Efficient wide-area surface deformation monitoring method and system

Abstract

The invention discloses an efficient wide-area surface deformation monitoring method and system, and the method comprises the steps: obtaining time sequence images based on a synthetic aperture radar sensor, carrying out the grouping and compression of an image set, and generating a local interference data set based on each group of images; spatial filtering is carried out based on phase components of the interference data, and a pixel local complex coherence matrix model is established based on the filtered data; in combination with the compressed image set, performing phase composition decomposition and connection based on a pixel local complex coherence matrix, and recovering a complete time sequence phase; and carrying out deformation distance inversion and geodetic coordinate system conversion on each pixel time sequence phase to obtain a wide-area earth surface deformation result. The method can achieve the efficient calculation of the wide-area surface deformation data of the synthetic aperture radar, effectively improves the quality of phase data, and improves the inversion density and monitoring precision of a monitoring point.

Description

Technical Field

[0001] The present invention belongs to the field of synthetic aperture radar deformation monitoring, and in particular relates to an efficient wide-area surface deformation monitoring method and system. Background Art

[0002] In recent years, deformation monitoring technology based on synthetic aperture radar sensors has demonstrated significant advantages in the field of engineering monitoring. In the long-term stability monitoring of large-scale infrastructure (such as bridges and dams) and geotechnical engineering projects (such as slopes, mining areas, and underground projects), traditional contact measurement methods are limited by single-point monitoring modes and environmental interference, making it difficult to achieve dynamic deformation sensing with millimeter-level accuracy over a wide range and in all weather conditions. However, synthetic aperture radar deformation monitoring technology can obtain millimeter-level deformation displacement information of the target area at a very low cost. Its spatially continuous coverage allows simultaneous capture of the overall deformation field of the project and local anomalies, effectively overcoming the spatial discreteness and timeliness limitations of traditional monitoring methods, thereby enabling refined analysis of complex engineering deformation characteristics. The phase information of the radar sensor is directly related to the detected deformation distance. However, given that most engineering monitoring scenarios are located in non-artificial surface areas, the phase signals of surrounding scatterers are prone to decoherence effects. Therefore, the recovery and processing of subsequent phase data are crucial.

[0003] Traditional phase data processing methods are generally based on the simultaneous processing of synthetic aperture radar (SAR) full-time series imagery. They construct a complex coherence matrix model of the full interferometric data by initially identifying homogeneous pixels in the target area. Phase calculation is then performed on the model using nonlinear iteration or eigenvalue decomposition strategies. For example, the PTA method performs nonlinear phase calculation by maximizing the logarithm of the probability density function of the complex coherence matrix, while the EMI method outputs phase information by transforming the nonlinear optimization problem of the maximum likelihood estimator into an eigenvalue decomposition problem. However, these traditional phase processing methods are prone to two drawbacks: First, the complex coherence matrix model established based on the identification of homogeneous pixels is susceptible to interference from heterogeneous pixels, and the calculated phase information may contain deformation phase ambiguity. This phenomenon can be more severe in complex engineering monitoring scenarios, resulting in a low spatial density and poor accuracy of the deformed monitoring points derived from the subsequent inversion. Second, the full-interferometric dimensional complex coherence matrix model constructed based on the full-time series imagery is prone to time-consuming matrix operations during phase calculation, resulting in low efficiency and a long cycle time for the entire engineering monitoring process. Therefore, how to improve data processing efficiency while ensuring the quality of phase and monitoring points, especially how to efficiently obtain high-density and high-precision deformation monitoring results in complex large-scale engineering monitoring scenarios, has become an important research task. Summary of the Invention

[0004] To solve the above technical problems, the application provides a high-efficiency wide-area ground surface deformation monitoring method and system, which can effectively alleviate the problem that the phase quality improvement and data processing efficiency improvement are difficult to be considered in the complex engineering monitoring environment of synthetic aperture radar, and effectively improve the density and monitoring accuracy of subsequent monitoring points.

