Space-Ground Collaborative Rapid Information Service System and Method

Abstract

This application provides a space-ground collaborative rapid information service system and method, which can be applied to the field of remote sensing information technology. The method includes: a space-ground integrated central cloud network for receiving external remote sensing tasks, determining the space-based satellite platform and ground computing nodes that perform the remote sensing tasks, parsing the task parameters of the remote sensing tasks, generating a space-ground collaborative processing link based on the task parameters, and distributing the processing link to the space-based satellite platform and ground computing nodes; the space-based satellite platform for collecting raw remote sensing data according to the instructions of the on-orbit processing tasks in the processing link, generating valid data based on the raw remote sensing data, and distributing the valid data according to the instructions of the receiver address in the processing link; and the ground computing nodes for receiving valid data according to the instructions of the ground processing tasks in the processing link, performing ground refinement processing on the valid data to obtain refined target coordinates, verifying and generating the final results, and distributing the final results according to the processing link.

Description

Technical Field

[0001] This application relates to the field of remote sensing information technology, and more specifically to a satellite-ground collaborative rapid information service system and method. Background Technology

[0002] Early remote sensing information services long adopted a "space-to-ground computing" model, meaning that the satellite platform only undertook the function of remote sensing data acquisition and did not have on-orbit data processing capabilities. The raw remote sensing images, spectral data, radar data, etc. acquired by the satellite had to be completely transmitted to the ground data center through the satellite-to-ground data transmission link, and then the ground computing cluster would perform a full-process processing operation, including data decompression, noise reduction, feature extraction, target identification, coordinate refinement, etc.

[0003] The "sky-sensing-ground-computing" model has the following technical defects: First, the data transmission link is lengthy, the satellite transit window is short, and the data transmission bandwidth is limited. The transmission of massive amounts of raw data takes several hours or even days, resulting in extremely high processing latency. Second, the ground processing pressure is concentrated, and massive amounts of redundant data occupy a large amount of storage and computing resources, resulting in low processing efficiency. Third, the response speed is lagging, and it cannot meet the rapid information support needs of "minute-level" or "second-level" in scenarios such as sudden disasters and target maneuvers. Summary of the Invention

[0004] In view of the above problems, this application provides a satellite-ground collaborative rapid information service system and method.

[0005] According to the first aspect of this application, a space-ground collaborative rapid information service system is provided, comprising: a space-ground integrated central cloud network, at least one space-based satellite platform, and at least one ground computing node;

[0006] The integrated space-ground central cloud network is communicatively connected to the space-based satellite platform and the ground computing node, respectively.

[0007] The integrated space-ground central cloud network is used to receive external remote sensing tasks, determine the space-based satellite platform and ground computing nodes that will execute the remote sensing tasks, parse the task parameters of the remote sensing tasks, generate a space-ground collaborative processing link based on the task parameters, and send the processing link to the space-based satellite platform and the ground computing nodes. The processing link includes at least the on-orbit processing tasks that the space-based satellite platform needs to execute, the ground processing tasks that the ground computing nodes need to execute, and the address of the receiver for the valid data transmitted by the space-based satellite platform.

[0008] The space-based satellite platform is used to collect raw remote sensing data according to the instructions of the on-orbit processing task in the processing link, generate valid data based on the raw remote sensing data, and download the valid data according to the instructions of the receiver address in the processing link.

[0009] The ground computing node is used to receive the valid data according to the instructions of the ground processing task in the processing link, perform ground refinement processing on the valid data to obtain refined target coordinates, verify the refined target coordinates, generate the final result, and distribute the final result according to the processing link.

[0010] According to an embodiment of this application, determining the space-based satellite platform and ground computing node for performing the remote sensing mission includes:

[0011] For each connected computing node, a normalized computing power value is calculated based on the actual computing power value of the computing node and the node type weight, wherein the weight of space-based computing nodes is lower than that of ground-based core computing nodes.

[0012] The computing power of the remote sensing task is allocated based on the normalized computing power value of each computing node, thereby determining the space-based satellite platform and ground computing nodes that will execute the remote sensing task.

[0013] According to an embodiment of this application, the step of parsing the mission parameters of the remote sensing mission and generating a satellite-ground collaborative processing link based on the mission parameters includes:

[0014] The timeliness level and accuracy requirements are extracted from the remote sensing task.

[0015] The first weighting coefficient for the processing delay optimization objective is determined based on the aforementioned timeliness level;

[0016] Determine the second weighting coefficient of the computing power utilization optimization target based on the aforementioned accuracy requirements;

[0017] The processing link is generated with three optimization objectives: minimizing processing latency, maximizing computing power utilization, and minimizing execution cost. The first weight coefficient, the second weight coefficient, and a preset third weight coefficient are assigned to each of these objectives respectively.

[0018] Specifically, when the remote sensing task is an emergency task, the first weighting coefficient is greater than the second weighting coefficient and the third weighting coefficient; when the remote sensing task is a routine task, the second weighting coefficient is greater than the first weighting coefficient and the third weighting coefficient.

[0019] According to an embodiment of this application, the step of generating valid data based on the original remote sensing data and transmitting the valid data according to the indication of the receiver address in the processing link includes:

[0020] The raw remote sensing data is preprocessed and target detection is performed sequentially in orbit. The preprocessing includes radiometric calibration, atmospheric correction and noise removal. The target detection is used to mark the location and category of suspected targets.

[0021] The effective data obtained after preprocessing and target detection is downloaded, while the original remote sensing data is retained on the space-based satellite platform and not downloaded.

[0022] The volume of the effective data is less than 10% of the volume of the original remote sensing data.

[0023] According to an embodiment of this application, the process of performing ground refinement to obtain refined target coordinates includes:

[0024] Receive valid data from the space-based satellite platform, the valid data containing the target's coordinates with an accuracy of ten meters;

[0025] Retrieve pre-stored ground control point data;

[0026] Using bundle adjustment or georeferencing algorithms, the coordinates in the effective data are matched and calibrated with the ground control point data to output refined target coordinates. The accuracy of the calibrated target coordinates reaches the sub-meter level.

[0027] According to an embodiment of this application, the step of verifying and confirming the refined target coordinates to generate the final result includes:

[0028] The refined target coordinates are then pushed to the manual review console.

[0029] The manual review platform performs three levels of review in sequence: the first level of review is used to check the accuracy of target identification, the second level of review is used to check the positioning accuracy and data integrity, and the third level of review is used to check the compliance of the results.

