A seamless positioning management method and system for personnel indoors and outdoors of a converter station

Abstract

The present invention discloses a seamless positioning management method and system for indoor and outdoor personnel of a converter station, which relates to the technical field of personnel positioning, including: obtaining the topological structure of the converter station, deploying video monitoring equipment based on the topological structure of the converter station, and forming a seamless monitoring network; performing real-time monitoring based on the seamless monitoring network to obtain a real-time video stream; preprocessing the real-time video stream to obtain a standard real-time video stream, which is divided into a standard indoor real-time video stream of the converter station and a standard outdoor real-time video stream of the converter station; identifying personnel on the standard real-time video stream and marking personnel; tracking based on the personnel identification to obtain the movement trajectory of the personnel, and positioning the personnel based on the movement trajectory of the personnel. The present invention solves the technical problem of insufficient real-time personnel positioning in the prior art, and achieves the technical effect of accurate positioning and real-time tracking of personnel.

Description

Technical Field

[0001] The present invention relates to the technical field of personnel positioning, and in particular to a seamless positioning management method and system for personnel inside and outside a converter station. Background Art

[0002] With the development of modern power systems, converter stations play a vital role in power transmission and distribution. Converter stations are the core part of high-voltage direct current transmission systems and are responsible for the conversion and control of electric energy. Due to the complex equipment and special environment in converter stations, real-time monitoring and management of their operating status have become particularly important. Especially in terms of safety management, real-time grasp of the location and movements of personnel inside and outside the converter station and ensuring safe operation of personnel in key areas are the key to ensuring the safe operation of converter stations.

[0003] Traditional personnel positioning methods mainly rely on technologies such as RFID, GPS or Bluetooth. These methods can locate and track personnel to a certain extent, but they have the problem of insufficient real-time performance. Summary of the invention

[0004] The present application provides a method and system for seamless positioning management of personnel indoors and outdoors in a converter station, which is used to solve the technical problem of insufficient real-time positioning of personnel in the prior art.

[0005] In view of the above problems, the present application provides a method and system for seamless positioning management of personnel inside and outside a converter station.

[0006] In a first aspect of the present application, a method for seamless positioning management of personnel indoors and outdoors of a converter station is provided, the method comprising:

[0007] The converter station topology structure is obtained, and video surveillance equipment is deployed based on the converter station topology structure to form a seamless monitoring network; real-time monitoring is performed based on the seamless monitoring network to obtain real-time video streams, and the real-time video streams are divided into real-time video streams indoors of the converter station and real-time video streams outdoors of the converter station; the real-time video streams are pre-processed to obtain standard real-time video streams, and the standard real-time video streams are divided into real-time video streams indoors of the standard converter station and real-time video streams outdoors of the standard converter station; personnel recognition is performed on the standard real-time video streams to identify personnel; tracking is performed based on the personnel identification to obtain the movement trajectory of the personnel, and personnel positioning is performed based on the movement trajectory of the personnel.

[0008] A second aspect of the present application provides a seamless positioning management system for personnel indoors and outdoors of a converter station, the system comprising:

[0009] A seamless monitoring network establishment module, the seamless monitoring network establishment module obtains the converter station topology structure, and deploys video monitoring equipment based on the converter station topology structure to form a seamless monitoring network; a real-time video stream acquisition module, the real-time video stream acquisition module performs real-time monitoring based on the seamless monitoring network to obtain real-time video streams, and the real-time video streams are divided into real-time video streams indoors in the converter station and real-time video streams outdoors in the converter station; a preprocessing module, the preprocessing module preprocesses the real-time video streams to obtain standard real-time video streams, and the standard real-time video streams are divided into real-time video streams indoors in the standard converter station and real-time video streams outdoors in the standard converter station; a personnel identification module, the personnel identification module recognizes personnel in the standard real-time video streams and performs personnel identification; a personnel positioning module, the personnel positioning module tracks based on the personnel identification, obtains personnel movement trajectories, and locates personnel based on the personnel movement trajectories.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] The present application obtains the converter station topology structure, and arranges video monitoring equipment based on the converter station topology structure to form a seamless monitoring network; performs real-time monitoring based on the seamless monitoring network to obtain real-time video streams, which are divided into indoor real-time video streams of the converter station and outdoor real-time video streams of the converter station; pre-processes the real-time video streams to obtain standard real-time video streams, which are divided into indoor real-time video streams of standard converter stations and outdoor real-time video streams of standard converter stations; performs personnel recognition and personnel identification on the standard real-time video streams; tracks based on personnel identification to obtain personnel movement trajectories, and locates personnel based on personnel movement trajectories. The present invention solves the technical problem of insufficient real-time personnel positioning in the prior art, and achieves the technical effect of accurate positioning and real-time tracking of personnel by reasonably arranging monitoring equipment, image processing and tracking algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A flow chart of a seamless positioning management method for personnel inside and outside a converter station provided in an embodiment of the present application;

