Multi-modal SAR satellite data-based cross-industry application optimization method and system

Abstract

The invention discloses a multi-modal SAR satellite data-based cross-industry application optimization method and system, and relates to the technical field of satellite remote sensing. The method comprises the steps of performing unified processing on multi-modal SAR satellite data, including radiation, geometry, time correction and noise suppression; carrying out regional seamless splicing on the data after unified processing; the data quality is automatically optimized according to different industry requirements; according to the industry customization module, carrying out sub-industry application; and performing intelligent scheduling and resource optimization on each industry task. The system comprises a data receiving and preprocessing module, an industry customization application module, an intelligent scheduling module, a resource optimization module and the like, and a satellite-ground collaborative architecture is adopted. According to the method, the comprehensive utilization efficiency of multi-modal SAR data is improved, cross-industry accurate and efficient service is realized, the satellite resource scheduling and system adaptive capability is optimized, and the method is suitable for multiple fields of disaster monitoring, agriculture, national defense and the like.

Description

Technical Field

[0001] This invention relates to the field of cross-industry application technology of SAR satellite data, and in particular to an optimization method and system for cross-industry application of multimodal SAR satellite data. Background Technology

[0002] With the acceleration of global digital transformation, the demand for timely and high-precision Earth observation data is exploding in fields such as disaster management, environmental monitoring, national defense and security, and smart agriculture. Traditional single-mode synthetic aperture radar (SAR) satellite systems are limited by fixed observation modes and rigid resource scheduling, making it difficult to meet the differentiated and dynamic application needs of multiple industries. Research on the cross-industry application performance evaluation and optimization system of AIRSAT constellation multimodal SAR satellites, by constructing an intelligent satellite resource scheduling and performance optimization system, can not only significantly improve the comprehensive service capabilities of remote sensing satellite systems, but also have a profound impact on national security, economic development, and social governance, possessing significant scientific value and practical significance.

[0003] Current multimodal SAR satellite systems suffer from several key technical shortcomings in cross-industry applications: First, at the system architecture level, traditional satellite constellations lack intelligent coordination mechanisms, with fixed task allocation among satellite nodes, making it difficult to achieve optimal allocation of observation resources and resulting in low overall system efficiency. Second, at the data processing level, existing technologies mostly adopt static task planning models, failing to dynamically adjust observation strategies based on real-time needs, leading to insufficient matching between observation capabilities and industry application requirements. Third, in terms of performance evaluation, the lack of a unified cross-industry evaluation index system makes it difficult to quantify system performance under different application scenarios, hindering the scientific nature of resource optimization decisions. Furthermore, existing systems generally suffer from weak satellite-ground coordination capabilities, strong ground dependence, and insufficient real-time processing capabilities, resulting in significant delays from data acquisition to application services. Regarding intelligence, traditional systems lack adaptive learning capabilities and cannot continuously optimize task scheduling strategies through historical data accumulation. These technical shortcomings severely limit the service efficiency and responsiveness of multimodal SAR satellite systems in dealing with complex and ever-changing cross-industry application needs.

[0004] For example, invention application No. 202410257805.8 discloses a method for early warning of secondary disasters based on ground-based SAR monitoring deformation data. This solution can quickly and comprehensively identify the development stage of secondary disasters and their deformation evolution trend. However, this solution also lacks an intelligent collaborative mechanism and cannot dynamically adjust the observation strategy according to real-time needs.

[0005] Therefore, a novel cross-industry application optimization method and system based on multimodal SAR satellite data is needed to address the many existing technical shortcomings. Summary of the Invention

[0006] To address the aforementioned problems, the present invention aims to provide a method and system for optimizing cross-industry applications based on multimodal SAR satellite data. This method and system resolve the technical deficiencies of multimodal SAR satellite systems in terms of system architecture, data processing, performance evaluation, satellite-ground collaboration, and intelligence level, thereby improving the service efficiency and responsiveness of the system in cross-industry applications.

[0007] This invention provides a method and system for optimizing cross-industry applications based on multimodal SAR satellite data.

[0008] First aspect: A method for optimizing cross-industry applications based on multimodal SAR satellite data, including:

[0009] S1. Perform unified data processing on multimodal SAR satellite data;

[0010] S2. Perform seamless regional data stitching processing on unified multimodal SAR satellite data;

[0011] S3. Automatically optimize the data quality of seamlessly stitched multimodal SAR satellite data to meet the needs of different industries;

[0012] S4. Based on industry-customized application modules, call automatically optimized multimodal SAR satellite data for industry-specific task applications;

[0013] S5. Performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

[0014] In one embodiment of the present invention, the data unification processing of multimodal SAR satellite data in S1 includes:

[0015] A unified data receiving and preprocessing center will be established to perform radiometric correction, geometric correction, time correction, and noise suppression on Ku / X / C / L band SAR satellite data.

[0016] Radiation correction includes: sensor gain correction, atmospheric attenuation correction, and surface roughness correction;

[0017] Geometric correction includes: basic geometric correction and terrain correction;

[0018] Time correction includes: timestamp standardization and timing consistency optimization;

[0019] Noise suppression includes eliminating speckle noise and impulse interference in SAR data of different frequency bands, thereby improving the data signal-to-noise ratio.

[0020] In one embodiment of the present invention, the seamless stitching process for regional data in step S2 includes:

[0021] A deep learning-based spatiotemporal registration algorithm seamlessly stitches together image data of the same area taken by different satellites at different times, including the following steps:

[0022] S11. Perform pre-processing before splicing;

[0023] S12, Perform AI automatic matching to extract stable feature points;

[0024] S13. Align image data of the same area based on effective stable feature points;

[0025] S14. For the aligned regions, perform radiation consistency optimization and overlap fusion strategies to achieve seamless splicing.

[0026] In one embodiment of the present invention, the industry-specific applications include:

[0027] Disaster monitoring and emergency response, marine and polar environmental monitoring, surface deformation and infrastructure monitoring, agricultural and ecological environment monitoring, national defense and military support, and urban planning.

[0028] In one embodiment of the present invention, the automatic data quality optimization in step S3 to adapt to the needs of different industries includes:

[0029] Disaster monitoring and emergency response: Enhance the ability to monitor minute surface deformations and improve the accuracy of landslide and earthquake early warning;

[0030] Marine and polar environment monitoring: Optimize wave and ice floe identification, and enhance ship and iceberg tracking capabilities;

[0031] Agricultural and ecological environment monitoring: Enhance soil moisture and crop growth status analysis to improve the accuracy of yield prediction.

