Ground-based interferometric radar inhomogeneous atmosphere correction method, device, medium and product

Abstract

The application discloses a foundation interferometric radar non-homogeneous atmosphere correction method, equipment, medium and product, relates to the microwave remote sensing measurement technical field, and the method comprises the following steps: determining a non-deformation area in a differential interferogram of a monitoring area; calculating an atmospheric coefficient of each high-coherent point in the non-deformation area; classifying the high-coherent points in the non-deformation area according to the atmospheric coefficient, and respectively calculating an atmospheric delay coefficient of each category; respectively determining a distance function model of each category according to the atmospheric delay coefficient of each category; respectively calculating an atmospheric delay phase of each high-coherent point in each category according to the distance function model of each category; removing the atmospheric delay phase of each high-coherent point from the differential interferogram to obtain an atmospheric correction result. The application clusters the high-coherent points through the atmospheric coefficient of the high-coherent points, calculates the atmospheric delay phase according to the category, and avoids the influence of the non-homogeneous atmosphere.

Description

Technical Field

[0001] The present application relates to the field of microwave remote sensing measurement technology, and in particular to a method, device, medium and product for non-homogeneous atmosphere correction of ground-based interferometric radar. Background Art

[0002] Ground-based Interferometric Synthetic Aperture Radar (GB-InSAR), a technology developed over the past two decades, enables continuous, high-precision monitoring of large areas around the clock and in all weather conditions. While demonstrating significant advantages and value in many fields, GB-InSAR's accuracy remains hampered by numerous challenges, particularly atmospheric effects. Therefore, atmospheric delay correction is essential.

[0003] Domestic and foreign researchers have conducted extensive research on atmospheric delay correction. There are two general atmospheric effects on the monitoring area: homogeneous atmosphere and non-homogeneous atmosphere. A homogeneous atmosphere means that at a certain moment, the meteorological parameters are the same at all locations in the monitoring area, that is, the atmospheric refractive index is the same at all locations. The atmospheric delay phase of the target point increases or decreases constantly with increasing distance. A non-homogeneous atmosphere means that two or more atmospheric effects appear in the monitoring area, and the meteorological parameters at all locations in the monitoring area are no longer the same. At this time, the atmospheric refractive index is no longer a unique value, and the atmospheric homogeneity is no longer satisfied. The atmospheric delay effect on ground-based interferometric radar is mainly caused by changes in meteorological parameters (temperature, relative humidity, and atmospheric pressure) in the monitoring scene. However, existing atmospheric correction methods work well in areas that meet the homogeneous atmosphere assumption. As GB-InSAR application scenarios become more and more extensive, the impact of atmospheric delay is becoming more and more complex, and the assumption of atmospheric homogeneity cannot be met in some scenarios. Summary of the Invention

[0004] The purpose of this application is to provide a method, equipment, medium and product for non-homogeneous atmosphere correction of ground-based interferometric radar, which takes into account the complex situation of non-homogeneous atmosphere in the monitoring area, divides the non-homogeneous atmospheric area into several different homogeneous atmospheric areas, and effectively improves the atmospheric correction accuracy of GB-InSAR in non-homogeneous atmospheric areas.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for correcting non-homogeneous atmosphere of a ground-based interferometric radar, the method comprising:

[0007] Obtaining a differential interferogram of the monitoring area;

[0008] determining a high coherence point in the differential interferogram;

[0009] determining a non-deformed region in the differential interferogram in combination with the high coherence point;

[0010] Calculating an atmospheric coefficient of each of the high coherence points in the non-deformed area; the atmospheric coefficient is calculated based on the atmospheric delay phase and radar distance of the high coherence point; the radar distance is the distance between the high coherence point and the radar device;

[0011] Classifying the high coherence points in the non-deformed area according to the atmospheric coefficient to obtain a plurality of categories, and allocating the high coherence points in the deformed area in the differential interferogram to the corresponding category nearby;

[0012] Calculating the atmospheric delay coefficient of each category according to the atmospheric delay phase and radar distance of the high coherence point in each category;

[0013] Determine a distance function model for each category according to the atmospheric delay coefficient of each category;

[0014] Calculating the atmospheric delay phase of each of the high coherence points in each of the categories according to the distance function model of each of the categories;

[0015] The atmospheric delay phase of each of the high coherence points is removed from the differential interferogram to obtain an atmospheric correction result.

[0016] In a second aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any one of the above-described methods for non-homogeneous atmosphere correction of ground-based interferometric radar.

[0017] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any one of the above-described methods for non-homogeneous atmosphere correction of ground-based interferometric radar.

[0018] In a fourth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements any one of the above-mentioned methods for non-homogeneous atmosphere correction of ground-based interferometric radar.

