Large-range earth surface deformation monitoring method based on PS-InSAR technology

Abstract

The invention discloses a large-range earth surface deformation monitoring method based on a PS-InSAR technology. The method comprises the following steps: step 1, collecting SAR images of a multi-scene time sequence of a target area; step 2, processing the SAR image; step 3, PS points of a first layer of network in the processed SAR image are selected; step 4, performing grid sparsification on PS points of the first-layer network, and selecting PS control points; step 5, constructing a network by PS points of the first-layer network; step 6, resolving PS point parameters of the first-layer network; step 7, selecting PS points of a second-layer network in the processed SAR image; step 8, constructing a second-layer network PS point network; step 9, resolving PS point parameters of the second-layer network; step 10, performing PS link on the double-layer network; and step 11, analyzing the PS point time sequence of the double-layer network. According to the invention, large-range, high-precision, high-density and high-efficiency surface deformation monitoring can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a large-scale surface deformation monitoring method based on PS-InSAR technology. Background Art

[0002] Interferometric Synthetic Aperture Radar (InSAR) is an advanced satellite geodetic and remote sensing technology, which has gradually become one of the standard technical means for large-scale surface deformation monitoring. Time-series InSAR technology is a new InSAR technology that has recently developed and matured. Through time-series processing and error analysis of multi-scene SAR images, time-series InSAR technology can suppress the influence of atmospheric artifacts, terrain errors and decorrelation noise, and has the ability to stably obtain millimeter-level deformation of the surface. Among them, PS-InSAR technology, as a classic method of time-series InSAR technology, can achieve refined dynamic deformation measurement of ground targets by using full-resolution SAR images, and has a resolution advantage that other InSAR technologies cannot match.

[0003] However, limited by algorithm design and hardware configuration, traditional PS-InSAR technology has encountered key technical bottlenecks in terms of accuracy and efficiency, which are mainly manifested in the following aspects: (1) With the continuous improvement of parameters such as the number, resolution and spatial coverage of SAR images, the number of targets that can be monitored has increased exponentially. Traditional PS-InSAR algorithms are difficult to adapt to this development situation, resulting in a decrease in accuracy in large-scale deformation monitoring; (2) With the continuous deepening of people's understanding of surface deformation, the demand for deformation monitoring tasks has developed from local monitoring to municipal, provincial and even national levels. How to effectively reduce the deformation solution time and memory requirements and realize large-scale and efficient monitoring of PS-InSAR technology has become an urgent problem to be solved. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a large-scale surface deformation monitoring method based on PS-InSAR technology.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a large-scale surface deformation monitoring method based on PS-InSAR technology, comprising the following steps:

[0006] Step 1: Collect multi-view time series SAR images of the target area;

[0007] Step 2: Process the SAR images, perform image registration, geocoding and differential interferometry on the SAR images;

[0008] Step 3, select the PS points of the first layer network in the processed SAR image;

[0009] Step 4: Grid thinning the PS points of the first layer network and selecting PS control points;

[0010] Step 5: Constructing the PS points of the first layer network, connecting the selected PS control points in the first layer network using an irregular triangular network to form a PS spatial network;

[0011] Step 6: Calculate the parameters of the PS points in the first layer network, and calculate the parameters of the PS points in the PS space network;

[0012] Step 7, select the PS points of the second layer network in the processed SAR image;

[0013] Step 8: The second layer network PS points are networked, and the selected second layer network PS points are connected with the PS control points of the first layer network to form multiple local star networks;

[0014] Step 9: Calculate the parameters of the PS points of the second-layer network. Establish a functional relationship between the differential interference phase and the deformation rate, terrain error and other parameters on each arc segment of the star network. Estimate the arc segment parameters of the star network from the arc segment observation values ​​through the arc segment parameter estimator. Based on the arc segment parameters, take the absolute parameters of the PS points estimated by the first-layer network as a reference, directly integrate the parameters of the PS points of the second-layer network, and obtain the absolute parameters of the PS points in the second-layer network.

