Wide-area coverage area high-precision deformation monitoring method based on InSAR technology

Abstract

The invention relates to a wide-area coverage area high-precision deformation monitoring method based on an InSAR technology, and the method comprises the steps: obtaining the InSAR monitoring data of a plurality of data sources for a wide-area coverage area, carrying out the registration of the InSAR monitoring data, screening interference pairs according to the coherence, carrying out the interference pair combination of a single-view complex image set obtained through registration, and obtaining a single-view complex image set; performing time sequence analysis according to the interference pair combination result to generate a time sequence interference image; performing phase filtering and phase unwrapping on the time sequence interference image; for the time sequence interference image after phase unwrapping, baseline parameters are corrected based on a nonlinear least square method, and atmospheric delay errors are corrected based on a GACOS atmospheric delay model; according to the processed image, constructing a wide-area surface deformation rate map and a preset application scene surface deformation rate map, and analyzing the spatio-temporal evolution law of the wide-area surface deformation rate map and the preset application scene surface deformation rate map; and finally, visually displaying the wide-area surface deformation rate diagram, the preset application scene surface deformation rate diagram and the spatio-temporal evolution law analysis result.

Description

Technical Field

[0001] The present application relates to the field of geological disaster monitoring technology, and in particular to a high-precision deformation monitoring method for a wide-area coverage area based on InSAR technology. Background Art

[0002] Surface deformation refers to regional surface displacement caused by the consolidation and compression of loose underground strata under the influence of natural or human factors. It will cause serious harm to land use, urban planning, transportation, flood control and drainage, building safety, etc., and may further trigger various geological disasters such as collapse, landslides and mudslides.

[0003] Wide-area coverage is a spatial scope concept used in surface deformation monitoring using InSAR (synthetic aperture radar) interferometry. It refers to a target area with a large coverage area and wide spatial distribution, typically encompassing provincial, river basin, national, or even global scales. In recent years, wide-area coverage areas have frequently experienced geological disasters such as landslides, debris flows, and earthquakes. Traditional monitoring methods struggle to fully cover these areas and cannot capture subtle deformations in a timely manner.

[0004] Currently, there are numerous deformation monitoring products based on InSAR technology on the market, but they are insufficient for wide-area coverage. Firstly, the surface scattering characteristics in wide-area coverage areas are complex and have poor coherence. The interference fringes acquired by conventional InSAR technology are of low quality, which limits the accuracy of deformation monitoring. Secondly, the atmospheric conditions in this area are complex, making it difficult to accurately eliminate the effects of atmospheric delay on phase, which can introduce significant errors. Summary of the Invention

[0005] In order to at least to some extent overcome the problems of limited accuracy and large errors in deformation monitoring of wide-area coverage areas based on InSAR technology in related technologies, the present application provides a high-precision deformation monitoring method for wide-area coverage areas based on InSAR technology.

[0006] The scheme of this application is as follows:

[0007] A high-precision deformation monitoring method for a wide-area coverage area based on InSAR technology, comprising:

[0008] Acquire InSAR monitoring data from multiple data sources covering a wide area;

[0009] Downloading the InSAR monitoring data and the external reference digital elevation model to a supercomputing server, registering the InSAR monitoring data, and obtaining a registered single-view complex image set;

[0010] Screening interference pairs according to coherence, performing interference pair combination on the single-view complex image set, performing time series analysis on the interference pair combination results, and generating a time series interference image;

[0011] performing phase filtering and phase unwrapping on the time-series interference image;

[0012] For the phase-unwrapped time-series interferometric images, the baseline parameters are corrected using the nonlinear least squares method, and the atmospheric delay error is corrected using the GACOS (Generic Atmospheric Correction Online Service) atmospheric delay model.

[0013] constructing a wide-area surface deformation rate map based on the processed image, screening a preset application scenario from the wide-area surface deformation rate map, and constructing a surface deformation rate map for the preset application scenario with higher accuracy;

[0014] Analyze the spatiotemporal evolution of deformation areas in wide-area surface deformation rate maps and surface deformation rate maps for pre-defined application scenarios. This includes analyzing the temporal trend of surface deformation and the spatial distribution of surface deformation in different geographic regions.

[0015] Visualize the wide-area surface deformation rate map, the surface deformation rate map for preset application scenarios, and the results of the spatiotemporal evolution law analysis.

[0016] Preferably, the different data sources include remote sensing satellites with different parameters;

[0017] The InSAR monitoring data provided by different data sources have at least one difference in band, resolution and imaging time.

[0018] Preferably, registering the InSAR monitoring data includes:

[0019] The image pair registration accuracy is better than 0.02 pixels in azimuth, and the initial registration is performed based on the preset InSAR image pair registration method;

[0020] Double differential interferometry is formed using the overlapping areas of each burst of images and the master and slave images, and fine registration between the bursts of images is performed according to the principle of minimizing the difference of double differential phase spectra.

[0021] Preferably, screening the interference pairs according to coherence and combining the interference pairs on the single-view complex image set comprises:

[0022] For coherent interferometric pairs, the interferometric pair combination is performed directly;

[0023] For the interference pairs with temporal incoherence, the average coherence of the interference pairs is used as a reference index to screen the time-series interference pairs, and the interference pairs after screening are combined;

[0024] For the spatially incoherent interference pairs, the spatial baseline threshold or the spatiotemporal combined baseline threshold is used as a reference index to screen the temporal interference pairs, and the interference pairs after screening are combined.

