A SAR offset three-dimensional deformation estimation method and system with additional DEM constraints

Abstract

The present invention discloses a SAR offset three-dimensional deformation estimation method and system with additional DEM constraints, which belongs to the field of surface three-dimensional deformation monitoring technology, and can solve the problem that only single-track SAR data can not be used to estimate three-dimensional deformation displacement in the prior art. The method comprises: obtaining two-view SAR data of the monitoring area, and preprocessing the two-view SAR data to obtain two-view single-view complex images; calculating the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex images; the two-dimensional deformation displacement includes azimuth deformation displacement and line-of-sight deformation displacement; according to the external DEM and two-dimensional deformation displacement of the monitoring area, establishing an observation equation for estimating three-dimensional deformation displacement from two-dimensional deformation displacement; the three-dimensional deformation displacement includes deformation displacement in the north direction, east direction and zenith direction; and performing parameter estimation on the observation equation to obtain the three-dimensional deformation displacement of the monitoring area. The present invention is used for monitoring geological disasters.

Description

Technical Field

[0001] The invention relates to a SAR offset three-dimensional deformation estimation method and system with additional DEM constraints, belonging to the technical field of surface three-dimensional deformation monitoring. Background Art

[0002] Synthetic aperture radar interferometry (InSAR) has been widely used in deformation monitoring of geological disasters such as landslides, earthquakes, and volcanoes due to its advantages of high precision, high resolution, and all-weather performance. InSAR can monitor deformation displacement along the line of sight (LOS) direction with a spatial resolution of tens of meters to meters and an accuracy of centimeters to millimeters under favorable conditions. However, the traditional synthetic aperture radar interferometry method has the limitation of rapid decoherence in dense vegetation and large gradient displacement areas, making it only suitable for monitoring landslides with sparse vegetation and slow movement. When the landslide movement is too fast to be measured by the interferometric synthetic aperture radar method, the SAR offset tracking method is an effective alternative. This method uses two SAR images with cross-correlated amplitude information to achieve two-dimensional deformation displacement measurement in the radar azimuth and upward direction of the line of sight, with an accuracy of approximately 1 / 10 to 1 / 30 pixels.

[0003] The 3D deformation field can provide special clues for understanding the type, failure mode and underground sliding geometry of active landslides. The 2D deformation displacement of the active part of the landslide can be estimated by the offset tracking method using single-orbit SAR data, which makes it possible to invert the 3D deformation displacement using SAR images from two or more observations with different imaging geometries. However, in most areas of the Earth, multi-orbit satellite synthetic aperture radar observations are usually unsatisfactory. Summary of the invention

[0004] The present invention provides a SAR offset three-dimensional deformation estimation method and system with additional DEM constraints, which can solve the problem in the prior art that only single-track SAR data cannot be used to estimate three-dimensional deformation displacement.

[0005] In one aspect, the present invention provides a SAR offset three-dimensional deformation estimation method with additional DEM constraints, the method comprising:

[0006] Step 11, acquiring two-view SAR data of the monitoring area, and preprocessing the two-view SAR data to obtain two-view single-view complex images;

[0007] Step 12: Calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex image; the two-dimensional deformation displacement includes the deformation displacement in the azimuth direction and the deformation displacement in the sight line direction;

[0008] Step 13: establishing an observation equation for estimating three-dimensional deformation displacement from two-dimensional deformation displacement according to the external DEM of the monitoring area and the two-dimensional deformation displacement; the three-dimensional deformation displacement includes deformation displacement in the north direction, the east direction and the zenith direction;

[0009] Step 14: perform parameter estimation on the observation equation to obtain the three-dimensional deformation displacement of the monitoring area.

[0010] Optionally, the step 12 specifically includes:

[0011] The SAR offset tracking method is used to calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-scene single-view complex images.