[0005] The technical solution for achieving the object of the application is as follows:

[0006] The first aspect of the application is to provide a high-efficiency wide-area ground surface deformation monitoring method, comprising:

[0007] Based on the time-series images obtained by the synthetic aperture radar sensor, the image set is grouped and compressed, and the local interference data set is generated based on each group of images;

[0008] Based on the phase component of the interference data, spatial filtering is performed, and a pixel local complex coherence matrix model is established based on the filtered data;

[0009] Combined with the compressed image set, the phase component solution and connection are performed based on the pixel local complex coherence matrix to restore the complete time-series phase;

[0010] The deformation distance inversion and geodetic coordinate system conversion are performed on each pixel time-series phase to obtain the wide-area ground surface deformation result.

[0011] Further, based on the time-series images obtained by the synthetic aperture radar sensor, the image set is grouped and compressed, and the local interference data set is generated based on each group of images, comprising:

[0012] Extract N scenes of synthetic aperture radar time-series images of a certain area, perform time-series registration and cutting and other preprocessing operations on the images, and group and compress the preprocessed time-series images;

[0013] Through data component analysis on each group of local images, the image set is conjugate multiplied in a free combination method to generate a local interference data set.

[0014] Further, the images in a certain local group comprise:

[0015]

[0016] Wherein, M s is the s scene complex image, a s is the real part of the s scene image, b s is the imaginary part of the s scene image, i represents the imaginary unit, A s is the amplitude component of the s scene image, represents the phase component of the s scene image, which is related to the detection distance, and N1 represents the number of images in the local group;

[0017] Interferometric data generated based on images in a local group include:

[0018]

[0019] Among them, D s,t is the interference data generated based on the s-th and t-th scene images, φ s,t represents the interference phase, * represents the conjugate operation, M t is the t-th scene complex image, A t is the amplitude component of the t-th image.

[0020] Furthermore, performing spatial filtering based on the phase component of the interference data includes:

[0021] According to the principle of two-dimensional fast Fourier transform, the interference data generated by each group is converted to the frequency domain, the filtering intensity parameter is set, and the frequency domain filtering control result is output. The frequency domain filtering control result is multiplied by the power spectrum and then converted back to the spatial domain to realize frequency domain filtering. Based on the result after the frequency domain filtering, cumulative summation is performed in the vertical and horizontal directions in the spatial domain to complete the spatial filtering of the interference data.

[0022] Furthermore, establishing a pixel local complex coherence matrix model based on the filtered data includes:

[0023]

[0024] in, is the complex coherence matrix model of a pixel in the mth group, and its matrix dimension is N1×N1, represents the interferometric phase extracted based on the filtered interferometric data set, is the coherence coefficient based on the filtered interference phase output.

[0025] Furthermore, restoring the complete time series phase of the pixels includes:

[0026] For a certain pixel, performing eigenvalue decomposition and solving on the local complex coherence matrix model of the pixels of each group to obtain the local solved phase of the pixel;

[0027] Combine the above-mentioned groups of compressed image data to establish a reference complex coherence matrix model again, perform phase solution on the reference model, obtain reference phase components, and assign the reference components to the local solution phase, thereby completing the time series phase connection, traversing all pixels, and realizing phase solution of the entire area.

[0028] Furthermore, performing phase calculation on the local complex coherence matrix of a pixel includes:

[0029]

[0030] in, is the local solution phase vector, is the coherence estimation matrix, |·| is the modulo operation, x and y represent the image positions, N1 represents the number of images in the local group, v is the eigenvector corresponding to the minimum eigenvalue, (·) T is the transpose operation.

[0031] Furthermore, the time series phase connection of pixels includes:

[0032]

[0033] in, is the consistent time series phase of a certain pixel, is the local solution phase vector of a pixel in the mth group, is the phase component of a pixel in the mth group for reference connection.

[0034] Furthermore, the surface deformation distance of the phase inversion to the geodetic coordinate system includes:

[0035]

[0036] Wherein, Δd is the relative deformation of the ground surface during the monitoring period, is the deformed phase after removing other interfering phase components, and λ is the sensor wavelength.