[0032] In one embodiment of the present invention, step S5 involves intelligent task scheduling and resource optimization for various industry application tasks, automatically selecting the best SAR satellite data, including:

[0033] Dynamically adjust task priorities to optimize satellite observation plans and resource allocation;

[0034] An adaptive optimal observation strategy intelligently recommends the optimal observation parameters based on the task.

[0035] The second aspect: A cross-industry application optimization system based on multimodal SAR satellite data, including:

[0036] The data receiving and preprocessing module collects multimodal SAR satellite data and performs data unification, seamless regional data stitching, and automatic data quality optimization.

[0037] Industry-specific customized application modules allow for the use of automatically optimized multimodal SAR satellite data for industry-specific tasks.

[0038] The intelligent scheduling and resource optimization module performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

[0039] In one embodiment of the present invention, the system further includes:

[0040] The performance feedback module iteratively optimizes the system algorithm based on user feedback data;

[0041] The industry database update module is used to customize and update industry applications, keeping them up-to-date.

[0042] In one embodiment of the present invention, the system further includes:

[0043] The visualization and analysis module provides interactive SAR satellite data browsing and decision support.

[0044] In one embodiment of the present invention, the system adopts a satellite-ground collaborative architecture, wherein a lightweight preprocessing model is deployed on the satellite to achieve on-orbit data screening, and a high-performance processing center is built on the ground station to complete complex analysis algorithms.

[0045] Third aspect: An electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of, for example, the method provided in the first aspect.

[0046] Fourth aspect: A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of, for example, the method provided in the first aspect.

[0047] The beneficial effects of this invention are:

[0048] 1. This invention establishes a unified data receiving and preprocessing center to achieve standardized processing of Ku / X / C / L multi-band data. By combining seamless stitching technology and intelligent enhancement algorithms, it solves the problems of redundancy, conflict, and insufficient industry adaptability in traditional data processing, significantly improving the information extraction capability of data in complex scenarios, providing a high-quality data foundation for cross-industry applications, and enhancing the comprehensive utilization efficiency of multimodal SAR satellite data.

[0049] 2. This invention relies on six industry-customized application modules to provide targeted and optimized data services for the differentiated needs of fields such as disaster monitoring, agricultural production, and national defense security. It changes the fragmented situation of traditional system applications, greatly improves the execution efficiency and scientific decision-making of tasks in various industries, and realizes precise and efficient cross-industry application services.

[0050] 3. This invention achieves optimal allocation of satellite resources through dynamic priority adjustment and adaptive observation strategies; combined with a space-ground collaborative architecture and performance feedback closed loop, the system can continuously iterate algorithms based on user feedback, enhancing its responsiveness to complex and ever-changing needs, upgrading the constellation system from a simple data acquisition platform to an intelligent service hub, and optimizing satellite resource scheduling and system self-adaptation capabilities. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating the cross-industry application optimization method of the present invention;

[0052] Figure 2 This is a schematic diagram of the cross-industry application optimization system of the present invention;

[0053] Figure 3 This is a sample image of 0.5mX band SAR data from the AIRSAT-08 satellite, as described in an embodiment of the present invention.

[0054] Figure 4 This is a sample image of 0.5mX band SAR data from the AIRSAT-08 satellite, as described in an embodiment of the present invention.

[0055] Figure 5 This is a sample image of 0.5mKu-band SAR data from the AIRSAT-01 satellite, as described in an embodiment of the present invention.

[0056] Figure 6 This is a sample image of 0.5mKu-band SAR data from the AIRSAT-01 satellite, as described in an embodiment of the present invention.

[0057] Figure 7 This is a sample image of 1mX band SAR data from the AIRSAT-08 satellite, as described in an embodiment of the present invention.

[0058] Figure 8 This is a sample image of 1mX band SAR data from the AIRSAT-08 satellite, as described in an embodiment of the present invention.

[0059] Figure 9 This is a sample image of 2mKu-band SAR data from the AIRSAT-01 satellite, as described in an embodiment of the present invention.

[0060] Figure 10 This is a sample image of 2mKu-band SAR data from the AIRSAT-01 satellite, as described in an embodiment of the present invention.

[0061] Figure 11This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0062] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar symbols denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0063] For ease of understanding, the AIRSAT constellation multimodal architecture of this invention will be introduced. Multimodal refers to the fact that this satellite system can support multiple different observation modes, including:

[0064] Multi-band: Satellites can use radar waves of different frequencies for observation. For example, Ku-band, X-band, C-band, and L-band; radar waves of different frequency bands have different physical characteristics.

[0065] Ku and X bands: High resolution, suitable for fine target identification, such as identifying the outlines of aircraft, ships, and buildings; C and L bands have medium resolution and are often used for marine monitoring. They also have strong penetrating power and are suitable for monitoring surface deformation under vegetation cover, such as soil moisture and geological subsidence.

[0066] Multi-resolution satellites can provide images with different levels of clarity, such as 0.5 meters, 1 meter, 2 meters, etc. High resolution (e.g., 0.5 meters) can make out finer objects, while lower resolution may be used for large-scale surveys.

[0067] Multipolarization: refers to the direction of electromagnetic wave vibration when radar waves are transmitted and received. Different polarization modes (such as HH, VV, HV, VH) have different sensitivities to the scattering characteristics of ground objects. By combining and analyzing them, richer information about ground objects can be obtained, such as distinguishing different types of vegetation or surface materials.

[0068] like Figures 3 to 6 As shown, the 0.5m SAR data from the AIRSAT constellation can identify oil slicks and clearly distinguish the outlines of aircraft, ships, and buildings. It can clearly identify oil slicks in the Indian Ocean and aircraft and specific buildings at Salt Lake City International Airport.

[0069] like Figure 7 and Figure 8 As shown, the 1m resolution data from the AIRSAT constellation SAR satellites can clearly identify patchy farmland, ships sailing in the ocean, and airport runways.

[0070] like Figure 9 and Figure 10As shown, the 2m resolution data from the AIRSAT constellation SAR satellites can clearly identify different types of transport aircraft, hangars, and sheds at naval bases, as well as different types of ships and drifting sea ice in the Antarctic Ocean.

