Mountainous landslide surface deformation monitoring method and system

Abstract

This invention discloses a method and system for monitoring surface deformation of landslides in complex mountainous areas, belonging to the field of geological disaster monitoring technology. The method includes: determining the CR deformation parameters of the target area based on the coordinates of corner reflectors in SAR images using the CR-InSAR method; determining the permanent scatterers in the images based on the PS-InSAR method and performing temporal phase analysis to determine the PS deformation parameters of the target area using these permanent scatterers; determining the deformation parameters of distributed scatterers in the images based on the DS-InSAR method by performing differential interferometry on images at adjacent time points, using corner reflectors as control points and the PS deformation parameters as constraints; and monitoring the deformation of the target area in the mountainous area based on the CR, PS, and DS deformation parameters. This invention integrates the advantages of three InSAR algorithms, improving the reliability of surface deformation monitoring results for landslides in mountainous areas.

Description

Technical Field

[0001] This invention relates to the field of geological disaster monitoring technology, and in particular to a method and system for monitoring surface deformation of landslides in mountainous areas. Background Technology

[0002] Landslides are one of the most serious disasters, second only to earthquakes, in terms of their high frequency of occurrence and wide range of impact, affecting people's lives, property and public infrastructure. Landslides in complex mountainous areas are particularly typical, being highly concealed, extremely dangerous, and difficult to monitor and study.

[0003] Currently, Synthetic Aperture Radar Interferometry (InSAR) technology is widely used in landslide deformation monitoring due to its high monitoring accuracy, high spatial resolution, high time repetition frequency, wide coverage, and minimal impact from weather conditions. However, InSAR technology still faces many limitations in landslide monitoring in complex mountainous areas. In vegetated mountainous regions, conventional time-series InSAR technology struggles to capture a sufficient number of coherent radar targets, and the limited number of monitoring points cannot fully reflect the spatial deformation patterns of landslides, easily leading to underestimation or misjudgment of deformation. Summary of the Invention

[0004] The purpose of this invention is to provide a highly reliable method and system for monitoring surface deformation of landslides in mountainous areas.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] This invention provides a method for monitoring surface deformation of landslides in mountainous areas, comprising:

[0007] Acquire SAR imagery of target areas in mountainous regions;

[0008] Based on the CR-InSAR method, the coordinates of corner reflectors in SAR images are determined by utilizing the backscattering characteristics of corner reflectors. Furthermore, the CR deformation parameters of the mountainous target area are determined based on the coordinates of each corner reflector in the SAR images at adjacent time points. The corner reflectors are located within the mountainous target area.

[0009] Based on the PS-InSAR method, permanent scatterers in the SAR image are identified, and temporal phase analysis is performed on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region.

[0010] Using the corner reflector as the control point and the PS deformation parameter as the constraint, differential interferometry is performed on the SAR images at adjacent time points based on the DS-InSAR method to determine the deformation parameters of the distributed scatterer in the SAR image.

[0011] The deformation of the target area in the mountainous region is monitored based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter.

[0012] Optionally, determining the CR deformation parameters of the mountainous target area based on the coordinates of each corner reflector in the SAR image at different times specifically includes:

[0013] Based on the solution space search method, the optimal estimate of the CR deformation parameter is determined according to the prior information of each corner reflector; the solution space search method determines the size, position and search step size of the solution space according to the prior information of each corner reflector; the prior information is the position information and deformation information obtained from GPS monitoring data.

[0014] Optionally, before monitoring the deformation of the target area in the mountainous region based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter, the method further includes:

[0015] Based on the iterative IGG-III method, the observations are weighted to eliminate the influence of gross errors in the observations; the observations include the DS deformation parameters.

[0016] Optionally, monitoring the deformation of the target area in the mountainous region based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter specifically includes:

[0017] Construct a Delaunay triangulation network for CR, PS, and DS;

[0018] Based on the Delaunay triangulation network, the time series of deformation rates at each location in the target mountain area are determined according to the CR deformation parameters, the PS deformation parameters, and the DS deformation parameters.

[0019] Optionally, the CR deformation parameters are continuous, regional deformation monitoring results, including CR deformation rate and CR time series cumulative deformation.

[0020] Optionally, the PS deformation parameters include the PS deformation rate and the PS time series cumulative deformation.

[0021] Optionally, the DS deformation parameters include the DS deformation rate and the cumulative deformation of the DS time series.

[0022] This invention also provides a surface deformation monitoring system for landslides in mountainous areas, comprising:

[0023] The image acquisition module is used to acquire SAR images of target areas in mountainous regions.

