River slope landslide risk monitoring method based on corner reflector and PS-InSAR technology

Abstract

The present invention discloses a method for monitoring river slope landslide risk based on corner reflectors and PS-InSAR technology, comprising obtaining at least one SAR image of a target area to be monitored and selecting a main image; processing the main image based on PS-InSAR technology to obtain preliminary deformation monitoring results; analyzing the preliminary deformation monitoring results to determine a stable point adjacent to the target area to be monitored as the first corner reflector placement position; gridding the target area to be monitored, performing real-scene analysis in the divided grids, and determining the second corner reflector placement position; using the first corner reflector as an unwrapping starting point to obtain a deformation phase, and obtaining a final deformation monitoring result after correction; and dividing key landslide risk areas based on the final deformation monitoring result. The present invention combines the corner reflector placement method with PS-InSAR technology to provide a method that can effectively monitor river slope landslide risks. This method ensures both the optimal number of corner reflectors and the accuracy of target monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of SAR satellite image processing and analysis, and in particular to a river slope landslide risk monitoring method based on corner reflector and PS-InSAR technology. Background Art

[0002] Landslide disasters are the most widespread among the six types of geological disasters, including landslides, collapses, mudslides, ground collapses, ground fissures, and ground subsidence. According to statistics, landslides account for 69.34% of the total number of landslides. They are characterized by high frequency of occurrence, high distribution density, wide impact area, and serious hazards.

[0003] InSAR (Synthetic Aperture Radar Interferometry) is an active microwave remote sensing technology that enables wide-area geometric measurement of the Earth's surface and offers all-day, all-weather observation capabilities. Since the application of DInSAR technology to obtain deformation information from the Landers earthquake in 1993, InSAR technology has gradually been applied to surface deformation monitoring. To address the temporal and spatial decorrelation issues that severely hinder the widespread application of DInSAR technology, researchers have proposed time-series InSAR technology, pushing the application of InSAR technology to a new stage and providing a new approach for studying geological hazards such as landslides. Summary of the Invention

[0004] In view of the fact that the existing technology does not use InSAR technology to realize geological disaster monitoring methods, the present invention aims to provide a river slope landslide risk monitoring method based on corner reflectors and PS-InSAR technology, which not only ensures the optimal number of corner reflectors but also ensures the accuracy of monitoring.

[0005] To achieve the above technical objectives, the first aspect of the present invention provides a river slope landslide risk monitoring method based on corner reflectors and PS-InSAR, comprising:

[0006] Acquire at least one SAR image of the target area to be monitored and select the main image;

[0007] Processing the main image based on PS-InSAR technology to obtain preliminary deformation monitoring results, wherein the preliminary deformation monitoring results include deformation value and deformation rate;

[0008] Analyze the preliminary deformation monitoring results and determine the stable point close to the target area to be monitored as the first corner reflector placement location;

[0009] Divide the target area to be monitored into grids, perform real-scene analysis in the divided grids, and determine the placement position of the second corner reflector;

[0010] The first corner reflector is used as the unwrapping starting point to obtain the deformation phase, and the final deformation monitoring result is obtained after correction;

[0011] Key landslide risk areas are divided based on the final deformation monitoring results.

[0012] The PS-InSAR processing of the main image mainly includes: selection of the main image, identification of permanent scatterers (PS points), PS differential phase modeling and parameter estimation, and decomposition of nonlinear signals of the PS network.

[0013] Furthermore, the step of acquiring at least one SAR image of the target area to be monitored and selecting a main image includes:

[0014] Acquire multiple SAR images of the target area to be monitored at different time points;

[0015] Focus each SAR image to generate an SLC format image;

[0016] Generate intensity maps, geocode, and crop target areas of SLC format images to obtain cropped images.

[0017] The image with the largest average intensity value is selected from the cropped images as the final main image.

