Satellite remote sensing image data processing method and system

Abstract

The present invention relates to the technical field of remote sensing image processing, and in particular to a method and system for processing satellite remote sensing image data. The method comprises the following steps: acquiring satellite remote sensing observation signals; decoding the satellite remote sensing observation signals to generate satellite remote sensing observation data; integrating the satellite remote sensing observation denoised data to generate satellite remote sensing observation images; reconstructing the satellite remote sensing observation images with spectral super-resolution to generate satellite remote sensing reconstructed images; reconstructing the satellite remote sensing reconstructed images with hyperspectral to generate satellite remote sensing hyperspectral reconstructed images; and calculating the spectral index of the satellite remote sensing hyperspectral reconstructed images to obtain the satellite remote sensing spectral index. The present invention improves the accuracy and quality of satellite remote sensing images through systematic decoding, denoising, super-resolution reconstruction, hyperspectral reconstruction, spectral index calculation and optimization, edge computing and cloud storage.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image processing, and in particular to a method and system for processing satellite remote sensing image data. Background Art

[0002] Early remote sensing images were mainly used for military reconnaissance and meteorological observation. With the advancement of technology, the resolution and coverage of remote sensing images have been continuously improved, making them widely used in fields such as earth science, environmental monitoring and natural resource management. With the development of computer technology, digital image processing technology has gradually matured, and the processing of remote sensing images has shifted from analog to digital. During this period, basic image processing technologies such as image enhancement, image restoration and image segmentation have been widely used, greatly improving the quality and availability of remote sensing images. With the launch of high-resolution remote sensing satellites, such as the Landsat series in the United States, the SPOT series in France and the Gaofen series in China, the data volume and accuracy of remote sensing images have made a qualitative leap. At the same time, the advancement of image processing algorithms, such as the introduction of machine learning and deep learning, has made automatic classification and target recognition of remote sensing images possible, greatly improving the efficiency and accuracy of data processing. However, the current traditional low-resolution images cannot meet the needs of refined analysis. At the same time, the influence of environmental interference is often ignored when calculating the spectral index, resulting in low accuracy and quality of satellite remote sensing images. Summary of the invention

[0003] Based on this, it is necessary to provide a method and system for processing satellite remote sensing image data to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for processing satellite remote sensing image data is provided, the method comprising the following steps:

[0005] Step S1: acquiring satellite remote sensing observation signals; decoding the satellite remote sensing observation signals to generate satellite remote sensing observation data; integrating the satellite remote sensing observation denoised data to generate satellite remote sensing observation images;

[0006] Step S2: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image; performing hyperspectral reconstruction on the satellite remote sensing reconstructed image to generate a satellite remote sensing hyperspectral reconstructed image;

[0007] Step S3: Calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with a preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, perform hyperspectral feature data on the corresponding satellite remote sensing observation spectral feature data; use the hyperspectral feature data to perform atmospheric regional interference optimization on the satellite remote sensing spectral feature map to generate a satellite remote sensing atmospheric optimization image;

[0008] Step S4: Use edge computing technology to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage to obtain the satellite remote sensing atmosphere storage image; perform distributed processing on the satellite remote sensing atmosphere storage image to execute the processing operation of the satellite remote sensing image data.

[0009] The present invention provides an original data source for subsequent processing by receiving original remote sensing signals from satellites. The satellite remote sensing observation signal is decoded to generate available observation data, and the original signal is converted into digital data that can be further processed. The observation data is denoised to improve the quality of the data, reduce noise interference, and enhance the accuracy of the data. The denoised data is integrated to generate a remote sensing observation image to form a preliminary remote sensing image, which provides a basis for subsequent processing. The remote sensing observation image is spectrally super-resolution reconstructed to generate a high-resolution reconstructed image, improve the resolution of the image, and enhance the details and clarity of the image. The reconstructed image is hyperspectrally reconstructed to generate a hyperspectral reconstructed image, increase the spectral information of the image, and improve the spectral resolution and accuracy of the image. The spectral index of the hyperspectral reconstructed image is calculated to extract key spectral feature parameters for subsequent analysis. The calculated spectral index is compared with a preset threshold value to screen out spectral feature data that meets specific conditions for key analysis. The hyperspectral feature data generated for atmospheric optimization can provide more refined feature data and lay the foundation for the optimization process. Using hyperspectral feature data to optimize the interference of the atmospheric region of the image and generate an optimized image can reduce atmospheric interference and improve image quality and authenticity. Uploading the optimized image to the cloud platform through edge computing technology can achieve efficient data transmission and preliminary processing, and reduce the computing burden of the cloud platform. Achieving efficient data transmission and preliminary processing, reducing the computing burden of the cloud platform, can provide a safe and reliable image storage solution. Distributed processing of images stored in the cloud platform can improve the efficiency and scalability of image processing and support large-scale data processing operations. Therefore, the present invention improves the accuracy and quality of satellite remote sensing images through systematic decoding, denoising, super-resolution reconstruction, hyperspectral reconstruction, spectral index calculation and optimization, edge computing and cloud storage.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire satellite remote sensing observation signals;

[0012] Step S12: Decoding the satellite remote sensing observation signal to generate satellite remote sensing observation data;

[0013] Step S13: performing radiation correction on the satellite remote sensing observation data to generate satellite remote sensing observation radiation correction data; performing geometric correction on the satellite remote sensing observation radiation correction data to generate satellite remote sensing observation geometric correction data; performing data denoising on the satellite remote sensing observation geometric correction data to generate satellite remote sensing observation denoised data;

[0014] Step S14: performing remote sensing image integration on the satellite remote sensing observation denoised data to generate a satellite remote sensing observation image.

[0015] The present invention obtains electromagnetic signals reflected or radiated from the surface of the earth through a satellite receiving device, provides basic raw data for subsequent data processing, and makes it possible to obtain remote sensing information. The obtained raw observation signal is decoded and converted into a processable digital data format, and the raw signal that cannot be directly used is converted into processable digital data, laying a foundation for subsequent correction and processing. Radiation correction is performed on satellite remote sensing observation data, satellite remote sensing observation radiation correction data is generated, and radiation errors in the observation data are adjusted to ensure that the radiation value of the data accurately reflects the real radiation characteristics of the ground object, improve the radiation accuracy of the data, ensure the radiation authenticity of the observation data, and provide accurate basic data for scientific research and application. Correct the geometric distortion in the observation data so that the spatial position of the image accurately corresponds to the geographic coordinate system, improve the spatial accuracy of the data, ensure the accurate correspondence between the image and the geographic location, and provide accurate spatial data for applications such as geographic information systems (GIS). Remove noise in the observation data, improve the signal-to-noise ratio of the data, make the image clearer, improve the data quality, make the image clearer and more usable, reduce errors, and improve the reliability of the data. Integrate the denoised remote sensing data to generate complete remote sensing images, generate high-quality remote sensing images that can be used for analysis and application, provide clear surface information, and provide important data support for environmental monitoring, resource management and other fields.

[0016] Preferably, step S14 includes the following steps:

[0017] Step S141: performing spectral extraction on the satellite remote sensing observation denoised data to obtain satellite remote sensing spectral data; performing spectral band analysis on the satellite remote sensing spectral data to generate satellite remote sensing spectral band data;

[0018] Step S142: using true color synthesis technology to perform spectral image synthesis on satellite remote sensing spectral band data to generate a satellite remote sensing spectral synthesis image;

[0019] Step S143: performing image preprocessing on the satellite remote sensing spectral composite image to generate a standard satellite remote sensing spectral composite image, wherein the image preprocessing includes histogram averaging, image brightness enhancement and contrast stretching;

[0020] Step S144: Perform geographic information encoding on the standard satellite remote sensing spectrum composite image according to GIS technology to generate a satellite remote sensing observation image.

