A Real-Time Gaussian Splash Modeling System and Method Based on 5G Communication

Abstract

This invention relates to a real-time Gaussian splash modeling system and method based on 5G communication. The real-time Gaussian splash modeling system may include: a data acquisition terminal, a 5G transmission module, and a server. The data acquisition terminal is a lightweight terminal responsible only for acquiring raw scene data. The 5G transmission module is used for data transmission between the data acquisition terminal and the server. The server receives the raw scene data transmitted by the data acquisition terminal and completes all Gaussian splash modeling, real-time rendering, and result feedback operations. This invention adopts an innovative architecture of "5G transmission + data acquisition terminal only + centralized modeling on the server," which, through architectural optimization, solves the pain points of bulky data acquisition terminals and poor real-time performance in existing technologies.

Description

Technical Field

[0001] This invention belongs to the field of 3D modeling technology, specifically relating to a real-time Gaussian splash modeling method and system based on 5G communication, which is suitable for application scenarios such as 3D scene reconstruction, virtual simulation, and real-time visualization that require high real-time performance and lightweight acquisition terminals. Background Technology

[0002] Gaussian Splatting (GS) technology, as a highly efficient 3D scene reconstruction and rendering technique, has been widely applied in fields such as virtual scene construction, digital twins, and AR / VR due to its superior real-time rendering efficiency and image fidelity compared to traditional voxel modeling and point cloud modeling. Existing Gaussian splatting modeling techniques mainly fall into two categories: local modeling, where the acquisition device and modeling / rendering module are integrated into the same terminal, and the acquired data is directly used for Gaussian splatting modeling and rendering locally; and distributed modeling based on traditional networks (4G, wired networks), where the acquisition terminal and server transmit data through a common network, and the server completes the modeling and rendering.

[0003] However, both of the above implementation methods have obvious technical drawbacks: 1. Local modeling method: Gaussian splash modeling and real-time rendering require a lot of computing resources (CPU, GPU computing power), and the acquisition terminal needs to be configured with high-performance hardware, resulting in large size, high cost and high power consumption of acquisition devices, which cannot be used for portable acquisition scenarios (such as handheld acquisition, drone acquisition), and it is difficult to achieve collaborative modeling of multiple acquisition terminals; 2. Traditional network distributed modeling methods: 4G networks suffer from high latency (usually above 100ms) and unstable bandwidth, while wired networks are limited by cabling range and cannot achieve flexible deployment of mobile acquisition terminals; the long data transmission latency between the acquisition terminal and the server will lead to modeling and rendering delays, which cannot meet real-time requirements (such as real-time scene monitoring and dynamic scene reconstruction). Moreover, in existing technologies, most acquisition terminals still need to undertake some data preprocessing or preliminary modeling tasks, which fails to achieve a complete lightweighting of the acquisition terminal. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing Gaussian splash modeling technologies, such as bulky and costly acquisition terminals, poor real-time performance, and inflexible deployment. It provides a real-time Gaussian splash modeling method and system based on 5G. Leveraging the low latency and high bandwidth of 5G, the acquisition terminal is only responsible for acquiring the original scene data, while all modeling and rendering tasks are completed by the server. This achieves lightweight acquisition terminals, real-time modeling, and flexible deployment, while also expanding the application scenarios of Gaussian splash modeling. Therefore, the technical solution adopted by this invention is as follows: According to one aspect of the present invention, a real-time Gaussian splash modeling system based on 5G communication is provided, which may include: a data acquisition terminal, a 5G transmission module, and a server, wherein the data acquisition terminal is a lightweight terminal and is only responsible for acquiring raw scene data; the 5G transmission module is used for data transmission between the data acquisition terminal and the server; the server is used to receive the raw scene data transmitted by the data acquisition terminal and complete all Gaussian splash modeling, real-time rendering, and result feedback operations.

[0005] In one embodiment, the acquisition terminal includes a sensor unit and a 5G communication unit. The sensor unit is used to acquire raw scene data of the target scene, including one or more of image data, depth data, and raw point cloud data. The 5G communication unit is a built-in 5G module that is communicatively connected to the 5G transmission module.

[0006] In one embodiment, the sensor unit includes one or more of an RGB camera, a depth camera, and a LiDAR.

