A method and device for estimating precision of MT-InSAR based on a sparse parameter model
By constructing a sparse observation matrix and an iterative least squares estimation through a sparse parameter model and a double-difference observation model, the limitations of the applicability of accuracy assessment and the influence of errors in multi-temporal InSAR technology are solved, and more accurate deformation monitoring results are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN UNIV
- Filing Date
- 2023-07-10
- Publication Date
- 2026-05-19
AI Technical Summary
Existing multi-temporal InSAR technology accuracy assessment methods have limited applicability and cannot accurately analyze data, especially in geological disaster monitoring where errors can lead to inaccurate deformation monitoring results.
A sparse parameter model is adopted, and a sparse observation matrix is constructed by multi-temporal InSAR deformation interferograms. A sparse parameter model is established, a double-difference observation model is constructed, iterative least squares estimation is performed, and the influence of atmospheric factors is removed through variance test. The deformation time series and accuracy of InSAR observation points are evaluated.
It effectively suppresses the influence of errors and noise in data processing, improves the accuracy of InSAR deformation monitoring, solves the problem of limited applicability of existing methods, and provides higher accuracy deformation time series assessment.
Smart Images
Figure CN116879854B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of surveying and mapping, and in particular to an MT-InSAR accuracy estimation method and apparatus based on a sparse parameter model. Background Technology
[0002] Multi-temporal InSAR, as a novel geodetic technique for monitoring surface deformation, has been widely used in various applications related to ground deformation monitoring for geological hazards and in urban infrastructure safety assessments. However, errors such as satellite orbit inaccuracies, atmospheric effects, and decorrelation noise can adversely affect the accuracy of deformation monitoring. Therefore, assessing the errors in InSAR measurements is crucial to the overall accuracy of deformation results. Currently, the accuracy assessment of multi-temporal InSAR deformation monitoring compares it with field measurements (e.g., leveling and Global Navigation Satellite Systems). However, in most cases, the spatial density of field measurement results is very limited and insufficient to prove the accuracy of InSAR deformation monitoring results. Furthermore, the difficulty in accurately modeling various error sources in multi-temporal InSAR data processing leads to unreliable stochastic models. For example, using an inappropriate deformation model during multi-temporal InSAR data processing may generate unmodeled deformation information, thus affecting the stochastic model and causing significant deviations in the final parameter estimates.
[0003] Furthermore, when monitoring ground deformation associated with various geological hazards, such as seismic activity and land subsidence, the accuracy assessment of multi-temporal InSAR deformation monitoring typically requires comparison with field measurements (such as leveling and GNSS measurements). However, since field measurements usually have very limited coverage, they often fail to reflect the reliability of each InSAR observation. On the other hand, using stochastic models is one of the traditional methods for assessing the reliability of multi-temporal InSAR deformation monitoring results; however, stochastic models themselves are affected by various factors, such as unknown atmospheric effects, which can lead to significant biases in accuracy assessment.
[0004] Therefore, existing methods for assessing the accuracy of multi-temporal InSAR technology need further improvement. Summary of the Invention
[0005] The technical problem to be solved by the present invention is to provide an MT-InSAR accuracy estimation method and apparatus based on a sparse parameter model, in order to address the shortcomings of existing InSAR measurement accuracy assessment methods, which have limited applicability and cannot accurately analyze the data.
[0006] The technical solution adopted by this invention to solve the technical problem is as follows:
[0007] In a first aspect, the present invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model, comprising:
[0008] A sparse observation matrix is constructed based on multi-temporal InSAR deformation interferometry, and a sparse parameter model is established by performing stability screening based on the sparse observation matrix.
[0009] Based on the arc segments of the spatial network in the sparse parameter model, a double-difference observation model is constructed.
[0010] Iterative least squares estimation is performed based on the sparse parameter model, and the variance is tested. Atmospheric factors are removed by the double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0011] In one implementation, constructing a sparse observation matrix based on the multi-temporal InSAR deformation interferogram includes:
[0012] The time buffer in the multi-temporal InSAR deformation interferogram is determined, and multiple sparse observation matrices are established based on the time buffer.
[0013] In one implementation, determining the time buffer in the multi-temporal InSAR deformation interferogram includes:
[0014] A time buffer is established in the multi-temporal InSAR deformation interferogram with a preset time period as the center.