[0030] If all three levels of review are passed, the refined target coordinates will be output as the final result.

[0031] If any level of review fails, the refined target coordinates will be returned for reprocessing, and the review comments will be synchronized to the integrated space-ground central cloud network.

[0032] According to a second aspect of this application, a method for providing rapid information services via satellite-ground collaboration using the system described in the first aspect is provided, comprising:

[0033] A space-ground integrated central cloud network is established, and at least one space-based satellite platform and at least one ground computing node are connected to the space-ground integrated central cloud network as heterogeneous computing resources.

[0034] The system receives external remote sensing tasks through the integrated space-ground central cloud network, determines the space-based satellite platform and ground computing node to execute the remote sensing tasks, parses the task parameters of the remote sensing tasks, generates a space-ground collaborative processing link based on the task parameters, and sends the processing link to the space-based satellite platform and the ground computing node. The processing link includes at least the on-orbit processing tasks to be executed by the space-based satellite platform, the ground processing tasks to be executed by the ground computing node, and the address of the receiver for the valid data transmitted by the space-based satellite platform.

[0035] The space-based satellite platform collects raw remote sensing data according to the on-orbit processing task instructions in the processing link, generates valid data based on the raw remote sensing data, and downloads the valid data according to the receiver address instructions in the processing link.

[0036] The ground computing node receives the valid data according to the instructions of the ground processing task in the processing link, performs ground refinement processing on the valid data to obtain refined target coordinates, verifies the refined target coordinates, generates the final result, and distributes the final result according to the processing link.

[0037] According to an embodiment of this application, determining the space-based satellite platform and ground computing node for performing the remote sensing mission includes:

[0038] For each connected computing node, a normalized computing power value is calculated based on the actual computing power value of the computing node and the node type weight, wherein the weight of space-based computing nodes is lower than that of ground-based core computing nodes.

[0039] The computing power of the remote sensing task is allocated based on the normalized computing power value of each computing node, thereby determining the space-based satellite platform and ground computing nodes that will execute the remote sensing task.

[0040] According to an embodiment of this application, the step of parsing the mission parameters of the remote sensing mission and generating a satellite-ground collaborative processing link based on the mission parameters includes:

[0041] The timeliness level and accuracy requirements are extracted from the remote sensing task.

[0042] The first weighting coefficient for the processing delay optimization objective is determined based on the aforementioned timeliness level;

[0043] Determine the second weighting coefficient of the computing power utilization optimization target based on the aforementioned accuracy requirements;

[0044] The processing link is generated with three optimization objectives: minimizing processing latency, maximizing computing power utilization, and minimizing execution cost. The first weight coefficient, the second weight coefficient, and a preset third weight coefficient are assigned to each of these objectives respectively.

[0045] Specifically, when the remote sensing task is an emergency task, the first weighting coefficient is greater than the second weighting coefficient and the third weighting coefficient; when the remote sensing task is a routine task, the second weighting coefficient is greater than the first weighting coefficient and the third weighting coefficient.

[0046] According to an embodiment of this application, the step of generating valid data based on the original remote sensing data and transmitting the valid data according to the instruction of the receiver address in the processing link includes:

[0047] The raw remote sensing data is preprocessed and target detection is performed sequentially in orbit. The preprocessing includes radiometric calibration, atmospheric correction and noise removal. The target detection is used to mark the location and category of suspected targets.

[0048] The effective data obtained after preprocessing and target detection is downloaded, while the original remote sensing data is retained on the space-based satellite platform and not downloaded.

[0049] The volume of the effective data is less than 10% of the volume of the original remote sensing data. Attached Figure Description

[0050] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0051] Figure 1 This schematic diagram illustrates the structure of a satellite-ground collaborative rapid information service system according to an embodiment of this application.

[0052] Figure 2 A flowchart illustrating a satellite-ground collaborative rapid information service method according to an embodiment of this application is shown. Detailed Implementation

[0053] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0054] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0055] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0056] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0057] Figure 1 The diagram illustrates the structure of a satellite-ground collaborative rapid information service system according to an embodiment of this application.

[0058] like Figure 1 As shown, the space-ground collaborative rapid information service system includes a space-ground integrated central cloud network 100, at least one space-based satellite platform (e.g., space-based satellite platform 200a, space-based satellite platform 200b, space-based satellite platform 200c), and at least one ground computing node 300. The space-ground integrated central cloud network 100 is communicatively connected to both the space-based satellite platform and the ground computing node 300.

[0059] According to the embodiments of this application, the space-ground integrated central cloud network 100 is the core control unit of the entire system. It is used to receive external remote sensing tasks, determine the space-based satellite platform and ground computing nodes that perform the remote sensing tasks, parse the task parameters of the remote sensing tasks, generate a space-ground collaborative processing link based on the task parameters, and send the processing link to the space-based satellite platform and ground computing nodes.

[0060] Specifically, the integrated space-ground cloud network 100 includes a ground-based core cloud platform, space-based relay cloud nodes, and a space-to-ground communication gateway. The ground-based core cloud platform can be deployed in ground-based data centers, such as national remote sensing centers, regional big data centers, or cloud computing service platforms; the space-based relay cloud nodes can be deployed as backbone nodes in geostationary orbit satellites or low-Earth orbit satellite constellations; and the space-to-ground communication gateway is used to accommodate various types of space-to-ground data transmission links, such as S-band links, Ka-band links, or laser communication links.

[0061] External remote sensing missions can originate from various sources, such as emergency command platforms, land management systems, environmental monitoring systems, or civilian and commercial remote sensing service platforms. Mission parameters include, but are not limited to: timeliness level (e.g., second-level, minute-level, hour-level), accuracy requirements (e.g., sub-meter-level, meter-level, ten-meter-level), target area range (e.g., latitude and longitude coordinate range), data type (e.g., optical imagery, synthetic aperture radar data, hyperspectral data), and deliverable format requirements.

[0062] According to the embodiments of this application, the processing link refers to the set of space-ground collaborative instructions generated by the integrated space-ground central cloud network 100 for performing remote sensing tasks. The processing link includes at least the following three types of information: First, the on-orbit processing tasks that the space-based satellite platform needs to perform, such as collecting raw remote sensing data of the target area, performing radiometric calibration, atmospheric correction, noise removal, target detection, etc. on the raw data; Second, the ground processing tasks that the ground computing node 300 needs to perform, such as target coordinate refinement, feature verification, image enhancement, dynamic analysis, etc.; Third, the address of the receiver of the valid data transmitted by the space-based satellite platform, such as the network address of the ground computing node 300 or the gateway address of the integrated space-ground central cloud network 100.