[0021] The present invention extracts data of each spectral band from the denoised remote sensing data and analyzes it to generate spectral band data. Through spectral extraction and analysis, detailed spectral information can be obtained, which is very important for identifying and classifying land features, and improves the scientific value and application potential of the data. The true color synthesis technology is used to synthesize multiple spectral band data into a color image, so that the image is closer to the effect of visible light of the human eye. The true color synthesis technology makes the image more intuitive and easy to understand, enhances the visualization effect of the image, and helps to quickly identify and analyze surface features. The brightness distribution of the image is balanced, the overall contrast of the image is improved, the brightness of the image is increased, the dark details are clearer, and the contrast of the image is stretched, so that the distinction between different parts of the image is more obvious. The preprocessing step improves the visual quality and detail expression of the image, making the image easier to use and more accurate in the analysis and application process. The preprocessed image is encoded with geographic information using geographic information system (GIS) technology, so that the image has accurate geographic location information. Through GIS encoding, the image and the geographic location are accurately corresponding, ensuring that the image can be used in various geographic information system applications and providing accurate spatial data support.

[0022] Preferably, step S2 comprises the following steps:

[0023] Step S21: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image;

[0024] Step S22: performing spectral interpolation on the satellite remote sensing reconstructed image to generate a satellite remote sensing spectral interpolation image;

[0025] Step S23: performing hyperspectral reconstruction on the satellite remote sensing spectral interpolation image to generate a satellite remote sensing hyperspectral reconstructed image.

[0026] The present invention utilizes spectral super-resolution technology to process low-resolution satellite remote sensing images through algorithms, thereby improving their spectral and spatial resolutions and generating higher-resolution images. Improving the resolution of the image makes the details of the image clearer, and can identify smaller-scale features of land objects, providing a basis for high-precision geographic information applications. The reconstructed image is subjected to spectral interpolation processing, and the spectral resolution of the image is improved by increasing spectral data points. Through spectral interpolation, the gaps in the spectral data are filled, the spectral continuity and accuracy of the image are improved, and it is helpful to more accurately analyze and classify land objects. Using hyperspectral reconstruction technology, the interpolated image is further processed to generate a hyperspectral image, so that the image has richer spectral information at each pixel. Hyperspectral reconstruction enables each pixel to contain more spectral band information, greatly enhancing the spectral resolution ability of the image, and facilitating more detailed and accurate analysis and classification of land objects.

[0027] Preferably, performing hyperspectral reconstruction on the satellite remote sensing spectral interpolation image includes:

[0028] The spectral regional resolution analysis is performed on the satellite remote sensing spectral interpolation image to generate the satellite remote sensing spectral image resolution data; the resolution of the satellite remote sensing spectral image resolution data is restored according to the non-negative matrix decomposition to generate the satellite remote sensing hyperspectral reconstructed image.

[0029] The present invention analyzes the interpolated spectral image, evaluates the resolution of different spectral regions, and generates corresponding resolution data. This process generally includes identifying and quantifying the resolution of each spectral region in the image. Through analysis, the resolution of each spectral region can be quantified to understand the resolution distribution of the image. Provide basic data for subsequent resolution recovery, so that the resolution improvement process is more targeted and effective. Using non-negative matrix decomposition (NMF) technology, the spectral image resolution data is processed to restore and improve the resolution of the image and generate a hyperspectral reconstructed image. NMF is a matrix decomposition technology that can decompose the data matrix into the product of two non-negative matrices to extract the characteristic information in the image. Through NMF technology, the resolution of the spectral image can be effectively restored and improved, making the image more detailed in spectrum and space. NMF technology can maintain the non-negativity of the original data, ensure that the reconstructed image has high fidelity, and will not introduce unreal negative information. NMF can extract characteristic information in the image, enhance the details and resolution of the image, and contribute to further analysis and classification of land objects. Through spectral region resolution analysis, detailed resolution distribution data is obtained to provide a basis for subsequent processing. Using non-negative matrix decomposition technology, the image resolution can be effectively restored and improved, while maintaining the non-negativity of the image data and ensuring the high fidelity of the reconstructed image. NMF technology can extract and enhance the feature information in the image and improve the image's detail expression and resolution.

[0030] Preferably, step S3 comprises the following steps:

[0031] Step S31: performing spectral decomposition on the satellite remote sensing hyperspectral reconstructed image to generate satellite remote sensing observation spectral feature data;

[0032] Step S32: Calculate the spectral index of the satellite remote sensing observation spectral feature data to obtain the satellite remote sensing spectral index;

[0033] Step S33: Compare the satellite remote sensing spectral index with a preset spectral index threshold value. When the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold value, perform hyperspectral marking on the corresponding satellite remote sensing observation spectral feature data to generate hyperspectral feature data; when the satellite remote sensing spectral index is less than the preset spectral index threshold value, perform spectral feature combination on the hyperspectral feature data to generate a satellite remote sensing spectral feature map; perform atmospheric transmittance evaluation on the satellite remote sensing observation denoised data using the satellite remote sensing spectral feature map to generate atmospheric transmittance evaluation data;

[0034] Step S34: Optimizing the atmospheric regional interference of the satellite remote sensing spectral characteristic map based on the atmospheric transmittance evaluation data to generate a satellite remote sensing atmospheric optimization image.

[0035] The present invention provides basic data for subsequent spectral index calculation and analysis by extracting important spectral features from hyperspectral images. Complex hyperspectral data is decomposed into processable feature data to simplify subsequent processing steps. Through the calculation of spectral index, different ground objects can be quantitatively analyzed and classified. Spectral index can be used in a variety of remote sensing applications such as vegetation monitoring, water body analysis, and soil classification. Corresponding spectral feature data are hyperspectrally labeled to generate hyperspectral feature data. Corresponding spectral feature data are eliminated. Hyperspectral feature data are combined to generate satellite remote sensing spectral feature maps. Satellite remote sensing spectral feature maps are used to evaluate the atmospheric transmittance of observed denoised data to generate atmospheric transmittance evaluation data. Through threshold comparison, important spectral feature data can be effectively screened out and irrelevant or noise data can be removed. Important spectral features are marked to generate more representative and analytically valuable hyperspectral feature data. Atmospheric transmittance is evaluated through spectral feature maps to understand the impact of the atmosphere on remote sensing data and improve data quality. By using the atmospheric transmittance assessment data, the atmospheric regional interference of the spectral characteristic map is optimized to reduce the interference of the atmosphere on the remote sensing data, generate optimized images, effectively reduce the interference of the atmosphere on the remote sensing images, and improve the clarity and authenticity of the images.

[0036] Preferably, using the satellite remote sensing spectral characteristic map to evaluate the atmospheric transmittance of the satellite remote sensing observation denoised data includes:

[0037] Collect historical data of satellite remote sensing observation denoised data to obtain satellite remote sensing observation historical data;

[0038] The spectral characteristics impact analysis is performed on the satellite remote sensing spectral characteristic map through the 6S radiation transfer model to generate initial atmospheric composition data; the atmospheric transmittance is calculated on the initial atmospheric composition data to obtain atmospheric transmittance data; the atmospheric assessment of the atmospheric transmittance data is optimized through satellite remote sensing observation denoising data and satellite remote sensing observation historical data to generate atmospheric transmittance assessment data.