[0037] A second aspect of the present invention is to provide an efficient wide-area surface deformation monitoring system for implementing the above method, comprising:

[0038] An image acquisition and preprocessing module is configured to acquire time-series images using a synthetic aperture radar sensor and to group and compress the time-series image sets;

[0039] a local interference processing module, connected to the image acquisition and preprocessing module, configured to generate a local interference data set based on each group of images after group processing;

[0040] a spatial filtering and coherence modeling module, connected to the local interference processing module, configured to perform spatial filtering on the phase components in the local interference data set and establish a pixel local complex coherence matrix model based on the filtered data;

[0041] a phase solution and connection module, connected to the spatial filtering and coherence modeling module and the image acquisition and preprocessing module, configured to combine the compressed image set and the pixel-local complex coherence matrix model to perform phase component solution and connection to recover the complete time series phase of each pixel;

[0042] A deformation inversion and output module is connected to the phase solution and connection module, configured to perform deformation distance inversion and geodetic coordinate system conversion on the time-series phase of each pixel, and obtain and output wide-area ground surface deformation results.

[0043] Advantages and beneficial effects of the present application:

[0044] The present application discloses a kind of high-efficiency wide-area ground surface deformation monitoring method and system, by grouping processing to synthetic aperture radar radar full time series image, each local subgroup is independently generated interference data set, reduce the time cost of subsequent data processing, and each group image is compressed storage;Based on the spatial filtering of the phase component of each group interference data, the local complex coherent matrix model of each pixel is established by extracting the filtered phase data.It is worth mentioning that the traditional pixel complex coherent matrix is mainly constructed based on homogeneous pixel set, and considering that existing homogeneous pixel recognition algorithm mostly depends on radar amplitude data for pixel selection, the consideration of phase information is ignored, and heterogeneous pixel interference is easy to produce, the complex coherent matrix model construction method described in the present application can effectively reduce the dependence on homogeneous pixels during model construction, effectively alleviate the negative effects such as deformation phase stripe blur phenomenon that output phase is prone to, improve the inversion density of subsequent deformation monitoring points.In addition, the local complex coherent matrix model is separately solved, which can maximize the influence of time dephasing error, and the complete time series phase is recovered by combining the compressed image set after outputting local phase;Finally, the deformation distance inversion is completed based on the time-series phase of each pixel and converted to geodetic coordinate system, and the final monitoring product is output.The method disclosed in the present application can effectively realize high-coverage density and high-precision deformation monitoring of target area while maintaining high data processing efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0045] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this application. The embodiments of this application and of the related description are used to explain the application and are not intended to limit the application. In the drawings:

[0046] Figure 1 The flowchart of the high-efficiency wide-area ground surface deformation monitoring method and system of the embodiment of the present application is shown in the figure.

[0047] Figure 2 The interference phase solution result of the real ground surface deformation data 60-day time interval used in the embodiment of the present application is shown in the figure, wherein (a) is the interference phase solution result of PTA, (b) is the interference phase solution result of EMI, and (c) is the interference phase solution result of the method of the present application.

[0048] Figure 3Schematic diagram of the posterior mean coherence distribution generated for the real surface deformation data used in the embodiments of the present invention based on different phase solution methods, where (a) is the posterior mean coherence result of PTA, (b) is the posterior mean coherence result of EMI, and (c) is the posterior mean coherence result of the method of the present invention.

[0049] Figure 4 Schematic diagram of the long-term surface deformation monitoring results of the mining area according to the embodiment of the invention, wherein (a) is the surface deformation monitoring result of the StaMPS technology, and (b) is the surface deformation monitoring result of the method of the present invention. DETAILED DESCRIPTION

[0050] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0051] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0052] like Figure 1 As shown, this embodiment provides an efficient wide-area surface deformation monitoring method and system, including:

[0053] Acquire time-series images based on synthetic aperture radar sensors, group and compress the image sets, and generate local interferometric datasets based on each group of images;

[0054] Perform spatial filtering based on the phase component of the interference data, and establish a pixel local complex coherence matrix model based on the filtered data;

[0055] Combined with the compressed image set, phase component solution and connection are performed based on the pixel local complex coherence matrix to restore the complete time series phase;

[0056] The deformation distance inversion and geodetic coordinate system transformation are performed on the phase of each pixel time series to obtain the wide-area surface deformation results.

