Systems and methods for processing of satellite image data

Abstract

A computer-implemented method of processing satellite image data for an area of interest (AOI) is provided. The method comprises: retrieving, by a processor, the satellite image data based on location coordinates of the AOI; generating, by the processor, ground deformation data based on the retrieved satellite image data; post-processing, by the processor, the generated ground deformation data to remove deformation data point irrelevant to the AOI; and providing, by the processor, a ground deformation analysis output based on post-processing the generated ground deformation data.

Description

[0001] Title: SYSTEMS AND METHODS FOR PROCESSING OF SATELLITE IMAGE DATA

[0002] Cross-Reference to Related Application

[0003] [1] This application claims the benefit of United States Provisional Patent Application No. 63 / 670,495 filed July 12, 2024, and the contents of United States Provisional Patent Application No. 63 / 670,495 is incorporated herein in its entirety.

[0004] Field

[0005] [2] The described embodiments relate to systems and methods for processing satellite image data.

[0006] Background

[0007] [3] Satellite image analysis is used in many industrial applications and can provide various insights into Earth’s surface dynamics. The satellite image data may be provided by any suitable satellite system. As one example, Sentinel-1 is a satellite system created by the European Union to monitor Earth's surface using radar technology. The Sentinel- 1 system can operate continuously (day / night), regardless of weather conditions, to provide continuous imaging of the planet's surface. The Sentinel-1 system can use radar to “see” through clouds, rain, and darkness, capturing detailed images of the Earth's surface. The radar images provide by the Sentinel-1 system or other satellite systems can help monitor changes in the environment, such as tracking land movements.

[0008] Summary

[0009] [4] The following summary is provided to introduce the reader to the more detailed discussion to follow. The summary is not intended to limit or define any claimed or as yet unclaimed invention. One or more inventions may reside in any combination or subcombination of the elements or process steps disclosed in any part of this document including its claims and figures.

[0010] [5] In a first aspect, there is provided a computer-implemented method of processing satellite image data for an area of interest (AOI). The method comprises: retrieving, by a processor, the satellite image data based on location coordinates of the AOI; generating, by the processor, ground deformation data based on the retrieved satellite image data; post-processing, by the processor, the generated ground deformation data to remove deformation data point irrelevant to the AOI; and providing, by the processor, a ground deformation analysis output based on post-processing the generated ground deformation data.

[0011] [6] In one or more embodiments, retrieving the satellite image data is further based on a coverage threshold defined for the AOI.

[0012] [7] In one or more embodiments, the AOI includes railway infrastructure.

[0013] [8] In one or more embodiments, retrieving the satellite image data includes filtering out irrelevant images from the satellite image data based on one or more parameters associated with the satellite image data.

[0014] [9] In one or more embodiments, the one or more parameters include an orbit type, a resolution or a swath width associated with the satellite image data.

[0015]

[0010] In one or more embodiments, the retrieved satellite image data includes synthetic aperture radar (SAR) images and generating the ground deformation data includes generating an interferogram based on the SAR images.

[0016]

[0011] In one or more embodiments, generating the ground deformation data further includes identifying and mitigating noise sources in the SAR images.

[0017]

[0012] In one or more embodiments, the method further comprises using a trained Convolutional Neural Network (CNN) to identify and mitigate the noise sources.

[0018]

[0013] In one or more embodiments, generating the ground deformation data further includes inputting the interferogram and associated metadata into a trained neural network based on a reslI-Net architecture.

[0019]

[0014] In one or more embodiments, post-processing the generated ground deformation data includes removing anomalous deformation data points and deformation data points corresponding to water bodies.

[0020]

[0015] In one or more embodiments, post-processing the generated ground deformation data includes using a haversine distance calculation to include deformation data points relevant to the AOI and exclude other irrelevant deformation data points.

[0016] In one or more embodiments, post-processing the generated ground deformation data includes retrieving climate and / or weather data corresponding to the generated ground deformation data.

[0021]

[0017] In one or more embodiments, providing the ground deformation analysis output includes providing a preventative maintenance recommendation for an infrastructure located in the AOI.

[0022]

[0018] In a second aspect, there is provided a system for processing satellite image data for an area of interest (AOI). The system comprises: a memory storing processorexecutable instructions; and a processor communicatively coupled to the memory, the instructions configuring the processor to perform any of the methods described herein.