[0012] Optionally, the step 12 specifically includes:

[0013] Performing orthorectification and precise registration on the two-scene single-view complex images to obtain a corrected image;

[0014] The two-dimensional deformation displacement of the ground surface in the monitoring area in the corrected image is calculated.

[0015] Optionally, the step 13 specifically includes:

[0016] According to the external DEM of the monitoring area and the two-dimensional deformation displacement, and adding DEM constraint conditions, a geometric relationship between the two-dimensional deformation displacement and the three-dimensional deformation displacement is obtained;

[0017] The geometric relationship is:

[0018]

[0019] Where θ and α are the incident angles of the SAR satellite, α is the SAR satellite flight azimuth, and D N , D E and D U are the deformation displacements in the north, east and zenith directions respectively, and are the first-order derivatives of the terrain elevation H in the east direction E and the north direction N, respectively. SR and D AZI is the deformation displacement observation value in the radar line of sight and azimuth, Δ SR and Δ AZI are the corresponding deformation displacement observation errors in the line of sight and azimuth directions respectively;

[0020] Convert the geometric relationship into a matrix form to obtain an observation equation;

[0021] The observation equation is:

[0022] AX = L + ΔL;

[0023] In the formula, is the design matrix, Middle D N , D E and D U are the three-dimensional deformation displacements in the north, east and zenith directions respectively, L = [D SR D AZI 0] T Middle D SR and D AZI is the observed value of deformation displacement in the radar line of sight and azimuth, ΔL=[ΔD SR ΔD AZI 0] is the deformation displacement observation error in the radar line of sight and azimuth.

[0024] Optionally, after step 13, the method further includes:

[0025] According to the external DEM error, the observation equation is corrected using the EIV model to obtain a corrected equation;

[0026] The corrected equation is:

[0027]

[0028] Where ΔA is the random error matrix of the design matrix A;

[0029] Accordingly, the step 14 is:

[0030] Parameters of the corrected equation are estimated to obtain the three-dimensional deformation displacement of the monitoring area.

[0031] Optionally, step 14 specifically includes:

[0032] The total least square method is used to estimate the parameters of the corrected equation to obtain the three-dimensional deformation displacement of the monitoring area.

[0033] Optionally, step 14 specifically includes:

[0034] Performing singular value decomposition on the augmented matrix G of the corrected equation to obtain a first formula;

[0035] The first formula is:

[0036]

[0037] Where U = [U1 U2] is the matrix [AL][AL] T The orthogonal matrix composed of n eigenvectors of is a matrix The orthogonal matrix composed of m+1 eigenvectors, S=diag(σ1,σ2,···,σ m+1 ) is composed of the singular values ​​σ1,σ2,···,σ m+1 The diagonal array composed of

[0038] In combination with the first formula, a low-rank approximation method is used to obtain the total least squares solution of the three-dimensional deformation displacement X to obtain the three-dimensional deformation displacement of the monitoring area.

[0039] Optionally, the low-rank approximation method is used to obtain the total least squares solution of the three-dimensional deformation displacement X to obtain the three-dimensional deformation displacement of the monitoring area, specifically:

[0040] The total least squares solution of the three-dimensional deformation displacement X is calculated according to the second formula;

[0041] The second formula is:

[0042]

[0043] Optionally, when solving the three-dimensional deformation displacement X, the overall least squares criterion used is vec(ΔA) T vec(ΔA)+(ΔL) T ΔL = min;

[0044] Where vec(·) represents the column vectorization of the matrix.

[0045] On the other hand, an embodiment of the present invention provides a SAR offset three-dimensional deformation estimation system with additional DEM constraints, the system comprising:

[0046] A preprocessing unit is used to obtain two-view SAR data of the monitoring area and preprocess the two-view SAR data to obtain two-view single-view complex images;

[0047] A calculation unit, used to calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex image; the two-dimensional deformation displacement includes the deformation displacement in the azimuth direction and the deformation displacement in the sight direction;

[0048] An equation building unit is used to build an observation equation for estimating a three-dimensional deformation displacement from a two-dimensional deformation displacement according to an external DEM of the monitoring area and the two-dimensional deformation displacement; the three-dimensional deformation displacement includes deformation displacements in the north direction, the east direction and the zenith direction;

[0049] The parameter estimation unit is used to perform parameter estimation on the observation equation to obtain the three-dimensional deformation displacement of the monitoring area.