[0018] Furthermore, before selecting the image with the largest average intensity value from the cropped images as the final main image, the method further includes:

[0019] Preliminary selection of candidate main images from the cropped images based on the time baseline and the space baseline;

[0020] Calculate the average intensity of the target area in the candidate main image, where the intensity I satisfies: I = A 2 , A is the image amplitude.

[0021] Furthermore, the first corner reflector is arranged as follows:

[0022] Extracting a stable area based on the deformation value in the preliminary deformation monitoring result and a preset first deformation rate threshold;

[0023] A relatively flat area that is relatively close to the target area to be monitored is selected as the first corner reflector placement point.

[0024] Furthermore, the first corner reflector is used as the unwrapping starting point to obtain the deformation phase, and the final deformation monitoring result is obtained after correction, including:

[0025] The first corner reflector is used as the unwrapping starting point, and a triangulated network is formed by selecting PS points from the target area to be monitored based on the amplitude deviation method. The first deformation phase is obtained using conventional PS-InSAR.

[0026] The first corner reflector is used as the unwrapping starting point, and a triangulated network is formed by selecting PS points from the target area to be monitored based on the amplitude deviation and correlation coefficient method. The second deformation phase is obtained using conventional PS-InSAR.

[0027] Correcting the first deformation phase using the second deformation phase;

[0028] The corrected first deformation phase is transformed into deformation phase, and the final deformation monitoring result is obtained after geocoding.

[0029] Furthermore, the first deformation phase is obtained specifically according to the following method:

[0030] Acquire multiple SAR images of the target area to be monitored;

[0031] Calculate the amplitude deviation value for each SAR image;

[0032] Determine the amplitude deviation threshold suitable for the target area to be monitored, select the PS points and unwrapping starting points that meet the conditions to form a path integral network;

[0033] Finally, conventional PS-InSAR is used to obtain the first deformation phase.

[0034] Furthermore, the second deformation phase is obtained specifically according to the following method:

[0035] Acquire multiple SAR images of the target area to be monitored;

[0036] Calculate the amplitude deviation value for each SAR image;

[0037] Filter out the first high-quality PS point set according to the amplitude deviation threshold;

[0038] Perform precise registration of the master image and the slave image one by one to obtain the correlation coefficient;

[0039] Set a correlation coefficient threshold, and select PS points whose target point correlation coefficient sequence remains above the correlation coefficient threshold from the first high-quality PS point set to form a second high-quality PS point set;

[0040] Combining the second highest quality PS point set into a PS point network for disentanglement;

[0041] Finally, conventional PS-InSAR is used to obtain the second deformation phase.

[0042] Furthermore, the correcting the first deformation phase by using the second deformation phase includes:

[0043] Extracting deformation points that can be matched in the first deformation phase and the second deformation phase monitoring;

[0044] Calculate the deformation rate difference between each set of matched deformation points;

[0045] The error correction network is constructed based on the deformation rate difference using bilinear interpolation method.

[0046] According to the error correction network, the deformation values of the PS points obtained based on the amplitude deviation method are corrected one by one.

[0047] Furthermore, the phase-conversion of the corrected first deformation phase is performed, and a final deformation monitoring result is obtained after geocoding, specifically including:

[0048] Using the parameters obtained from the pulse echo time delay and the Doppler parameters obtained during the imaging process, the coordinate values in the inertial coordinate system are obtained through the slant range Doppler positioning equation;

[0049] The corresponding geographic longitude and latitude are obtained according to the geographic coordinate conversion method.

[0050] Furthermore, the key landslide risk areas are divided according to the final deformation monitoring results, including:

[0051] Obtain the PS points whose annual average deformation rate is above the second deformation rate threshold, and record them as the first abnormal deformation PS point set;

[0052] Calculate the deformation rates of the PS points in the first abnormal deformation PS point set in a short period in sequence, extract the points where the deformation rate in the short period continuously increases, and divide these points into the second abnormal deformation PS point set;

[0053] The points in the second abnormal deformation PS point set are superimposed on the optical satellite image to draw the corresponding risk area vector;

[0054] Identify key landslide risk areas based on risk area vectors.