[0014] Figure 2 A schematic diagram of the structure of a seamless positioning management system for personnel indoors and outdoors of a converter station provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: seamless monitoring network establishing module 11, real-time video stream acquiring module 12, pre-processing module 13, personnel identification module 14, personnel positioning module 15. DETAILED DESCRIPTION

[0016] The present application provides a seamless positioning management method and system for personnel indoors and outdoors of a converter station, aiming to solve the technical problem of insufficient real-time positioning of personnel in the prior art, and achieves the technical effect of accurate positioning and real-time tracking of personnel through reasonable deployment of monitoring equipment, image processing and tracking algorithms.

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0018] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0019] Embodiment 1

[0020] like Figure 1 As shown, the present application provides a method for seamless positioning management of personnel indoors and outdoors of a converter station, the method comprising:

[0021] Step S100: Acquire a converter station topology structure, and deploy video monitoring equipment based on the converter station topology structure to form a seamless monitoring network.

[0022] In an embodiment of the present application, the converter station topology is obtained based on the design drawings in the historical database. Based on the topology, the internal and external environment of the converter station is analyzed to determine the key areas that need to be monitored. These key areas include personnel entrances and exits, important equipment areas, work areas, etc. According to the monitoring requirements of each key area, select the appropriate type and number of cameras. Determine the specific installation position and angle of each camera based on the topology to ensure that the camera can cover all key areas and avoid monitoring blind spots. Lay the necessary network lines inside and outside the converter station to ensure that all cameras can be stably connected to the monitoring network. Through the above steps, video surveillance equipment is deployed inside and outside the converter station to form a seamless monitoring network.

[0023] Furthermore, step S100 in the method provided in the application embodiment also includes:

[0024] Acquire a plan view and a three-dimensional model of the converter station, and determine key monitoring points inside and outside the converter station based on the plan view and the three-dimensional model; deploy the video monitoring equipment at the key monitoring points to form a seamless monitoring network.

[0025] In the embodiment of the present application, the plan view of the converter station is first extracted from the architectural design drawings, and the three-dimensional model is obtained by extracting from the historical database.

[0026] After obtaining the floor plan and 3D model, use professional software such as AutoCAD, Revit or other BIM software to process the floor plan and 3D point cloud data to generate an accurate digital model. Mark important areas, facilities and channels in the 3D model. Based on these data, analyze the internal and external environment of the converter station and determine the key areas that need to be monitored, including entrances and exits, important equipment areas, work areas and safety areas.

[0027] After determining the key areas, select the appropriate camera type based on monitoring needs. For example, high-definition cameras are used to capture details, night vision cameras are used in low-light environments, and 360-degree panoramic cameras are used for wide-area monitoring. Simulate the installation position and angle of the cameras in the 3D model to ensure that each camera can cover the entire monitoring area and avoid monitoring blind spots. When laying out video monitoring equipment, lay the necessary network lines inside and outside the converter station to ensure that all cameras can be stably connected to the monitoring network.

[0028] Through the above process, the deployment of video surveillance equipment is completed to form a seamless monitoring network.

[0029] Step S200: Real-time monitoring is performed based on the seamless monitoring network to obtain a real-time video stream, wherein the real-time video stream is divided into a real-time video stream indoors at the converter station and a real-time video stream outdoors at the converter station.

[0030] In the embodiment of the present application, after the seamless monitoring network is deployed, the entire monitoring system is started, and all installed cameras start working to continuously capture and transmit live video.

[0031] Real-time video streams inside and outside the converter station are continuously acquired through cameras in the seamless monitoring network. These video streams can be classified into two categories according to the layout of the cameras and the monitoring area: indoor real-time video streams of the converter station and outdoor real-time video streams of the converter station. The indoor real-time video stream of the converter station captures the activities of personnel and the operation of equipment in various areas inside the converter station, including important equipment areas, work areas, and safety areas. The outdoor real-time video stream of the converter station captures the entry and exit of personnel and the surrounding environment outside the converter station, including the main entrances and exits and the peripheral safety area.

