Unmanned aerial vehicle flow field rendering method based on asynchronous rasterization and multi-band fusion

Abstract

Embodiments of the present disclosure disclose a UAV flow field rendering method based on asynchronous rasterization and multi-band fusion. A specific embodiment of the method comprises: collecting, by a UAV, thermal flow data of an airport to be tested to obtain a UAV flow field data set; performing spatiotemporal blocking on the UAV flow field data set to obtain a UAV flow field data block set; encapsulating each UAV flow field data block to obtain an encapsulated flow field data block set; performing asynchronous loading on the encapsulated flow field data block set according to window range information of an interactive terminal to obtain a decoded flow field data block set; performing rasterization processing on each decoded flow field data block to obtain a rasterized flow field data block set; performing multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point to obtain a rendered flow field view frame sequence; and performing time interpolation on the rendered flow field view frame sequence to obtain a smoothed flow field video. The embodiment can improve the efficiency of flow field data rendering.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of flow field visualization technology, specifically to a UAV flow field rendering method based on asynchronous rasterization and multi-band fusion. Background Technology

[0002] Unmanned aerial vehicle (UAV) flow field visualization is a technique that transforms abstract flow field data into intuitive graphics or animations. Its core purpose is to help analysts quickly identify the distribution patterns, anomalies, and dynamic trends of the flow field. In low-altitude heat flow monitoring at airports, UAV-based 3D flow field rendering has become a key research direction. Currently, when rendering flow field data collected by UAVs, the common approach is to load the complete mesh data into the video memory at once, and then use the video memory for flow field rendering.

[0003] However, when using the above methods, a common technical problem is that the technology based on 3D volume rendering needs to process massive amounts of 3D volume data. Loading the 3D data into memory or video memory at once can easily lead to memory overflow and excessive video memory usage, resulting in rendering interruption or crash. At the same time, the 3D volume data processing process takes a long time, thereby reducing visualization efficiency and making it difficult to adapt to the lightweight and real-time requirements of UAV field operations. Summary of the Invention

[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] Some embodiments of this disclosure propose a UAV flow field rendering method based on asynchronous rasterization and multi-band fusion to solve the technical problems mentioned in the background section above.

[0006] In a first aspect, some embodiments of this disclosure provide a UAV flow field rendering method based on asynchronous rasterization and multi-band fusion. The method includes: collecting thermal flow data from an airport to be tested using a UAV to obtain a UAV flow field dataset; performing spatiotemporal segmentation on the UAV flow field dataset to obtain a UAV flow field data block set; encapsulating each UAV flow field data block in the UAV flow field data block set to generate encapsulated flow field data blocks, resulting in an encapsulated flow field data block set; asynchronously loading the encapsulated flow field data block set according to the viewport information of the interactive terminal to obtain a decoded flow field data block set; rasterizing each decoded flow field data block in the decoded flow field data block set to generate rasterized flow field data blocks, resulting in a rasterized flow field data block set; performing multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set to generate a rendered flow field view frame, resulting in a rendered flow field view frame sequence; and performing temporal interpolation on the rendered flow field view frame sequence to obtain a smoothed flow field video.

[0007] Secondly, some embodiments of this disclosure provide a UAV flow field rendering device based on asynchronous rasterization and multi-band fusion. The device includes: a heat flow data acquisition unit configured to acquire heat flow data from an airport to be tested using a UAV, thereby obtaining a UAV flow field dataset; a spatiotemporal segmentation unit configured to perform spatiotemporal segmentation on the aforementioned UAV flow field dataset, thereby obtaining a UAV flow field data block set; an encapsulation unit configured to encapsulate each UAV flow field data block in the aforementioned UAV flow field data block set to generate encapsulated flow field data blocks, thereby obtaining an encapsulated flow field data block set; and an asynchronous loading unit configured to load according to the viewport range of the interactive terminal. The system asynchronously loads the aforementioned encapsulated flow field data block set to obtain a decoded flow field data block set; a rasterization processing unit is configured to rasterize each decoded flow field data block in the aforementioned decoded flow field data block set to generate rasterized flow field data blocks, thus obtaining a rasterized flow field data block set; a multi-band fusion rendering unit is configured to perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the aforementioned rasterized flow field data block set to generate rendered flow field view frames, thus obtaining a rendered flow field view frame sequence; and a time interpolation unit is configured to perform time interpolation on the rendered flow field view frame sequence to obtain a smoothed flow field video.

[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0010] The various embodiments of this disclosure have the following beneficial effects: the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion in some embodiments of this disclosure has high visualization efficiency. Specifically, the reason for the low flow field rendering efficiency is that: the technology based on 3D volume rendering needs to process massive amounts of 3D volume data. Loading the 3D data into memory or video memory at once can easily lead to memory overflow, excessive video memory usage, rendering interruption or crash. At the same time, the 3D volume data processing process is time-consuming, thereby reducing visualization efficiency and making it difficult to adapt to the lightweight and real-time requirements of UAV field operations. Based on this, the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion in some embodiments of this disclosure firstly collects thermal flow data from the airport to be tested using a UAV, obtaining a UAV flow field dataset. Then, the above UAV flow field dataset is spatiotemporally divided into blocks to obtain a UAV flow field data block set. Thus, the global data can be divided into voxel blocks and time-time slices, which facilitates multi-threaded and distributed processing and improves rendering efficiency. Secondly, each UAV flow field data block in the aforementioned UAV flow field data block set is encapsulated to generate encapsulated flow field data blocks, resulting in an encapsulated flow field data block set. This allows for overall data compression and encapsulation, reducing storage and transmission overhead and providing feasibility for edge node caching and low-bandwidth environments. Next, based on the viewport information of the interactive terminal, the encapsulated flow field data block set is asynchronously loaded to obtain a decoded flow field data block set. This avoids the rendering computation burden caused by full loading, preventing rendering latency and stuttering. Then, each decoded flow field data block in the aforementioned decoded flow field data block set is rasterized to generate rasterized flow field data blocks, resulting in a rasterized flow field data block set. This converts the raw data into rasterized data that can be processed in parallel by the GPU, facilitating GPU shader processing. Finally, multi-band fusion rendering is performed on the rasterized flow field data blocks corresponding to the same time point in the aforementioned rasterized flow field data block set to generate rendered flow field view frames, resulting in a rendered flow field view frame sequence. Therefore, rendering multiple parameters in a single pass can improve rendering quality and performance while meeting real-time interaction requirements. Finally, time interpolation is performed on the rendered flow field view frame sequence to obtain a smooth flow field video. This allows for low-frequency sampling, saving storage and transmission costs while ensuring the continuity of the rendered flow field video. This enables high frame rate output without increasing the drone's load, making it suitable for low-power, low-bandwidth outdoor operation scenarios. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a schematic diagram of an application scenario of the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion, which is one of the embodiments of this disclosure;

[0013] Figure 2 This is a flowchart of some embodiments of the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion according to the present disclosure;

[0014] Figure 3 This is a rendering flowchart of the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion according to this disclosure;

[0015] Figure 4 This is a schematic diagram of the structure of some embodiments of the UAV flow field rendering device based on asynchronous rasterization and multi-band fusion according to the present disclosure;

[0016] Figure 5 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0017] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0018] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0019] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0020] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0021] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0022] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] Figure 1 This is a schematic diagram of an application scenario of the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion, which is one of the embodiments of this disclosure.