[0057] Furthermore, based on the synthetic aperture radar sensor to obtain time-series images, the image sets are grouped and compressed, and the local interferometric data sets are generated based on each group of images, including:

[0058] Extracting N synthetic aperture radar time series images of a certain area, performing preprocessing operations such as time series registration and cropping on the images, and grouping and compressing the preprocessed time series images;

[0059] By performing data component analysis on each group of local images, the image sets are paired and conjugate multiplied in pairs using a free combination method to generate a local interference data set;

[0060] Furthermore, the images in a local group include:

[0061]

[0062] Among them, M s is the complex image of the sth scene, a s is the real part of the s-th scene image, b s is the imaginary part of the s-th scene image, i represents the imaginary unit, A s is the amplitude component of the s-th scene image, represents the phase component of the s-th image, which is related to the detection distance, and N1 represents the number of images in the local group;

[0063] Interferometric data generated based on images in a local group include:

[0064]

[0065] Among them, D s,t is the interference data generated based on the s-th and t-th scene images, φ s,t represents the interference phase, * represents the conjugate operation, M t is the t-th scene complex image, A t is the amplitude component of the t-th image.

[0066] Furthermore, the compressed image generated based on a certain local group includes:

[0067]

[0068] in, The compressed image generated for this group, M n represents the nth original complex image of the group, η n is the compressed variation basis component.

[0069] Furthermore, spatial filtering based on the phase component of the interferometric data includes:

[0070] According to the principle of two-dimensional fast Fourier transform, the interference data generated by each group is converted to the frequency domain, the filtering intensity parameter is set, and the frequency domain filtering control result is output. The frequency domain filtering control result is multiplied by the power spectrum and then converted back to the spatial domain to realize frequency domain filtering. Based on the result after the frequency domain filtering, cumulative summation is performed in the vertical and horizontal directions in the spatial domain to complete the spatial filtering of the interference data.

[0071] Specifically, a space-frequency filtering method was introduced to filter noise and smooth out pseudo-fringe errors in the interferometric data generated by each group. By converting the spatial data to the frequency domain and setting the filter spectrum intensity control parameters, phase noise was effectively filtered out. The frequency-domain filtered data was then converted back to the spatial domain and smoothed to effectively reduce the impact of pseudo-fringe errors.

[0072] The conversion of the original interference data in space to the frequency domain can be expressed as:

[0073]

[0074] Where D(x,y) is the local window interferometric phase data centered at the coordinate (x,y), FFT2{·} represents the two-dimensional fast Fourier transform operation, is the frequency spectrum after conversion, θ is the spatial frequency. The frequency domain filtering control result can be expressed as:

[0075]

[0076] in, is the frequency domain filtering control result, is a two-dimensional Gaussian kernel, α represents the filter strength parameter, Represents the convolution operation. By multiplying the above formula with the power spectrum, the final frequency domain filtering result is:

[0077]

[0078] in, is the interference data frequency wave filtering result, FFT2 -1 {·} is a two-dimensional inverse fast Fourier transform. Based on the filtering result in the frequency domain, with the pixel at coordinate (x, y) as the center, by setting a certain sliding window, the window sum in the vertical direction of the spatial data is calculated as:

[0079]

[0080] in, is the cumulative sum of the current pixel (x, y) along the vertical direction of the window. On this basis, the window sum in the horizontal direction of the current spatial data is calculated as:

[0081]

[0082] in, This is the final spatial filtering result of the interference data.

[0083] Furthermore, establishing a pixel local complex coherence matrix model based on the filtered data includes:

[0084] The spatial filtered interference data is subjected to phase component extraction, and a coherence coefficient corresponding to a pixel is obtained based on the filtered phase; a phase matrix and a coherence coefficient matrix in a local spatial form (image length x image width) are converted into a pixel existing form (N1xN1); and the pixel existing form coherence matrix and the phase matrix are subjected to point multiplication to complete complex coherence matrix model construction. The pixel local complex coherence matrix model can be expressed as:

[0085]

[0086] wherein, is a complex coherence matrix model of a certain pixel in the mth group, and the matrix dimension is N1xN1, is a pixel phase matrix, is a pixel coherence coefficient matrix, represents an interference phase extracted based on the filtered interference data set, is a coherence coefficient output based on the filtered interference phase, represents point multiplication operation.

[0087] Further, combined with the compressed image set, phase component solving and connection are performed based on the pixel local complex coherence matrix to restore the complete time sequence phase, including:

[0088] For a certain determined pixel, the pixel local complex coherence matrix models of the groups are subjected to eigenvalue decomposition solving to obtain the local solving phase of the pixel.