[0055] In a second aspect, the present invention further provides a river slope landslide risk monitoring system based on corner reflectors and PS-InSAR, comprising:

[0056] A main image acquisition module is used to acquire at least one SAR image of the target area to be monitored and select a main image;

[0057] A first processing module processes the main image based on the PS-InSAR technology to obtain preliminary deformation monitoring results, wherein the preliminary deformation monitoring results include deformation value and deformation rate;

[0058] A first determination module is used to analyze the preliminary deformation monitoring results and determine a stable point adjacent to the target area to be monitored as the first corner reflector placement position;

[0059] The second determination module divides the target area to be monitored into grids, performs real-scene analysis in the divided grids, and determines the placement position of the second corner reflector;

[0060] The second processing module is used to obtain the deformation phase by taking the first corner reflector as the unwrapping starting point, and obtain the final deformation monitoring result after correction;

[0061] Identification module, which divides key landslide risk areas according to the final deformation monitoring results.

[0062] This application uses corner reflector interferometry (corner reflactor-InSAR) technology to deploy a certain number of corner reflectors in the target monitoring area to reduce the impact of spatiotemporal incoherence on differential InSAR technology, and achieve the goal of being able to monitor the deformation of complex target areas with high precision. At the same time, PS-InSAR technology uses multiple SAR images of the target area, analyzes all SAR images according to certain rules, and identifies those points that are not easily affected by temporal baseline changes or spatial baseline changes as permanent scatterers (PS points). The selected PS points are then constructed into a PS network according to certain rules. Deformation information is extracted by performing differential processing on any two adjacent points in the network, and based on this, the deformation rate difference and elevation difference between each adjacent PS point are obtained. The calculated deformation parameter value and elevation difference of the PS point are regarded as known numbers, and a model is constructed to solve the deformation parameter value and elevation parameter value of all PS points in the target area. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 Schematic diagram of the flow of the river slope landslide risk monitoring method based on corner reflector and PS-InSAR in this embodiment;

[0064] Figure 2 Flowchart of the screening method for primary images;

[0065] Figure 3 A flow chart of a method for obtaining a first deformation phase using a first corner reflector as an unwrapping starting point;

[0066] Figure 4 A flow chart of a method for obtaining a second deformation phase using a first corner reflector as an unwrapping starting point;

[0067] Figure 5 A flow chart for correcting a first deformation phase using a second deformation phase;

[0068] Figure 6 Flowchart for converting the corrected first deformation phase into deformation and obtaining the final deformation monitoring result after geocoding;

[0069] Figure 7 A flow chart for demarcating key landslide risk areas based on the final deformation monitoring results;

[0070] Figure 8 The temporal and spatial relationship distribution diagrams of the master image and the slave image in the selected SAR image;

[0071] Figure 9 This is the average intensity value change diagram of the 7 best SAR images after time baseline and space baseline selection;

[0072] Figure 10 This is the distribution map of the preliminary deformation monitoring results of the target area based on PS-InSAR technology;

[0073] Figure 11 Schematic diagram of the position selection for corner reflector No. 1;

[0074] Figure 12 is the angular reflection distribution diagram of the final layout;

[0075] Figure 13 The deformation distribution map obtained by PS-InSAR technology is based on the PS points and corner reflectors obtained under relatively loose selection conditions;

[0076] Figure 14 The deformation result distribution map obtained based on PS-InSAR technology is based on PS points and corner reflectors obtained under relatively strict selection conditions;

[0077] Figure 15 The final deformation monitoring result distribution map is obtained;

[0078] Figure 16 It is a distribution map of local areas that need to be monitored based on the deformation rate value and the deformation rate change trend. DETAILED DESCRIPTION

[0079] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and drawings. The contents mentioned in the embodiments are not intended to limit the present invention.