[0024] The first calculation module is used to determine the coordinates of corner reflectors in SAR images based on the CR-InSAR method and utilizing the backscattering characteristics of corner reflectors, and to determine the CR deformation parameters of the mountainous target area based on the coordinates of each corner reflector in the SAR images at adjacent time points; the corner reflectors are set in the mountainous target area.

[0025] The second calculation module is used to determine the permanent scatterers in the SAR image based on the PS-InSAR method, and to perform temporal phase analysis on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region.

[0026] The third calculation module is used to perform differential interferometry processing on the SAR images at adjacent time points based on the DS-InSAR method, using the corner reflector as the control point and the PS deformation parameter as the constraint, to determine the deformation parameter of the distributed scatterer in the SAR image.

[0027] The monitoring module is used to monitor the deformation of the target area in the mountainous region based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter.

[0028] Optionally, the first computing module specifically includes:

[0029] The first calculation unit is used to determine the optimal estimate of the CR deformation parameter based on the prior information of each corner reflector using a solution space search method; the solution space search method determines the size, position and search step size of the solution space based on the prior information of each corner reflector; the prior information is the position information and deformation information obtained from GPS monitoring data.

[0030] Optional, also includes:

[0031] The weight reduction module is used to reduce the weight of the observations based on the iterative IGG-III method to eliminate the influence of gross errors in the observations; the observations include the DS deformation parameters.

[0032] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0033] This invention provides a method for monitoring surface deformation of landslides in mountainous areas, comprising: acquiring SAR images of a target area in a mountainous region; determining the coordinates of corner reflectors in the SAR images based on the CR-InSAR method and utilizing the backscattering characteristics of corner reflectors, and determining the CR deformation parameters of the target area in the mountainous region based on the coordinates of each corner reflector in the SAR images at adjacent time points; setting corner reflectors in the target area in the mountainous region; determining permanent scatterers in the SAR images based on the PS-InSAR method and performing temporal phase analysis on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region; using corner reflectors as control points and the PS deformation parameters as constraints, performing differential interferometry processing on SAR images at adjacent time points based on the DS-InSAR method to determine the deformation parameters of distributed scatterers in the SAR images; and monitoring the deformation of the target area in the mountainous region based on the CR deformation parameters, PS deformation parameters, and DS deformation parameters. This invention combines the advantages of CR-InSAR, PS-InSAR and DS-InSAR analysis methods. By using DS, it effectively increases the number of monitoring points in vegetation-covered areas, and by using PS, it identifies permanent scatterers with high phase stability on the surface of landslides in mountainous areas, thereby improving the reliability of deformation monitoring results on the surface of landslides in mountainous areas. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart of a method for monitoring surface deformation of landslides in mountainous areas, provided in an embodiment of the present invention.

[0036] Figure 2 A framework diagram for monitoring surface deformation of landslides in mountainous areas provided in an embodiment of the present invention;

[0037] Figure 3 is a Delaunay triangular network simulation diagram provided in the embodiment of the present invention; Figure 3(a) is a CR simulation triangular network diagram; Figure 3(b) is a CR / PS simulation triangular network diagram; Figure 3(c) is a CR / PS / DS simulation triangular network diagram.

[0038] Figure 4 is a schematic diagram of the landslide disaster deformation monitoring results provided in the embodiment of the present invention; Figure 4(a) is a schematic diagram of the CR-InSAR solution results; Figure 4(b) is a schematic diagram of the PS-InSAR solution results with CR as the control point; Figure 4(c) is a schematic diagram of the DS-InSAR solution results with CR as the control point and PS as the constraint condition. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] my country is one of the countries in the world most frequently affected by geological disasters, suffering extremely severe losses. These disasters are characterized by their diverse types, wide distribution, and high hazard. The fragile and sensitive geological environment, complex and variable meteorological conditions, and the impact of human activities are the main reasons for the frequent occurrence of geological disasters in my country. Landslides, due to their high frequency and wide range of impact, are one of the most serious disasters affecting people's lives, property, and public infrastructure, second only to earthquakes. Landslides in complex mountainous areas are particularly typical, characterized by their high degree of concealment, great hazard, and difficulty in monitoring and research. With the increasing frequency and scale of climate change and human activities, landslide disasters are showing a trend of increasing severity year by year, causing continuous sudden disasters that seriously threaten the safety of people's lives and property and critical infrastructure in mountainous areas.