[0007] In one embodiment, the 5G transmission module supports simultaneous access by 10 to 50 of the acquisition terminals.

[0008] In one embodiment, the server includes: A data receiving unit, adapted to the 5G transmission module, is used to receive raw scene data transmitted by the acquisition terminal and perform data decryption and integrity verification. A data preprocessing unit performs standardization processing on the received raw scene data; A Gaussian splash modeling unit, which uses a Gaussian splash algorithm to complete real-time modeling of the target scene based on preprocessed standardized data, wherein the modeling frame rate is consistent with the acquisition frame rate of the acquisition terminal; A real-time rendering unit is used to render the modeled 3D scene in real time and generate a rendered image, wherein the rendering frame rate is ≥30 frames / second. The result feedback unit is used to feed back the modeling results and rendered images to the acquisition terminal or the designated visualization terminal in real time through the 5G transmission module, wherein the feedback latency is ≤5ms.

[0009] In one embodiment, the standardization process includes format conversion, noise reduction, and data alignment.

[0010] In one embodiment, the real-time rendering unit employs a rasterization rendering algorithm.

[0011] In one embodiment, the acquisition frame rate is 10 to 60 frames per second.

[0012] In one embodiment, the visualization terminal includes AR glasses and / or a monitoring screen.

[0013] According to another aspect of the present invention, a real-time Gaussian splash modeling method based on 5G communication is also provided, which may include the following steps: S1. Provide a real-time Gaussian splash modeling system based on 5G communication as described above; S2. Initialize the acquisition terminal: Start the acquisition terminal, set the acquisition frame rate and transmission parameters, and ensure that the acquisition terminal communicates normally with the 5G transmission module and the server. S3. Raw Scene Data Acquisition: The sensor unit acquires raw scene data of the target scene in real time. After acquisition, the data is directly transmitted to the server through the 5G communication unit and 5G transmission module without any preprocessing. S4. Server-side data processing: The server-side data receiving unit receives the original scene data, completes data decryption and integrity verification, and then the data preprocessing unit performs standardization processing on the original scene data. S5. Real-time modeling and rendering: The server-side Gaussian splash modeling unit completes the Gaussian splash modeling of the target scene in real time based on standardized data, and the real-time rendering unit renders the modeling results in real time to generate rendered images. S6. Result Feedback: The server uses the result feedback unit to transmit the modeling results and rendered images to the acquisition terminal or visualization terminal in real time via the 5G transmission module, thus completing a real-time modeling process. S7. Repeat steps S3-S6 to achieve continuous real-time Gaussian splash modeling of the target scene until the user stops data collection or the modeling task ends.

[0014] The beneficial effects of adopting the above technical solution in this invention are: 1. Completely lightweight acquisition terminal: The acquisition terminal is only responsible for acquiring raw scene data and does not participate in any preprocessing, modeling, or rendering operations. It does not require high-performance CPUs or GPUs, which greatly reduces the size, cost, and power consumption of the acquisition terminal. It can be deployed in a portable manner (such as handheld acquisition or drone acquisition) and supports collaborative modeling of multiple acquisition terminals. 2. Significantly improved modeling real-time performance: Leveraging the low latency (≤10ms) and high bandwidth advantages of 5G technology, the problem of high latency in traditional network transmission is solved. Combined with high-performance computing on the server side, the real-time performance of the entire process of acquisition, transmission, modeling, rendering, and feedback is achieved (total latency ≤20ms). The modeling frame rate is synchronized with the acquisition frame rate, meeting the needs of dynamic scene reconstruction, real-time monitoring, and other requirements. 3. Flexible deployment and strong scalability: 5G transmission requires no wiring, supports mobile deployment of data collection terminals, and can be adapted to various indoor and outdoor scenarios; it also supports concurrent access of multiple data collection terminals, and the number of data collection terminals can be increased or decreased according to scenario requirements, making it highly scalable; 4. Excellent modeling accuracy and visualization effects: The server adopts standardized data preprocessing and Gaussian splashing modeling algorithm, combined with real-time rasterization rendering, to ensure the geometric accuracy of the 3D model and the fidelity of the rendered image, adapting to high-requirement application scenarios (such as digital twins, AR / VR interaction). Attached Figure Description