[0025] Preferably, performing phase filtering on the time-series interferometric image comprises:

[0026] The initial phase filtering process was performed using a nonlinear filtering index of 0.5 and a filtering window of 32 × 32;

[0027] determining a phase loss portion in an initial phase filtering process;

[0028] The phase loss is compensated by using a nonlinear filter index of 0.4 and a filter intensity of 5×5 in the iterative differential interference process.

[0029] Preferably, performing phase unwrapping on the time-series interference images comprises:

[0030] The phase unwrapping problem is transformed into a network model for the minimum cost flow problem. Each pixel in the image or each sampling point in the interference fringe is regarded as a node in the network. Directed edges are established between adjacent nodes based on the phase relationship and preset rules. The capacity and cost of the edge are determined based on information such as phase difference.

[0031] Determine the traffic rules in the network based on the characteristics and constraints of the phase unwrapping problem;

[0032] Configure the cost of a path based on the phase change of the path;

[0033] By continuously adjusting the traffic distribution in the network, the flow plan with the lowest cost is found, and the phase untangling result is obtained according to the flow plan with the lowest cost.

[0034] Preferably, the baseline parameters of the phase-unwrapped time-series interferometric images are corrected based on a nonlinear least squares method, including:

[0035] Uniformly sampling the image point phases in the time-series interferometric image after phase unwrapping;

[0036] Based on the elevation value corresponding to the phase of each image point, the baseline parameter estimation value is determined by nonlinear least square method, and the baseline parameter estimation value is corrected;

[0037] The baseline parameters include: baseline length and baseline inclination.

[0038] Preferably, the atmospheric delay error of the phase unwrapped time-series interferometric image is corrected based on the GACOS atmospheric delay model, including:

[0039] The numerical weather prediction model provides global atmospheric three-dimensional structure data, including air pressure, temperature, and humidity;

[0040] Numerical simulation and analysis of global atmospheric three-dimensional structure data to determine atmospheric delay estimates;

[0041] Calibrate and validate atmospheric delay estimates using ground-based weather station observations;

[0042] The atmospheric delay parameters are inverted by integrating multi-source data through the GACOS model, and the estimated atmospheric delay values ​​are corrected.

[0043] Preferably, the parameter requirements for the wide-area surface deformation rate map and the preset application scenario surface deformation rate map include:

[0044] The resolution of the wide-area surface deformation rate map is better than 20m, and the resolution of the surface deformation rate map for the preset application scenario is better than 5m. The coverage rate of the preset application scenario reaches 100%, the georeferencing accuracy is better than 1 / 8 pixel, the internal coincidence accuracy reaches the millimeter level, the interferometric registration accuracy is better than 0.1 pixel, and the deformation accuracy reaches the centimeter level.

[0045] The preset application scenarios include at least:

[0046] Identification of ground collapse in mining areas, ground subsidence in urban areas, transportation (bridges, roads) deformation, water conservancy facility deformation, energy infrastructure deformation and geological disaster risks.

[0047] Preferably, the method further comprises:

[0048] Build a surface deformation monitoring, display and analysis platform and provide a user interactive interface;

[0049] The user interface supports:

[0050] Data retrieval, data analysis, data sharing, as well as wide-area surface deformation rate maps and surface deformation rate maps for preset application scenarios, and visualization of the results of spatiotemporal evolution law analysis.

[0051] The technical solution provided by this application may have the following beneficial effects:

[0052] In this technical solution, InSAR monitoring data of a wide coverage area are obtained from multiple data sources. The InSAR monitoring data from different data sources have their own characteristics in terms of band, resolution, imaging time, etc. Through multi-source data fusion technology, the advantages can be complementary to improve data reliability and monitoring accuracy.

[0053] According to the characteristics of wide-area coverage, the interference pairs are screened according to coherence, and the interference pairs are combined for the single-view complex image set. Time series analysis is performed based on the interference pair combination results to improve the coherence of the generated time-series interference images, thereby improving the quality of the interference fringes.

[0054] Correcting baseline parameters using nonlinear least squares can eliminate systematic errors due to orbital errors. Correcting atmospheric delay errors based on the GACOS atmospheric delay model, taking into account the unique atmospheric conditions in areas with low vegetation cover, can more accurately eliminate atmospheric delay errors.

[0055] This technical solution not only focuses on the acquisition of surface deformation data, but also deeply analyzes the spatiotemporal evolution of deformation areas in the wide-area surface deformation rate map and the surface deformation rate map of the preset application scenario.

[0056] The wide-area surface deformation rate map and the surface deformation rate map of the preset application scenario, as well as the analysis results of the spatiotemporal evolution law are visualized to form a complete service process. By combining multi-source data such as geology, meteorology, and human activities, it provides various industries with more targeted and decision-making support value geological disaster monitoring information.

[0057] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0059] Figure 1 This is a flow chart of a method for high-precision deformation monitoring of a wide-area coverage area based on InSAR technology provided by an embodiment of the present application;

[0060] Figure 2 This is a flowchart of phase unwrapping of time-series interference images provided by an embodiment of the present application. DETAILED DESCRIPTION

[0061] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0062] Figure 1This is a flow chart of a method for high-precision deformation monitoring of a wide-area coverage area based on InSAR technology provided by an embodiment of the present application, with reference to Figure 1 A high-precision deformation monitoring method for a wide-area coverage area based on InSAR technology, comprising:

[0063] S11: Acquire InSAR monitoring data from multiple data sources covering a wide area;

[0064] S12: Download the InSAR monitoring data and the external reference digital elevation model to the supercomputing server, perform registration on the InSAR monitoring data, and obtain a registered single-view complex image set;

[0065] S13: Screen the interference pairs according to coherence, combine the interference pairs on the single-view complex image set, perform time series analysis on the interference pair combination results, and generate a time series interference image;

[0066] S14: Phase filtering and phase unwrapping of the time-series interferometric images;

[0067] S15: For the phase-unwrapped time-series interferometric images, the baseline parameters are corrected based on the nonlinear least squares method, and the atmospheric delay error is corrected based on the GACOS atmospheric delay model.