[0019] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0020] The present application provides a method, device, medium and product for non-homogeneous atmosphere correction of a ground-based interferometric radar, the method comprising: obtaining a differential interferogram of a monitoring area; determining high coherence points in the differential interferogram; determining a non-deformed area in the differential interferogram in combination with the high coherence points; calculating an atmospheric coefficient for each high coherence point in the non-deformed area; the atmospheric coefficient being calculated by the atmospheric delay phase and radar distance of the high coherence point; the radar distance being the distance between the high coherence point and the radar device; classifying the high coherence points in the non-deformed area according to the atmospheric coefficient to obtain a plurality of categories, and allocating the high coherence points in the deformed area of ​​the differential interferogram to the corresponding category. In the present application, the atmospheric coefficients of the high coherence points in the distance direction are calculated to classify the high coherence points, and the high coherence points affected by similar atmospheric effects are classified into one category, that is, the non-homogeneous atmospheric region is divided into several different homogeneous atmospheric regions; then, the atmospheric delay coefficient of each category is calculated according to the atmospheric delay phase and radar distance of the high coherence points in each category; the distance function model of each category is determined according to the atmospheric delay coefficient of each category; the atmospheric delay phase of each high coherence point in each category is calculated according to the distance function model of each category; the atmospheric delay phase of each high coherence point in each category is removed from the differential interferogram to obtain the atmospheric correction result. The present application takes into account the complex situation of non-homogeneous atmosphere in the monitoring area. By calculating the atmospheric coefficients of the high coherence points in the distance direction, the atmospheric coefficients can reflect the degree to which the high coherence points are affected by the atmosphere. Then, the high coherence points are classified, and the high coherence points with similar atmospheric effects are classified into one category, that is, the non-homogeneous atmospheric region is divided into several different homogeneous atmospheric regions. Then, different categories use the distance function model for atmospheric correction. The atmospheric correction accuracy of GB-InSAR in non-homogeneous atmospheric regions is effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 This is a diagram showing the application environment of a non-homogeneous atmosphere correction method for ground-based interferometric radar in one embodiment of the present application;

[0023] Figure 2 A schematic flow chart of a method for correcting non-homogeneous atmosphere in a ground-based interferometric radar according to an embodiment of the present application;

[0024] Figure 3A schematic flow chart of a method for correcting non-homogeneous atmosphere in a ground-based interferometric radar according to another embodiment of the present application;

[0025] Figure 4 A block diagram of a non-homogeneous atmosphere correction system for ground-based interferometric radar provided in one embodiment of the present application;

[0026] Figure 5 A schematic diagram of a radar device scanning result provided in one embodiment of the present application;

[0027] Figure 6 A schematic diagram of the distribution of highly coherent points provided in one embodiment of the present application;

[0028] Figure 7 A schematic diagram of phase unwrapping of selected high coherence points provided in one embodiment of the present application;

[0029] Figure 8 A schematic diagram of performing statistics on all highly coherent points in the form of a histogram provided in one embodiment of the present application;

[0030] Figure 9 A schematic diagram of classifying highly coherent points using the K-means clustering algorithm provided in one embodiment of the present application;

[0031] Figure 10 A schematic diagram of atmospheric delay simulated according to the solved atmospheric delay coefficient provided in one embodiment of the present application;

[0032] Figure 11 An interferogram corrected according to the solved atmospheric delay coefficient is provided in one embodiment of the present application.

[0033] Figure 12 A schematic diagram of atmospheric delay simulated by a traditional elevation function model provided in one embodiment of the present application;

[0034] Figure 13 The interference diagram after correction of the traditional elevation function model provided in one embodiment of the present application;

[0035] Figure 14 A schematic diagram of atmospheric delay simulated by a three-dimensional atmospheric delay model provided in one embodiment of the present application;

[0036] Figure 15 The interference diagram after correction of the three-dimensional atmospheric delay model provided in one embodiment of the present application;

[0037] Figure 16 A schematic diagram of the residuals after correction of the distance-elevation function model provided in one embodiment of the present application;

[0038] Figure 17 A schematic diagram of the residual after correction of the three-dimensional atmospheric delay correction model provided in one embodiment of the present application;

[0039] Figure 18 A schematic diagram of the residual after correction by the method proposed in this application is provided in one embodiment of this application. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0041] Some scholars have proposed an atmospheric correction method based on meteorological data. et al. collected temperature, humidity, and air pressure data for the study area and simulated a distance-dependent atmospheric phase model based on a functional model of the interferometric phase and refractive index. This method requires manual meteorological data collection, and the accuracy of atmospheric correction is affected by the acquisition location. In response, some researchers have proposed an atmospheric correction method based on a functional model. This method selects highly coherent points as known points, establishes a functional model using the phase of each highly coherent point and its distance from the coordinate origin, and uses the least squares method to solve for the atmospheric delay coefficient, thereby obtaining the atmospheric delay phase for all points. Luzi et al. used known stable points as PS points and established a functional model to simulate the atmospheric phase, achieving relatively ideal correction results. However, these two methods are effective in areas where the homogeneous atmosphere assumption is met. As GB-InSAR applications expand, the effects of atmospheric delay are becoming increasingly complex, and the assumption of atmospheric homogeneity cannot be met in some scenarios. Therefore, this application proposes clustering highly coherent points by calculating the atmospheric coefficients related to the range dimension of highly coherent points. This clustering allows highly coherent points affected by similar atmospheric effects to be grouped together, essentially dividing the non-homogeneous atmospheric region into several distinct homogeneous atmospheric regions, thereby removing atmospheric effects from the entire monitoring area. The purpose of this application is to provide an atmospheric correction method and system for ground-based interferometric radar taking into account the non-homogeneous atmospheric delay, so as to improve the correction accuracy of atmospheric delay.

