GB-SAR atmospheric delay phase correction method based on clustering analysis

Abstract

The invention provides a GB-SAR atmospheric delay phase correction method based on clustering analysis, and the method comprises the steps: collecting a plurality of single-view complex images of a plurality of time phases through GB-SAR equipment, generating a short-baseline interferogram pair, and carrying out the screening of a permanent scatterer point; based on preset large-scale atmospheric phase correction, performing model fitting on the permanent scatterer points to obtain a large-scale estimated value; based on preset small-scale atmospheric phase correction, clustering the permanent scatterer points through spatial dimension clustering and time dimension clustering to obtain a space-time sample cluster, and calculating or complementing the space-time sample cluster to obtain a small-scale estimated value; and adding the large-scale estimation value and the small-scale estimation value to obtain a total atmospheric phase estimation value. According to the method, the large-scale fitting advantage of the PSM method and the phase stacking thought of the CSS method are combined, homogeneous samples are obtained through space and time two-dimensional clustering analysis, and accurate removal of small-scale turbulent atmosphere is achieved.

Description

Technical Field

[0001] This invention relates to the field of radar measurement technology, and in particular to a GB-SAR atmospheric delay phase correction method based on cluster analysis. Background Technology

[0002] Since its development in the late 20th century, Ground-Based Synthetic Aperture Radar (GB-SAR) has become one of the most important tools in the field of surface deformation monitoring. This technology achieves high-precision measurement of minute surface displacements by comparing the phase differences of radar images from different time phases. Compared with spaceborne InSAR, GB-SAR has significant advantages such as flexible deployment, short image acquisition cycles (typically several minutes to tens of minutes), and high measurement accuracy (down to sub-millimeter level). It has been widely used in open-pit mine slope stability monitoring, landslide early warning, and deformation detection of man-made structures such as bridges and dams. For example, in open-pit mine scenarios, GB-SAR can monitor minute slope displacements in real time to prevent collapse accidents; in landslide monitoring, it can provide high-frequency monitoring data to support emergency response decisions; and in man-made structure monitoring, it can detect subtle changes caused by bridge vibrations or dam seepage.

[0003] However, the sub-millimeter measurement accuracy of GB-SAR is severely affected by atmospheric interference. Atmospheric phase delay effects are the main source of error, including large-scale slowly varying atmospheric components (such as uniform changes caused by overall temperature or pressure gradients) and small-scale turbulent atmospheric components (such as rapid fluctuations caused by local water vapor vortices or wind disturbances). These atmospheric errors stem from the spatial and temporal inhomogeneities of atmospheric refractive index, causing phase delays in the radar signal propagation path and superimposing spurious deformations in the interferometric phase. Especially in scenarios with monitoring distances of several kilometers, high humidity areas near rivers with variable environments, or complex weather conditions, these errors can reach several millimeters or even higher, severely limiting the accuracy of practical engineering applications. Domestic and international research has verified the severity of atmospheric influences through controlled deformation accuracy verification experiments; for example, some tests show that the atmosphere can introduce deformation errors of about 2 mm or phase deviations of 1 rad.

[0004] In existing technologies, atmospheric correction methods are mainly divided into external data methods and internal data methods. External data methods rely on parameters such as temperature, pressure, and humidity collected by meteorological stations, simulating atmospheric refractive index changes through linear or exponential models. However, the low spatial and temporal sampling rates make it difficult to meet the requirements of GB-SAR's small-scale, high-precision monitoring. Internal data methods are based on information inherent in the GB-SAR interferometric data itself, such as utilizing the phase information of permanent scatterers (PS) or artificial corner reflectors, and estimating the atmosphere through polynomial model fitting combined with spatial interpolation (hereinafter referred to as the PSM method). The PSM method performs well in removing large-scale, slowly varying atmospheric elements, but it processes each interferogram individually without considering temporal information, making it difficult to remove locally residual small-scale turbulent atmospheric elements. The co-date phase stacking method (hereinafter referred to as the CSS method), commonly used in spaceborne InSAR, removes turbulent components in the temporal dimension, but it assumes linear stability in deformation, making it unsuitable for GB-SAR's common scenarios of sudden slope deformation or artificially controlled deformation experiments, leading to the loss of abrupt change information.

