the issue handling process.

[0126] For example, in an airport scenario, when patrol personnel are performing the "Check-in Area Inspection" project, they may find that the display screen of a check-in counter is black, which does not meet the standard of "equipment screen display is normal" in the inspection content. The patrol personnel click "Report Problem" in the system, select "Equipment Failure" as the problem type, fill in the fault description as "Check-in Island Counter No. 3 display screen is black", and attach a screenshot of the black screen as an attachment, generate an inspection problem work order, and send the inspection problem work order to the airport repair system (i.e., alarm system).

[0127] In fact, the target inspection task can determine the task time based on the suggested time for its corresponding multiple inspection items. When the task time corresponding to the target inspection task is reached, the next task can be selected as the target inspection task from the candidate inspection tasks under the target account.

[0128] By controlling the suggested time, the inspection pace of each inspection item was standardized, preventing the overall task progress from being affected by excessive time spent on a single inspection, and ensuring the efficiency and quality of the inspection work.

[0129] In one or more embodiments of this specification, the inspection duration is displayed on the inspection interface. When the inspection duration reaches the corresponding suggested time, the inspection of the target inspection item ends; or, if the inactivity time on the inspection interface corresponding to the target inspection item exceeds a preset threshold, the target inspection task automatically exits and resumes execution upon re-entry. Specific implementation methods are as follows: The step of selecting a target inspection project from the plurality of inspection projects when the corresponding suggested time has been reached includes: When the inspection time reaches the suggested time in the inspection interface corresponding to the target inspection project, an end reminder message will be displayed; In response to the inspection end command, the inspection of the target inspection item ends, and the step of selecting the target inspection item from the multiple inspection items continues to be executed, wherein the inspection interface displays the monitoring video corresponding to the monitoring equipment information; And / or, Perform useless time detection; if the inactivity time on the inspection interface corresponding to the target inspection item exceeds a preset threshold, automatically exit the target inspection task. Upon re-entering the target inspection task, the inspection resumes from the target inspection item where the task was paused upon exiting, and if the corresponding recommended time is reached, the inspection of the target inspection item ends, and the step of selecting a target inspection item from the multiple inspection items continues.

[0130] The inspection duration can be understood as the time elapsed from the start of the inspection of the current target project to the current moment; the inspection interface can be understood as the system interaction page used to display monitoring video, inspection content, and operation controls; the end reminder message can be understood as a pop-up reminder to the inspection personnel that the current target project has reached the suggested time.

[0131] Inactive time detection can be understood as a monitoring mechanism by which the system checks whether patrol personnel are performing effective operations on the inspection interface, used to determine whether the patrol personnel are inactive. The preset threshold can be understood as the critical inactivity duration that triggers automatic exit; for example, a preset threshold of 3 minutes.

[0132] Specifically, during the inspection of a target inspection item, the inspection system continuously times the inspection duration of the current target inspection item. When the inspection duration reaches the preset recommended time for the target inspection item, a reminder pop-up can be displayed on the current inspection interface. This reminder pop-up can contain a "Execute Next Inspection" control. When the inspection personnel click this control, an inspection end command is triggered, thereby ending the inspection of the current target inspection item and continuing to select the next target inspection item from the remaining inspection items, thus realizing the sequential advancement of each inspection item.

[0133] Building upon this foundation or as a standalone solution, the inspection system also performs inactive time detection: real-time monitoring of the operator's actions on the inspection interface, including mouse movement, clicks, and keyboard input. If no operation is detected within a preset threshold time, the system determines that the operator is inactive, automatically exits the current target inspection task, and saves the task progress.

[0134] When the inspector re-enters the target inspection task, the system automatically locates the target inspection item where the exit was completed, restores the inspection interface, and allows the inspector to continue inspecting the target inspection item. After the inspection time reaches the recommended time, the target inspection item ends and the next item is selected.

[0135] The task-driven video inspection method provided in this specification standardizes the inspection rhythm and activity level of inspection personnel through two mechanisms: suggested time control and useless time detection. Suggested time ensures that each inspection item receives a reasonable inspection time, avoiding the impact on the overall task progress due to excessive time taken by individual items; while useless time detection and automatic exit and recovery mechanisms ensure the continuity of task progress.

[0136] In one or more embodiments of this specification, a re-inspection time can be set for the detected inspection problems, thereby enabling the tracking of inspection problems. Specific implementation methods are described below: After obtaining the inspection problem corresponding to the target inspection project and sending the inspection problem to the alarm system, the method further includes: Set a review time for the inspection problem, and if no repair response is received from the alarm system and the review time has arrived, trigger a tracking reminder for the inspection problem; Upon receiving a repair response from the alarm system, the problem status of the inspection issue is marked as resolved.

[0137] The review period can be understood as the expected timeframe for rectification of an inspection problem set by the inspection personnel when reporting the problem, and is used to follow up on the implementation of the rectification of the inspection problem.