[0068] S16: constructing a wide-area surface deformation rate map based on the processed image, screening a preset application scenario from the wide-area surface deformation rate map, and constructing a surface deformation rate map for the preset application scenario with higher accuracy;

[0069] S17: Analyze the spatiotemporal evolution of deformation regions in the wide-area surface deformation rate map and the surface deformation rate map for the pre-defined application scenario. This includes analyzing the temporal trend of surface deformation and the spatial distribution of surface deformation in different geographic regions.

[0070] This technical solution analyzes the temporal trends of surface deformation, such as whether there are periodic changes or the timing of sudden deformation. Spatially, it analyzes the distribution of deformation across different geographic regions and its relationship to factors such as topography and geological structure. This in-depth analysis reveals the underlying mechanisms of surface deformation, providing a scientific basis for early warning and targeted prevention of geological disasters.

[0071] S18: Visualize the wide-area surface deformation rate map, the surface deformation rate map for the preset application scenario, and the results of the spatiotemporal evolution law analysis.

[0072] It should be noted that different data sources include remote sensing satellites with different parameters;

[0073] The InSAR monitoring data provided by different data sources have at least one difference in band, resolution and imaging time.

[0074] In specific practice, the data sources and their key parameters are as follows:

[0075] (1) Sentinel-1

[0076] The Sentinel-1 satellite is a C-band synthetic aperture radar (SAR) satellite. The C-band has a moderate wavelength, capable of penetrating a certain degree of cloud cover, rainfall, and vegetation cover to effectively acquire surface information. Its central wavelength is approximately 5.6 cm, making it highly adaptable to various surface monitoring tasks. It features multiple polarization modes, including single polarization (HH or VV) and dual polarization (HH+HV or VH+VV). Different polarization modes have varying sensitivities to different surface characteristics. For example, HH polarization is particularly effective in monitoring urban buildings, rocks, and other landforms, while dual polarization data provides richer surface information, facilitating applications such as land cover classification and target identification. The spatial resolution varies depending on the imaging mode. In Stripmap Mode, the spatial resolution can reach 5m×20m (azimuth×range), which is suitable for detailed observation of specific areas. The spatial resolution of Interferometric Wide-Swap Mode is approximately 5m×20m to 20m×40m, which can monitor a larger area and provide strong support for large-area terrain mapping and surface deformation monitoring. Sentinel-1 is designed with a short revisit cycle. Under a specific orbital configuration, it can revisit the same area every 6 to 12 days. This feature gives it a unique advantage in monitoring rapidly changing surface phenomena, such as flood evolution and tracking surface deformation after earthquake disasters. It can capture surface change information in a timely manner and provide high-frequency data support for disaster emergency response and research.

[0077] (2) AIRSAT 01 and 02

[0078] AIRSAT series 01 and 02 satellites occupy a key position in the data source of this technical solution. They are the only Ku-band radar satellites in China. This unique band characteristic gives them significant advantages in specific application scenarios. The Ku-band has a shorter wavelength and, compared to other bands, has higher sensitivity and resolution potential in detecting small targets and fine structures on the surface. Its antenna is the first parabolic phased array used in China. This advanced antenna form gives the satellite excellent beam control capabilities and flexible imaging mode switching capabilities. In actual applications, the phased array antenna can quickly adjust the beam direction and shape electronically to achieve efficient scanning and precise imaging of different areas. In terms of spatial resolution, the performance is particularly outstanding. The sliding beam mode can achieve an ultra-high resolution of 0.5m. In this mode, the satellite can conduct detailed observations of key areas of focus, such as high-precision mapping of urban core areas, detailed reconnaissance of military facilities, and other tasks, and can clearly distinguish extremely small land features and structural details; the strip mode can achieve a resolution of 2m. While maintaining a certain level of resolution, it can cover a relatively wider area. It is suitable for application scenarios such as linear land feature monitoring and preliminary screening of large-scale land use changes. It can quickly obtain surface information over a large range and effectively identify obvious land feature changes and anomalies.

[0079] (3) AIRSAT 03 and 04

[0080] AIRSAT series 03 and 04 satellites also occupy an important position in the data source of this technical solution. These two satellites are the flagship models of lightweight and small X-band SAR satellites. Their strong attitude maneuverability enables the satellites to quickly and flexibly adjust observation angles and directions, and conduct efficient imaging and monitoring of specific areas. The satellites have high observation flexibility and mission adaptability, and their technical indicators are at the leading level in China. Utilizing its high-resolution SAR imaging capability and high-precision orbit determination and velocity measurement capabilities, the satellite can quickly and accurately obtain topographic information in terrain mapping and geographic information updating, assisting in map production and urban planning; in environmental monitoring and resource management, it can perform long-term monitoring of forests, vegetation, water resources, etc., providing a basis for ecological protection and resource utilization; in disaster monitoring and emergency response, it can promptly detect disaster situations and provide key information for rescue and assessment; in agricultural monitoring and yield estimation, it can assist in agricultural production management and food security; in the field of marine monitoring and shipping management, it can provide services such as marine meteorological forecasting and shipping safety; in infrastructure monitoring and safety assessment, it can promptly detect hidden dangers and ensure public safety. Its advanced technical indicators provide strong support and guarantee for economic and social development and national security.