[0024] exist Figure 1 In the application scenario, firstly, the computing device 101 can receive the drone flow field dataset collected by the drone 102. Then, the computing device 101 can perform spatiotemporal segmentation on the aforementioned drone flow field dataset to obtain a drone flow field data block set. Next, each drone flow field data block in the aforementioned drone flow field data block set is encapsulated to generate encapsulated flow field data blocks, resulting in an encapsulated flow field data block set. When receiving window range information from the interactive terminal 103, the computing device 101 can asynchronously load the encapsulated flow field data block set through a preset decoding thread to obtain a decoded flow field data block set. The interactive terminal can be used to display the rendered flow field data and interact with the user. The decoded flow field data block set can be stored in the memory 104. The interactive terminal 103, the memory 104, and the computing device 101 can communicate with each other. Next, the computing device 101 can perform rasterization processing on each decoded flow field data block in the aforementioned decoded flow field data block set through a preset computing thread to generate rasterized flow field data blocks, thus obtaining a rasterized flow field data block set. Then, the computing device 101 can perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the aforementioned rasterized flow field data block set through a preset rendering thread to generate rendered flow field view frames, thus obtaining a rendered flow field view frame sequence. Finally, the computing device 101 can perform time interpolation on the rendered flow field view frame sequence to obtain a smoothed flow field video. The computing device 101 can then transmit the smoothed flow field video to the interactive terminal 103 for display.

[0025] It should be noted that the aforementioned computing device 101 can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed in the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here. It should be understood that... Figure 1The number of computing devices in the system can be arbitrary, depending on the implementation requirements.

[0026] Continue to refer to Figure 2 The diagram illustrates a flowchart 200 of some embodiments of a UAV flow field rendering method based on asynchronous rasterization and multi-band fusion according to this disclosure. This UAV flow field rendering method based on asynchronous rasterization and multi-band fusion includes the following steps:

[0027] Step 201: Use a drone to collect thermal flow data at the airport to be tested, and obtain a drone flow field dataset.

[0028] In some embodiments, the execution entity of the UAV flow field rendering method based on asynchronous rasterization and multi-band fusion (e.g.) Figure 1 The computing device 101 shown can collect thermal flow data from an airport to be tested using a UAV, obtaining a UAV flow field dataset. The UAV can establish a wireless connection with the execution entity. The airport to be tested can be any airport requiring thermal flow visualization. Each UAV flow field data point in the dataset can include longitude, latitude, altitude, time, and parameters. These parameters can include eastward velocity components, northward velocity components, vertical velocity components, and temperature. The eastward velocity component can be the east-west velocity component of the wind speed vector in the low-altitude region. The northward velocity component can be the north-south velocity component of the wind speed vector. The vertical velocity component can be the vertical velocity component of the wind speed vector. For example, the eastward velocity component can be +5 m / s, a positive value indicating that the wind is blowing from the west and moving eastward; 5 m / s represents the wind speed in the eastward direction.

[0029] In practice, the aforementioned UAV can collect thermal flow data in the low-altitude region of the airport, obtaining a UAV flow field dataset. This low-altitude region can be the area within a 0-15m height range of the airport. Subsequently, the UAV can wirelessly transmit the UAV flow field dataset to the executing entity.

[0030] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra wideband) connections, and other currently known or future wireless connection methods.

[0031] Step 202: Spatiotemporally divide the above UAV flow field dataset into blocks to obtain a UAV flow field data block set.

[0032] In some embodiments, the aforementioned execution entity can perform spatiotemporal partitioning on the aforementioned UAV flow field dataset to obtain a UAV flow field data block set. Each UAV flow field data block in the aforementioned UAV flow field data block set may include the flow field data of each UAV within a single voxel within a single time period. In practice, the aforementioned UAV flow field dataset can be spatiotemporally partitioned based on a preset spatio-temporal cube chunking technique to obtain the UAV flow field data block set.

[0033] Optionally, the aforementioned execution entity may perform spatiotemporal segmentation of the aforementioned UAV flow field dataset to obtain a UAV flow field data block set, which may include the following steps:

[0034] The first step is to divide the three-dimensional space corresponding to the airport where the heat flow to be tested is located into multiple voxels, obtaining a voxel set. This three-dimensional space can be the low-altitude region of the airport. For example, it can be an airspace 15m above the airport ground level. Each voxel in the voxel set can be a cubic structure. In practice, the three-dimensional space can be uniformly divided into multiple voxels to obtain the voxel set.

[0035] The second step involves combining the flow field data of each drone corresponding to the same voxel in the aforementioned drone flow field dataset into a voxel drone flow field data set. Each voxel drone flow field data set can include flow field data from all drones at all time points within the same voxel. In practice, firstly, for each drone flow field data in the aforementioned drone flow field dataset, the corresponding voxel can be determined based on the longitude, latitude, and altitude included in the flow field data. Then, the combination of flow field data within the same voxel can be defined as a voxel drone flow field data set.

[0036] The third step involves combining the flow field data of each voxel drone within the aforementioned voxel drone flow field data set, corresponding to the same time period, into a single-frame voxel drone flow field data set. Each single-frame voxel drone flow field data set can include flow field data from various drones within the same voxel during the same time period.

[0037] The fourth step is to determine each single-frame voxel UAV flow field data group in the obtained single-frame voxel UAV flow field data group set as a UAV flow field data block, thus obtaining a UAV flow field data block set.

[0038] In practice, the aforementioned optional steps systematically slice the UAV flow field dataset by establishing a five-dimensional data model (longitude, latitude, altitude, time, and parameters). This allows for the location and loading of corresponding data based on specific spatiotemporal contexts, avoiding full dataset scanning and improving data transmission performance. Secondly, the segmented UAV flow field data blocks facilitate distributed computing and parallel processing. Furthermore, the segmented data blocks can be compressed more effectively, simplifying storage and transmission.

[0039] Step 203: Encapsulate each UAV flow field data block in the above UAV flow field data block set to generate encapsulated flow field data blocks, thus obtaining an encapsulated flow field data block set.

[0040] In some embodiments, the execution entity may encapsulate each UAV flow field data block in the aforementioned UAV flow field data block set to generate encapsulated flow field data blocks, thus obtaining an encapsulated flow field data block set. Each UAV flow field data block may include a longitude range, latitude range, altitude range, time period, and parameter set. The parameter set may include a velocity component dataset and a temperature dataset. The velocity component dataset may include the eastward velocity component, northward velocity component, and vertical velocity component corresponding to each UAV flow field data point within the same voxel in the same time period. The temperature dataset may include the temperature corresponding to each UAV flow field data point within the same voxel in the same time period. As an example, the UAV flow field data block can be: "Longitude range: [117.5000°E, 117.5005°E], Latitude range: [30.2000°N, 30.2005°N], Altitude range: [10m, 15m], Time period: 2024-05-01T14:00:00Z to 2024-05-01T14:01:00Z, Parameter set: (2024-05-01T14:00:43Z, 2.3m / s, 1.1m / s, 0.5m / s, 35.2℃) ... (2024-05-01T14:00:51Z, 2.4m / s, 1.2m / s, 0.7m / s, 34.2℃)". Each parameter in the above parameter set includes the eastward velocity component, the northward velocity component, the vertical velocity component, and temperature.