[0042] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0043] The non-homogeneous atmosphere correction method for ground-based interferometric radar provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the differential interference map of the monitoring area to be processed to the server 104. After the server 104 receives the differential interference map of the monitoring area to be processed, it determines the high coherence points in the differential interference map; determines the non-deformation area in the differential interference map in combination with the high coherence points; calculates the atmospheric coefficient of each high coherence point in the non-deformation area; the atmospheric coefficient is calculated by the atmospheric delay phase and radar distance of the high coherence point; the radar distance is the distance between the high coherence point and the radar equipment; the high coherence points in the non-deformation area are classified according to the atmospheric coefficient , obtaining several categories, and assigning the high coherence points in the deformation area in the differential interferogram to the corresponding categories nearby; calculating the atmospheric delay coefficient of each category according to the atmospheric delay phase and radar distance of the high coherence points in each category; determining the distance function model of each category according to the atmospheric delay coefficient of each category; calculating the atmospheric delay phase of each high coherence point in each category according to the distance function model of each category; removing the atmospheric delay phase of each high coherence point from the differential interferogram to obtain the atmospheric correction result. The server 104 can feed back the obtained atmospheric correction result to the terminal 102. In addition, in some embodiments, the ground-based interferometric radar non-homogeneous atmosphere correction method can also be implemented separately by the server 104 or the terminal 102, such as the terminal 102 can directly process the differential interferogram of the monitoring area to be processed, or the server 104 can obtain the differential interferogram of the monitoring area to be processed from the data storage system and process it.

[0044] Terminal 102 may include, but is not limited to, various desktop computers, laptops, smartphones, tablet computers, IoT devices, and portable wearable devices. IoT devices may include smart speakers, smart TVs, smart air conditioners, and smart car devices. Portable wearable devices may include smart watches, smart bracelets, and head-mounted devices. Server 104 may be implemented as a standalone server or a server cluster consisting of multiple servers, or may be a cloud server.

[0045] In an exemplary embodiment, Figure 2 As shown, a method for non-homogeneous atmosphere correction of ground-based interferometric radar is provided. The method is executed by a computer device, specifically a computer device such as a terminal or a server, or a terminal and a server. In the embodiment of the present application, the method is applied to Figure 1The server 104 in the example is used to illustrate the process, including the following steps S1 to S9.

[0046] S1. Obtain the differential interferogram of the monitoring area.

[0047] The differential interferogram is obtained by performing differential interferometry on the original single-view complex images in the time series of ground-based synthetic aperture radar images.

[0048] S2. Determine a high coherence point in the differential interference pattern.

[0049] A high coherence point in the differential interference pattern is selected, where the high coherence point is selected based on a preset amplitude deviation and a coherence coefficient threshold.

[0050] In this embodiment, the amplitude dispersion and coherence coefficient of each point in the differential interferogram are calculated, and points that simultaneously satisfy the conditions of amplitude dispersion less than a preset threshold and coherence coefficient greater than a preset threshold are selected as high-coherence points. The preset thresholds for amplitude dispersion and coherence coefficient are 0.15 and 0.95, respectively.

[0051] S3. Determine a non-deformed area in the differential interference pattern based on the high coherence points.

[0052] Phase unwrapping is performed on the high coherence points, and the high coherence points in the deformation area are identified. The deformation area is identified by using the unwrapped interference pattern or by quickly processing all images using stacking technology. The area other than the deformation area in the interference pattern is determined as the non-deformation area.

[0053] S4. Calculate the atmospheric coefficient of each high coherence point in the non-deformed area; the atmospheric coefficient is calculated based on the atmospheric delay phase and radar distance of the high coherence point; the radar distance is the distance between the high coherence point and the radar device.

[0054] The atmospheric coefficient of each non-deformed high coherence point can be obtained by the atmospheric delay phase (the default high coherence phase value of the pre-processed non-deformed area is the atmospheric delay phase and the noise phase. The noise phase component is very small, so it is mainly the atmospheric delay phase by default) and the distance between it and the radar device according to formula a i =φ i / r i Find; where a i is the atmospheric coefficient of the desired high coherence point i, φ i is the atmospheric delay phase of high coherence point i, r i is the distance from the high coherence point i to the radar equipment.

[0055] S5. Classify the high coherence points in the non-deformed area according to the atmospheric coefficient to obtain a plurality of categories, and assign the high coherence points in the deformed area in the differential interference pattern to the corresponding category nearby.

[0056] The atmospheric coefficients of the high-coherence points reflect the degree to which the atmosphere affects the radar equipment. When all high-coherence points satisfy the same Gaussian distribution, they are similarly affected by atmospheric delay and belong to the same atmospheric delay. Conversely, if they satisfy multiple Gaussian distributions, they belong to different atmospheric delays. Therefore, the histogram distribution of the atmospheric coefficients of all high-coherence points is first calculated. The number of clusters is determined based on the number of Gaussian distributions. Then, the K-means clustering algorithm is used to divide the high-coherence points in the non-deformed area into several categories. For the high-coherence points in the deformed area, the high-coherence points are assigned to the corresponding category based on the spatial correlation characteristics of the atmosphere.