[0005] In summary, the shortcomings of existing technologies are mainly reflected in the following aspects: (1) the external data method has a low sampling rate and strong dependence, which cannot adapt to the real-time requirements of GB-SAR; (2) the PSM method does not completely remove small-scale atmospheric data, and residual errors affect accuracy; (3) although the CSS method considers the time dimension, it ignores deformation nonlinearity and is prone to losing high-frequency abrupt change information; (4) there is a lack of refined methods that integrate the spatial and temporal dimensions, making it impossible to simultaneously achieve atmospheric removal and deformation protection. These problems often lead to false alarms or decreased accuracy in practical engineering, and new methods are urgently needed to solve them. Summary of the Invention

[0006] This invention provides a GB-SAR atmospheric delay phase correction method based on cluster analysis to address the shortcomings of existing technologies. It achieves the precise removal of small-scale turbulent atmosphere by obtaining homogeneous samples through spatial and temporal cluster analysis without relying on external meteorological data, combining the large-scale fitting advantages of the PSM method and the phase stacking concept of the CSS method. At the same time, it effectively protects the abrupt deformation information in the high-frequency acquisition of GB-SAR.

[0007] In a first aspect, the present invention provides a GB-SAR atmospheric delay phase correction method based on cluster analysis, comprising: Multiple temporal single-view complex images were acquired using GB-SAR equipment, short baseline interferogram pairs were generated, and permanent scatterer points were obtained by screening. Based on a preset large-scale atmospheric phase correction, the permanent scatterer points are model-fitted to obtain large-scale estimates. Based on the preset small-scale atmospheric phase correction, the permanent scatterer points are clustered by spatial and temporal clustering to obtain spatiotemporal sample clusters. The spatiotemporal sample clusters are then calculated or completed to obtain small-scale estimates. The large-scale estimate and the small-scale estimate are added together to obtain the total atmospheric phase estimate.

[0008] According to the present invention, a GB-SAR atmospheric delay phase correction method based on cluster analysis is provided, which utilizes GB-SAR equipment to acquire multiple temporal single-look complex images, generates short baseline interferogram pairs, and filters out permanent scatterer points, including: Based on the device's preset scanning mode, short baseline interferogram pairs are generated from single-view complex images of multiple time phases; Permanent scattering points were obtained by screening using amplitude deviation index and coherence threshold; The number of integer cycles of the wrapped phase is determined by unwrapping the short baseline interferogram pair, and the unwrapped interferometric phase sequence is output.

[0009] According to the present invention, a GB-SAR atmospheric delay phase correction method based on cluster analysis is provided. Based on a preset large-scale atmospheric phase correction, a model is fitted to the permanent scatterer points to obtain large-scale estimates, including: The distance between any permanent scatterer point and the radar is obtained. The four-dimensional modeling polynomial function of the permanent scatterer point is obtained from the three-dimensional coordinates of any permanent scatterer point, the distance between any permanent scatterer point and the radar, and the atmospheric property function at any time. Using the coordinates of all permanent scatterer points and the unwrapped interferometric phase sequence, the optimal weight parameters of the four-dimensional modeling polynomial function are solved by the least squares method to obtain atmospheric phase estimates at any two time points.

[0010] According to the present invention, a GB-SAR atmospheric delay phase correction method based on cluster analysis is provided, which clusters the permanent scatterer points through spatial dimensional clustering, including: Using any central time as a reference, determine 2N symmetrical short baseline phases as the first cluster sample; The clustering range is determined, and the KD tree method is used to search for the nearest permanent scatterer points around the current permanent scatterer point within the clustering range to form the first spatiotemporal sample cluster with similar phase sequences in a short time range; If the number of samples in the first spatiotemporal sample cluster is greater than the first preset condition, then it is determined that there are similar phase signals in space.