[0041] Currently, among the methods for monitoring deformation in landslide disasters, precision leveling and GNSS technologies offer high monitoring accuracy, but suffer from low spatial resolution and limited monitoring range. They are typically deployed on known landslide bodies, making them highly susceptible to terrain conditions and posing significant challenges for instrument deployment and on-site measurements. Unmanned aerial vehicle (UAV) remote sensing is simple to operate and suitable for emergency response and real-time disaster assessment, but UAV platforms have low stability, limited image range, and face difficulties in image processing. LiDAR and ground-based synthetic aperture radar offer high-frequency observations and can perform high-precision continuous observations, but their monitoring range is small and suitable for monitoring single landslides. Spaceborne optical remote sensing is suitable for large-scale landslide detection and can be used to identify landslides with significant deformation, but it cannot observe landslides with potential slow deformation, and its monitoring accuracy is relatively low and highly susceptible to weather conditions.

[0042] Synthetic aperture radar interferometry (SAR) technology has been widely used in landslide deformation monitoring due to its high monitoring accuracy, high spatial resolution, high time repetition frequency, wide coverage, and minimal impact from weather conditions. It has unique advantages, particularly in large-scale potential landslide detection, long-term deformation monitoring of high-risk key targets, and reconstruction of historical landslide deformation. Therefore, conducting large-scale, long-term, and high-precision dynamic monitoring of landslide hazards in complex mountainous areas is of great significance for the early identification and monitoring of major geological disaster risks, protecting the lives and property of people in mountainous areas, and maintaining rapid economic development and social harmony and stability.

[0043] While InSAR technology has been widely applied to time-series monitoring of landslide surface deformation and has achieved significant progress, it still has many limitations in monitoring landslides in complex mountainous areas. In vegetated mountainous areas, conventional time-series InSAR technology struggles to detect a sufficient number of coherent radar targets. A sparse number of monitoring points cannot fully reflect the spatial deformation patterns of landslides and can easily lead to underestimation or even misjudgment of deformation. Existing time-series InSAR research on landslide monitoring in complex mountainous areas mostly employs Distributed Scatter Interferometry (DS-InSAR) to address the insufficient number of monitoring points. However, due to the lack of necessary control point information and constraints within the study area, the monitoring results are unreliable and contain large errors, severely hindering the interpretation and mechanism analysis of landslide disasters and greatly limiting the application of InSAR technology in landslide monitoring.

[0044] The purpose of this invention is to provide a highly reliable method and system for monitoring surface deformation of landslides in mountainous areas.

[0045] like Figure 1 As shown, the present invention provides a method for monitoring surface deformation of landslides in mountainous areas, comprising:

[0046] Step 1: Acquire SAR imagery of the target area in the mountainous region;

[0047] Step 2: Based on the CR-InSAR method, the coordinates of the corner reflectors in the SAR image are determined by utilizing the backscattering characteristics of the corner reflectors. Then, based on the coordinates of each corner reflector in the SAR image at adjacent time points, the CR deformation parameters of the mountainous target area are determined. The corner reflectors are located in the mountainous target area.

[0048] Step 3: Based on the PS-InSAR method, identify permanent scatterers in the SAR image and perform temporal phase analysis on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region;

[0049] Step 4: Using the corner reflector as the control point and the PS deformation parameter as the constraint, perform differential interferometry on the SAR images at adjacent time points based on the DS-InSAR method to determine the deformation parameters of the distributed scatterer in the SAR image.

[0050] Step 5: Monitor the deformation of the target area in the mountainous region based on the CR deformation parameters, the PS deformation parameters, and the DS deformation parameters.

[0051] This invention combines the advantages of CR, PS, and DS technologies: such as the high coherence of CR, the high phase quality of PS, and the high spatial density of DS. By constructing a three-level network with corner reflectors (CR) as control points, permanent scatterers (PS) as constraints, and distributed scatterers (DS) as unknown parameters, a deformation monitoring method that comprehensively considers the advantages of CR-, PS-, and DS-InSAR algorithms is realized.

[0052] In SAR imagery, the signal of each resolution cell is a coherent superposition of the signals from all discrete scatterers within that cell. If the phase of a cell is dominated by a target with strong scattering characteristics and maintains high phase stability over a long time series, that resolution cell is called a permanent scatterer. A PS (Permanent Scatterer) is a ground feature that maintains strong and stable scattering characteristics to radar waves, and its recorded phase information has a high signal-to-noise ratio and is less affected by spatiotemporal uncorrelation. PS targets are widely present in urban areas, and in densely vegetated mountainous areas, they appear as exposed rocks, sparse buildings, etc.

