3D visual medical rescue command system

Abstract

The invention relates to the field of rescue systems, and discloses a 3D visual medical rescue command system, which is used for solving the problems of untimely and incomplete information acquisition, low processing efficiency, insufficient decision support and the like in a traditional marathon event medical rescue command system, and provides an efficient and intelligent solution for modern medical rescue command. The obtaining module obtains personnel positioning and on-site video streams in real time and constructs a three-dimensional digital twinborn model. The synchronization module carries out timestamp alignment and spatial registration on the multi-source data; the distribution module generates a resource distribution scheme, a disaster diffusion prediction map and an unmanned aerial vehicle obstacle avoidance flight path according to the unified coordinate data; the path module dynamically renders multiple layers on the 3D display terminal; and the execution module issues a rescue path instruction, and controls the unmanned aerial vehicle swarm to execute tasks and deliver materials. According to the invention, intelligence and visualization of medical rescue command of the marathon event are realized, the rescue efficiency and accuracy are effectively improved, and scientific and efficient decision support is provided for emergency rescue.

Description

Technical Field

[0001] This invention relates to the field of rescue systems, and more particularly to a 3D visualization medical rescue command system. Background Technology

[0002] Marathon events, due to their long course distances, dense participant populations, and highly open environments, present significant challenges in terms of the spatial dispersion, time urgency, and complexity of emergency medical rescue needs. Traditional medical rescue models, relying on manual patrols, handheld GPS positioning, and two-dimensional electronic maps for command, face three major technological bottlenecks:

[0003] Insufficient spatial perception accuracy: Ordinary GPS positioning error reaches 5-10 meters, and it is easily affected by multipath effects from buildings in complex urban racetracks, leading to deviations in the location of injured persons. Existing 3D modeling technologies mostly use satellite remote sensing data, with a resolution of only 0.5-2 meters, making it difficult to accurately identify detailed terrain features such as narrow alleys and guardrails.

[0004] The fusion of dynamic information is lagging: there is a minute-level delay between manual reporting and drone video streams, and disaster spread prediction relies on empirical formulas, making it impossible to couple multi-dimensional dynamic parameters such as fire spread and personnel movement in real time. Resource scheduling often uses static path planning algorithms, which do not consider nonlinear factors such as real-time congestion and obstacle changes.

[0005] Fragmented command system: Positioning terminals, communication equipment, and visualization platforms operate independently, requiring manual input of casualty status updates across systems, easily creating information silos. Unmanned aerial vehicle (UAV) mission execution and resource allocation lack intelligent linkage, making it difficult to form a closed-loop collaborative system for material delivery and search and rescue.

[0006] Therefore, we propose a 3D visualization medical rescue command system to solve the above problems. Summary of the Invention

[0007] This invention provides a 3D visualization medical rescue command system to solve the problems of untimely and incomplete information acquisition, low processing efficiency and insufficient decision support in traditional medical rescue command systems, and provides an efficient and intelligent solution for modern medical rescue command.

[0008] The first aspect of this invention provides a 3D visualization medical rescue command system, comprising: an acquisition module for real-time acquisition of latitude, longitude, and elevation data of marathon emergency personnel and injured individuals, generating a BeiDou positioning data stream, collecting multi-view video streams from the accident scene, generating a UAV multispectral video stream, and generating a centimeter-level precision three-dimensional digital twin track model via a 3D real-scene GPS ground-based augmentation station; a synchronization module for timestamping the BeiDou positioning data stream, the UAV multispectral video stream, and the three-dimensional digital twin track model to generate a spatiotemporal synchronization dataset, generating a coordinate system dataset based on the BeiDou positioning data stream, and generating a spatial registration dataset based on the UAV multispectral video stream and the three-dimensional digital twin track model; and an allocation module for solving a system of differential equations for medical resource flow based on the coordinate system dataset as boundary conditions, generating a real-time resource allocation scheme, and using spatial... The thermal imaging data in the registration dataset serves as input parameters to define the fire spread rules for each grid cell, generating a disaster spread prediction map. Based on the building coordinates in the 3D digital twin track model, a drone obstacle avoidance flight path is generated. The path module dynamically renders the base layer, dynamic overlay layer, and interactive layer in the electronically controlled circularly polarized 3D display terminal based on the spatiotemporal synchronization dataset, real-time resource allocation scheme, disaster spread prediction map, and drone obstacle avoidance flight path. When the deviation of the injured person's location in the real-time resource allocation scheme exceeds a set threshold, the drone obstacle avoidance flight path is automatically replanned. The execution module distributes the optimized path in the real-time resource allocation scheme to the Beidou dual-mode positioning terminal, controlling the drone swarm to perform injured person search tasks and deliver medical supplies according to the drone obstacle avoidance flight path. In the electronically controlled circularly polarized 3D display terminal, the location and task status of marathon emergency personnel in the Beidou positioning data stream are updated in real time, forming a command closed loop.