[0080] like Figure 1 As shown, this embodiment provides a river slope landslide risk monitoring method based on corner reflector and PS-InSAR, comprising the following steps:

[0081] S100: Acquire at least one SAR image of the target area to be monitored and select the main image

[0082] SAR (Synthetic Aperture Radar) images can be acquired from multiple satellites and the main image can be selected based on the principle of optimal quality. Figure 2 As shown, the specific screening method of the main image is as follows:

[0083] S101: Acquire multiple SAR images of the target area to be monitored at different time points;

[0084] S102: Focusing each SAR image to generate an SLC (Single Look Complex) image, where the SLC image contains focused SAR data.

[0085] S103: Generating an intensity map, geocoding, and cropping a target area of the SLC image to obtain a cropped image.

[0086] S104: Select the image with the largest average intensity value from the cropped images as the final main image.

[0087] The intensity of the image I satisfies: I = A 2 , where A is the image amplitude. The average intensity value in each cropped image is calculated, and the cropped image with the maximum value is selected as the final main image.

[0088] In order to reduce the amount of intensity value calculation, before S104, a candidate main image can be preliminarily selected from the cropped image. The selection of the candidate main image is based on the time baseline and the spatial baseline. This avoids the need to calculate the average intensity value of each image. Instead, only the average intensity value of the preliminarily selected candidate main image needs to be calculated.

[0089] In one embodiment, six candidate main images are screened out based on the above-mentioned judgment criteria. If the total number of cropped images is the base number, seven images are selected.

[0090] S200: Processing the main image based on the PS-InSAR technology to obtain preliminary deformation monitoring results, wherein the preliminary deformation monitoring results include deformation values and deformation rates.

[0091] The basic processing based on PS-InSAR technology is:

[0092] After selecting a master image from the time series of SAR images, the remaining SAR images (slave images) are projected into the geometric space of the only master image through image registration to form a long time series of interferometric phase maps. Finally, differential interferometric processing is performed on all the interferometric phase maps.

[0093] Persistent Scatter (PS) Identification: Persistent scatterers are typically much smaller than the spatial resolution unit of a SAR image, and these PS points maintain stable scattering characteristics over long periods of time. Common PS point identification methods include the amplitude deviation index threshold method, the phase deviation threshold method, and the phase noise stability-based identification method.

[0094] PS differential phase modeling and parameter estimation: A PS point network model is constructed based on the Delaunay triangulation network. A differential phase model of adjacent PS points is established. By solving the model parameters, the elevation error increment and deformation rate increment between two adjacent PS points in the differential phase are estimated. This problem is thus transformed into a maximization problem of the correlation coefficient function. Maximizing the objective function can obtain the elevation error increment and deformation rate increment between adjacent PS points.

[0095] Decomposition of nonlinear signals in PS networks: In the residual phase, the effects of atmospheric delay and nonlinear deformation phase exhibit different characteristics in the time and spatial domains of the interferogram. It is generally believed that atmospheric delay phase appears as a low-frequency signal in the spatial domain. Using time-space filtering, the atmospheric phase, nonlinear deformation components, and noise components can be separated from the phase. The linear and nonlinear deformation components at the PS points are summed to obtain the deformation values at discrete PS points on the bridge deck.

[0096] S300: Analyze the preliminary deformation monitoring results and determine a stable point adjacent to the target area to be monitored as the placement position of the first corner reflector (corner reflector No. 1).

[0097] Specifically: according to the deformation value in the preliminary deformation monitoring results and the preset first deformation rate threshold, a stable area is extracted, and then a relatively flat area that is closest to or relatively close to the target area to be monitored is selected, and the first corner reflector is deployed in this area. The choice of stability is confirmed based on the dual principles of distance and stability.

[0098] S400: Divide the target area to be monitored into grids, perform real-scene analysis in the divided grids, and determine the placement position of the second corner reflector (corner reflector No. 2).