[0138] A repair response can be understood as a confirmation message returned by the alarm system to the inspection system after a problem has been repaired, indicating that the problem has been resolved. A follow-up reminder can be understood as a notification sent by the inspection system to relevant inspection personnel or managers after the scheduled re-inspection time, prompting them to verify the rectification status of the inspection problem. This follow-up reminder can be displayed via a pop-up window. The problem status indicates the processing stage of the inspection problem, including "Pending," "Processing," and "Resolved."

[0139] Specifically, when inspection issues are sent to the alarm system, the inspection personnel can set a review time for the reported issues. The review time can be determined based on the urgency of the issue and the difficulty of rectification, and there is no limit to this.

[0140] The inspection system records the review time and continuously monitors for the repair response returned by the alarm system. If the inspection system receives the repair response from the alarm system before the review time arrives, it marks the problem status of the inspection issue as resolved, indicating that the inspection issue has been rectified and no further follow-up is required.

[0141] If the inspection system has not received a rectification response for the inspection issue by the scheduled review time, a follow-up reminder will be triggered, notifying the inspection personnel to review the issue. Upon receiving the follow-up reminder, the inspection personnel can retrieve the corresponding inspection footage through the inspection system to verify the rectification status. If the issue is confirmed to be resolved, the inspection personnel can manually turn off the follow-up reminder; if the issue remains unresolved, the review time can be adjusted, and follow-up monitoring can continue until the issue is resolved.

[0142] The task-driven video inspection method provided in this specification achieves tracking and supervision of the rectification of inspection problems by setting review time and monitoring the repair response of the alarm system. Furthermore, by providing tracking reminders, it reduces the repetitive verification work of inspection personnel and strengthens the closed-loop management capability for problems in inspection scenarios.

[0143] In one or more embodiments of this specification, upon completion of a target inspection task, inspection data can be collected to control the inspection quality. Specific implementation methods are described below: After obtaining the inspection problem corresponding to the target inspection project and sending the inspection problem to the alarm system, the method further includes: Upon completion of the target inspection task, an inspection quality control table is generated based on the inspection data. The inspection quality control table includes at least one statistical component from the following: task execution log, problem management log, and employee quality statistics. The task execution log includes at least one data point from the following: task completion progress, number of problems found, executor, and actual execution time. The problem management log includes at least one data point from the following: inspection problem description and inspection problem status. The employee quality statistics include at least one data point from the following: effective inspection time statistics, problem reporting statistics, and number of incomplete tasks statistics.

[0144] Among them, the inspection data can be understood as all relevant information automatically collected by the inspection system during the execution of the target inspection task, including the inspection duration of each inspection item, the actual inspection time, the problems found, and the completion rate of the project.

[0145] The inspection quality control form can be understood as a statistical report generated based on inspection data, used to evaluate and improve the quality of inspection work. Through the summarization and analysis of multidimensional data, it reflects the execution of inspection tasks and the work performance of inspection personnel.

[0146] The task execution log records the details of each inspection task, used to track task progress and completion quality. The issue management log records detailed information and current processing status of all reported inspection issues, used for issue tracing and statistical analysis. Employee quality statistics summarize work performance indicators by inspection personnel, used for personnel evaluation and work quality assessment.

[0147] Specifically, once the inspection personnel have completed the entire target inspection task (or inspection project), the inspection system automatically collects all inspection data generated during the execution of this target inspection task (or inspection project). The inspection system then categorizes, summarizes, and statistically calculates this inspection data to generate an inspection quality control table.

[0148] The inspection quality control form may include one or more statistical components from task execution logs, problem management logs, and employee quality statistics. The task execution log records the percentage of task completion, the total number of problems found during the task, the name or account of the inspector who performed the task, and the total time actually consumed. The problem management log records a detailed description of each inspection problem and its current processing status. Employee quality statistics provide performance indicators such as the inspector's effective inspection time, the number of reported problems, and the number of unfinished tasks, either by time period or by task. The generated inspection quality control form can be used by managers for quality audits, problem tracing, and personnel performance evaluation.

[0149] For example, in an airport scenario, an inspector completed the "T2 Terminal Morning Peak Process Node Supervision Task." The system collected data showing that all three inspection items for this task were completed, with a completion rate of 100%, and one issue was found (the display screen at check-in island counter 3 was black). The executor was "Inspector A," and the actual total execution time was 7 minutes and 30 seconds (the recommended total time is 7 minutes). The system generated a task execution log in the inspection quality control table, recording the above data. The issue management log recorded the issue description "the display screen at check-in island counter 3 was black" and its current status "resolved." Employee quality statistics, based on this task, showed "Inspector A's" effective inspection time of 7 minutes and 30 seconds, 1 reported issue, and 0 incomplete tasks.