[0041] In practice, each UAV flow field data block in the aforementioned UAV flow field data block set can be encapsulated using a pre-defined encapsulation technique to generate encapsulated flow field data blocks, thus obtaining an encapsulated flow field data block set. This encapsulation technique can be a JSON compression tool.

[0042] Optionally, the aforementioned execution entity encapsulates each UAV flow field data block in the aforementioned UAV flow field data block set to generate encapsulated flow field data blocks, thereby obtaining an encapsulated flow field data block set. This may include the following steps:

[0043] The first step involves predictively encoding each velocity component data included in the aforementioned UAV flow field data block to generate coded velocity component data, resulting in a coded velocity component dataset. Each coded velocity component in this dataset can include an coded eastward velocity component, a coded northward velocity component, and a coded vertical velocity component. In practice, a pre-defined predictive encoding algorithm can be used to predictively encode the eastward velocity component in each velocity component data included in the aforementioned UAV flow field data block to generate a coded eastward velocity component set. This predictive encoding algorithm can be a differential coding algorithm or a Huffman coding algorithm. Secondly, the aforementioned predictive encoding algorithm can be used to predictively encode the northward velocity component and the vertical velocity component in each velocity component data included in the aforementioned UAV flow field data block, respectively, to generate coded northward velocity components and coded vertical velocity components, resulting in coded northward velocity component sets and coded vertical velocity component sets.

[0044] The second step involves performing wavelet transform compression on each temperature data point included in the aforementioned UAV flow field data block to generate compressed temperature data, thus obtaining a compressed temperature dataset. In practice, a preset wavelet transform algorithm can be used to perform wavelet transform compression on each temperature data point included in the aforementioned UAV flow field data block to generate compressed temperature data, thus obtaining a compressed temperature dataset. The aforementioned wavelet transform algorithm can be the Daubechies-4 algorithm.

[0045] The third step involves updating the UAV flow field data block based on the aforementioned coded velocity component dataset and compressed temperature dataset, resulting in an updated UAV flow field data block, which serves as the compressed flow field data block. In practice, the coded velocity component dataset and compressed temperature dataset can be used to replace the velocity component data and temperature data in the parameter set of the UAV flow field data block to obtain the updated UAV flow field data block. This updated UAV flow field data block can then be designated as the compressed flow field data block.

[0046] The fourth step is to encapsulate the compressed flow field data block to generate an encapsulated flow field data block. In practice, the above encapsulation technique can be used to encapsulate the compressed flow field data block to generate an encapsulated flow field data block.

[0047] Step 204: Based on the window range information of the interactive terminal, the above-mentioned encapsulated flow field data block set is asynchronously loaded to obtain the decoded flow field data block set.

[0048] In some embodiments, the aforementioned execution entity can asynchronously load the encapsulated flow field data block set according to the viewport range information of the interactive terminal to obtain a decoded flow field data block set. The interactive terminal corresponds to a viewport. The viewport can be used to display the rendered UAV flow field video data. The viewport can correspond to viewpoint information, viewport range information, and viewport frame rate. The viewpoint information can characterize the observer's position and viewing direction relative to the rendered flow field's three-dimensional space. The viewpoint information can include viewpoint coordinates and viewpoint direction. The viewport range information can be the range of the visualized flow field viewed by the user through the interactive terminal. The viewport frame rate can be the rendering frame rate of the visualized flow field viewed by the user through the interactive terminal. In practice, the aforementioned execution entity can determine each encapsulated flow field data block in the encapsulated flow field data block set corresponding to the aforementioned viewport range information as a target flow field data block, obtaining a target flow field data block set. Then, a decoding thread, based on a preset decoding tool, can decode each target flow field data block in the target flow field data block set to obtain a decoded flow field data block set. The aforementioned decoding thread can be a thread used to perform data decoding tasks. This decoding thread can run independently of other threads, or in parallel with other threads. The aforementioned decoding tool can be JSON.parse(). Simultaneously, each encapsulated flow field data block in the encapsulated flow field data block set that does not correspond to the aforementioned window range information can be stored at an edge node. This edge node can be a cache server deployed at the network edge, or a CDN edge computing node.

[0049] Optionally, the execution entity asynchronously loads the encapsulated flow field data block set based on the window range information of the interactive terminal to obtain the decoded flow field data block set, which may include the following steps:

[0050] The first step involves, in response to receiving window range information from the interactive terminal, identifying each encapsulated flow field data block in the encapsulated flow field data block set corresponding to the aforementioned window range information as a target flow field data block, thus obtaining a target flow field data block set. The aforementioned window range information may include the window longitude range, window latitude range, window height range, and window time period. In practice, the interactive terminal can send the window range information to the execution entity in real time. When the execution entity receives the window range information from the interactive terminal, it can determine the target flow field data block set corresponding to the aforementioned window range information within the encapsulated flow field data blocks based on the window longitude range, window latitude range, window height range, and window time period included in the window range information.

[0051] The second step involves identifying each packaged flow field data block in the aforementioned packaged flow field data block set that is temporally adjacent to each target flow field data block in the aforementioned target flow field data block set as an adjacent temporal flow field data block, thus obtaining a set of adjacent temporal flow field data blocks. In practice, each packaged flow field data block that is one time unit before and after each of the aforementioned target flow field data blocks can be queried as an adjacent temporal flow field data block to obtain a set of adjacent temporal flow field data blocks.

[0052] The third step involves using a decoding thread to obtain the target flow field data block set and the adjacent temporal flow field data block set from edge nodes or local nodes. Each target flow field data block and each adjacent temporal flow field data block is then decoded to generate a decoded flow field data block set, which is then stored in a video memory node. This video memory node can be a GPU device connected to the execution entity. In practice, the execution entity first enables the decoding thread to obtain the target flow field data block set and the adjacent temporal flow field data block set from the edge nodes. Then, the decoding thread uses a decapsulation technique corresponding to the encapsulation technique to decode each target flow field data block in the target flow field data block set and each adjacent temporal flow field data block in the adjacent temporal flow field data block set, obtaining the decoded flow field data block set. Finally, the decoding thread stores the decoded flow field data block set in the video memory node.

[0053] In practice, to improve the efficiency of flow field rendering, the raw UAV flow field data can be divided into blocks and compressed. After determining the target data block to be rendered, the target data block and adjacent temporal data blocks are written into the video memory node for subsequent rasterization and rendering stages. This improves the efficiency and real-time performance of UAV flow field rendering.