[0015] Step 10: Double-layer network PS link, combining the first-layer network PS point parameter estimation results with the second-layer network PS point parameter estimation results to obtain the deformation rate results and terrain error results of the entire target area;

[0016] Step 11, double-layer network PS point time series analysis, the deformation rate phase and terrain error phase are removed to obtain the secondary differential phase, the secondary differential phase is composed of nonlinear deformation, atmospheric error and random noise; according to the different characteristics of different components of the secondary differential phase in the time domain and space domain, first use time high-pass filtering to filter the secondary differential phase to separate nonlinear deformation, atmospheric phase and random error; in order to separate the atmospheric phase and noise, take spatial low-pass filtering of the high-frequency phase; after two filterings in the time and space domains, nonlinear deformation, atmospheric phase and random noise signals are obtained; combined with linear deformation and nonlinear deformation, the deformation time series of the double-layer network PS point is restored.

[0017] As a further description of the above technical solution:

[0018] The SAR images of the multi-view time series of the target area collected in step 1 include but are not limited to RadarSat-2, ALOS-2, Sentinel-1A, TerraSAR-X / TanDEM-X, COSMO_SkyMed, ICEYE, Lutan-1, Gaofen-3, Macro Figure 1 One or more of No. 1, Luojia No. 2 and Fucheng No. 1.

[0019] As a further description of the above technical solution:

[0020] Processing the SAR image in step 2 includes the following steps:

[0021] According to the multi-view time series SAR images of the target area, the time baseline, spatial baseline and Doppler centroid frequency difference between the multi-view SAR images are compared first, and the SAR image corresponding to the maximum correlation coefficient is selected as the common main image according to the established comprehensive correlation function;

[0022] The remaining SAR images are used as slave images and intensity cross-correlation registration is performed with the master image respectively;

[0023] Image pairs were constructed in single master image mode, the temporal baseline and spatial baseline between the master and slave images were calculated respectively, and conventional interference processing was performed on all image pairs;

[0024] Geocode the external DEM in the geographic coordinate system and convert it to the SAR image coordinate system;

[0025] The interferogram and DEM in the SAR coordinate system are differentially processed to obtain a series of differential interferograms.

[0026] As a further description of the above technical solution:

[0027] The steps for selecting the PS points of the first layer network in the processed SAR image in step 3 are as follows:

[0028] In the first layer of the network, the amplitude deviation index threshold method is used to detect high coherent scatterers and obtain the PS point target of the first layer of the network; the amplitude deviation can be expressed as follows:

[0029]

[0030] Among them, m A and σ A represent the temporal mean value and standard deviation of the pixel amplitude, respectively;

[0031] The first layer network PS point grid sparseness in step 4 includes:

[0032] PS points selected by amplitude deviation index threshold are used to select the highest quality and evenly distributed PS control points;

[0033] By presetting the grid size, only the PS point with the relatively smallest amplitude deviation index in each grid is selected as the PS control point in the grid;

[0034] The number of PS points in the first layer of the network is reduced by sparse PS point grid.

[0035] As a further description of the above technical solution:

[0036] Each arc segment of the PS space network represents an arc segment observation value; the PS space network is constructed by using Delaunay triangulation, local Delaunay triangulation and free connection network.

[0037] As a further description of the above technical solution:

[0038] In the step 6, the first layer network PS point parameter solution comprises two steps of arc parameter estimation and network adjustment;

[0039] The arc segment parameter estimation is calculated using the following formula:

[0040]

[0041] in is the winding operator; Δv(x, y) and Δz(x, y) are the difference in deformation rate and terrain error between PS points x and y, respectively; t j represents the time baseline of the image pair; λ is the wavelength; B ⊥ is the vertical baseline; R and θ are the radar slant range and incident angle respectively; represents the residual phase difference, and

[0042]

[0043] in, represents the overall coherence coefficient, M is the number of interference pairs, e j It is a plural form;

[0044] The network adjustment is calculated using the following formula:

[0045] B*X=L+R

[0046] X=(B T PB) -1 B T PL

[0047] Among them, B is the coefficient matrix, L and R are the arc segment observations and residuals respectively, P is the weight value set according to the complex overall coherence coefficient on the arc segment, and X is the absolute parameter of all PS points.

[0048] As a further description of the above technical solution:

[0049] The steps for selecting the PS points of the second layer network in the processed SAR image in step 7 are as follows:

[0050] In the second layer network, the remaining PS points are selected, wherein the remaining PS points include the PS points deleted after the PS point grid of the first layer network is sparsely processed and the remaining PS points detected using the average coherence threshold;

[0051] The coherence threshold is used to select the coherent point targets with medium quality on the image;

[0052] Within a fixed window, the coherence of two master-slave complex signals S1 and S2 is defined as:

[0053]

[0054] Where E(x) represents the expected value.