[0039] The present invention forms a rich sample library through the accumulation of historical data, provides a reference for the evaluation of atmospheric transmittance, can analyze the atmospheric change trend under different periods and conditions, and enhance the accuracy of the evaluation. Use the 6S (Second Simulation of a Satellite Signal in the Solar Spectrum) radiation transmission model, combine the satellite remote sensing spectral characteristic map, analyze the influence of spectral characteristics on atmospheric components, and generate initial atmospheric composition data. The 6S model can simulate the process of sunlight transmission in the atmosphere and accurately analyze the influence of atmospheric components on spectral characteristics. Based on the initial atmospheric composition data, the atmospheric transmittance is calculated, the transmittance data on each spectral band is obtained, the atmospheric transmittance of each spectral band is accurately calculated, the influence of the atmosphere on the remote sensing signal is understood, and basic data is provided for subsequent transmittance evaluation optimization. Combined with the current denoising data and historical data, the initially calculated atmospheric transmittance data is optimized and evaluated, more accurate atmospheric transmittance evaluation data is generated, and historical data is used for correction to reduce errors and improve the accuracy of the evaluation. The generated transmittance evaluation data is more accurate and can be used for further atmospheric correction and image optimization.

[0040] Preferably, performing atmospheric regional interference optimization on the satellite remote sensing spectral characteristic map based on the atmospheric transmittance evaluation data comprises:

[0041] Conduct time-series image acquisition of satellite remote sensing spectral characteristic images to obtain satellite remote sensing atmospheric multi-temporal images;

[0042] Perform time series alignment on satellite remote sensing atmospheric multi-temporal images to generate satellite remote sensing atmospheric multi-temporal phase alignment images; perform atmospheric change region edge segmentation on satellite remote sensing atmospheric multi-temporal phase alignment images to generate atmospheric change edge region images;

[0043] Based on the atmospheric transmittance assessment data, the visible area of ​​the atmospheric change edge area image is screened to obtain the atmospheric change visible area image; according to the atmospheric change visible area image, the atmospheric interference area of ​​the satellite remote sensing spectral characteristic map is filled to generate the satellite remote sensing atmospheric optimization image.

[0044] The present invention can analyze the atmospheric conditions and change trends at different time points through multi-phase images, increase the data of the time dimension, and improve the recognition and processing capabilities of atmospheric changes. The multi-phase images are aligned to ensure that the images at different time points are aligned in space for subsequent analysis, to ensure that the images at different time points are accurately aligned, to reduce the spatial error in the time series analysis, and the aligned images have higher consistency, which is convenient for comparison and analysis. Through image processing technology, the regional edges of atmospheric changes in the multi-phase aligned images are segmented, and the specific areas of atmospheric changes are identified. The atmospheric change areas can be accurately identified, and the detection capabilities of atmospheric interference areas are improved. The generated edge area images clearly mark the boundaries of the atmospheric change areas, which is convenient for subsequent processing. The areas with high visibility are screened out by evaluating the data, ensuring that the retained data quality is higher, eliminating the areas with low visibility, and reducing the interference of atmospheric noise on the image. Using the visible area images of atmospheric changes, the atmospheric interference areas in the spectral characteristic graph are filled and corrected, and the optimized images are generated, which can effectively fill and correct the atmospheric interference areas, improve the overall quality of the images, and the generated atmospheric optimized images have higher clarity and authenticity, which is convenient for further analysis and application.

[0045] Preferably, step S4 comprises the following steps:

[0046] Step S41: using edge computing technology to compress the satellite remote sensing atmospheric optimization image to generate a satellite remote sensing atmospheric compression image;

[0047] Step S42: Upload the satellite remote sensing atmospheric compressed image to the cloud platform for image storage to obtain the satellite remote sensing atmospheric storage image; perform distributed processing on the satellite remote sensing atmospheric storage image to execute the processing operation of the satellite remote sensing image data.

[0048] The present invention uses edge computing technology to compress the atmospheric optimization image on the edge device close to the data source to reduce the size of the image file while retaining the key features and information of the image. Edge computing technology allows real-time processing near the data source to reduce data transmission delay. Through image compression, the bandwidth required for data transmission is significantly reduced and the transmission efficiency is improved. Edge computing reduces the burden on the central server and improves the overall processing efficiency. The compressed image is uploaded to the cloud platform for storage to utilize the high capacity and high reliability storage facilities of the cloud platform. The cloud platform provides highly reliable storage services to ensure the security and persistence of data. Cloud storage has good scalability and can process a large amount of satellite remote sensing image data. Through the cloud platform, users can easily access and manage the stored remote sensing images. Distributed computing technology is used on the cloud platform to process the stored remote sensing image data, such as analysis, classification, monitoring, etc., to perform various remote sensing image processing operations. Using the distributed computing resources of the cloud platform, large-scale remote sensing image data can be quickly processed. The distributed processing system has a high degree of flexibility and can dynamically allocate computing resources according to demand. Distributed processing improves the efficiency of image processing tasks and shortens processing time.

[0049] In this specification, a satellite remote sensing image data processing system is provided, which is used to execute the above-mentioned satellite remote sensing image data processing method. The satellite remote sensing image data processing system includes:

[0050] The remote sensing signal conversion module is used to obtain satellite remote sensing observation signals; decode the satellite remote sensing observation signals to generate satellite remote sensing observation data; perform remote sensing image integration on the satellite remote sensing observation denoising data to generate satellite remote sensing observation images;

[0051] The spectral reconstruction module is used to perform spectral super-resolution reconstruction on satellite remote sensing observation images to generate satellite remote sensing reconstructed images; perform hyperspectral reconstruction on satellite remote sensing reconstructed images to generate satellite remote sensing hyperspectral reconstructed images;

[0052] The atmospheric interference optimization module is used to calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with the preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, the corresponding satellite remote sensing observation spectral feature data is subjected to hyperspectral feature data; the satellite remote sensing spectral feature map is subjected to atmospheric regional interference optimization using the hyperspectral feature data to generate a satellite remote sensing atmospheric optimization image;

[0053] The image storage module is used to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage using edge computing technology to obtain the satellite remote sensing atmosphere storage image; the satellite remote sensing atmosphere storage image is distributedly processed to execute the processing operation of the satellite remote sensing image data.

[0054] The beneficial effect of the present invention is that by obtaining the original satellite remote sensing data, basic data is provided for subsequent processing. Spectral super-resolution reconstruction and hyperspectral reconstruction are performed on satellite remote sensing observation images. These technologies can improve the resolution and spectral information of the image, making the observation image clearer and more detailed, and help to extract ground object information more accurately. The spectral index of the satellite remote sensing hyperspectral reconstruction image is calculated and compared with the preset spectral index threshold. This step can provide spectral index information for analyzing and evaluating environmental indicators such as vegetation growth and soil quality in different regions, which is helpful for research and decision-making in the fields of agriculture and environmental monitoring. The satellite remote sensing atmospheric optimization image is uploaded to the cloud platform and stored using edge computing technology, and then distributed processing is performed. This processing method can make full use of the computing and storage resources of the cloud platform, improve processing efficiency and data processing capabilities, and provide users with fast and efficient satellite remote sensing data processing services. Therefore, the present invention improves the accuracy and quality of satellite remote sensing images through systematic decoding, denoising, super-resolution reconstruction, hyperspectral reconstruction, spectral index calculation and optimization, edge computing and cloud storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic diagram of the steps of a method for processing satellite remote sensing image data;

[0056] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0057] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0058] Figure 4 for Figure 1 Detailed implementation steps of step S4 in FIG.