[0011] According to the present invention, a GB-SAR atmospheric delay phase correction method based on cluster analysis is provided, which clusters the permanent scatterer points through time-dimensional clustering, including: Using the phase of the permanent scatterer points in any interference pair of the spatial dimension of the first spatiotemporal sample cluster as the clustering data, 2N samples are formed, constituting the second cluster sample; The atmospheric phase obtained by stacking the first spatiotemporal sample cluster through conventional phase stacking is used as the cluster center of the second spatiotemporal sample cluster. This results in a second spatiotemporal sample cluster composed of atmospheric components similar to those in the first cluster, which is homogeneous with the current space. If the number of samples in the second spatiotemporal sample cluster is not empty, then the average stationarity condition is satisfied. Phase stacking is calculated using samples from the first and second spatiotemporal sample clusters to obtain the average estimate of atmospheric composition at any central moment; The small-scale atmospheric estimate in the interferogram is obtained by taking the difference between the average estimates of atmospheric composition at any two center times.

[0012] According to the present invention, a GB-SAR atmospheric delay phase correction method based on cluster analysis is provided, which calculates or completes the spatiotemporal sample clusters to obtain small-scale estimates, including: An iterative K-Means binary classification method is used to classify the spatiotemporal sample clusters based on a given morphological coefficient threshold. If the number of clusters is determined to be less than the given morphological coefficient threshold, then phase stacking is used to calculate the optimal number of clusters. If the first or second spatiotemporal sample cluster is insufficient or does not meet the calculation conditions, then cubic interpolation is performed using the surrounding calculated permanent scattering points to complete the calculation and output a small-scale estimate.

[0013] According to the present invention, a GB-SAR atmospheric delay phase correction method based on cluster analysis is provided, which adds the large-scale estimate and the small-scale estimate to obtain a total atmospheric phase estimate, including: The large-scale estimate and the small-scale estimate are added together and then phase winding is performed to obtain the total atmospheric phase estimate.

[0014] Secondly, the present invention also provides a GB-SAR atmospheric delay phase correction system based on cluster analysis, comprising: The filtering module is used to acquire single-view complex images of multiple time phases using GB-SAR equipment, generate short baseline interferogram pairs, and filter out permanent scatterer points. The first correction module is used to perform model fitting on the permanent scatterer points based on a preset large-scale atmospheric phase correction to obtain large-scale estimated values. The second correction module is used to cluster the permanent scatterer points based on the preset small-scale atmospheric phase correction, obtain spatiotemporal sample clusters by spatial and temporal clusters, and calculate or complete the spatiotemporal sample clusters to obtain small-scale estimates. The synthesis module is used to add the large-scale estimate and the small-scale estimate to obtain the total atmospheric phase estimate.

[0015] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the GB-SAR atmospheric delay phase correction method based on cluster analysis as described above.

[0016] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the GB-SAR atmospheric delay phase correction method based on cluster analysis as described above.

[0017] The GB-SAR atmospheric delay phase correction method based on cluster analysis provided by this invention has the following beneficial effects: (1) Effective removal of small-scale turbulent atmosphere: spatial clustering utilizes local clustering to increase the sample size; temporal clustering screens similar atmospheric components to avoid smoothing abrupt signal changes, thereby achieving fine correction of residual small-scale atmosphere and improving GB-SAR deformation monitoring accuracy to the sub-millimeter level.

[0018] (2) Protecting mutation information in high-frequency observations: Unlike the linear assumption of the CSS method, this method preserves sudden deformation or human-controlled experimental signals in time-dimensional clustering, which is suitable for emergency monitoring scenarios.