[0055] As a further description of the above technical solution:

[0056] The second layer network PS point networking step in step 8 includes:

[0057] In the second layer network, the PS control points of the first layer network are used as reference to perform the fragmentary measurement of the remaining PS points;

[0058] By connecting each remaining PS point of the second layer network to the nearest PS control point in the first layer network, a plurality of local star networks are formed, and each arc segment of the star network represents an arc segment observation value of the second layer network.

[0059] The present invention has the following beneficial effects:

[0060] 1. Compared with the prior art, the large-scale surface deformation monitoring method based on PS-InSAR technology proposed in the present invention can realize large-scale, high-precision, high-density and high-efficiency surface deformation monitoring. The main innovation is that the PS control points with the highest quality and uniform distribution are selected through the first-layer network PS point grid sparseness step, and the optimal processing and solution of these PS control points improve the calculation efficiency and parameter inversion accuracy. Secondly, the present invention adopts the classic bisection search method when estimating arc segment parameters, which further improves the efficiency of arc segment parameter interval search.

[0061] 2. Compared with the prior art, the present invention adopts a double-layer network PS link method to perform control measurement and fragment measurement of PS points, effectively increasing the number and density of InSAR monitoring points. The double-layer network method establishes different levels of spatial networks for InSAR monitoring points of different qualities, and error propagation can be suppressed through step-by-step control. In addition, the hierarchical solution also greatly reduces the computational burden. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 A flow chart of the technical solution proposed by the present invention;

[0063] Figure 2 This is a schematic diagram of PS point grid thinning in the present invention;

[0064] Figure 3 It is a schematic diagram of the double-layer network PS link of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] Reference Figure 1-3 The present invention provides a large-scale surface deformation monitoring method based on PS-InSAR technology, comprising the following steps:

[0067] Step 1: Collect multi-view time series SAR images of the target area; The SAR images collected in the target area include but are not limited to RadarSat-2, ALOS-2, Sentinel-1A, TerraSAR-X / TanDEM-X, COSMO_SkyMed, ICEYE, Lutan-1, Gaofen-3, Macro Figure 1 One or more of RadarSat-2, ALOS-2, TerraSAR-X / TanDEM-X, COSMO_SkyMed, ICEYE, Lutan-1, Gaofen-3, Macro Figure 1The No. 1, Luojia-2 and Fucheng-1 are high-resolution commercial satellites that can provide SAR images at the 5-meter, 3-meter and sub-meter levels. In order to ensure the measurement accuracy of PS-InSAR technology, the number of SAR images should generally be greater than 20, which is determined according to user needs.

[0068] Step 2, process the SAR images, and perform image registration, geocoding and differential interferometry on the SAR images; the processing of SAR images specifically includes the following steps: based on the collected multi-scene time series SAR images of the target area, first compare the time baseline, spatial baseline and Doppler centroid frequency difference between the multi-scene SAR images, and select the SAR image corresponding to the maximum correlation coefficient as the common master image according to the established comprehensive correlation function; use the remaining SAR images as slave images and perform intensity cross-correlation registration with the master image respectively; construct image pairs in single master image mode, calculate the time baseline and spatial baseline between the master and slave images respectively, and perform conventional interferometry processing on all image pairs; geocode the external DEM in the geographic coordinate system and convert it to the SAR image coordinate system; perform differential processing on the interferogram in the SAR coordinate system and the DEM to obtain a series of differential interferograms.

[0069] Step 3, select the PS points of the first layer network in the processed SAR image, the specific steps are as follows: Since the existing PS-InSAR technology directly processes a series of single main image interference pairs, without considering the limitation of the baseline, the interference pattern is seriously affected by the spatial decorrelation factor, and the criterion based on spatial coherence estimation cannot select PS points with high phase quality. Compared with the coherence coefficient threshold, the amplitude deviation index does not average the data in an estimation window, but analyzes the amplitude sequence to find the point target that is stable in time, and can obtain the local features with the highest spatial resolution. When this method is applied to a large number of SAR images (such as the number of images is greater than 20 scenes), it can reliably select high-quality coherent point targets. Therefore, in the first layer of the network, the present invention uses the amplitude deviation index threshold method to detect high-coherence scatterers and obtain the PS point targets of the first layer of the network; the amplitude deviation can be expressed by the following formula:

[0070]

[0071] Among them, m A and σ A represent the temporal mean value and standard deviation of the pixel amplitude, respectively.