[0081] (4) AIRSAT 08 (Haishao-1)

[0082] Haishao-1 (AIRSAT 08) is China's only low-inclination satellite. Its payload boasts several key features: first, its unique ultra-low-orbit, low-inclination X-band orbit. This low-inclination imaging mode is highly advantageous in east-west regions such as the Qilian Mountains and the Yellow River. For example, in the Qilian Mountains, it enables efficient and comprehensive data acquisition for tasks such as snow depth monitoring, ecological change assessment, and land feature classification. Its low-inclination orbit allows for observations from a more appropriate angle, minimizing the adverse effects of terrain obstruction. Furthermore, it covers a larger area in a single pass, enabling the capture of overall changes in ecosystem components and the distribution characteristics of different land feature types within the region. Furthermore, the satellite features an advanced payload featuring fully polarized and integrated optical SAR. The fully polarized mode provides rich ground scattering information, facilitating in-depth analysis of the physical structure and properties of land features. For glacier monitoring, it can better identify different surface material compositions and ice cracks. In maritime applications, it can accurately detect and classify targets such as wind, waves, and ships. The integrated optical SAR design combines the advantages of both optical and synthetic aperture radar detection methods, enabling the acquisition of multi-source data under varying weather and lighting conditions. This further enhances data richness and usability, providing more comprehensive and accurate data support for maritime applications such as marine resource exploration, maritime traffic regulation, and glacier monitoring in the northwest region. Its 1m swath mode is also relatively rare, offering high resolution while maintaining a certain coverage width, enabling the acquisition of more detailed terrain information in both maritime and land applications, such as fine-grained coastal landforms and small features in inland areas.

[0083] (5) RADARSAT-2

[0084] RADARSAT-2 operates in the C-band and offers a wide range of polarization modes, including full polarization and various flexible polarization combinations. Full polarization mode captures a complete target scattering matrix, providing a rich data foundation for studying ground object scattering mechanisms and high-precision object classification. Its flexible polarization combination modes allow targeted polarization configuration selection based on specific application requirements, such as ocean monitoring and ice and snow research, improving data sensitivity and information extraction accuracy for specific targets. It also offers multiple resolution imaging modes. In fine resolution mode, spatial resolution can exceed 3m×3m, enabling high-resolution applications such as detailed monitoring of urban infrastructure and military target identification. Standard mode offers a spatial resolution of approximately 5m×8m, making it widely used in land use change monitoring and agricultural irrigation infrastructure monitoring. Wide-swath imaging mode enables observations over a larger area. Although its resolution is relatively low, it is crucial for rapid screening of large areas and macro-change monitoring, with a resolution of approximately 25m×25m. The satellite's orbital design ensures a revisit period of approximately 24 days. Although the revisit period is relatively long, it can still play an important role in the study of long-term surface changes by combining it with other satellite data, such as long-term monitoring of coastal evolution and periodic assessment of large-scale engineering construction areas, providing stable data support for understanding the slow changes of large-scale surface processes.

[0085] (7)COSMO

[0086] COSMO satellite radar data boasts high resolution, reaching resolutions up to 1 meter, enabling clear discrimination of subtle features such as buildings and roads, providing a foundation for accurate identification. Furthermore, it features multiple polarization modes, such as HH, HV, VH, and VV, which can capture richer information about features and improve classification and identification accuracy. It also possesses interferometry capabilities, processing images taken at different times or angles to obtain high-precision terrain and surface deformation information. In topographic mapping and geological exploration, this data can be used to rapidly generate high-precision digital elevation models, providing detailed data for topographic mapping, assisting in the investigation of potential geological disasters, monitoring subtle surface deformations in geologically active areas, and providing early warning of geological disasters. In environmental monitoring and resource management, COSMO satellite radar data can be used for long-term monitoring of ecosystems such as forests and grasslands, obtaining information on vegetation growth and cover changes, and assessing the ecological environment. It can also monitor water resource distribution and marine environmental factors, such as wind speed and sea ice distribution. COSMO satellite radar data offers significant advantages in disaster monitoring and emergency response. When disasters such as earthquakes and floods occur, images of the affected areas can be quickly acquired to understand the scope, extent, and development trends of the disaster, providing key support for rescue and post-disaster assessments. In fire monitoring, it can also penetrate smoke and assist in firefighting operations. In agricultural monitoring and yield estimation, it can monitor crop planting, growth, and pest and disease conditions, helping farmers to rationally arrange their farming activities. It combines multiple types of data to build models to predict yields and ensure food security. In ocean monitoring and shipping management, its data is used to monitor marine meteorological elements, providing guarantees for navigation safety. It can also promptly detect marine pollution incidents and assist in marine environmental protection. For infrastructure monitoring and safety assessments, it can regularly monitor urban buildings, bridges, and water conservancy projects, promptly detect structural safety hazards, and ensure the safe and stable operation of various facilities.

[0087] (8) Haisi No. 1 / Chaohu No. 1 / Fucheng No. 1

[0088] The Haisi-1 / Chaohu-1 / Fucheng-1 satellites offer advantages such as low cost, rapid deployment, strong coverage, and flexible scheduling. They provide high resolution, wide coverage, continuous monitoring, and rapid remote sensing information processing, offering commercial users diverse, high-quality, and cost-effective SAR satellite data services. Benchmarked against internationally advanced standards, the satellites employ active phased array antennas and are C-band SAR remote sensing satellites, capable of all-day, all-weather interferometric imaging. Fucheng-1 utilizes a high-precision quad-frequency GPS + Beidou GNSS receiver, with an expected orbit determination accuracy of better than 10 cm. Using long-time series InSAR technology, they can monitor surface deformation to the millimeter level.