[0023]

[0019] In a third aspect, there is provided a non-transitory computer readable medium storing thereon program instructions that are executable by a processor for performing a any of the methods described herein of processing satellite image data for an area of interest (AOI).

[0024] Brief Description of the Drawings

[0025]

[0020] The drawings included herewith are for illustrating various examples of systems, methods, and devices of the teaching of the present specification and are not intended to limit the scope of what is taught in any way.

[0026]

[0021] FIG. 1 is a schematic diagram showing a system for processing satellite image data, in accordance with one or more embodiments.

[0027]

[0022] FIG. 2 is a block diagram showing components of the system of FIG. 1 , in accordance with one or more embodiments.

[0028]

[0023] FIG. 3 is a flowchart showing a computer-implemented method of processing satellite image data, in accordance with one or more embodiments.

[0029]

[0024] FIG. 4 is an example ground deformation analysis output generated by the system of FIG. 1 , in accordance with one or more embodiments.

[0030]

[0025] FIG. 5 is another example ground deformation analysis output generated by the system of FIG. 1 , in accordance with one or more embodiments.

[0031]

[0026] FIG. 6A is another example ground deformation analysis output generated by the system of FIG. 1 , in accordance with one or more embodiments.

[0027] FIG. 6B is another example ground deformation analysis output generated by the system of FIG. 1 , in accordance with one or more embodiments.

[0032] Detailed Description

[0033]

[0028] Several example embodiments are described herein. It will be appreciated that numerous specific details are set forth in order to provide a thorough understanding of the example embodiments described herein. However, it will be understood by those of ordinary skill in the art that the embodiments described herein may be practiced without these specific details. In other instances, well-known methods, procedures and components have not been described in detail so as not to obscure the embodiments described herein. Furthermore, this description and the drawings are not to be considered as limiting the scope of the embodiments described herein in any way, but rather as merely describing the implementation of the various embodiments described herein.

[0034]

[0029] It should be noted that terms of degree such as “substantially”, “about” and “approximately” when used herein mean a reasonable amount of deviation of the modified term such that the end result is not significantly changed. These terms of degree should be construed as including a deviation of the modified term if this deviation would not negate the meaning of the term it modifies.

[0035]

[0030] In addition, as used herein, the wording “and / or” is intended to represent an inclusive-or. That is, “X and / or Y” is intended to mean X or Y or both, for example. As a further example, “X, Y, and / or Z” is intended to mean X or Y or Z or any combination thereof.

[0036]

[0031] The embodiments of the systems and methods described herein may be implemented in hardware or software, or a combination of both. These embodiments may be implemented in computer programs executing on programmable computers, each computer including at least one processor, a data storage system (including volatile memory or non-volatile memory or other data storage elements or a combination thereof), and at least one communication interface. For example and without limitation, the programmable computers (referred to below as computing devices) may be a server, network appliance, embedded device, computer expansion module, personal computer, laptop, personal data assistant, cellular telephone, smart-phone device, tablet computer, wireless device or any other computing device capable of being configured to carry out the methods described herein.

[0037]

[0032] In some embodiments, the communication interface may be a network communication interface. In embodiments in which elements are combined, the communication interface may be a software communication interface, such as those for inter-process communication (IPC). In still other embodiments, there may be a combination of communication interfaces implemented as hardware, software, and a combination thereof.

[0038]

[0033] Program code may be applied to input data to perform the functions described herein and to generate output information. The output information is applied to one or more output devices, in known fashion.

[0039]

[0034] Each program may be implemented in a high-level procedural or object-oriented programming and / or scripting language, or both, to communicate with a computer system. However, the programs may be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Each such computer program may be stored on a storage media or a device (e.g. ROM, magnetic disk, optical disc) readable by a general or special purpose programmable computer, for configuring and operating the computer when the storage media or device is read by the computer to perform the procedures described herein. Embodiments of the system may also be considered to be implemented as a non-transitory computer-readable storage medium, configured with a computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner to perform the functions described herein.

[0040]

[0035] Furthermore, the systems, processes and methods of the described embodiments are capable of being distributed in a computer program product comprising a computer readable medium that bears computer usable instructions for one or more processors. The medium may be provided in various forms, including one or more diskettes, compact disks, tapes, chips, wireline transmissions, satellite transmissions, internet transmission or downloads, magnetic and electronic storage media, digital and analog signals, and the like. The computer useable instructions may also be in various forms, including compiled and non-compiled code.