[0072] Step 4: Grid thinning of the first layer of network PS points and selecting PS control points; Grid thinning of the first layer of network PS points includes: Figure 2For the PS points selected by the amplitude deviation index threshold, the highest quality and evenly distributed PS control points can be further selected. By presetting the grid size (such as 100×100 pixels), only the PS points with the relatively smallest amplitude deviation index in each grid are selected as the PS control points in the grid. The number of PS points in the first layer of the network can be reduced by sparse PS point grids, thereby reducing the computational burden of subsequent network adjustment and greatly improving the processing efficiency of PS-InSAR.

[0073] Step 5: The first layer network PS point network is constructed. The PS control points in the selected first layer network are connected using an irregular triangular network to form a PS space network. The first layer network PS point network construction steps are as follows:

[0074] Reference Figure 3 , a robust PS spatial network is constructed by connecting PS control points using an irregular triangulated network. Each arc segment of the PS spatial network represents the arc segment observation value; the PS spatial network is constructed using Delaunay triangulation, local Delaunay triangulation and free connection network.

[0075] Step 6, first layer network PS point parameter solution, the PS points in the PS space network are subjected to parameter solution; the first layer network PS point parameter solution includes two steps: arc parameter estimation and network adjustment;

[0076] On the arc segments of the PS space network, a functional relationship between the differential interferometry phase and parameters such as deformation rate and terrain error is established, and the arc parameters (mainly including deformation rate and terrain error parameters) are estimated from the arc segment observations through arc segment parameter estimators (such as periodograms, integer least squares and least squares with phase ambiguity detection).

[0077] Assuming that the two adjacent PS points in the jth interference pattern are x and y, the phase observation value on the arc segment connecting them can be calculated using the following formula:

[0078]

[0079] in is the winding operator; Δv(x, y) and Δz(x, y) are the difference in deformation rate and terrain error between PS points x and y, respectively; t j represents the time baseline of the image pair; λ is the wavelength; B ⊥ is the vertical baseline; R and θ are the radar slant range and incident angle respectively; represents the residual phase difference. Since there is generally no significant difference in the atmospheric characteristics and nonlinear deformation of adjacent PS points, it is generally believed that under the arc condition of the PS space network

[0080] Represents the residual phase. At this time, the absolute value of the complex population coherence coefficient can be used as a reliable standard for parameter estimation. The specific formula is:

[0081]

[0082] in, represents the overall coherence coefficient, M is the number of interference pairs, e j In plural form. The value range of is [0, 1], The higher the value, the more reliable the corresponding estimates of Δv(x, y) and Δz(x, y). In the traditional solution space search and solution, the range and step size of Δv(x, y) and Δz*x, y are first set, and then the corresponding values ​​of each set of solutions are calculated in a two-dimensional solution space. Then record The maximum value of and its corresponding parameter solution. In order to ensure the accuracy of the solution, it is usually Arcs below a certain threshold are discarded (the threshold is generally 0.7). Considering the inefficiency of traditional sequential search, the present invention adopts a bisection search method to estimate the arc parameter value. Each time the bisection interval is divided once, the scope of the bisection search method is reduced by half, and the corresponding consumption time is also reduced by half, which effectively improves the efficiency of arc parameter estimation.

[0083] Finally, the absolute parameters of each PS point are retrieved using a least squares-based network adjustment method through empirically selected reference points.

[0084] The network adjustment is calculated using the following formula:

[0085] B*X=L+R

[0086] X=(B T PB) -1 B T PL

[0087] Among them, B is the coefficient matrix, L and R are the arc segment observations and residuals respectively, P is the weight value set according to the complex overall coherence coefficient on the arc segment, and X is the absolute parameter of all PS points.

[0088] Step 7: Select the PS points of the second layer network in the processed SAR image. The specific steps are as follows:

[0089] In the second layer network, the remaining PS points are selected, including the PS points deleted after the PS point grid is sparsely processed in the first layer network and the remaining PS points detected using the average coherence threshold. The coherence threshold is used to select the coherent point targets with medium quality on the image. In a fixed window, the coherence of the two master-slave complex signals S1 and S2 is defined as:

[0090]

[0091] Where E(x) represents the expected value.