[0089] This technical solution acquires InSAR monitoring data from multiple data sources covering a wide area. These data sources each have their own unique characteristics in terms of wavelength, resolution, and imaging time. Multi-source data fusion technology can complement these advantages, improving data reliability and monitoring accuracy. For example, by combining the varying sensitivity of data from different wavelengths to different surface features, surface deformation information can be more comprehensively captured.

[0090] It should be noted that the registration of InSAR monitoring data includes:

[0091] The image pair registration accuracy is better than 0.02 pixels in azimuth, and the initial registration is performed based on the preset InSAR image pair registration method;

[0092] Double differential interferometry is formed using the overlapping areas of each burst of images and the master and slave images, and fine registration between the bursts of images is performed according to the principle of minimizing the difference of double differential phase spectra.

[0093] In this technical solution, the InSAR monitoring data and the external reference digital elevation model (DEM) - ALOS ASTER 12.5m DEM are first downloaded to the supercomputing server, and then the InSAR monitoring data are aligned.

[0094] Registration is a fundamental and crucial step in InSAR monitoring data processing. To meet interferometric requirements, InSAR image registration accuracy is generally required to be within 0.1 pixel. To ensure that the interferometric phase deviation is less than 3°, the registration accuracy must be within 0.001 pixel.

[0095] To ensure the stability of precise registration between bursts of SAR data, image pair registration accuracy should be better than 0.02 pixels in azimuth. An improved InSAR image pair registration method is used for initial registration, enabling high-precision InSAR image pair registration in complex terrain or low coherence conditions. Subsequently, fine registration is performed based on this foundation. Double-difference interferometry is formed using the overlapping regions of each burst and between the master and slave images. Fine-scale registration between bursts is performed based on the principle of minimizing the difference in double-difference phase spectra. This improves the algorithm's sensitivity to registration offsets, meets the accuracy requirements of data registration, and ultimately results in a set of registered single-look complex images.

[0096] It should be noted that the interference pairs are screened according to coherence, and the interference pairs are combined on the single-view complex image set, including:

[0097] For coherent interferometric pairs, the interferometric pair combination is performed directly;

[0098] For the interference pairs with temporal incoherence, the average coherence of the interference pairs is used as a reference index to screen the time-series interference pairs, and the interference pairs after screening are combined;

[0099] For the spatially incoherent interference pairs, the spatial baseline threshold or the spatiotemporal combined baseline threshold is used as a reference index to screen the temporal interference pairs, and the interference pairs after screening are combined.

[0100] When performing time series analysis, it is necessary to perform interferometric pair combination on the registered images. However, due to the influence of various decoherence factors, only a portion of the interferometric pairs have good coherence and can provide reliable observation information. For these coherent interferometric pairs, interferometric pair combination is performed directly.

[0101] Incoherence factors primarily include radar system thermal noise, Doppler centroid drift of SAR image pairs, volume scattering incoherence, incoherence introduced during data processing, temporal incoherence, and spatial (geometric) incoherence. Currently, the incoherence that can be mitigated through interferometric baseline combination primarily consists of temporal and spatial incoherence. Therefore, interferometric pairs with good coherence are selected for subsequent processing by setting a temporal baseline threshold, a spatial baseline threshold, or a combination of the two. The temporal baseline threshold is primarily used to prevent temporal incoherence, while the spatial baseline threshold is primarily used to prevent geometric incoherence.

[0102] Preferably, for wide-area InSAR processing, especially provincial surface deformation extraction, since the processed data range encompasses diverse geographic environments, it is unreasonable to use a unified spatiotemporal baseline threshold as a constraint for interferometric pair combination. Therefore, from a coherence perspective, this embodiment uses the average coherence of the interferometric pairs instead of the temporal baseline as a reference indicator for optimizing time-series interferometric pairs. That is, for temporally incoherent interferometric pairs, the average coherence of the interferometric pairs is used as a reference indicator for screening the temporal interferometric pairs, and the selected interferometric pairs are then combined. For spatially incoherent interferometric pairs, the spatial baseline threshold or the combined spatiotemporal baseline threshold is used as a reference indicator for screening the temporal interferometric pairs, and the selected interferometric pairs are then combined.

[0103] In this technical solution, based on the characteristics of the wide-area coverage area, the interference pairs are screened according to coherence, the interference pairs are combined for the single-view complex image set, and time series analysis is performed based on the interference pair combination results to improve the coherence of the generated time-series interference images, thereby improving the quality of the interference fringes.

[0104] It should be noted that phase filtering of time-series interferometric images includes:

[0105] The initial phase filtering process was performed using a nonlinear filtering index of 0.5 and a filtering window of 32 × 32;

[0106] determining a phase loss portion in an initial phase filtering process;

[0107] The phase loss is compensated by using a nonlinear filter index of 0.4 and a filter intensity of 5×5 in the iterative differential interference process.

[0108] In order to reduce the impact of phase noise and improve the stability of phase unwrapping, phase filtering is required before preliminary phase unwrapping.

[0109] Preferably, in this embodiment, the Goldstein adaptive filtering method is used to perform phase filtering to effectively reduce the residual phase component.

[0110] In order to minimize the influence of phase noise and avoid the adverse effects of phase unwrapping errors on subsequent baseline corrections, deformation rate, and elevation error estimation, a strong filter with a nonlinear filter index of 0.5 and a filter window of 32 × 32 was used to process the differential interferogram before the initial phase unwrapping.