[0036] Various embodiments have been described herein by way of example only. Various modifications and variations may be made to these example embodiments without departing from the spirit and scope of the invention, which is limited only by the appended claims. Also, in the various user interfaces illustrated in the figures, it will be understood that the illustrated user interface text and controls are provided as examples only and are not meant to be limiting. Other suitable user interface elements may be possible.

[0041]

[0037] InSAR (Interferometric Synthetic Aperture Radar) is an analysis technique used in combination with satellite image data to detect ground deformation. By comparing satellite radar images taken at different times, InSAR can identify changes in the surface, such as sinking or rising ground. InSAR images can measure millimeter-scale deformation of surfaces including buildings, coastline, linear infrastructure such as bridges, ground surface along railway tracks, etc. using a time-series multi-image approach. The displacement velocity along with timestamps obtained from the InSAR analysis can be used for proactive maintenance management, optimize on-condition monitoring and improve infrastructure resilience. The disclosed systems, processes and methods can use InSAR techniques to analyze ground deformation trends near locations of interest, for example, near infrastructure locations like railway tracks and bridges. The disclosed systems, processes and methods can provide high resolution and high accuracy ground deformation measurements and predict ground deformation trends.

[0042]

[0038] There can be shortcomings associated with the typical reactive or preventive maintenance strategies used by infrastructure owners / operators to repair and maintain the infrastructure. For example, poor resource allocation can cause maintenance delays. Repairs may only be performed in response to a failure or after the preventive maintenance schedule is overdue. This can cause safety issues, higher repair downtimes (e.g., higher downtime associated with replacement of a failed component compared with conducting timely preventive maintenance of the component), and / or faster degradation of the infrastructure. As another example, excessive preventive maintenance can increase operational costs associated with higher preventive maintenance downtimes, higher labor costs, higher operational challenges (e.g., associated with conducting preventive maintenance in remote locations) and / or lower operational lifetime of the infrastructure components (e.g., associated with replacing components more frequently than necessary).

[0043]

[0039] The disclosed systems, processes and methods can identify current risks and make future predictions related to ground deformation issues, such as land subsidence or uplift, which could affect the infrastructure and / or cause safety issues. This can enable proactive measures to be undertaken to improve the safety, reliability and operational efficiency of various infrastructure, for example, railway infrastructure.

[0044]

[0040] In some cases, processing satellite image data can require a large amount of manual interventions / processing at various stages of the analysis, for example, manual data selection, manual annotations etc. The disclosed systems, processes and methods can improve the efficiency of the data analysis systems by fully autonomous processing of the satellite image data. No manual interventions / processing may be required at any stage of the processing.

[0045]

[0041] The disclosed systems, processes and methods can autonomously process the satellite image data and generate comprehensive and concise analysis reports. The analysis reports may be customized for each user and / or specific location (e.g., specific location of railway infrastructure). Further, the analysis reports may include ground deformation trends analyzed on a grid level basis, coordinate level basis, through 3D plots, longitudinal plots, and / or scatter plots.

[0046]

[0042] The disclosed systems, processes and methods may use one or more Al models to generate insights and / or infrastructure maintenance recommendations based on the processing of the satellite image data. Further, the disclosed systems, processes and methods may use one or more Al models to provide output text in natural language to provide actionable insights to users.

[0047]

[0043] Reference is first made to FIG. 1. FIG. 1 is a schematic diagram showing an example system 100 for processing satellite image data. System 100 can provide an end- to-end ground deformation monitoring system using satellite image data The satellite image data may be acquired by any suitable satellite system 20. For example, satellite system 20 may include the Sentinel-1 satellite system.

[0048]

[0044] User device 30 may include any suitable device usable by a user to provide user inputs to system 100. In some embodiments, user device 30 may be usable by a user to view / receive analysis outputs generated by system 100. User device 30 may include, for example, a desktop computer, a laptop computer, a tablet device, a smartphone, a workstation etc. The user can be, for example, an owner, an operator, a maintenance technician etc. of an infrastructure facility (e.g., railway infrastructure). The analysis outputs generated by system 100 may include, for example, real-time insights, analytics, actionable data, maintenance recommendations etc.

[0049]

[0045] Server 40 may include any suitable networked computing device. Server 40 may provide additional storage capacity and / or computational resources to system 100.