[0150] The task-driven video inspection method provided in this specification automatically generates an inspection quality control table after the task is completed, thereby realizing quantitative management and quality traceability of inspection work and improving the management refinement and quality control capabilities of video inspection.

[0151] The following is in conjunction with the appendix Figure 2 Taking the application of the task-driven video inspection method provided in this specification in an airport inspection scenario as an example, the task-driven video inspection method will be further explained. Figure 2 This diagram illustrates the processing flow of a task-driven video inspection method according to an embodiment of this specification.

[0152] The data processing system corresponding to this task-driven video inspection method consists of video surveillance devices, a patrol system, and a data transmission network. The front-end video surveillance devices achieve comprehensive coverage of public places, and the video signals from the video surveillance devices are transmitted to the patrol system in real time through the data transmission network. The patrol system then inspects all areas covered by the surveillance devices.

[0153] To achieve standardized and efficient video inspection, a series of metadata configuration tasks must first be completed. Starting with metadata configuration, the process proceeds through project type configuration, inspection content configuration, camera data maintenance, and inspection account configuration, followed by project and camera configuration. After completing the metadata configuration, the subsequent processes of task formulation, task execution, and problem reporting and tracking are carried out.

[0154] The project type configuration includes configuring the inspection types of the project, which include four categories: process nodes, regional supervision, regular supervision, and spot checks. Based on the different characteristics and focuses of public place supervision, the video supervision element perception points, i.e., inspection points, are sorted out, and the supervision content is configured for each inspection point, and inspection projects are generated. The inspection project includes the content name, type, suggested time, and inspection content description.

[0155] For example, by conducting a regulatory review of the airport, the inspection points include the pre-security check area. At this time, the inspection items are generated by configuring the inspection content of the pre-security check area as follows: "Content Name: Pre-security Check Area; Type: Process Node Inspection; Suggested Time: 5 seconds; Inspection Content Description: ① Whether the staff are on duty, whether the service is standardized, and whether there is any refusal of passengers ② Whether the number of people queuing for each pre-security check gate in international and regional areas is greater than 8, and whether the number of people queuing for each pre-security check gate in domestic areas is greater than 15 ③ Whether the queuing is orderly, whether it obstructs the passage of people and the flow of passengers, and whether the staff maintains the queuing order."

[0156] In practical applications, configuration or management personnel can manage inspection projects, such as displaying them in a tree-like node structure. See also... Figure 3 , Figure 3 This document illustrates a schematic diagram of an inspection project management system provided in one embodiment. Specifically, inspection projects are organized according to inspection type, forming primary and secondary categories. The primary category refers to the aforementioned inspection types, including process node supervision, regional supervision, periodic supervision, and spot checks. The secondary category refers to the information on regulatory area nodes, i.e., specific inspection projects. Multiple secondary categories constitute the primary category. Specifically, users can customize and create secondary categories based on the characteristics of the location.

[0157] For example, under "Process Nodes", you can set up "T1 Joint Inspection Area", "T1L3 Boarding Gate", "Manual Check-in Counter" and multiple process department nodes (such as Process Department 1, Process Department 2); under "Regional Supervision", you can set up secondary categories such as "T2L4 Office Area Corridor" and "T2 Remote Stand Boarding Gate". Similarly, you can set up corresponding secondary categories under regular supervision and spot check items.

[0158] Regarding camera (i.e., surveillance equipment) data maintenance, such as Figure 4 As shown, Figure 4This specification illustrates a processing diagram of a camera configuration provided in one embodiment. Specifically, the camera configuration requires the intelligent security integration system (i.e., the patrol system in the above embodiment) to work collaboratively with the video surveillance system and map service. The video surveillance system is used to manage ledger information, and the map service is used to implement map service construction.

[0159] In practice, the intelligent security integration system requests camera log data from the video surveillance system daily to obtain basic camera log information, including camera identifiers and IP addresses. After the data is stored in the database, the intelligent security integration system further requests floor and location data of the cameras from the map service based on the camera identifiers and IP addresses. Once the floor and location data are obtained from the map service and the data is stored in the database, the daily device information synchronization is complete. Subsequently, the system interface of the intelligent security integration system can perform extended information self-maintenance and device display.

[0160] When it is necessary to use the monitoring footage of the camera for inspection, the intelligent security integration system can request the corresponding video stream data (i.e., monitoring video) from the video surveillance system based on the camera's UUID (Universally Unique Identifier, i.e., the camera identifier in the above embodiment). After obtaining the video stream data, the system can perform business assembly and video playback according to the inspection items.

[0161] Regarding the configuration of inspection accounts, it is confirmed that inspection accounts can be created and account permissions can be set. Once the inspection content configuration is completed and the inspection projects are obtained, and the camera configuration is also completed, the binding configuration between inspection projects and cameras (i.e., project and camera configuration) can be achieved. Specifically, depending on the actual business scenario and user habits, there are two implementation schemes: project-based "task configuration" (i.e., forward binding in the above embodiment) and camera-based "camera configuration" (i.e., directional binding in the above embodiment). These will not be elaborated upon here.