[0050] The beneficial effects that the present invention can produce include:

[0051] The SAR offset three-dimensional deformation estimation method with additional DEM constraints provided by the present invention eliminates the rank deficiency problem of the conversion equation between two-dimensional deformation displacement and three-dimensional deformation displacement by adding external DEM constraints, and realizes the estimation of three-dimensional deformation displacement from two-dimensional deformation displacement. The advantage of this method is that the azimuth and line-of-sight offsets obtained by single-track SAR data can realize three-dimensional deformation monitoring of the monitoring area, reducing the requirement of SAR data for multiple different tracks for three-dimensional deformation monitoring. At the same time, after the external DEM is introduced, the error of the external DEM is taken into consideration, and the total least squares method is used for parameter estimation, thereby improving the solution accuracy of the three-dimensional deformation displacement. This method can be applied to the research in the field of gravity-driven landslide geological disaster monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A flow chart of a SAR offset three-dimensional deformation estimation method with additional DEM constraints provided in an embodiment of the present invention;

[0053] Figure 2 A deformation rate diagram of the Laojingbian landslide in the radar azimuth provided by an embodiment of the present invention;

[0054] Figure 3 A deformation rate diagram of the Laojingbian landslide in the radar line of sight provided by an embodiment of the present invention;

[0055] Figure 4 A deformation rate diagram of the Laojingbian landslide in the north direction provided by an embodiment of the present invention;

[0056] Figure 5 A deformation rate diagram of the Laojingbian landslide in the east direction provided by an embodiment of the present invention;

[0057] Figure 6 A deformation rate diagram of the Laojingbian landslide in the zenith direction provided by an embodiment of the present invention;

[0058] Figure 7 A horizontal movement rate diagram of the Laojingbian landslide provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The present invention is described in detail below in conjunction with embodiments, but the present invention is not limited to these embodiments.

[0060] The embodiment of the present invention provides a SAR offset three-dimensional deformation estimation method with additional DEM constraints, such as Figure 1 As shown, the method includes:

[0061] Step 11, obtain two-view SAR data of the monitoring area, and pre-process the two-view SAR data to obtain two-view single-look complex images. Among them, SAR (Synthetic Aperture Radar) is synthetic aperture radar; single-look complex image is SLC (single look complex) image; digital elevation model (Digital Elevation Model), referred to as DEM.

[0062] For example, two-view SAR data are two-view data of the same area before and after the landslide.

[0063] Step 12: Calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-scene single-view complex image; the two-dimensional deformation displacement includes the deformation displacement in the azimuth direction and the deformation displacement in the sight line direction.

[0064] In practical applications, the SAR offset tracking method can be used to calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex image.

[0065] Specifically, the method includes: (1) orthorectifying and precisely registering the two-view single-view complex images to obtain the corrected images; and (2) calculating the two-dimensional deformation displacement of the surface of the monitoring area in the corrected images.

[0066] The specific idea of ​​the SAR offset tracking method is: select one of the scenes as the main image, register and resample the other slave image to the main image, so that the main image and the slave image have the same spatial geometric structure; use the SAR amplitude information to roughly register the two SAR images, calculate the initial offset between the main and slave images, and then select a search window of a certain size on the basis of the rough registration and calculate the normalized mutual correlation coefficient ρ(x,y) between the main and slave images, as shown in the following formula:

[0067]

[0068] Where x and y are the main image coordinates, x s and s is the offset in both directions, m and n are the window sizes, and f m and f s They are the master image amplitude and the slave image amplitude respectively.