[0050]

[0046] System 100 may communicate with satellite system 20 using network 50. Network 50 may include any suitable communication network. For example, network 50 may include a communication network such as the Internet, a Wide-Area Network (WAN), a Local-Area Network (LAN), or another type of network. In some embodiments, network 50 may include a point-to-point connection, or another communications connection between two nodes. The illustrated example shows a single network 50 connecting system 100, satellite system 20, user device 30 and server 40. In other examples, a combination of different networks may be used to provide interconnections between system 100, satellite system 20, user device 30 and server 40.

[0051]

[0047] Reference is now additionally made to FIG. 2. FIG. 2 is a block diagram showing components of system 100. In the illustrated embodiment, system 100 includes a communication unit 205, a display 210, a processor unit 215, a memory unit 220, an I / O unit 225, a user interface engine 230 and a power unit 235. In other examples, system 100 may include a different combination of components. One or more components of system 100 may be at a remote location and communicate with system 100 using a communication network.

[0052]

[0048] Communication unit 205 can include wired or wireless connection capabilities. Communication unit 205 can be used by system 100 to communicate with other devices or computers. For example, system 100 may use communication unit 205 to receive, via network 50, satellite image data from satellite system 20. As another example, system 100 may use communication unit 205 to provide (e.g., via network 50) recommendations or insights to user device 30.

[0049] Display 210 may be a LED or LCD based display and may be a touch sensitive screen that supports gestures. Display 210 may be integrated into system 100. Alternatively, display 210 may be located physically remote from system 100 and communicate with system 100 using a communication network, for example, network 50. In some embodiments, system 100 may not include a dedicated display 210. For example, system 100 may provide output displays via a display of user device 30.

[0053]

[0050] I / O unit 225 can include at least one of a mouse, a keyboard, a touch screen, a thumbwheel, a trackpad, a trackball, a card-reader, voice recognition software and the like, depending on the particular implementation of system 100. In some embodiments, some of these components can be integrated with one another. I / O unit 225 can enable a user to interact with system 100.

[0054]

[0051] Power unit 235 can be any suitable power source that provides power to system 100 such as a power adaptor or a rechargeable battery pack depending on the implementation of system 100, as is known by those skilled in the art.

[0055]

[0052] Processor unit 215 can control the operation of system 100. Processor unit 215 can be any suitable processor, controller or digital signal processor that can provide sufficient processing power depending on the configuration, purposes and requirements of system 100, as is known by those skilled in the art. For example, processor unit 215 may be a high-performance general processor. For example, processor unit 215 may include a standard processor, such as an Intel® processor, or an AMD® processor. Alternatively, processor unit 215 can include more than one processor with each processor being configured to perform different dedicated tasks. Alternatively, specialized hardware (e.g., graphical processing units (GPUs)) can be used provide some of the functions provided by processor unit 215.

[0056]

[0053] Processor unit 215 can execute a user interface engine 230 that may be used to generate various user interfaces. User interface engine 230 may be configured to provide a user interface on display 210. Optionally, system 100 may be in communication with external displays via network 50, for example, a display of user device 30. User interface engine 230 may generate user interface data for the external displays that are in communication with system 100. 154] User interface engine 230 can be configured to provide a user interface to enable set-up and initialization of system 100. User interface engine 230 can be configured to provide a user interface to receive various user inputs, for example, user inputs related to selection of locations for ground deformation analysis (e.g., locations corresponding to infrastructure like railway tracks). User interface engine 230 can be configured to provide a user interface to provide satellite image analysis outputs to a user of system 100. The satellite image analysis outputs may include, for example, ground deformation trends, real-time insights, analytics, actionable data, maintenance recommendations etc.

[0057]

[0055] Memory unit 220 can include software code for implementing an operating system 240, programs 245, a database 250, a model generation engine 255, a model training engine 260, a satellite image analysis module 265 and a recommendation engine 270. Memory unit 220 can include RAM, ROM, one or more hard drives, one or more flash drives or some other suitable data storage elements such as disk drives, etc.

[0058]

[0056] Memory unit 220 can be used to store an operating system 240 and programs 245, as is commonly known by those skilled in the art. For instance, operating system 240 provides various basic operational processes for system 100. For example, the operating system 240 may be an operating system such as Windows® Server operating system, or Red Hat® Enterprise Linux (RHEL) operating system, or another operating system. Programs 245 can include various programs so that system 100 can perform various functions such as, but not limited to, retrieving satellite image data, perform InSAR analysis, generate analysis reports, etc.