[0009] Optionally, in the first implementation of the first aspect of the present invention, the method includes: real-time acquisition of latitude, longitude, and elevation data of marathon emergency responders and injured persons through a Beidou dual-mode positioning terminal to eliminate multipath interference and generate a Beidou positioning data stream; acquisition of visible light video stream and thermal imaging video stream of the accident scene through an infrared-visible light dual-mode gimbal mounted on a DJI drone, and synthesis of the dual-mode video stream into a drone multispectral video stream based on hardware-level timestamp alignment technology; and scanning of the race track area through a 3D real-scene GPS ground-based augmentation station, and generation of a three-dimensional digital twin race track model using a multi-sensor fusion reconstruction algorithm.

[0010] Optionally, in a second implementation of the first aspect of the present invention, the method includes: converting the geodetic coordinate system of the BeiDou positioning data stream into the local northeast-sky coordinate system of the three-dimensional digital twin track model based on the Lie group SE transform algorithm; extracting thermal imaging feature points from the UAV multispectral video stream using the SIFT feature point matching algorithm; and performing geometric projection matching between the image coordinates of the thermal imaging feature points and the spatial coordinates of the three-dimensional digital twin track model to generate a spatial registration dataset.

[0011] Optionally, in the third implementation of the first aspect of the present invention, the method includes: based on the system dynamics model, using the coordinate system dataset as boundary conditions, constructing a set of differential equations including resource supply, transportation speed, and path resistance to generate a real-time resource allocation scheme; based on the cellular automata model, using thermal imaging data in the spatial registration dataset as input parameters, defining fire spread rules for each grid cell, with a temperature threshold ≥150℃ and a wind direction weight factor of 0.3-0.7, to generate a disaster spread prediction map; and using a von Lono diagram segmentation algorithm with obstacle constraints to generate an obstacle avoidance flight path for the UAV based on the building coordinates in the three-dimensional digital twin track model.

[0012] Optionally, in a fourth implementation of the first aspect of the present invention, the method includes: dynamically rendering a base layer, a dynamic overlay layer, and an interaction layer in an electrically controlled circularly polarized 3D display terminal.

[0013] Base layer: Based on a coordinate system and a centralized 3D digital twin track model, the track terrain and building outlines are displayed with centimeter-level precision;

[0014] Dynamic overlay:

[0015] i) BeiDou positioning personnel marking: Based on the BeiDou positioning data stream in the spatiotemporal synchronized dataset, dots are used to mark the locations of marathon emergency personnel and injured people;

[0016] ii) Resource flow vector field: Based on the real-time resource allocation scheme, the arrow length represents the transportation speed and the color gradient represents the drug allocation priority;

[0017] iii) Fire risk heat map: Based on the disaster spread prediction map, the range of fire spread in the next 10 minutes is represented by a red gradient area;

[0018] Interaction layer:

[0019] i) UAV multispectral video stream pop-up: Embeds the real-time image of the UAV multispectral video stream into the corresponding geographic coordinates;

[0020] ii) Countdown label for estimated arrival time: Based on the drone's obstacle avoidance flight path, the arrival time is displayed floating above the target point;

[0021] The system monitors the positioning deviation of the wounded in the real-time resource allocation scheme in real time. When the deviation exceeds the threshold, it automatically sends a replanning command to the UAV obstacle avoidance flight path generation module. The replanned path is updated and displayed in real time through the countdown label in the interaction layer.

[0022] Optionally, in the fifth implementation of the first aspect of the present invention, the method includes: sending the optimized path instruction from the real-time resource allocation scheme to the Beidou dual-mode positioning terminal, wherein the optimized path instruction includes target coordinates, priority identifier and estimated arrival time; controlling the DJI drone swarm to perform the task according to the drone obstacle avoidance flight path: the main drone carries a medical supply bin and flies along the path to the coordinates of the injured person to perform airdrop; the auxiliary drone simultaneously activates the infrared-visible light dual-mode gimbal and transmits the search image back to the three-dimensional visualization command interface in real time; in the three-dimensional visualization command interface, the Beidou positioning data stream is analyzed in real time, the location markers of the marathon event emergency personnel are updated, the task status identifier is dynamically displayed, and when the airdrop of supplies is detected to be completed, the target point is automatically marked as rescued, and the status update data is synchronously fed back to the resource allocation model to trigger a new round of resource scheduling.