[0162] During the task planning phase, the planning of patrol tasks mainly involves configuring patrol accounts, patrol time periods, and patrol project types. Specifically, based on the selected patrol projects, the total control time and number of cameras for the corresponding patrol task can be automatically calculated.

[0163] See Figure 5 , Figure 5 The illustration shows a schematic diagram of the process of adding a new patrol task according to an embodiment of this specification. When a new patrol task is added, a task number (i.e., task identifier) ​​is automatically generated first. The administrator can complete the task name and set the target account, that is, the new patrol task will be assigned to the target account.

[0164] The inspection system can automatically retrieve the first-level inspection type (such as process node supervision, area supervision, etc.) and supports selecting multiple inspection types. After selecting the first-level inspection type, the system automatically loads the optional inspection items. For example, when selecting process node supervision, the system automatically loads multiple inspection items under that process node supervision (multiple inspection items under process node supervision refer to inspection items of the process node supervision type).

[0165] When selecting inspection projects, all lower-level third-level types (i.e., sub-tasks under the inspection project) are selected by default. After the administrator selects inspection projects, they configure the suggested time period for each inspection project. Since each inspection project has a corresponding suggested time and number of cameras, when multiple inspection projects are selected, the total control time (i.e., suggested task time) and number of cameras for the newly added inspection task can be calculated in a linked manner. Finally, the task date range and suggested task time are finalized. Before saving and exiting, the system will perform a date conflict check. If there are no date conflicts, the inspection task is completed.

[0166] It should be noted that, to improve the ease of use for system users, the system includes copy and drag-and-drop sorting functions for the selection of inspection projects. After an inspection task is created, its status is updated based on whether it is started, manually terminated, automatically terminated upon expiration, or copied. For example, an inspection task can change from an unstarted state to a started or terminated state.

[0167] Regarding task execution, inspection tasks can be manually initiated by administrators or automatically generated by the background server within the planned date range. After the inspection task is generated, the inspection personnel log in to the system, respond to the login operation of the target account, obtain the target inspection task under the target account, and execute the inspection of the target inspection task to achieve work monitoring.

[0168] See Figure 6 , Figure 6 This specification illustrates a schematic diagram of task execution according to an embodiment. Specifically, after a task is generated manually or automatically, and the inspector selects a seat to enter (i.e., logs into the target account) and starts the task, the inspection system can automatically resume the task from the previous interrupted position. After starting a sub-task inspection (which can be a sub-task under a regional supervision project or an inspection project), if a problem is found, the problem is reported and tracked; if all tasks are completed, the task ends; if not all tasks are completed and not forcibly ended, the next task is executed. During the inspection, the system can choose to forcibly end the task execution.

[0169] The specific process for executing sub-task inspections is as follows: Figure 7 As shown, Figure 7A schematic diagram of a subtask inspection process provided in one embodiment of this specification is shown.

[0170] Specifically, when a patrol project is selected, if the patrol project has not yet been completed, the patrol will begin or continue; if the patrol project has been completed, the relevant information of the patrol project can be viewed, but the total working time will be recorded.

[0171] When starting or continuing a patrol, the system typically selects the first unexecuted subtask in the patrol project by default and changes its status to "in progress". If the task has not been entered before, the system records the subtask's start time; if it has been entered before, the actual start time of the task is not changed.

[0172] After selecting a subtask, if the subtask is incomplete, the subtask countdown will begin. During the patrol, the patrol personnel will check against the inspection standards. If they click "Return to Previous", select another menu, or force close the browser, they will exit the patrol.

[0173] After the subtask countdown ends, the subtask status is changed to "Completed," and a selection countdown begins (e.g., 5 minutes). After the countdown ends, you can exit the screen patrol or select the next subtask. After selection, the screen patrol for the current subtask ends. At this point, the screen patrol time for the subtask can be accumulated. If the subtask is not completed, the system can change its status to "Not Started." If the subtask is completed, the system can record the actual end time of the subtask.

[0174] See Figure 8 , Figure 8 This specification illustrates a schematic diagram of the state changes of a subtask inspection according to an embodiment of the present specification.

[0175] In the illustrated front-end and back-end interaction flow, the front-end triggers the start of a subtask by clicking the start or continue button of a node task, clicking the next task button of a subtask, or clicking to select a task, and then requests subtask information from the back-end. The front-end determines the subtask status based on the subtask information returned by the back-end.

[0176] For incomplete subtasks, screen patrol is performed within the countdown; for completed subtasks, the task can be viewed by performing a viewing action. In the case of incomplete subtasks with empty start times, the front end asynchronously sends start time information to the back end and modifies the start time of the corresponding subtask in the database table.