[0053] If a pixel contains no dominant scatterer, and all discrete scatterers have similar properties, their contributions to the pixel's phase are roughly equal, and they exhibit strong randomness, this pixel is called a distributed scatterer. Its phase exhibits instability over time. Distributed scatterers exist in natural scenes, with low backscattering rates, corresponding to areas with low grayscale values ​​in SAR images, such as crops, grasslands, and deserts. If a pixel in a SAR image contains only one stable, strong scatterer, the pixel's amplitude and phase information signal-to-noise ratio are very high, maintaining very high coherence over a long period. Deploying artificial corner reflectors in areas with low coherence can effectively compensate for situations where monitoring targets are scarce or impossible to detect in complex mountainous regions.

[0054] In some embodiments, based on the CR-InSAR method, the coordinates of corner reflectors in SAR images are determined using the backscattering characteristics of corner reflectors. Then, based on the coordinates of each corner reflector in the SAR images at adjacent time points, the CR deformation parameters of the mountainous target area are determined. Specifically, this can be done as follows:

[0055] Based on the solution space search method, the optimal estimate of the CR deformation parameter is determined according to the prior information of each corner reflector; the solution space search method determines the size, position and search step size of the solution space according to the prior information of each corner reflector; the prior information is the position information and deformation information obtained from GPS monitoring data.

[0056] Specifically, the nonlinear CR-InSAR method based on phase coherence uses the backscattering characteristics of corner reflectors to identify targets, determines the optimal estimate of deformation parameters through solution space search and prior information, treats atmospheric phase and noise phase as random errors, and uses the least squares method to solve for the parameters of nonlinear deformation, thereby obtaining the results of CR nonlinear time-series deformation.

[0057] Corner reflectors are ground features with special geometries that can be identified in synthetic aperture radar (SAR) intensity images because they produce strong backscattering in the echo. This backscattering characteristic makes corner reflectors a type of target that can be used to measure surface deformation.

[0058] Specifically, in the CR-InSAR method, a solution space search is used, combining prior information such as location and deformation information from GPS monitoring data, to determine the solution space range and search step size for deformation parameters. Solution space search is an iterative process; by tracking the optimal solution within the solution space, the optimal value of the deformation parameters can be quickly estimated.

[0059] CR deformation parameters refer to Corner Reflector (CR) deformation parameters, including CR deformation rate and CR time-series cumulative deformation. These parameters are used to describe surface deformation or crustal movement. CR deformation parameters are commonly used for deformation monitoring and research using InSAR (Interferometric Synthetic Aperture Radar) technology.

[0060] Specifically, CR deformation parameters are typically based on phase information acquired using InSAR technology. By analyzing interferometric images at different times and calculating the phase difference, changes in surface deformation or crustal movement are revealed. CR deformation parameters aim to provide continuous, regional deformation monitoring results to understand the deformation process of the target area's surface.

[0061] Specifically, CR deformation parameters typically involve the following key concepts:

[0062] Regionality: The CR deformation parameter employs a regional calculation method, extending the focus to the entire region rather than just local deformation changes. Therefore, it can provide more comprehensive information on surface deformation, covering a wider area.

[0063] Continuity: CR deformation parameters can provide continuous time-series deformation information, that is, monitor the evolution of surface deformation over a period of time. By acquiring multi-temporal interferometric image data and performing phase analysis and time-series analysis, continuous deformation parameters can be calculated.

[0064] Deformation parameters: CR deformation parameters typically include deformation rate and cumulative deformation over time series, used to describe the rate and spatial distribution of surface deformation. These parameters can reflect changes in surface deformation, such as surface uplift, subsidence, and horizontal displacement.

[0065] In some embodiments, based on the PS-InSAR method, permanent scatterers in the SAR image are identified, and temporal phase analysis is performed on the permanent scatterers to determine the PS deformation parameters of the mountainous target area. Specifically, this may include:

[0066] 1) Selecting the main image: Select a main image as the reference based on the temporal baseline (the time interval between adjacent images), the spatial baseline (the positional difference between adjacent images), and the Doppler center.

[0067] 2) Generate a single master image differential interferogram: Perform differential interferometry on the master image and other images to obtain an interferogram, which records the deformation information of the Earth's surface.

[0068] 3) Remove flat and topographic phases: Interferograms contain interference from flat and topographic phases, which need to be removed in order to extract surface deformation information more accurately.

[0069] 4) Selecting PS candidate points: Using the amplitude dispersion exponential thresholding method, select some candidate points (Persistent Scatterers, PS) in the intensity map. These points exhibit relatively stable amplitude characteristics in multiple images.

[0070] 5) Estimate spatially correlated and uncorrelated viewpoint errors: For PS candidate points, estimate their spatially correlated and uncorrelated viewpoint errors. These errors are caused by interferometric tasks with different orientations and line-of-sight distances.