[0015] To further illustrate the various embodiments, the present invention provides accompanying drawings. These drawings are part of the disclosure of the present invention, primarily used to illustrate the embodiments and to explain the operating principles of the embodiments in conjunction with the relevant descriptions in the specification. With reference to these drawings, those skilled in the art should be able to understand other possible implementations and the advantages of the present invention. Components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0016] Figure 1 This is an overall architecture diagram of a real-time Gaussian splash modeling system based on 5G communication according to an embodiment of the present invention; Figure 2 yes Figure 1 The diagram shows the data acquisition terminal structure of a real-time Gaussian splash modeling system based on 5G communication. Figure 3 yes Figure 1 The diagram shows the server-side architecture of a real-time Gaussian splash modeling system based on 5G communication. Figure 4 This is a flowchart of a real-time Gaussian splash modeling method based on 5G communication according to an embodiment of the present invention. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings to provide a clearer understanding of the purpose, features, and advantages of the present invention. It should be understood that the embodiments shown in the drawings are not intended to limit the scope of the present invention, but are merely illustrative of the essential spirit of the technical solution of the present invention.

[0018] In the following description, certain specific details are set forth for the purpose of illustrating various disclosed embodiments in order to provide a thorough understanding of the various disclosed embodiments. However, those skilled in the art will recognize that embodiments may be practiced without one or more of these specific details. In other instances, well-known apparatuses, structures, and techniques associated with this application may not have been shown or described in detail to avoid unnecessarily obscuring the description of the embodiments.

[0019] Unless the context requires otherwise, throughout the specification and claims, the word “comprising” and its variations, such as “including” and “having”, shall be understood to have an open, inclusive meaning, that is, to be interpreted as “including, but not limited to”.

[0020] Throughout this specification, references to "an embodiment" or "an embodiment" indicate that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Therefore, the appearance of "in an embodiment" or "an embodiment" in various places throughout the specification does not necessarily refer to the same embodiment. Furthermore, a particular feature, structure, or characteristic may be combined in any manner in one or more embodiments.

[0021] The singular forms “a” and “the” used in this specification and the appended claims include plural references unless otherwise expressly stated herein. It should be noted that the term “or” is generally used to mean “and / or” unless otherwise expressly stated herein.

[0022] In the following description, in order to clearly demonstrate the structure and working method of the present invention, a number of directional terms will be used. However, terms such as "front", "back", "left", "right", "outside", "inside", "outward", "inward", "up", and "down" should be understood as convenient terms and not as limiting terms.

[0023] Furthermore, terms such as "horizontal," "vertical," and "sag" do not imply that components must be absolutely horizontal or suspended, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0024] In the description of this application, it should also be noted that, unless otherwise expressly specified and limited, the terms "set up," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0025] like Figures 1-3 As shown, the real-time Gaussian splash modeling system based on 5G communication of the present invention may include three core parts: a data acquisition terminal 1, a 5G transmission module 2, and a server 3. Each part functions independently but works collaboratively, and the specific structure is as follows: Acquisition terminal 1 is a lightweight terminal, responsible only for acquiring raw scene data. It does not participate in any data preprocessing, modeling, or rendering operations, reducing the hardware configuration requirements of the acquisition terminal and achieving lightweight deployment. Acquisition terminal 1 includes at least one sensor unit 11 and a 5G communication unit 12 (adapted to the 5G transmission module 2). The sensor unit 11 is used to acquire raw scene data of the target scene, including but not limited to one or more of image data, depth data, and raw point cloud data. The sensor type can be selected according to the application scenario, such as an IMU attitude sensor, RGB camera, depth camera, or LiDAR. The acquisition frame rate can be adjusted according to real-time requirements, for example, 10~60 frames / second. The 5G communication unit 12 is a built-in 5G module, responsible only for transmitting the raw scene data acquired by the sensor unit 11 to the server in real time via the 5G transmission module 2, without performing any data compression, format conversion, or other preprocessing operations (to avoid consuming the computing power of the acquisition terminal). Encryption protocols (such as AES encryption) can be used during transmission to ensure data security.