[0089] The reference complex coherence matrix model is re-established combined with the compressed image data of the groups, the phase of the reference model is solved to obtain a reference phase component, the reference component is distributed to the local solving phase, and then the time sequence phase connection is completed, all pixels are traversed, and the phase solving of the entire region is realized.

[0090] Further, the phase solving of the local complex coherence matrix of a certain pixel includes:

[0091]

[0092] wherein, is a local solving phase vector, is a coherence estimation matrix, |·| is a modulus operation, x and y represent image positions, N1 represents the number of images in a local group, v is a characteristic vector corresponding to a minimum eigenvalue, and (·) T is a transposition operation. The solving of the local complex coherence matrix with a lower dimension greatly reduces the time incoherence error and the time cost of matrix calculation.

[0093] Further, the time sequence phase connection of the pixel includes:

[0094]

[0095] in, is the local solution phase vector of a pixel in the mth group, is the phase component of a pixel in the mth group for reference connection, is the consistent time series phase of a certain pixel.

[0096] Furthermore, deformation distance inversion and geodetic coordinate system transformation are performed on the phase of each pixel time series to obtain wide-area surface deformation results including:

[0097]

[0098] Wherein, Δd is the relative deformation of the ground surface during the monitoring period, is the deformed phase after removing other interfering phase components, and λ is the sensor wavelength.

[0099] In order to verify the technical effect of the present invention, the real surface deformation interference data of the mining area was selected and phase solution processing and analysis were carried out based on PTA, EMI and the method of the present invention under the same conditions. The long-term mining area surface deformation monitoring was carried out based on StaMPS technology and the method of the present invention. Among them, the interference results of the real surface deformation data with a time interval of 60 days after phase solution by PTA, EMI and the method of the present invention are shown in the figure below. Figure 2 As shown; Figure 2 (a) is the original interference phase result, Figure 2 (b) is the interference phase result after PTA phase solution, Figure 2 (c) is the interference phase result after EMI phase solution, Figure 2 (d) is the interference phase result after phase solution by the method of the present invention; Figure 2 From the interference results of the target area shown, it can be found that the original phase signal detected by the radar sensor is seriously affected by decoherence factors such as phase noise due to the extensive seasonal vegetation in the non-artificial surface area. After phase solution by PTA and EMI methods, although the phase noise in the target area is suppressed to a certain extent, the deformation center of the mining area still shows a relatively obvious deformation phase fringe aliasing and blurring phenomenon. The deformation phase blurring degree of EMI is slightly better than that of PTA. In comparison, the method of the present invention can not only maintain a more stable and excellent noise filtering ability, but also the degree of phase fringe signal recovery in the deformation area is far superior to that of PTA and EMI methods. The overall phase continuity is stronger and the deformation phase fringe is smoother. In addition, through the analysis of phase performance evaluation indicators, the number of residual points of PTA is 2.32×10 4 , the sum of phase differences (SPD) is 4.69×105 ; For EMI, the number of residual points of the interference phase is 2.05×10 4 , SPD is 4.4×10 5 ; The number of residual points of the method of the present invention is only 0.63×10 4 , SPD is 1.78×10 5 It can be found that the method of the present invention is in a leading position in terms of both the ability to suppress phase noise and the smoothness of deformation phase stripes. At the same time, in terms of time series phase solution efficiency, PTA takes the longest time, about 3 hours, and EMI takes about 0.23 hours. The phase solution of the method of the present invention takes only 0.19 hours. The method of the present invention can output high-quality phase while having a faster processing efficiency. The posterior mean coherence results for screening deformation monitoring points based on the solution phase are as follows: Figure 3 As shown; Figure 3 (a) is the posterior mean coherence result of PTA, Figure 3 (b) is the posterior mean coherence result of EMI, Figure 3 (c) is the posterior average coherence result of the method of the present invention; the posterior average coherence index directly determines the selection of subsequent deformation monitoring points. The higher the coherence (the closer to 1), the higher the density of the corresponding extracted monitoring points. Figure 3 Observation and statistics show that the posterior quality of PTA in the deformation area is relatively the worst, followed by EMI. The coherence quality of the method of the present invention in the two deformation areas is far superior to the above two methods. Specifically, for deformation area A, when the posterior average coherence threshold is 0.75, the corresponding pixel proportion of PTA is 32.37%, the corresponding pixel proportion of EMI is 34.21%, and the corresponding pixel proportion of the method of the present invention is 69.80%. For deformation area B, when the posterior average coherence threshold is 0.75, the corresponding pixel proportion of PTA is 29.78%, the corresponding pixel proportion of EMI is 31.41%, and the corresponding pixel proportion of the method of the present invention is 79.22%. The statistical results show that the method of the present invention can effectively improve the spatial coverage density of surface deformation monitoring points, especially in the deformation center area, the density of monitoring points is more significantly improved. Figure 4 It is a deformation product obtained by long-term monitoring of the mining area surface based on different monitoring technologies. Figure 4 (a) is the surface deformation monitoring result obtained by StaMPS technology, Figure 4 (b) is the surface deformation monitoring result obtained by the method of the present invention; Figure 4 The analysis shows that the total number of monitoring points extracted by StaMPS during the monitoring period of the target area is about 115,086, and the spatial density of monitoring points is 948 points / km 2 The total number of monitoring points that can be extracted by the method of the present invention reaches 654,772, and the spatial density of monitoring points is 5,400 points / km 2More effective monitoring points can correspond to obtaining more surface deformation monitoring details. During the monitoring period, there are three significant deformation areas in the target area. The StaMPS technology is relatively difficult to restore the deformation areas, while the method of the present invention basically detects a relatively complete surface deformation field in the target area.