[0071] Due to their different multi-band, multi-resolution, and multi-polarization methods, current multi-modal SAR satellite systems have shortcomings in cross-industry applications: First, in terms of system architecture, it is difficult to achieve optimal allocation of observation resources, resulting in low system efficiency; second, in terms of data processing, it is impossible to adjust observation strategies according to real-time needs, leading to insufficient matching between observation capabilities and industry requirements; third, in terms of efficiency evaluation, there is a lack of a unified cross-industry evaluation index system, making it difficult to quantify system performance and restricting the scientific nature of resource optimization decisions. These shortcomings severely limit the system's service efficiency and responsiveness to cross-industry demands.

[0072] Based on this, the present invention discloses a method and system for cross-industry application optimization based on multimodal SAR satellite data.

[0073] Example 1:

[0074] This embodiment discloses an optimization method for cross-industry applications based on multimodal SAR satellite data, for example... Figure 1 As shown, the method includes the following steps:

[0075] S1. Perform unified data processing on multimodal SAR satellite data.

[0076] This includes establishing a unified data receiving and preprocessing center to perform radiometric correction, geometric correction, time correction, and noise suppression on Ku / X / C / L band SAR satellite data.

[0077] To achieve standardized reception and preprocessing of Ku, X, C, and L multi-band SAR data, a unified data reception and preprocessing center architecture is established, forming a central system integrating hardware, software, and protocols. Specifically, this includes:

[0078] Deploy multi-band compatible receiving equipment and configure signal receiving terminals that support Ku (12-18GHz), X (8-12GHz), C (4-8GHz), and L (1-2GHz) bands to ensure that data from different satellite payloads (such as the Ku band of AIRSAT-01 satellite and the X band of AIRSAT-08 satellite) can be accessed in real time or near real time.

[0079] Deploy a high-performance computing cluster equipped with GPU / CPU hybrid computing nodes to meet the needs of multi-band data parallel processing, and also equip it with a redundant storage array to support long-term archiving of raw and processed data.

[0080] Develop a unified data access interface that is compatible with various satellite data transmission protocols and automatically parses the frame structure, checksum, and metadata of satellite downlink data, such as satellite number, shooting time, imaging mode, and resolution.

[0081] A preprocessing workflow engine is built, employing a modular design and integrating core algorithm modules such as radiometric correction, geometric correction, temporal correction, and noise suppression. It supports automatic invocation of adaptive parameters based on frequency band characteristics. Furthermore, an AI-powered automatic matching module seamlessly stitches together data from the same area captured by different satellites, avoiding duplication or omissions. An intelligent data enhancement module is also included, which automatically optimizes data quality according to the needs of different industries.

[0082] Establish a data format conversion standard to uniformly convert the raw data into the system's self-developed multimodal SAR general format. This format includes data volume, metadata, and index information, which facilitates cross-frequency band data retrieval and comparison.

[0083] Based on core algorithm modules such as radiometric correction, geometric correction, time correction, and noise suppression, the acquired multimodal SAR satellite data undergoes radiometric correction, geometric correction, time correction, and noise suppression processing. Specifically, this includes:

[0084] The radiometric correction target mainly eliminates the radiometric deviations in SAR data of different frequency bands caused by factors such as sensor gain, atmospheric attenuation, and surface roughness, so as to make the data reflectivity physically consistent.

[0085] For the Ku / X band, the correction methods used include sensor gain correction and atmospheric attenuation correction.

[0086] Among them, sensor gain correction, based on satellite payload calibration parameters, eliminates gain drift caused by instrument aging, and the formula is:

[0087]

[0088] in, The compensation coefficient is calculated in real time from the calibration curve provided by the satellite manufacturer. The initial reflectivity, To correct the reflectivity.

[0089] Atmospheric attenuation correction uses synchronous atmospheric sounding data, such as air pressure and humidity, to calculate the propagation loss of electromagnetic waves in the atmosphere. The Ku band is greatly affected by water vapor, so a cloud cover correction factor is introduced.

[0090] For the C / L band, the correction methods used include surface roughness correction.

[0091] In surface roughness correction, the L-band has strong penetration into vegetated areas, and radiation distortion caused by surface undulations needs to be eliminated through a scattering model.

[0092] For the C-band, the differences in radiation characteristics between water and land are distinguished. Water surfaces have low reflectivity, so specular reflection correction is enhanced, and incident angle normalization is performed.

[0093] To address the differences in radiation caused by different satellite observation angles, the data were normalized to reflectance values ​​at an incident angle of 30°, using the following formula:

[0094]

[0095] in, The initial reflectivity, To correct reflectivity, This is the actual observed incident angle.

[0096] Geometric correction unifies the spatial coordinates of data from different frequency bands to the geodetic coordinate system, eliminating spatial offsets caused by satellite orbit errors, Earth curvature, and terrain undulations, and ensuring the spatial alignment accuracy of data in the same area. For example, the alignment error of 0.5m resolution data is ≤0.5m.

[0097] Geometric correction includes basic geometric correction and terrain correction.

[0098] The basic geometric correction includes orbit parameter optimization: using satellite ephemeris data and ground control points, an accurate orbit model is fitted using the least squares method to correct orbital offsets. It also includes projection transformation, converting the original slant range projection data to Gauss-Kruger projection to ensure consistency of plane coordinates (x,y).

[0099] Terrain correction targets complex terrain areas such as mountainous and polar regions. It introduces altimetry data to construct a digital elevation model and performs terrain slope correction on the Ku / X band, which is susceptible to terrain shading.

[0100] For the L-band, which has strong penetrating power and is sensitive to shallow surface topography, surface roughness is additionally introduced to correct the geometric distortion of vegetation cover areas.

[0101] Time correction unifies the timestamps of observation data from different satellites and at different times to UTC time, eliminating time deviations caused by satellite clock differences and data transmission delays, and ensuring the time accuracy of time series analysis.

[0102] Time correction includes timestamp standardization and timing consistency optimization.

[0103] Timestamp standardization includes raw data time parsing, extracting the onboard clock time from the satellite downlink data, combining it with the local clock of the ground receiving station, calculating the transmission delay, and correcting it to the actual observed UTC time. It also includes cross-satellite time alignment, which uses a master satellite anchoring method to address the clock differences of different satellites in the AIRSAT constellation (e.g., satellites 01 and 08). For example, using the highly stable onboard clock of satellite 08 as a reference, clock difference models of other satellites are calculated using historical data, and timestamps are corrected in real time.