[0057] S6. Calculate the atmospheric delay coefficient of each category according to the atmospheric delay phase and radar distance of the high coherence point in each category.

[0058] S7. Determine a distance function model for each category according to the atmospheric delay coefficient of each category.

[0059] The atmospheric delay coefficient of each category of distance function model is determined according to the atmospheric delay phase of the high coherence point in each category and the distance to the radar position, and the distance function model determined by the coefficient of each category is obtained.

[0060] Since the high coherence points in each category are all points in the non-deformed area, after filtering to eliminate some random noise phases, it can be considered that the phases of the selected high coherence points are mainly composed of atmospheric phases and a small amount of residual noise phases. Therefore, the phases of the high coherence points in each category of the unwrapped graph and the distances to the coordinate origin are respectively substituted into the distance function formula model: The least squares algorithm is used to solve the atmospheric delay coefficient in the distance function model of each category to obtain the first distance function model of each category. is the atmospheric delay coefficient of the kth type of high coherence point i, is the atmospheric delay phase of the kth type of high coherence point i, r i k is the distance from the kth high coherence point i to the radar equipment.

[0061] The categories in the untangle graph do not satisfy The high coherence points of each category are eliminated, and the atmospheric delay phase of the remaining high coherence points in each category in the unwrapped graph and the distance to the coordinate origin are used to re-solve the atmospheric delay coefficient in the distance function model of each category, and the distance function model of each category is obtained respectively, where, is the atmospheric delay phase of the high coherence point i calculated according to the k-th distance function model, is the atmospheric delay phase of the kth type of high coherence point i, σ k is the error threshold of the kth class.

[0062] The calculation formula of the error threshold is:

[0063]

[0064] in, are the atmospheric delay phases of the kth type of high coherence point 1, high coherence point 2, high coherence point i and high coherence point q, respectively. The atmospheric delay phases of high coherence point 1, high coherence point 2, high coherence point i and high coherence point q are calculated for the first distance function model of the kth category respectively. q is the number of high coherence points. The value of q varies for different categories.

[0065] S8. Calculate the atmospheric delay phase of each of the high coherence points in each of the categories according to the distance function model of each of the categories.

[0066] S9. Remove the atmospheric delay phase of each of the high coherence points from the differential interferogram to obtain an atmospheric correction result.

[0067] The present embodiment provides a method for correcting the non-homogeneous atmosphere of a ground-based interferometric radar, comprising: obtaining a differential interferogram obtained by differentially interfering the original single-view complex image of a time series of ground-based synthetic aperture radar images; selecting high coherence points in the differential interferogram; performing phase unwrapping on the high coherence points and identifying the high coherence points in the deformation area; solving the atmospheric coefficient of each high coherence point in the non-deformation area; classifying the high coherence points in the non-deformation area according to the solved atmospheric coefficient, and assigning the high coherence points in the deformation area to the corresponding category according to the position; determining the atmospheric delay coefficient of each category distance function model according to the atmospheric delay phase of the high coherence points in each category and the distance to the radar position, and obtaining the distance function model determined by the coefficient of each category; determining the atmospheric delay phase of each high coherence point in each category according to the distance function model determined by the coefficient of each category; removing the atmospheric delay phase of each high coherence point from the differential interferogram to obtain the high coherence point after atmospheric correction. The method and system for correcting the non-homogeneous atmosphere of a ground-based interferometric radar provided by the present application solve the problem of heterogeneous atmospheric delay in the monitoring area and improve the correction accuracy of atmospheric delay.

[0068] In another exemplary embodiment, Figure 3 As shown, a method for non-homogeneous atmosphere correction of ground-based interferometric radar is provided, comprising the following steps:

[0069] Step 101: Obtain a differential interferogram of a time series of original single-view complex images of a ground-based synthetic aperture radar image after differential interferometry.

[0070] Step 102: selecting a high coherence point in the differential interferogram, wherein the high coherence point may be a high coherence point selected based on a preset amplitude deviation and a coherence coefficient threshold.

[0071] Step 103: performing phase unwrapping on the high coherence points and identifying the high coherence points in the deformation area.

[0072] Step 104: Calculate the atmospheric coefficients in the distance direction for all high coherence points in the non-deformed area. The calculation formula is a i =φ i / r i , where a i is the atmospheric coefficient of the desired high coherence point i, φ i is the atmospheric delay phase of high coherence point i, r i is the distance from the high coherence point i to the radar equipment.

[0073] Step 105: Using the K-means clustering algorithm to classify the non-deformed high coherence points according to the solved atmospheric coefficients, and assigning the deformed high coherence points to the corresponding categories according to their locations.

