Efficiency analysis method and system based on business data and video monitoring

Abstract

According to the business data and video monitoring-based efficiency analysis method and system, early warning rules are set for core dimensions such as personnel on-duty states, reservation resource configuration, mass affair handling experience and the like, and are combined with workday configuration, so that the problems of off-duty, unreasonable resource configuration, multi-pass running and the like are automatically judged, the dependence on manual patrol is reduced, and the working efficiency is improved. The efficiency and the accuracy of service supervision are improved; after early warning, a feedback link is checked in a linkage mode, timely problem rectification is promoted, and continuous optimization of service efficiency is promoted.

Description

Technical Field

[0001] This application relates to the field of performance analysis, specifically to a performance analysis method and system based on business data and video surveillance. Background Technology

[0002] With the deepening of government service reform, the need for effective management of service venues within the jurisdiction is becoming increasingly urgent. Traditional supervision methods rely heavily on on-site methods such as overt and covert inspections, which have problems such as limited coverage, insufficient timeliness, and passive response. They are no longer able to meet the requirements of refined and intelligent management. Government supervision urgently needs to be upgraded to off-site supervision to realize the transformation from passively discovering problems to proactively identifying risks.

[0003] Video, as the most direct means of reflecting on-site situations, is increasingly widely used in the supervision and management of service venues. In existing technology, patent number CN106846221A, entitled "An Electronic Supervision System," discloses a video monitoring module, an electronic supervision module, a user feedback module, and a monitoring platform. The video monitoring module captures office video through camera units in the office area and sends it to the monitoring platform; the electronic supervision module collects and uploads approval information through a data acquisition unit; the user feedback module transmits user feedback information to the monitoring platform; and the monitoring platform stores all types of data through a supervision database. This technical solution relies on manual inspection and monitoring of video data for supervision, resulting in a single means of oversight. When faced with a large number of video monitoring points, the traditional "manual inspection" model is inefficient in covering all areas and is prone to oversights. Furthermore, it lacks corresponding early warning mechanisms for common service venue issues affecting service quality, such as the practice of manipulating service volume through high-frequency queuing, the phenomenon of people making multiple trips for the same matter, and insufficient online appointment numbers. This results in a lack of intelligent early warning capabilities for these key supervisory scenarios, failing to meet the needs of upgrading non-on-site supervision of government services. Summary of the Invention

[0004] In view of the above problems, this application provides a performance analysis method and system based on business data and video surveillance to solve the problem that existing performance supervision cannot meet the needs of off-site supervision of government services.

[0005] To achieve the above objectives, the inventors provide a performance analysis method based on business data and video surveillance, which includes the following steps:

[0006] Efficiency data is obtained from the early warning database, including appointment number data, video surveillance data, and case processing data.

[0007] Based on the workday configuration and preset early warning rules, determine whether the performance data meets the early warning conditions; if the early warning conditions are met, issue an early warning and verify the feedback.

[0008] The preset rules specifically include:

[0009] The algorithm for detecting absenteeism is used to determine whether absenteeism exists in video surveillance data. If it does, an early warning condition is met; or

[0010] The reservation number data is used to determine whether resource allocation is reasonable; if it is unreasonable, the warning conditions are met.

[0011] The system uses case processing data to determine if there are any "multiple trips" incidents; if so, it meets the warning criteria.

[0012] Furthermore, the step of using the personnel absence / absence algorithm to determine whether there is absence / absence in the video surveillance data includes the following steps:

[0013] Define at least one warning zone and / or a post work area;

[0014] Based on the YOLOv7 target detection model and DeepSORT multi-target tracking algorithm, the system locates and tracks people in video surveillance data to obtain the pixel coordinates of people in the video surveillance data.

[0015] The pixel coordinates of people in video surveillance data are mapped to actual planar coordinates based on the perspective transformation matrix;

[0016] Determine whether personnel have entered the restricted area and / or left their work area based on actual planar coordinates;

[0017] If so, determine whether the time spent entering the restricted area and / or leaving the work area exceeds the preset time;

[0018] If so, it is determined that there has been a case of being absent from one's post.

[0019] Furthermore, the step of using reservation number data to determine whether resource allocation is reasonable includes judging one or more of the following conditions:

[0020] Determine if there are any restrictions on vehicle registration; if so, determine if the resource allocation is unreasonable.