[0032] Step S300: pre-processing the real-time video stream to obtain a standard real-time video stream, wherein the standard real-time video stream is divided into a standard converter station indoor real-time video stream and a standard converter station outdoor real-time video stream.

[0033] In an embodiment of the present application, the real-time video stream collected is denoised, and filtering algorithms such as Gaussian filtering and median filtering are used to remove noise in the video to improve the clarity and quality of the image. Image enhancement techniques such as histogram equalization and contrast stretching are used to enhance the video. Frame synchronization is performed on video streams from different cameras, and the time sequence of video frames is adjusted according to the timestamp information of the camera to ensure the temporal consistency of different video streams. The video streams from different cameras are uniformly converted to a standard format, including the unification of resolution, frame rate and encoding format.

[0034] Finally, the preprocessed standard real-time video stream is classified into standard converter station indoor real-time video stream and standard converter station outdoor real-time video stream.

[0035] Step S400: performing personnel recognition and identification on the standard real-time video stream.

[0036] Furthermore, step S400 in the method provided in the application embodiment also includes:

[0037] A convolutional neural network is used to detect people in the standard real-time video stream, identify and locate people in the video; and the detected people are uniquely identified and assigned ID tags.

[0038] In an embodiment of the present application, the convolutional neural network is a pre-trained target detection model, such as YOLO, SSD or Faster R-CNN. Personnel detection is performed on a standard real-time video stream through a convolutional neural network, the real-time video stream is split into continuous frames, and the CNN model is applied to each frame for personnel detection. Use target tracking algorithms such as SORT and DeepSORT to track the detected personnel across frames to obtain their continuous position information. A unique ID tag is assigned to each person who is detected and successfully tracked. When a person temporarily leaves the field of view and then reappears, the original ID tag is restored by matching the saved feature information, such as the appearance model.

[0039] Furthermore, in the method provided in the embodiment of the application, the use of a convolutional neural network to perform personnel detection on the standard real-time video stream also includes:

[0040] An image data set containing people is collected, and an image recognition model is obtained for the convolutional neural network based on the image data set; the standard real-time video stream is inferred by the image recognition model to identify and locate the people in the video.

[0041] In an embodiment of the present application, a dataset of images of people at different angles, different postures, and different lighting conditions is collected from a historical database. These datasets cover various possible scenes of people's activities indoors and outdoors in the converter station. Professional annotation tools are used to finely annotate the collected images, and the annotation content includes the bounding box and category label of the person. The annotated images are normalized to ensure that the pixel value of the image is between 0 and 1.

[0042] For the task of personnel detection, select the existing mature CNN architecture, such as YOLO, SSD, Faster R-CNN, etc. Use the preprocessed data set to train the CNN model. During the training process, select the appropriate training strategy according to actual needs and computing resources, such as setting the appropriate batch size, learning rate, number of iterations, etc. The image recognition model is obtained through training.

[0043] Next, the standard real-time video stream is split into consecutive frames, and each frame is preprocessed, such as scaling and normalization. The preprocessed frames are input into the image recognition model for inference, and the bounding box of each detected person is output. Finally, a unique ID tag is assigned to each detected person by recording the bounding box corresponding to each ID tag in the database. At the same time, the specific location and time of each detected person in the video are determined by combining the timestamp and frame number in the video stream.

[0044] Step S500: Tracking is performed based on the personnel identification to obtain the movement trajectory of the personnel, and the personnel is located based on the movement trajectory of the personnel.

[0045] Furthermore, step S500 in the method provided in the application embodiment also includes:

[0046] According to the personnel identification, a multi-target tracking algorithm is used to track the personnel to obtain continuous position data of each identified person; the indoor video stream and the outdoor video stream of the converter station are spliced ​​to perform seamless tracking inside and outside the converter station; based on the tracking data in the spliced ​​video stream, the movement trajectory of each identified person is generated.

[0047] In an embodiment of the present application, when a multi-target tracking algorithm is used to track a person according to the person identification, an algorithm suitable for multi-target tracking, such as SORT, Deep SORT, FairMOT, etc., is selected to track each person and update their position information. These algorithms are based on the motion model of the target, such as the Kalman filter and appearance features, such as color and shape, to associate the target. For each tracked person, its position information in each frame is recorded, such as the coordinates of the center point of the bounding box. These position information will be saved in chronological order to form continuous position data of the person.