[0015] In one implementation, the step of performing stability screening based on the sparse observation matrix and establishing a sparse parameter model includes:
[0016] Select the matrix that satisfies the conditions from multiple sparse observation matrices to obtain the final sparse matrix;
[0017] The sparse parameter model is established based on the final sparse matrix, the preset time period, and the multi-temporal InSAR deformation interferogram.
[0018] In one implementation, constructing a double-difference observation model based on arc segments of the spatial network in the sparse parameter model includes:
[0019] Obtain the arc segments of the spatial network in the sparse parameter model, use the arc segments in the spatial network as observations, and determine the parameters to be estimated on the arc segments;
[0020] The parameters to be estimated and their corresponding covariances are propagated from the time interval to the sparse sequence, and the deformation and covariance of the points are obtained through the adjustment of the spatial network to construct a double-difference observation model.
[0021] In one implementation, the iterative least-squares estimation based on the sparse parameter model, followed by a variance test to obtain the deformation time series of each InSAR observation point and the corresponding accuracy assessment results, includes:
[0022] Based on the sparse parameter model, the sparse parameter estimation method is used to estimate the sparse parameters.
[0023] The variance of the parameters obtained by the iterative least squares estimation method is verified to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0024] In one implementation, the step of performing variance verification on the parameters obtained through the iterative least squares estimation method further includes:
[0025] The parameters that have been estimated are then passed through a double-difference observation model to remove the influence of atmospheric factors, in order to obtain more accurate deformation time series of each InSAR observation point and the corresponding accuracy assessment results.
[0026] Secondly, the present invention provides an MT-InSAR accuracy estimation device based on a sparse parameter model, comprising:
[0027] The sparse parameter model module is used to construct a sparse observation matrix based on multi-temporal InSAR deformation interferograms, perform stability screening based on the sparse observation matrix, and establish a sparse parameter model.
[0028] The double-difference observation model module is used to construct a double-difference observation model based on the arc segments of the spatial network in the sparse parameter model.
[0029] The verification result module is used to perform iterative least squares estimation based on the sparse parameter model and verify it based on the variance. Atmospheric factors are removed through the double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0030] Thirdly, the present invention provides a terminal, comprising: a processor and a memory, wherein the memory stores an MT-InSAR accuracy estimation program based on a sparse parameter model, and the MT-InSAR accuracy estimation program based on a sparse parameter model is executed by the processor to implement the operation of the MT-InSAR accuracy estimation method based on a sparse parameter model as described in the first aspect.
[0031] Fourthly, the present invention also provides a medium, which is a computer-readable storage medium, storing an MT-InSAR accuracy estimation program based on a sparse parameter model, wherein the MT-InSAR accuracy estimation program based on a sparse parameter model, when executed by a processor, is used to implement the operation of the MT-InSAR accuracy estimation method based on a sparse parameter model as described in the first aspect.
[0032] The present invention, by employing the above technical solution, has the following effects:
[0033] This invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model. A sparse parameter model is established using multi-temporal InSAR deformation interferograms to evaluate the accuracy of InSAR deformation time series. The sparse parameter model estimates the sparsed time series deformation results, and redundant InSAR time series data can be generated during the estimation process for accuracy assessment of the estimated results. A double-difference observation model is constructed using arc segments of the spatial network in the sparse parameter model to remove the influence of atmospheric factors. By constructing the sparse parameter model and the double-difference observation model, this invention effectively suppresses the influence of errors and noise during data processing, thus solving the problems of limited applicability and inaccurate analysis in existing InSAR measurement accuracy assessment methods. 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 description of the embodiments or the prior art 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 the structures shown in these drawings without creative effort.
[0035] Figure 1 This is a flowchart of an MT-InSAR accuracy estimation method based on a sparse parameter model in one implementation of the present invention.
[0036] Figure 2 This is a schematic diagram of the condition number for establishing a sparse model design matrix using different CTPs in one implementation of the present invention.
[0037] Figure 3 This is a schematic diagram of the deformation rate of the Houston area in one implementation of the present invention.
[0038] Figure 4 This is a schematic diagram comparing the time series InSAR (SBAS and SPM) and GPS deformation time series and their uncertainty range in one implementation of the present invention.