[0021] Determine if there is an insufficient number of available numbers; if so, determine if resource allocation is unreasonable.

[0022] Determine if there is congestion in the processing of applications; if so, determine if resource allocation is unreasonable.

[0023] Furthermore, the step of determining whether a restriction number exists includes the following steps:

[0024] If the number of numbers issued is the same in N consecutive working days, then there is a limit to the number of numbers issued; or

[0025] If no number is recorded after the preset time before the end of the workday in N consecutive workdays, then there is a limit on the number of tickets that can be taken; or

[0026] If, within N consecutive workdays, a single person has multiple records of taking a number after a preset time before the end of the workday, then a limit on the number of numbers is applied; or

[0027] If, within N consecutive workdays, there are records of taking numbers after a preset time before the end of the workday, and no numbers are called or all numbers have passed, then there is a limit on the number of numbers taken.

[0028] If, in N consecutive workdays, a certain window or a certain matter exists for M workdays, and in the call log after a preset time before the end of the workday, there are A consecutive call intervals of less than B seconds, then there is a limit on the number of calls.

[0029] Furthermore, the step of determining whether there is an insufficient number of license plates to be issued includes the following steps:

[0030] If, within N consecutive working days, there are M working days where the appointment slots are full, then there is a shortage of available slots.

[0031] If, within N consecutive working days, the reservation slots for M working days are fully booked within C minutes, then there is an insufficient number of slots to be released.

[0032] Furthermore, the step of determining whether the application is congested includes the following steps:

[0033] The queuing time is obtained based on the time of obtaining the number and the time of being called.

[0034] If the queuing time exceeds the preset waiting time, it will be recorded as crowded.

[0035] Determine if there are M consecutive congested numbers; if so, the application is congested.

[0036] If the number of queuing numbers exceeds the preset number, and the window opening rate is less than 50%, and the queuing time exceeds the preset waiting time, then the application process will be crowded.

[0037] Furthermore, the step of using case data to determine whether a "multiple trips" incident has occurred includes the following steps:

[0038] If, in N consecutive working days, there are M working days in which the same person handles the same type of matter more times than the preset number;

[0039] If the item is not on the "multiple trips" warning whitelist, then the "multiple trips" incident has occurred.

[0040] Furthermore, the video surveillance data is obtained from the video supervision platform and is included in the basic database for early warning; the appointment number data is obtained from the queuing and calling system and is included in the basic database for early warning; the case processing data is obtained from the service item database and is included in the basic database for early warning.

[0041] Furthermore, the early warning includes generating an alarm record that meets the corresponding early warning conditions, and sending an alarm notification through at least one of the following methods: an audible and visual alarm device, SMS, email, or system message.

[0042] A performance analysis system based on business data and video surveillance is provided to implement the aforementioned performance analysis method based on business data and video surveillance.

[0043] Unlike existing technologies, the above-mentioned technical solution sets early warning rules based on core dimensions such as staff on-duty status, appointment resource allocation, and public service experience. Combined with weekday configuration, it automatically identifies problems such as absenteeism, unreasonable resource allocation, and multiple trips, reducing reliance on manual inspections and improving the efficiency and accuracy of service supervision. After the early warning, it links with the verification and feedback process to promote timely rectification of problems and continuously optimize service efficiency.

[0044] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0045] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of the present invention and other related contents, and should not be considered as limitations on this application.

[0046] In the accompanying drawings of the instruction manual:

[0047] Figures 1-3 A flowchart illustrating the performance analysis method based on business data and video surveillance as described in the specific implementation method;

[0048] Figure 4 This is a schematic diagram of the modules of the performance analysis system based on business data and video surveillance, as described in a specific implementation. Detailed Implementation

[0049] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0050] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0051] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0052] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.

[0053] In this application, terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual quantity, hierarchy or order relationship between these entities or operations.

[0054] Without further limitations, the use of terms such as “comprising,” “including,” “having,” or other similar open-ended expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.

[0055] Similar to the understanding in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.

[0056] In the description of the embodiments of this application, the space-related expressions used, such as "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "vertical," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential," indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiments or drawings. They are only for the purpose of describing the specific embodiments of this application or for the reader's understanding, and do not indicate or imply that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application.