[0026] The 5G transmission module 2 employs 5G private network or public network enhancement technology to establish a low-latency, high-bandwidth data transmission link between the data acquisition terminal and the server. Its core function is to transmit the raw scene data from the data acquisition terminal to the server without delay or loss, while also supporting concurrent transmission from multiple data acquisition terminals (supporting 10-50 data acquisition terminals to access simultaneously). Specific parameter requirements: end-to-end transmission latency ≤10ms, bandwidth ≥1Gbps, and packet loss rate ≤0.01%, ensuring the real-time performance and integrity of raw scene data transmission and providing data support for real-time modeling on the server side.

[0027] The 5G transmission module 2 may include a 5G base station, and the 5G communication units 12 and 31 of the acquisition terminal 1 and the server 3 are respectively connected to the 5G base station for communication. The 5G base station can be a public network or a private network 5G base station.

[0028] Server-side 3 is a high-performance computing terminal responsible for receiving raw scene data transmitted from the acquisition terminal, completing all Gaussian splash modeling, real-time rendering, and result feedback operations. It is the core computing unit of the entire system. Server-side 3 includes a 5G communication unit 31, a data receiving unit 32, a data preprocessing unit 32, a Gaussian splash modeling unit 33, a real-time rendering unit 34, and a result feedback unit 35. The functions of each unit are as follows: The 5G communication unit 31 of the server-side 3 can be a 5G board plugged into the server and communicates with the 5G transmission module 2. That is, the server-side 2 and the acquisition terminal 1 communicate with the 5G transmission module 2 through their respective built-in 5G modules (5G communication units 11 and 31), thereby realizing data transmission between them. It should be understood that the 5G communication unit 31 of the server-side 3 can also be integrated into the gateway, and the server and the gateway can be connected via a gigabit network.

[0029] The data receiving unit 32 receives the raw scene data transmitted by the acquisition terminal and performs data decryption and integrity verification. If data loss is detected, a retransmission request is sent to the acquisition terminal via the 5G transmission module to ensure the integrity of the raw scene data.

[0030] The data preprocessing unit 33 performs standardization processing on the received raw scene data, including format conversion (converting the raw format collected by the sensor into a format supported by Gaussian splash modeling), noise reduction (removing interference points and noise in the raw scene data), and data alignment (in the case of multiple acquisition terminals, realizing spatial alignment of data from each acquisition terminal). The preprocessing time is ≤5ms / frame, ensuring that it does not affect the overall real-time performance.

[0031] The Gaussian splash modeling unit 34 uses the Gaussian splash algorithm to complete the real-time modeling of the target scene based on the preprocessed standardized data. This includes Gaussian kernel initialization, Gaussian parameter optimization, and scene geometry reconstruction. The modeling frame rate is consistent with the acquisition frame rate of the acquisition terminal (10~60 frames / second) to ensure the real-time performance of the modeling. At the same time, it supports dynamic adjustment of modeling parameters (such as the number of Gaussian kernels and modeling accuracy) to adapt to the needs of different application scenarios.

[0032] The real-time rendering unit 35 is used to render the modeled 3D scene in real time. It adopts a rasterization rendering algorithm and combines scene lighting and texture information to generate high-fidelity real-time rendered images with a rendering frame rate of ≥30 frames / second to ensure visualization effects.

[0033] The result feedback unit 36 ​​is used to feed back the modeling results (3D model data) and rendered images to the acquisition terminal or the designated visualization terminal 4 (such as AR glasses or monitoring screen) in real time via the 5G transmission module, so that users can view and adjust them. The feedback latency is ≤5ms.

[0034] The data receiving unit 32, data preprocessing unit 33, Gaussian splash modeling unit 34, real-time rendering unit 35, and result feedback unit 36 ​​can be corresponding program modules installed on the server. They can be different software programs or different functional modules within the same software program. The structures of these units are well-known and will not be described in detail here.