[0023] Optionally, in the sixth implementation of the first aspect of the present invention, it further includes: acquiring centimeter-level three-dimensional point cloud data of the area where the injured person is located in real time, registering the point cloud data with the three-dimensional digital twin track model in real time, and generating a high-precision positioning correction command; when the main UAV hovers at the target point, activating the image recognition mode of the visible light gimbal, performing sub-pixel-level positioning based on preset medical markers, and controlling the material warehouse release mechanism to automatically open the warehouse and deliver materials at a height of 5 meters above the ground in conjunction with the high-precision positioning correction command; monitoring the flight trajectory of the auxiliary aircraft group in real time, calculating the three-dimensional Euclidean distance between each UAV based on the obstacle avoidance flight path of the UAV, and automatically triggering a collision avoidance maneuver command when the distance between the UAVs is detected to be less than a safety threshold, and synchronously updating the path display of the interaction layer.

[0024] The mechanism of this invention is as follows:

[0025] Construct a cross-scale coupled model of system dynamics and cellular automata;

[0026] A two-layer path control mechanism combining von Lonoie diagram segmentation and dynamic Euclidean distance monitoring is proposed.

[0027] Establish a strong closed-loop architecture of perception → decision-making → execution → feedback.

[0028] Beneficial effects:

[0029] By using the Beidou dual-mode positioning terminal and the UAV multispectral gimbal, centimeter-level precision positioning of the location of emergency personnel and injured people in marathon events was achieved, as well as real-time acquisition of multi-angle video streams at accident sites, providing comprehensive and accurate data support for rescue command.

[0030] Based on system dynamics and cellular automata models, intelligent allocation of medical resources and prediction of disaster spread were realized, providing scientific decision support for rescue command.

[0031] A von Lono diagram segmentation algorithm with obstacle constraints was adopted to generate and replan the obstacle avoidance flight path of the UAV, ensuring the safe flight of the UAV in complex environments.

[0032] The electronically controlled circularly polarized 3D display terminal enables dynamic rendering and interaction of a three-dimensional visualization command interface, providing an intuitive and convenient operating platform for rescue command.

[0033] By updating the BeiDou positioning data stream and mission status in real time, a closed-loop command system was formed, ensuring the continuous optimization and adjustment of the rescue operation. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of one embodiment of the 3D visualization medical rescue command system of the present invention. Detailed Implementation

[0035] This invention provides a 3D visualization-based medical rescue command system to address the problems of untimely and incomplete information acquisition, low processing efficiency, and insufficient decision support in traditional medical rescue command systems, offering an efficient and intelligent solution for modern medical rescue command. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0036] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the 3D visualization medical rescue command system in this invention includes:

[0037] 101. Acquisition module, used to acquire latitude, longitude and elevation data of marathon emergency personnel and injured people in real time, generate Beidou positioning data stream, collect multi-view video stream of accident scene, generate UAV multispectral video stream, and generate a three-dimensional digital twin track model with centimeter-level accuracy through 3D real scene GPS ground-based augmentation station.

[0038] It is understood that the executing entity of this invention can be a 3D visualization medical rescue command device, a terminal, or a server; no specific limitation is made here. This embodiment of the invention will be described using a server as an example.

[0039] Specifically, the BeiDou positioning data stream is generated by real-time collection of latitude, longitude, and elevation data of marathon emergency personnel and injured individuals through a BeiDou dual-mode positioning terminal (integrated with BD3 / GPS dual-frequency chip, positioning accuracy ±0.3 meters); an adaptive Kalman filter algorithm is used to eliminate multipath interference and generate the BeiDou positioning data stream (data format is NMEA-0183 protocol, 1Hz update rate).

[0040] The drone multispectral video stream was generated by simultaneously acquiring visible light video streams and thermal imaging video streams from the accident scene at a frequency of 30Hz using an infrared-visible light dual-mode gimbal (thermal imaging resolution 640×512, visible light resolution 3840×2160) mounted on a DJI Matrice 350RTK drone. Based on hardware-level timestamp alignment technology, the dual-mode video streams were synthesized into a drone multispectral video stream (H.265 encoding, time synchronization error ≤1ms).