[0092] Step 8, the second layer network PS point networking, the selected second layer network PS points are connected with the PS control points of the first layer network to form multiple local star networks; the specific steps of the second layer network PS point networking include: in the second layer network, with the PS control points of the first layer network as a reference, the remaining PS points are measured in fragments; by connecting each remaining PS point of the second layer network to the nearest PS control point in the first layer network, multiple local star networks are formed, and each arc segment of the star network represents the arc segment observation value of the second layer network.

[0093] Step 9, solving the parameters of the PS points of the second-layer network; specifically including: establishing a functional relationship between the differential interference phase and deformation rate, terrain error and other parameters on each arc segment of the star network, estimating the arc segment parameters of the star network from the arc segment observation values ​​through the arc segment parameter estimator; based on the arc segment parameters, taking the absolute parameters of the PS points estimated by the first-layer network as a reference, directly integrating the parameters of the PS points of the second-layer network, and obtaining the absolute parameters of the PS points in the second-layer network.

[0094] Step 10, double-layer network PS link; the double-layer network PS link represents combining the PS point parameter estimation results of the double-layer network to obtain the deformation rate results and terrain error results of the entire target area.

[0095] Step 11: Double-layer network PS point time series analysis, specifically including:

[0096] The deformation rate and terrain error are the main components of the differential interferogram. The secondary differential phase can be obtained by removing the deformation rate phase and the terrain error phase. Only nonlinear deformation, atmospheric error and random noise remain in the secondary differential phase. According to the different characteristics of different components of the secondary differential phase in the time domain and space domain, the secondary differential phase is first filtered using a time high-pass filter to separate nonlinear deformation, atmospheric phase and random error. In order to separate the atmospheric phase from noise, a spatial low-pass filter is performed on the high-frequency phase. After two filterings in the time and space domains, nonlinear deformation, atmospheric phase and random noise signals are obtained. The linear deformation and nonlinear deformation are combined to restore the deformation time series of the double-layer network PS points.

[0097] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A large-scale surface deformation monitoring method based on PS-InSAR technology, characterized by: The following steps are involved: Step 1: Collect multi-view time series SAR images of the target area; Step 2: Process the SAR images, perform image registration, geocoding and differential interferometry on the SAR images; Step 3, select the PS points of the first layer network in the processed SAR image; Step 4: Grid thinning the PS points of the first layer network and selecting PS control points; Step 5: Constructing the PS points of the first layer network, connecting the selected PS control points in the first layer network using an irregular triangular network to form a PS spatial network; Step 6: Calculate the parameters of the PS points in the first layer network, and calculate the parameters of the PS points in the PS space network; Step 7, select the PS points of the second layer network in the processed SAR image; Step 8: The second layer network PS points are networked, and the selected second layer network PS points are connected with the PS control points of the first layer network to form multiple local star networks; Step 9: Calculate the parameters of the PS points of the second-layer network. Establish a functional relationship between the differential interference phase and the deformation rate, terrain error and other parameters on each arc segment of the star network. Estimate the arc segment parameters of the star network from the arc segment observation values ​​through the arc segment parameter estimator. Based on the arc segment parameters, take the absolute parameters of the PS points estimated by the first-layer network as a reference, directly integrate the parameters of the PS points of the second-layer network, and obtain the absolute parameters of the PS points in the second-layer network. Step 10: Double-layer network PS link, combining the first-layer network PS point parameter estimation results with the second-layer network PS point parameter estimation results to obtain the deformation rate results and terrain error results of the entire target area; Step 11, double-layer network PS point time series analysis, the deformation rate phase and terrain error phase are removed to obtain the secondary differential phase, the secondary differential phase is composed of nonlinear deformation, atmospheric error and random noise; according to the different characteristics of different components of the secondary differential phase in the time domain and space domain, first use time high-pass filtering to filter the secondary differential phase to separate nonlinear deformation, atmospheric phase and random error; in order to separate the atmospheric phase and noise, take spatial low-pass filtering of the high-frequency phase; after two filterings in the time and space domains, nonlinear deformation, atmospheric phase and random noise signals are obtained; combined with linear deformation and nonlinear deformation, the deformation time series of the double-layer network PS point is restored.