[0069] When the correlation coefficient ρ(x, y) reaches the maximum, the accurate sub-pixel offset can be obtained, and finally the azimuthal deformation D of the surface can be obtained. AZI and the line of sight deformation D SR .

[0070] Step 13: Based on the external DEM and two-dimensional deformation displacement of the monitoring area, an observation equation for estimating three-dimensional deformation displacement from two-dimensional deformation displacement is established; the three-dimensional deformation displacement includes deformation displacement in the north direction, east direction and zenith direction.

[0071] The step 13 specifically includes:

[0072] (1) Based on the external DEM and two-dimensional deformation displacement of the monitoring area, and adding DEM constraints, the geometric relationship between two-dimensional deformation displacement and three-dimensional deformation displacement is obtained;

[0073] The geometric relationship is:

[0074]

[0075] Where θ and α are the incident angles of the SAR satellite, α is the SAR satellite flight azimuth, and D N , D E and D U are the deformation displacements in the north, east and zenith directions respectively, and are the first-order derivatives of the terrain elevation H in the east direction E and the north direction N, respectively. SR and D AZI is the deformation displacement observation value in the radar line of sight and azimuth, Δ SR and Δ AZI are the corresponding deformation displacement observation errors in the line of sight and azimuth directions respectively;

[0076] Among them, the first-order derivative of the terrain elevation H in the monitoring area in the east direction E and the north direction N is and It is calculated using an external DEM.

[0077] (2) Convert the geometric relationship into a matrix form to obtain the observation equation;

[0078] The observation equation is:

[0079] AX = L + ΔL;

[0080] In the formula, is the design matrix, Middle D N , D E and D U are the three-dimensional deformation displacements in the north, east and zenith directions respectively, L = [D SR D AZI 0] T Middle D SR and D AZI is the observed value of deformation displacement in the radar line of sight and azimuth, ΔL=[ΔD SRΔD AZI 0] is the deformation displacement observation error in the radar line of sight and azimuth.

[0081] Step 14: Estimate the parameters of the observation equation to obtain the three-dimensional deformation displacement of the monitoring area.

[0082] Since the external DEM contains errors, the design matrix A also contains errors. Therefore, in order to improve the accuracy of parameter estimation, the observation equation needs to be corrected. Therefore, after step 13, the method further includes:

[0083] According to the external DEM error, the EIV (errors-in-variables) model is used to add the random error matrix ΔA to the design matrix A of the observation equation to correct the observation equation and obtain the corrected equation.

[0084] The corrected equation is:

[0085] (A+ΔA)X=L+ΔL;

[0086] It can also be written as:

[0087] Where ΔA is the random error matrix of the design matrix A;

[0088] Correspondingly, step 14 is: perform parameter estimation on the corrected equation to obtain the three-dimensional deformation displacement.

[0089] In practical applications, the total least squares method can be used to estimate the parameters of the corrected equation to obtain the three-dimensional deformation displacement.

[0090] The mean vector and variance-covariance matrix of the error vector can be expressed as:

[0091]

[0092] Where vec(·) represents the column vectorization of the matrix, Σ L and Σ A are the variance-covariance matrices of ΔL and ΔA, respectively.

[0093] Step 14 specifically includes: (1) performing singular value decomposition on the augmented matrix G of the corrected equation to obtain a first formula;

[0094] The first formula is:

[0095]

[0096] Where U = [U1 U2] is the matrix [AL][AL] T The orthogonal matrix composed of n eigenvectors of is a matrix The orthogonal matrix composed of m+1 eigenvectors, S=diag(σ1,σ2,···,σ m+1 ) is composed of the singular values ​​σ1,σ2,···,σ m+1 The diagonal array composed of

[0097] (2) Combined with the first formula, the low-rank approximation method (Eckart–Young–Mirsky theorem) is used to obtain the total least squares solution of the three-dimensional deformation displacement X to obtain the three-dimensional deformation displacement.