[0074] Step 106: Determine the atmospheric delay coefficient of the distance function model of each category according to the atmospheric delay phase of the high coherence point in each category and the distance to the radar position, and obtain the distance function model determined by the coefficient of each category. The distance function model formula is: is the atmospheric delay coefficient of the kth type of high coherence point i, is the atmospheric delay phase of the kth type of high coherence point i, r i k is the distance from the kth high coherence point i to the radar equipment.

[0075] Step 107: Determine the atmospheric delay phase of each high coherence point in each category according to the distance function model determined by the coefficient of each category.

[0076] Step 108: removing the atmospheric delay phase of each high coherence point from the differential interferogram to obtain the atmospherically corrected high coherence point.

[0077] In practical applications, high coherence points can be selected based on the preset amplitude deviation and coherence coefficient thresholds, or they can be the union of the above-mentioned points. The calculation formula for amplitude deviation is:

[0078]

[0079] where σ A and mA They represent the standard deviation and mean of the amplitude of the pixel points in the time series.

[0080] The calculation formula of the coherence coefficient is:

[0081]

[0082] Where S and M represent the phases of different high coherence points, (i, j) represents the row and column numbers of the high coherence point, and * represents the conjugate operation of the complex number.

[0083] The principle of the non-homogeneous atmosphere correction method in this embodiment is as follows:

[0084] According to the electromagnetic wave propagation theory, an electromagnetic wave with a wavelength of λ is emitted at time t and propagates from the emission point through a distance r i When reaching the target pixel i and returning, the echo phase of its differential interferogram can be expressed as equation (3), where φ(t) is the differential interferogram obtained at time t.

[0085]

[0086] Where n is the atmospheric delay coefficient, which is related to temperature, humidity and air pressure. When the atmosphere in the scene is assumed to be homogeneous, it can be considered that n(r,t) is only related to time t. The phase related to the atmosphere in the original interferogram can be simply expressed as follows:

[0087]

[0088] Among them, φ t is the differential interference phase at time t, n(t) is the atmospheric delay coefficient at time t, and r is the distance from the radar transmitting point to the target point.

[0089] The phase of the original interference pattern is mainly composed of the deformation phase φ dis , atmospheric phase φ atm and random noise phase φ noi Composition, as shown in formula (5):

[0090] φ=φ dis +φ atm +φ noi (5)

[0091] As can be seen from Equation (4), the atmospheric coefficient of any high coherence point can be obtained from the atmospheric phase and the distance between it and the radar device. For high coherence points without deformation, after filtering to eliminate some random noise phase, it can be considered that the phase of the selected high coherence point is mainly composed of the atmospheric phase and a small amount of residual noise phase. Therefore, the phase difference and slant range difference between each high coherence point and the radar device can be calculated. Equation (4) can then be used to estimate the atmospheric coefficient of each high coherence point in the relevant distance direction. The atmospheric coefficient of the high coherence point represents the degree of atmospheric influence on the radar device. When the atmospheric coefficients of two points are the same, it means that the atmospheric refractive index of the two points is the same. Then, the two points can be considered to be affected by the same atmospheric effect. Therefore, they can be classified according to the difference in atmospheric coefficients. However, due to the influence of random noise, even two points belonging to the same atmospheric effect cannot have exactly the same atmospheric coefficients. Instead, they have a small difference. When there are enough points, they will satisfy the same Gaussian distribution. Therefore, the distribution of the atmospheric coefficients of all points can be used to determine whether they are affected by the same atmospheric effect. If the atmospheric coefficients at all points in the monitoring area satisfy the same Gaussian distribution, the area is considered to have a homogeneous atmosphere. However, if multiple Gaussian distributions appear in the monitoring area, the area is considered to have a non-homogeneous atmosphere. Therefore, this paper divides the non-homogeneous area into several homogeneous regions by calculating the distance-dependent atmospheric coefficients. The number of homogeneous regions can be determined by the number of Gaussian distributions that the atmospheric coefficients satisfy.

[0092] However, for highly coherent points with deformation, classifying them by solving the atmospheric coefficient based on the relationship between phase difference and slant range difference will lead to inaccurate classification, because the main phase component of the deformation point is the deformation phase. Therefore, the highly coherent points in the deformation area should be eliminated before using the atmospheric coefficient classification. This paper uses the unwrapped interferogram to identify the deformation area, and then masks the deformation area. After classification, the highly coherent points in the deformation area are assigned to the corresponding category according to the atmospheric space correlation characteristics.

[0093] The classification algorithm uses the K-means clustering algorithm, which is simple and easy to implement and has good clustering effect. The specific steps are as follows:

[0094] (1) Count the distribution of atmospheric coefficients of all highly coherent points, determine the number of clusters according to the number of Gaussian distributions, and set the peak of each Gaussian distribution as the centroid of each cluster, denoted as a i (i=1,2,…m);

[0095] (2) Calculate the distance between the atmospheric coefficients of all high coherence points and each center of mass. Since this paper only uses the atmospheric coefficient data, it is only necessary to solve all atmospheric coefficients a k The absolute value of the difference between (k=1,2,…n) and each centroid can be calculated using the following formula (6):

[0096]

[0097] (3) Assign the point with the smallest absolute value of the difference between the atmospheric coefficient and the center of mass to the corresponding category. Assign the kth high coherence point to the jth category through formula (7);

[0098]

[0099] (IV) Calculate the average value in each class as the new centroid, S i Is a i The centered class, a i It can be obtained by formula (8): a is the atmospheric coefficient, N is S i The number of a in the class.