[0104] Temporal consistency optimization involves not only unifying the UTC time for multiple periods of data from the same region, but also labeling the solar altitude angle to facilitate the elimination of the impact of time-dimension differences in illumination during subsequent analysis.

[0105] Noise suppression is the process of eliminating speckle noise and impulse interference in SAR data of different frequency bands, improving the signal-to-noise ratio (SNR) of the data, and ensuring the identification accuracy of key targets, such as ships and farmland boundaries.

[0106] For the Ku / X band, the main approach is to eliminate speckle noise. By employing adaptive nonlocal mean filtering, the filter window size is adaptively adjusted based on the statistical characteristics of similar pixels. This suppresses noise while preserving edge details, resulting in an SNR improvement of ≥10dB.

[0107] For the C / L bands, the primary focus is on eliminating interference noise. The L band, with its strong penetrating power, is susceptible to vegetation scattering noise. A threshold denoising technique based on wavelet transform can be used to decompose the data into different frequency bands. Soft threshold suppression is applied to the noise-dominated high-frequency bands, while preserving surface structure information in the low-frequency bands. The C band is susceptible to atmospheric noise; therefore, multi-view processing technology is introduced. This involves coherently averaging multiple images of the same area from different observation angles to reduce noise variance.

[0108] S2. Perform seamless regional data stitching processing on unified multimodal SAR satellite data.

[0109] First, preprocessing is performed before stitching. The consistency of the Ku / X / C / L multi-band SAR data that has already undergone unified processing (radiometric, geometric, time correction and noise suppression) is further enhanced before stitching to lay the foundation for seamless stitching. Preprocessing before stitching includes resolution normalization and regional range definition and overlapping area extraction.

[0110] Resolution normalization addresses the differences in resolution between various satellite data (e.g., AIRSAT-08 satellite's 0.5m spotlight mode and 1m strip mode, AIRSAT-01 satellite's 2m strip mode), requiring unification to a target resolution. For example, 0.5m can be dynamically selected based on the application scenario for disaster monitoring / military support, while 1m can be used for agricultural / ecological monitoring. For low-resolution data, super-resolution reconstruction algorithms can be used, such as those based on Generative Adversarial Networks (GANs), to upscale to the target resolution; for high-resolution data, Gaussian downsampling can be used to achieve the target resolution, avoiding information redundancy.

[0111] like Figure 3 As shown, Hefei High-tech Zone, based on the latitude and longitude coordinates in the metadata, performs regional boundary definition and overlap area extraction, such as... Figure 4 The boundary coordinates of areas such as Pearl Harbor Naval Base are shown. A Geographic Information System (GIS) is used to delineate the minimum bounding rectangle of the area to be stitched, thus defining the stitching range. Overlapping areas between different satellite data are automatically identified. A spatial indexing algorithm is used to compare the latitude and longitude ranges of the multi-source data, calculate the overlap rate, and filter out valid data pairs with high overlap rates while discarding invalid data with low overlap rates.

[0112] Secondly, AI-powered automatic matching technology is employed for cross-satellite data spatial alignment. Deep learning-driven feature matching can achieve precise spatial alignment of data from different satellites (e.g., AIRSAT-01 and 08) and different frequency bands (e.g., Ku and X-band). The steps include:

[0113] First, stable feature points are extracted. For model selection, an improved convolutional neural network (CNN) and a Transformer hybrid model can be used to optimize feature extraction capabilities for speckle noise and strong scattering characteristics of SAR data.

[0114] The model prioritizes the extraction of stable feature points, which include cross-frequency band invariant artificial and natural features. Artificial features include building corners, bridge edges, and power transmission towers; natural features include ridgelines, river bends, and lake boundaries. The extraction density of stable feature points can be ≥500 points per square kilometer to ensure sufficient matching anchor points even in sparse areas.

[0115] Then, for stable feature points, feature matching and mismatch removal are performed. First, the K-Nearest Neighbor (KNN) matching algorithm combined with cosine similarity calculation can be used to initially screen similar feature point pairs. Then, the random sampling consensus algorithm is used to remove mismatches, such as false matches caused by shadows or noise, and retain valid matching pairs.

[0116] Based on effective stable feature points, the spatial transformation matrix is ​​solved by the least squares method to achieve accurate alignment of data coordinates and ensure alignment error and alignment rate.

[0117] Secondly, for the aligned overlapping areas, a seamless visual and physical interface is achieved through radiation consistency optimization and overlap fusion strategies.

[0118] Radiometric normalization based on feature points in the overlapping area addresses the potential deviation in radiometric values ​​in the overlapping area caused by differences in the gain of different satellite sensors and observation angles. It calculates the ratio of the mean radiometric values ​​of the same feature points in the overlapping area to generate a radiometric correction coefficient, and then linearly stretches the low-radiation data to keep the deviation at a low level.

[0119] Overlapping area fusion strategies can employ a weighted average fusion method, assigning weights to overlapping area pixels based on their distance from the stitching boundary. For example, the closer a pixel is to the boundary of data A, the higher its weight.

[0120] Secondly, the overlapping areas after alignment are checked repeatedly to ensure the integrity of full-area coverage.

[0121] By constructing a coverage heatmap, the number of times each pixel in the stitched area is covered can be counted in units of the target resolution pixels, and overlapping areas can be marked.

[0122] S3 automatically optimizes the data quality of seamlessly stitched multimodal SAR satellite data to meet the needs of different industries.

[0123] A demand-driven intelligent optimization framework is constructed for seamlessly stitched multimodal SAR data. This framework establishes a closed-loop system encompassing demand analysis, strategy matching, dynamic optimization, and quality feedback, executing automatic data quality optimization strategies tailored to specific industries. The industry-demand-driven intelligent optimization framework includes modules for industry demand analysis, optimization strategy libraries, dynamic optimization engines, and quality assessment.

[0124] The system comprises several modules: an industry requirements analysis module, a dynamic optimization engine module, and a quality assessment module. The former receives requirements from industry applications such as disaster monitoring and marine monitoring, and extracts key optimization objectives. The latter stores customized algorithms for six major industries and associates algorithm parameters. The former invokes algorithms from the strategy library based on the analyzed requirements, adaptively processing the spliced ​​data and supporting parallel / serial combinations of multiple algorithms. The latter outputs quantitative metrics, such as signal-to-noise ratio and target recognition rate, which are fed back to the engine for iterative parameter optimization.