[0071] 6) Evaluation of phase stability and selection of final PS points: Phase stability is evaluated by pixel time correlation, and stable PS points are selected. These points can provide high-quality deformation measurements.

[0072] 7) Three-dimensional phase unwrapping: Using the interferometric phase that has been corrected for DEM error and viewing angle error, three-dimensional phase unwrapping is performed to obtain accurate measurement results of surface deformation.

[0073] 8) Time and spatial domain filtering: The unwrapped phase is filtered by high-pass filtering in the time domain and low-pass filtering in the spatial domain to eliminate noise and smooth the deformation field, thus obtaining the deformation rate and time series results.

[0074] Among them, the PS deformation parameter refers to the Persistent Scatterer (PS) deformation parameter, which is used to describe the deformation information of the Persistent Scatterer points extracted by the Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR) technology.

[0075] Specifically, PS-InSAR is a technique for deformation monitoring and research by analyzing Persistent Scatterer Points (PS points) in remote sensing interferometric synthetic aperture radar (InSAR) data, including PS deformation rate and PS time-series cumulative deformation. Persistent Scatterer Points are surface points that stably scatter signals at multiple time points, typically composed of man-made and natural scatterers such as buildings, bridges, and rocks. PS deformation parameters are obtained through temporal phase analysis and change detection of Persistent Scatterer Points (PS points / permanent scatterers).

[0076] Specifically, the PS deformation parameters include the following key parameters:

[0077] 1) Deformation rate: The PS deformation rate represents the rate of deformation at a certain location on the Earth's surface, usually expressed in units such as millimeters or centimeters per year. It is calculated by observing the interferometric phase changes of the Persistent Scatterer over a time series.

[0078] 2) Time Series: PS deformation parameters are typically presented as a time series, showing the deformation of the Persistent Scatterer point at different times. This provides a comprehensive understanding of the deformation evolution process, including fluctuations, trends, and seasonal variations in deformation.

[0079] In some embodiments, using the corner reflector as a control point and the PS deformation parameter as a constraint, differential interferometry is performed on the SAR images at adjacent time points based on the DS-InSAR method to determine the deformation parameters of the distributed scatterer in the SAR image. Specifically, this can be done as follows:

[0080] CR (Constant Resonance) can be used as a control point for InSAR phase inversion because it maintains good coherence over long periods and has high signal-to-noise ratios for both amplitude and phase. PS (Peripheral Targets) are typically ground targets with strong and stable scattering characteristics, and their signal-to-noise ratios are less affected by spatial / temporal decorrelation; therefore, the PS deformation parameters are used as constraints. Then, DS (Distributed Targets) are connected with CR and PS through a Delaunay triangulation, and the deformation parameters of DS are used as unknown parameters to be solved, as shown in Figures 3(a), (b), and (c), which are simulated triangulation networks combining CR(a), PS(b), and DS(c). Black triangles represent CR, dark gray circles represent PS, and light gray rectangles represent DS.

[0081] Specifically, the robust estimation-based CR-, PS-, and DS-InSAR three-level fusion method is a way to obtain more accurate deformation information by utilizing different types of InSAR data. This method combines CR data with phase stability and high signal-to-noise ratio, PS data of stable scattering targets, and DS data, and uses Delaunay triangulation to connect these data to estimate deformation parameters.

[0082] First, CR data is obtained from multiple consecutive SAR observations, exhibiting good coherence and a high signal-to-noise ratio. Therefore, CR data can serve as control points for InSAR phase inversion, remaining relatively stable over time. The phase information from CR data can be used to estimate surface deformation.

[0083] PS (Polarization Point) data typically originates from stable scattering targets, such as buildings and bridges, which possess strong and stable scattering characteristics. PS points have a high spatial density and are relatively unaffected by spatial / temporal decorrelation. Therefore, PS data can be used as a constraint on deformation parameters, providing stable deformation information.

[0084] Distributed scattering (DS) data consists of scattering from point targets over a large area, such as crops, grasslands, and deserts. DS data has a high density of scattering points but a low phase signal-to-noise ratio. By connecting DS points with CR and PS data using a Delaunay triangulation, the deformation parameters of the DS data can be solved as unknowns.

[0085] Delaunay triangulation is a non-overlapping triangulation network, characterized by all triangles having circumcircles that do not contain other data points. In the three-level fusion method, CR, PS, and DS data points are connected using Delaunay triangulation to form a network. Then, the deformation parameters in this network are used for calculation to obtain more accurate deformation information.