[0100]

[0101] (5) When the centroid change no longer exceeds the set threshold, the iteration stops, otherwise go to the second step to continue the cycle.

[0102] After K-means classification, the non-homogeneous atmospheric region is divided into several different homogeneous atmospheric regions. It is only necessary to perform atmospheric correction on each homogeneous atmospheric region. Since this method is based on the atmospheric coefficient related to the distance upward, the classified homogeneous atmospheric regions satisfy the linear change in the distance upward. Therefore, the correction method of the distance function model is selected, as shown in formula (9): atm is the atmospheric delay phase.

[0103] φatm=r·β (9)

[0104] Taking a certain category as an example, the high coherence points are brought into the model to construct a linear equation system, such as formula (10):

[0105]

[0106] Where n represents the number of high coherence points. i 、r i and are the unwrapped phase and slant range of the i-th high coherence point, respectively. β is the coefficient matrix to be estimated. The unknown vector β is estimated by least squares regression (11):

[0107]

[0108] Where T represents the matrix transpose. The estimated component of the atmospheric phase is given by equation (12): represents the obtained true value, X represents the matrix containing r in formula (10), and Δφ represents the phase matrix in formula (10).

[0109]

[0110] For the deformation points among the high coherence points, two steps are still performed. First, all the high coherence points are brought into the model and the least squares principle is used for preliminary estimation. Then, the phase standard deviation σ is calculated by formula (13). The standard deviation can reflect the degree of dispersion of the data. When the initial estimated phase is far from the original phase, it means that it may be affected by the deformation phase or gross error. Therefore, twice the standard deviation is set as the threshold to eliminate the gross error points that are greatly affected by noise and deformation. Then, the remaining stable high coherence points are substituted into the model for calculation.

[0111] Δφ-Δφ atm |<2·σ

[0112]

[0113] The interferogram after atmospheric phase correction in step 108 can be expressed as:

[0114] φ corr =φ meas -φ atm (14)

[0115] Among them, φ meas is the measured phase. φ atm is the atmospheric delay phase.

[0116] Based on the same inventive concept, the embodiment of the present application also provides a device for implementing the above-mentioned ground-based interferometric radar non-homogeneous atmosphere correction method. The solution provided by the device is similar to the solution described in the above-mentioned method, so the specific limitations in one or more embodiments provided below can refer to the limitations above and will not be repeated here. Figure 4 , the system comprises:

[0117] The differential interferogram acquisition module 201 is used to obtain a differential interferogram obtained by performing differential interferometry on the original single-view complex images in the time series of ground-based synthetic aperture radar images.

[0118] The high coherence point selection module 202 is configured to select high coherence points in the differential interferogram, wherein the high coherence points are selected based on a preset amplitude deviation and a coherence coefficient threshold.

[0119] The phase unwrapping module 203 is used to perform phase unwrapping on the high coherence points and identify the high coherence points in the deformation area.

[0120] The high coherence point atmospheric coefficient calculation module 204 is used to calculate the atmospheric coefficients of all high coherence points in the non-deformation area in the relevant distance direction. The calculation formula is a i =φ i / r i, where a i is the atmospheric coefficient of the desired high coherence point i, φ i is the atmospheric delay phase of high coherence point i, r i is the distance from the high coherence point i to the radar equipment.

[0121] The high coherence point classification module 205 is used to classify high coherence points with different atmospheric delay effects into different categories according to the atmospheric coefficients of each high coherence point using the K-means clustering algorithm, that is, to classify non-homogeneous atmospheric regions into different homogeneous atmospheric regions.

[0122] The distance function model solving module 206 is used to solve the atmospheric delay coefficient in the distance function model according to the atmospheric delay phase of each high coherence point in different categories and the distance to the coordinate origin, and obtain the distance function model determined by the coefficient of each category. The distance function model formula is: in is the atmospheric delay coefficient of the kth type of high coherence point i, is the atmospheric delay phase of the kth type of high coherence point i, r i k is the distance from the kth high coherence point i to the radar equipment.

[0123] The atmospheric delay phase solving module 207 is used to calculate the atmospheric delay phase of each high coherence point in the interference pattern according to the distance function model determined by each category coefficient.

[0124] The atmospheric correction module 208 is configured to remove the atmospheric delay phase of each point from the differential interferogram to obtain an interferogram after atmospheric phase correction.

[0125] As an optional implementation manner, the high coherence point selection module 202 specifically includes:

[0126] The high coherence point selection unit is used to calculate the amplitude deviation and coherence coefficient of each point in the differential interferogram and select points that simultaneously meet the requirements of amplitude deviation less than a preset threshold and coherence coefficient greater than a preset threshold as high coherence points. The preset thresholds for amplitude deviation and coherence coefficient can be 0.15 and 0.95, respectively.