[0098] Among them, the low-rank approximation method is used to obtain the total least squares solution of the three-dimensional deformation displacement X to obtain the three-dimensional deformation displacement, which is specifically:

[0099] The total least squares solution of the three-dimensional deformation displacement X is calculated according to the second formula;

[0100] The second formula is:

[0101]

[0102] When solving the three-dimensional deformation displacement X, the overall least squares criterion used is vec(ΔA) T vec(ΔA)+(ΔL) T ·ΔL=min.

[0103] The present invention eliminates the rank deficiency problem of the conversion equation between two-dimensional deformation displacement and three-dimensional deformation displacement by adding external DEM constraints, and realizes the estimation of three-dimensional deformation displacement from two-dimensional deformation displacement. The advantage of this method is that the three-dimensional deformation monitoring of the monitoring area is realized by the azimuth and line of sight offsets obtained by single-track SAR data, which reduces the requirement of SAR data for multiple different tracks for three-dimensional deformation monitoring. At the same time, after the external DEM is introduced, the error of the external DEM is taken into account, and the total least squares method is used for parameter estimation, which improves the solution accuracy of the three-dimensional deformation displacement. This method can be applied to the research in the field of gravity-driven landslide geological disaster monitoring.

[0104] Another embodiment of the present invention provides a specific SAR offset three-dimensional deformation estimation method with additional DEM constraints. This embodiment selects the Laojingbian landslide in the Wudongde Reservoir area located in the lower reaches of the Jinsha River in Yunnan Province, China. The Wudongde Reservoir area is located on the southeast edge of the Qinghai-Tibet Plateau, forming a transition zone between the Yunnan-Guizhou Plateau and the Sichuan Basin, thus producing a special geological environment including deep valleys and narrow valleys and high mountains. In addition, the area is one of the most intense tectonic activities and earthquake-prone areas in China, belonging to the subtropical monsoon climate zone, with concentrated rainfall, strong solar radiation, and large daily temperature differences, resulting in strong weathering of slope materials. Due to the coupling effect of gravity, physical weathering, seismic activity and heavy rainfall, the area is very prone to landslides.

[0105] The SAR data used in this embodiment are 12 ALOS / PALSAR-2 images collected from September 2014 to May 2020, and the DEM data used is the ALOS global digital surface model (AW3D30DSM) data with a horizontal resolution of 30 meters (1 arc second). The directional derivative of the terrain is calculated and the system offset error caused by the terrain undulation is eliminated. The method specifically includes:

[0106] Step 21, first pre-process the acquired 12 ALOS / PALSAR-2 images to generate single-look complex images (SLC).

[0107] Step 22, orthorectification and precise registration of different SLC images. Terrain undulation can cause pixel misalignment of SAR images, which can significantly reduce the registration accuracy of SAR images. Therefore, based on SAR imaging geometry and external DEM, orthorectification is performed to remove the terrain undulation effect. Then, based on the orthorectified SAR images, the primary and secondary images are precisely registered. The basic principle and processing method are as follows:

[0108] (1) The best primary image among different SAR images is determined based on the Doppler center frequency variation and the spatiotemporal baseline. Then, based on the SAR imaging geometry and the external DEM, a direct function is established to record the position of each pixel in the primary image to the corresponding pixel position in the secondary image.

[0109] (2) The established direct function is refined to handle the deviations caused by the SAR image orbit state vector error and the external DEM error.

[0110] (3) According to the corrected direct function, all auxiliary SAR images are resampled into the coordinate system of the main image to further obtain the corrected SAR image.

[0111] Step 23, using SAR offset technology, calculate the deformation rate of the monitoring area along the radar azimuth and line of sight from September 2014 to May 2020, and finally obtain the deformation rate in the azimuth and line of sight as follows: Figure 2 and Figure 3 As shown. Among them, Figure 2 is the azimuthal deformation rate diagram, Figure 3 This is the line of sight deformation rate diagram.