[0059] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0060] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0061] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0062] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0063] To achieve this, please refer to Figures 1 to 4 , a method for processing satellite remote sensing image data, the method comprising the following steps:

[0064] Step S1: acquiring satellite remote sensing observation signals; decoding the satellite remote sensing observation signals to generate satellite remote sensing observation data; integrating the satellite remote sensing observation denoised data to generate satellite remote sensing observation images;

[0065] Step S2: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image; performing hyperspectral reconstruction on the satellite remote sensing reconstructed image to generate a satellite remote sensing hyperspectral reconstructed image;

[0066] Step S3: Calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with a preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, perform hyperspectral feature data on the corresponding satellite remote sensing observation spectral feature data; use the hyperspectral feature data to perform atmospheric regional interference optimization on the satellite remote sensing spectral feature map to generate a satellite remote sensing atmospheric optimization image;

[0067] Step S4: Use edge computing technology to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage to obtain the satellite remote sensing atmosphere storage image; perform distributed processing on the satellite remote sensing atmosphere storage image to execute the processing operation of the satellite remote sensing image data.

[0068] The present invention provides an original data source for subsequent processing by receiving original remote sensing signals from satellites. The satellite remote sensing observation signal is decoded to generate available observation data, and the original signal is converted into digital data that can be further processed. The observation data is denoised to improve the quality of the data, reduce noise interference, and enhance the accuracy of the data. The denoised data is integrated to generate a remote sensing observation image to form a preliminary remote sensing image, which provides a basis for subsequent processing. The remote sensing observation image is spectrally super-resolution reconstructed to generate a high-resolution reconstructed image, improve the resolution of the image, and enhance the details and clarity of the image. The reconstructed image is hyperspectrally reconstructed to generate a hyperspectral reconstructed image, increase the spectral information of the image, and improve the spectral resolution and accuracy of the image. The spectral index of the hyperspectral reconstructed image is calculated to extract key spectral feature parameters for subsequent analysis. The calculated spectral index is compared with a preset threshold value to screen out spectral feature data that meets specific conditions for key analysis. The hyperspectral feature data generated for atmospheric optimization can provide more refined feature data and lay the foundation for the optimization process. Using hyperspectral feature data to optimize the interference of the atmospheric region of the image and generate an optimized image can reduce atmospheric interference and improve image quality and authenticity. Uploading the optimized image to the cloud platform through edge computing technology can achieve efficient data transmission and preliminary processing, and reduce the computing burden of the cloud platform. Achieving efficient data transmission and preliminary processing, reducing the computing burden of the cloud platform, can provide a safe and reliable image storage solution. Distributed processing of images stored in the cloud platform can improve the efficiency and scalability of image processing and support large-scale data processing operations. Therefore, the present invention improves the accuracy and quality of satellite remote sensing images through systematic decoding, denoising, super-resolution reconstruction, hyperspectral reconstruction, spectral index calculation and optimization, edge computing and cloud storage.

[0069] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a method for processing satellite remote sensing image data of the present invention. In this example, the method for processing satellite remote sensing image data includes the following steps:

[0070] Step S1: acquiring satellite remote sensing observation signals; decoding the satellite remote sensing observation signals to generate satellite remote sensing observation data; integrating the satellite remote sensing observation denoised data to generate satellite remote sensing observation images;

[0071] In an embodiment of the present invention, by starting the remote sensing equipment on the satellite, such as a multispectral sensor, a hyperspectral sensor, a radar, etc., the observation signal of the ground or the ocean is collected. The collected original observation signal is transmitted to the ground receiving station through the satellite communication link. The ground receiving station uses a high-gain antenna to receive the observation signal from the satellite. The received signal is preliminarily processed to filter out noise and interference to ensure the signal quality. The received signal is decoded using special decoding software and algorithms to extract valid observation data. The decoding process includes steps such as synchronization, error correction, and data packet parsing. The decoded observation data is converted into a standard format, such as GeoTIFF, HDF5, etc., for subsequent processing and analysis. By analyzing the observation data, the noise and outliers therein are detected and marked. This can be achieved by methods such as statistical analysis and filters. Different denoising algorithms, such as mean filtering, median filtering, wavelet transform, etc., are applied to remove noise from the observation data. Ensure that the denoised data maintains the integrity and accuracy of the original information. The observation data of multiple frequency bands or multiple time points are fused to generate comprehensive remote sensing data. Data fusion can improve the spatial resolution, spectral resolution, and temporal resolution of the image. Use image processing software to convert the fused remote sensing data into images. This includes steps such as stitching, correction, and registration. Apply color coding and enhancement techniques to improve the visual effect of the image. Perform geometric correction and radiation calibration on the generated remote sensing images to ensure their accuracy and consistency.

[0072] Step S2: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image; performing hyperspectral reconstruction on the satellite remote sensing reconstructed image to generate a satellite remote sensing hyperspectral reconstructed image;

[0073] In an embodiment of the present invention, the image data quality and format consistency are ensured by collecting and collating satellite remote sensing observation images. If there are images of multiple phases, phase registration is performed to align the images in the time dimension. The image is standardized to remove the existing radiation distortion and geometric distortion. Image enhancement techniques, such as histogram equalization, are applied to improve the visual effect and contrast of the image. A specific spectral super-resolution reconstruction algorithm is selected, such as a convolutional neural network (CNN) based on deep learning, a generative adversarial network (GAN), a sparse representation, etc. The selected algorithm is trained or adjusted to adapt to specific remote sensing data and application scenarios. The preprocessed image is input into the spectral super-resolution reconstruction model. The model generates a high-resolution reconstructed image by processing the low-resolution image. The reconstructed image is evaluated to ensure that the reconstruction quality meets the expected standard. The generated high-resolution image is post-processed to further remove noise and artifacts. Image sharpening and detail enhancement are performed to improve the clarity and resolution of the image. The reconstructed high-resolution images are collected and sorted. The quality and consistency of the image data are ensured to provide a basis for hyperspectral reconstruction. Select a specific hyperspectral reconstruction algorithm, such as sparse representation, dictionary learning, deep learning model, etc. If there is training data, the model can be trained through supervised learning; if there is no training data, unsupervised learning or semi-supervised learning methods can be used. Input the reconstructed high-resolution image into the hyperspectral reconstruction model. The model generates a hyperspectral reconstructed image by processing the image, which contains richer spectral information. Evaluate the reconstructed image to ensure the accuracy and completeness of the spectral information. Post-process the generated hyperspectral image, such as noise removal and spectral smoothing. Extract and analyze spectral features to ensure that the spectral information of the image can meet the application requirements.

[0074] Step S3: Calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with a preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, perform hyperspectral feature data on the corresponding satellite remote sensing observation spectral feature data; use the hyperspectral feature data to perform atmospheric regional interference optimization on the satellite remote sensing spectral feature map to generate a satellite remote sensing atmospheric optimization image;

[0075] In an embodiment of the present invention, a specific spectral index is selected according to application requirements, such as the Normalized Difference Vegetation Index (NDVI), the Enhanced Vegetation Index (EVI), the Soil Adjusted Vegetation Index (SAVI), etc. The selected spectral index is calculated by a formula. Taking NDVI as an example, the calculation formula is: NDVI = (NIR-Red) / (NIR + Red); wherein NIR and Red are the reflectances of the near-infrared and red light bands, respectively. For each pixel of the hyperspectral reconstructed image, the selected spectral index formula is calculated to generate a corresponding spectral index image. The spectral index threshold is set according to the research purpose and historical data analysis. For example, the NDVI threshold can be set to 0.3 to identify vegetation coverage areas. The spectral index value of each pixel is compared with a preset threshold. If the spectral index value is greater than or equal to the threshold, the hyperspectral feature data of the pixel is extracted. The hyperspectral feature data includes reflectance values ​​of multiple spectral bands. A specific atmospheric correction model is selected, such as the radiative transfer model (RTM), the dark pixel method (DOS), the fast atmospheric correction algorithm (QUAC), etc. If there is ground measurement data, it can be used to verify and adjust the correction model. The extracted hyperspectral feature data is input into the selected atmospheric correction model. The atmospheric correction model removes atmospheric interference and restores the true reflectance of the surface by inverting atmospheric components (such as water vapor, aerosols, etc.). Generate optimized remote sensing images through atmospherically corrected hyperspectral data. Visually and quantitatively evaluate the optimized images to ensure that atmospheric interference is effectively removed and image quality is improved.