[0100] In summary, the above aspects show that the method of the present invention can more efficiently improve the phase output quality of synthetic aperture radar used in large-scale surface monitoring projects, enhance the phase to deformation interpretation accuracy, and ultimately increase the monitoring point density in deformation monitoring, thereby obtaining richer deformation detail information.

[0101] The second aspect of the present invention is to provide an efficient wide-area surface deformation monitoring system for the above method, comprising:

[0102] An image acquisition and preprocessing module is configured to acquire time-series images using a synthetic aperture radar sensor and to group and compress the time-series image sets;

[0103] a local interference processing module, connected to the image acquisition and preprocessing module, configured to generate a local interference data set based on each group of images after group processing;

[0104] a spatial filtering and coherence modeling module, connected to the local interference processing module, configured to perform spatial filtering on the phase components in the local interference data set and establish a pixel local complex coherence matrix model based on the filtered data;

[0105] a phase solution and connection module, connected to the spatial filtering and coherence modeling module and the image acquisition and preprocessing module, configured to combine the compressed image set and the pixel-local complex coherence matrix model to perform phase component solution and connection to recover the complete time series phase of each pixel;

[0106] The deformation inversion and output module is connected to the phase solution and connection module and is configured to perform deformation distance inversion and geodetic coordinate system conversion for the temporal phase of each pixel to obtain and output wide-area surface deformation results.

[0107] Compared with commonly used phase calculation methods such as PTA and EMI, the method of the present invention has higher phase output efficiency and phase quality. Compared with traditional StaMPS monitoring technology, the deformation information monitored by the method of the present invention has higher coverage and richness.

[0108] Through real surface deformation experiments, the results show that the efficient wide-area surface deformation monitoring method and system proposed in the present invention can effectively and efficiently improve the problems of synthetic aperture radar deformation monitoring phase data quality, monitoring point density and low monitoring accuracy.

[0109] The above merely provides the preferred embodiments of the present application, and the protection scope of the present application is not limited thereto, and any changes or substitutions within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An efficient wide-area surface deformation monitoring method, characterized in that: include: Acquire time-series images based on synthetic aperture radar sensors, group and compress the image sets, and generate local interferometric datasets based on each group of images; Perform spatial filtering based on the phase component of the interference data, and establish a pixel local complex coherence matrix model based on the filtered data; Combined with the compressed image set, phase component solution and connection are performed based on the pixel local complex coherence matrix to restore the complete time series phase; The deformation distance inversion and geodetic coordinate system transformation are performed on the phase of each pixel time series to obtain the wide-area surface deformation results.

2. The efficient wide-area surface deformation monitoring method according to claim 1, characterized in that: Acquire time-series images based on synthetic aperture radar sensors, group and compress the image sets, and generate local interferometric datasets based on each group of images, including: Extracting N synthetic aperture radar time series images of a certain area, performing time series registration and cropping preprocessing operations on the images, and grouping and compressing the preprocessed time series images; By performing data component analysis on each group of local images, the image sets are paired and conjugate multiplied in pairs using a free combination method to generate a local interference data set.