[0057] The processor described in the embodiments of this application can be implemented by hardware, firmware, software, or a combination thereof. It can be a circuit, one or more of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, or a microprocessor. It also includes other physical, biological, or chemical structures that can implement the same or equivalent functions as the processors listed above, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute some or all of the steps in the computer program or method involved in the various embodiments of this application, or any combination of the steps mentioned therein.

[0058] The computer program involved in the embodiments can be stored in a computer device readable storage medium, which includes, but is not limited to, disks, magnetic tapes, magnetic cards, floppy disks, flash memory, optical disks, optical cards, read-only memory (ROM), random access memory (RAM), erasable programmable ROM (EPROM), and electrically erasable programmable ROM (EEPROM), etc., and also includes other biological, physical, or chemical structures that can achieve the same or equivalent functions as the storage media listed above, such as DNA, RNA, proteins, and other units with information storage capabilities. In specific embodiments, the storage medium involved can be one of the above-mentioned media types, or a combination of the above-mentioned media types. In different embodiments, the computer program involved in the embodiments can be centrally stored in a single medium, or distributed and stored in multiple media. The memory containing the computer device readable storage medium can be non-volatile memory or random access memory. These computer device readable storage media can be built into the device, or can be connected to the device involved in the embodiments as an external device or part of an external device. In some embodiments, the memory having a computer device readable storage medium is deployed locally; in other embodiments, the memory may be deployed remotely from the processor, for example, as a network-attached memory accessed via RF circuitry or an external port and a communication network, wherein the communication network may be the Internet, one or more intranets, a local area network (LAN), a wide area network (WLAN), a storage area network (SAN), or a suitable combination thereof, as long as computer device access to the memory is enabled. Furthermore, the computer program involved in the embodiments may be stored in plaintext / ciphertext form, or it may be designed as training data, integrated and recombined through model training and implicitly stored in the parameter states of a deep neural network or other machine learning model.

[0059] An efficiency analysis method based on business data and video surveillance sets early warning rules for core dimensions such as staff on-duty status, appointment resource allocation, and public service experience. Combined with weekday configuration, it automatically identifies problems such as absenteeism, unreasonable resource allocation, and multiple trips, reducing reliance on manual inspections and improving the efficiency and accuracy of service supervision. After the early warning, it links with the verification and feedback process to promote timely rectification of problems and continuously optimize service efficiency.

[0060] The following combination Figures 1-3 This paper provides an implementation method for performance analysis based on business data and video surveillance, which includes the following steps:

[0061] Efficiency data is obtained from the early warning database, including appointment number data, video surveillance data, and case processing data.

[0062] Based on the workday configuration and preset early warning rules, determine whether the performance data meets the early warning conditions; if the early warning conditions are met, issue an early warning and verify the feedback.

[0063] The preset rules specifically include:

[0064] The algorithm for detecting absenteeism is used to determine whether absenteeism exists in video surveillance data. If it does, an early warning condition is met; or

[0065] The reservation number data is used to determine whether resource allocation is reasonable; if it is unreasonable, the warning conditions are met.

[0066] The system uses case processing data to determine if there are any "multiple trips" incidents; if so, it meets the warning criteria.

[0067] The aforementioned performance data is obtained from the early warning database. Performance data includes, but is not limited to, appointment number data, video surveillance data, and case handling data. Different data come from the corresponding applications.

[0068] The aforementioned video surveillance data can be obtained from the video supervision platform and stored in the basic database for early warning. The video supervision platform collects video surveillance data collected through on-site monitoring equipment and is mainly used to monitor the on-duty status of staff, service scenarios, and other real-time conditions.