[0039] Figure 5This is a schematic diagram comparing the mean square error and LOS deformation time series between InSAR (SBAS and SPM) and GPS deformation results in one implementation of the present invention.
[0040] Figure 6 This is a functional schematic diagram of the terminal in one implementation of the present invention.
[0041] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0042] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0043] Exemplary methods
[0044] In existing methods for deformation monitoring using multi-temporal InSAR, the unreliability of stochastic models arises because various error sources in multi-temporal InSAR data processing are difficult to accurately model. For example, using an inappropriate deformation model during multi-temporal InSAR data processing may generate unmodeled deformation information, thus affecting the stochastic model and leading to significant deviations in the final parameter estimation. Furthermore, the spatial density of field measurement results is very limited, insufficient to guarantee the accuracy of InSAR time-series deformation results.
[0045] To address the aforementioned problems, this invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model. A sparse parameter model is established using multi-temporal InSAR deformation interferograms to evaluate the accuracy of InSAR deformation time series. The sparse parameter model estimates the sparsed time series deformation results, and redundant InSAR time series data can be generated during the estimation process for accuracy assessment of the estimated results. A double-difference observation model is constructed using arc segments of the spatial network in the sparse parameter model to remove the influence of atmospheric factors. This invention, by establishing a sparse parameter model and a double-difference observation model, effectively suppresses the influence of errors and noise during data processing, thereby solving the problems of limited applicability and inaccurate analysis in existing InSAR measurement accuracy assessment methods.
[0046] like Figure 1 As shown, this embodiment of the invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model, comprising the following steps:
[0047] Step S100: Construct a sparse observation matrix based on multi-temporal InSAR deformation interferograms, perform stability screening based on the sparse observation matrix, and establish a sparse parameter model.
[0048] In this embodiment, the MT-InSAR accuracy estimation method based on the sparse parameter model is applied to a terminal, which includes, but is not limited to, devices such as computers and mobile terminals; the terminal is equipped with a training and transfer platform for the MT-InSAR accuracy estimation model based on the sparse parameter model.
[0049] The MT-InSAR accuracy estimation method based on a sparse parameter model provided in this invention constructs a sparse observation matrix through multi-temporal InSAR deformation interferograms, and establishes a sparse parameter model based on the stability screening of the sparse observation matrix.
[0050] In this embodiment, before constructing a sparse observation matrix using multi-temporal InSAR deformation interferograms, it is necessary to generate SLC (Single Look Complex) images from raw data, perform SLC registration to generate high-quality SAR images, generate interferograms using an interferometric network, and generate deformation interferograms using an external DEM (Digital Elevation Model). After filtering the interferograms, phase unwrapping is performed to obtain the multi-temporal InSAR deformation interferograms.
[0051] In this embodiment, N SAR images are acquired, and M differential deformation interferograms are generated based on conventional short-spatial baseline thresholds and considering the connectivity of the interferogram network; wherein All differential deformation interferograms are generated by phase unwrapping, where the unwrapped phase of the i-th pixel in one interferogram is... It can be represented as:
[0052]
[0053]
[0054]
[0055] in, It is the residual topographic phase (DEM error phase) after the surface topography is compensated by external DEM; It is the deformation phase component corresponding to the deformation between two SAR image acquisitions; It is an atmospheric phase screen (APS) that can be compensated by the joint model method; It is the phase corresponding to the satellite orbit error; Represents the phase of the decorrelated noise; It is the residual phase, which includes APS, orbital error, and decorrelation noise phase; andΔh i The unknown parameters to be estimated are deformation and DEM error along the line of sight (LOS); B ⊥ ρ is the vertical baseline of the interferogram; ρ, θ, and λ are the distance between the satellite and the ground target, the angle of incidence, and the radar wavelength.