[0054] Optionally, the aforementioned implementing entity may also perform the following steps:

[0055] The first step is to determine the set of prefetched flow field data blocks corresponding to the aforementioned window range information, and to store this set of prefetched flow field data blocks to the edge node. This set of prefetched flow field data blocks can be each encapsulated flow field data block corresponding to the window range after spatial translation, spatial scaling, or temporal translation of the aforementioned window range information. In practice, the executing entity can translate, scale, or enlarge the window longitude range, window latitude range, and window height range corresponding to the aforementioned window range information to obtain the prefetched longitude range, prefetched latitude range, and prefetched height range. The time interval between two consecutive time units corresponding to the window time interval of the aforementioned window range information is determined as the prefetched time interval. Next, each encapsulated flow field data block in the encapsulated flow field data block set corresponding to the aforementioned prefetched longitude range, prefetched latitude range, prefetched height range, and the aforementioned prefetched time interval can be determined as a prefetched flow field data block, resulting in the prefetched flow field data block set. Finally, the prefetched flow field data block set can be stored to the aforementioned edge node.

[0056] The second step involves identifying each encapsulated flow field data block in the aforementioned encapsulated flow field data block set that meets the preset historical viewing conditions as a historical viewing flow field data block, thus obtaining a historical viewing flow field data block set. This historical viewing data block set is then stored on a local node. The aforementioned historical viewing conditions can be the target flow field data blocks corresponding to various window ranges within a historical time period. The aforementioned historical time period can be the previous time unit. The aforementioned local node can be local storage (such as an SSD or memory cache) connected to the aforementioned execution entity. In practice, the aforementioned execution entity can identify each data block determined as a target flow field data block in the previous time unit as the historical viewing flow field data block set and store this set on a local node, allowing users to quickly review historical flow field scenarios.

[0057] Step 205: Rasterize each decoded flow field data block in the above decoded flow field data block set to generate rasterized flow field data blocks, thus obtaining a rasterized flow field data block set.

[0058] In some embodiments, the execution entity may perform rasterization processing on each decoded flow field data block in the decoded flow field data block set to generate rasterized flow field data blocks, thus obtaining a rasterized flow field data block set. Each rasterized flow field data block in the rasterized flow field data block set may include raster flow velocity information and color information. The raster flow velocity information may include horizontal direction angle, average vertical elevation angle, and average flow velocity magnitude.

[0059] In practice, for each decoded flow field data block in the aforementioned decoded flow field data block set, the average value among the various velocity components included in the decoded flow field data block can be determined as the raster flow velocity information. Secondly, the average value among the various temperature data included in the aforementioned decoded flow field data block can be determined as the average temperature. Then, based on a preset color texture mapping table, the color information corresponding to the average temperature is determined. The aforementioned color texture mapping table can be a predefined temperature-color correspondence table. For example, the aforementioned color texture mapping table may include: "(12℃, blue, RGB: 173, 216, 230)...(37℃, orange, RGB: 255, 165, 0)".

[0060] Optionally, the execution entity may perform rasterization processing on each decoded flow field data block in the decoded flow field data block set to generate rasterized flow field data blocks, which may include the following steps:

[0061] The first step is to obtain the set of decoded flow field data blocks stored in the video memory nodes. In practice, this can be done using a pre-defined computation thread. This computation thread can be the thread used to perform rasterization processing. The computation thread can run in parallel with the decoding thread. Specifically, after the decoding thread completes its decoding task, the computation thread can begin accessing and processing these data blocks. Because the computation thread and the decoding thread work in parallel, even if the decoding thread is still decoding new frames, the computation thread can continue processing already decoded frames.

[0062] The second step involves determining the velocity information corresponding to each velocity component data included in the decoded flow field data block, thus obtaining a velocity information set. Each velocity information in this set can include a horizontal direction angle, a vertical elevation angle, and a velocity magnitude. Each velocity information in the set can be represented as a three-dimensional vector. In practice, the execution entity can use the aforementioned computation thread to determine the velocity information corresponding to each velocity component data included in the decoded flow field data block, thereby obtaining the velocity information set. Specifically, firstly, the L2 norm between the eastward, northward, and vertical velocity components in the velocity component data can be determined as the velocity magnitude. Secondly, the horizontal direction angle between the eastward and northward velocity components can be determined based on the arctan2 function. Then, the vertical elevation angle between the horizontal direction angle and the vertical velocity component can be determined based on the arctan2 function.

[0063] The third step is to determine the grid velocity information corresponding to each velocity information in the aforementioned velocity information set. This grid velocity information includes the horizontal horizontal direction angle, the average vertical elevation angle, the average velocity magnitude, and the turbulence disturbance amplitude. In practice, the average value among the various velocity information in the aforementioned velocity information set can be determined as the grid average velocity information. This grid average velocity information can include the horizontal horizontal direction angle, the average vertical elevation angle, and the average velocity magnitude. Next, the turbulence disturbance amplitude corresponding to each velocity information in the aforementioned velocity information set can be determined. Specifically, firstly, the root mean square value among the various horizontal direction angles included in the aforementioned velocity information set can be determined as the root mean square value of the horizontal direction angle. Secondly, the root mean square value among the various vertical elevation angles included in the aforementioned velocity information set can be determined as the root mean square value of the vertical elevation angle. Thirdly, the root mean square value among the various velocity magnitudes included in the aforementioned velocity information set can be determined as the root mean square value of the velocity magnitude. Subsequently, the L2 norm between the above-mentioned root mean square values ​​of the horizontal direction angle, the above-mentioned root mean square values ​​of the vertical elevation angle, and the above-mentioned root mean square values ​​of the velocity modulus can be determined as the turbulence disturbance amplitude.

[0064] The fourth step is to determine the average value among the various temperature data included in the above decoded flow field data block as the grid temperature.

[0065] The fifth step is to determine the color information corresponding to the above grid temperature based on the preset color texture mapping table. The above color texture mapping table can be a predefined temperature-color correspondence table. For example, the above color texture mapping table may include: "(12℃, blue, RGB: 173, 216, 230)...(37℃, orange, RGB: 255, 165, 0)".

[0066] Step 6: Based on the aforementioned raster flow velocity information and color information, generate rasterized flow field data blocks. In practice, the aforementioned raster flow velocity information and color information can be used as parameters for the decoded flow field data blocks to generate rasterized flow field data blocks.

[0067] Step 206: Perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the above rasterized flow field data block set to generate a rendered flow field view frame, and obtain a sequence of rendered flow field view frames.

[0068] In some embodiments, the execution entity can perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set to generate a rendered flow field view frame, resulting in a sequence of rendered flow field view frames. In practice, a preset rendering thread, based on a preset rendering technology, can perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set to generate a rendered flow field view frame, resulting in a sequence of rendered flow field view frames. The rendering thread can be a thread used to render flow field data. The rendering thread can run in parallel with the decoding thread and the computation thread. The rendering technology can be WebGL technology. Specifically, for each rasterized flow field data block at the same time point, an independent rendering thread can be used to utilize the graphics shader in WebGL technology to map each corresponding velocity component data as a particle and each temperature data as the color of the particle, generating a rendered flow field view frame that reflects multi-band physical information.