[0086] By fusing different types of InSAR data in a three-level manner, the advantages of each data type can be fully utilized to improve the accuracy and reliability of deformation monitoring. CR data provides phase stability and high signal-to-noise ratio over long periods, PS data provides stable deformation constraints, and DS data provides deformation information for local areas. This fusion method can be applied to fields such as geological disaster monitoring and urban deformation monitoring, providing more accurate deformation analysis results.

[0087] In addition, during DS-InSAR processing, SAR data undergoes preprocessing steps such as cropping, registration, baseline estimation, removal of flat terrain effects and topographic phase, etc. These preprocessing steps are mainly performed in Doris software. Then, through a self-developed DS point target selection and optimal phase estimation module, the identified DS point targets are selected and the optimal phase is extracted. Finally, this data is imported as input into standard StaMPS software for further deformation analysis and interpretation.

[0088] Specifically, homogeneous pixel identification: In DS-InSAR, homogeneous pixel identification aims to increase the spatial density of measurement points and address the problem of sparse target points within the monitoring area. This step aims to identify pixels with similar phase evolution characteristics at multiple time points and define them as homogeneous pixels. By using homogeneous pixels, more stable scattering compensation points can be selected from the original interferometric image, thereby improving the spatial resolution and accuracy of deformation monitoring.

[0089] Optimal phase estimation: Optimal phase estimation is another crucial step in the DS-InSAR method. In the time domain, it is equivalent to filtering the data to improve the reliability of phase information for DS point targets. Optimal phase estimation calculates and models the phase of identified DS point targets, using appropriate algorithms and filtering techniques to extract phase-varying and deformation-related signals, thereby better capturing deformation signals and reducing non-deformation-related noise and errors.

[0090] In some embodiments, before monitoring the deformation of the target mountain landslide surface based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter, the method may further include:

[0091] Based on the iterative IGG-III method, the observations are weighted to eliminate the influence of gross errors in the observations; the observations include the DS deformation parameters.

[0092] Among them, the DS deformation parameters include the DS deformation rate and the cumulative deformation of the DS time series.

[0093] Specifically, to mitigate the impact of spatial correlation errors (such as atmospheric delay) and thus obtain accurate surface deformation information, the interferometric phase of adjacent monitoring points is differentially analyzed. Similar to leveling networks in geodesy, after identifying all monitoring points, observation edges are established for each pair of monitoring points that meet distance constraints. This forms a monitoring network consisting of the monitoring points and the observation baseline edges. Typically, a Delaunay triangulation network is used to construct the monitoring network. Based on the adjustment principle, the deformation rate and elevation correction for each monitoring point are obtained using indirect adjustment methods. A functional model for the indirect adjustment of the baseline network is established, and the optimal estimate of the deformation parameters can be obtained through least squares. However, the least squares criterion is the optimal unbiased estimate, which can balance errors but cannot resist gross errors. Therefore, this patent adopts a robust estimation method to reduce the influence of gross errors in the observations. It uses the IGGIII method based on the median in the weighted iterative method to reduce the weight of the observations containing gross errors. After multiple iterations, the weight function of the abnormal observations containing gross errors is 0 or close to 0, so as to achieve the optimal or near-optimal under the premise of robustness. This not only resists model bias but also resists gross errors, making the results more reliable.

[0094] In some embodiments, such as Figure 2 As shown, the monitoring process for surface deformation of landslides in mountainous areas is as follows:

[0095] 1) Acquire SAR data and identify corner reflectors from the SAR intensity map by utilizing their strong backscattering characteristics. Using a solution space search method, determine the size, location, and search step size of the solution space based on the location and deformation information provided by GPS monitoring data, which can quickly obtain the optimal estimate of the CR deformation rate.

[0096] 2) Candidate PS points are selected using the amplitude discrete exponential thresholding method, spatially correlated and uncorrelated viewpoint errors are estimated, and the final PS point is selected by evaluating phase stability through pixel temporal correlation. The interferometric phase, after correcting for DEM and viewpoint errors, undergoes three-dimensional phase unwrapping. The unwrapped phase is then subjected to high-pass filtering in the time domain and low-pass filtering in the spatial domain to finally obtain the PS deformation rate and time series results.

[0097] 3) DS identification can use homogeneous pixel identification method, which can increase the spatial density of measurement points and solve the problem of the small number of monitoring point targets in the study area. The optimal phase estimation can be equivalent to the filtering operation in the time domain, which can improve the reliability of the phase information of DS point targets.