[0125] Industry-specific automatic data quality optimization strategies include:

[0126] Disaster monitoring is optimized to enhance the detection capability of minute surface deformations, highlight millimeter-level surface deformation signals, suppress noise and atmospheric interference, and improve the spatiotemporal accuracy of landslide and earthquake early warnings. Optimization strategies include phase noise suppression and deformation signal enhancement.

[0127] Phase noise suppression employs a combination algorithm of multi-view processing and adaptive filtering: reducing speckle noise in the stitched SAR satellite data; and reducing the phase standard deviation by dynamically adjusting the filtering window.

[0128] Deformation signal enhancement is applied to landslide / earthquake areas. It uses small baseline set technology to perform differential processing on time-series stitched data to extract the cumulative deformation trend. Low-frequency deformation signals and high-frequency noise are decomposed by wavelet transform to enhance linear deformation features, such as the direction of landslide movement.

[0129] Ocean monitoring optimizes wave and ice floe identification and target tracking, improves wave texture clarity and ice floe edge recognition, enhances the contrast between ships / icebergs and the sea surface background, and ensures continuous tracking of dynamic targets.

[0130] The optimization strategies include enhancing wave and ice floe features and improving ship and iceberg tracking.

[0131] Among them, the enhancement of wave and ice features is achieved by extracting wave textures, using a two-dimensional Fourier transform to decompose the wave spectrum of the stitched ocean data, retaining the effective wavelength wave signals, and reconstructing the texture through inverse transform.

[0132] Ship and iceberg tracking enhancement improves target contrast. It can suppress sea clutter, perform adaptive threshold segmentation on strongly scattering targets such as ships and icebergs, eliminate polarization interference, and improve the contrast between the target and the background.

[0133] Agricultural monitoring enhances the analysis of soil moisture and crop growth status, improves the accuracy of soil moisture inversion, and strengthens the distinguishability of differences in crop growth.

[0134] Optimization strategies include optimizing soil moisture signals and enhancing crop growth status.

[0135] Among them, soil moisture signal optimization is aimed at bare soil / low vegetation areas, separating soil backscattering and vegetation scattering: extracting soil dielectric constant based on L-band data (strong penetration), and inverting soil volumetric water content by combining C-band data (sensitive to humidity); reducing moisture inversion error through surface roughness correction.

[0136] Enhanced crop growth status involves calculating polarization vegetation index and radar vegetation index from spliced ​​X / C band data, extracting crop growth trends through time-series analysis, and using a random forest model to fuse multiple index features to distinguish crop types and growth stages, thereby improving classification accuracy.

[0137] Defense and military support optimization combines high-resolution band data with super-resolution reconstruction to enhance target details, such as ship deck equipment and aircraft tail fins.

[0138] Urban planning optimization involves stitching together data from urban areas and using object-oriented classification algorithms to enhance the distinction between building areas, green spaces, and roads, improve the completeness of building outline extraction, and support urban expansion analysis.

[0139] For the aforementioned industry-specific automatic data quality optimization strategies, intelligent decision-making and dynamic iteration mechanisms can be employed.

[0140] By pre-setting industry demand tags, after receiving industry application requests, the system calls corresponding optimization strategies through keyword matching. For example, it can automatically trigger deformation enhancement and edge detection optimization in disaster emergencies. Moreover, it supports multiple strategy combinations for complex requirements.

[0141] While invoking the corresponding optimization strategy, the parameters are adaptively adjusted. Based on the scene characteristics of the spliced ​​data, the optimization engine automatically adjusts the algorithm parameters, and a quality report is generated synchronously when the optimized data is output.

[0142] S4. Based on industry-customized application modules, call automatically optimized multimodal SAR satellite data for industry-specific task applications.

[0143] Industry-customized application modules include: disaster monitoring and emergency response, marine and polar environment monitoring, surface deformation and infrastructure monitoring, agricultural and ecological environment monitoring, national defense and military support, and urban planning.

[0144] Disaster monitoring and emergency response include rapid disaster assessment, which uses multimodal SAR satellite data to quickly image and analyze affected areas, accurately determining the scope and severity of the disaster and providing a scientific basis for the allocation of relief resources. It can also monitor the development of disasters in real time, such as the direction of flood spread and the dynamics of landslides, to allow for timely adjustments to emergency response plans. For example, after an earthquake or flood, it automatically compares pre- and post-disaster data and generates a damage map within 10 minutes.

[0145] Marine and polar environmental monitoring, including illegal fishing monitoring, utilizes the high-resolution imaging capabilities of multimodal SAR satellite data to clearly identify vessel activity in the ocean and accurately determine whether illegal fishing activities exist. Satellites can continuously monitor large areas of the sea, promptly detecting abnormal activity patterns of suspicious vessels, such as frequent stops or operations in prohibited fishing areas. For example, suspicious vessels can be automatically identified and compared with AIS (Automatic Identification System) data.

[0146] Simultaneously, for the polar environment, this monitoring can be used to observe changes in glaciers. Through long-term collection and analysis of satellite data, it is possible to accurately grasp information such as the melting rate, area changes, and direction of glacier movement, which is of great significance for studying global climate change. It can also monitor the distribution and thickness changes of sea ice in polar oceans, providing safety assurance for polar shipping, early warning of potential ice zone hazards, and preventing ships from encountering danger. Furthermore, for pollution in the ocean such as oil spills and floating debris, satellite data can be used to quickly locate and monitor their spread, providing strong support for marine environmental protection and pollution control.

[0147] Surface deformation and infrastructure monitoring includes dam / bridge health monitoring and mining area settlement analysis. Utilizing the high-precision monitoring capabilities of multimodal SAR satellite data, even minute deformations of dams and bridges can be captured in real time. Satellites continuously acquire topographic data of dams and bridges, and PS-InSAR technology is used to achieve millimeter-level deformation monitoring. This allows for the timely detection of potential structural safety hazards, such as crack propagation and foundation settlement, enabling millimeter-level precision deformation monitoring and early warning of potential collapse risks. This allows relevant departments to take timely maintenance and reinforcement measures to ensure the safe operation of dams and bridges.

[0148] For mining area subsidence analysis, satellites can provide comprehensive and long-term monitoring of the mining area. Real-time monitoring of the impact of mining activities on the ground helps prevent geological disasters. It allows for precise measurement of the degree and extent of ground subsidence in the mining area, and analysis of subsidence trends and rates. This helps mining companies understand geological changes in the mining area in a timely manner, rationally plan mining activities, and avoid geological disasters such as ground collapse caused by over-mining. At the same time, it also provides important data for the protection of the surrounding ecological environment and infrastructure, reducing the adverse impact of mining area subsidence on the lives of surrounding residents and other facilities.