[0041] The construction of a 3D digital twin track model involves scanning the track area using a 3D real-scene GPS ground-based augmentation station (integrating a millimeter-wave radar array and a binocular stereo vision module); and using a multi-sensor fusion reconstruction algorithm (including ICP registration of radar point clouds and visible light images) to generate a 3D digital twin track model (spatial resolution ≤2cm, coordinate system is local ENU).

[0042] 102. Synchronization module, used to timestamp align the BeiDou positioning data stream, UAV multispectral video stream, and 3D digital twin track model to generate a spatiotemporal synchronization dataset, generate a coordinate system dataset based on the BeiDou positioning data stream, and generate a spatial registration dataset based on the UAV multispectral video stream and the 3D digital twin track model.

[0043] Specifically, the spatiotemporal synchronization dataset is generated by using the IEEE 1588v2 precision clock protocol to align the timestamps of the BeiDou positioning data stream (1Hz update rate), the UAV multispectral video stream (30fps frame rate), and the 3D digital twin track model (5Hz update rate) to generate a spatiotemporal synchronization dataset (synchronization error ≤2ms).

[0044] The coordinate system is generated by a dataset. Based on the Lie group SE(3) transformation algorithm, the geodetic coordinate system (WGS84) of the Beidou positioning data stream is transformed into the local northeast-sky coordinate system (ENU) of the three-dimensional digital twin track model. The transformation residual is ≤0.1 meters.

[0045] Spatial registration dataset generation: Thermal imaging feature points (temperature gradient ≥ 5℃ / pixel region) are extracted from the UAV multispectral video stream using the SIFT feature point matching algorithm; the image coordinates of the thermal imaging feature points are geometrically projected and matched with the spatial coordinates of the 3D digital twin track model to generate a spatial registration dataset (matching accuracy ≤ 3 pixels).

[0046] 103. Allocation module, used to solve the differential equations of medical resource flow based on a coordinate system dataset as boundary conditions, generate a real-time resource allocation scheme, define the fire spread rules for each grid cell using thermal imaging data in the spatial registration dataset as input parameters, generate a disaster spread prediction map, and generate the drone obstacle avoidance flight path based on the building coordinates in the 3D digital twin track model.

[0047] Specifically, the real-time resource allocation scheme is generated based on the system dynamics model, using a coordinate system dataset as boundary conditions, and constructing a system of differential equations including resource supply, transportation speed and path resistance. The solution generates a real-time resource allocation scheme (including ambulance dispatch priority and drug allocation ratio).

[0048] The disaster spread prediction map is generated based on the cellular automata model. It uses thermal imaging data in the spatial registration dataset as input parameters, defines the fire spread rules for each grid cell (temperature threshold ≥150℃, wind direction weight factor 0.3-0.7), and generates a disaster spread prediction map (including the fire spread range in the next 10 minutes).

[0049] The obstacle avoidance flight path generation for UAVs employs a von Lono diagram segmentation algorithm with obstacle constraints. Based on the building coordinates (accuracy ±0.1 meters) in the 3D digital twin track model, the UAV obstacle avoidance flight path is generated (the minimum distance between the path and obstacles is ≥5 meters).

[0050] 104. Path module, used in the electronically controlled circularly polarized 3D display terminal to dynamically render the base layer, dynamic overlay layer and interactive layer based on the spatiotemporal synchronization dataset, real-time resource allocation scheme, disaster spread prediction map and UAV obstacle avoidance flight path. When the positioning deviation of the wounded in the real-time resource allocation scheme exceeds the set threshold, the replanning of the UAV obstacle avoidance flight path is automatically triggered.