[0112] Step 24: Calculate the elevation of the monitoring area in the east direction based on the AW3D30 DSM data of the monitoring area. E and north direction N Directional derivative of and

[0113] Step 25, based on the geometric relationship, establish an observation equation for solving the three-dimensional (north, east, and zenith) deformation rate from the two-dimensional (azimuth and line of sight) deformation rate, and obtain the overall least squares solution of the parameters according to the first formula to obtain the three-dimensional deformation rate as follows: Figure 4 , Figure 5 and Figure 6 As shown. Among them, Figure 4 is the deformation rate diagram in the north direction, Figure 5 is the deformation rate diagram in the east direction, Figure 6 This is the deformation rate diagram in the zenith direction.

[0114] Step 26, plot the horizontal movement rate of the landslide based on the three-dimensional deformation rate. The three-dimensional deformation rate result provides an accurate description of the real movement of the landslide and can better reveal the spatial deformation pattern of the Laojingbian landslide. Therefore, the horizontal movement rate of the landslide is synthesized using the north and east deformation rates of the three-dimensional deformation rate, as shown in Figure 7 shown.

[0115] It should be noted that, in practical applications, the deformation displacement can be used to measure the deformation of the landslide, and the deformation rate can also be used to measure the deformation of the landslide, which is not limited in the embodiments of the present invention.

[0116] Another embodiment of the present invention provides a SAR offset three-dimensional deformation estimation system with additional DEM constraints, the system comprising:

[0117] A preprocessing unit is used to obtain two-view SAR data of the monitoring area and preprocess the two-view SAR data to obtain two-view single-view complex images;

[0118] A calculation unit, used to calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex image; the two-dimensional deformation displacement includes the deformation displacement in the azimuth direction and the deformation displacement in the sight direction;

[0119] An equation building unit is used to build an observation equation for estimating a three-dimensional deformation displacement from a two-dimensional deformation displacement according to an external DEM and a two-dimensional deformation displacement of the monitoring area; the three-dimensional deformation displacement includes deformation displacements in the north direction, the east direction and the zenith direction;

[0120] The parameter estimation unit is used to estimate the parameters of the observation equation and obtain the three-dimensional deformation displacement.

[0121] The specific description of each unit in the above estimation system can refer to the description of each step in the estimation method, which will not be repeated here. The above estimation system can achieve the same function as the estimation method.

[0122] The above are only a few embodiments of the present application and do not constitute any form of limitation to the present application. Although the present application is disclosed as above with preferred embodiments, it is not intended to limit the present application. Any technician familiar with the profession, without departing from the scope of the technical solution of the present application, using the technical content disclosed above to make slight changes or modifications are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A SAR offset three-dimensional deformation estimation method with additional DEM constraints, characterized in that: The method comprises: Step 11, acquiring two-view SAR data of the monitoring area, and preprocessing the two-view SAR data to obtain two-view single-view complex images; Step 12: Calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex image; the two-dimensional deformation displacement includes the deformation displacement in the azimuth direction and the deformation displacement in the sight line direction; Step 13: establishing an observation equation for estimating three-dimensional deformation displacement from two-dimensional deformation displacement according to the external DEM of the monitoring area and the two-dimensional deformation displacement; the three-dimensional deformation displacement includes deformation displacement in the north direction, the east direction and the zenith direction; Step 14: perform parameter estimation on the observation equation to obtain the three-dimensional deformation displacement of the monitoring area.

2. The method according to claim 1, characterized in that The step 12 specifically includes: The SAR offset tracking method is used to calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-scene single-view complex image.

3. The method according to claim 2, characterized in that The step 12 specifically includes: Performing orthorectification and precise registration on the two-scene single-view complex images to obtain a corrected image; The two-dimensional deformation displacement of the ground surface in the monitoring area in the corrected image is calculated.