2. The method for monitoring large-scale surface deformation based on PS-InSAR technology according to claim 1, characterized in that: The multi-view time series SAR images of the target area collected in step 1 include but are not limited to one or more of RadarSat-2, ALOS-2, Sentinel-1A, TerraSAR-X / TanDEM-X, COSMO_SkyMed, ICEYE, Lutan-1, Gaofen-3, Hongtu-1, Luojia-2 and Fucheng-1.

3. The large-scale surface deformation monitoring method based on PS-InSAR technology according to claim 1 is characterized in that: Processing the SAR image in step 2 includes the following steps: According to the multi-view time series SAR images of the target area, the time baseline, spatial baseline and Doppler centroid frequency difference between the multi-view SAR images are compared first, and the SAR image corresponding to the maximum correlation coefficient is selected as the common main image according to the established comprehensive correlation function; The remaining SAR images are used as slave images and intensity cross-correlation registration is performed with the master image respectively; Image pairs were constructed in single master image mode, the temporal baseline and spatial baseline between the master and slave images were calculated respectively, and conventional interference processing was performed on all image pairs; Geocode the external DEM in the geographic coordinate system and convert it to the SAR image coordinate system; The interferogram and DEM in the SAR coordinate system are differentially processed to obtain a series of differential interferograms.

4. The large-scale surface deformation monitoring method based on PS-InSAR technology according to claim 1 is characterized in that: The steps for selecting the PS points of the first layer network in the processed SAR image in step 3 are as follows: In the first layer of the network, the amplitude deviation index threshold method is used to detect high coherent scatterers and obtain the PS point target of the first layer of the network; the amplitude deviation can be expressed as follows: Among them, m A and σ A represent the temporal mean value and standard deviation of the pixel amplitude, respectively; The first layer network PS point grid sparseness in step 4 includes: PS points selected by amplitude deviation index threshold are used to select the highest quality and evenly distributed PS control points; By presetting the grid size, only the PS point with the relatively smallest amplitude deviation index in each grid is selected as the PS control point in the grid; The number of PS points in the first layer of the network is reduced by sparse PS point grid.

5. The large-scale surface deformation monitoring method based on PS-InSAR technology according to claim 1 is characterized in that: Each arc segment of the PS space network represents an arc segment observation value; the PS space network is constructed by using Delaunay triangulation, local Delaunay triangulation and free connection network.

6. The large-scale surface deformation monitoring method based on PS-InSAR technology according to claim 1 is characterized in that: In the step 6, the first layer network PS point parameter solution comprises two steps of arc parameter estimation and network adjustment; The arc segment parameter estimation is calculated using the following formula: in is the winding operator; Δv(x, y) and Δz(x, y) are the difference in deformation rate and terrain error between PS points x and y, respectively; t j represents the time baseline of the image pair; λ is the wavelength; B ⊥ is the vertical baseline; R and θ are the radar slant range and incident angle respectively; represents the residual phase difference, and in, represents the overall coherence coefficient, M is the number of interference pairs, e j It is a plural form; The network adjustment is calculated using the following formula: B*X=L+R X=(B T PB) -1 B T PL Among them, B is the coefficient matrix, L and R are the arc segment observations and residuals respectively, P is the weight value set according to the complex overall coherence coefficient on the arc segment, and X is the absolute parameter of all PS points.

7. The large-scale surface deformation monitoring method based on PS-InSAR technology according to claim 1 is characterized in that: The steps for selecting the PS points of the second layer network in the processed SAR image in step 7 are as follows: In the second layer network, the remaining PS points are selected, wherein the remaining PS points include the PS points deleted after the PS point grid of the first layer network is sparsely processed and the remaining PS points detected using the average coherence threshold; The coherence threshold is used to select the coherent point targets with medium quality on the image; Within a fixed window, the coherence of two master-slave complex signals S1 and S2 is defined as: Where E(x) represents the expected value.

8. The large-scale surface deformation monitoring method based on PS-InSAR technology according to claim 1 is characterized by: The second layer network PS point networking step in step 8 includes: In the second layer network, the PS control points of the first layer network are used as reference to perform the fragmentary measurement of the remaining PS points; By connecting each remaining PS point of the second layer network to the nearest PS control point in the first layer network, a plurality of local star networks are formed, and each arc segment of the star network represents an arc segment observation value of the second layer network.

Images