[0111] This technical solution takes into account that overly strong filtering may filter out part of the deformation phase, resulting in phase loss. This part of the lost phase will be compensated by reducing the filtering intensity (nonlinear filtering index 0.4, filtering window 5×5) during the iterative differential interference process.

[0112] It should be noted that, referring to Figure 2 , perform phase unwrapping on time-series interferometric images, including:

[0113] S21: A network model that transforms the phase unwrapping problem into a minimum cost flow problem. Each pixel in the image or each sampling point in the interference fringe is considered a node in the network. Directed edges are established between adjacent nodes based on the phase relationship and preset rules. The capacity and cost of the edges are determined based on information such as phase difference.

[0114] S22: Determine the traffic rules in the network based on the characteristics and constraints of the phase unwrapping problem;

[0115] S23: configuring the cost of the path according to the phase change of the path;

[0116] S24: By continuously adjusting the traffic distribution in the network, a flow plan with the lowest cost is found, and a phase untangling result is obtained according to the flow plan with the lowest cost.

[0117] Preferably, in this embodiment, an MCF (Minimum Cost Flow) unwrapping method is used to perform phase unwrapping.

[0118] The MCF unwrapping method has strong noise immunity and can, to a certain extent, handle noise and errors in phase data, resulting in relatively accurate phase unwrapping results. Furthermore, its network flow-based model has a clear mathematical structure and can be efficiently solved using mature network flow algorithms, demonstrating excellent performance when handling large-scale phase unwrapping problems.

[0119] The MCF untangling method is a classic model for network flow problems. In a weighted directed graph, each edge has a capacity constraint and a unit cost. The minimum cost flow problem is to find a flow allocation scheme that minimizes the total cost while meeting certain flow requirements. The basic principle is as follows:

[0120] Network construction: This involves transforming the phase unwrapping problem into a network model for the minimum cost flow problem. Each pixel in an image or each sampling point in an interference pattern is typically considered a node in the network. Directed edges are established between adjacent nodes based on phase relationships and certain rules. The capacity and cost of an edge are determined based on information such as the phase difference. For example, if the phase difference between two adjacent points is within a reasonable range, the corresponding edge may have a low cost, indicating that the phase change between the two points is reasonable and continuous. However, if the phase difference exceeds a certain range, it may indicate a phase jump, resulting in a higher cost for the corresponding edge.

[0121] Flow rules: Based on the characteristics of the problem and the constraints, flow rules within the network are determined. Generally speaking, the total flow from the source node to the sink node must meet certain conditions. These conditions are often related to the physical meaning or mathematical requirements of the entire phase unwrapping problem. For example, in interferometry-based phase unwrapping, the flow through the entire network may be required to correspond to a physical quantity such as the total change in the interference fringes.

[0122] Cost calculation: The design of the cost function is crucial for MCF phase unwrapping. Cost is typically related to factors such as phase difference and noise level. A common cost calculation method is based on the square of the phase difference between adjacent pixels or other similar metrics. This approach assigns a lower cost to paths with smooth phase changes, while paths with larger phase jumps have a higher cost. This approach favors paths with continuous and reasonable phase changes when solving for the minimum cost flow.

[0123] Solving for minimum-cost flows: Leveraging established minimum-cost flow algorithms, such as the primal-dual algorithm and the network simplex method, we solve for minimum-cost flows on constructed networks. These algorithms continuously adjust network traffic distribution to find the flow solution with the lowest cost. The resulting minimum-cost flow solution corresponds to a phase untangling solution. This solution determines the phase value of each node based on the traffic distribution along the edges in the network, achieving phase untangling.

[0124] It should be noted that the baseline parameters of the phase-unwrapped time-series interferometric images are corrected based on the nonlinear least squares method, including:

[0125] Uniformly sampling the image point phases in the time-series interferometric image after phase unwrapping;

[0126] Based on the elevation value corresponding to the phase of each image point, the baseline parameter estimation value is determined by nonlinear least square method, and the baseline parameter estimation value is corrected;

[0127] Baseline parameters include: baseline length and baseline inclination.

[0128] It should be noted that the unwrapped phase mainly includes surface deformation, terrain error, baseline error, atmospheric effect, and phase error introduced by data processing steps such as phase filtering:

[0129]

[0130] in, represents the unwrapping phase; is the surface deformation phase; is the terrain error phase; is the baseline error phase; is the atmospheric effect phase; is the remaining phase component.

[0131] For baseline error, the phase of the image points is uniformly sampled in the interferogram, and the baseline parameters, including baseline length and baseline inclination α, are estimated using the nonlinear least squares method based on the corresponding elevation values, and then corrections are made.

[0132] The formula for estimating baseline parameters using nonlinear least squares method is as follows:

[0133]

[0134] Among them, h represents the elevation value; B ⊥ is the vertical baseline; θ is the radar incident angle; P is the radar slant range.

[0135] It should be noted that the atmospheric delay error of the phase-unwrapped time-series interferometric images is corrected based on the GACOS atmospheric delay model, including:

[0136] The numerical weather prediction model provides global atmospheric three-dimensional structure data, including air pressure, temperature, and humidity;

[0137] Numerical simulation and analysis of global atmospheric three-dimensional structure data to determine atmospheric delay estimates;

[0138] Calibrate and validate atmospheric delay estimates using ground-based weather station observations;

[0139] The atmospheric delay parameters are inverted by integrating multi-source data through the GACOS model, and the estimated atmospheric delay values ​​are corrected.

[0140] Numerical weather prediction models provide information on the three-dimensional structure of the global atmosphere, including pressure, temperature, and humidity. Complex numerical simulations and analyses of this data allow for preliminary estimates of atmospheric delays. Observations from ground-based weather stations also provide a crucial supplement for calibrating and validating NWP model results.