[0051] Specifically, a 3D visualization command interface is generated, and the following layers are dynamically rendered in an electronically controlled circularly polarized 3D display terminal (120Hz refresh rate, 3840×2160 resolution):

[0052] Base layer: Based on the generated coordinate system and a dataset of 3D digital twin track models, displaying track terrain and building outlines with centimeter-level precision;

[0053] Dynamic overlay:

[0054] i) BeiDou positioning personnel marking: Based on the BeiDou positioning data stream in the spatiotemporal synchronized dataset, the locations of marathon race emergency personnel and injured persons are marked with dots with an accuracy of ±0.3 meters;

[0055] ii) Resource flow vector field: Based on the real-time resource allocation scheme, the arrow length represents the transportation speed and the color gradient represents the drug allocation priority;

[0056] iii) Fire risk heat map: a disaster spread prediction map, with red gradient areas indicating the range of fire spread in the next 10 minutes;

[0057] Interaction layer:

[0058] i) UAV multispectral video stream pop-up: Embeds the real-time image of the UAV multispectral video stream into the corresponding geographic coordinates (offset error ≤ 3 pixels);

[0059] ii) Countdown label for estimated arrival time: Based on the drone's obstacle avoidance flight path, the arrival time is displayed floating above the target point (accuracy ±1 second);

[0060] The path replanning trigger mechanism monitors the casualty positioning deviation in the real-time resource allocation scheme (default threshold 0.5 meters). When the deviation exceeds the threshold, it automatically sends a replanning command to the UAV obstacle avoidance flight path generation module.

[0061] The replanned path is displayed in real time via a countdown label in the interactive layer.

[0062] 105. The execution module is used to send the optimized path in the real-time resource allocation scheme to the Beidou dual-mode positioning terminal, control the drone swarm to perform the casualty search mission according to the drone obstacle avoidance flight path, and deliver medical supplies. In the electronically controlled circularly polarized 3D display terminal, the location and mission status of marathon emergency personnel in the Beidou positioning data stream are updated in real time to form a command closed loop.

[0063] Specifically, the rescue route instructions are issued by sending the optimized route instructions in the real-time resource allocation scheme to the Beidou dual-mode positioning terminal through the Beidou RDSS short message protocol (working frequency band 1615.68MHz±4.08MHz, coding efficiency ≥95%) (transmission delay ≤500ms).

[0064] The optimized path instruction includes the target coordinates (accuracy ±0.3 meters), priority indicator, and estimated arrival time (error ±1 second);

[0065] The drone swarm works in a coordinated manner, controlling a DJI drone swarm (including one Matrice 350RTK main drone and four Mavic 3E auxiliary drones) to perform tasks according to the drone obstacle avoidance flight path:

[0066] The main drone carries a medical supply bin (5kg payload) and flies along the path to the coordinates of the injured (deviation ≤1 meter) to perform the airdrop;

[0067] The auxiliary machine simultaneously activates the infrared-visible light dual-mode gimbal and transmits the search image back to the three-dimensional visualization command interface in real time.

[0068] The command closed-loop status is updated in real time. In the three-dimensional visualization command interface, the Beidou positioning data stream is analyzed in real time to update the location markers of marathon emergency personnel (refresh rate 1Hz); the mission status markers (including "arrived" and "in transit") are dynamically displayed. When the airdrop of supplies is detected to be completed, the target point is automatically marked as "rescued"; the status update data is synchronously fed back to the resource allocation model to trigger a new round of resource scheduling.

[0069] It should be noted that the LiDAR (scanning frequency 20Hz, ranging accuracy ±2cm) carried by the Matrice 350RTK main UAV can acquire centimeter-level three-dimensional point cloud data of the area where the wounded are located in real time.

[0070] The point cloud data is registered with the 3D digital twin track model in real time to generate high-precision positioning correction instructions (coordinate correction amount ≤ 0.1 meters);

[0071] When the main UAV hovers over the target point, it activates the image recognition mode of the visible light gimbal and performs sub-pixel-level positioning based on the preset medical marker (reflective cross mark) (recognition error ≤ 3 pixels);

[0072] Based on the high-precision positioning correction command, the material warehouse release mechanism is controlled to automatically open and deliver materials when the material warehouse is 5 meters above the ground (the ground contact position deviation is ≤0.5 meters);

[0073] Real-time monitoring of the flight trajectory of the auxiliary aircraft group; based on the obstacle avoidance flight path of the UAVs, calculate the three-dimensional Euclidean distance between each UAV (update rate 10Hz);

[0074] When the distance between aircraft is detected to be less than the safety threshold (default 15 meters), a collision avoidance maneuver command is automatically triggered (yaw angle adjustment ≤ 5°), and the path display in the interactive layer is updated simultaneously.