4. The method according to claim 1, characterized in that: The step 13 specifically includes: According to the external DEM of the monitoring area and the two-dimensional deformation displacement, and adding DEM constraint conditions, a geometric relationship between the two-dimensional deformation displacement and the three-dimensional deformation displacement is obtained; The geometric relationship is: Where θ and α are the incident angles of the SAR satellite, α is the SAR satellite flight azimuth, and D N , D E and D U are the deformation displacements in the north, east and zenith directions respectively, and are the first-order derivatives of the terrain elevation H in the east direction E and the north direction N, respectively. SR and D AZI is the deformation displacement observation value in the radar line of sight and azimuth, Δ SR and Δ AZI are the corresponding deformation displacement observation errors in the line of sight and azimuth directions respectively; Convert the geometric relationship into a matrix form to obtain an observation equation; The observation equation is: AX = L + ΔL; In the formula, is the design matrix, Middle D N , D E and D U are the three-dimensional deformation displacements in the north, east and zenith directions respectively, L = [D SR D AZI 0] T Middle D SR and D AZI is the observed value of deformation displacement in the radar line of sight and azimuth, ΔL=[ΔD SR ΔD AZI 0] is the deformation displacement observation error in the radar line of sight and azimuth.

5. The method according to claim 4, characterized in that After step 13, the method further includes: According to the external DEM error, the observation equation is corrected using the EIV model to obtain a corrected equation; The corrected equation is: Where ΔA is the random error matrix of the design matrix A; Accordingly, the step 14 is: Parameters of the corrected equation are estimated to obtain the three-dimensional deformation displacement of the monitoring area.

6. The method according to claim 5, characterized in that The step 14 specifically includes: The total least square method is used to estimate the parameters of the corrected equation to obtain the three-dimensional deformation displacement of the monitoring area.

7. The method according to claim 6, characterized in that The step 14 specifically includes: Performing singular value decomposition on the augmented matrix G of the corrected equation to obtain a first formula; The first formula is: Where U = [U1U2] is the matrix [AL][AL] T The orthogonal matrix composed of n eigenvectors of is a matrix The orthogonal matrix composed of m+1 eigenvectors, S=diag(σ1,σ2,···,σ m+1 ) is composed of the singular values ​​σ1,σ2,···,σ m+1 The diagonal array composed of In combination with the first formula, a low-rank approximation method is used to obtain the total least squares solution of the three-dimensional deformation displacement X to obtain the three-dimensional deformation displacement of the monitoring area.

8. The method according to claim 7, characterized in that The low-rank approximation method is used to obtain the total least squares solution of the three-dimensional deformation displacement X to obtain the three-dimensional deformation displacement of the monitoring area, specifically: The total least squares solution of the three-dimensional deformation displacement X is calculated according to the second formula; The second formula is:

9. The method according to claim 6 or 7, characterized in that: When solving the three-dimensional deformation displacement X, the overall least squares criterion used is vec(ΔA) T vec(ΔA)+(ΔL) T ΔL = min; Where vec(·) represents the column vectorization of the matrix.

10. A SAR offset three-dimensional deformation estimation system with additional DEM constraints, characterized in that: The system comprises: A preprocessing unit is used to obtain two-view SAR data of the monitoring area and preprocess the two-view SAR data to obtain two-view single-view complex images; A calculation unit, used to calculate the two-dimensional deformation displacement of the surface of the monitoring area in the two-view single-view complex image; the two-dimensional deformation displacement includes the deformation displacement in the azimuth direction and the deformation displacement in the sight direction; An equation building unit is used to build an observation equation for estimating a three-dimensional deformation displacement from a two-dimensional deformation displacement according to an external DEM of the monitoring area and the two-dimensional deformation displacement; the three-dimensional deformation displacement includes deformation displacements in the north direction, the east direction and the zenith direction; The parameter estimation unit is used to perform parameter estimation on the observation equation to obtain the three-dimensional deformation displacement of the monitoring area.

Images