[0069] In practice, a common technical challenge in UAV flow field rendering technology for airports is the large volume of 3D flow field data, which slows down rendering speed and hinders real-time interaction and high-precision flow field rendering. This reduces the visualization accuracy of low-altitude areas of the airport, consequently impacting operational efficiency and safety. Therefore, the following solution is proposed.

[0070] Optionally, the execution entity may perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set to generate a rendered flow field view frame, which may include the following steps:

[0071] The first step is to generate the overall particle resolution based on the viewport range information and viewport frame rate of the interactive terminal. The overall particle resolution characterizes the average particle density corresponding to the rasterized flow field data block set. The unit of the overall particle resolution can be the number of particles per voxel grid. In practice, the overall particle resolution corresponding to the viewport range information and viewport frame rate can be determined based on a pre-defined viewport-particle resolution mapping table. This mapping table can be a predefined table of correspondence between viewport range information, viewport frame rate, and particle resolution. For example, the mapping table could include: "Viewport range information: Longitude range [117.5000°E, 117.5100°E], Latitude range [30.2000°N, 30.2055°N], Altitude range [1m, 15m]; Viewport frame rate: 0.3 frames per second; Particle resolution: 18 particles per voxel grid."

[0072] The second step involves performing the following steps on each rasterized flow field data block corresponding to the same time point in the above rasterized flow field data block set:

[0073] The first sub-step involves determining the rasterized flow field data blocks corresponding to the same time point as a single-frame rasterized flow field data block set. This single-frame rasterized flow field data block set can be a collection of all rasterized flow field data blocks required to construct the current viewport.

[0074] The second sub-step involves generating a corresponding local particle resolution for each single-frame rasterized flow field data block in the aforementioned single-frame rasterized flow field data block set, based on the overall particle resolution and the relationship between the single-frame rasterized flow field data block and the viewport. The local particle resolution can be defined as the number of particles in a single rasterized flow field data block. In practice, firstly, the distance between the single-frame rasterized flow field data block and the viewpoint coordinates corresponding to the viewport can be determined. Next, the square of this distance can be increased by one, and the reciprocal can be taken to obtain the viewport distance weight. Finally, the product of the overall particle resolution and the viewport distance weight can be rounded up to determine the local particle resolution.

[0075] The third sub-step involves determining the particle motion information set corresponding to each single-frame rasterized flow field data block in the aforementioned single-frame rasterized flow field data block set, based on the generated local particle resolutions. Each particle motion information in the aforementioned particle motion information set includes the particle motion direction and particle motion intensity. The particle motion direction can include the particle's horizontal and vertical angles. The particle motion intensity can be the particle's velocity magnitude. Each particle motion information in the aforementioned particle motion information set corresponds to three-dimensional coordinates. In practice, the local particle resolution corresponding to the aforementioned single-frame rasterized flow field data block can be determined as the number of particle motion information in the particle motion information set, and corresponding coordinates can be generated. These coordinates belong to the voxel grid of the aforementioned single-frame rasterized flow field data block. Subsequently, based on the turbulence disturbance amplitude corresponding to the aforementioned single-frame rasterized flow field data block, random perturbations can be added to the horizontal direction angle, average vertical elevation angle, and average velocity magnitude included in the aforementioned single-frame rasterized flow field data block to generate particle motion information corresponding to the aforementioned coordinates. Specifically, firstly, the turbulence disturbance amplitude can be multiplied by a preset disturbance percentage to obtain a disturbance percentage threshold. For example, the disturbance percentage can be 5%. Secondly, for the horizontal direction angle included in the single-frame rasterized flow field data block, a random number between the negative disturbance percentage threshold and the disturbance percentage threshold can be generated using a preset random number generator, serving as the random disturbance percentage. Then, the random disturbance percentage can be incremented by one and multiplied by the horizontal direction angle to obtain the horizontal wind direction angle after adding the disturbance. Similarly, the average vertical elevation angle and the average velocity magnitude after adding the disturbance can be generated. Finally, the horizontal wind direction angle and the average vertical elevation angle after adding the disturbance can be used as the particle motion direction. The average velocity magnitude after adding the disturbance is used as the particle motion intensity.

[0076] The fourth sub-step involves slicing the 3D mesh corresponding to the aforementioned single-frame rasterized flow field data block set to obtain a 2D mesh set. The aforementioned 3D mesh can be composed of voxel meshes corresponding to each of the single-frame rasterized flow field data blocks in the aforementioned single-frame rasterized flow field data block set. In practice, firstly, the 3D mesh corresponding to each single-frame rasterized flow field data block in the aforementioned single-frame rasterized flow field data block set can be determined. Then, the aforementioned 3D mesh can be sliced ​​along the vertical direction to obtain individual 2D meshes.

[0077] The fifth sub-step involves deforming each two-dimensional mesh in the aforementioned two-dimensional mesh set based on the determined particle motion information sets, thereby generating a two-dimensional deformed mesh set. In practice, firstly, the particle motion information sets can be sliced ​​using the same method as in the fourth sub-step to obtain the particle motion information set corresponding to each two-dimensional mesh. Next, for each two-dimensional mesh in the aforementioned two-dimensional mesh set, the particle motion information can be mapped back to the mesh vertices based on a preset mesh deformation method. The displacement increment of each vertex can be calculated, and its coordinates updated to generate a two-dimensional deformed mesh that reflects the dynamic characteristics of the flow field. This mesh deformation method can be based on particle-driven mesh deformation.

[0078] The sixth sub-step involves fusing the two-dimensional deformable mesh, a preset noise texture map, and a historical texture map for each two-dimensional deformable mesh in the aforementioned two-dimensional deformable mesh set to obtain a two-dimensional flow field map. The noise texture map can be a noise map used to enhance flow field details. This noise texture map can be a Perlin or Worley noise map. The historical texture map can be the previous frame's two-dimensional flow field map corresponding to the aforementioned two-dimensional deformable mesh. This historical texture map can be stored in the video memory node. In practice, a preset flow field visualization technique can be used to fuse the two-dimensional deformable mesh, the preset noise texture map, and the historical texture map to obtain a two-dimensional flow field map. This flow field visualization technique can be the IBFV (Image-Based Flow Visualization) algorithm. Specifically, firstly, the noise texture map can be mapped onto the two-dimensional deformable mesh using mesh cell texture mapping to obtain a mapped noise texture map. Then, the historical texture map and the mapped noise texture map can be weighted and mixed to generate a two-dimensional flow field map. The rendering flowcharts corresponding to the fourth to sixth sub-steps are as follows: Figure 3 As shown.

[0079] The seventh sub-step involves generating corresponding three-dimensional flow field models based on the determined two-dimensional flow field diagrams. These three-dimensional flow field models can represent the flow field state of the current frame. In practice, the determined two-dimensional flow field diagrams can be stacked in spatial order to obtain the three-dimensional flow field model.

[0080] The eighth sub-step involves rendering the 3D flow field model using the color information corresponding to each single-frame rasterized flow field data block in the aforementioned single-frame rasterized flow field data block set, resulting in a rendered 3D flow field model. This rendered 3D flow field model can be a 3D model with color textures. In practice, for each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set, the color information corresponding to that single-frame rasterized flow field data block can be used as the color of the corresponding voxel in the 3D flow field model, resulting in a rendered voxel mesh. Subsequently, the obtained rendered voxel meshes can be integrated to obtain the rendered 3D flow field model.