[0056] In this embodiment, when all M unwrapped differential deformation interferograms are generated, equation (2) can be expressed by equation (3), becoming a linear inversion problem with the following observation equation and stochastic model:
[0057]
[0058]
[0059] Where A represents the design matrix; This represents the deformation of a SAR image within a continuous time interval. For example, if m-th interferograms cover n (n∈1,2,…N-1) intervals, then only the corresponding column of the m rows in matrix A is equal to 1. The rank of matrix A is designed to be R(A)=N; Q is the auxiliary factor matrix of the observations L. This is the initial variance factor. |D L |≠0 and R(D) L ) = N. Δh is the DEM error. In equation (3), due to the correlation between the M interferograms, the number of unknowns (N) is greater than the number of independent observations (N-1). Therefore, the solution to equation (3) is not unique.
[0060] In this embodiment, the multi-temporal InSAR deformation interferogram obtained by the above method can be used as observation data for subsequent time-series InSAR data processing; then, multiple sparse observation matrices can be established based on the multi-temporal InSAR deformation interferogram, thereby establishing the sparse parameter model.
[0061] Currently, various technical methods (such as SBAS) have been proposed to reduce the number of unknowns in X by predefining a specific deformation model and to obtain a unique solution through least squares estimation. However, due to the influence of the aforementioned error factors on the stochastic model, traditional methods are difficult to perform accurate reliability analysis on deformation sequences or have low accuracy in reliability analysis.
[0062] In this embodiment, the proposed Sparse Parameter Model (SPM) does not predefine a specific deformation or random model, but reduces the unknown X in equation (3) by estimating a sparse deformation sequence, such as a monthly deformation sequence. For time series of SAR images with high time sampling rates, such as Sentinel-1A / B, the above method is particularly effective in slowly deformed regions.
[0063] Specifically, in one implementation of this embodiment, step S100 includes the following steps:
[0064] Step S110: Determine the time buffer in the multi-temporal InSAR deformation interferogram, and establish multiple sparse observation matrices based on the time buffer.
[0065] In this embodiment, a preset time period is established for the multi-temporal InSAR deformation interferogram. A time buffer for the multi-temporal InSAR deformation interferogram is established with the preset time period as the center. Multiple sparse observation matrices are established based on the time buffer to facilitate the subsequent construction of a sparse parameter model.
[0066] Specifically, in one implementation of this embodiment, step S110 includes the following steps:
[0067] Step S111: Establish a time buffer in the multi-temporal InSAR deformation interferogram centered on a preset time period.
[0068] Step S112: Select the matrix that meets the conditions from the multiple sparse observation matrices to obtain the final sparse matrix;
[0069] Step S113: Establish the sparse parameter model based on the final sparse matrix, the preset time period, and the multi-temporal InSAR deformation interferogram.
[0070] In this embodiment, a deformation-related time period (CTP) is established (i.e., a preset time period). A time buffer is established with the established CTP as the center. Multiple sparse matrices are established based on all CTPs in the time buffer. The matrix with the smallest condition number is selected as the final sparse matrix. Based on the selected CTP value, it is assumed that the deformation within the time period is linear. The sparse parameter model is constructed based on the interferometric network diagram and the sparse observation matrix.
[0071] Specifically, firstly, to define a sparse time series, we can assume that ground deformation is correlated within a specific period, and the corresponding time period is called the Correlated Time Period (CTP). Without loss of generality, we assume that deformation and velocity within the CTP are correlated. It is linear, and then the interval deformation rate in the sparse sequence can be regarded as an unknown. Equation (3) can be represented by the sparse design matrix in equation (4).
[0072]
[0073] Where S is an unknown vector containing the DEM error and interval deformation rate between sparse deformation sequences. G is the new design matrix. The rank of matrix G is R(G) = <-1 and |DL |≠0,R(D L Therefore, matrix G is full rank, and more importantly, it has redundant independent observations, which helps in the uncertainty analysis of the parameters to be estimated; Equation (5) gives the original unknown parameters when using SAR data with a six-day revisit period ( ) and new sparse unknown parameters ( Example comparison of design matrices corresponding to (CTP = 12 days).
[0074]
[0075]
[0076] Secondly, to enhance the stability of the constructed sparse design matrix, a time buffer is set around the predefined CTP to generate multiple sparse design matrices G, and the design matrix with the smallest condition number (CN) is selected to avoid ill-conditioned matrices. For example... Figure 2 As shown, an example of generating sparse design matrices by using different CTPs on the same interferogram network is illustrated. Initially, the CTP is predefined as 35 days to construct the design matrix, and then different design matrices are constructed using a + / - 5 day time buffer (i.e., from 30 to 40 days). After calculating their condition numbers, the design matrix with the smallest condition number coefficients is selected for estimation in equation (4), i.e., with a CTP of 37 days and a box in the diagram. Figure 2 Highlighted in the middle.