[0127] As an optional implementation, the distance function model solving module 206 specifically includes:

[0128] A first distance function model solving unit is used to substitute the phase of each high coherence point in each category and the distance to the coordinate origin into the distance function model, solve the atmospheric delay coefficient in the distance function model, and obtain the first atmospheric delay model of each category;

[0129] The second distance function model solving unit is used to separate the categories that do not meet The high coherence points of each category are removed, and the atmospheric delay phase of the remaining high coherence points of each category in the unwrapped graph and the distance to the coordinate origin are used to re-solve the atmospheric delay coefficient in each distance function model to obtain the second atmospheric delay model, which is the atmospheric delay model determined by the coefficients of each category, where is the atmospheric delay phase of the high coherence point i calculated according to the k-th distance function model, is the atmospheric delay phase of the kth type of high coherence point i, σ k is the error threshold of the kth class.

[0130] The error threshold determination unit is based on Calculate the error threshold σ k ,in, are respectively the atmospheric delay phases of the high coherence point 1, high coherence point 2, high coherence point i and high coherence point q obtained from the kth class of the unwrapped graph, The atmospheric delay phases of high coherence point 1, high coherence point 2, high coherence point i and high coherence point q are calculated for the first distance function model of the kth category respectively. q is the number of high coherence points. The value of q varies for different categories.

[0131] The following is a verification of the correction effect of the non-homogeneous atmosphere correction method and system for ground-based interferometric radar provided in this application:

[0132] The upper reservoir of a pumped storage power station in Shaanxi Province was used as the experimental area. The area is basin-shaped, with high sides and low in the middle. A radar device was set up in the middle of the upper reservoir and a 360° full-scale scan was performed. Figure 5 The radar equipment used was the GPRI-II, manufactured by GAMMA, Switzerland. GB-InSAR interferograms with a 2.5-hour interval were selected for the atmospheric correction experiment. The experimental area has a wide scanning range and a complex, non-homogeneous atmosphere, making it suitable for non-homogeneous atmospheric correction experiments. The specific implementation steps are as follows:

[0133] Step 1: 206,279 high coherence points were selected based on the intersection of the amplitude deviation method and the coherence coefficient method, with the thresholds set at 0.15 and 0.95 respectively. The distribution of high coherence points is shown in Figure 6 .

[0134] Step 2: Perform phase unwrapping on the selected high coherence points. Figure 7 .

[0135] Step 3: Calculate the atmospheric coefficients of all high coherence points in the range direction, and make statistics of all high coherence points in the form of a histogram, see Figure 8 .

[0136] Step 4: Use K-means clustering algorithm to classify high coherence points. Figure 9 .

[0137] Step 5: Establish distance function models for different categories.

[0138] Step 6: Solve the atmospheric delay coefficients of different categories separately.

[0139] Step 7: Calculate and correct the atmospheric delay of each category. Based on the atmospheric delay coefficients obtained, the atmospheric model is:

[0140]

[0141] The simulated atmospheric delay is shown in Figure 10 , the corrected interference pattern is shown in Figure 11 In order to verify the accuracy of the proposed method, the calculation results of the traditional elevation function model are compared with the distance elevation function model and the rectangular coordinate three-dimensional atmospheric delay model.

[0142]

[0143] The simulated atmospheric delay is shown in Figure 12 , the corrected interference pattern is shown in Figure 13 .

[0144] The calculation results of the three-dimensional atmospheric delay model are:

[0145]

[0146] The simulated atmospheric delay is shown in Figure 14 , the corrected interference pattern is shown in Figure 15 .

[0147] Comparing the atmospheric phase simulated by the method provided by the present application and the two traditional methods, it can be clearly seen that the atmospheric phase simulated by the method provided by the present application is obviously more consistent with the interference pattern, and the high coherence point area after correction is also near 0, while the traditional method is not the case. The interference pattern after correction obviously has residual atmosphere. In order to quantify its accuracy, the high-quality residuals after correction are used for accuracy assessment. Since the selected high-coherence points are invisible, the residuals after correction should theoretically be 0, but due to the presence of noise, they will appear as a Gaussian distribution with a mean of 0. Obviously, the smaller the standard deviation of the residuals, the better the correction result. The histogram of the high-coherence point residuals after correction by the traditional method and the method provided by the present application is shown in Figure 16-Figure 18 . Figure 16 and Figure 17 are the residuals after correction by the distance-elevation function model and the three-dimensional atmospheric delay correction model, respectively. The standard deviations of the residuals are 0.61 mm and 0.42 mm, Figure 18The residual error after correction by the proposed method is 0.33 mm, and its standard deviation is 0.33 mm. It can be seen that the method proposed in this application has significantly improved accuracy compared to traditional methods and can effectively improve the accuracy of atmospheric correction.

[0148] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output (I / O) interface, and a communication interface. The processor, memory, and I / O interface are connected via a system bus, and the communication interface is connected to the system bus via the I / O interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The I / O interface of the computer device is configured to exchange information between the processor and an external device. The communication interface of the computer device is configured to communicate with an external terminal via a network connection. When executed by the processor, the computer program implements a method for non-homogeneous atmosphere correction for a ground-based interferometric radar.