[0177] During the inspection or review process, the front end asynchronously monitors the countdown timer (i.e., whether the suggested time has been reached). If the countdown ends, the front end sends the end time to the back end, causing the back end to change the sub-task status and modify the actual time. To effectively monitor the quality and efficiency of the inspection work, the system records and analyzes the task status, start time, end time, and actual time of each sub-task during the inspection process for subsequent query and display.

[0178] Regarding issue reporting and follow-up, the attributes for issue reporting include, but are not limited to, issue name and category, regional location information, fault information, issue tracking and time limit settings, and images. Issue tracking and time limit settings refer to the tracking and reminders for the rectification of the issue. After setting the reminder time limit according to the specifications, the inspection personnel will continue the task inspection. When the time limit is reached, the system will pop up an issue screen. After verifying the rectification status, the inspection personnel can continue tracking or close the loop.

[0179] See Figure 9 , Figure 9 A schematic diagram of a problem reporting process provided in one embodiment of this specification is shown.

[0180] After the inspection issues are reported, the work order is transmitted to the corresponding alarm system (in this implementation, the passenger operation system is used as the alarm system) for processing. The specific interaction process is as follows: the security system WEB service receives the reported information filled in by the inspection personnel and uploads attachments (such as monitoring screenshots) to the passenger operation system. The passenger operation system can send feedback to the security system WEB service that the upload was successful. The security system WEB service saves the reported information in the database and synchronizes the reported information to the security system SYNCDATA service. The SYNCDATA service then reports it to the passenger operation system. After the report is successful, an event identifier is returned to the security system WEB service and written back to the database.

[0181] The security system's SYNCDATA service calls the query interface cyclically at fixed time intervals (e.g., every 2 minutes) to query the event status from the passenger operation system, obtain the maintenance results, descriptions, status codes, and times, and then write them back to the database. The security system's WEB service page can also query the processing status.

[0182] In terms of inspection quality control, the system automatically collects inspection data and generates relevant reports, including task execution logs, problem management logs, and employee quality statistics. The task execution log includes task completion progress, number of problems found, executor, and actual execution time. The problem management log includes problem descriptions, reporting personnel, and closed-loop status. Employee quality statistics include statistics on effective inspection time within the selected time period, problem reporting statistics, and the number of incomplete tasks. Through this comprehensive configuration, execution, reporting, tracking, and quality control mechanism, the system can fully utilize human resources, standardize inspection criteria, reduce inspection time, improve inspection efficiency and quality, significantly alleviate inspector fatigue, enhance control over on-site operations, and ensure the safe and efficient operation of large public places, thereby achieving optimal on-site situation control with minimal manpower and time costs.

[0183] The task-driven video inspection method provided in this specification, through in-depth analysis of the on-site operational elements of public places, extracts key on-site operational information (such as inspection points, inspection content, and inspection content), and combines the coverage of surveillance cameras. Through the fine planning and flexible allocation of cameras, it fully realizes close integration and efficient linkage between humans and machines, improves the strength and efficiency of video surveillance inspection, and reduces manpower and time costs.

[0184] Corresponding to the above method embodiments, this specification also provides embodiments of a task-driven video inspection device. Figure 10 A schematic diagram of a task-driven video inspection device according to one embodiment of this specification is shown. Figure 10 As shown, the device includes: The task acquisition module 1002 is configured to respond to the login operation of the target account, acquire the target inspection task corresponding to the current time under the target account, and determine multiple inspection items corresponding to the target inspection task and the suggested time for each inspection item. The inspection time period corresponding to the target inspection task corresponds to the current time, and the inspection time period is determined based on the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the inspection time period is automatically calculated and generated according to external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. The project inspection module 1004 is configured to select a target inspection project from the plurality of inspection projects, analyze the monitoring video corresponding to the target inspection project based on the inspection content corresponding to the target inspection project and in combination with video analysis algorithms, and obtain analysis results. The result generation module 1006 is configured to compare the analysis results with the qualification standards in the inspection content, and generate the inspection results of the target inspection project based on the comparison results. The problem reporting module 1008 is configured to continue executing the step of selecting the target inspection project from the multiple inspection projects when the corresponding suggested time is reached, and to send the inspection problem to the alarm system if the inspection result contains an inspection problem.

[0185] The device further includes: The binding module is configured to acquire monitoring equipment information of the target area and bind the monitoring equipment information with the patrol projects of the target area, and determine the monitoring equipment information corresponding to each patrol project. The monitoring equipment information corresponding to each patrol project is configured through forward binding or reverse binding. Forward binding includes binding monitoring equipment information to patrol projects, and reverse binding includes binding monitoring equipment information to corresponding patrol projects. The attribute information of each patrol project includes project name, patrol type, suggested time, and inspection content.