[0019] (3) No external data required: Independent correction is achieved based on the internal interference data itself, which is suitable for complex environments. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is one of the flowcharts of the GB-SAR atmospheric delay phase correction method based on cluster analysis provided by the present invention; Figure 2 This is the second flowchart of the GB-SAR atmospheric delay phase correction method based on cluster analysis provided by the present invention; Figure 3 These are comparison images of interferograms after atmospheric correction provided by this invention; Figure 4 This is a comparison diagram of the residual phase gradient of the interferogram provided by the present invention; Figure 5 This is a comparison diagram of the artificial deformation sequence of the corner reflector provided by the present invention; Figure 6 This is a schematic diagram of the GB-SAR atmospheric delay phase correction system based on cluster analysis provided by the present invention; Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0023] Figure 1 This is one of the flowcharts illustrating the GB-SAR atmospheric delay phase correction method based on cluster analysis provided in this embodiment of the invention, such as... Figure 1 As shown, it includes: Step 100: Use GB-SAR equipment to acquire single-view complex images of multiple time phases, generate short baseline interferogram pairs, and screen to obtain permanent scatterer points; Step 200: Based on the preset large-scale atmospheric phase correction, perform model fitting on the permanent scatterer points to obtain large-scale estimates; Step 300: Based on the preset small-scale atmospheric phase correction, the permanent scatterer points are clustered by spatial and temporal clustering to obtain spatiotemporal sample clusters. The spatiotemporal sample clusters are calculated or completed to obtain small-scale estimates. Step 400: Add the large-scale estimate and the small-scale estimate to obtain the total atmospheric phase estimate.

[0024] like Figure 2 As shown, the overall processing flow of this embodiment of the invention includes data acquisition, PSM correction, spatial clustering, temporal clustering, and final estimation, specifically including the following steps: GB-SAR Interferometric Data Acquisition and PS Point Preprocessing: First, multiple temporal single-view complex images (SLC) are acquired using a GB-SAR device. Short baseline interferogram pairs are generated based on the device's scanning mode (angle scan or line scan). Amplitude deviation index (DA) thresholding is then applied. Coherence (COH) threshold High-quality target area (PS) points are selected using a dual-threshold screening process; these PS points typically correspond to stable ground features or artificial corner reflectors. Subsequently, the interferogram is unwrapped to determine the integer number of cycles of the wrapped phase, ensuring phase continuity. This step outputs the unwrapped interferometric phase sequence, providing fundamental data for subsequent correction.

[0025] Large-scale atmospheric phase correction: for wavelengths of The GB-SAR transmitted signal, assuming the atmospheric composition is uniform and constant, then at two certain moments of GB-SAR ( In the interferogram formed by the SLC image of ), any PS point ( The large-scale atmospheric phase obtained by integrating the radar signal along the path in the atmosphere can be simplified as follows: (1) in, , A function representing the atmospheric properties at a given moment, determined by parameters related to temperature, pressure, and humidity. For the first The distance between each PS point and the radar. In reality, the atmosphere is not homogeneous, and it is usually modeled as a polynomial function of the multiple coordinates of the PS point using a four-dimensional model: (2) in, , and The first The three-dimensional coordinates of all PS points are used. The optimal four-dimensional coordinates are then determined using the least squares method with the coordinates of all PS points and the unwrapped interference phase. The parameters will give you the time. Estimated atmospheric phase.

[0026] Small-scale atmospheric phase correction: For residual small-scale turbulent atmosphere, the idea of ​​co-date phase stacking is introduced and optimized through clustering in both spatial and temporal dimensions.

[0027] Spatial Dimension Clustering: First, construct clustering samples. For a certain central moment In other words, symmetrically, the total before and after that moment Short baseline phases are used as clustering data: (3) Given clustering range The KD-tree method is used to quickly search for the current PS point. The nearest surrounding A PS point is formed. One sample. After clustering, clusters are obtained where the current PS point has a similar phase sequence within a short time range. When the number of samples within a cluster If the conditions are greater than a certain threshold, it is considered that there are similar phase signals in space.