[0075] In this embodiment of the invention, centimeter-level precision positioning of marathon emergency responders and injured personnel was achieved through a BeiDou dual-mode positioning terminal and a UAV multispectral gimbal, as well as real-time acquisition of multi-view video streams from accident scenes, providing comprehensive and accurate data support for rescue command. This improved the accuracy and efficiency of rescue operations, reducing delays caused by inaccurate or delayed information. The use of the IEEE 1588v2 precision clock protocol and multi-sensor fusion reconstruction algorithm enabled spatiotemporal synchronization and spatial registration of multi-source data, ensuring precise matching and consistency between data sources. This provided a reliable data foundation for subsequent intelligent analysis and decision-making, improving the scientific rigor and rationality of rescue command. Based on system dynamics and cellular automata models, intelligent allocation of medical resources and prediction of disaster spread were achieved, providing scientific decision support for rescue command. Resource allocation was optimized, rescue efficiency was improved, and disaster losses were reduced. The use of a von Lono diagram segmentation algorithm with obstacle constraints enabled the generation and replanning of UAV obstacle avoidance flight paths, ensuring the UAV's ability to avoid obstacles during complex events. Enhancing safe flight in complex environments; improving the efficiency and safety of drone rescue missions, reducing rescue interruptions caused by collisions and other accidents; achieving dynamic rendering and interaction of a three-dimensional visualization command interface through an electronically controlled circularly polarized 3D display terminal, providing an intuitive and convenient operating platform for rescue command; improving command efficiency and reducing command difficulty, enabling rescuers to make decisions more quickly and accurately; forming a command closed loop through real-time updates of BeiDou positioning data stream and mission status, ensuring continuous optimization and adjustment of rescue operations; improving the flexibility and adaptability of marathon rescue operations, enabling timely adjustments to rescue strategies based on actual conditions, and increasing the success rate of rescues.

[0076] The present invention also provides a 3D visualization medical rescue command device, which includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor performs the steps of the 3D visualization medical rescue command system in the above embodiments.

[0077] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the 3D visualization medical rescue command system.

[0078] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0079] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0080] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A 3D visualization medical rescue command system, characterized in that, The 3D visualization medical rescue command system includes: The acquisition module is used to acquire the latitude, longitude, and elevation data of marathon emergency personnel and injured people in real time, generate Beidou positioning data stream, collect multi-view video streams of the accident scene, generate UAV multispectral video streams, and generate a three-dimensional digital twin track model with centimeter-level accuracy through a 3D real-scene GPS ground-based augmentation station. The synchronization module is used to align the timestamps of the BeiDou positioning data stream, the UAV multispectral video stream, and the 3D digital twin track model to generate a spatiotemporal synchronization dataset, generate a coordinate system dataset based on the BeiDou positioning data stream, and generate a spatial registration dataset based on the UAV multispectral video stream and the 3D digital twin track model. The allocation module is used to solve a set of differential equations for the flow of medical resources based on a coordinate system dataset as boundary conditions, generate a real-time resource allocation scheme, define the fire spread rules for each grid cell using thermal imaging data in the spatial registration dataset as input parameters, generate a disaster spread prediction map, and generate an obstacle avoidance flight path for drones based on the building coordinates in the 3D digital twin track model. The path module is used to dynamically render the base layer, dynamic overlay layer and interactive layer in the electronically controlled circularly polarized 3D display terminal based on the spatiotemporal synchronization dataset, real-time resource allocation scheme, disaster spread prediction map and UAV obstacle avoidance flight path. When the positioning deviation of the wounded in the real-time resource allocation scheme exceeds the set threshold, the replanning of the UAV obstacle avoidance flight path is automatically triggered. The execution module is used to send the optimized path in the real-time resource allocation scheme to the Beidou dual-mode positioning terminal, control the drone swarm to perform the casualty search mission according to the drone obstacle avoidance flight path, and deliver medical supplies. In the electronically controlled circularly polarized 3D display terminal, the location and mission status of marathon emergency personnel in the Beidou positioning data stream are updated in real time, forming a command closed loop.

2. The 3D visualization medical rescue command system according to claim 1, characterized in that, include: The latitude, longitude, and elevation data of marathon emergency personnel and injured people are collected in real time through the Beidou dual-mode positioning terminal to eliminate multipath interference and generate Beidou positioning data stream. By using the infrared-visible dual-mode gimbal on a DJI drone, visible light video streams and thermal imaging video streams from the accident scene are collected. Based on hardware-level timestamp alignment technology, the dual-mode video streams are combined into a multispectral video stream from the drone. By scanning the track area with a 3D real-scene GPS ground-based augmentation station and using a multi-sensor fusion reconstruction algorithm, a three-dimensional digital twin track model is generated.