[0048] Next, use video processing software or libraries such as OpenCV to stitch the indoor and outdoor video streams together. In the stitched video streams, continue to use the multi-target tracking algorithm to track people. Since the indoor and outdoor video streams have been stitched together, seamless tracking inside and outside the converter station is achieved.

[0049] For each tracked person, its action trajectory is recorded based on its continuous position data in the spliced ​​video stream. The trajectory consists of a series of position points and timestamps.

[0050] Furthermore, in the method provided in the embodiment of the application, the indoor video stream and the outdoor video stream of the converter station are spliced ​​and processed to perform seamless tracking inside and outside the converter station, and further includes:

[0051] The indoor video stream and the outdoor video stream are time-synchronized according to the timestamps of the cameras inside and outside the converter station; a unified coordinate system is established based on the plan view and three-dimensional model of the converter station, and the coordinates in the indoor video stream and the outdoor video stream are converted; the indoor video stream and the outdoor video stream that have undergone time synchronization and coordinate conversion are spliced ​​and merged to generate a unified spliced ​​video stream.

[0052] In an embodiment of the present application, first, the timestamp information of the camera is used to synchronize the indoor and outdoor video streams. Each camera will be accompanied by a timestamp when shooting a video, and these timestamps ensure the accurate recording of the video frames in time. By analyzing and aligning these timestamps, it is ensured that the corresponding frames of the video streams of different cameras at the same time point are synchronized, thereby achieving temporal consistency. Next, a unified coordinate system is established based on the floor plan and three-dimensional model of the converter station. The floor plan and three-dimensional model provide detailed spatial layout and structural information of the converter station. Using this information, a global coordinate system is defined in three-dimensional space. The position, viewing angle, and capture range of each camera can be mapped to this global coordinate system. In this process, each camera is accurately calibrated to determine its position and direction in the unified coordinate system. Through this coordinate conversion step, the coordinate data in the indoor and outdoor video streams are unified so that the data of all cameras are represented in the same coordinate system.

[0053] After completing time synchronization and coordinate conversion, the video streams are spliced ​​and fused. First, the synchronized indoor and outdoor video frames are spliced ​​according to a unified timestamp. Since the position and viewing angle of each camera have been calibrated in a unified coordinate system, the positions of people in the video frames can be seamlessly connected, ensuring that their positions can be continuously tracked when they cross indoor and outdoor areas. In the process of video splicing, fusion processing is required, including color correction, brightness adjustment, and edge smoothing, to ensure that the spliced ​​video is visually coherent and eliminate splicing marks.

[0054] Through the above steps, the seamless splicing and fusion of indoor and outdoor video streams are achieved to generate a unified spliced ​​video stream.

[0055] Furthermore, in the method provided in the embodiment of the application, based on the tracking data in the spliced ​​video stream, generating the movement trajectory of each identified person, further includes:

[0056] Based on the spliced ​​video stream, the position information of each identified person is extracted to generate original trajectory data; the original trajectory data is smoothed to obtain a smoothed person movement trajectory; based on the smoothed person movement trajectory, the real-time position of the person is determined to complete the positioning of the person.

[0057] In the embodiment of the present application, a multi-target tracking algorithm, such as DeepSORT, is first used to identify and track the position of each marked person in the spliced ​​video stream, including the bounding box coordinates and the corresponding timestamp. Through this information, the original trajectory data of each marked person is generated, which contains the position information of the person at different time points, forming a continuous movement trajectory.

[0058] After the raw trajectory data is generated, it is smoothed to reduce noise and errors and ensure the continuity and accuracy of the trajectory. Commonly used smoothing methods include Kalman filtering and sliding average. Kalman filtering can predict and update each position point, reduce measurement errors and noise in the trajectory, and make the trajectory smoother and more stable. Sliding average further smoothes the processed trajectory to eliminate short-term fluctuations and make the trajectory more natural and continuous. Through these smoothing steps, a smoothed action trajectory of each identified person is obtained, eliminating noise and jump points in the raw data.

[0059] After obtaining the smoothed trajectory, the real-time location of each identified person is determined based on the trajectory data. The determination of the real-time location depends on the latest location points in the trajectory data, which contain the current coordinates and timestamp of the person. By analyzing the latest points in the smoothed trajectory, the current real-time location of each identified person is determined. Furthermore, by mapping these location data to the floor plan and three-dimensional model of the converter station, the specific area where the person is located can be determined and accurate positioning can be achieved.