[0099] The grid size of the target area to be monitored can be set to 200m×200m. The grid size can be enlarged for areas with low complexity, and can be reduced for areas with high complexity.

[0100] Certain real-scene analysis is performed in the divided grids. For example, if the grid of the target area to be monitored is basically full of trees with high density, then two corner reflectors are generally required. The specific positions need to refer to the layout of adjacent corner reflectors, and ultimately achieve approximately uniform distribution of corner reflectors.

[0101] For grid areas with relatively many PS points, one corner reflector may be deployed, or even no corner reflector may be deployed.

[0102] S500: Using the first corner reflector as the unwrapping starting point to obtain the deformation phase, and obtaining the final deformation monitoring result after correction.

[0103] Specifically, the deformation phase is obtained by taking the first corner reflector as the unwrapping starting point to obtain the first deformation phase and the second deformation phase respectively.

[0104] The first deformation phase is obtained by triangulating the PS points selected from the target area to be monitored based on the amplitude deviation method, and then using conventional PS-InSAR. Figure 3 , the specific steps are as follows:

[0105] S511: Acquire multiple SAR images of the target area to be monitored;

[0106] S512: Calculate the amplitude deviation value for each SAR image;

[0107] The calculation of amplitude deviation value adopts the classic algorithm in PS point selection algorithm - amplitude deviation method. Under high signal-to-noise ratio conditions, the phase standard deviation σ φ Approximately to the amplitude deviation D A , so the amplitude deviation D can be used A To determine the stability of the target point. The mathematical expression of amplitude deviation is:

[0108] σφ≈σ nR ≈σ nl

[0109]

[0110] Among them, σ nR , σ nl are the standard deviations of the real and imaginary parts of the noise, σ A 、μ A are the standard deviation and mean of the amplitude respectively. φ is the phase standard deviation, D AThe amplitude dispersion method utilizes only the amplitude information of the image. It is computationally simple and highly efficient, but it requires a large amount of data to obtain statistical significance in the time series. Therefore, it is only suitable for PS-InSAR experiments conducted over long periods of time and with extensive SAR data.

[0111] S513: Determine an amplitude deviation threshold suitable for the target area to be monitored, select PS points and unwrapping starting points that meet the conditions to form a path integral network;

[0112] S514: Finally, conventional PS-InSAR is used to obtain the first deformation phase.

[0113] The second deformation phase is obtained by triangulating the PS points selected from the target area to be monitored based on the amplitude deviation and correlation coefficient method, and then using conventional PS-InSAR. Figure 4 , the specific steps are as follows:

[0114] S521: Acquire multiple SAR images of the target area to be monitored;

[0115] S522: Calculate the amplitude deviation value for each SAR image;

[0116] Steps S521 and S522 are the same as S511 and S512.

[0117] S523: Filtering out a first high-quality PS point set according to the amplitude deviation threshold;

[0118] When screening the first high-quality point set, the amplitude deviation threshold needs to be appropriately increased.

[0119] S524: Perform precise registration on the master image and the slave image one by one to obtain correlation coefficients;

[0120] S525: Setting a correlation coefficient threshold, selecting PS points whose target point correlation coefficient sequence remains above the correlation coefficient threshold from the first high-quality PS point set to form a second high-quality PS point set;

[0121] S526: composing the second high-quality PS point set into a PS point network for disentanglement;

[0122] S527: Finally, conventional PS-InSAR is used to obtain the second deformation phase.