[0141] In this embodiment, the atmospheric delay parameters are inverted by integrating multi-source data through the GACOS model, and the estimated atmospheric delay value is corrected. The GACOS model can more accurately match and fuse GACOS data with InSAR data in time and space. During the data acquisition stage, the process of obtaining atmospheric delay products from the GACOS database is optimized to ensure that the acquired data accurately corresponds to the time and space range of the InSAR data. In the data preprocessing link, through efficient resampling and projection conversion algorithms, the GACOS data can better adapt to the resolution and coordinate system of the InSAR interferogram, greatly improving the accuracy of atmospheric delay extraction and phase correction.

[0142] This technical solution uses a nonlinear least squares approach to correct baseline parameters, eliminating systematic errors due to orbital errors. Correcting atmospheric delay errors based on the GACOS atmospheric delay model takes into account the unique atmospheric conditions in areas with low vegetation cover, enabling more accurate elimination of atmospheric delay errors.

[0143] It should be noted that the parameter requirements for the wide-area surface deformation rate map and the surface deformation rate map for the preset application scenario include:

[0144] The resolution of the wide-area surface deformation rate map is better than 20m, and the resolution of the surface deformation rate map for the preset application scenario is better than 5m. The coverage rate of the preset application scenario reaches 100%, the georeferencing accuracy is better than 1 / 8 pixel, the internal coincidence accuracy reaches the millimeter level, the interferometric registration accuracy is better than 0.1 pixel, and the deformation accuracy reaches the centimeter level.

[0145] The preset application scenarios include at least:

[0146] Identification of ground collapse in mining areas, ground subsidence in urban areas, transportation (bridges, roads) deformation, water conservancy facility deformation, energy infrastructure deformation and geological disaster risks.

[0147] In this technical solution, preset application scenarios are screened in the wide-area surface deformation rate map, and a surface deformation rate map of the preset application scenarios is constructed with higher precision. The preset application scenarios are key monitoring areas, including ground collapse in mining areas, ground subsidence in urban areas, transportation (bridges, roads) deformation, water conservancy facility deformation, energy infrastructure deformation, and identification of geological disaster hazards.

[0148] In this technical solution, monitoring parameters for these key monitoring areas have been optimized, namely, the resolution of the wide-area surface deformation rate map is better than 20m, the resolution of the surface deformation rate map for the preset application scenario is better than 5m, the coverage rate of the preset application scenario reaches 100%, the geographic registration accuracy is better than 1 / 8 pixel, the internal coincidence accuracy reaches the millimeter level, the interference pair registration accuracy is better than 0.1 pixel, and the deformation accuracy reaches the centimeter level.

[0149] Wide-area resolution refers to the spatial resolution of a wide-area monitoring image, which determines the spatial accuracy of the overall regional data.

[0150] The resolution of the surface deformation rate map in the preset application scenario refers to the high-resolution requirement of the key monitoring area, which supports high-precision analysis of local engineering facilities.

[0151] The preset application scenario coverage refers to the coverage completeness of key scenarios, which is used to ensure that the target area monitoring is not missed.

[0152] Georeferencing accuracy refers to the degree of match with the map coordinate system and is used to improve geolocation accuracy.

[0153] Internal consistency refers to the internal consistency of multi-period deformation time series, which is used to ensure the stability and comparability of time series products.

[0154] Interferometric registration accuracy refers to the registration error between the primary and secondary SAR images, which is used to improve the interferometric phase accuracy and reduce the unwrapping error.

[0155] Deformation accuracy refers to the error between the actual surface deformation and the monitored value, which is used to ensure the credibility and reliability of the final product.

[0156] Understandably, the surface characteristics of low-vegetation areas are complex, and different application scenarios have different requirements for deformation results. Therefore, this technical solution constructs and studies a universal InSAR data processing process for the study area, and conducts core technology improvements and parameter optimization work in conjunction with the study scenario. For example, to address the problem of easy loss of coherence in detection of large-gradient deformation scenes in mining areas, a better unwrapping method is constructed; a high-coherence point selection algorithm is developed for small deformation scenes of roads and bridges; and an atmospheric model application algorithm is developed for geological disaster hazard identification scenarios, thereby improving the application effectiveness of InSAR technology.

[0157] It should be noted that the method also includes:

[0158] Build a surface deformation monitoring, display and analysis platform and provide a user interactive interface;

[0159] User interface support:

[0160] Data retrieval, data analysis, data sharing, as well as wide-area surface deformation rate maps and surface deformation rate maps for preset application scenarios, and visualization of the results of spatiotemporal evolution law analysis.

[0161] This technical solution presents technical achievements through a platform (system). This platform (system) must not only have a user-friendly interface to facilitate researchers' rapid retrieval of required results, but also support diverse data analysis functions, such as data visualization and statistical analysis, to meet the needs of different users in different application scenarios. Furthermore, through data sharing, it will promote cross-disciplinary research and cooperation, providing comprehensive and accurate technical support for regional geological disaster prevention and control, resource development, infrastructure construction, and ecological and environmental protection.

[0162] The wide-area surface deformation rate map and the surface deformation rate map of the preset application scenario, as well as the analysis results of the spatiotemporal evolution law are visualized to form a complete service process. By combining multi-source data such as geology, meteorology, and human activities, it provides various industries with more targeted and decision-making support value geological disaster monitoring information.