[0076] Step S4: Use edge computing technology to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage to obtain the satellite remote sensing atmosphere storage image; perform distributed processing on the satellite remote sensing atmosphere storage image to execute the processing operation of the satellite remote sensing image data.

[0077] In an embodiment of the present invention, edge computing devices are deployed at ground receiving stations or locations close to data collection points. These devices have powerful computing and storage capabilities and can perform preliminary data processing and optimization. Satellite remote sensing atmospheric optimization images are transmitted from the receiving device to the edge computing device. Data preprocessing is performed on the edge computing device, including compression, format conversion, metadata addition, etc., to facilitate subsequent storage and processing. The atmospheric optimization image is encrypted to ensure the security of the data during transmission. The encrypted image data is uploaded to the cloud platform using a secure transmission protocol (such as HTTPS, FTPS). The preprocessed and encrypted image data is uploaded to the storage system of the cloud platform. Cloud storage services (such as Amazon S3, Google Cloud Storage, Azure Blob Storage) can be used for storage. Ensure the integrity and reliability of the data upload process and verify it through a checksum or hash value. The uploaded image data is stored and managed on the cloud platform, and specific storage policies and permission controls are set. Create and maintain a data directory structure for subsequent data retrieval and management. Distributed computing resources such as virtual machines, containers, distributed storage systems, etc. are configured on the cloud platform. Select a specific distributed computing framework (such as Hadoop, Spark, Flink) to support large-scale data processing. Use the task scheduling system of the cloud platform to assign image data processing tasks to different computing nodes. Dynamically adjust computing resources to ensure efficient execution of processing tasks and rational use of resources. Divide the atmospheric storage image data into blocks and assign these blocks to different computing nodes for parallel processing. Perform image data processing jobs on each computing node, including image enhancement, classification, target detection, change detection, etc. Integrate the processing results of distributed computing nodes to generate complete processing results. Further analyze and process the integrated data to extract useful information and features. Store the processing results on the cloud platform for subsequent access and application. Use the visualization tools of the cloud platform (such as Tableau, PowerBI, Google Data Studio) to visualize the processing results and generate reports and charts.

[0078] Preferably, step S1 comprises the following steps:

[0079] Step S11: Acquire satellite remote sensing observation signals;

[0080] Step S12: Decoding the satellite remote sensing observation signal to generate satellite remote sensing observation data;

[0081] Step S13: performing radiation correction on the satellite remote sensing observation data to generate satellite remote sensing observation radiation correction data; performing geometric correction on the satellite remote sensing observation radiation correction data to generate satellite remote sensing observation geometric correction data; performing data denoising on the satellite remote sensing observation geometric correction data to generate satellite remote sensing observation denoised data;

[0082] Step S14: performing remote sensing image integration on the satellite remote sensing observation denoised data to generate a satellite remote sensing observation image.

[0083] In an embodiment of the present invention, by starting the remote sensing equipment (such as multispectral sensor, hyperspectral sensor, radar, etc.) on the satellite, the observation signal of the ground or the ocean is collected. The satellite communication link is used to transmit the collected original observation signal to the ground receiving station. The observation signal from the satellite is received using a high-gain antenna. The received signal is preliminarily processed to filter out noise and interference to ensure signal quality. The received signal is decoded using special decoding software and algorithms to extract valid observation data. The decoding process includes steps such as synchronization, error correction, and data packet parsing. The decoded observation data is converted into a standard format, such as GeoTIFF, HDF5, etc., for subsequent processing and analysis. The observation data is radiometrically corrected using a radiation correction model (such as 6S model, MODTRAN, etc.) to remove the influence of the atmosphere and instruments on the observation signal. The observation data after radiation correction is generated to accurately reflect the surface reflectivity. The radiation correction data is geometrically corrected to correct the geometric distortion in the image. Image registration and reprojection are performed using geographic reference data (such as DEM, GPS data) to ensure the spatial accuracy of the image. The observation data after geometric correction is generated. Use filtering methods (such as mean filtering, median filtering, wavelet transform, etc.) to denoise the geometrically corrected data. Remove random noise and systematic noise from the observation data to generate denoised observation data. Fuse observation data from multiple time points or multiple sensors to generate comprehensive remote sensing data. Data fusion can improve the spatial resolution, spectral resolution, and temporal resolution of the image. Use image processing software to convert the fused data into an image for splicing, correction, and registration. Apply color coding and enhancement techniques to enhance the visual effect of the image. Perform geometric and radiometric calibration on the generated remote sensing image to ensure its accuracy and consistency. Verify and calibrate the image by comparing it with existing high-precision maps or field measurement data.

[0084] Preferably, step S14 includes the following steps:

[0085] Step S141: performing spectral extraction on the satellite remote sensing observation denoised data to obtain satellite remote sensing spectral data; performing spectral band analysis on the satellite remote sensing spectral data to generate satellite remote sensing spectral band data;

[0086] Step S142: using true color synthesis technology to perform spectral image synthesis on satellite remote sensing spectral band data to generate a satellite remote sensing spectral synthesis image;

[0087] Step S143: performing image preprocessing on the satellite remote sensing spectral composite image to generate a standard satellite remote sensing spectral composite image, wherein the image preprocessing includes histogram averaging, image brightness enhancement and contrast stretching;

[0088] Step S144: Perform geographic information encoding on the standard satellite remote sensing spectrum composite image according to GIS technology to generate a satellite remote sensing observation image.

[0089] In an embodiment of the present invention, spectral data is generated by extracting the reflectance value of each spectral band from the denoised satellite remote sensing data. Using spectral separation technology, data of different bands are independently extracted to form multiple single-band images. The extracted spectral data is analyzed to identify the characteristics and applications of each band. For example, the near infrared (NIR) band is analyzed for vegetation index calculation, and the mid-infrared (SWIR) band is analyzed for moisture content monitoring. Spectral band data containing characteristic information of each band is generated. Bands suitable for true color synthesis are selected, usually including red (Red), green (Green) and blue (Blue) bands. Ensure that the selected bands can accurately reflect the true color and characteristics of the ground object. Use a true color synthesis algorithm to combine the data of the red, green and blue bands to generate a true color image. Ensure that the data of each band is aligned in space to avoid color difference and distortion. Adjust the hue and saturation of the synthesized image to make its visual effect more natural. Perform histogram averaging on the synthesized image to balance the grayscale distribution of the image. Enhance the contrast of the image to make the distinction between different ground objects more obvious. Adjust the brightness of the image to ensure the visibility of the image under different lighting conditions. Use brightness enhancement algorithms, such as adaptive histogram equalization (CLAHE), to improve the brightness and details of the image. Perform contrast stretching on the image to expand the grayscale range of the image. Use linear or nonlinear contrast stretching algorithms to make the dark and bright parts of the image clearer. Align the preprocessed standard satellite remote sensing image with geographic reference data (such as DEM, GPS data). Use geographic registration algorithms (such as image registration, spatial reprojection) to ensure the accurate position of the image in the geographic coordinate system. Encode geographic information (such as latitude and longitude, elevation, and feature category) into the image to generate a geocoded image. Use geographic information system (GIS) software (such as ArcGIS, QGIS) for encoding and management. Save the geocoded image in a standard format (such as GeoTIFF) to ensure its compatibility and accessibility. Verify and quality check the generated satellite remote sensing observation image to ensure its geographic and radiometric accuracy.