[0063] According to the embodiments of this application, the space-based satellite platform is the front-end sensing and rapid processing unit of the system, which is used to collect raw remote sensing data according to the instructions of the on-orbit processing tasks in the processing link, generate valid data based on the raw remote sensing data, and download the valid data according to the instructions of the receiver address in the processing link.

[0064] Specifically, the space-based satellite platform includes a remote sensing payload, an on-orbit computing module, and a space-to-ground communication module. The remote sensing payload can be an optical camera, a synthetic aperture radar, or a hyperspectral imager; the on-orbit computing module is equipped with a lightweight intelligent recognition algorithm; and the space-to-ground communication module is used to receive the processing links issued by the space-ground integrated hub cloud network 100 and send the valid data to the designated receiver via the space-to-ground link.

[0065] Raw remote sensing data refers to the unprocessed raw data collected by remote sensing payloads, such as uncorrected raw pixel matrices of satellite images, raw echo signal data, or raw spectral curve data.

[0066] Effective data refers to the data obtained by a space-based satellite platform after in-orbit processing of raw remote sensing data, and is used for downlink transmission. Compared with the raw remote sensing data, effective data removes a large amount of redundant information, such as cloud-obscured areas, invalid pixels, and repetitive textures, retaining only key information, such as the location coordinates, category labels, and feature vectors of targets. For example, when the raw remote sensing data is a 100MB satellite image, the effective data generated after in-orbit processing may only be 5MB to 10MB, containing the location and category information of all detected targets.

[0067] According to the embodiments of this application, the ground computing node 300 is the back-end precision processing unit of the system, which is used to receive valid data according to the instructions of the ground processing task in the processing link, perform ground refinement processing on the valid data to obtain refined target coordinates, verify the refined target coordinates, generate the final result, and distribute the final result according to the processing link.

[0068] Specifically, the ground computing node 300 includes a high-precision computing cluster and a manual review console. The high-precision computing cluster can be a GPU server cluster or a high-performance computing cluster deployed in a ground data center; the manual review console can include a monitor, input devices, and review software running on the terminal device. The ground computing node 300 can be deployed in a tiered manner, such as a national-level core computing center and regional-level edge computing nodes. The national-level core computing center can be deployed in supercomputing centers or cloud data centers in first-tier cities, while the regional-level edge computing nodes can be deployed in provincial or municipal remote sensing application centers, close to the user side to ensure response speed.

[0069] Ground-based refinement refers to further processing and optimizing effective data transmitted from space using abundant ground-based computing resources. Specifically, this includes: employing bundle adjustment or georegistration algorithms, combined with ground control point data, to optimize target coordinates in the effective data from ten-meter accuracy obtained from on-board processing to sub-meter accuracy; sharpening, mosaicking, or removing shadows from remote sensing images to improve image quality; refining and verifying target features to distinguish real targets from interference; and generating specific deliverables based on mission requirements, such as target trajectories, change monitoring maps, or area calculation data.

[0070] Verification and validation refer to the quality inspection of the results after ground-based refinement. Specifically, a three-level review can be performed sequentially through a manual review console: the first level checks the accuracy of target identification, determining whether there are any misidentifications or omissions; the second level checks the positioning accuracy and data integrity, confirming whether the coordinate deviation is within the allowable range and whether the data is complete; the third level checks the compliance of the results, ensuring that the final results meet industry standards or mission requirements. For results that fail the review, the system returns them for reprocessing and feeds back the review comments to the integrated space-ground hub cloud network 100, forming a closed-loop management system.

[0071] The final deliverable refers to the deliverable generated after refinement and verification by 300 ground computing nodes. The format of the final deliverable can be specified by the deliverable format requirements in the processing chain, such as GIS vector data (Shapefile format), remote sensing image map (GeoTIFF format), PDF inspection report, Excel data report, or dynamic situation map.

[0072] According to the embodiments of this application, a unified scheduling mechanism is established between the space-based satellite platform and ground computing nodes via a central cloud network. The central cloud network generates a processing link that includes on-orbit processing tasks, ground processing tasks, and receiver addresses. This enables the space-based platform to collect data, generate valid data, and transmit it to the designated receiver according to the link instructions. Ground nodes receive data according to the link instructions and sequentially perform refinement processing, verification, and result distribution, thus constructing a complete closed loop from task reception to result distribution. Compared with related technologies, this application achieves precise division of labor and seamless connection between space-based and ground-based tasks through the processing link, avoiding the low scheduling efficiency caused by the independent operation of space-based and ground-based resources. After the space-based platform generates valid data through on-orbit processing, it transmits it down while the original data remains on the satellite, significantly reducing the amount of data transmitted between space and ground. Ground nodes perform refinement processing and verification on the data transmitted from the space-based platform, balancing the dual requirements of rapid response and high-precision output. The receiver addresses in the processing link ensure that data flows directly to the designated nodes, reducing intermediate forwarding links and further improving timeliness.

[0073] In one embodiment, the integrated space-ground central cloud network 100 determines the space-based satellite platform and ground computing nodes that perform remote sensing tasks in the following way:

[0074] First, the integrated space-ground cloud network 100 calculates a normalized computing power value for each connected computing node. The computing nodes are divided into two categories: space-based satellite platforms and 300 ground-based computing nodes. Space-based computing nodes refer to satellite platforms operating in orbit, whose computing power is strictly constrained by size, weight, and power consumption (SWaP), resulting in relatively limited processing capabilities. Ground-based core computing nodes refer to high-performance server clusters deployed in ground data centers, whose computing power is abundant and not limited by power consumption or the space environment.

[0075] The normalized computing power value is calculated as follows: obtain the actual computing power value of the computing node, multiply it by the weight coefficient corresponding to the node type, and then divide it by the system's baseline peak computing power. The weight of space-based computing nodes is lower than that of ground-based core computing nodes. For example, the weight of a space-based computing node can be 0.85, the weight of a ground-based core computing node can be 1.0, and the weight of a ground-based edge computing node can be 0.9. By introducing node type weights, heterogeneous space-based and ground-based computing power can be mapped to the same dimension, achieving unified scheduling.