[0059]

[0057] Database 250 may include a Structured Query Language (SQL) database such as PostgreSQL or MySQL or a not only SQL (NoSQL) database such as MongoDB, or Graph Databases, etc. Database 250 may be integrated with system 100. Alternatively, database 250 may run independently on a database server in network communication with system 100.

[0060]

[0058] Database 250 may store satellite image data (e.g., retrieved from satellite system 20) and user inputs received from user device 30. Database 250 may store one or more machine learning models used by system 100. In some embodiments, database 250 may store analysis outputs generated by system 100.

[0061] - I Q -

[0059] Model generation engine 255 may generate one or more machine learning models used by system 100. For example, model generation engine 255 may generate machine learning models for identifying and mitigating noise sources in the synthetic aperture radar (SAR) data, generating deformation maps based on interferogram data, data processing and cleaning for linear infrastructures, generating text annotations (e.g., natural language explanations) for analysis results etc. In some embodiments, system 100 may not include a model generation engine 255. System 100 may receive the machine learning models from an external device (e.g., server 40). The generated and / or received models may be stored in database 250. In some embodiments, the generated and / or received models may be stored in an external storage device that is in network communication with system 100 (e.g., at server 40).

[0062]

[0060] Model training engine 260 may train one or more machine learning models used by system 100. The machine learning models may include models generated by model generation engine 255 and / or models received from external devices.

[0063]

[0061] Model training engine 260 may be configured to perform training of the machine learning models at various times. For example, model training engine 260 may be configured to train the models when they are initially generated by model generation engine 255. As another example, model training engine 260 may be configured to train the models based on a time-based schedule. The time-based schedule may be based on a training time period parameter stored in database 250. As another example, model training engine 260 may be configured to train the models in response to the model output accuracy falling below a threshold level. In some embodiments, model training engine 260 may train a model in response to a training request. The training request may be received, for example, from user device 30.

[0064]

[0062] Model training engine 260 may use any suitable training method to train the machine learning models. The training method may be selected, for example, based on factors including the type of model, available computational resources, accuracy requirements, and / or size / type of training dataset available. Model training engine 260 may use historical data to train the machine learning models. In some embodiments, model training engine 260 may use simulated or synthetic training data to train the machine learning models.

[0063] Satellite image analysis module 265 may include any suitable components (e.g., one or more Al models) to analyze retrieved satellite image data. Satellite image analysis module 265 may generate an interferogram based on the SAR images and generate ground deformation trends based on the interferograms.

[0065]

[0064] Recommendation engine 270 may include any suitable Al or machine learning model trained to generate one or more recommendations based on the analysis results of the satellite image data. The recommendations may include, for example, preventive maintenance recommendations for infrastructure. In some embodiments, recommendation engine 270 may receive climate / weather information / predictions associated with the analyzed locations. Recommendation engine 270 may generate the recommendations based on a combination of the satellite image analysis results and the climate / weather information / predictions.

[0066]

[0065] Reference is now made to FIG. 3. FIG. 3 is a flowchart showing a computer- implemented method 300 of processing satellite image data, in accordance with one or more embodiments. Method 300 may be implemented using any suitable system. For example, method 300 may be implemented using system 100 shown in FIGS. 1 and 2, and concurrent reference is made in the below description to FIGS. 1 and 2. Method 300 can autonomously process the satellite image data without requiring manual processing / interventions.

[0067]

[0066] Method 300 may be performed at various times. For example, method 300 may be performed based on an automated schedule. As another example, method 300 may be automatically performed when input image data is received from a satellite system (e.g., satellite system 20). As another example, method 300 may be performed in response to a user input (e.g., from user device 30).

[0068]

[0067] At act 310, method 300 includes retrieving satellite image data. For example, system 100 may automatically retrieve satellite image data based on a user input received from user device 30. The user input may include location coordinates of an area of interest (AOI). The AOI can include, for example, infrastructure like railway tracks, roadways, runways, pipelines, dams, bridges etc.