[0186] Optionally, the task acquisition module 1002 is further configured to: In response to a login operation of the target account, obtain multiple candidate patrol tasks corresponding to the current time under the target account; Based on the current time and the inspection period of each candidate inspection task, at least one inspection task corresponding to the current time is obtained from the plurality of candidate inspection tasks, and a target inspection task is determined from the at least one inspection task to be executed, wherein the inspection period of each candidate inspection task is determined when the corresponding candidate inspection task is created.

[0187] The device further includes: A creation module is configured to respond to a request to add a patrol task, generate a task identifier corresponding to the patrol task, and obtain a patrol task template; generate an initial patrol task based on the patrol task parameters and the patrol task template, wherein the patrol task parameters include a target account, at least one patrol item, and a task trigger mode; determine the patrol time period corresponding to the initial patrol task based on the task trigger mode, wherein the task trigger mode includes a fixed trigger mode and a dynamic trigger mode. In the dynamic trigger mode, the patrol time period is automatically calculated and generated based on external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans; determine whether there is a time period conflict between the patrol time period of the initial patrol task and the patrol time period of an existing candidate patrol task under the target account. If there is a time period conflict, update the patrol time period corresponding to the initial patrol task so that the updated initial patrol task does not have a time period conflict with the existing candidate patrol tasks under the target account, and determine the updated initial patrol task as a candidate patrol task under the target account. If there is no time period conflict, the initial patrol task is determined as a candidate patrol task under the target account.

[0188] Optionally, the project inspection module 1004 is further configured to: When the inspection time reaches the suggested time in the inspection interface corresponding to the target inspection project, an end reminder message will be displayed; In response to the inspection end command, the inspection of the target inspection item ends, and the step of selecting the target inspection item from the multiple inspection items continues to be executed, wherein the inspection interface displays the monitoring video corresponding to the monitoring equipment information; And / or, Perform useless time detection; if the inactivity time on the inspection interface corresponding to the target inspection item exceeds a preset threshold, automatically exit the target inspection task. Upon re-entering the target inspection task, the inspection resumes from the target inspection item where the task was paused upon exiting, and if the corresponding recommended time is reached, the inspection of the target inspection item ends, and the step of selecting a target inspection item from the multiple inspection items continues.

[0189] Optionally, the project inspection module 1004 is further configured to: The inspection content corresponding to the target inspection project is determined according to the preset configuration information, wherein the inspection content includes inspection items for inspecting the on-site conditions and corresponding qualification standards.

[0190] The device further includes: The review module is configured to set a review time for the inspection problem, and if no repair response is received from the alarm system and the review time has arrived, trigger a tracking reminder for the inspection problem; if a repair response is received from the alarm system, mark the problem status of the inspection problem as resolved.

[0191] The device further includes: The statistics module is configured to generate an inspection quality control table based on the inspection data upon completion of the target inspection task. The inspection quality control table includes at least one statistical component from the following: task execution log, problem management log, and employee quality statistics. The task execution log includes at least one data point from the following: task completion progress, number of problems found, executor, and actual execution time. The problem management log includes at least one data point from the following: inspection problem description and inspection problem status. The employee quality statistics include at least one data point from the following: effective inspection time statistics, problem reporting statistics, and number of incomplete tasks statistics.

[0192] Optionally, the project inspection module 1004 is further configured to: At least one regional subtask corresponding to the target inspection project is determined. Based on the inspection content corresponding to each regional subtask and combined with video analysis algorithms, the monitoring videos corresponding to each regional subtask are analyzed to obtain the analysis results corresponding to each regional subtask. Each regional subtask corresponds to a combination of different monitoring devices within the target area where the target inspection project is located. The target inspection project is obtained by connecting and combining each regional subtask in series. The monitoring devices of each regional subtask are configured according to the splitting principle.

[0193] The above is an illustrative scheme of a task-driven video inspection device according to this embodiment. It should be noted that the technical solution of this task-driven video inspection device and the technical solution of the task-driven video inspection method described above belong to the same concept. Details not described in detail in the technical solution of the task-driven video inspection device can be found in the description of the technical solution of the task-driven video inspection method described above.

[0194] Figure 11 A structural block diagram of a computing device 1100 according to one embodiment of this specification is shown. The components of the computing device 1100 include, but are not limited to, a memory 1110 and a processor 1120. The processor 1120 is connected to the memory 1110 via a bus 1130, and a database 1150 is used to store data.

[0195] The computing device 1100 also includes an access device 1140, which enables the computing device 1100 to communicate via one or more networks 1160. Examples of these networks include Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or combinations of communication networks such as the Internet. The access device 1140 may include one or more of any type of wired or wireless network interface (e.g., a network interface card (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Wi-MAX (Worldwide Interoperability for Microwave Access) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, or a Near Field Communication (NFC) interface.