[0090] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0091] Step S21: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image;

[0092] Step S22: performing spectral interpolation on the satellite remote sensing reconstructed image to generate a satellite remote sensing spectral interpolation image;

[0093] Step S23: performing hyperspectral reconstruction on the satellite remote sensing spectral interpolation image to generate a satellite remote sensing hyperspectral reconstructed image.

[0094] In an embodiment of the present invention, spectral super-resolution reconstruction aims to recover high spectral resolution details from low-resolution multispectral data. A super-resolution reconstruction algorithm (such as a learning-based method, a regularization method, etc.) is used to enhance low-resolution spectral information to high resolution. Reconstruction is performed based on the correlation between pixels and spectral features to improve the spatial and spectral resolution of the image. Spectral interpolation is performed on the image after super-resolution reconstruction to fill in the existing spectral missing or incompleteness. Interpolation operations are performed on the spectral dimension using interpolation algorithms (such as bilinear interpolation, polynomial interpolation, spline interpolation, etc.). The purpose is to generate complete and continuous spectral information so that the image has a uniform distribution and continuity in each spectral band. Hyperspectral reconstruction refers to extracting higher-dimensional spectral information from the interpolated spectral image. Using a hyperspectral reconstruction algorithm (such as principal component analysis, linear unmixing, non-negative matrix decomposition, etc.), more spectral features are extracted from the interpolated image. The reconstructed image can contain richer spectral information than the original image, which is used for fine object classification and spectral analysis. A deep learning model, such as a convolutional neural network (CNN), is used to train and reconstruct low-resolution multispectral images. The input is multiple bands of a low-resolution image, and the output is a reconstructed high-resolution spectral image. The correlation of spectral features and local information in space are considered during model training. Interpolation is performed on the high-resolution spectral image reconstructed by deep learning to fill the gaps and discontinuities in the spectrum. Bilinear interpolation or other interpolation methods are used to ensure the continuity and consistency of the spectrum across the entire band. Principal component analysis (PCA) or other hyperspectral reconstruction techniques are used to extract more spectral information from the interpolated image. The output hyperspectral reconstructed image has higher dimensions and richer spectral features, which is suitable for complex land object classification and spectral analysis tasks. Regularization methods (such as Tikhonov regularization, total variation regularization, etc.) are used to perform super-resolution reconstruction of low-resolution multispectral images. By optimizing the regularization parameters, the spectral and spatial fidelity of the reconstructed image are balanced. Spectral interpolation operations are performed on the reconstructed high-resolution image to fill the gaps in the spectrum and ensure continuity. Specific interpolation techniques are used to ensure the uniform distribution of the image in the spectrum and the continuity of the spectral bands. Use linear unmixing or other hyperspectral reconstruction techniques to extract richer spectral information from the interpolated image. Generate a hyperspectral reconstructed image suitable for applications such as land object classification and land use monitoring.

[0095] Preferably, performing hyperspectral reconstruction on the satellite remote sensing spectral interpolation image includes:

[0096] The spectral regional resolution analysis is performed on the satellite remote sensing spectral interpolation image to generate the satellite remote sensing spectral image resolution data; the resolution of the satellite remote sensing spectral image resolution data is restored according to the non-negative matrix decomposition to generate the satellite remote sensing hyperspectral reconstructed image.

[0097] In an embodiment of the present invention, a resolution analysis is performed on a satellite remote sensing image after spectral interpolation processing. The spatial resolution of each spectral band, that is, the resolution of the spectral region, is determined. The spatial resolution and spectral resolution of each band are calculated for subsequent hyperspectral reconstruction processing. The resolution data of the satellite remote sensing spectral image is represented as a non-negative matrix. Using the NMF technology, the spectral data matrix is ​​decomposed into the product of two non-negative matrices, one of which represents the spectral component and the other represents the spatial component. On the basis of the NMF decomposition, the hyperspectral information of the image is restored by reconstructing the spectral components and spatial components. The spectral component matrix and the spatial component matrix reconstructed by NMF are used to generate a hyperspectral reconstructed image. It is ensured that the reconstructed image has high resolution and accurate spectral information in the spectrum, which can support more accurate land object classification and environmental monitoring tasks.

[0098] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0099] Step S31: performing spectral decomposition on the satellite remote sensing hyperspectral reconstructed image to generate satellite remote sensing observation spectral feature data;

[0100] Step S32: Calculate the spectral index of the satellite remote sensing observation spectral feature data to obtain the satellite remote sensing spectral index;

[0101] Step S33: Compare the satellite remote sensing spectral index with a preset spectral index threshold value. When the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold value, perform hyperspectral marking on the corresponding satellite remote sensing observation spectral feature data to generate hyperspectral feature data; when the satellite remote sensing spectral index is less than the preset spectral index threshold value, perform spectral feature combination on the hyperspectral feature data to generate a satellite remote sensing spectral feature map; perform atmospheric transmittance evaluation on the satellite remote sensing observation denoised data using the satellite remote sensing spectral feature map to generate atmospheric transmittance evaluation data;

[0102] Step S34: Optimizing the atmospheric regional interference of the satellite remote sensing spectral characteristic map based on the atmospheric transmittance evaluation data to generate a satellite remote sensing atmospheric optimization image.

[0103] In an embodiment of the present invention, a satellite remote sensing hyperspectral reconstructed image is spectrally decomposed by using techniques such as principal component analysis (PCA) or factor analysis. Complex hyperspectral data is reduced to fewer feature dimensions to extract key spectral features. Based on the spectral feature data obtained from the spectral decomposition, various spectral indices, such as vegetation index (such as NDVI), soil index, etc., are calculated. Spectral indexes usually describe specific attributes of target objects by combining bands, such as vegetation health or soil moisture. The calculated spectral index is compared with a preset spectral index threshold. When the spectral index is greater than or equal to the preset threshold, the corresponding spectral feature data is marked as hyperspectral feature data for subsequent analysis. When the spectral index is less than the preset threshold, the corresponding spectral feature data is removed to ensure the accuracy and reliability of the analysis. The marked hyperspectral feature data is combined into a satellite remote sensing spectral feature map. These data reflect the spectral characteristics of the target object under specific spectral index conditions. Atmospheric transmittance is evaluated using satellite remote sensing observation denoised data and previously calculated spectral feature maps. The influence of the atmosphere on the spectral features is analyzed to evaluate the degree of interference in the atmospheric region. According to the results of atmospheric transmittance assessment, the satellite remote sensing spectral characteristic map is optimized to reduce or correct the interference in the atmospheric region. The satellite remote sensing atmospheric optimized image is generated after atmospheric optimization processing to improve the image quality and availability.

[0104] Preferably, using the satellite remote sensing spectral characteristic map to evaluate the atmospheric transmittance of the satellite remote sensing observation denoised data includes:

[0105] Collect historical data of satellite remote sensing observation denoised data to obtain satellite remote sensing observation historical data;

[0106] The spectral characteristics impact analysis is performed on the satellite remote sensing spectral characteristic map through the 6S radiation transfer model to generate initial atmospheric composition data; the atmospheric transmittance is calculated on the initial atmospheric composition data to obtain atmospheric transmittance data; the atmospheric assessment of the atmospheric transmittance data is optimized through satellite remote sensing observation denoising data and satellite remote sensing observation historical data to generate atmospheric transmittance assessment data.