[0149] Agricultural and ecological environment monitoring includes precision agriculture and deforestation monitoring.

[0150] In precision agriculture, satellites can provide high-resolution images of farmland, helping farmers accurately monitor crop growth. By analyzing the image data, information such as crop health, soil moisture, and fertility distribution can be obtained. Farmers can use this data to analyze soil moisture and crop growth, guiding irrigation and fertilization, reducing water waste, and improving crop yield and quality.

[0151] For monitoring deforestation, satellites can monitor changes in forest cover in real time, automatically detect illegal logging activities, protect the ecological environment, and promptly identify illegal logging activities, providing forest protection departments with accurate location and extent information. Through long-term monitoring data, it is also possible to analyze deforestation trends and causes, providing a scientific basis for formulating forest protection policies and sustainable development strategies.

[0152] National defense and military support includes monitoring of military facilities and border patrols.

[0153] In the area of ​​military facility monitoring, satellites, with their precise remote sensing technology and high-resolution imaging, can identify suspicious military deployments or facility construction, enabling comprehensive and high-frequency monitoring of various military facilities. They can acquire real-time information on the construction, maintenance, and use of facilities such as military bases, missile silos, and radar stations. Through comparative analysis of satellite imagery from different periods, changes such as new construction, expansion, or demolition of facilities can be detected promptly, providing strong support for military command departments to accurately grasp the dynamics of their own military force deployments.

[0154] For border patrols, satellites can provide uninterrupted monitoring of the long border, real-time surveillance of border dynamics, and the detection of illegal border crossings or smuggling activities. They can clearly identify the movement of people, vehicles, and goods in border areas, effectively detecting illegal border crossings and smuggling. Even in complex geographical environments and harsh weather conditions, satellites are not limited by terrain or weather, providing stable and reliable monitoring information. By cooperating with ground patrol forces, the efficiency and effectiveness of border patrols are greatly improved, safeguarding national territorial security and border stability. Furthermore, satellites can also monitor changes in the natural environment of border areas, such as natural disasters like floods and landslides, providing crucial information for disaster prevention and mitigation efforts in border regions.

[0155] Urban planning includes urban expansion analysis. By analyzing multimodal SAR satellite data, it is possible to accurately identify changes in urban boundaries at different times, determine the direction and speed of urban expansion, monitor urban building growth trends, and assist in land resource management. This helps urban planners to scientifically and rationally formulate land use plans, avoiding resource waste and environmental problems caused by blind urban expansion.

[0156] S5. Performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

[0157] Construct an intelligent scheduling and resource optimization system to form a full-process intelligent scheduling system that includes task perception, priority decision-making, resource matching, dynamic execution, and feedback optimization.

[0158] The intelligent scheduling and resource optimization system mainly includes:

[0159] The task receiving and parsing module receives application tasks from various industries and extracts key parameters. The priority evaluation engine module dynamically generates priority weights based on dimensions such as task urgency, social value, and resource consumption. The satellite resource management library module updates the status of AIRSAT constellation satellites in real time and establishes a resource-task matching model. The execution monitoring and adjustment module tracks task execution progress in real time and automatically triggers resource reallocation when satellite status changes, such as sudden malfunctions or the insertion of new tasks, such as in emergency events.

[0160] By leveraging an intelligent scheduling and resource optimization system to achieve a dynamic priority adjustment mechanism, differentiated priority rules can be formulated based on task type (emergency / routine) to achieve precise allocation and efficient utilization of resources.

[0161] In emergency mode, when receiving official disaster warnings, industry emergency requests, or abnormal events detected automatically by the system, resources are seized and a rapid response is initiated according to high-priority tasks.

[0162] Priority rules are set with emergency missions as the highest priority, superseding all regular missions. Sub-priorities are further subdivided according to urgency. Resource preemption strategies are implemented based on mission priority, prioritizing high-resolution, high-time-efficiency configurations. Regular missions currently in progress are immediately interrupted, and the nearest satellite to the target area is reassigned to emergency observation mode. For example, if AIRSAT-08 is conducting urban planning imaging after an earthquake, the mission is immediately terminated, and its orbit is adjusted to over the disaster area. Simultaneously, the ground processing center activates an emergency channel; data is downloaded but bypasses routine verification and directly enters a rapid processing flow.

[0163] Furthermore, when multiple priority tasks compete for resources, a priority conflict resolution mechanism can be adopted, which can sort them as emergency tasks > quasi-emergency tasks > regular tasks. For tasks of the same level, they can be sorted according to the ratio of social value to resource consumption, for example, those with higher value and lower consumption are given priority.

[0164] In the standard mode, the planned execution and reserved flexible resources cover the mission scope. Satellite resources can be allocated in advance based on the needs of various industries, and baseline observation tables (clearly specifying satellites, frequency bands, and time windows) can be generated.

[0165] The optimal observation strategy involves precisely matching mission data with satellite data. Based on mission requirements and satellite characteristics, a multi-dimensional matching rule base is established to automatically select the best satellite and parameters, such as frequency band, resolution, and polarization mode.

[0166] Frequency band selection can be based on the mission's requirements for penetration and resolution. For example, X / Ku bands are used for high-resolution requirements; L bands are used for strong penetration; and C bands are used for balanced requirements.

[0167] Resolution selection can be based on target size. For example, small targets (ships, aircraft) can use a 0.5m resolution; medium-sized targets (farmland, bridges) can use a 1m resolution; and large areas (urban expansion, polar ice caps) can use a 2m resolution.

[0168] The polarization mode can be selected based on the target's scattering characteristics. For example, artificial features (buildings, ships) use VV single polarization, while natural features (vegetation, ice) use VV and VH dual polarization. Dam deformation monitoring uses VV single polarization, while deforestation monitoring uses VV and VH dual polarization.

[0169] The further intelligent scheduling and resource optimization system adopts a dynamic adjustment and resource optimization guarantee mechanism.

[0170] It can send status messages from the satellite to the ground in real time and update the resource management database in real time; if the satellite suddenly fails, the system can identify and trigger the backup plan in a short time.