[0060] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0061] The present application obtains the converter station topology structure, and arranges video monitoring equipment based on the converter station topology structure to form a seamless monitoring network; performs real-time monitoring based on the seamless monitoring network to obtain real-time video streams, which are divided into indoor real-time video streams of the converter station and outdoor real-time video streams of the converter station; pre-processes the real-time video streams to obtain standard real-time video streams, which are divided into indoor real-time video streams of standard converter stations and outdoor real-time video streams of standard converter stations; performs personnel recognition and personnel identification on the standard real-time video streams; tracks based on personnel identification to obtain personnel movement trajectories, and locates personnel based on personnel movement trajectories. The present invention solves the technical problem of insufficient real-time personnel positioning in the prior art, and achieves the technical effect of accurate positioning and real-time tracking of personnel by reasonably arranging monitoring equipment, image processing and tracking algorithms.

[0062] Embodiment 2

[0063] Based on the same inventive concept as the seamless positioning management method for indoor and outdoor personnel of a converter station in the aforementioned embodiment, Figure 2 As shown, the present application provides a seamless positioning management system for personnel inside and outside a converter station. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0064] A seamless monitoring network establishing module 11, wherein the seamless monitoring network establishing module 11 obtains a converter station topology structure, and deploys video monitoring equipment based on the converter station topology structure to form a seamless monitoring network;

[0065] A real-time video stream acquisition module 12, wherein the real-time video stream acquisition module 12 performs real-time monitoring based on the seamless monitoring network to obtain a real-time video stream, wherein the real-time video stream is divided into a real-time video stream indoors at the converter station and a real-time video stream outdoors at the converter station;

[0066] A preprocessing module 13, wherein the preprocessing module 13 preprocesses the real-time video stream to obtain a standard real-time video stream, wherein the standard real-time video stream is divided into a standard converter station indoor real-time video stream and a standard converter station outdoor real-time video stream;

[0067] A personnel identification module 14, wherein the personnel identification module 14 performs personnel recognition and personnel identification on the standard real-time video stream;

[0068] The personnel positioning module 15 tracks the personnel identification to obtain the movement trajectory of the personnel, and locates the personnel based on the movement trajectory of the personnel.

[0069] Furthermore, the system is also used to implement the following functions:

[0070] Obtaining a plan view and a three-dimensional model of the converter station, and determining key monitoring points inside and outside the converter station based on the plan view and the three-dimensional model;

[0071] The video surveillance equipment is deployed at the key monitoring points to form a seamless monitoring network.

[0072] Furthermore, the system is also used to implement the following functions:

[0073] Using a convolutional neural network to perform personnel detection on the standard real-time video stream, and identify and locate the personnel in the video;

[0074] Uniquely identify the detected person and assign an ID tag.

[0075] Furthermore, the system is also used to implement the following functions:

[0076] Collecting an image dataset containing people, and obtaining an image recognition model based on the convolutional neural network.

[0077] The standard real-time video stream is inferred by the image recognition model to identify and locate people in the video.

[0078] Furthermore, the system is also used to implement the following functions:

[0079] According to the personnel identification, a multi-target tracking algorithm is used to track the personnel to obtain continuous position data of each identified person;

[0080] Splice the indoor and outdoor video streams of the converter station to achieve seamless tracking inside and outside the converter station;

[0081] Based on the tracking data in the spliced ​​video stream, the movement trajectory of each identified person is generated.

[0082] Furthermore, the system is also used to implement the following functions:

[0083] Performing time synchronization on the indoor video stream and the outdoor video stream according to the timestamps of the cameras inside and outside the converter station;

[0084] Based on the plan view and the three-dimensional model of the converter station, a unified coordinate system is established to convert the coordinates in the indoor video stream and the outdoor video stream;

[0085] The indoor video stream and the outdoor video stream that have undergone time synchronization and coordinate transformation are spliced ​​and merged to generate a unified spliced ​​video stream.

[0086] Furthermore, the system is also used to implement the following functions:

[0087] Extracting the location information of each identified person based on the spliced ​​video stream to generate original trajectory data;

[0088] Smoothing the original trajectory data to obtain a smoothed personnel motion trajectory;

[0089] According to the smoothed movement trajectory of the personnel, the real-time position of the personnel is determined to complete the positioning of the personnel.