[0069]

[0068] In some embodiments, a coverage threshold may be defined around the AOI. The coverage threshold can be any suitable threshold distance from the AOI to include in the analysis. For example, the coverage threshold may be in a range from 3m - 5000m. In other examples, the coverage threshold may be smaller than 3m. Smaller coverage threshold may enable faster data analysis centered around the AOI. In other examples, the coverage threshold may be greater than 5000m. Larger coverage thresholds may enable higher accuracy predictions and better insights for the AOI. The coverage threshold may be included in the user input. In some embodiments, the coverage threshold may be stored in database 250 and system 100 may automatically use the stored coverage threshold.

[0070]

[0069] System 100 may automatically connect to application programming interfaces (APIs) corresponding to various satellite image providers (e.g., European Space Agency, Copernicus, Natural Resources Canada (NRCAN), etc.) to retrieve satellite image data based on the location coordinates of the AOI and coverage threshold (if any). System 100 can retrieve different types of satellite image data including, for example, Sentinel - 1 , Radarsat 2, and TerraSAR - X.

[0071]

[0070] In some embodiments, system 100 can automatically filter out irrelevant images based on coordinates, threshold, and other parameters linked to satellite images, such as orbit type, resolution, swath width, etc. Method 300 can reduce the overall processing time of the satellite image data and improve the efficiency of processing system 100 by filtering out irrelevant images. Additionally, system 100 can improve processing speed and / or efficiency by sorting the retrieved images based on the various parameters and storing the sorted images in corresponding folders (e.g., in database 250).

[0072]

[0071] At act 320, method 300 includes generating ground deformation data based on the retrieved satellite image data. System 100 may perform any suitable analysis of the satellite image data to generate the ground deformation data. For example, system 100 may perform automated InSAR analysis based on the filtered satellite image data. System 100 may first automatically convert the satellite data from a Georeferenced Tagged Image File Format (GeoTIFF) file format to a Flexible Image Transport System (FITS) file format. This conversion can enable handling larger datasets due to the superior capability of the FITS file format to store and manage multi-dimensional arrays.

[0073]

[0072] System 100 may further generate an interferogram based on the SAR images in the satellite image data. The interferogram can represent phase differences between successive SAR acquisitions. System 100 may use Range Doppler Terrain Correction (RDTC) to mitigate geometric distortions due to topographic variations. This can enable accurate alignment of the SAR images.

[0074]

[0073] Further, system 100 can perform topographic corrections and range and azimuth processing at act 320. System 100 may use any suitable method to perform the topographic corrections. For example, system 100 may use digital elevation models (DEMs) to correct for topographic distortions. System 100 may perform the range and azimuth processing by adjusting the SAR image coordinates to a uniform grid, enhancing the accuracy of subsequent interferometric analysis.

[0075]

[0074] In some embodiments, system 100 may perform phase unwrapping at act 320. The phase unwrapping can convert the wrapped phase values, constrained within a range of [IT, IT], into continuous phase values. Phase unwrapping can be essential for accurately measuring the absolute displacement. System 100 may use any suitable method to perform the phase unwrapping. For example, system 100 may use a Minimum Cost Flow (MCF) algorithm for phase unwrapping. This can provide high precision in displacement measurements, thereby enabling reliable deformation analysis.

[0076]

[0075] After the interferogram is created, system 100 identify and mitigate noise sources in the SAR data. System 100 may use any suitable method to identify and mitigate the noise sources, for example, system 100 may use a trained neural network. The neural network can be a Convolutional Neural Network (CNN) generated by model generation engine 255. The CNN may be trained using large datasets of SAR images annotated with known noise characteristics. For example, the CNN may be trained using model training engine 260. System 100 can use the trained CNN to identify and mitigate various noise sources including, for example, thermal noise, speckle noise, and atmospheric disturbances. The thermal noise may be random noise introduced by sensor electronics. The speckle noise may be interference caused by coherent processing of SAR signals. The atmospheric disturbances can be variations in the atmosphere that affect signal propagation. The trained CNN can effectively identify and suppress noise patterns while preserving signal integrity.

[0077]

[0076] In some embodiments, system 100 may further use a neural network to perform the ground deformation analysis. Any suitable neural network may be used. For example, the neural network may be based on a reslI-Net architecture. The reslI-Net model is a variant of the ll-Net architecture that incorporates residual connections, enhancing its ability to learn complex patterns in the SAR data. This architecture can combine the strengths of Il-Net's encoder-decoder structure with residual learning, facilitating efficient training and improved performance on deformation analysis tasks. The model may be initialized with pretrained weights, enabling rapid convergence and superior generalization to new data.