[0069] The aforementioned appointment number data can be obtained from the queuing and calling system and is included in the basic data for early warning. This appointment number data can include, but is not limited to, data on number release, appointment, number retrieval, and number calling. To ensure the accuracy of subsequent early warning judgments, the appointment number data can be extracted from the queuing and calling system for targeted cleaning before entering the early warning basic database. Specific operations include: calculating the difference between the calling time and the number retrieval time; if this difference is less than 0 or greater than 4 hours, the corresponding record is marked as dirty data, which may be due to system failure or operational errors causing time record anomalies; when a number retrieval timestamp is detected to be later than the calling timestamp (which does not conform to normal business logic, which should normally be number retrieval before calling), the system will automatically correct the timestamp (e.g., adjusting it to a reasonable order according to business process logic); if it cannot be automatically corrected, it is marked as abnormal data. For all records marked as dirty data and abnormal data, the system will classify and store them (e.g., by anomaly type, occurrence time, etc.) and support manual review. Staff can verify and correct the data by viewing detailed records to ensure that the appointment number data entering the early warning basic database is accurate and standardized, providing a reliable basis for judging the rationality of subsequent resource allocation. Specifically, an HTTP interface can be developed to obtain queuing machine data from the queuing and calling system, and the transmitted data can be fully encrypted using the national cryptographic SM2 elliptic curve algorithm; data ETL services can be used to clean the business source data and store it in the database as basic data for early warning.

[0070] The aforementioned case processing data is obtained from the service item database. Specifically, data ETL services can be used to extract and transform service data and positive / negative feedback data from the service item database, and the results are then stored in the early warning database. Case processing data records the entire process information of business processing, including but not limited to processing time, required materials, approval steps, and number of times the user has processed the application.

[0071] The above-mentioned steps for determining whether personnel are absent from their posts in video surveillance data using the personnel absence from their posts algorithm include the following steps:

[0072] Define at least one warning zone and / or a post work area;

[0073] Based on the YOLOv7 target detection model and DeepSORT multi-target tracking algorithm, the system locates and tracks people in video surveillance data to obtain the pixel coordinates of people in the video surveillance data.

[0074] The pixel coordinates of people in video surveillance data are mapped to actual planar coordinates based on the perspective transformation matrix;

[0075] Determine whether personnel have entered the restricted area and / or left their work area based on actual planar coordinates;

[0076] If so, determine whether the time spent entering the restricted area and / or leaving the work area exceeds the preset time;

[0077] If so, it is determined that there has been a case of being absent from one's post.

[0078] The aforementioned work area refers to the fixed area where staff are supposed to work (such as the physical area corresponding to a window counter or service desk), usually marked on a plane coordinate system with polygonal boundaries (such as rectangles or irregular quadrilaterals). The aforementioned restricted area is the area where staff are prohibited or restricted from entering during working hours; non-work areas can also be marked with polygonal boundaries.

[0079] The aforementioned method of locating and tracking people in video surveillance data using the YOLOv7 target detection model and DeepSORT multi-target tracking algorithm to obtain the pixel coordinates of people in the video surveillance data refers to using YOLOv7 to identify "person" targets in each frame of the video surveillance data, and DeepSORT assigning a unique ID to each person and tracking their movement trajectory through correlation analysis of consecutive frames. This ensures that even in the event of brief occlusion or personnel movement, the pixel coordinates of the people in the video surveillance data are output. Specifically, the input resolution of YOLOv7 can be set to 640x640: this size strikes a balance between "detection accuracy" and "computational efficiency." The 640x640 resolution can retain sufficient details (such as the outline features of people at a distance) to meet the detection needs of indoor monitoring scenarios (such as service windows and office areas), while avoiding a surge in computation due to excessively high resolution, thus ensuring real-time processing capabilities (suitable for continuous analysis of monitoring images). Weighted Non-Maximum Suppression (W-NMS) can be used. In object detection, the same person may be output by the model with multiple overlapping bounding boxes. Non-Maximum Suppression (NMS) retains the best box and removes redundant boxes. W-NMS, on the other hand, assigns higher weights to high-confidence boxes, reducing false deletions caused by occlusion or object blurring (for example, when multiple people are standing side by side, ordinary NMS may falsely delete the boxes of people at the edge, while W-NMS can more accurately retain valid detection results), thereby improving the accuracy of person detection. DeepSORT is a tracking model optimized based on the SORT algorithm. Its core is to solve the problem of "stable association of the same person in consecutive frames". DeepSORT's appearance feature extractor can use the ResNet50 network. As a deep convolutional neural network, ResNet50 can extract robust appearance feature vectors (converting the appearance of people into quantifiable values), improving the ability to distinguish different people and reducing the problem of "ID switching" (i.e., the same person being misidentified as multiple people). Mahalanobis distance is used to measure the similarity (including motion features such as position and speed) between the current detection bounding box and historical tracking bounding boxes. When the distance is less than or equal to 0.8, the model determines that "the currently detected person and the historically tracked person are the same person." This threshold setting needs to balance "association accuracy" and "tracking continuity": if the threshold is too low (e.g., 0.5), rapid movement of people may cause association failure (misjudging as a new target); if it is too high (e.g., 1.2), different people may be mistakenly associated with the same person. Setting it to 0.8 ensures tracking stability and accuracy in most monitoring scenarios (e.g., fixed indoor areas, moderate personnel movement speed).