[0171] For routine missions of the same type in neighboring areas, the same satellite is scheduled to continuously capture images, reducing energy consumption from orbit adjustments and optimizing resource utilization. Simultaneously, the response to emergency missions is assessed, prioritizing emergency mission response rates.

[0172] Example 2:

[0173] Based on the optimization method of Example 1, such as Figure 2 As shown, this utility model discloses a cross-industry application optimization system based on multimodal SAR satellite data.

[0174] This system addresses the cross-industry application needs of multimodal SAR satellite data by constructing a full-link intelligent system encompassing data processing, industry applications, scheduling optimization, and feedback iteration. It employs a space-ground collaborative architecture to achieve high-efficiency data services. The core modules include: a data reception and preprocessing module, an industry-customized application module, and an intelligent scheduling and resource optimization module, supplemented by an efficiency feedback module, an industry database update module, and a visualization analysis module, forming a closed-loop operation mechanism.

[0175] The data receiving and preprocessing module is used to collect multimodal SAR satellite data and perform data unification, seamless regional data stitching, and automatic data quality optimization to provide a high-quality data foundation for subsequent industry applications.

[0176] The data receiving and preprocessing module is equipped with signal receiving terminals compatible with Ku, X, C, and L bands, and accesses AIRSAT constellation satellite data in real time (such as Ku band for satellite 01 and X band for satellite 08), parses metadata such as satellite number, imaging time, and resolution, and stores it in a distributed database.

[0177] Sensor gain and atmospheric interference are eliminated through radiometric correction; data is unified to the WGS84 coordinate system through geometric correction; data is synchronized to UTC time through time correction; and data is uniformly processed by using adaptive filtering through noise suppression.

[0178] Based on a deep learning-based spatiotemporal registration algorithm, feature points are matched for data from the same region across different satellites and time phases. Seams are eliminated through radiometric consistency correction and weighted fusion. Intelligent enhancement is performed to meet industry needs, achieving seamless stitching and quality optimization.

[0179] Industry-specific application modules, based on optimized data, provide customized monitoring and decision support services for various sub-sectors, specifically including:

[0180] Disaster monitoring and emergency response: By integrating change detection algorithms, it automatically compares pre- and post-disaster data for disasters such as earthquakes and floods, generating damage maps in a short time and pushing them to emergency management departments. Marine and polar environment monitoring: Using ship detection models (combined with AIS data) to identify illegal fishing vessels; long-term tracking of polar ice layer changes using L-band data to quantify glacier melting rates. Surface deformation and infrastructure monitoring: It can achieve millimeter-level deformation monitoring of dams and bridges, automatically generating risk warning reports; it can perform real-time analysis of mining area subsidence, marking high-risk areas of subsidence rate. Agriculture and ecological environment monitoring: It can use L-band data to invert soil moisture, combined with C-band vegetation indices to guide precision irrigation; it can detect illegal deforestation using X-band data. National defense and military support: Based on high-resolution data, it can identify military facilities (such as aircraft and radar stations) and abnormal border activities (such as illegal border crossings). Urban planning: It can extract building area boundaries using C-band data, analyze urban expansion trends, and assist in land resource management.

[0181] The intelligent scheduling and resource optimization module is used to perform intelligent task scheduling and resource optimization for application tasks in various industries, and automatically select the best SAR satellite data.

[0182] The intelligent scheduling and resource optimization module can dynamically allocate satellite resources to ensure efficient mission execution and achieve the scheduling goal of prioritizing emergency response and maintaining orderly routine operations.

[0183] Dynamic priority adjustment includes: In emergency mode, when receiving signals such as earthquake early warnings or border alerts, the highest priority is activated, regular tasks are interrupted, and the nearest satellite is scheduled for rapid imaging. In regular mode, tasks such as agriculture and urban planning can be scheduled on a quarterly basis, reserving flexible resources, making plans, and responding to quasi-emergency needs.

[0184] Optimal observation matching automatically selects satellite parameters based on mission requirements. For example, military reconnaissance uses the X-band 0.5m spotlight mode; soil moisture monitoring uses the L-band 1m strip mode; and polar ice monitoring uses the C-band 2m dual polarization mode.

[0185] Furthermore, the system also includes a performance feedback module, which collects user evaluations of the results from various industries and regularly iterates and optimizes the algorithm. The industry library is updated in a timely manner to customize and update industry applications, keeping them timely.

[0186] The performance feedback module meticulously records all user feedback issues and suggestions, categorizing them according to the characteristics of different industries. For issues with concentrated feedback, a professional team conducts in-depth analysis to explore potential shortcomings in the algorithm. By simulating task execution under different scenarios, the algorithm undergoes multiple adjustments and tests to ensure that the optimized algorithm better meets the actual needs of various industries.

[0187] The industry database update module monitors the latest developments and trends across various industries in real time. For example, in agriculture, it updates relevant monitoring parameters and analytical models promptly as planting techniques and crop varieties improve. By maintaining close communication with experts and institutions in various industries, it obtains firsthand industry information, ensuring the accuracy and cutting-edge nature of its industry application content.

[0188] Furthermore, the system also includes a visualization and analysis module, which provides interactive SAR satellite data browsing and decision support.

[0189] The visualization analysis module provides an interactive platform that supports data layer overlay (such as pre- and post-disaster image comparison) and trend chart generation, helping decision-makers to intuitively obtain information.

[0190] By dynamically displaying data changes across different time series, decision-makers can clearly understand the underlying patterns in the data. In the field of urban planning, images of urban construction from different periods can be overlaid to analyze the direction of urban expansion and the speed of development. Furthermore, this module also boasts powerful filtering and query functions, allowing users to precisely filter data for specific regions and time ranges according to their needs.

[0191] Furthermore, the system adopts a space-ground collaborative architecture, in which a lightweight preprocessing model is deployed on the satellite to achieve on-orbit data screening, and a high-performance processing center is built on the ground station to complete complex analysis algorithms.

[0192] On the satellite side: Lightweight preprocessing models are deployed to enable on-orbit data filtering, such as removing cloud-obscured areas and reducing the amount of invalid data transmitted. On the ground station side: High-performance processing centers are built to run complex algorithms (such as SAR data deformation analysis and deep learning target detection), and data services are provided to industry users through cloud service interfaces.