[0076] Then, the integrated space-ground central cloud network 100 allocates computing power to the remote sensing task based on the normalized computing power of each computing node. It selects a space-based satellite platform from at least one such platform and a ground computing node from at least one ground computing node 300. The principle of computing power allocation is to prioritize nodes with higher normalized computing power, while also considering the current load status of the nodes and the geographical location requirements of the task. For example, for a task requiring the acquisition of images of a specific latitude and longitude region, the system will prioritize a space-based satellite platform with a transit time matching that region and a higher normalized computing power; for tasks requiring high-precision processing, the system will prioritize ground core computing nodes with higher normalized computing power.

[0077] In one embodiment, the specific method by which the integrated space-ground cloud network 100 parses the mission parameters of a remote sensing mission and generates a processing link based on the mission parameters is as follows:

[0078] First, the timeliness level and accuracy requirements are extracted from the remote sensing mission. The timeliness level indicates the mission's required response speed, such as seconds, minutes, or hours; the accuracy requirements indicate the mission's required target positioning accuracy, such as sub-meter, meter, or ten-meter accuracy.

[0079] Then, the first weighting coefficient for optimizing processing latency is determined based on the timeliness level. The higher the latency requirement (i.e., the faster the response is required), the larger the value of the first weighting coefficient. For example, for emergency tasks requiring a response time within seconds, the first weighting coefficient can be set to 0.6.

[0080] The second weighting coefficient for optimizing computing power utilization is determined based on the accuracy requirements. Higher accuracy requirements necessitate the use of more ground-based computing resources, thus requiring a larger second weighting coefficient. For example, for fine observation tasks requiring sub-meter accuracy, a second weighting coefficient of 0.5 could be used.

[0081] The integrated space-ground cloud network 100 takes minimizing processing latency, maximizing computing power utilization, and minimizing execution cost as its three optimization objectives, assigning a first weight coefficient (corresponding to latency), a second weight coefficient (corresponding to computing power utilization), and a preset third weight coefficient (corresponding to cost) to each. The sum of the three weight coefficients is 1.

[0082] Based on the above three optimization objectives and their weight coefficients, the space-ground integrated central cloud network 100 selects the optimal combination from the candidate space-based satellite platforms and ground computing nodes 300 to generate processing links.

[0083] When remote sensing tasks are emergency tasks (such as earthquake disaster monitoring, forest fire early warning, and flood range assessment), timeliness is the most important indicator. Therefore, the first weight coefficient is set to be greater than the second and third weight coefficients, for example, α is 0.6, β is 0.2, and γ is 0.2, to ensure that the system prioritizes the combination of satellite and ground resources that can return results the fastest.

[0084] When remote sensing tasks are routine tasks (such as annual land surveys, agricultural planting area statistics, and urban expansion monitoring), computing power utilization is a more important indicator (to reduce costs). Therefore, the second weighting coefficient is set to be greater than the first and third weighting coefficients, for example, α is 0.2, β is 0.6, and γ is 0.2, to ensure that the system prioritizes the combination that can maximize the use of existing computing power resources.

[0085] In one embodiment, the space-based satellite platform generates valid data and transmits it according to the recipient's address in the following specific manner:

[0086] First, the space-based satellite platform performs preprocessing and target detection on the raw remote sensing data collected in orbit.

[0087] Preprocessing includes radiometric calibration, atmospheric correction, and noise removal. Radiometric calibration is the process of converting the raw digital quantized values ​​recorded by the sensor into physically meaningful radiance or reflectance values. Atmospheric correction is the process of eliminating the influence of atmospheric scattering and absorption on remote sensing signals, ensuring that the image accurately reflects surface information. Noise removal uses filtering algorithms to remove random noise, stripe noise, and other noise introduced by the sensor itself or during transmission. In addition, preprocessing may also include bad pixel repair and invalid data removal, such as removing image areas obscured by clouds or data segments with severe signal interference.

[0088] Target detection is the process of identifying and locating targets of interest on preprocessed data. Space-based satellite platforms carry lightweight deep learning models to process images in real time, extracting the contour features, texture features, or spectral features of targets, and marking the location (e.g., pixel coordinates or geographic coordinates) and category (e.g., "ships", "aircraft", "buildings", "fire spots").

[0089] After preprocessing and target detection, the space-based satellite platform generates valid data. Valid data includes target detection results (location and category information) and necessary auxiliary parameters (such as acquisition time and satellite orbit parameters). Meanwhile, the raw remote sensing data is retained on the space-based satellite platform and not transmitted.

[0090] Through the aforementioned on-orbit processing, the space-based satellite platform reduces the size of the effective data to less than 10% of the original remote sensing data. For example, when the original remote sensing data is a 200MB high-resolution optical image, the effective data generated after preprocessing and target detection may only be 15MB to 20MB, containing the location coordinates and category labels of all detected targets. This compression ratio significantly reduces the bandwidth pressure on the satellite-to-ground data transmission link and shortens the downlink time.

[0091] Finally, the space-based satellite platform transmits the valid data via the satellite-to-ground link according to the receiver address specified in the processing link (such as the network address of the ground computing node 300 or the gateway address of the space-ground integrated central cloud network 100).

[0092] In one embodiment, the ground computing node 300 performs ground refinement processing to obtain the refined target coordinates in the following specific manner:

[0093] First, the ground computing node 300 receives valid data from the space-based satellite platform. This valid data must contain at least the target's coordinates with an accuracy of at least ten meters. Ten-meter accuracy means that the target's position error obtained through on-board processing is within the range of 10 to 100 meters. Due to limitations in on-board computing resources and factors such as satellite attitude measurement errors and orbit determination errors, on-board processing cannot achieve higher accuracy.

[0094] Then, the ground computing node 300 acquires pre-stored ground control point data. Ground control points are ground features with precise geographic coordinates (typically at the centimeter or decimeter level), such as road intersections, building corners, and GPS survey markers. The ground control point data is pre-stored in the database of the ground computing node 300.

[0095] Finally, the ground computing node 300 uses a bundle adjustment algorithm or a georeferencing algorithm to match and calibrate the target coordinates in the effective data with the ground control point data.

[0096] Bundle adjustment is a classic algorithm in photogrammetry that optimizes the exterior orientation elements (position and attitude) of satellite imaging at a given time and the three-dimensional coordinates of ground points to ensure that all observation rays converge optimally at the ground points. Georegistration is the process of registering satellite imagery with ground control points of known coordinates. It maps image pixel coordinates to a geographic coordinate system by calculating affine or polynomial transformation parameters.