[0078]

[0077] System 100 may input the interferogram and associated metadata into the neural network. The neural network can produce high-resolution deformation maps, accurately reflecting ground displacement. System 100 can optimize the neural network for parallel processing on GPU hardware, ensuring fast and efficient analysis. In some embodiments, system 100 may provide the deformation output in additional formats including, for example, excel sheets with deformation points for further analysis.

[0079]

[0078] System 100 may evaluate the performance of the neural network using metrics like mean absolute error (MAE) and root mean square error (RMSE). The neural network may be retrained (e.g., by model training engine 260) in response to the metrics not meeting a threshold condition (e.g., the RMSE is greater than a threshold value).

[0080]

[0079] At act 330, method 300 includes post-processing the deformation data generated at act 320. For example, system 100 may automatically perform data cleaning and remove irrelevant deformation data points. For example, the AOI may correspond to an infrastructure location and system 100 may remove irrelevant deformation data points including, for example, points corresponding to water bodies and anomalous deformation data points from further analysis. Anomalous points may be defined as points that have deformation values that do not coincide with the neighboring points by a margin greater than a threshold margin. The threshold margin may be pre-defined or provided as an input to system 100. Anomalous points may be caused by errors associated with the satellite data measurements due to issues including, for example, earth surface deformations, drastic elevation changes, atmospheric disturbances etc. In some embodiments, system 100 may further use a haversine distance calculation to include deformation data points relevant to the AOI and exclude the other irrelevant deformation data points.

[0080] System 100 may use any suitable method to remove the irrelevant deformation data points. For example, system 100 may use a neural network model trained to identify the irrelevant deformation data points. This can improve the speed and / or efficiency of system 100 in processing the satellite image data by automatically removing the irrelevant deformation data points from further analysis. Further, system 100 can provide greater consistency and reproducibility in removing the irrelevant deformation data points for different scenarios compared with a manual processing / intervention to identify and remove irrelevant deformation data points.

[0081]

[0081] In some embodiments, system 100 may further retrieve climate / weather data corresponding to the relevant deformation data points. The climate data may include longer-term data compared with the weather data. The climate / weather data may be received from weather satellite systems, local weather stations, etc. In some embodiments, the climate / weather data may include predicted climate / weather data from climate / weather prediction models. The climate / weather data may include, for example, temperature data, humidity data, rainfall data, snowfall data etc.

[0082]

[0082] As an example, the AOI for the deformation data may include railway infrastructure. The climate / weather conditions may impact the railway infrastructure in different ways, including directly causing defects in the railway infrastructure or affecting the surrounding subgrade thereby indirectly causing defects in the railway infrastructure. For example, ice jacking can be caused by accumulation and pooling of ice-melt water at the base of the tracks.

[0083]

[0083] System 100 can retrieve the climate / weather data for the spatiotemporal deformation data points for further analysis. System 100 can use any suitable method for combined analysis of the spatiotemporal deformation data and climate / weather data. System 100 can provide automated and seamless integration of the ground deformation data with corresponding climate / weather information at specific locations and times. This can enable a comprehensive analysis of the relationship between environmental factors and infrastructure performance, and generation of valuable insights for predictive maintenance and risk mitigation.

[0084]

[0084] At act 340, method 300 includes providing the ground deformation analysis output. System 100 can automatically generate the ground deformation analysis output based on the ground deformation analysis performed. System 100 can provide the ground deformation analysis output in one or more output formats. For example, system 100 can generate data visualizations (e.g., via user interface engine 230) including interactive scatter maps and graphs that can provide summarized insights (e.g., average ground deformations over time) and / or granular ground deformation analysis at the grid level.

[0085]

[0085] System 100 may further use one or more Al models to generate text explanations (e.g., in natural language) for each graph / grid segment explaining the observed ground deformation pattern. This can enable system 100 to provide actionable insights based on complex ground deformation data. In some examples, the AOI may include infrastructure and system 100 can generate ground deformation analysis output that includes one or more preventive maintenance recommendations.