[0163] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0164] It should be noted that, in the description of this application, the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0165] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0166] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0167] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0168] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0169] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0170] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0171] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A high-precision deformation monitoring method for a wide-area coverage area based on InSAR technology, characterized in that: include: Acquire InSAR monitoring data from multiple data sources covering a wide area; Downloading the InSAR monitoring data and the external reference digital elevation model to a supercomputing server, registering the InSAR monitoring data, and obtaining a registered single-view complex image set; Screening interference pairs according to coherence, performing interference pair combination on the single-view complex image set, performing time series analysis on the interference pair combination results, and generating a time series interference image; performing phase filtering and phase unwrapping on the time-series interference image; For the phase-unwrapped time-series interferometric images, the baseline parameters are corrected based on the nonlinear least squares method, and the atmospheric delay error is corrected based on the GACOS atmospheric delay model. constructing a wide-area surface deformation rate map based on the processed image, screening a preset application scenario from the wide-area surface deformation rate map, and constructing a surface deformation rate map for the preset application scenario with higher accuracy; Analyze the spatiotemporal evolution of deformation areas in wide-area surface deformation rate maps and surface deformation rate maps for pre-defined application scenarios. This includes analyzing the temporal trend of surface deformation and the spatial distribution of surface deformation in different geographic regions. Visualize the wide-area surface deformation rate map, the surface deformation rate map for preset application scenarios, and the results of the spatiotemporal evolution law analysis.

2. The method according to claim 1, characterized in that Different data sources include remote sensing satellites with different parameters; The InSAR monitoring data provided by different data sources have at least one difference in band, resolution and imaging time.

3. The method according to claim 1, characterized in that The InSAR monitoring data is registered, including: The image pair registration accuracy is better than 0.02 pixels in azimuth, and the initial registration is performed based on the preset InSAR image pair registration method; Double differential interferometry is formed using the overlapping areas of each burst of images and the master and slave images, and fine registration between the bursts of images is performed according to the principle of minimizing the difference of double differential phase spectra.

4. The method according to claim 1, wherein Screening the interference pairs according to coherence, and combining the interference pairs on the single-view complex image set, including: For coherent interferometric pairs, the interferometric pair combination is performed directly; For the interference pairs with temporal incoherence, the average coherence of the interference pairs is used as a reference index to screen the time-series interference pairs, and the interference pairs after screening are combined; For the spatially incoherent interference pairs, the spatial baseline threshold or the spatiotemporal combined baseline threshold is used as a reference index to screen the temporal interference pairs, and the interference pairs after screening are combined.

5. The method according to claim 1, wherein Performing phase filtering on the time-series interference image, comprising: The initial phase filtering process was performed using a nonlinear filtering index of 0.5 and a filtering window of 32 × 32; determining a phase loss portion in an initial phase filtering process; The phase loss is compensated by using a nonlinear filter index of 0.4 and a filter intensity of 5×5 in the iterative differential interference process.

6. The method according to claim 1, characterized in that Performing phase unwrapping on the time-series interference image includes: The phase unwrapping problem is transformed into a network model for the minimum cost flow problem. Each pixel in the image or each sampling point in the interference fringe is regarded as a node in the network. Directed edges are established between adjacent nodes based on the phase relationship and preset rules. The capacity and cost of the edge are determined based on information such as phase difference. Determine the traffic rules in the network based on the characteristics and constraints of the phase unwrapping problem; Configure the cost of a path based on the phase change of the path; By continuously adjusting the traffic distribution in the network, the flow plan with the lowest cost is found, and the phase untangling result is obtained according to the flow plan with the lowest cost.

7. The method according to claim 1, characterized in that For the phase-unwrapped time-series interferometric images, the baseline parameters are corrected based on the nonlinear least squares method, including: Uniformly sampling the image point phases in the time-series interferometric image after phase unwrapping; Based on the elevation value corresponding to the phase of each image point, the baseline parameter estimation value is determined by nonlinear least square method, and the baseline parameter estimation value is corrected; The baseline parameters include: baseline length and baseline inclination.

8. The method according to claim 1, characterized in that For the phase unwrapped time-series interferometric images, the atmospheric delay error is corrected based on the GACOS atmospheric delay model, including: The numerical weather prediction model provides global atmospheric three-dimensional structure data, including air pressure, temperature, and humidity; Numerical simulation and analysis of global atmospheric three-dimensional structure data to determine atmospheric delay estimates; Calibrate and validate atmospheric delay estimates using ground-based weather station observations; The atmospheric delay parameters are inverted by integrating multi-source data through the GACOS model, and the estimated atmospheric delay values ​​are corrected.

9. The method according to claim 1, characterized in that The parameter requirements for the wide-area surface deformation rate map and the preset application scenario surface deformation rate map include: The resolution of the wide-area surface deformation rate map is better than 20m, and the resolution of the surface deformation rate map for the preset application scenario is better than 5m. The coverage rate of the preset application scenario reaches 100%, the georeferencing accuracy is better than 1 / 8 pixel, the internal coincidence accuracy reaches the millimeter level, the interferometric registration accuracy is better than 0.1 pixel, and the deformation accuracy reaches the centimeter level. The preset application scenarios include at least: Identification of ground collapse in mining areas, ground subsidence in urban areas, transportation (bridges, roads) deformation, water conservancy facility deformation, energy infrastructure deformation and geological disaster risks.

10. The method according to claim 1, characterized in that The method further comprises: Build a surface deformation monitoring, display and analysis platform and provide a user interactive interface; The user interface supports: Data retrieval, data analysis, data sharing, as well as wide-area surface deformation rate maps and surface deformation rate maps for preset application scenarios, and visualization of the results of spatiotemporal evolution law analysis.

Images