[0028] Temporal clustering: Traditional methods require stable or linear deformation to obtain reliable estimates through averaging. However, when used in short-term sudden deformation or controlled deformation experiments, they can smooth out abrupt signal changes. Therefore, it is necessary to perform temporal re-clustering on the short baseline moments involved in averaging to obtain moments with similar atmospheric composition. On the contrary, Using the phase of a certain interference pair as clustering data, a spatially dimensional PS sample is formed. 1 sample, construct cluster samples .by This interference is an example: (4) It is important to note that in clustering At that time, clusters can be The atmospheric phase obtained by conventional phase stacking is taken as the initial estimate. The cluster centers are determined. Clusters of homogeneity in the current spatial region are obtained. Clusters formed by moments with similar internal atmospheric composition When the number of samples When the value is not empty, it can be considered to meet the stationarity condition required for averaging. Using , By calculating the phase stack of the samples within the range, the atmospheric composition at the center time can be estimated by averaging as follows: (5) The estimated value of the small-scale atmosphere in the interferogram is... It should be noted that, in choosing the clustering method, this invention employs an iterative K-Means binary classification method. Given a morphological coefficient threshold, samples are continuously divided into two clusters. The process iterates until the morphological coefficient is less than the threshold, thereby automatically determining the optimal number of clusters. Furthermore, if the PS point is due to… or When insufficient data results in failure to meet the calculation conditions, cubic interpolation of the surrounding calculated PS points is used to obtain complete and continuous atmospheric results.

[0029] Final atmospheric phase estimate: The large-scale and small-scale estimates are added together to obtain the total atmospheric phase estimate. (6) in This indicates a phase wrapping operation; removing the estimated total atmospheric phase from the original wrapped interferogram yields a clean interferogram containing only the deformation and noise phases. The results are optimized by iteratively estimating both large-scale and small-scale phases.

[0030] Based on the above embodiments, the present invention will illustrate the technical solution with specific examples.

[0031] (1) On-site data acquisition: A Ku-band true aperture GB-SAR instrument was deployed in the observation shed on the right bank of a large-scale inter-basin water transfer project to observe the left bank slope. On November 27, 2024, 104 SLC images were acquired at a frequency of 14 min / scene to simulate normal monitoring; on November 30, 2024, the frequency was increased to 1 min / scene to acquire 161 SLC images, and a human-controlled deformation experiment was conducted: the adjustable corner reflector deployed on the left bank slope was manually adjusted 10 times at 10-minute intervals, and each time it was moved precisely by a certain number of mm through a three-dimensional sliding rail, and returned to its original position after the experiment.

[0032] (2) Atmospheric correction comparison: Atmospheric correction was performed using both conventional methods and the method of this invention, and the clustering range was set. Sample size The comparison results of the interferograms are as follows: Figure 3 As shown, a comparison is presented between the original interferogram, the results of the PSM method, the results of the CSS method, and the results of the method of the present invention. It can be seen that the method of the present invention effectively removes the remaining atmospheric components from the PSM method, resulting in a stable residual phase while preserving spatial abrupt changes. Figure 4 As shown, the phase gradient statistics processed by each method are illustrated. The average gradient calculation results for all interferogram pairs show that the gradients of the original interferogram, the PSM method result, the CSS method result, and the result of the method of this invention are 0.15, 0.13, 0.06, and 0.11, respectively. This indicates that the phase gradient of this method is between that of the PSM method and the CSS method, and based on the polynomial model fitting, the remaining atmospheric components are further reduced by about 15%. The CSS method has a lower gradient due to its excessive smoothness. The phase gradient results of this method show that, based on the polynomial model fitting, the remaining small-scale atmospheric components are further reduced.

[0033] (3) Time series comparison: The time series deformation results obtained in the experiment of artificially adjusting the corner reflector are as follows Figure 5 As shown, the root mean square error (RMSE) of the deformation obtained by the original interferogram, the PSM method result, the CSS method result, and the method of the present invention, compared with the true recorded value, are 0.22, 0.16, 0.25, and 0.16, respectively. This indicates that compared with the original uncorrected result, the accuracy of the method of the present invention is improved, and it can avoid the phenomenon that the deformation sequence of the conventional phase stacking method is smoothed, and the accuracy calculation result is actually lower than that of the uncorrected result.

[0034] The above embodiments demonstrate the effectiveness of the method of the present invention in removing small-scale turbulent atmospheric components in GB-SAR measurements.

[0035] The GB-SAR atmospheric delay phase correction system based on cluster analysis provided by this invention is described below. The GB-SAR atmospheric delay phase correction system based on cluster analysis described below can be referred to in correspondence with the GB-SAR atmospheric delay phase correction method based on cluster analysis described above.