[0081] The ninth sub-step involves determining the rendered flow field view frame corresponding to the rendered 3D flow field model based on the aforementioned viewport range information. In practice, the rendered 3D flow field model can be mapped onto the 2D plane corresponding to the aforementioned viewport range information to obtain the rendered flow field view frame. Specifically, the execution entity can use the aforementioned rendering thread to call a preset graphics API to project the rendered 3D flow field model into a 2D view frame that conforms to the user's viewing angle, which will be the final output rendered flow field view frame.

[0082] The first and second steps and related content of the above-mentioned optional solutions constitute an inventive point of this disclosure, solving the aforementioned technical problem: "affecting the operational efficiency and safety of airports." The reason for reduced operational efficiency and safety of airports is often the large amount of 3D flow field data, leading to slow rendering speeds and difficulty in achieving real-time interaction and high-precision flow field rendering, thereby reducing the visualization accuracy of low-altitude areas of airports. Solving these factors can improve the operational efficiency and safety of airports. To achieve this effect, firstly, an overall particle resolution is generated based on the current viewport range and viewport frame rate. This allows for dynamic load balancing, achieving accurate rendering while avoiding excessive computation in invisible or non-critical areas. Secondly, data blocks within the current time point and current viewport can be loaded and rendered via a rendering thread, rather than global data across all time periods, significantly reducing memory usage and data processing volume. Then, based on the aforementioned overall particle resolution, a higher particle count is allocated to data blocks closer to the viewpoint, and a lower particle count to those farther away, using a view distance weighting system. This reduces the number of particles without sacrificing visual perception, while simultaneously improving local rendering details and global rendering effects. Next, based on the raster velocity information and local particle resolution corresponding to each flow field data block, a particle motion information set is generated to facilitate the subsequent rendering of the flow field motion state. Then, the particle motion information can be used to directly drive the displacement of 3D mesh vertices, enabling real-time and intuitive representation of the flow field morphology. Then, by using preset 3D noise, turbulent details are added to the flow field, and the results from the previous frame are blended into the current frame with a certain weight, effectively smoothing abrupt changes between frames, reducing flickering and blurring, and significantly improving the coherence and realism of the rendered flow field video. Finally, the 3D model is projected into a 2D view for interactive display. The rendering thread allows data loading, physical information calculation, and graphics rendering to be performed in a pipelined, overlapping manner, eliminating waiting time during sequential processing. The overall solution achieves a balance between flow field rendering accuracy and time.

[0083] Step 207: Perform time interpolation on the above-rendered flow field view frame sequence to obtain a smooth flow field video.

[0084] In some embodiments, the execution entity can perform temporal interpolation on the rendered flow field view frame sequence to obtain a smooth flow field video. In practice, a preset video interpolation algorithm can be used to perform temporal interpolation on the rendered flow field view frame sequence to obtain a smooth flow field video. The video interpolation algorithm can be a motion estimation and compensation (MEMC) algorithm or a temporal linear interpolation method.

[0085] In practice, a common technical problem encountered when generating rendered flow field videos is that the difference between sampling time points leads to discontinuities between two rendered flow field view frames. Increasing the original sampling frequency significantly increases the GPU load, making it difficult to improve the continuity and flow field characteristics between frames while reducing computational load. Therefore, the following solution is proposed.

[0086] Optionally, the execution entity may perform time interpolation on the rendered flow field view frame sequence to obtain a smooth flow field video, which may include the following steps:

[0087] The first step is to perform the following steps for each rendered flow field view frame in the above rendered flow field view frame sequence:

[0088] The first sub-step involves determining the rendered flow field view frame and the corresponding next rendered flow field view frame as the subsequent rendered flow field view frame.

[0089] The second sub-step generates the flow field change amplitude based on the rendered flow field view frame and the subsequently rendered flow field view frame. This flow field change amplitude characterizes the intensity of the flow field change between the two flow field view frames. In practice, the flow field change amplitude between the rendered and subsequently rendered flow field view frames can be determined based on a preset optical flow algorithm. This optical flow algorithm can be the Lucas-Kanade optical flow method. Specifically, firstly, the motion vector matrix between the rendered and subsequently rendered flow field view frames can be determined using the optical flow algorithm. The width and height of the motion vector matrix can be the same as those of the rendered and subsequently rendered flow field view frames. Each element in the motion vector matrix can include a horizontal displacement component and a vertical displacement component. Next, for each element in the motion vector matrix, the L2 norm between the horizontal and vertical displacement components can be determined as the motion vector magnitude. Finally, the average value among the determined motion vector magnitudes can be determined as the flow field change amplitude.

[0090] The third sub-step involves determining the interpolation frame rate corresponding to the aforementioned flow field change amplitude. This interpolation frame rate can be the number of video frames per unit time. In practice, the interpolation frame rate can be determined based on a pre-defined heat flow to interpolation frame mapping table. This heat flow to interpolation frame mapping table can be a table showing the correspondence between heat flow and interpolation frames.

[0091] The fourth sub-step involves performing bilinear interpolation on the rendered flow field view frames and the subsequently rendered flow field video frames, based on the aforementioned interpolation frame rate, to obtain an interpolated frame set. In practice, multiple bilinear interpolations can be performed between the rendered flow field view frames and the subsequently rendered flow field video frames using a preset bilinear interpolation technique to obtain the interpolated frame set. The ratio of the number of interpolated frames in the interpolated frame set to the time interval is equal to the aforementioned interpolation frame rate.

[0092] The fifth sub-step involves performing velocity field correction on each interpolated frame in the aforementioned interpolated frame set to generate a corrected interpolated frame set. In practice, for each interpolated frame in the aforementioned interpolated frame set, the motion vector matrix between the interpolated frame and the previous frame view can be determined based on the aforementioned optical flow algorithm. Then, for the motion vector magnitude corresponding to each element in the aforementioned motion vector matrix, the ratio between the motion vector magnitude and a preset time interval can be determined as the pixel displacement velocity. The aforementioned time interval is the time interval between the interpolated frame and the previous frame view. Afterwards, for each determined pixel displacement velocity, if it is determined that each pixel displacement velocity is less than or equal to a preset motion threshold, then the interpolated frame is determined as a corrected interpolated frame. The unit of the aforementioned motion threshold can be pixels per second. For example, the aforementioned motion threshold can be 25 pixels per second. When it is determined that any pixel displacement velocity among the aforementioned pixel displacement velocities exceeds a preset motion threshold, for pixels in the interpolated frame whose displacement velocities exceed the preset motion threshold, the corresponding horizontal and vertical displacement components are multiplied by the ratio between the motion threshold and the pixel displacement velocity, respectively, to obtain corrected horizontal and vertical displacement components. Based on these corrected horizontal and vertical displacement components, the new coordinates corresponding to the pixel are determined, resulting in an updated pixel. The interpolated frame is then updated based on these updated pixels to obtain an updated interpolated frame, which serves as the corrected interpolated frame.