[0080] After obtaining the pixel coordinates of personnel through video surveillance, it is necessary to further combine them with the planar coordinates of the actual scene to determine whether the personnel have entered the restricted area or left their post work area. In some embodiments, the ray casting method is used to determine whether personnel have entered the restricted area and / or left their post work area. That is, a ray is cast from the personnel coordinates in any direction (usually horizontally to the right), and the number of intersections between the ray and the boundary of the restricted area or the boundary of the post work area is counted. The parity of the number of intersections determines whether the point is within the restricted area or the post work area. If the "post work area" is a polygon P, the ray casting method is used to detect whether the personnel coordinates are outside P. If so, it is determined that the personnel have "left their post". If the "restricted area" is a polygon Q, the ray casting method is used to detect whether the personnel coordinates are inside Q. If so, it is determined that the personnel have "entered the restricted area".

[0081] From the moment a person is first identified as having "entered the restricted area" or "left their post," the system automatically times and calculates the duration of their presence in that abnormal position. This duration is then compared to a "preset time" (defined according to business rules, such as 10 minutes or 15 minutes). If the duration does not exceed the preset time, it is considered a "brief absence," and no "absent from post" warning is triggered (to avoid false alarms). If the duration exceeds the preset time, it is considered an unreasonable absence, classified as "absent from post," and a warning is triggered. This avoids false alarms caused by momentary positional deviations (such as standing up and stretching, or brief movements) or reasonable brief absences, ensuring that warnings are only issued for genuine violations of work regulations, such as "prolonged absence from post" or "prolonged stay in the restricted area."

[0082] The steps described above for determining the rationality of resource allocation using reservation number data include judging one or more of the following conditions:

[0083] Determine if there are any restrictions on vehicle registration; if so, determine if the resource allocation is unreasonable.

[0084] Determine if there is an insufficient number of available numbers; if so, determine if resource allocation is unreasonable.

[0085] Determine if there is congestion in the processing of applications; if so, determine if resource allocation is unreasonable.

[0086] The above steps for determining whether there are vehicle restrictions include the following:

[0087] If the number of numbers issued is the same in N consecutive working days, then there is a limit to the number of numbers issued; or

[0088] If no number is recorded after the preset time before the end of the workday in N consecutive workdays, then there is a limit on the number of tickets that can be taken; or

[0089] If, within N consecutive workdays, a single person has multiple records of taking a number after a preset time before the end of the workday, then a limit on the number of numbers is applied; or

[0090] If, within N consecutive workdays, there are records of taking numbers after a preset time before the end of the workday, and no numbers are called or all numbers have passed, then there is a limit on the number of numbers taken.

[0091] If, in N consecutive workdays, a certain window or a certain matter exists for M workdays, and in the call log after a preset time before the end of the workday, there are A consecutive call intervals of less than B seconds, then there is a limit on the number of calls.

[0092] The steps described above for determining whether there is an insufficient number of available numbers include the following:

[0093] If, within N consecutive working days, the reservation slots for M working days are already full, then there is a shortage of available slots.

[0094] If, within N consecutive working days, the reservation slots for M working days are fully booked within C minutes, then there is an insufficient number of slots to be released.

[0095] The steps described above for determining whether there is a congestion of applications include the following:

[0096] The queuing time is obtained based on the time of obtaining the number and the time of being called.

[0097] If the queuing time exceeds the preset waiting time, it will be recorded as crowded.

[0098] Determine if there are M consecutive congested numbers; if so, the application is congested.

[0099] If the number of queuing numbers exceeds the preset number, and the window opening rate is less than 50%, and the queuing time exceeds the preset waiting time, then the application process will be crowded.