[0123] Then, the first deformation phase is corrected using the second deformation phase. Figure 5 As shown:

[0124] S531: extracting deformation points that can be matched in the first deformation phase and the second deformation phase monitoring;

[0125] S532: Calculating the deformation rate difference between each set of matched deformation points;

[0126] S533: Using bilinear interpolation to construct an error correction network for the deformation rate difference;

[0127] The principle of bilinear interpolation is as follows: use linear interpolation to estimate f(x,y), where (x1,y1,f(x1,y1)), (x1,y2,f(x1,y2)), (x2,y1,f(x2,y1)), (x2,y2,f(x2,y2)) are known. To use bilinear interpolation to estimate f(x,y), you need to first interpolate x to find f(x,y1) and f(x,y2), as shown below:

[0128]

[0129] According to the above formula, we can find f(x,y):

[0130]

[0131] If x2=1, x1=0, y2=1, y1=0, then:

[0132] f(x,y)=(1-y)·(1-x)·f(0,0)+(1-y)·x·f(1,0)+y·(1-x)·f(1,0)+y·x·f(1,1)

[0133] S534: According to the error correction network, the deformation values of the PS points obtained based on the amplitude deviation method are corrected one by one to reduce the influence of low-quality PS points on the unwrapping results.

[0134] The first deformation phase after correction is transformed into deformation phase, and the final deformation monitoring result is obtained after geocoding. Figure 6 , specifically including:

[0135] S541: Using the parameters obtained from the pulse echo time delay and the Doppler parameters obtained during the imaging process, the coordinate values in the inertial coordinate system can be obtained through the slant range Doppler positioning equation;

[0136] S542: According to the geographic coordinate conversion method, the corresponding geographic longitude and geographic latitude can be obtained to form the final deformation monitoring result.

[0137] S600: Divide the key landslide risk areas based on the final deformation monitoring results. Figure 7 As shown:

[0138] S601: Obtain PS points whose annual average deformation rate is above a second deformation rate threshold, and record them as a first abnormal deformation PS point set;

[0139] In this embodiment, the second deformation rate threshold is set to 15 mm / year.

[0140] S602: sequentially calculating the deformation rates of the PS points in the first abnormal deformation PS point set within a short period (e.g., half a year), analyzing their changes, extracting points where the deformation rates continuously increase in the short period, and classifying these points into the second abnormal deformation PS point set;

[0141] S603: superimposing the points in the second abnormal deformation PS point set onto the optical satellite image, and drawing the corresponding risk area vector;

[0142] S604: Identify key landslide risk areas based on risk area vectors.

[0143] In this embodiment, a river slope landslide risk monitoring system based on corner reflectors and PS-InSAR is also provided, including:

[0144] A main image acquisition module is used to acquire at least one SAR image of the target area to be monitored and select a main image;

[0145] A first processing module processes the main image based on the PS-InSAR technology to obtain preliminary deformation monitoring results, wherein the preliminary deformation monitoring results include deformation value and deformation rate;

[0146] A first determination module is used to analyze the preliminary deformation monitoring results and determine a stable point adjacent to the target area to be monitored as the first corner reflector placement position;

[0147] The second determination module divides the target area to be monitored into grids, performs real-scene analysis in the divided grids, and determines the placement position of the second corner reflector;

[0148] The second processing module is used to obtain the deformation phase by taking the first corner reflector as the unwrapping starting point, and obtain the final deformation monitoring result after correction;

[0149] Identification module, which divides key landslide risk areas according to the final deformation monitoring results.

[0150] The following is combined with Figures 8 to 16 An embodiment of the present invention is described.

[0151] The study area is located in a landslide area in the Jinsha River Basin. The area is mainly mountainous, with large undulating terrain, large changes in vegetation and tree cover, and the influence of rainfall and geological conditions. The area is prone to many landslides. The monitoring method of this application can be used to monitor large landslides in the basin, effectively avoiding the impact of barrier lakes caused by landslides and protecting the lives and property of people in the basin. The case of the present invention uses medium-to-high-resolution SAR satellite images. The main image information and the slave image information are shown in Table 1.

[0152] Table 1 Collection table of main and secondary images of the Jinsha River Basin

[0153]

[0154]

[0155] Figure 2 and Figure 3 The temporal baseline and spatial baseline distributions of the SAR images selected in the test area and the average intensity changes of the seven best SAR images are displayed. Based on these two conditions, the image acquired on March 13, 2018 was finally selected as the main image.