[0086]

[0086] The preventive maintenance recommendations for the infrastructure may be generated, for example, by recommendation engine 270. The preventive maintenance recommendations may include current / predicted infrastructure condition description and / or current / predicted infrastructure risk analysis. Recommendation engine 270 may be trained (e.g., by model training engine 260) using a supervised learning method. The training data may include samples having a specific input data combination (e.g., ground deformation data and corresponding climate / weather data) and a successful response action for that specific input data combination. Model training engine 260 may use a large and diverse training dataset to train recommendation engine 270 to capture complex relationships between input data combinations and optimal / successful response actions.

[0087]

[0087] Reference is now made to FIG. 4. FIG. 4 shows an example ground deformation analysis output 400, in accordance with one or more embodiments. Ground deformation analysis output 400 includes a scatter plot showing line of sight deformation data for specific points on the Earth’s surface. The line of sight deformation data can provide insights into ground movement, subsidence and deformation patterns.

[0088]

[0088] Reference is now made to FIG. 5. FIG. 5 shows an example ground deformation analysis output 500, in accordance with one or more embodiments. Ground deformation analysis output 500 includes a deformation heatmap generated to visualize spatial patterns of ground movement. The deformation heatmap can provide a comprehensive overview of deformation intensity across a region, aiding in the identification of areas with significant changes.

[0089]

[0089] Reference is now made to FIGS. 6A and 6B showing examples of ground deformation analysis outputs, in accordance with one or more embodiments. Ground deformation analysis output 600, shown in FIG. 6A, includes an interactive map where users can view markers (e.g., markers 610-630) representing analyzed points. A user can interact with the map by clicking on a marker to reveal a detailed line graph 650 shown in FIG. 6B depicting the temporal evolution of ground deformation at that specific location. The interactive map can enhance the accessibility and interpretability of the ground deformation analysis results.

[0090]

[0090] The present invention has been described here by way of example only. Various modification and variations may be made to these exemplary embodiments without departing from the spirit and scope of the invention, which is limited only by the appended claims.

Claims

We claim:

1. A computer-implemented method of processing satellite image data for an area of interest (AOI), the method comprising: retrieving, by a processor, the satellite image data based on location coordinates of the AOI; generating, by the processor, ground deformation data based on the retrieved satellite image data; post-processing, by the processor, the generated ground deformation data to remove deformation data point irrelevant to the AOI; and providing, by the processor, a ground deformation analysis output based on postprocessing the generated ground deformation data.

2. The method of claim 1 , wherein retrieving the satellite image data is further based on a coverage threshold defined for the AOI.

3. The method of claim 1 , wherein the AOI includes railway infrastructure.

4. The method of claim 1 , wherein retrieving the satellite image data includes filtering out irrelevant images from the satellite image data based on one or more parameters associated with the satellite image data.

5. The method of claim 4, wherein the one or more parameters include an orbit type, a resolution or a swath width associated with the satellite image data.

6. The method of claim 1 , wherein the retrieved satellite image data includes synthetic aperture radar (SAR) images and generating the ground deformation data includes generating an interferogram based on the SAR images.

7. The method of claim 6, wherein generating the ground deformation data further includes identifying and mitigating noise sources in the SAR images.

8. The method of claim 7 further comprising using a trained Convolutional Neural Network (CNN) to identify and mitigate the noise sources.

9. The method of claim 6, wherein generating the ground deformation data further includes inputting the interferogram and associated metadata into a trained neural network based on a reslI-Net architecture.

10. The method of claim 1 , wherein post-processing the generated ground deformation data includes removing anomalous deformation data points and deformation data points corresponding to water bodies.1 1 .The method of claim 1 , wherein post-processing the generated ground deformation data includes using a haversine distance calculation to include deformation data points relevant to the AOI and exclude other irrelevant deformation data points.

12. The method of claim 1 , wherein post-processing the generated ground deformation data includes retrieving climate and / or weather data corresponding to the generated ground deformation data.

13. The method of claim 1 , wherein providing the ground deformation analysis output includes providing a preventative maintenance recommendation for an infrastructure located in the AOI.

14. A system for processing satellite image data for an area of interest (AOI), the system comprising: a memory storing processor-executable instructions; and a processor communicatively coupled to the memory, the instructions configuring the processor to perform the method according to any one of claims 1 to 13.

15. A non-transitory computer readable medium storing thereon program instructions that are executable by a processor for performing a method of processing satellite image datafor an area of interest (AOI), wherein the method is defined according to any one of claims 1 to 13.

Images