[0036] Figure 6 This is a schematic diagram of the GB-SAR atmospheric delay phase correction system based on cluster analysis provided in an embodiment of the present invention, as shown below. Figure 6 As shown, it includes: a screening module 61, a first correction module 62, a second correction module 63, and a comprehensive module 64, wherein: The filtering module 61 is used to acquire multiple temporal single-view complex images using GB-SAR equipment, generate short baseline interferogram pairs, and filter out permanent scatterer points; the first correction module 62 is used to perform model fitting on the permanent scatterer points based on a preset large-scale atmospheric phase correction to obtain large-scale estimates; the second correction module 63 is used to cluster the permanent scatterer points through spatial and temporal clustering based on a preset small-scale atmospheric phase correction to obtain spatiotemporal sample clusters, and calculate or complete the spatiotemporal sample clusters to obtain small-scale estimates; the synthesis module 64 is used to add the large-scale estimates and the small-scale estimates to obtain the total atmospheric phase estimate.

[0037] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include a processor 710, a communication interface 720, a memory 730, and a communication bus 740. The processor 710, communication interface 720, and memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute a GB-SAR atmospheric delay phase correction method based on clustering analysis. This method includes: acquiring multiple temporal single-view complex images using a GB-SAR device, generating short baseline interferogram pairs, and filtering to obtain permanent scatterer points; performing model fitting on the permanent scatterer points based on a preset large-scale atmospheric phase correction to obtain large-scale estimates; clustering the permanent scatterer points using spatial and temporal clustering based on a preset small-scale atmospheric phase correction to obtain spatiotemporal sample clusters, calculating or completing the spatiotemporal sample clusters to obtain small-scale estimates; and adding the large-scale estimates and the small-scale estimates to obtain a total atmospheric phase estimate.

[0038] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0039] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the GB-SAR atmospheric delay phase correction method based on cluster analysis provided by the above methods. The method includes: acquiring multiple temporal single-view complex images using a GB-SAR device, generating short baseline interferogram pairs, and filtering to obtain permanent scatterer points; performing model fitting on the permanent scatterer points based on a preset large-scale atmospheric phase correction to obtain large-scale estimates; clustering the permanent scatterer points into spatiotemporal sample clusters through spatial and temporal clustering based on a preset small-scale atmospheric phase correction to obtain spatiotemporal sample clusters, calculating or completing the spatiotemporal sample clusters to obtain small-scale estimates; and adding the large-scale estimates and the small-scale estimates to obtain a total atmospheric phase estimate.

[0040] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0041] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0042] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A GB-SAR atmospheric delay phase correction method based on cluster analysis, characterized in that, include: Multiple temporal single-view complex images were acquired using GB-SAR equipment, short baseline interferogram pairs were generated, and permanent scatterer points were obtained by screening. Based on a preset large-scale atmospheric phase correction, the permanent scatterer points are model-fitted to obtain large-scale estimates. Based on the preset small-scale atmospheric phase correction, the permanent scatterer points are clustered by spatial and temporal clustering to obtain spatiotemporal sample clusters. The spatiotemporal sample clusters are then calculated or completed to obtain small-scale estimates. The large-scale estimate and the small-scale estimate are added together to obtain the total atmospheric phase estimate.

2. The GB-SAR atmospheric delay phase correction method based on cluster analysis according to claim 1, characterized in that, Multiple temporal single-look complex images were acquired using GB-SAR equipment, short baseline interferogram pairs were generated, and permanent scatterer points were selected, including: Based on the device's preset scanning mode, short baseline interferogram pairs are generated from single-view complex images of multiple time phases; Permanent scattering points were obtained by screening using amplitude deviation index and coherence threshold; The number of integer cycles of the wrapped phase is determined by unwrapping the short baseline interferogram pair, and the unwrapped interferometric phase sequence is output.