[0077] In this embodiment, by establishing deformation-related time periods and setting up a time buffer centered on the related time periods, multiple sparse observation matrices are established based on all related time periods within the time buffer. The matrix with the smallest condition number is selected as the final sparse matrix. The sparse parameter model is established based on the interferometric network diagram and the sparse observation matrix, so as to facilitate the subsequent construction of the double-difference model and the obtaining of the deformation time series of each InSAR observation point and the corresponding evaluation results.
[0078] like Figure 1 As shown, this embodiment of the invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model, comprising the following steps:
[0079] Step S200: Construct a double-difference observation model based on the arc segments of the spatial network in the sparse parameter model.
[0080] In this embodiment, the statistical model in equation (4) is usually assumed to be a Gaussian model. The Gaussian assumption of APS is sometimes invalid, and the expected value may be non-zero, i.e., E(Δ)≠0, so the accuracy assessment may be biased.
[0081] In this embodiment, the influence of APS is reduced by using the arc segments of the spatial network in the sparse parameter model as observation values.
[0082] Specifically, in one implementation of this embodiment, step S200 includes the following steps:
[0083] Step S210: Obtain the arc segments of the spatial network in the sparse parameter model, use the arc segments in the spatial network as observations, and determine the parameters to be estimated on the arc segments;
[0084] Step S220: Propagate the parameters to be estimated and their corresponding covariance from the time interval to the sparse sequence, and obtain the deformation and covariance of the points through the adjustment of the spatial network to construct a double-difference observation model.
[0085] In this implementation, to reduce the impact of APS, the differential phase between pixel pairs that are close to each other (i.e., arcs in the spatial network) is used as the observation, which is called network-based estimation; since the pixel pairs are very close (usually the distance is <1km), the impact of APS on the differential phase is not significant.
[0086] In this embodiment, the stable coefficient design matrix in equation (4) is combined with iterative weighted least squares (LS), and χ is used. 2 The reliability of the least squares estimation is verified by testing to estimate the parameters on the arc segment. After determining the parameters to be estimated on the arc segment, the parameters to be estimated and the corresponding covariance are propagated from the time interval to the sparse sequence, and the deformation and covariance of the points are obtained by spatial network adjustment, thereby constructing a double-difference observation model to overcome the influence of atmospheric effects.
[0087] like Figure 1 As shown, this embodiment of the invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model, comprising the following steps:
[0088] Step S300: Perform iterative least squares estimation based on the sparse parameter model and test it based on the variance. Remove atmospheric factors using the double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0089] In this embodiment, iterative least squares estimation is performed by combining a sparse parameter model and InSAR time series observations, and combined with χ²... 2 Variance testing was performed to maintain the reliability of the estimates; spatial network adjustment was carried out using a double-difference observation model to remove the influence of atmospheric effects; finally, the deformation time series of each InSAR observation point and the corresponding accuracy assessment were obtained.
[0090] Specifically, in one implementation of this embodiment, step S300 includes the following steps:
[0091] Step S310: Based on the sparse parameter model, perform sparse parameter estimation using the iterative least squares estimation method;
[0092] Step S320: Variance verification is performed on the parameters obtained by the iterative least squares estimation method to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0093] In this embodiment, after constructing the sparse parameter model, iterative weighted least squares (LS) is used to estimate the parameters for equation (4), as shown in equation (6).
[0094]
[0095] in This is the weight matrix; an identity matrix can be used initially. Estimated parameters. The complete uncertainty can be calculated based on the residual (i.e. r), as shown in equation (7).
[0096]
[0097]
[0098] In the above iterative estimation process, χ is used 2 The reliability of the least squares estimate is verified by a test, χ². 2 The test can be used to check whether the front variance and the back variance are statistically equal. This means at a certain confidence level (usually 95%).