[0090] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0091] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0092] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A seamless positioning management method for personnel inside and outside a converter station, characterized in that: The method comprises: Acquire a converter station topology structure, and deploy video surveillance equipment based on the converter station topology structure to form a seamless surveillance network; Real-time monitoring is performed based on the seamless monitoring network to obtain a real-time video stream, wherein the real-time video stream is divided into a real-time video stream indoors at the converter station and a real-time video stream outdoors at the converter station; Preprocessing the real-time video stream to obtain a standard real-time video stream, wherein the standard real-time video stream is divided into a standard converter station indoor real-time video stream and a standard converter station outdoor real-time video stream; Performing personnel recognition and identification on the standard real-time video stream; Tracking is performed based on the personnel identification to obtain the movement trajectory of the personnel, and the personnel is located based on the movement trajectory of the personnel.

2. A seamless positioning management method for indoor and outdoor personnel of a converter station as claimed in claim 1, characterized in that: The obtaining of the converter station topology structure and the deployment of video monitoring equipment based on the converter station topology structure to form a seamless monitoring network include: Obtaining a plan view and a three-dimensional model of the converter station, and determining key monitoring points inside and outside the converter station based on the plan view and the three-dimensional model; The video surveillance equipment is deployed at the key monitoring points to form a seamless monitoring network.

3. A seamless positioning management method for indoor and outdoor personnel of a converter station as claimed in claim 1, characterized in that: The performing personnel recognition and personnel identification on the standard real-time video stream includes: Using a convolutional neural network to perform personnel detection on the standard real-time video stream, and identify and locate the personnel in the video; Uniquely identify the detected person and assign an ID tag.

4. A method for seamless positioning management of personnel inside and outside a converter station as claimed in claim 3, characterized in that: The method of using a convolutional neural network to perform personnel detection on the standard real-time video stream includes: Collecting an image dataset containing people, and obtaining an image recognition model based on the convolutional neural network. The standard real-time video stream is inferred by the image recognition model to identify and locate people in the video.

5. A seamless positioning management method for personnel inside and outside a converter station as claimed in claim 1, characterized in that: The tracking based on the personnel identification to obtain the movement trajectory of the personnel includes: According to the personnel identification, a multi-target tracking algorithm is used to track the personnel to obtain continuous position data of each identified person; Splice the indoor and outdoor video streams of the converter station to achieve seamless tracking inside and outside the converter station; Based on the tracking data in the spliced ​​video stream, the movement trajectory of each identified person is generated.

6. A method for seamless positioning management of personnel inside and outside a converter station as claimed in claim 5, characterized in that: The indoor and outdoor video streams of the converter station are spliced ​​and processed to achieve seamless tracking inside and outside the converter station, including: Performing time synchronization on the indoor video stream and the outdoor video stream according to the timestamps of the cameras inside and outside the converter station; Based on the plan view and the three-dimensional model of the converter station, a unified coordinate system is established to convert the coordinates in the indoor video stream and the outdoor video stream; The indoor video stream and the outdoor video stream that have undergone time synchronization and coordinate transformation are spliced ​​and merged to generate a unified spliced ​​video stream.

7. A method for seamless positioning management of personnel inside and outside a converter station as claimed in claim 5, characterized in that: Based on the tracking data in the spliced ​​video stream, the movement trajectory of each identified person is generated, including: Extracting the location information of each identified person based on the spliced ​​video stream to generate original trajectory data; Smoothing the original trajectory data to obtain a smoothed personnel motion trajectory; According to the smoothed movement trajectory of the personnel, the real-time position of the personnel is determined to complete the positioning of the personnel.

8. A seamless positioning management system for personnel inside and outside a converter station, characterized in that: The system comprises: A seamless monitoring network establishment module, wherein the seamless monitoring network establishment module obtains a converter station topology structure, and deploys video monitoring equipment based on the converter station topology structure to form a seamless monitoring network; A real-time video stream acquisition module, wherein the real-time video stream acquisition module performs real-time monitoring based on the seamless monitoring network to obtain a real-time video stream, wherein the real-time video stream is divided into a real-time video stream indoors at the converter station and a real-time video stream outdoors at the converter station; A preprocessing module, wherein the preprocessing module preprocesses the real-time video stream to obtain a standard real-time video stream, wherein the standard real-time video stream is divided into a standard converter station indoor real-time video stream and a standard converter station outdoor real-time video stream; A personnel identification module, which performs personnel recognition and personnel identification on the standard real-time video stream; A personnel positioning module is used to track the personnel based on the personnel identification, obtain the movement trajectory of the personnel, and locate the personnel based on the movement trajectory of the personnel.

Images