[0035] As shown in Figure 4, a real-time Gaussian splash modeling method based on 5G communication according to the present invention may include the following steps: S1. Provide the real-time Gaussian splash modeling system based on 5G communication as described above. The specific structure of the system has been described above and will not be repeated here. S2. Initialize the acquisition terminal: Start the acquisition terminal, set the acquisition frame rate and transmission parameters, and ensure that the acquisition terminal communicates normally with the 5G transmission module and the server. S3. Raw Scene Data Acquisition: The sensor unit acquires raw scene data of the target scene in real time. After acquisition, the data is directly transmitted to the server through the 5G communication unit and 5G transmission module without any preprocessing. S4. Server-side data processing: The server-side data receiving unit receives the original scene data, completes data decryption and integrity verification, and then the data preprocessing unit performs standardization processing on the original scene data. S5. Real-time modeling and rendering: The server-side Gaussian splash modeling unit completes the Gaussian splash modeling of the target scene in real time based on standardized data, and the real-time rendering unit renders the modeling results in real time to generate rendered images. S6. Result Feedback: The server uses the result feedback unit to transmit the modeling results and rendered images to the acquisition terminal or visualization terminal in real time via the 5G transmission module, thus completing a real-time modeling process. S7. Repeat steps S3-S6 to achieve continuous real-time Gaussian splash modeling of the target scene until the user stops data collection or the modeling task ends.

[0036] The following two specific examples further illustrate the real-time Gaussian splash modeling method based on 5G communication of the present invention.

[0037] Example 1: Real-time modeling of indoor scenes using a single acquisition terminal This example is applicable to real-time 3D reconstruction of indoor scenes (such as conference rooms and exhibition halls). The specific implementation steps are as follows: 1. System Configuration: (1) Acquisition terminal: A handheld acquisition terminal is used, with a built-in RGB-D depth camera (acquisition image resolution 1920×1080, depth acquisition range 0.5-5m) and 5G module (supports 5G SA mode). Only a basic control unit is configured, and no high-performance computing module is configured. The weight is ≤200g and the power consumption is ≤10W. (2) 5G transmission module: adopts 5G private network, end-to-end transmission latency ≤5ms, bandwidth ≥2Gbps, transmission packet loss rate ≤0.005%, and supports stable transmission of a single acquisition terminal; (3) Server side: Configure high-performance CPU (Intel Xeon Platinum 8470C), GPU (NVIDIA A100), 64GB memory, 1TB storage, install data preprocessing software, Gaussian splash modeling software, and real-time rendering software. Preprocessing time ≤3ms / frame, modeling frame rate 30 frames / second, and rendering frame rate 60 frames / second.

[0038] 2. Modeling process: S1. Data Acquisition Terminal Initialization: Start the handheld data acquisition terminal, set the acquisition frame rate to 30 frames / second, adapt the 5G transmission parameters to the 5G private network, and ensure normal communication between the data acquisition terminal and the server. S2. Raw scene data acquisition: The acquisition personnel hold a handheld acquisition terminal and move around in the indoor scene to collect data. The RGB-D depth camera collects the image data and depth data of the scene in real time. After the acquisition is completed, it is directly transmitted to the server through the 5G module and 5G private network without any data processing. S3. Server-side data processing: After receiving the data, the server-side data receiving unit completes AES decryption and integrity verification. Then, the data preprocessing unit converts the image data into PNG format and the depth data into PLY format, while performing noise reduction processing (removing interference points in the depth data). The preprocessing time is 2ms / frame. S4. Real-time modeling and rendering: The Gaussian splash modeling unit initializes 10,000 Gaussian kernels based on preprocessed image and depth data, performs Gaussian parameter optimization and scene geometry reconstruction, and the modeling frame rate is 30 frames / second; the real-time rendering unit combines indoor lighting parameters to perform rasterization rendering and generate a high-fidelity rendering image at 60 frames / second. S5. Result Feedback: The server will transmit the reconstructed indoor 3D model data and rendered images to the display screen of the acquisition terminal via a 5G private network. The acquisition personnel can view the modeling effect in real time and adjust the acquisition angle and position as needed. S6. Repeat steps S2-S5 for 10 minutes to complete the real-time 3D reconstruction of the entire indoor scene, and finally generate a complete indoor 3D model and real-time rendered video.