[0097] In one embodiment, the ground computing node 300 verifies the refined target coordinates and generates the final result in the following specific manner:

[0098] First, the ground computing node 300 pushes the refined target coordinates to the manual review console. The manual review console can include input / output devices such as a monitor, mouse, and keyboard, as well as a review software interface running on the terminal device. The review interface can display the refined target coordinates in a list format or overlay them onto the satellite base map for a visual presentation, facilitating intuitive judgment by reviewers.

[0099] Then, the three-level review process is executed sequentially on the manual review platform:

[0100] The first level of review is used to verify the accuracy of target identification. Reviewers determine whether the space-based target detection results are correct and whether there are any misidentifications (e.g., misidentifying clouds as targets or shadows as water bodies) or omissions (e.g., missing real targets). For misidentified targets, reviewers can remove them; for omissions, reviewers can manually add them.

[0101] The second level of audit is used to verify positioning accuracy and data integrity. Auditors check whether the refined target coordinates are within the allowable error range and whether the valid data is complete (e.g., whether there is data loss, missing fields, etc.). For targets with coordinate deviations exceeding the limit, auditors can manually correct their coordinates.

[0102] The third level of review is used to verify the compliance of the results. Reviewers confirm whether the format, content, and accuracy of the final results meet industry standards or task requirements. For example, for inspection reports submitted to the land and resources department, it is necessary to confirm whether the results comply with the provisions of the "Land Survey Database Standard".

[0103] If all three levels of review are passed, ground computing node 300 will output the refined target coordinates as the final result. The format of the final result can be specified by the task requirements, such as GIS vector data (Shapefile format), remote sensing image map (GeoTIFF format), PDF inspection report, Excel data report, or dynamic situation map.

[0104] If any level of review fails (e.g., the first level of review finds numerous misidentifications, the second level of review finds coordinate deviations far exceeding the threshold, or the third level of review finds the format does not meet the requirements), the ground computing node 300 will return the refined target coordinates for reprocessing. Reprocessing may include: re-executing the ground refinement algorithm, adjusting algorithm parameters, or re-performing target detection. Simultaneously, the review comments (e.g., "Target A is a misidentification, please remove it," "Target B's coordinate deviation is too large, please adjust the registration parameters") are synchronized to the integrated space-ground central cloud network 100. The central cloud network can optimize the scheduling strategy and processing links of subsequent tasks based on the review comments, such as adjusting the confidence threshold of the target detection model for that area, or increasing the density of ground control points in that area.

[0105] The three-level review and closed-loop feedback mechanism ensured the reliability of the final results and the system's ability to continuously optimize.

[0106] This application also provides a systematic method for rapid information service through space-ground collaboration, employing... Figure 1 The system implementation method includes: establishing a space-ground integrated central cloud network, and connecting at least one space-based satellite platform and at least one ground computing node as heterogeneous computing resources to the space-ground integrated central cloud network. The space-ground integrated central cloud network receives external remote sensing tasks, determines the space-based satellite platform and ground computing node to perform the remote sensing tasks, parses the task parameters of the remote sensing tasks, generates a space-ground collaborative processing link based on the task parameters, and distributes the processing link to the space-based satellite platform and ground computing node. The processing link includes at least the on-orbit processing tasks to be performed by the space-based satellite platform, the ground processing tasks to be performed by the ground computing node, and the address of the recipient for the valid data downloaded from the space-based satellite platform. The space-based satellite platform collects raw remote sensing data according to the instructions of the on-orbit processing tasks in the processing link, generates valid data based on the raw remote sensing data, and downloads the valid data according to the instructions of the recipient address in the processing link. The ground computing node receives the valid data according to the instructions of the ground processing tasks in the processing link, performs ground refinement processing on the valid data to obtain refined target coordinates, verifies the refined target coordinates, generates the final result, and distributes the final result according to the processing link.

[0107] Figure 2 A flowchart illustrating a satellite-ground collaborative rapid information service method according to an embodiment of this application is shown.

[0108] like Figure 2 As shown, the satellite-ground collaborative rapid information service method includes operations S1-S6.

[0109] In operation S1, an integrated space-ground central cloud network is built, and at least one space-based satellite platform and at least one ground computing node are connected to the integrated space-ground central cloud network as heterogeneous computing resources.

[0110] According to the embodiments of this application, the hardware and software deployment of the integrated space-ground central cloud network is divided into two parts: the ground-based end and the space-based end. The ground-based end builds a core cloud platform server cluster, configured with multi-core processors, large-capacity memory, and high-speed storage units, and is equipped with a distributed scheduling operating system and resource monitoring software. The space-based end deploys lightweight relay cloud nodes, adapted to the communication protocols of the satellite service system. A space-ground communication gateway is built, compatible with multiple types of data transmission links such as S-band and Ka-band, to achieve two-way communication interconnection between space and ground.

[0111] According to the embodiments of this application, the on-orbit computing modules, detection payloads, and communication modules of various satellite platforms, such as on-orbit optical remote sensing satellites, synthetic aperture radar (SAR) satellites, and hyperspectral satellites, are successively connected to the cloud network to complete unique identity (ID) registration and computing power parameter filing (including peak computing power, power consumption threshold, and processing latency). Simultaneously, multiple levels of ground nodes, including provincial ground big data centers, regional edge computing stations, and portable ground terminals, are connected to record parameters such as computing capacity, processing accuracy, and response speed. A dynamically updated global computing power resource ledger is established to achieve real-time visibility, manageability, and controllability of resource status.

[0112] According to the embodiments of this application, a task priority strategy (emergency tasks > real-time tasks > routine public service tasks), computing power allocation algorithm, fault switching threshold, and data transmission encryption rules can be built in. Differentiated scheduling windows are set for satellites in different orbits (low Earth orbit, medium Earth orbit, and high Earth orbit), and exclusive processing permissions are set for different ground nodes, completely breaking down the barriers between space and ground resources and realizing unified scheduling of computing power across the entire domain.