[0196] In one embodiment of this specification, the aforementioned components of the computing device 1100 and Figure 11 Other components, not shown, can also be connected to each other, for example, via a bus. It should be understood that... Figure 11 The block diagram of the computing device shown is for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0197] The computing device 1100 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or personal computers (PCs). The computing device 1100 can also be a mobile or stationary server.

[0198] The processor 1120 is used to execute the following computer program / instruction, which, when executed by the processor, implements the steps of the above-described task-driven video inspection method.

[0199] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, the computing device embodiments are relatively simple in description because they are fundamentally similar to the task-driven video inspection method embodiments; relevant parts can be referred to in the descriptions of the task-driven video inspection method embodiments.

[0200] An embodiment of this specification also provides a computer-readable storage medium storing a computer program / instructions that, when executed by a processor, implement the steps of the task-driven video inspection method described above.

[0201] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, the computer-readable storage medium embodiment is described simply because it is substantially similar to the task-driven video inspection method embodiment; relevant parts can be found in the description of the task-driven video inspection method embodiment.

[0202] An embodiment of this specification also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the task-driven video inspection method described above.

[0203] The above is an illustrative scheme of a computer program product according to this embodiment. It should be noted that the technical solution of this computer program product and the technical solution of the task-driven video inspection method described above belong to the same concept. For details not described in detail in the technical solution of the computer program product, please refer to the description of the technical solution of the task-driven video inspection method described above.

[0204] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0205] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added or removed according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0206] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments in this specification are not limited to the described order of actions, because according to the embodiments in this specification, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments in this specification.

[0207] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0208] The preferred embodiments disclosed above are merely illustrative of this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments described herein. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the embodiments, thereby enabling those skilled in the art to better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. A task-driven video inspection method, characterized in that, Applications include airport inspection scenarios, including: In response to the login operation of the target account, the system obtains the target inspection task corresponding to the current time under the target account, and determines multiple inspection items corresponding to the target inspection task and the suggested time for each inspection item. The inspection time period corresponding to the target inspection task corresponds to the current time, and the inspection time period is determined based on the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the inspection time period is automatically calculated and generated according to external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. Select a target inspection project from the multiple inspection projects, and analyze the monitoring video corresponding to the target inspection project based on the inspection content and video analysis algorithm to obtain the analysis results; The analysis results are compared with the pass / fail standards in the inspection content, and the inspection results of the target inspection project are generated based on the comparison results. If the recommended time is reached, continue to execute the step of selecting the target inspection project from the multiple inspection projects, and if the inspection results contain inspection problems, send the inspection problems to the alarm system.

2. The task-driven video inspection method according to claim 1, characterized in that, Before responding to the login operation of the target account and obtaining the target patrol task corresponding to the current time under the target account, the following steps are included: The system acquires monitoring equipment information for the target area and binds this information to patrol projects within the target area. It then determines the monitoring equipment information corresponding to each patrol project. This monitoring equipment information is configured through forward binding or reverse binding. Forward binding involves binding monitoring equipment information to patrol projects, while reverse binding involves binding monitoring equipment information to corresponding patrol projects. The attribute information for each patrol project includes the project name, patrol type, suggested time, and inspection content.

3. The task-driven video inspection method according to claim 1, characterized in that, The step of responding to a login operation of the target account and obtaining the target patrol task corresponding to the current time under the target account includes: In response to a login operation of the target account, obtain multiple candidate patrol tasks corresponding to the current time under the target account; Based on the current time and the inspection period of each candidate inspection task, at least one inspection task corresponding to the current time is obtained from the plurality of candidate inspection tasks, and a target inspection task is determined from the at least one inspection task to be executed, wherein the inspection period of each candidate inspection task is determined when the corresponding candidate inspection task is created.

4. The task-driven video inspection method according to claim 3, characterized in that, Before responding to the login operation of the target account and obtaining the target patrol task corresponding to the current time under the target account, the method further includes: In response to a request to add an inspection task, a task identifier corresponding to the inspection task is generated, and an inspection task template is obtained; An initial inspection task is generated based on the inspection task parameters and the inspection task template, wherein the inspection task parameters include the target account, at least one inspection item, and the task triggering mode. According to the task triggering mode, the patrol period corresponding to the initial patrol task is determined. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the patrol period is automatically calculated and generated based on external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. Determine whether there is a time conflict between the patrol time period of the initial patrol task and the patrol time period of the existing candidate patrol tasks under the target account. If a time slot conflict exists, update the patrol time slot corresponding to the initial patrol task to ensure that the updated initial patrol task does not conflict with any existing candidate patrol tasks under the target account. Then, designate the updated initial patrol task as a candidate patrol task under the target account. If there is no time conflict, the initial patrol task will be identified as a candidate patrol task under the target account.