[0107] In the embodiment of the present invention, historical records of satellite remote sensing observation denoising data are collected and sorted. It is ensured that the historical data contains a sufficiently long time span and a comprehensive geographical coverage to support the subsequent atmospheric transmittance evaluation and optimization. The 6S model (Second Simulation of the Satellite Signal in the Solar Spectrum) is used to perform spectral feature impact analysis. The satellite remote sensing spectral feature map is input to simulate the radiation transmission process of different bands, including atmospheric absorption, scattering and reflection effects. The degree of influence of the spectral characteristics of each band on the atmosphere is obtained according to the model calculation. Based on the spectral feature impact analysis results simulated by the 6S model, the initial atmospheric composition data is derived. The atmospheric transmittance of each band is calculated using these initial data, that is, the ratio of the atmosphere to the radiation transmitted in a specific band. Considering the absorption, scattering and transmission characteristics of the earth's atmosphere in different bands, the accuracy and reliability of the calculation are ensured. The calculated atmospheric transmittance data is compared and verified with the actual satellite remote sensing observation denoising data and historical data. The deviations and errors in the data are analyzed, and the atmospheric transmittance data is optimized and adjusted to improve its accuracy and applicability. Generate the final atmospheric transmittance assessment data for subsequent satellite remote sensing data processing and analysis.

[0108] Preferably, performing atmospheric regional interference optimization on the satellite remote sensing spectral characteristic map based on the atmospheric transmittance evaluation data comprises:

[0109] Conduct time-series image acquisition of satellite remote sensing spectral characteristic images to obtain satellite remote sensing atmospheric multi-temporal images;

[0110] Perform time series alignment on satellite remote sensing atmospheric multi-temporal images to generate satellite remote sensing atmospheric multi-temporal phase alignment images; perform atmospheric change region edge segmentation on satellite remote sensing atmospheric multi-temporal phase alignment images to generate atmospheric change edge region images;

[0111] Based on the atmospheric transmittance assessment data, the visible area of ​​the atmospheric change edge area image is screened to obtain the atmospheric change visible area image; according to the atmospheric change visible area image, the atmospheric interference area of ​​the satellite remote sensing spectral characteristic map is filled to generate the satellite remote sensing atmospheric optimization image.

[0112] In an embodiment of the present invention, multiple time-series images are obtained from satellite remote sensing data to cover observations at different time points or time periods. Ensure that the time-series images contain sufficient time span and frequency to capture dynamic information of atmospheric composition and changes. Align the collected time-series images to ensure that the images at different time points correspond in space and spectrum. Use image registration technology, such as a method based on feature point matching or a method based on image histogram, to ensure accurate alignment of the time-series images. On the aligned time-series images, use edge detection or segmentation algorithms to identify and extract the edges of the atmospheric change area. Edge segmentation can determine the boundaries of the atmospheric change area based on pixel-level differences, color changes or other features. Use the atmospheric transmittance evaluation data calculated previously to analyze and screen the images of the edge areas of atmospheric changes. Determine those areas where the atmospheric transmittance changes significantly as visible areas of atmospheric changes. Based on the determined visible area images of atmospheric changes, fill or correct the atmospheric interference areas in the satellite remote sensing spectral feature map. Use image repair or filling algorithms, such as interpolation methods based on neighboring pixels or repair methods based on image content, to fill in the missing or distortion caused by atmospheric interference.

[0113] As an example of the present invention, refer to Figure 4 As shown, in this example, step S4 includes:

[0114] Step S41: using edge computing technology to compress the satellite remote sensing atmospheric optimization image to generate a satellite remote sensing atmospheric compression image;

[0115] Step S42: Upload the satellite remote sensing atmospheric compressed image to the cloud platform for image storage to obtain the satellite remote sensing atmospheric storage image; perform distributed processing on the satellite remote sensing atmospheric storage image to execute the processing operation of the satellite remote sensing image data.

[0116] In an embodiment of the present invention, the satellite remote sensing image that has been processed by atmospheric optimization is compressed by using edge computing technology. Compression can use standard image compression algorithms, such as JPEG, JPEG-2000 or WebP, to reduce the size of the image file and maintain image quality. Edge computing capabilities ensure that compression is performed immediately close to the data source, reducing data transmission and storage costs. The compressed satellite remote sensing atmospheric optimization image is uploaded to the cloud platform for storage. Cloud storage services (such as AWS S3, Azure Blob Storage or Google Cloud Storage) are used for safe and reliable image storage management. Distributed processing is performed on the cloud platform for the stored satellite remote sensing atmospheric images. Distributed processing can include large-scale data analysis, image processing, feature extraction and other operations. The computing resources of the cloud platform and distributed computing frameworks (such as Hadoop and Spark) are used to perform efficient data processing and analysis tasks.

[0117] In this specification, a satellite remote sensing image data processing system is provided, which is used to execute the above-mentioned satellite remote sensing image data processing method. The satellite remote sensing image data processing system includes:

[0118] The remote sensing signal conversion module is used to obtain satellite remote sensing observation signals; decode the satellite remote sensing observation signals to generate satellite remote sensing observation data; perform remote sensing image integration on the satellite remote sensing observation denoising data to generate satellite remote sensing observation images;

[0119] The spectral reconstruction module is used to perform spectral super-resolution reconstruction on satellite remote sensing observation images to generate satellite remote sensing reconstructed images; perform hyperspectral reconstruction on satellite remote sensing reconstructed images to generate satellite remote sensing hyperspectral reconstructed images;

[0120] The atmospheric interference optimization module is used to calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with the preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, the corresponding satellite remote sensing observation spectral feature data is subjected to hyperspectral feature data; the satellite remote sensing spectral feature map is subjected to atmospheric regional interference optimization using the hyperspectral feature data to generate a satellite remote sensing atmospheric optimization image;

[0121] The image storage module is used to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage using edge computing technology to obtain the satellite remote sensing atmosphere storage image; the satellite remote sensing atmosphere storage image is distributedly processed to execute the processing operation of the satellite remote sensing image data.

[0122] The beneficial effect of the present invention is that by obtaining the original satellite remote sensing data, basic data is provided for subsequent processing. Spectral super-resolution reconstruction and hyperspectral reconstruction are performed on satellite remote sensing observation images. These technologies can improve the resolution and spectral information of the image, making the observation image clearer and more detailed, and help to extract ground object information more accurately. The spectral index of the satellite remote sensing hyperspectral reconstruction image is calculated and compared with the preset spectral index threshold. This step can provide spectral index information for analyzing and evaluating environmental indicators such as vegetation growth and soil quality in different regions, which is helpful for research and decision-making in the fields of agriculture and environmental monitoring. The satellite remote sensing atmospheric optimization image is uploaded to the cloud platform and stored using edge computing technology, and then distributed processing is performed. This processing method can make full use of the computing and storage resources of the cloud platform, improve processing efficiency and data processing capabilities, and provide users with fast and efficient satellite remote sensing data processing services. Therefore, the present invention improves the accuracy and quality of satellite remote sensing images through systematic decoding, denoising, super-resolution reconstruction, hyperspectral reconstruction, spectral index calculation and optimization, edge computing and cloud storage.