[0100] The aforementioned window open rate uses a sliding window mechanism, with a granularity of 5 minutes to calculate the window open rate. The window open rate is the ratio of the number of open windows to the total number of windows.

[0101] The steps described above for determining whether a "multiple trips" incident has occurred using case processing data include the following:

[0102] If, in N consecutive working days, there are M working days in which the same person handles the same type of matter more times than the preset number;

[0103] If the item is not on the "multiple trips" warning whitelist, then the "multiple trips" incident has occurred.

[0104] The aforementioned "multiple trips" warning whitelist may include the type of matter, the person handling the matter (such as exemption records for specific groups), and the service window (such as pilot windows during the system testing phase).

[0105] The aforementioned early warnings include generating alarm records that meet the corresponding warning conditions and sending alarm notifications via at least one of the following methods: audible and visual alarm devices, SMS, email, or system messages. For example, if an alarm is issued for absenteeism, the alarm record may include the time, location, and relevant video clips of the absence. The alarm record can also be stored in a database for later querying and analysis. If an alarm is issued for unreasonable resource allocation, the alarm record may include the reasons for the unreasonable allocation and relevant data. After an alarm record is generated, it is sent to the relevant departments for verification and feedback on rectification. Alarm records can be uniformly displayed through a visual interface. In addition to aggregating monitoring videos from all administrative service centers, the alarm data will also be uniformly displayed according to different dimensions such as service quality and service efficiency. The verification department can view the video access status and alarm data for its area through the visual interface.

[0106] Taking absenteeism as an example, the specific methods and technical details for displaying early warning information from a spatiotemporal perspective are as follows: A heatmap is used to present the early warning density, generated using the kernel density estimation (KDE) algorithm to calculate the spatial distribution of early warning points. Color levels are divided according to the number of early warnings: green (0-3 times), yellow (4-6 times), and red (≥7 times), intuitively reflecting the concentration of absenteeism early warnings in different areas. The timeline has a backtracking function, supporting both natural days and working days on a dual time scale, facilitating the viewing of early warning situations across different time dimensions. Furthermore, the α-β pruning algorithm optimizes the display of overlapping labels for key events, ensuring clear and readable timeline information. Simultaneously, an early warning penetration query function is designed. Clicking on the early warning icon on the heatmap or timeline allows users to drill down to view the corresponding original video clips (restoring the scene of absenteeism) and business data snapshots (such as job information and staff schedules for that period), enabling rapid tracing from macro-level early warning distribution to micro-level specific events, improving the efficiency of visual analysis and verification of absenteeism early warnings.

[0107] To achieve closed-loop management of early warning systems and improve compatibility in responding to emergencies, it is necessary to improve the verification and feedback mechanism and disaster recovery plan. In the rectification and feedback mechanism, a state machine for early warning work orders is designed, dividing the work order status into three stages: "Pending Confirmation," "Rectified," and "Review Passed," forming a complete process from early warning triggering to problem resolution. Simultaneously, timeout escalation rules are set (if a work order remains unprocessed in the "Pending Confirmation" state for more than a preset time, it is automatically escalated to a higher level of supervision). Regarding the disaster recovery plan, the video access layer adopts an exponential backoff retry strategy (initial retry interval of 2 seconds, maximum of 5 retries, with each retry interval increasing exponentially) to ensure rapid recovery after video surveillance data transmission interruptions. Furthermore, a circular buffer is used for video caching, retaining the most recent 5 minutes of data by default to prevent video surveillance data loss. Appointment number data and case data are stored locally with a retransmission mechanism during network fluctuations, and then re-uploaded after network recovery, ensuring data integrity. These designs improve the system's feedback efficiency and emergency response capabilities.

[0108] See Figure 4 As shown, this application also provides a performance analysis system based on business data and video surveillance, implementing the aforementioned performance analysis method based on business data and video surveillance. Specifically, it may include a data interface service module, a data ETL service module, a data analysis service module, an early warning service module, a visualization service module, a weekday service module, a service center resource module, a task resource module, a camera resource module, a service window resource module, a message announcement module, and a system management module, etc.