[0093] The second step is to insert the obtained sets of corrected interpolated frames into the above-mentioned rendered flow field view frame sequence in chronological order to obtain a smooth flow field video.

[0094] The first and second steps and related content of the above-mentioned optional solutions constitute an inventive point of this disclosure, solving the aforementioned technical problem: "difficulty in improving the continuity and flow field characteristics between frames while reducing computational load." The reduced continuity and flow field characteristics between frames are often caused by the gap between sampling time points, resulting in discontinuities between two rendered flow field view frames. Increasing the original sampling frequency significantly increases GPU load. Solving these factors allows for improving the continuity and flow field characteristics between frames while reducing computational load. To achieve this, firstly, the amplitude of flow field change between two adjacent frames is determined. This allows the amplitude of flow field change to reflect the intensity of flow field changes. Then, based on the amplitude of flow field change, the interpolation density is adaptively determined. This allows for more interpolation frames in areas of intense flow field change and fewer interpolation frames in stable areas, avoiding wasted computational power in static regions and achieving precise resource allocation. Next, a set of interpolated frames between two adjacent frames is determined using bilinear interpolation technology. Then, velocity field correction is applied to each interpolated frame in the interpolated frame set to improve the rationality of particle motion in the interpolated frames. Finally, the interpolated frames and the rendered flow field view frame sequence are arranged in chronological order to generate a smooth flow field video. This allows for smoothing between two adjacent frames in a two-dimensional view, reducing computational load while improving the continuity and smoothness between each frame.

[0095] Further reference Figure 4 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a UAV flow field rendering device based on asynchronous rasterization and multi-band fusion. These device embodiments are similar to... Figure 2 Corresponding to the method embodiments shown, this UAV flow field rendering device based on asynchronous rasterization and multi-band fusion can be specifically applied to various electronic devices.

[0096] like Figure 4As shown, a UAV flow field rendering device 400 based on asynchronous rasterization and multi-band fusion in some embodiments includes: a thermal flow data acquisition unit 401, a spatiotemporal segmentation unit 402, an encapsulation unit 403, an asynchronous loading unit 404, a rasterization processing unit 405, a multi-band fusion rendering unit 406, and a time interpolation unit 407. The thermal flow data acquisition unit 401 is configured to acquire thermal flow data from an airport to be tested using a UAV, obtaining a UAV flow field dataset; the spatiotemporal segmentation unit 402 is configured to perform spatiotemporal segmentation on the aforementioned UAV flow field dataset, obtaining a UAV flow field data block set; the encapsulation unit 403 is configured to encapsulate each UAV flow field data block in the aforementioned UAV flow field data block set to generate encapsulated flow field data blocks, obtaining an encapsulated flow field data block set; and the asynchronous loading unit 404 is configured to asynchronously load the aforementioned encapsulated flow field data block set according to the viewport information of the interactive terminal. A set of decoded flow field data blocks is obtained; a rasterization processing unit 405 is configured to perform rasterization processing on each decoded flow field data block in the above-mentioned set of decoded flow field data blocks to generate rasterized flow field data blocks, thereby obtaining a set of rasterized flow field data blocks; a multi-band fusion rendering unit 406 is configured to perform multi-band fusion rendering on each rasterized flow field data block in the above-mentioned set of rasterized flow field data blocks corresponding to the same time point, thereby generating a rendered flow field view frame, thereby obtaining a sequence of rendered flow field view frames; a time interpolation unit 407 is configured to perform time interpolation on the above-mentioned sequence of rendered flow field view frames to obtain a smooth flow field video.

[0097] It is understandable that the units and references described in the UAV flow field rendering device 400 based on asynchronous rasterization and multi-band fusion are... Figure 2 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the UAV flow field rendering device 400 based on asynchronous rasterization and multi-band fusion and the units contained therein, and will not be repeated here.

[0098] The following is for reference. Figure 5 It shows a schematic diagram of the structure of an electronic device 500 (e.g., a computing device) suitable for implementing some embodiments of the present disclosure. Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0099] like Figure 5As shown, the electronic device 500 may include a processing unit 501 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the electronic device 500. The processing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0100] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows electronic device 500 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 An electronic device 500 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 5 Each box shown can represent a device or multiple devices as needed.

[0101] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 509, or installed from storage device 508, or installed from ROM 502. When the computer program is executed by processing device 501, it performs the functions defined in the methods of some embodiments of this disclosure.

[0102] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0103] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0104] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs. When the electronic device executes the aforementioned one or more programs, the electronic device causes the following to occur: It collects thermal flow data from an airport under test using a UAV to obtain a UAV flow field dataset; it performs spatiotemporal segmentation on the aforementioned UAV flow field dataset to obtain a UAV flow field data block set; it encapsulates each UAV flow field data block in the aforementioned UAV flow field data block set to generate encapsulated flow field data blocks, thus obtaining an encapsulated flow field data block set; it asynchronously loads the aforementioned encapsulated flow field data block set according to the viewport information of the interactive terminal to obtain a decoded flow field data block set; it rasterizes each decoded flow field data block in the aforementioned decoded flow field data block set to generate rasterized flow field data blocks, thus obtaining a rasterized flow field data block set; it performs multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the aforementioned rasterized flow field data block set to generate a rendered flow field view frame, thus obtaining a rendered flow field view frame sequence; and it performs temporal interpolation on the aforementioned rendered flow field view frame sequence to obtain a smoothed flow field video.

[0105] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0106] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0107] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0108] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for rendering UAV flow fields based on asynchronous rasterization and multi-band fusion, characterized in that, include: The UAV flow field dataset is obtained by collecting heat flow data from the airport to be tested using a UAV. The UAV flow field dataset is spatiotemporally divided into blocks to obtain a UAV flow field data block set; Each UAV flow field data block in the UAV flow field data block set is encapsulated to generate an encapsulated flow field data block, thus obtaining an encapsulated flow field data block set. Based on the viewport information of the interactive terminal, the encapsulated flow field data block set is asynchronously loaded to obtain the decoded flow field data block set; Each decoded flow field data block in the decoded flow field data block set is rasterized to generate a rasterized flow field data block, thus obtaining a rasterized flow field data block set; Multi-band fusion rendering is performed on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set to generate a rendered flow field view frame, thus obtaining a sequence of rendered flow field view frames. Time interpolation is performed on the rendered flow field view frame sequence to obtain a smooth flow field video; The process of multi-band fusion rendering of each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set includes: The overall particle resolution is generated based on the viewport range information and viewport frame rate of the interactive terminal. For each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set, the following steps are performed: Each rasterized flow field data block corresponding to the same time point is determined as a single-frame rasterized flow field data block set; For each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set, a corresponding local particle resolution is generated based on the overall particle resolution and the relationship between the single-frame rasterized flow field data block and the view window. Based on the resolution of each generated local particle, determine the particle motion information set corresponding to each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set; The three-dimensional mesh corresponding to the single-frame rasterized flow field data block set is sliced ​​to obtain a two-dimensional mesh set; Based on the determined set of particle motion information, each two-dimensional grid in the two-dimensional grid set is deformed to generate a two-dimensional deformed grid, thus obtaining a set of two-dimensional deformed grids. For each two-dimensional deformable mesh in the two-dimensional deformable mesh set, the two-dimensional deformable mesh, the preset noise texture map, and the historical texture map are fused to obtain a two-dimensional flow field map; Based on the determined two-dimensional flow field diagrams, generate the corresponding three-dimensional flow field models; Based on the color information corresponding to each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set, the three-dimensional flow field model is color-rendered to obtain the rendered three-dimensional flow field model. Based on the viewport range information, determine the rendered flow field view frame corresponding to the rendered 3D flow field model.