[0039] Example 2: Real-time modeling of outdoor scenes with multiple acquisition terminals This example is applicable to real-time 3D reconstruction of outdoor scenes (such as parks and squares), using three acquisition terminals working collaboratively. The specific implementation steps are as follows: 1. System Configuration: (1) Data acquisition terminal: 3 drone data acquisition terminals, each terminal is equipped with a built-in lidar (point cloud acquisition frame rate of 10 frames / second, point cloud density of 1000 points / frame) and 5G module, lightweight design, and can last for a long time; (2) 5G transmission module: adopts 5G public network enhancement technology, with end-to-end transmission latency ≤8ms, bandwidth ≥1.5Gbps, supports concurrent transmission of 3 acquisition terminals, and transmission packet loss rate ≤0.01%; (3) Server side: Configured with dual GPUs (NVIDIA A100), CPU Intel Xeon Platinum 8470C, memory 128GB, supports multi-source data alignment and parallel modeling, preprocessing time ≤4ms / frame, modeling frame rate 10 frames / second, rendering frame rate 30 frames / second.

[0040] 2. Modeling process: S1. Data Acquisition Terminal Initialization: Start 3 drone data acquisition terminals, set the acquisition frame rate to 10 frames / second, plan the acquisition path of each drone (avoid excessive acquisition overlap), and ensure that the 3 data acquisition terminals communicate normally with the 5G transmission module and the server. S2. Raw scene data acquisition: Three drones take off simultaneously and collect raw point cloud data of the outdoor scene according to the planned path. After the data is collected, it is directly transmitted to the server through the 5G module and the 5G public network without any preprocessing. S3. Server-side data processing: The data receiving unit receives point cloud data from three acquisition terminals, completes decryption and integrity verification, and the data preprocessing unit performs format conversion (converts to XYZ format) and noise reduction on the three sets of point cloud data. At the same time, it performs data alignment (based on spatial coordinates to achieve precise alignment of the three sets of data). The preprocessing time is 4ms / frame. S4. Real-time modeling and rendering: The Gaussian splash modeling unit, based on the aligned point cloud data, adopts a parallel modeling approach, initializes 15,000 Gaussian kernels, and completes real-time modeling of the outdoor scene at a modeling frame rate of 10 frames / second; the real-time rendering unit combines outdoor lighting and shadow effects to perform real-time rendering, generating a rendered image at 30 frames / second. S5. Results Feedback: The server will send the reconstructed outdoor 3D model and rendered images to the ground visualization terminal (monitoring screen), allowing staff to view the modeling progress and effects in real time and adjust the drone's data collection path. S6. Repeat steps S2-S5 for 30 minutes to complete the real-time 3D reconstruction of the outdoor scene and generate a complete 3D model of the outdoor scene, which supports multi-angle viewing and detail magnification.

[0041] Compared with the prior art, the present invention has the following advantages: 1. Completely lightweight acquisition terminal: The acquisition terminal is only responsible for acquiring raw scene data and does not participate in any preprocessing, modeling, or rendering operations. It does not require high-performance CPUs or GPUs, which greatly reduces the size, cost, and power consumption of the acquisition terminal. It can be deployed in a portable manner (such as handheld acquisition or drone acquisition) and supports collaborative modeling of multiple acquisition terminals. 2. Significantly improved modeling real-time performance: Leveraging the low latency (≤10ms) and high bandwidth advantages of 5G technology, the problem of high latency in traditional network transmission is solved. Combined with high-performance computing on the server side, the real-time performance of the entire process of acquisition, transmission, modeling, rendering, and feedback is achieved (total latency ≤20ms). The modeling frame rate is synchronized with the acquisition frame rate, meeting the needs of dynamic scene reconstruction, real-time monitoring, and other requirements. 3. Flexible deployment and strong scalability: 5G transmission requires no wiring, supports mobile deployment of data collection terminals, and can be adapted to various indoor and outdoor scenarios; it also supports concurrent access of multiple data collection terminals, and the number of data collection terminals can be increased or decreased according to scenario requirements, making it highly scalable; 4. Excellent modeling accuracy and visualization effects: The server-side adopts standardized data preprocessing and Gaussian splashing modeling algorithm, combined with real-time rasterization rendering, to ensure the geometric accuracy of the 3D model and the fidelity of the rendered image, adapting to high-requirement application scenarios (such as digital twins, AR / VR interaction). 5. This invention adopts an innovative architecture of "5G transmission + data acquisition terminal only acquisition + centralized modeling on the server side". Through architecture optimization, it solves the pain points of bulky acquisition terminals and poor real-time performance in existing technologies.