[0193] This satellite-ground collaborative model significantly improves the system's operational efficiency and data processing capabilities. On-orbit data filtering at the satellite end effectively reduces the data processing burden on ground stations, allowing them to focus on processing more valuable data. Removing cloud-obscured areas ensures cleaner downlink data, preventing invalid data from interfering with subsequent analysis. Meanwhile, the ground station's high-performance processing center, with its powerful computing capabilities, can perform in-depth analysis of the filtered data.

[0194] Application examples:

[0195] In response to the emergency, a level 6 earthquake warning was issued. The system automatically activated the emergency mode, interrupted the regular urban imaging mission of the AIRSAT-08 satellite, and switched it to the X-band 0.5m spotlight mode to complete the imaging of the disaster area within 15 minutes.

[0196] After receiving the data, the ground center completes radiometric correction and stitching with pre-disaster data within 10 minutes to enhance surface deformation characteristics. Based on disaster monitoring and emergency response, a damage map is generated, marking areas of house damage and road interruptions.

[0197] Furthermore, after emergency departments use the system, they can adjust the algorithm parameters based on feedback on any missed areas to improve the accuracy of the next response.

[0198] This embodiment discloses a cross-industry application optimization system based on multimodal SAR satellite data. It employs a unified multi-band data processing method, achieving seamless stitching accuracy down to the sub-pixel level, thus improving the precision of industry services and meeting the diverse needs of multiple fields. Simultaneously, it enables flexible resource scheduling and a near 100% emergency task response rate. Through modular design and the integration of intelligent algorithms, this system optimizes the entire process of multimodal SAR satellite data acquisition and application, providing efficient and accurate technical support for disaster management, national defense, and other fields.

[0199] The present invention also provides an electronic device, Figure 11 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, for example... Figure 11 As shown, the electronic device may include a processor, a communications interface, memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor can invoke logical instructions from the memory, for example, to execute the following method:

[0200] S1. Perform unified data processing on multimodal SAR satellite data;

[0201] S2. Perform seamless regional data stitching processing on unified multimodal SAR satellite data;

[0202] S3. Automatically optimize the data quality of seamlessly stitched multimodal SAR satellite data to meet the needs of different industries;

[0203] S4. Based on industry-customized application modules, call automatically optimized multimodal SAR satellite data for industry-specific task applications;

[0204] S5. Performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

[0205] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0206] This invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the methods provided in the above embodiments, including, for example:

[0207] S1. Perform unified data processing on multimodal SAR satellite data;

[0208] S2. Perform seamless regional data stitching processing on unified multimodal SAR satellite data;

[0209] S3. Automatically optimize the data quality of seamlessly stitched multimodal SAR satellite data to meet the needs of different industries;

[0210] S4. Based on industry-customized application modules, call automatically optimized multimodal SAR satellite data for industry-specific task applications;

[0211] S5. Performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

[0212] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0213] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0214] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing cross-industry applications based on multimodal SAR satellite data, characterized in that, include: S1. Perform unified data processing on multimodal SAR satellite data; S2. Perform seamless regional data stitching processing on unified multimodal SAR satellite data; S3. Automatically optimize the data quality of seamlessly stitched multimodal SAR satellite data to meet the needs of different industries; S4. Based on industry-customized application modules, call automatically optimized multimodal SAR satellite data for industry-specific task applications; S5. Performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

2. The optimization method according to claim 1, characterized in that, The S1 step involves unified data processing of multimodal SAR satellite data, including: A unified data receiving and preprocessing center will be established to perform radiometric correction, geometric correction, time correction, and noise suppression on Ku / X / C / L band SAR satellite data. Radiation correction includes: sensor gain correction, atmospheric attenuation correction, and surface roughness correction; Geometric correction includes: basic geometric correction and terrain correction; Time correction includes: timestamp standardization and timing consistency optimization; Noise suppression includes eliminating speckle noise and impulse interference in SAR data of different frequency bands, thereby improving the data signal-to-noise ratio.

3. The optimization method according to claim 1, characterized in that, The S2 process involves seamless stitching of regional data, including: A deep learning-based spatiotemporal registration algorithm seamlessly stitches together image data of the same area taken by different satellites at different times, including the following steps: S11. Perform pre-processing before splicing; S12, Perform AI automatic matching to extract stable feature points; S13. Align image data of the same area based on effective stable feature points; S14. For the aligned regions, perform radiation consistency optimization and overlap fusion strategies to achieve seamless splicing.

4. The optimization method according to claim 1, characterized in that, The industry-specific applications include: Disaster monitoring and emergency response, marine and polar environmental monitoring, surface deformation and infrastructure monitoring, agricultural and ecological environment monitoring, national defense and military support, and urban planning.

5. The optimization method according to claim 4, characterized in that, The S3 section performs automatic data quality optimization to adapt to the needs of different industries, including: Disaster monitoring and emergency response: Enhance the ability to detect minute surface deformations and improve the accuracy of landslide and earthquake early warning; Marine and polar environment monitoring: Optimize wave and ice floe identification, and enhance ship and iceberg tracking capabilities; Agricultural and ecological environment monitoring: Enhance soil moisture and crop growth status analysis to improve the accuracy of yield prediction.

6. The optimization method according to claim 1, characterized in that, In S5, intelligent task scheduling and resource optimization are performed for various industry application tasks, and the best SAR satellite data is automatically selected, including: Dynamically adjust task priorities to optimize satellite observation plans and resource allocation; An adaptive optimal observation strategy intelligently recommends the optimal observation parameters based on the task.

7. An optimization system, characterized in that, The system, applied to the optimization method according to any one of claims 1 to 6, comprises: The data receiving and preprocessing module collects multimodal SAR satellite data and performs data unification, seamless regional data stitching, and automatic data quality optimization. Industry-specific customized application modules allow for the use of automatically optimized multimodal SAR satellite data for industry-specific tasks. The intelligent scheduling and resource optimization module performs intelligent task scheduling and resource optimization for various industry application tasks, and automatically selects the best SAR satellite data.

8. The system according to claim 7, characterized in that, The system also includes: The performance feedback module iteratively optimizes the system algorithm based on user feedback data; The industry database update module is used to customize and update industry applications, keeping them up-to-date.

9. The system according to claim 7, characterized in that, The system also includes: The visualization and analysis module provides interactive SAR satellite data browsing and decision support.

10. The system according to claim 7, characterized in that, The system adopts a space-ground collaborative architecture, in which a lightweight preprocessing model is deployed on the satellite to achieve on-orbit data screening, and a high-performance processing center is built on the ground station to complete complex analysis algorithms.

Images