[0123] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0124] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for processing satellite remote sensing image data, characterized in that: The following steps are involved: Step S1: acquiring satellite remote sensing observation signals; decoding the satellite remote sensing observation signals to generate satellite remote sensing observation data; integrating the satellite remote sensing observation denoised data to generate satellite remote sensing observation images; Step S2: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image; performing hyperspectral reconstruction on the satellite remote sensing reconstructed image to generate a satellite remote sensing hyperspectral reconstructed image; Step S3: Calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with a preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, perform hyperspectral feature data on the corresponding satellite remote sensing observation spectral feature data; Using the hyperspectral feature data, the satellite remote sensing spectral feature map is subjected to atmospheric regional interference optimization to generate a satellite remote sensing atmospheric optimization image; step S3 comprises the following steps: Step S31: performing spectral decomposition on the satellite remote sensing hyperspectral reconstructed image to generate satellite remote sensing observation spectral feature data; Step S32: Calculate the spectral index of the satellite remote sensing observation spectral feature data to obtain the satellite remote sensing spectral index; Step S33: Compare the satellite remote sensing spectral index with a preset spectral index threshold value. When the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold value, perform hyperspectral marking on the corresponding satellite remote sensing observation spectral feature data to generate hyperspectral feature data; when the satellite remote sensing spectral index is less than the preset spectral index threshold value, perform spectral feature combination on the hyperspectral feature data to generate a satellite remote sensing spectral feature map; perform atmospheric transmittance evaluation on the satellite remote sensing observation denoised data using the satellite remote sensing spectral feature map to generate atmospheric transmittance evaluation data; Step S34: Optimizing the atmospheric regional interference of the satellite remote sensing spectral characteristic graph based on the atmospheric transmittance evaluation data to generate a satellite remote sensing atmospheric optimization image; Optimizing the atmospheric regional interference of the satellite remote sensing spectral characteristic graph based on the atmospheric transmittance evaluation data includes: Conduct time-series image acquisition of satellite remote sensing spectral characteristic images to obtain satellite remote sensing atmospheric multi-temporal images; Perform time series alignment on satellite remote sensing atmospheric multi-temporal images to generate satellite remote sensing atmospheric multi-temporal phase alignment images; perform atmospheric change region edge segmentation on satellite remote sensing atmospheric multi-temporal phase alignment images to generate atmospheric change edge region images; Based on the atmospheric transmittance assessment data, the visible area of ​​the atmospheric change edge area image is screened to obtain the atmospheric change visible area image; based on the atmospheric change visible area image, the satellite remote sensing spectral characteristic map is filled with atmospheric interference areas to generate the satellite remote sensing atmospheric optimization image; Step S4: Use edge computing technology to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage to obtain the satellite remote sensing atmosphere storage image; perform distributed processing on the satellite remote sensing atmosphere storage image to execute the processing operation of the satellite remote sensing image data.

2. The method for processing satellite remote sensing image data according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire satellite remote sensing observation signals; Step S12: Decoding the satellite remote sensing observation signal to generate satellite remote sensing observation data; Step S13: performing radiation correction on the satellite remote sensing observation data to generate satellite remote sensing observation radiation correction data; performing geometric correction on the satellite remote sensing observation radiation correction data to generate satellite remote sensing observation geometric correction data; performing data denoising on the satellite remote sensing observation geometric correction data to generate satellite remote sensing observation denoised data; Step S14: performing remote sensing image integration on the satellite remote sensing observation denoised data to generate a satellite remote sensing observation image.

3. The method for processing satellite remote sensing image data according to claim 2, characterized in that: Step S14 includes the following steps: Step S141: performing spectral extraction on the satellite remote sensing observation denoised data to obtain satellite remote sensing spectral data; performing spectral band analysis on the satellite remote sensing spectral data to generate satellite remote sensing spectral band data; Step S142: using true color synthesis technology to perform spectral image synthesis on satellite remote sensing spectral band data to generate a satellite remote sensing spectral synthesis image; Step S143: performing image preprocessing on the satellite remote sensing spectral composite image to generate a standard satellite remote sensing spectral composite image, wherein the image preprocessing includes histogram averaging, image brightness enhancement and contrast stretching; Step S144: Perform geographic information encoding on the standard satellite remote sensing spectrum composite image according to GIS technology to generate a satellite remote sensing observation image.

4. The method for processing satellite remote sensing image data according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing spectral super-resolution reconstruction on the satellite remote sensing observation image to generate a satellite remote sensing reconstructed image; Step S22: performing spectral interpolation on the satellite remote sensing reconstructed image to generate a satellite remote sensing spectral interpolation image; Step S23: performing hyperspectral reconstruction on the satellite remote sensing spectral interpolation image to generate a satellite remote sensing hyperspectral reconstructed image.

5. The method for processing satellite remote sensing image data according to claim 4, characterized in that: Hyperspectral reconstruction of satellite remote sensing spectral interpolation images includes: The spectral regional resolution analysis is performed on the satellite remote sensing spectral interpolation image to generate the satellite remote sensing spectral image resolution data; the resolution of the satellite remote sensing spectral image resolution data is restored according to the non-negative matrix decomposition to generate the satellite remote sensing hyperspectral reconstructed image.

6. The method for processing satellite remote sensing image data according to claim 1, characterized in that: The atmospheric transmittance assessment of satellite remote sensing observation denoised data using satellite remote sensing spectral characteristic maps includes: Collect historical data of satellite remote sensing observation denoised data to obtain satellite remote sensing observation historical data; The spectral characteristics impact analysis is performed on the satellite remote sensing spectral characteristic map through the 6S radiation transfer model to generate initial atmospheric composition data; the atmospheric transmittance is calculated on the initial atmospheric composition data to obtain atmospheric transmittance data; the atmospheric assessment of the atmospheric transmittance data is optimized through satellite remote sensing observation denoising data and satellite remote sensing observation historical data to generate atmospheric transmittance assessment data.

7. The method for processing satellite remote sensing image data according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: using edge computing technology to compress the satellite remote sensing atmospheric optimization image to generate a satellite remote sensing atmospheric compression image; Step S42: Upload the satellite remote sensing atmospheric compressed image to the cloud platform for image storage to obtain the satellite remote sensing atmospheric storage image; perform distributed processing on the satellite remote sensing atmospheric storage image to execute the processing operation of the satellite remote sensing image data.

8. A satellite remote sensing image data processing system, characterized in that: For executing the satellite remote sensing image data processing method as claimed in claim 1, the satellite remote sensing image data processing system comprises: The remote sensing signal conversion module is used to obtain satellite remote sensing observation signals; decode the satellite remote sensing observation signals to generate satellite remote sensing observation data; perform remote sensing image integration on the satellite remote sensing observation denoising data to generate satellite remote sensing observation images; The spectral reconstruction module is used to perform spectral super-resolution reconstruction on satellite remote sensing observation images to generate satellite remote sensing reconstructed images; perform hyperspectral reconstruction on satellite remote sensing reconstructed images to generate satellite remote sensing hyperspectral reconstructed images; The atmospheric interference optimization module is used to calculate the spectral index of the satellite remote sensing hyperspectral reconstructed image to obtain the satellite remote sensing spectral index; compare the satellite remote sensing spectral index with the preset spectral index threshold, and when the satellite remote sensing spectral index is greater than or equal to the preset spectral index threshold, the corresponding satellite remote sensing observation spectral feature data is subjected to hyperspectral feature data; the satellite remote sensing spectral feature map is subjected to atmospheric regional interference optimization using the hyperspectral feature data to generate a satellite remote sensing atmospheric optimization image; The image storage module is used to upload the satellite remote sensing atmosphere optimization image to the cloud platform for image storage using edge computing technology to obtain the satellite remote sensing atmosphere storage image; the satellite remote sensing atmosphere storage image is distributedly processed to execute the processing operation of the satellite remote sensing image data.

Images