[0042] The preferred embodiments of the present invention have been described in detail above. However, it should be understood that after reading the above teachings, those skilled in the art can make various alterations or modifications to the present invention. These equivalent forms also fall within the scope defined by the appended claims.

Claims

1. A real-time Gaussian splash modeling system based on 5G communication, characterized in that, include: The system includes a data acquisition terminal, a 5G transmission module, and a server. The data acquisition terminal is a lightweight terminal that is only responsible for acquiring raw scene data. The 5G transmission module is used for data transmission between the acquisition terminal and the server; the server is used to receive the raw scene data transmitted by the acquisition terminal and complete all Gaussian splash modeling, real-time rendering and result feedback operations.

2. The real-time Gaussian splash modeling system based on 5G communication as described in claim 1, characterized in that, The acquisition terminal includes a sensor unit and a 5G communication unit. The sensor unit is used to acquire raw scene data of the target scene, including one or more of image data, depth data and raw point cloud data. The 5G communication unit is a built-in 5G module that communicates with the 5G transmission module.

3. The real-time Gaussian splash modeling system based on 5G communication as described in claim 2, characterized in that, The sensor unit includes one or more of an RGB camera, a depth camera, and a LiDAR.

4. The real-time Gaussian splash modeling system based on 5G communication as described in claim 1, characterized in that, The 5G transmission module supports simultaneous access for 10 to 50 of the aforementioned acquisition terminals.

5. The real-time Gaussian splash modeling system based on 5G communication as described in claim 1, characterized in that, The server-side includes: A data receiving unit, adapted to the 5G transmission module, is used to receive raw scene data transmitted by the acquisition terminal and perform data decryption and integrity verification. A data preprocessing unit performs standardization processing on the received raw scene data; A Gaussian splash modeling unit, which uses a Gaussian splash algorithm to complete real-time modeling of the target scene based on preprocessed standardized data, wherein the modeling frame rate is consistent with the acquisition frame rate of the acquisition terminal; A real-time rendering unit is used to render the modeled 3D scene in real time and generate a rendered image, wherein the rendering frame rate is ≥30 frames / second. The result feedback unit is used to feed back the modeling results and rendered images to the acquisition terminal or the designated visualization terminal in real time through the 5G transmission module, wherein the feedback latency is ≤5ms.

6. The real-time Gaussian splash modeling system based on 5G communication as described in claim 5, characterized in that, The standardization process includes format conversion, noise reduction, and data alignment.

7. The real-time Gaussian splash modeling system based on 5G communication as described in claim 5, characterized in that, The real-time rendering unit uses a rasterization rendering algorithm.

8. The real-time Gaussian splash modeling system based on 5G communication as described in claim 5, characterized in that, The acquisition frame rate is 10~60 frames / second.

9. The real-time Gaussian splash modeling system based on 5G communication as described in claim 5, characterized in that, The visualization terminal includes AR glasses and / or a monitoring screen.

10. A real-time Gaussian splash modeling method based on 5G communication, characterized in that, Includes the following steps: S1. Provide a real-time Gaussian splash modeling system based on 5G communication as described in any one of claims 1 to 9; S2. Initialize the acquisition terminal: Start the acquisition terminal, set the acquisition frame rate and transmission parameters, and ensure that the acquisition terminal communicates normally with the 5G transmission module and the server. S3. Raw Scene Data Acquisition: The sensor unit acquires raw scene data of the target scene in real time. After acquisition, the data is directly transmitted to the server through the 5G communication unit and 5G transmission module without any preprocessing. S4. Server-side data processing: The server-side data receiving unit receives the original scene data, completes data decryption and integrity verification, and then the data preprocessing unit performs standardization processing on the original scene data. S5. Real-time modeling and rendering: The server-side Gaussian splash modeling unit completes the Gaussian splash modeling of the target scene in real time based on standardized data, and the real-time rendering unit renders the modeling results in real time to generate rendered images. S6. Result Feedback: The server uses the result feedback unit to transmit the modeling results and rendered images to the acquisition terminal or visualization terminal in real time via the 5G transmission module, thus completing a real-time modeling process. S7. Repeat steps S3-S6 to achieve continuous real-time Gaussian splash modeling of the target scene until the user stops data collection or the modeling task ends.

Images