[0156] Figure 4 The deformation monitoring results of the target slope obtained through preliminary processing based on PS-InSAR technology were displayed. The results showed that there was no obvious deformation near the location of a village at the lower left foot of the slope in the recent period, and it remained stable.

[0157] By analyzing the initial deformation results obtained based on PS-InSAR and grasping the dual principles of distance and stability, the position of corner reflector No. 1 was finally selected as follows: Figure 5 .

[0158] Figure 6 The specific distribution of corner reflectors arranged according to the corner reflector layout principle of the present invention is demonstrated, thereby ensuring that the target slope can be fully monitored in the end.

[0159] There are three main principles for layout: first, there are no effective PS points or very few PS points in the deformation monitoring area; second, the distance between the corner reflector and other PS point clusters or corner reflectors should be moderate; third, the location where the corner reflector is laid should be convenient for construction and the geological conditions should be relatively stable.

[0160] Figure 7 The deformation result distribution map obtained based on the PS-InSAR technology is based on the PS points and corner reflectors obtained under relatively loose selection conditions.

[0161] The deformation rate in this area ranges from -23.58mm / y to 25.15mm / y, with an average deformation rate of -1.57mm / y. There is an obvious deformation area in the local area, but it is temporarily far away from the river channel. Combined with factors such as slope and slope direction, the risk is also high.

[0162] Figure 8 The deformation result distribution map obtained based on PS-InSAR technology is based on the PS points and corner reflectors obtained under relatively strict selection conditions.

[0163] The deformation rate in this area ranges from -23.63 mm / y to 22.60 mm / y, with an average deformation rate of -1.18 mm / y. The distribution diagram shows that high-quality and corner reflectors are basically scattered throughout the target area, which can play a comprehensive control role.

[0164] Figure 9 The deformation monitoring results of the entire area after correction with high-quality deformation monitoring results are displayed. The results show that there are local areas of continuous deformation in the monitoring area, which require special attention.

[0165] Figure 10 For local areas that need to be monitored based on the deformation rate value and the deformation rate change trend, personnel need to be dispatched to carry out governance if necessary.

[0166] The above describes in detail the river slope landslide risk monitoring method based on corner reflectors and PS-InSAR provided by this application. The description of the specific embodiments is only intended to help understand the method and core concept of this application. It should be noted that for those skilled in the art, various improvements and modifications can be made to this application without departing from the principles of this application, and such improvements and modifications also fall within the scope of protection of the claims of this application.

Claims

1. A river slope landslide risk monitoring method based on corner reflector and PS-InSAR technology is characterized by: include: Acquire at least one SAR image of the target area to be monitored and select the main image; Processing the main image based on PS-InSAR technology to obtain preliminary deformation monitoring results, wherein the preliminary deformation monitoring results include deformation value and deformation rate; Analyze the preliminary deformation monitoring results and determine the stable point close to the target area to be monitored as the first corner reflector placement location; Divide the target area to be monitored into grids, perform real-scene analysis in the divided grids, and determine the placement position of the second corner reflector; The first corner reflector is used as the unwrapping starting point to obtain the deformation phase, and the final deformation monitoring result is obtained after correction; Key landslide risk areas are divided based on the final deformation monitoring results.

2. The river slope landslide risk monitoring method according to claim 1, characterized in that: The step of acquiring at least one SAR image of the target area to be monitored and selecting a main image includes: Acquire multiple SAR images of the target area to be monitored at different time points; Focus each SAR image to generate an SLC format image; Generate intensity maps, geocode, and crop target areas of SLC format images to obtain cropped images. The image with the largest average intensity value is selected from the cropped images as the final main image.

3. The river slope landslide risk monitoring method according to claim 2, characterized in that: Before selecting the image with the largest average intensity value from the cropped images as the final main image, the method further includes: Preliminary selection of candidate main images from the cropped images based on the time baseline and the space baseline; Calculate the average intensity of the target area in the candidate main image, where the intensity I satisfies: I = A 2 , A is the image amplitude.