[0109] The aforementioned data interface service module serves as the connection hub between the system and external data sources. It is responsible for connecting to various external system interfaces, including business systems (such as appointment systems and case handling systems), video surveillance equipment (cameras), etc., to achieve standardized access to raw data such as appointment number data, case handling data, and video stream data, ensuring that data can enter the system stably and efficiently.

[0110] The aforementioned data ETL service module is responsible for data extraction, transformation, and loading. After obtaining raw data from the data interface service module, it performs cleaning (such as processing dirty data in appointment number data and correcting abnormal timestamps), transformation (such as converting video pixel coordinates to actual planar coordinates and unifying data formats), and integration (linking the time / spatial dimensions of business data and video data). Finally, the processed data is loaded into the system database to provide high-quality data for subsequent analysis.

[0111] The aforementioned data analysis service module is responsible for executing various performance analysis algorithms. Based on video surveillance data, it runs YOLOv7 object detection and DeepSORT multi-object tracking algorithms to output personnel pixel coordinates; it uses ray casting to determine whether personnel have entered the restricted area or left their posts. Based on business data, it analyzes appointment number data to determine the rationality of resource allocation (limited number of appointments, insufficient number of appointments, congestion of cases, etc.), analyzes case data to identify "multiple trips" events; and, combined with workday configuration, calculates the duration of personnel leaving their posts to determine whether early warning conditions have been triggered.

[0112] Based on the results of the data analysis service module and the preset warning rules (such as excessive absence from duty, unreasonable resource allocation, and "multiple trips" events), the aforementioned early warning service module triggers an early warning and generates an early warning work order; it manages the work order status machine (pending confirmation → rectified → approved), executes the timeout escalation rules, and pushes the early warning information to relevant modules (such as the message announcement module and the visualization service module).

[0113] The aforementioned visualization service module presents the analysis results and early warning information in an intuitive form, generating heatmaps (displaying early warning density based on kernel density estimation) and timelines (supporting dual time scales and key event markers); it provides an early warning penetration query function, allowing users to drill down to view original video clips and business data snapshots by clicking on the early warning icon; and it displays data reports such as resource configuration analysis results and case handling efficiency statistics to help users quickly understand the performance status.

[0114] The aforementioned workday service module manages the time base configuration, including the division of workdays / holidays and the business characteristics of different time periods (such as peak hours and off-peak hours), providing a time dimension basis for data analysis and early warning judgment (such as resource configuration rules that distinguish between workdays and holidays).

[0115] The aforementioned resource management modules (Service Center / Items / Camera / Service Window Resource Modules) are responsible for managing the entity resource information involved in the system, providing context for analysis.

[0116] Service Center Resource Module: Stores basic information about the storage center (such as name, address, and jurisdiction), and associates it with its included resources such as service windows and cameras;

[0117] The Item Resources module manages information on available business items (such as item name, required materials, processing procedures, and time limits), serving as a benchmark for analyzing "multiple trips" incidents.

[0118] Camera resource module: Records the camera's location, coverage area, and parameters (such as resolution and installation angle), and associates them with its corresponding video stream data to provide device attributes for video analysis;

[0119] Service Window Resource Module: Maintains the service window number, corresponding position, responsibilities, and staff information, and defines the planar coordinates of its work area as a spatial reference for determining whether a person is absent from their post.

[0120] The aforementioned message announcement module is responsible for pushing messages and publishing announcements within the system, pushing early warning notifications (such as "a staff member at a certain window is absent from their post"), work order status change reminders (such as "the work order has been upgraded"); and publishing system announcements (such as rule updates and maintenance notices) to ensure that relevant personnel can obtain key information in a timely manner.

[0121] The aforementioned system management module is the foundation for ensuring the normal operation of the system. It includes functions such as user management (account and permission allocation), system configuration (e.g., early warning rule parameters, ETL processing thresholds), log management (operation records, error logs), and data backup and recovery, ensuring system security, stability, and maintainability.

[0122] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. A performance analysis method based on business data and video surveillance, characterized in that, Includes the following steps: Efficiency data is obtained from the early warning database, including appointment number data, video surveillance data, and case processing data. The system determines whether performance data meets the warning conditions based on the weekday configuration and preset warning rules. If the warning conditions are met, an early warning will be issued, and verification and feedback will be provided. The preset rules specifically include: The algorithm for detecting absenteeism is used to determine whether absenteeism exists in video surveillance data. If it does, an early warning condition is met; or The reservation number data is used to determine whether resource allocation is reasonable; if it is unreasonable, the warning conditions are met. The system uses case processing data to determine if there are any "multiple trips" incidents. If so, the system meets the warning criteria.