3. The efficient wide-area surface deformation monitoring method according to claim 2, characterized in that: Images in a local group include: Among them, M s is the complex image of the sth scene, a s is the real part of the s-th scene image, b s is the imaginary part of the s-th scene image, i represents the imaginary unit, A s is the amplitude component of the s-th image, represents the phase component of the s-th image, which is related to the detection distance, and N1 represents the number of images in the local group; Interferometric data generated based on images in a local group include: Among them, D s,t is the interference data generated based on the s-th and t-th scene images, φ s,t represents the interference phase, * represents the conjugate operation, M t is the t-th scene complex image, A t is the amplitude component of the t-th image.

4. The efficient wide-area surface deformation monitoring method according to claim 1, characterized in that: Spatial filtering based on the phase component of interferometric data includes: According to the principle of two-dimensional fast Fourier transform, the interference data generated by each group is converted to the frequency domain, the filtering intensity parameter is set, and the frequency domain filtering control result is output. The frequency domain filtering control result is multiplied by the power spectrum and then converted back to the spatial domain to realize frequency domain filtering. Based on the result after the frequency domain filtering, cumulative summation is performed in the vertical and horizontal directions in the spatial domain to complete the spatial filtering of the interference data.

5. The efficient wide-area surface deformation monitoring method according to claim 1 is characterized in that: The pixel local complex coherence matrix model is established based on the filtered data, including: in, is the complex coherence matrix model of a pixel in the mth group, and its matrix dimension is N1×N1, represents the interferometric phase extracted based on the filtered interferometric data set, is the coherence coefficient based on the filtered interference phase output.

6. The efficient wide-area surface deformation monitoring method according to claim 1, characterized in that: Restoring the pixel-complete time series phase includes: For a certain pixel, performing eigenvalue decomposition and solving on the local complex coherence matrix model of the pixels of each group to obtain the local solved phase of the pixel; Combine the above-mentioned groups of compressed image data to establish a reference complex coherence matrix model again, perform phase solution on the reference model, obtain reference phase components, and assign the reference components to the local solution phase, thereby completing the time series phase connection, traversing all pixels, and realizing phase solution of the entire area.

7. The efficient wide-area surface deformation monitoring method according to claim 6, characterized in that: The phase solution of the local complex coherence matrix of a pixel includes: in, is the local solution phase vector, is the coherence estimation matrix, |·| is the modulo operation, x and y represent the image positions, N1 represents the number of images in the local group, v is the eigenvector corresponding to the minimum eigenvalue, (·) T is the transpose operation.

8. The efficient wide-area surface deformation monitoring method according to claim 6, characterized in that: The time series phase connection of pixels includes: in, is the consistent time series phase of a certain pixel, is the local solution phase vector of a pixel in the mth group, is the phase component of a pixel in the mth group for reference connection.

9. The efficient wide-area surface deformation monitoring method according to claim 1, characterized in that: The surface deformation distance from phase inversion to the geodetic coordinate system includes: Wherein, Δd is the relative deformation of the ground surface during the monitoring period, is the deformed phase after removing other interfering phase components, and λ is the sensor wavelength.

10. An efficient wide-area surface deformation monitoring system for implementing the method according to any one of claims, characterized in that: include: An image acquisition and preprocessing module is configured to acquire time-series images using a synthetic aperture radar sensor and to group and compress the time-series image sets; a local interference processing module, connected to the image acquisition and preprocessing module, configured to generate a local interference data set based on each group of images after group processing; a spatial filtering and coherence modeling module, connected to the local interference processing module, configured to perform spatial filtering on the phase components in the local interference data set and establish a pixel local complex coherence matrix model based on the filtered data; a phase solution and connection module, connected to the spatial filtering and coherence modeling module and the image acquisition and preprocessing module, configured to combine the compressed image set and the pixel-local complex coherence matrix model to perform phase component solution and connection to recover the complete time series phase of each pixel; The deformation inversion and output module is connected to the phase solution and connection module and is configured to perform deformation distance inversion and geodetic coordinate system conversion for the temporal phase of each pixel to obtain and output wide-area surface deformation results.

Images