4. The river slope landslide risk monitoring method according to claim 1, characterized in that: The first corner reflector is arranged as follows: Extracting a stable area based on the deformation value in the preliminary deformation monitoring result and a preset first deformation rate threshold; A relatively flat area that is relatively close to the target area to be monitored is selected as the first corner reflector placement point.

5. The river slope landslide risk monitoring method according to claim 1, characterized in that: The method of using the first corner reflector as the unwrapping starting point to obtain the deformation phase and obtaining the final deformation monitoring result after correction includes: The first corner reflector is used as the unwrapping starting point, and a triangulated network is formed by selecting PS points from the target area to be monitored based on the amplitude deviation method. The first deformation phase is obtained using conventional PS-InSAR. The first corner reflector is used as the unwrapping starting point, and a triangulated network is formed by selecting PS points from the target area to be monitored based on the amplitude deviation and correlation coefficient method. The second deformation phase is obtained using conventional PS-InSAR. Correcting the first deformation phase using the second deformation phase; The corrected first deformation phase is transformed into deformation phase, and the final deformation monitoring result is obtained after geocoding.

6. The river slope landslide risk monitoring method according to claim 5, characterized in that: The first deformation phase is obtained specifically according to the following method: Acquire multiple SAR images of the target area to be monitored; Calculate the amplitude deviation value for each SAR image; Determine the amplitude deviation threshold suitable for the target area to be monitored, select the PS points and unwrapping starting points that meet the conditions to form a path integral network; Finally, conventional PS-InSAR is used to obtain the first deformation phase.

7. The river slope landslide risk monitoring method according to claim 6, characterized in that: The second deformation phase is obtained specifically according to the following method: Acquire multiple SAR images of the target area to be monitored; Calculate the amplitude deviation value for each SAR image; Filter out the first high-quality PS point set according to the amplitude deviation threshold; Perform precise registration of the master image and the slave image one by one to obtain the correlation coefficient; Set a correlation coefficient threshold, and select PS points whose target point correlation coefficient sequence remains above the correlation coefficient threshold from the first high-quality PS point set to form a second high-quality PS point set; Combining the second highest quality PS point set into a PS point network for disentanglement; Finally, conventional PS-InSAR is used to obtain the second deformation phase.

8. The river slope landslide risk monitoring method according to claim 7, characterized in that: The correcting the first deformation phase by using the second deformation phase includes: Extracting deformation points that can be matched in the first deformation phase and the second deformation phase monitoring; Calculate the deformation rate difference between each set of matched deformation points; The error correction network is constructed based on the deformation rate difference using bilinear interpolation method. According to the error correction network, the deformation values of the PS points obtained based on the amplitude deviation method are corrected one by one.

9. The river slope landslide risk monitoring method according to claim 8, characterized in that: The phase-converting deformation of the corrected first deformation phase and obtaining the final deformation monitoring result after geocoding specifically includes: Using the parameters obtained from the pulse echo time delay and the Doppler parameters obtained during the imaging process, the coordinate values in the inertial coordinate system are obtained through the slant range Doppler positioning equation; The corresponding geographic longitude and latitude are obtained according to the geographic coordinate conversion method.

10. The river slope landslide risk monitoring method according to claim 1, characterized in that: The key landslide risk areas are divided according to the final deformation monitoring results, including: Obtain PS points whose annual average deformation rate is above the preset second deformation rate threshold, and record them as the first abnormal deformation PS point set; Calculate the deformation rates of the PS points in the first abnormal deformation PS point set in a short period in sequence, extract the points where the deformation rate in the short period continuously increases, and divide these points into the second abnormal deformation PS point set; The points in the second abnormal deformation PS point set are superimposed on the optical satellite image to draw the corresponding risk area vector; Identify key landslide risk areas based on risk area vectors.

Images