[0113] According to an embodiment of this application, for each connected computing node, a normalized computing power value is calculated based on its actual computing power value and node type weight. The calculation formula is as follows:

[0114]

[0115] In the formula: This is the normalized computing power value; The actual computing power of the node; This serves as the system's baseline peak computing power. The weights are determined by node type. Space-based computing nodes have lower weights than ground-based core computing nodes. Specifically, space-based nodes have a weight of 0.85, ground-based core nodes have a weight of 1.0, and ground-based edge nodes have a weight of 0.9. By introducing node type weights, heterogeneous space-based and ground-based computing power can be mapped to the same dimension, enabling unified scheduling.

[0116] During S2 operation, the integrated space-ground central cloud network receives external remote sensing missions, determines the space-based satellite platform and ground computing nodes that will execute the remote sensing missions, analyzes the mission parameters of the remote sensing missions, generates a space-ground collaborative processing link based on the mission parameters, and distributes the processing link to the space-based satellite platform and ground computing nodes.

[0117] According to the embodiments of this application, the integrated space-ground central cloud network receives external task instructions through dedicated API interfaces and dedicated network communication ports. Task sources include emergency command platforms, control systems, and public service platforms. Upon receipt, the intelligent task parsing module is immediately activated, employing natural language processing and rule matching algorithms to decompose the core task parameters: task type (emergency monitoring, target detection, routine patrol), timeliness level (seconds, minutes, hours), accuracy requirements (sub-meter, meter, ten-meter), target area latitude and longitude range, data type (optical imagery, SAR data, hyperspectral data), and output format requirements.

[0118] According to the embodiments of this application, computing power is allocated based on the normalized computing power value of each computing node. The cloud network scheduling module retrieves real-time computing power status data across the entire domain, including parameters such as satellite transit time, on-board computing power load rate, remaining power consumption, idle computing power of ground nodes, satellite-to-ground link bandwidth, and transmission latency. Based on the comparison results of the normalized computing power values, a space-based platform is selected from at least one space-based satellite platform to perform this mission, and a ground node is selected from at least one ground computing node to perform this mission.

[0119] According to the embodiments of this application, an improved genetic algorithm and a greedy algorithm are used to perform multi-objective optimization scheduling calculations, taking into account three major indicators: timeliness, cost, and computing load, and automatically generating the optimal task processing link for unified scheduling of satellite and ground computing resources. The multi-objective scheduling fitness function is as follows:

[0120]

[0121] In the formula: The fitness value (the higher the value, the better the solution); Estimate the latency for the entire process; This represents computing power utilization; Cost represents the task execution cost. , , Let be the weighting coefficient, satisfying .

[0122] The weighting coefficients are dynamically adjusted based on the task type. When the remote sensing task is an emergency task, timeliness is the most important factor, so α is set to 0.6, β to 0.2, and γ to 0.2. That is, the first weighting coefficient (for optimizing processing latency) is greater than the second weighting coefficient (for optimizing computing power utilization) and the third weighting coefficient (for optimizing execution cost). When the remote sensing task is a routine task, computing power utilization is more important, so α is set to 0.2, β to 0.6, and γ to 0.2. That is, the second weighting coefficient is greater than the first and third weighting coefficients.

[0123] A clear division of labor between the space and ground systems is established: the spacecraft is responsible for raw data acquisition, lightweight preprocessing, and initial target identification; the ground system is responsible for refined processing, manual review, and result generation. Simultaneously, the data transmission sequence, encryption method, verification rules, and fault compensation scheme are determined. Scheduling instruction packets are generated and sent to the corresponding space-based satellite platform and ground computing nodes, with the entire process time controlled within seconds. The processing link must include at least the following information: the on-orbit processing tasks to be performed by the space-based satellite platform (e.g., acquiring raw data of the target area, performing target detection), the ground processing tasks to be performed by the ground computing nodes (e.g., coordinate refinement, verification), and the recipient address of the valid data transmitted from the space-based satellite platform.

[0124] During S3 operation, the space-based satellite platform collects raw remote sensing data according to the instructions of the on-orbit processing tasks in the processing link, and generates valid data based on the raw remote sensing data.

[0125] According to the embodiments of this application, the space-based satellite platform receives scheduling command packets issued by the central cloud network through the space-to-ground communication module. After parsing the commands, the satellite service system coordinates the attitude control module, power supply module, and remote sensing payload to work together. The attitude control module initiates an attitude adjustment program, controlling the satellite's attitude through momentum wheels and magnetic torque devices to ensure that the remote sensing payload is precisely aligned with the target area, guaranteeing that the imaging tilt angle and scanning range meet mission requirements. The power supply module allocates the power of the solar panels and energy storage batteries, prioritizing power supply to the remote sensing payload and on-orbit computing module. Based on the command parameters, the remote sensing payload initiates optical imaging, SAR imaging, or hyperspectral detection modes to acquire raw remote sensing data of the target area, simultaneously recording auxiliary information such as acquisition time, satellite orbit parameters, attitude angle, and imaging resolution.

[0126] According to the embodiments of this application, the raw data is transmitted to the on-orbit computing module in real time. The module is equipped with a lightweight deep learning model (such as YOLOv5-tiny, MobileNet) and preprocessing algorithms to perform rapid on-orbit processing. First, data preprocessing is performed, including radiometric calibration, atmospheric correction, noise removal, and bad pixel repair of the raw image, removing invalid data such as cloud cover and signal interference. Then, feature extraction is performed, using a convolutional neural network to extract target contours, textures, and spectral features. Finally, initial target identification is performed, matching and classifying the extracted features to mark the location and category of suspected targets, completing the lightweight on-orbit processing.

[0127] After processing, only valid feature data, initial target identification results, and auxiliary parameters are retained, and the original massive data is compressed. The formula for calculating the on-orbit data compression ratio is as follows:

[0128]

[0129] In the formula: This refers to the data compression ratio. This refers to the original data volume; To ensure the effective data volume after on-orbit processing, this architecture controls the compression ratio to ≥90%, meaning the effective data volume is less than 10% of the original remote sensing data volume. The original remote sensing data is retained on the space-based satellite platform and is not downloaded. For example, when the original remote sensing data is a 200MB high-resolution optical image, the effective data generated after on-orbit processing is approximately 15MB to 20MB, containing the location coordinates (with ten-meter accuracy) and category labels of all detected targets.

[0130] During S4 operation, the space-based satellite platform transmits valid data according to the address of the receiver in the processing link, and the integrated space-ground central cloud network or ground computing node performs integrity verification on the received data.