[0149] In an exemplary embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0150] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0151] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0152] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0153] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0154] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for correcting non-homogeneous atmosphere of ground-based interferometric radar, characterized in that: The ground-based interferometric radar non-homogeneous atmosphere correction method comprises: Obtaining a differential interferogram of the monitoring area; determining a high coherence point in the differential interferogram; determining a non-deformed region in the differential interferogram in combination with the high coherence point; Calculating an atmospheric coefficient of each of the high coherence points in the non-deformed area; the atmospheric coefficient is calculated based on the atmospheric delay phase and radar distance of the high coherence point; the radar distance is the distance between the high coherence point and the radar device; Classifying the high coherence points in the non-deformed area according to the atmospheric coefficient to obtain a plurality of categories, and allocating the high coherence points in the deformed area in the differential interferogram to the corresponding category nearby; Calculating the atmospheric delay coefficient of each category according to the atmospheric delay phase and radar distance of the high coherence point in each category; Determine a distance function model for each category according to the atmospheric delay coefficient of each category; Calculating the atmospheric delay phase of each of the high coherence points in each of the categories according to the distance function model of each of the categories; removing the atmospheric delay phase of each of the high coherence points from the differential interferogram to obtain an atmospheric correction result; Classifying the high coherence points in the non-deformed area according to the atmospheric coefficient to obtain a plurality of categories, and allocating the high coherence points in the deformed area in the differential interferogram to the corresponding category nearby, specifically including: The number of clusters is determined according to the number of values ​​in the atmospheric coefficient that satisfy the Gaussian distribution; the atmospheric coefficient is the atmospheric coefficient of the high coherence point in the non-deformed area; Based on the number, the high coherence points in the non-deformed area are classified using a K-means clustering algorithm to obtain a plurality of categories; According to the spatial correlation characteristics of the atmosphere, the highly coherent points in the deformation area are assigned to the corresponding categories.

2. The method for non-homogeneous atmosphere correction of ground-based interferometric radar according to claim 1, characterized in that: Calculating the atmospheric delay coefficient of each category according to the atmospheric delay phase and radar distance of the high coherence point in each category; Determining the distance function model of each category according to the atmospheric delay coefficient of each category specifically includes: Substitute the atmospheric delay phase and radar distance of the high coherence point in each category into the distance function formula model to calculate the first atmospheric delay coefficient of each category respectively; the distance function formula model is: in, is the atmospheric delay coefficient of the kth type of high coherence point i, is the atmospheric delay phase of the kth type of high coherence point i, r i k is the distance from the kth high coherence point i to the radar equipment; Using the least squares method to fit the first atmospheric delay coefficient of each category respectively, to obtain a first distance function model of each category; Each of the categories that do not meet The high coherence points are eliminated; among them, is the atmospheric delay phase of the high coherence point i calculated according to the kth type first distance function model, σ k is the error threshold of the kth class; Substituting the atmospheric delay phase and radar distance of the remaining high coherence points in each of the categories into the distance function formula model, respectively calculating the second atmospheric delay coefficient of each of the categories; The second atmospheric delay coefficient of each category is fitted respectively by using the least square method to obtain a distance function model of each category.

3. The method for non-homogeneous atmosphere correction of ground-based interferometric radar according to claim 2, characterized in that: The calculation formula of the error threshold is: in, are the atmospheric delay phases of the kth type of high coherence point 1, high coherence point 2, high coherence point i and high coherence point q, respectively. The atmospheric delay phases of high coherence point 1, high coherence point 2, high coherence point i and high coherence point q are calculated for the first distance function model of the kth type respectively, where q is the number of high coherence points.

4. The method for non-homogeneous atmosphere correction of ground-based interferometric radar according to claim 1, characterized in that: Determining a high coherence point in the differential interferogram specifically includes: Calculating the amplitude deviation and coherence coefficient of each point in the differential interferogram; A point that satisfies both the conditions of amplitude deviation being less than a preset amplitude deviation and coherence coefficient being greater than a coherence coefficient threshold is selected as a high coherence point.

5. The method for non-homogeneous atmosphere correction of ground-based interferometric radar according to claim 1, characterized in that: Determining a non-deformed area in the differential interference pattern in combination with the high coherence point specifically includes: performing phase unwrapping on the differential interference pattern according to the high coherence points to obtain an interference pattern; identifying a deformation region in the interference pattern using an image processing method; The area other than the deformed area in the interference pattern is determined as a non-deformed area.

6. The method for non-homogeneous atmosphere correction of ground-based interferometric radar according to claim 1, characterized in that: The calculation formula of the atmospheric coefficient of the high coherence point is: a i =φ i / r i ; Among them, a i is the atmospheric coefficient of the high coherence point i, φ i is the atmospheric delay phase of high coherence point i, r i is the distance from the high coherence point i to the radar equipment.

7. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the non-homogeneous atmosphere correction method for ground-based interferometric radar according to any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the non-homogeneous atmosphere correction method for ground-based interferometric radar according to any one of claims 1 to 6 is implemented.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the non-homogeneous atmosphere correction method for ground-based interferometric radar according to any one of claims 1 to 6 is implemented.

Images