2. The method according to claim 1, characterized in that, The process of spatiotemporally partitioning the UAV flow field dataset to obtain a UAV flow field data block set includes: The three-dimensional space corresponding to the airport where the heat flow to be tested is to be tested is divided into multiple voxels to obtain a voxel set; The flow field data of each UAV corresponding to the same voxel in the UAV flow field dataset are combined into a voxel UAV flow field data group to obtain a voxel UAV flow field data group set. For each voxel drone flow field data group in the voxel drone flow field data group set, the voxel drone flow field data corresponding to the same time period in the voxel drone flow field data group are combined into a single frame voxel drone flow field data group to obtain a single frame voxel drone flow field data group set. Each single-frame voxel UAV flow field data group in the obtained single-frame voxel UAV flow field data set is determined as a UAV flow field data block, thus obtaining a UAV flow field data block set.

3. The method according to claim 1, characterized in that, Each UAV flow field data block includes a velocity component dataset and a temperature dataset; The encapsulation of each UAV flow field data block in the UAV flow field data block set to generate an encapsulated flow field data block includes: Each velocity component data included in the UAV flow field data block is predicted and encoded to generate encoded velocity component data, thus obtaining an encoded velocity component dataset. Wavelet transform compression is performed on each temperature data included in the UAV flow field data block to generate compressed temperature data, thus obtaining a compressed temperature dataset. The UAV flow field data block is updated based on the encoded flow velocity component dataset and the compressed temperature dataset to obtain the updated UAV flow field data block, which serves as the compressed flow field data block. The compressed flow field data block is encapsulated to generate an encapsulated flow field data block.

4. The method according to claim 1, characterized in that, The step of asynchronously loading the encapsulated flow field data block set based on the viewport range information of the interactive terminal to obtain the decoded flow field data block set includes: In response to receiving window range information from the interactive terminal, each encapsulated flow field data block in the encapsulated flow field data block set corresponding to the window range information is determined as a target flow field data block, thus obtaining a target flow field data block set; Each encapsulated flow field data block that is temporally adjacent to each target flow field data block in the encapsulated flow field data block set is determined as an adjacent temporal flow field data block, thus obtaining an adjacent temporal flow field data block set; The target flow field data block set and the adjacent time-series flow field data block set are obtained from the edge node or the local node through the decoding thread, and each target flow field data block and each adjacent time-series flow field data block are decoded to generate a decoded flow field data block, thereby obtaining a decoded flow field data block set, and the decoded flow field data block set is stored in the video memory node.

5. The method according to claim 4, characterized in that, The method further includes: Determine the set of prefetched flow field data blocks corresponding to the window range information, and store the set of prefetched flow field data blocks to the edge node; Each encapsulated flow field data block in the encapsulated flow field data block set that meets the preset historical viewing conditions is identified as a historical viewing flow field data block, thus obtaining a historical viewing flow field data block set, and storing the historical viewing flow field data block set to a local node.

6. The method according to claim 1, characterized in that, The step of rasterizing each decoded flow field data block in the decoded flow field data block set to generate a rasterized flow field data block includes: Obtain the set of decoded flow field data blocks stored in the video memory node; Determine the flow velocity information corresponding to each flow velocity component data included in the decoded flow field data block to obtain a flow velocity information set, wherein each flow velocity information in the flow velocity information set includes a horizontal direction angle, a vertical elevation angle, and a flow velocity magnitude. Determine the grid velocity information corresponding to each velocity information in the velocity information set, wherein the grid velocity information includes the horizontal direction angle, the average vertical elevation angle, the average velocity magnitude, and the turbulence disturbance amplitude; The average value among the various temperature data included in the decoded flow field data block is determined as the grid temperature; Based on a preset color texture mapping table, determine the color information corresponding to the grid temperature; Based on the grid flow velocity information and the color information, a rasterized flow field data block is generated.

7. A UAV flow field rendering device based on asynchronous rasterization and multi-band fusion, characterized in that, include: The heat flow data acquisition unit is configured to collect heat flow data from the airport to be tested using a drone, and obtain a drone flow field dataset. The spatiotemporal segmentation unit is configured to perform spatiotemporal segmentation on the UAV flow field dataset to obtain a UAV flow field data block set; The encapsulation unit is configured to encapsulate each UAV flow field data block in the UAV flow field data block set to generate encapsulated flow field data blocks, thereby obtaining an encapsulated flow field data block set. The asynchronous loading unit is configured to asynchronously load the encapsulated flow field data block set according to the window range information of the interactive terminal to obtain the decoded flow field data block set; The rasterization processing unit is configured to perform rasterization processing on each decoded flow field data block in the decoded flow field data block set to generate rasterized flow field data blocks, thereby obtaining a rasterized flow field data block set. A multi-band fusion rendering unit is configured to perform multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set, to generate a rendered flow field view frame, thus obtaining a sequence of rendered flow field view frames; wherein, performing multi-band fusion rendering on each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set includes: The overall particle resolution is generated based on the viewport range information and viewport frame rate of the interactive terminal. For each rasterized flow field data block corresponding to the same time point in the rasterized flow field data block set, the following steps are performed: Each rasterized flow field data block corresponding to the same time point is determined as a single-frame rasterized flow field data block set; For each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set, a corresponding local particle resolution is generated based on the overall particle resolution and the relationship between the single-frame rasterized flow field data block and the view window. Based on the resolution of each generated local particle, determine the particle motion information set corresponding to each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set; The three-dimensional mesh corresponding to the single-frame rasterized flow field data block set is sliced ​​to obtain a two-dimensional mesh set; Based on the determined set of particle motion information, each two-dimensional grid in the two-dimensional grid set is deformed to generate a two-dimensional deformed grid, thus obtaining a set of two-dimensional deformed grids. For each two-dimensional deformable mesh in the two-dimensional deformable mesh set, the two-dimensional deformable mesh, the preset noise texture map, and the historical texture map are fused to obtain a two-dimensional flow field map; Based on the determined two-dimensional flow field diagrams, generate the corresponding three-dimensional flow field models; Based on the color information corresponding to each single-frame rasterized flow field data block in the single-frame rasterized flow field data block set, the three-dimensional flow field model is color-rendered to obtain the rendered three-dimensional flow field model. Based on the window range information, determine the rendered flow field view frame corresponding to the rendered 3D flow field model; A time interpolation unit is configured to perform time interpolation on the rendered flow field view frame sequence to obtain a smooth flow field video.

8. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 6.

9. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images