[0099] In this embodiment, after iterative least squares estimation based on the sparse parameter model, variance testing is performed to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0100] In another implementation of this embodiment, step S320 further includes:
[0101] Step S321: The parameters that have been estimated are passed through a double-difference observation model to remove the influence of atmospheric factors, so as to obtain deformation time series of each InSAR observation point with higher accuracy and the corresponding accuracy assessment results.
[0102] In this embodiment, after iterative least squares estimation based on the sparse parameter model, the influence of atmospheric effects can be removed by using the double-difference observation model to obtain a more accurate deformation time series of each InSAR observation point and the corresponding accuracy assessment.
[0103] In this embodiment, tests were conducted using both simulated and real Sentinel-1A / B datasets, and the results are as follows: Figure 3 , Figure 4 , Figure 5 As shown: When using a real dataset, the derived deformation is verified by some GPS measurement results; compared with the GPS results, the overall standard deviation of the InSAR results is 5.4 mm, which is within the uncertainty range of the InSAR results; the test results show that the sparse parameter model provided by the present invention is an effective technical solution for evaluating the accuracy of InSAR time-series deformation results.
[0104] This embodiment achieves the following technical effects through the above technical solution:
[0105] This invention provides an MT-InSAR accuracy estimation method based on a sparse parameter model. A sparse parameter model is established using multi-temporal InSAR deformation interferograms to evaluate the accuracy of InSAR deformation time series. The sparse parameter model estimates the sparsed time series deformation results, and redundant InSAR time series data can be generated during the estimation process for accuracy assessment of the estimated results. A double-difference observation model is constructed using arc segments of the spatial network in the sparse parameter model to remove the influence of atmospheric factors. By constructing the sparse parameter model and the double-difference observation model, this invention effectively suppresses the influence of errors and noise during data processing, thus solving the problems of limited applicability and inaccurate analysis in existing InSAR measurement accuracy assessment methods.
[0106] Exemplary device
[0107] Based on the above embodiments, the present invention also provides an MT-InSAR accuracy estimation device based on a sparse parameter model, comprising:
[0108] The sparse parameter model module is used to construct a sparse observation matrix based on multi-temporal InSAR deformation interferograms, perform stability screening based on the sparse observation matrix, and establish a sparse parameter model.
[0109] The double-difference observation model module is used to construct a double-difference observation model based on the arc segments of the spatial network in the sparse parameter model.
[0110] The verification result module is used to perform iterative least squares estimation based on the sparse parameter model and verify it based on the variance. Atmospheric factors are removed through the double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
[0111] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 6 As shown.
[0112] The terminal includes: a processor, a memory, an interface, a display screen, and a communication module connected via a system bus; wherein, the processor of the terminal provides computing and control capabilities; the memory of the terminal includes a storage medium and internal memory; the storage medium stores the operating system and computer programs; the internal memory provides an environment for the operation of the operating system and computer programs in the storage medium; the interface is used to connect to external devices, such as mobile terminals and computers; the display screen is used to display relevant information; and the communication module is used to communicate with a cloud server or mobile terminal.
[0113] When executed by a processor, this computer program is used to implement an MT-InSAR accuracy estimation method based on a sparse parameter model.
[0114] It will be understood by those skilled in the art that Figure 6 The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0115] In one embodiment, a terminal is provided, comprising: a processor and a memory, the memory storing an MT-InSAR accuracy estimation program based on a sparse parameter model, the MT-InSAR accuracy estimation program based on a sparse parameter model being executed by the processor to implement the operation of the MT-InSAR accuracy estimation method based on a sparse parameter model as described above.
[0116] In one embodiment, a storage medium is provided, wherein the storage medium stores an MT-InSAR accuracy estimation program based on a sparse parameter model, which, when executed by a processor, is used to implement the operation of the MT-InSAR accuracy estimation method based on the sparse parameter model as described above.
[0117] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory.
[0118] In summary, this invention discloses an MT-InSAR accuracy estimation method and apparatus based on a sparse parameter model, comprising: constructing a sparse observation matrix based on multi-temporal InSAR deformation interferograms; performing stability screening based on the sparse observation matrix to establish a sparse parameter model; constructing a double-difference observation model based on the arc segments of the spatial network in the sparse parameter model; performing iterative least squares estimation based on the sparse parameter model and verifying it based on variance; removing atmospheric factors through the double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results; this invention solves the problems of limited applicability and inaccurate analysis of existing InSAR measurement accuracy evaluation methods by constructing a sparse parameter model and a double-difference observation model.