[0098] 4) CR, due to its high signal-to-noise ratio (SNR) for both amplitude and phase, can be used as a control point for InSAR phase inversion. PS's SNR is less affected by spatial / temporal decorrelation; therefore, PS deformation parameters are used as constraints. Then, DS, CR, and PS are connected together via a Delaunay triangulation network, and the deformation parameters of DS are solved as unknown parameters. Finally, robust estimation is used to mitigate the impact of gross errors in deformation phase inversion.

[0099] In some embodiments, Figures 4(a), (b), and (c) show the monitoring results of the Jiaju landslide in Sichuan. When monitoring the Jiaju landslide in Sichuan, the present invention first requires identifying CR targets. CR deformation information can be obtained by extracting the deformation of CR targets. Although CR has a low spatial resolution, its amplitude and phase signal-to-noise ratio is high, so it can be used as a control point. Secondly, PS targets can be identified by using the amplitude deviation index thresholding method. PS targets are selected by evaluating phase stability through pixel temporal correlation. PS has a high signal-to-noise ratio, and although its distribution is relatively sparse, it can be used as a constraint condition. Next, DS targets are identified by using a homogeneous pixel identification method. DS targets have high spatial resolution, which can meet the needs of large-scale monitoring. Using Delaunay triangulation, we connect DS with CR and PS, and solve for the deformation parameters of DS as unknown parameters. Finally, through this method, we can obtain relatively detailed spatial distribution characteristics and temporal evolution patterns of the deformation of the Jiaju landslide.

[0100] This invention also provides a surface deformation monitoring system for landslides in mountainous areas, comprising:

[0101] The image acquisition module is used to acquire SAR images of target areas in mountainous regions.

[0102] The first calculation module is used to determine the coordinates of corner reflectors in SAR images based on the CR-InSAR method and utilizing the backscattering characteristics of corner reflectors, and to determine the CR deformation parameters of the mountainous target area based on the coordinates of each corner reflector in the SAR images at adjacent time points; the corner reflectors are set in the mountainous target area.

[0103] The second calculation module is used to determine the permanent scatterers in the SAR image based on the PS-InSAR method, and to perform temporal phase analysis on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region.

[0104] The third calculation module is used to perform differential interferometry processing on the SAR images at adjacent time points using the corner reflector as the control point and the PS deformation parameters as constraints, based on the DS-InSAR method, to determine the deformation parameters of the distributed scatterers in the SAR images.

[0105] The monitoring module is used to monitor the deformation of the target area in the mountainous region based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter.

[0106] In some embodiments, the first computing module specifically includes:

[0107] The first calculation unit is used to determine the optimal estimate of the CR deformation parameter based on the prior information of each corner reflector using a solution space search method; the solution space search method determines the size, position and search step size of the solution space based on the prior information of each corner reflector; the prior information is the position information and deformation information obtained from GPS monitoring data.

[0108] In some embodiments, the system further includes:

[0109] The weight reduction module is used to reduce the weight of the observations based on the iterative IGG-III method to eliminate the influence of gross errors in the observations; the observations include the DS deformation parameters.

[0110] In summary, the present invention has the following advantages.

[0111] 1. This invention combines the advantages of CR-InSAR, PS-InSAR and DS-InSAR analysis methods. By using DS, it effectively increases the number of monitoring points in vegetation-covered areas. By using PS, it identifies permanent scatterers with high phase stability on the surface of landslides in mountainous areas. This allows for an accurate description of the temporal evolution and spatial distribution characteristics of deformation in complex mountainous landslide disasters, thereby improving the reliability of landslide surface deformation monitoring results in mountainous areas.

[0112] 2. The maximum likelihood estimation method in robust estimation is used to reduce the impact of gross errors in the observations. The IGGIII method based on the median in the weighted iterative method is used to reduce the weight of the observations containing gross errors. After multiple iterations, the optimal or near-optimal results are achieved under the premise of robustness, making the results more reliable.

[0113] 3. Landslides in complex mountainous areas are characterized by their high degree of concealment, significant hazard, and difficulty in monitoring and research. Traditional ground subsidence monitoring technologies based on methods such as leveling require extensive fieldwork, are intensive, time-consuming, and costly, and have limited coverage and sampling density, making them unsuitable for large-scale surface deformation surveys and monitoring. Synthetic Aperture Radar Interferometry (SAR) technology, with its high monitoring accuracy, high spatial resolution, high time repetition frequency, wide coverage, and minimal impact from weather conditions, has been widely applied in landslide deformation monitoring. It has unique advantages, particularly in large-scale potential landslide detection, long-term deformation monitoring of high-risk key targets, and reconstruction of historical landslide deformation. In vegetated mountainous areas, conventional time-series InSAR technology struggles to detect a sufficient number of coherent radar targets. A sparse number of monitoring points cannot fully reflect the spatial deformation patterns of landslides and can easily lead to underestimation or even misjudgment of deformation. This invention combines InSAR analysis methods using corner reflectors, permanent scatterers, and distributed scatterers. By extracting high-precision deformation of landslides, it facilitates the coupling analysis of SAR landslide deformation observation results with triggering factors, providing important decision support for early warning and forecasting of landslide disasters. High-precision landslide monitoring is crucial for the study of landslide disaster patterns.