3. The 3D visualization medical rescue command system according to claim 1, characterized in that, include: Based on the Lie group SE transform algorithm, the geodetic coordinate system of the BeiDou positioning data stream is transformed into the local northeast-sky coordinate system of the three-dimensional digital twin track model; The SIFT feature point matching algorithm is used to extract thermal imaging feature points from the UAV multispectral video stream. The image coordinates of the thermal imaging feature points are then geometrically projected and matched with the spatial coordinates of the three-dimensional digital twin track model to generate a spatial registration dataset.

4. The 3D visualization medical rescue command system according to claim 1, characterized in that, include: Based on the system dynamics model, and using the coordinate system dataset as boundary conditions, a set of differential equations including resource supply, transportation speed and path resistance is constructed to generate a real-time resource allocation scheme. Based on the cellular automata model, using thermal imaging data in the spatial registration dataset as input parameters, fire spread rules are defined for each grid cell, with a temperature threshold ≥150℃ and a wind direction weighting factor of 0.3-0.7, to generate a disaster spread prediction map. Using a von Lonoie diagram segmentation algorithm with obstacle constraints, an obstacle avoidance flight path for the UAV is generated based on the building coordinates in the three-dimensional digital twin track model.

5. The 3D visualization medical rescue command system according to claim 1, characterized in that, include: Dynamically rendering the base layer, dynamic overlay layer, and interactive layer in an electronically controlled circularly polarized 3D display terminal: Base layer: Based on a coordinate system and a centralized 3D digital twin track model, the track terrain and building outlines are displayed with centimeter-level precision; Dynamic overlay: i) BeiDou positioning personnel marking: Based on the BeiDou positioning data stream in the spatiotemporal synchronized dataset, dots are used to mark the locations of marathon emergency personnel and injured people; ii) Resource flow vector field: Based on the real-time resource allocation scheme, the arrow length represents the transportation speed and the color gradient represents the drug allocation priority; iii) Fire risk heat map: Based on the disaster spread prediction map, the range of fire spread in the next 10 minutes is represented by a red gradient area; Interaction layer: i) UAV multispectral video stream pop-up: Embeds the real-time image of the UAV multispectral video stream into the corresponding geographic coordinates; ii) Countdown label for estimated arrival time: Based on the drone's obstacle avoidance flight path, the arrival time is displayed floating above the target point; The system monitors the positioning deviation of the wounded in the real-time resource allocation scheme in real time. When the deviation exceeds the threshold, it automatically sends a replanning command to the UAV obstacle avoidance flight path generation module. The replanned path is updated and displayed in real time through the countdown label in the interaction layer.

6. The 3D visualization medical rescue command system according to claim 1, characterized in that, include: The optimized path instruction in the real-time resource allocation scheme is sent to the Beidou dual-mode positioning terminal. The optimized path instruction includes the target coordinates, priority identifier and estimated arrival time. Control the DJI drone swarm to perform the mission according to the drone obstacle avoidance flight path: The main drone carries a medical supply bin and flies along the path to the coordinates of the wounded to perform the airdrop; the auxiliary drone simultaneously activates the infrared-visible light dual-mode gimbal and transmits the search image back to the three-dimensional visualization command interface in real time. In the three-dimensional visualization command interface, the BeiDou positioning data stream is analyzed in real time, the location markers of marathon emergency personnel are updated, and the mission status indicators are dynamically displayed. When the airdrop of supplies is detected to be completed, the target point is automatically marked as rescued. The status update data is synchronously fed back to the resource allocation model, triggering a new round of resource scheduling.

7. The 3D visualization medical rescue command system according to claim 6, characterized in that, Also includes: Real-time acquisition of centimeter-level 3D point cloud data of the area where the injured are located; real-time registration of the point cloud data with the 3D digital twin track model; generation of high-precision positioning correction instructions. When the main UAV hovers over the target point, it activates the image recognition mode of the visible light gimbal, performs sub-pixel-level positioning based on the preset medical markers, and controls the material warehouse release mechanism to automatically open the warehouse and deliver the materials when it is 5 meters above the ground. The system monitors the flight trajectories of the auxiliary drone group in real time, calculates the three-dimensional Euclidean distance between each drone based on the drone obstacle avoidance flight path, and automatically triggers a collision avoidance maneuver command when the distance between drones is detected to be less than the safety threshold, and updates the path display in the interactive layer simultaneously.

Images