[0119] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.
Claims
1. A method for estimating the accuracy of MT-InSAR based on a sparse parameter model, characterized in that, include: A sparse observation matrix is constructed based on multi-temporal InSAR deformation interferometry, and a sparse parameter model is established by performing stability screening based on the sparse observation matrix. Based on the arc segments of the spatial network in the sparse parameter model, a double-difference observation model is constructed. Iterative least squares estimation is performed based on the sparse parameter model, and the variance is used for verification. Atmospheric factors are removed by a double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
2. The MT-InSAR accuracy estimation method based on a sparse parameter model according to claim 1, characterized in that, The construction of a sparse observation matrix based on the multi-temporal InSAR deformation interferogram includes: The time buffer in the multi-temporal InSAR deformation interferogram is determined, and multiple sparse observation matrices are established based on the time buffer.
3. The MT-InSAR accuracy estimation method based on a sparse parameter model according to claim 2, characterized in that, Determining the time buffer in the multi-temporal InSAR deformation interferogram includes: A time buffer is established in the multi-temporal InSAR deformation interferogram with a preset time period as the center.
4. The MT-InSAR accuracy estimation method based on a sparse parameter model according to claim 3, characterized in that, The step of performing stability screening based on the sparse observation matrix and establishing a sparse parameter model includes: Select the matrix that satisfies the conditions from multiple sparse observation matrices to obtain the final sparse matrix; The sparse parameter model is established based on the final sparse matrix, the preset time period, and the multi-temporal InSAR deformation interferogram.
5. The MT-InSAR accuracy estimation method based on a sparse parameter model according to claim 1, characterized in that, The construction of a double-difference observation model based on the arc segments of the spatial network in the sparse parameter model includes: Obtain the arc segments of the spatial network in the sparse parameter model, use the arc segments in the spatial network as observations, and determine the parameters to be estimated on the arc segments; The parameters to be estimated and their covariance are propagated from the time interval to the sparse sequence, and the deformation and covariance of the points are obtained through the adjustment of the spatial network to construct a double-difference observation model.
6. The MT-InSAR accuracy estimation method based on a sparse parameter model according to claim 1, characterized in that, The iterative least squares estimation based on the sparse parameter model, followed by variance testing, yields the deformation time series and corresponding accuracy assessment results for each InSAR observation point, including: Based on the sparse parameter model, the sparse parameter estimation method is used to estimate the sparse parameters. The variance of the parameters obtained by the iterative least squares estimation method is verified to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
7. The MT-InSAR accuracy estimation method based on a sparse parameter model according to claim 6, characterized in that, The process of performing variance verification on the parameters obtained through the iterative least squares estimation method further includes: The parameters that have been estimated are then passed through a double-difference observation model to remove the influence of atmospheric factors, in order to obtain more accurate deformation time series of each InSAR observation point and the corresponding accuracy assessment results.
8. An MT-InSAR accuracy estimation device based on a sparse parameter model, characterized in that, include: The sparse parameter model module is used to construct a sparse observation matrix based on multi-temporal InSAR deformation interferograms, perform stability screening based on the sparse observation matrix, and establish a sparse parameter model. The double-difference observation model module is used to construct a double-difference observation model based on the arc segments of the spatial network in the sparse parameter model. The detection results module is used to perform iterative least squares estimation based on the sparse parameter model, and to perform verification based on variance. Atmospheric factors are removed through the double-difference observation model to obtain the deformation time series of each InSAR observation point and the corresponding accuracy evaluation results.
9. A terminal, characterized in that, include: The processor and memory, wherein the memory stores an MT-InSAR accuracy estimation program based on a sparse parameter model, the MT-InSAR accuracy estimation program based on a sparse parameter model being executed by the processor to implement the operation of the MT-InSAR accuracy estimation method based on a sparse parameter model as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, The processor and memory, wherein the memory stores an MT-InSAR accuracy estimation program based on a sparse parameter model, the MT-InSAR accuracy estimation program based on a sparse parameter model being executed by the processor to implement the operation of the MT-InSAR accuracy estimation method based on a sparse parameter model as described in any one of claims 1-7.