[0131] According to the embodiments of this application, after the space-based satellite platform completes on-orbit processing, the space-to-ground communication module starts the high-speed data transmission mode, and uses LDPC channel coding and AES encryption algorithm to package and encrypt the processed data, the initial target identification results, and the collected auxiliary parameters, and transmits them to the designated receiver address in the processing link through a preset space-to-ground link (which can be the space-ground integrated central cloud network gateway or the ground computing node).

[0132] According to the embodiments of this application, after receiving data, the cloud network or ground computing node immediately initiates a dual integrity verification mechanism. The first step is to perform a CRC cyclic redundancy check, verifying the data frame checksum to determine if there is packet loss or corruption during transmission. The core generator polynomial (32-bit) for the CRC check is as follows:

[0133]

[0134] This polynomial is used to calculate the checksum, ensuring data transmission integrity with a checksum error rate of less than 10%. -9 The second step is to validate the data format, checking whether the data dimensions, parameter completeness, and identification information are compliant.

[0135] If the verification fails, the receiver immediately issues a retransmission command. The retransmission request only includes the identifier of the data packet that failed verification (e.g., the packet sequence number), rather than requesting the retransmission of all data. The space-based satellite platform quickly retransmits the corresponding data frame within the transit window, supporting breakpoint resumption. To address the short transit time of low-Earth orbit satellites, a data packet transmission and multi-link redundancy backup mechanism is adopted to ensure efficient and complete data delivery, with transmission latency controlled within seconds. If the verification passes, the cloud network accurately forwards the data to the designated ground computing node according to the scheduling link plan, while caching data copies for subsequent traceability and verification.

[0136] During operation S5, the ground computing node receives valid data according to the instructions of the ground processing task in the processing link, and performs ground refinement processing on the valid data to obtain the refined target coordinates.

[0137] According to the embodiments of this application, after receiving data forwarded by the cloud network, the ground computing node automatically decompresses and imports it into the high-precision processing system. The valid data includes the target coordinates obtained by the space-based satellite platform, with an accuracy of ten meters (i.e., an error within the range of 10 to 100 meters). Due to limitations in on-board computing resources and factors such as satellite attitude measurement errors and orbit determination errors, on-board processing cannot achieve higher precision.

[0138] According to the embodiments of this application, a layered refinement process is carried out based on the mission accuracy requirements. The first layer is target location refinement, which uses bundle adjustment and georegistration algorithms, combined with ground control point data, to calibrate and correct the coordinates of the initially identified targets on the satellite. Ground control points are ground features with precise geographic coordinates (usually reaching centimeter or decimeter accuracy), such as road intersections, building corners, GPS measurement markers, etc., which are pre-stored in the database of ground computing nodes. The target coordinate refinement and correction formula is as follows:

[0139]

[0140] In the formula: The refined geodetic coordinates; These are the original coordinates on the satellite; This refers to the control point deviation value; The first level is the georeferencing coefficient. Through the above calibration, the positioning accuracy is optimized from the ten-meter level on satellite to the sub-meter level. The second level is image and feature refinement, which involves sharpening and enhancing remote sensing images, mosaicking, removing shadows, and refining target features to distinguish real targets from interference. The third level is in-depth processing of results, which generates specialized results such as target trajectories, change monitoring maps, and area calculation data according to mission requirements.

[0141] After refinement, the system automatically pushes the data to the manual review console for multi-level review by professional technicians. The first level of review verifies the accuracy of target identification, eliminating misidentified and omitted targets and supplementing missing target information. The second level verifies positioning accuracy and data integrity, correcting coordinate deviations and formatting errors. The third level verifies the compliance of the results, ensuring they meet industry standards and task requirements. The review process supports annotation, commentary, and modification functions. Review comments are synchronized to the central cloud network in real time.

[0142] If all three levels of review pass, the refined target coordinates will be output as the approved result to the next stage. If any level of review fails, the refined target coordinates will be returned for reprocessing (e.g., re-executing the ground refinement algorithm, adjusting algorithm parameters, or re-performing target detection), and the review comments will be synchronized to the integrated space-ground central cloud network to form a closed-loop feedback mechanism.

[0143] In S6 operation, the ground computing node verifies the refined target coordinates to generate the final result, and distributes the final result according to the processing link.

[0144] Once approved, the ground computing node's results generation module is activated, integrating space-based on-orbit processing data, ground-based refinement results, and review comments to generate standardized final results in the user-specified format. User-specified formats may include, but are not limited to: GIS vector data (e.g., Shapefile format), remote sensing imagery (e.g., GeoTIFF format), PDF inspection reports, Excel data reports, dynamic situation maps, etc. Watermark information such as the result number and generation time is simultaneously overlaid.

[0145] After the results are generated, a distribution channel is selected based on the task level, and the transmitted data is encrypted and compressed. An identity authentication and access control mechanism is employed to push the data to authorized users and mobile terminals, supporting real-time download and online viewing. The formula for the encryption security factor of the results is as follows:

[0146] In the formula: To ensure encryption security; The length of the encryption key; This represents the transmission channel weight. A higher encryption security level indicates stronger data transmission security.

[0147] After distribution, the central cloud network archives the final results, processing logs, and scheduling records to the database, setting hierarchical storage permissions to support subsequent rapid querying, reuse, and traceability. This completes the closed loop of rapid processing and information services for the entire space-ground collaborative process.

[0148] 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 application. 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 a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may 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.

[0149] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A satellite-ground collaborative rapid information service system, characterized in that, include: A central cloud network integrating space and ground, at least one space-based satellite platform, and at least one ground computing node; The integrated space-ground central cloud network is communicatively connected to the space-based satellite platform and the ground computing node, respectively. The integrated space-ground central cloud network is used to receive external remote sensing tasks, determine the space-based satellite platform and ground computing nodes that will execute the remote sensing tasks, parse the task parameters of the remote sensing tasks, generate a space-ground collaborative processing link based on the task parameters, and send the processing link to the space-based satellite platform and the ground computing nodes. The processing link includes at least the on-orbit processing tasks that the space-based satellite platform needs to execute, the ground processing tasks that the ground computing nodes need to execute, and the address of the receiver for the valid data transmitted by the space-based satellite platform. The space-based satellite platform is used to collect raw remote sensing data according to the instructions of the on-orbit processing task in the processing link, generate valid data based on the raw remote sensing data, and download the valid data according to the instructions of the receiver address in the processing link. The ground computing node is used to receive the valid data according to the instructions of the ground processing task in the processing link, perform ground refinement processing on the valid data to obtain refined target coordinates, verify the refined target coordinates, generate the final result, and distribute the final result according to the processing link.