5. The task-driven video inspection method according to claim 1, characterized in that, The step of selecting a target inspection project from the plurality of inspection projects when the corresponding suggested time has been reached includes: When the inspection time reaches the suggested time in the inspection interface corresponding to the target inspection project, an end reminder message will be displayed; In response to the inspection end command, the inspection of the target inspection item ends, and the step of selecting the target inspection item from the multiple inspection items continues to be executed, wherein the inspection interface displays the monitoring video corresponding to the monitoring equipment information; And / or, Perform useless time detection; if the inactivity time on the inspection interface corresponding to the target inspection item exceeds a preset threshold, automatically exit the target inspection task. Upon re-entering the target inspection task, the inspection resumes from the target inspection item where the task was paused upon exiting, and if the corresponding recommended time is reached, the inspection of the target inspection item ends, and the step of selecting a target inspection item from the multiple inspection items continues.

6. The task-driven video inspection method according to claim 1 or 5, characterized in that, Before analyzing the monitoring video corresponding to the target inspection project based on the inspection content and video analysis algorithm to obtain the analysis results, the process also includes: The inspection content corresponding to the target inspection item is determined according to the preset configuration information. The inspection content includes inspection items for the on-site condition and corresponding qualification standards. The qualification standards include a number of people threshold, a queue length threshold, and / or an area occupancy threshold.

7. The task-driven video inspection method according to claim 1, characterized in that, After sending the inspection problem to the alarm system, the process also includes: Set a review time for the inspection problem, and if no repair response is received from the alarm system and the review time has arrived, trigger a tracking reminder for the inspection problem; Upon receiving a repair response from the alarm system, the problem status of the inspection issue is marked as resolved.

8. The task-driven video inspection method according to claim 1, characterized in that, After sending the inspection problem to the alarm system, the process also includes: Upon completion of the target inspection task, an inspection quality control table is generated based on the inspection data. The inspection quality control table includes at least one statistical component from the following: task execution log, problem management log, and employee quality statistics. The task execution log includes at least one data point from the following: task completion progress, number of problems found, executor, and actual execution time. The problem management log includes at least one data point from the following: inspection problem description and inspection problem status. The employee quality statistics include at least one data point from the following: effective inspection time statistics, problem reporting statistics, and number of incomplete tasks statistics.

9. The task-driven video inspection method according to any one of claims 1-5, characterized in that, The inspection type of the target inspection project is any one of process node supervision, regional supervision, periodic supervision, and spot check; The process node supervision refers to the inspection of process nodes divided according to the traffic flow. The regional inspection refers to the type of inspection conducted in the designated areas based on the regional division of the supervised premises; the regular inspection refers to the type of inspection with a fixed number of inspections required; and the spot check refers to the type of inspection without a fixed frequency.

10. The task-driven video inspection method according to claim 9, characterized in that, When the inspection type of the target inspection project is regional supervision, the step of analyzing the monitoring video corresponding to the target inspection project based on the inspection content and video analysis algorithm to obtain the analysis results includes: At least one regional subtask corresponding to the target inspection project is determined. Based on the inspection content corresponding to each regional subtask and combined with video analysis algorithms, the monitoring videos corresponding to each regional subtask are analyzed to obtain the analysis results corresponding to each regional subtask. Each regional subtask corresponds to a combination of different monitoring devices within the target area where the target inspection project is located. The target inspection project is obtained by connecting and combining each regional subtask in series. The monitoring devices of each regional subtask are configured according to the splitting principle.

11. A task-driven video inspection device, characterized in that, include: The task acquisition module is configured to respond to the login operation of the target account, acquire the target inspection task corresponding to the current time under the target account, and determine the multiple inspection items corresponding to the target inspection task and the suggested time for each inspection item. The inspection time period corresponding to the target inspection task corresponds to the current time, and the inspection time period is determined based on the task triggering mode. The task triggering mode includes a fixed triggering mode and a dynamic triggering mode. In the dynamic triggering mode, the inspection time period is automatically calculated and generated according to external operation plan data and preset time offset rules. The external operation plan data includes dynamically accessed flight operation plans. The project inspection module is configured to select a target inspection project from the multiple inspection projects, analyze the monitoring video corresponding to the target inspection project based on the inspection content corresponding to the target inspection project and in combination with video analysis algorithms, and obtain analysis results. The result generation module is configured to compare the analysis results with the pass / fail standards in the inspection content, and generate the inspection results of the target inspection project based on the comparison results. The problem reporting module is configured to continue executing the step of selecting the target inspection project from the multiple inspection projects when the corresponding suggested time is reached, and to send the inspection problem to the alarm system if the inspection result contains an inspection problem.

12. A computing device, characterized in that, include: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, they implement the steps of the task-driven video inspection method according to any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, It stores a computer program / instruction that, when executed by a processor, implements the steps of the task-driven video inspection method according to any one of claims 1-10.

14. A computer program product, characterized in that, Includes a computer program / instruction that, when executed by a processor, implements the steps of the task-driven video inspection method according to any one of claims 1-10.

Images