[0114] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for monitoring surface deformation of landslides in mountainous areas, characterized in that, include: Acquire SAR imagery of target areas in mountainous regions; Based on the CR-InSAR method, the coordinates of corner reflectors in SAR images are determined by utilizing the backscattering characteristics of corner reflectors, and the CR deformation parameters of the mountainous target area are determined based on the coordinates of each corner reflector in the SAR images at adjacent time points. The corner reflector is installed in the target area of ​​the mountainous region; Based on the PS-InSAR method, permanent scatterers in the SAR image are identified, and temporal phase analysis is performed on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region. Using the corner reflector as the control point and the PS deformation parameter as the constraint, differential interferometry is performed on the SAR images at adjacent time points based on the DS-InSAR method to determine the deformation parameters of the distributed scatterer in the SAR image. Based on the iterative IGG-III method, the observations are weighted to eliminate the influence of gross errors in the observations; the observations include DS deformation parameters. The deformation of the target area in the mountainous region is monitored based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter.

2. The method for monitoring surface deformation of landslides in mountainous areas according to claim 1, characterized in that, The step of determining the CR deformation parameters of the mountainous target area based on the coordinates of each corner reflector in the SAR image at adjacent time points specifically includes: Based on the solution space search method, the optimal estimate of the CR deformation parameter is determined according to the prior information of each corner reflector; the solution space search method determines the size, position and search step size of the solution space according to the prior information of each corner reflector; the prior information is the position information and deformation information obtained from GPS monitoring data.

3. The method for monitoring surface deformation of landslides in mountainous areas according to claim 1, characterized in that, The monitoring of deformation in the mountainous target area based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter specifically includes: Construct a Delaunay triangulation network for CR, PS, and DS; Based on the Delaunay triangulation network, the time series of deformation rates at each location in the target mountain area are determined according to the CR deformation parameters, the PS deformation parameters, and the DS deformation parameters.

4. The method for monitoring surface deformation of landslides in mountainous areas according to claim 1, characterized in that, The CR deformation parameters are continuous, regional deformation monitoring results, including CR deformation rate and CR time series cumulative deformation.

5. The method for monitoring surface deformation of landslides in mountainous areas according to claim 1, characterized in that, The PS deformation parameters include the PS deformation rate and the PS time series cumulative deformation.

6. The method for monitoring surface deformation of landslides in mountainous areas according to claim 1, characterized in that, The DS deformation parameters include the DS deformation rate and the cumulative deformation of the DS time series.

7. A surface deformation monitoring system for landslides in mountainous areas, characterized in that, include: The image acquisition module is used to acquire SAR images of target areas in mountainous regions. The first calculation module is used to determine the coordinates of corner reflectors in SAR images based on the CR-InSAR method and utilizing the backscattering characteristics of corner reflectors, and to determine the CR deformation parameters of the mountainous target area based on the coordinates of each corner reflector in the SAR images at adjacent time points. The corner reflector is installed in the target area of ​​the mountainous region; The second calculation module is used to determine the permanent scatterers in the SAR image based on the PS-InSAR method, and to perform temporal phase analysis on the permanent scatterers to determine the PS deformation parameters of the target area in the mountainous region. The third calculation module is used to perform differential interferometry processing on the SAR images at adjacent time points based on the DS-InSAR method, using the corner reflector as the control point and the PS deformation parameter as the constraint, to determine the deformation parameter of the distributed scatterer in the SAR image. The weight reduction module is used to reduce the weight of observations based on the iterative IGG-III method to eliminate the influence of gross errors in the observations; the observations include DS deformation parameters; The monitoring module is used to monitor the deformation of the target area in the mountainous region based on the CR deformation parameter, the PS deformation parameter, and the DS deformation parameter.

8. The mountain landslide surface deformation monitoring system according to claim 7, characterized in that, The first calculation module specifically includes: The first calculation unit is used to determine the optimal estimate of the CR deformation parameter based on the prior information of each corner reflector using a solution space search method; the solution space search method determines the size, position and search step size of the solution space based on the prior information of each corner reflector; the prior information is the position information and deformation information obtained from GPS monitoring data.

Images