2. The performance analysis method based on business data and video surveillance according to claim 1, characterized in that, The step of using the personnel absence / absence algorithm to determine whether there is absence / absence in video surveillance data includes the following steps: Define at least one warning zone and / or a post work area; Based on the YOLOv7 target detection model and DeepSORT multi-target tracking algorithm, the system locates and tracks people in video surveillance data to obtain the pixel coordinates of people in the video surveillance data. The pixel coordinates of people in video surveillance data are mapped to actual planar coordinates based on the perspective transformation matrix; Determine whether personnel have entered the restricted area and / or left their work area based on actual planar coordinates; If so, determine whether the time spent entering the restricted area and / or leaving the work area exceeds the preset time; If so, it is determined that there has been a case of being absent from one's post.

3. The performance analysis method based on business data and video surveillance according to claim 1, characterized in that, The step of using reservation number data to determine whether resource allocation is reasonable includes judging one or more of the following conditions: Determine if there are any restrictions on vehicle registration; if so, determine if the resource allocation is unreasonable. Determine if there is an insufficient number of available numbers; if so, determine if resource allocation is unreasonable. Determine if there is congestion in the processing of applications; if so, determine if resource allocation is unreasonable.

4. The performance analysis method based on business data and video surveillance according to claim 3, characterized in that, The step of determining whether a number is subject to a restriction includes the following steps: If the number of numbers issued is the same in N consecutive working days, then there is a limit to the number of numbers issued; or If no number is recorded after the preset time before the end of the workday in N consecutive workdays, then there is a limit on the number of tickets that can be taken; or If, within N consecutive workdays, a single person has multiple records of taking a number after a preset time before the end of the workday, then a limit on the number of numbers is applied; or If, within N consecutive workdays, there are records of taking numbers after a preset time before the end of the workday, and no numbers are called or all numbers have passed, then there is a limit on the number of numbers taken. If, in N consecutive workdays, a certain window or a certain matter exists for M workdays, and in the call log after a preset time before the end of the workday, there are A consecutive call intervals of less than B seconds, then there is a limit on the number of calls.

5. The performance analysis method based on business data and video surveillance according to claim 3, characterized in that, The step of determining whether there is an insufficient number of license plates to be issued includes the following steps: If, within N consecutive working days, there are M working days where the appointment slots are full, then there is a shortage of available slots. If, within N consecutive working days, the reservation slots for M working days are fully booked within C minutes, then there is an insufficient number of slots to be released.

6. The performance analysis method based on business data and video surveillance according to claim 3, characterized in that, The step of determining whether the application is congested includes the following steps: The queuing time is obtained based on the time of obtaining the number and the time of being called. If the queuing time exceeds the preset waiting time, it will be recorded as crowded. Determine if there are M consecutive congested numbers; if so, the application is congested. If the number of queuing numbers exceeds the preset number, and the window opening rate is less than 50%, and the queuing time exceeds the preset waiting time, then the application process will be crowded.

7. The performance analysis method based on business data and video surveillance according to claim 1, characterized in that, The step of using case data to determine whether a "multiple trips" incident has occurred includes the following steps: If, in N consecutive working days, there are M working days in which the same person handles the same type of matter more times than the preset number; If it is not on the "multiple trips" warning whitelist, then there is a "multiple trips" incident.

8. The performance analysis method based on business data and video surveillance according to claim 1, characterized in that, The video surveillance data is obtained from the video supervision platform and is included in the basic database for early warning; the appointment number data is obtained from the queuing and calling system and is included in the basic database for early warning; the case processing data is obtained from the service item database and is included in the basic database for early warning.

9. The performance analysis method based on business data and video surveillance according to claim 1, characterized in that, The early warning includes generating alarm records that meet the corresponding early warning conditions and sending alarm notifications through at least one of the following methods: audible and visual alarm devices, SMS, email, or system messages.

10. A performance analysis system based on business data and video surveillance, implementing the performance analysis method based on business data and video surveillance as described in any one of claims 1-9.

Images