2. The space-ground collaborative rapid information service system as described in claim 1, characterized in that, The determination of the space-based satellite platform and ground computing nodes for performing the remote sensing mission includes: For each connected computing node, a normalized computing power value is calculated based on the actual computing power value of the computing node and the node type weight, wherein the weight of space-based computing nodes is lower than that of ground-based core computing nodes. The computing power of the remote sensing task is allocated based on the normalized computing power value of each computing node, thereby determining the space-based satellite platform and ground computing nodes that will execute the remote sensing task.

3. The system as described in claim 1, characterized in that, The process of parsing the mission parameters of the remote sensing mission and generating a satellite-ground collaborative processing link based on the mission parameters includes: The timeliness level and accuracy requirements are extracted from the remote sensing task. The first weighting coefficient for the processing delay optimization objective is determined based on the aforementioned timeliness level; Determine the second weighting coefficient of the computing power utilization optimization target based on the aforementioned accuracy requirements; The processing link is generated with three optimization objectives: minimizing processing latency, maximizing computing power utilization, and minimizing execution cost. The first weight coefficient, the second weight coefficient, and a preset third weight coefficient are assigned to each of these objectives respectively. Specifically, when the remote sensing task is an emergency task, the first weighting coefficient is greater than the second weighting coefficient and the third weighting coefficient; when the remote sensing task is a routine task, the second weighting coefficient is greater than the first weighting coefficient and the third weighting coefficient.

4. The space-ground collaborative rapid information service system as described in claim 1, characterized in that, The step of generating valid data based on the original remote sensing data and transmitting the valid data according to the recipient address indicated in the processing link includes: The raw remote sensing data is preprocessed and target detection is performed sequentially in orbit. The preprocessing includes radiometric calibration, atmospheric correction and noise removal. The target detection is used to mark the location and category of suspected targets. The effective data obtained after preprocessing and target detection is downloaded, while the original remote sensing data is retained on the space-based satellite platform and not downloaded. The volume of the effective data is less than 10% of the volume of the original remote sensing data.

5. The space-ground collaborative rapid information service system as described in claim 1, characterized in that, The refined target coordinates obtained by performing ground refinement processing include: Receive valid data from the space-based satellite platform, the valid data containing the target's coordinates with an accuracy of ten meters; Retrieve pre-stored ground control point data; Using bundle adjustment or georeferencing algorithms, the coordinates in the effective data are matched and calibrated with the ground control point data to output refined target coordinates. The accuracy of the calibrated target coordinates reaches the sub-meter level.

6. The space-ground collaborative rapid information service system as described in claim 5, characterized in that, The step of verifying and confirming the refined target coordinates to generate the final result includes: The refined target coordinates are then pushed to the manual review console. The manual review platform performs three levels of review in sequence: the first level of review is used to check the accuracy of target identification, the second level of review is used to check the positioning accuracy and data integrity, and the third level of review is used to check the compliance of the results. If all three levels of review are passed, the refined target coordinates will be output as the final result. If any level of review fails, the refined target coordinates will be returned for reprocessing, and the review comments will be synchronized to the integrated space-ground central cloud network.

7. A method for providing rapid information services via satellite-ground collaboration using the system described in any one of claims 1 to 6, characterized in that, include: A space-ground integrated central cloud network is established, and at least one space-based satellite platform and at least one ground computing node are connected to the space-ground integrated central cloud network as heterogeneous computing resources. The system receives external remote sensing tasks through the integrated space-ground central cloud network, determines the space-based satellite platform and ground computing node to execute the remote sensing tasks, parses the task parameters of the remote sensing tasks, generates a space-ground collaborative processing link based on the task parameters, and sends the processing link to the space-based satellite platform and the ground computing node. The processing link includes at least the on-orbit processing tasks to be executed by the space-based satellite platform, the ground processing tasks to be executed by the ground computing node, and the address of the receiver for the valid data transmitted by the space-based satellite platform. The space-based satellite platform collects raw remote sensing data according to the on-orbit processing task instructions in the processing link, generates valid data based on the raw remote sensing data, and downloads the valid data according to the receiver address instructions in the processing link. The ground computing node receives the valid data according to the instructions of the ground processing task in the processing link, performs ground refinement processing on the valid data to obtain refined target coordinates, verifies the refined target coordinates, generates the final result, and distributes the final result according to the processing link.

8. The method as described in claim 7, characterized in that, The determination of the space-based satellite platform and ground computing nodes for performing the remote sensing mission includes: For each connected computing node, a normalized computing power value is calculated based on the actual computing power value of the computing node and the node type weight, wherein the weight of space-based computing nodes is lower than that of ground-based core computing nodes. The computing power of the remote sensing task is allocated based on the normalized computing power value of each computing node, thereby determining the space-based satellite platform and ground computing nodes that will execute the remote sensing task.

9. The method as described in claim 7, characterized in that, The process of parsing the mission parameters of the remote sensing mission and generating a satellite-ground collaborative processing link based on the mission parameters includes: The timeliness level and accuracy requirements are extracted from the remote sensing task. The first weighting coefficient for the processing delay optimization objective is determined based on the aforementioned timeliness level; Determine the second weighting coefficient of the computing power utilization optimization target based on the aforementioned accuracy requirements; The processing link is generated with three optimization objectives: minimizing processing latency, maximizing computing power utilization, and minimizing execution cost. The first weight coefficient, the second weight coefficient, and a preset third weight coefficient are assigned to each of these objectives respectively. Specifically, when the remote sensing task is an emergency task, the first weighting coefficient is greater than the second weighting coefficient and the third weighting coefficient; when the remote sensing task is a routine task, the second weighting coefficient is greater than the first weighting coefficient and the third weighting coefficient.

10. The method as described in claim 7, characterized in that, The step of generating valid data based on the original remote sensing data and transmitting the valid data according to the recipient address indicated in the processing link includes: The raw remote sensing data is preprocessed and target detection is performed sequentially in orbit. The preprocessing includes radiometric calibration, atmospheric correction and noise removal. The target detection is used to mark the location and category of suspected targets. The effective data obtained after preprocessing and target detection is downloaded, while the original remote sensing data is retained on the space-based satellite platform and not downloaded. The volume of the effective data is less than 10% of the volume of the original remote sensing data.

Images