3. The GB-SAR atmospheric delay phase correction method based on cluster analysis according to claim 1, characterized in that, Based on a preset large-scale atmospheric phase correction, the permanent scatterer points are model-fitted to obtain large-scale estimates, including: The distance between any permanent scatterer point and the radar is obtained. The four-dimensional modeling polynomial function of the permanent scatterer point is obtained from the three-dimensional coordinates of any permanent scatterer point, the distance between any permanent scatterer point and the radar, and the atmospheric property function at any time. Using the coordinates of all permanent scatterer points and the unwrapped interferometric phase sequence, the optimal weight parameters of the four-dimensional modeling polynomial function are solved by the least squares method to obtain atmospheric phase estimates at any two time points.

4. The GB-SAR atmospheric delay phase correction method based on cluster analysis according to claim 1, characterized in that, Clustering of the permanent scatterer points using spatial dimension clustering includes: Using any central time as a reference, determine 2N symmetrical short baseline phases as the first cluster sample; The clustering range is determined, and the KD tree method is used to search for the nearest permanent scatterer points around the current permanent scatterer point within the clustering range to form the first spatiotemporal sample cluster with similar phase sequences in a short time range; If the number of samples in the first spatiotemporal sample cluster is greater than the first preset condition, then it is determined that there are similar phase signals in space.

5. The GB-SAR atmospheric delay phase correction method based on cluster analysis according to claim 4, characterized in that, Clustering of the permanent scatterer points using time-dimensional clustering includes: Using the phase of the permanent scatterer points in any interference pair of the spatial dimension of the first spatiotemporal sample cluster as the clustering data, 2N samples are formed, constituting the second cluster sample; The atmospheric phase obtained by stacking the first spatiotemporal sample cluster through conventional phase stacking is used as the cluster center of the second spatiotemporal sample cluster. This results in a second spatiotemporal sample cluster composed of atmospheric components similar to those in the first cluster, which is homogeneous with the current space. If the number of samples in the second spatiotemporal sample cluster is not empty, then the average stationarity condition is satisfied. Phase stacking is calculated using samples from the first and second spatiotemporal sample clusters to obtain the average estimate of atmospheric composition at any central moment; The small-scale atmospheric estimate in the interferogram is obtained by taking the difference between the average estimates of atmospheric composition at any two center times.

6. The GB-SAR atmospheric delay phase correction method based on cluster analysis according to claim 5, characterized in that, The spatiotemporal sample cluster is calculated or completed to obtain small-scale estimates, including: An iterative K-Means binary classification method is used to classify the spatiotemporal sample clusters based on a given morphological coefficient threshold. If the number of clusters is determined to be less than the given morphological coefficient threshold, then phase stacking is used to calculate the optimal number of clusters. If the first or second spatiotemporal sample cluster is insufficient or does not meet the calculation conditions, then cubic interpolation is performed using the surrounding calculated permanent scattering points to complete the calculation and output a small-scale estimate.

7. The GB-SAR atmospheric delay phase correction method based on cluster analysis according to claim 1, characterized in that, The large-scale estimate and the small-scale estimate are added together to obtain the total atmospheric phase estimate, which includes: The large-scale estimate and the small-scale estimate are added together and then phase winding is performed to obtain the total atmospheric phase estimate.

8. A GB-SAR atmospheric delay phase correction system based on cluster analysis, characterized in that, include: The filtering module is used to acquire single-view complex images of multiple time phases using GB-SAR equipment, generate short baseline interferogram pairs, and filter out permanent scatterer points. The first correction module is used to perform model fitting on the permanent scatterer points based on a preset large-scale atmospheric phase correction to obtain large-scale estimated values. The second correction module is used to cluster the permanent scatterer points based on the preset small-scale atmospheric phase correction, obtain spatiotemporal sample clusters by spatial and temporal clusters, and calculate or complete the spatiotemporal sample clusters to obtain small-scale estimates. The synthesis module is used to add the large-scale estimate and the small-scale estimate to obtain the total atmospheric phase estimate.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the GB-SAR atmospheric delay phase correction method based on cluster analysis as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the GB-SAR atmospheric delay phase correction method based on cluster analysis as described in any one of claims 1 to 7.

Images