Ground-based radar adaptive threshold permanent scatterer identification method, device and storage medium

By grouping and clustering algorithms on ground-based radar timing SAR images, high-quality permanent scatterer points are identified, which solves the problem of unadaptive PS point selection in the prior art, and improves the accuracy and stability of deformation inversion sex.

CN113848550BActive Publication Date: 2024-12-06INNER MONGOLIA UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111122989.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-09-24
Publication Date
2024-12-06
Estimated Expiration
2041-09-24

AI Technical Summary

Technical Problem

In the prior art, the permanent scatterer point selection method relies on fixed thresholds, resulting in missed selection and missed selection of PS points during long-term observations, and the adaptability cannot be achieved, affecting the accuracy of deformation inversion results.

Method used

Adaptive threshold method of ground-based radar is used to group the timing SAR images, and the amplitude threshold method and clustering algorithm are used to identify candidate permanent scatterer points. Based on coherence coefficient, amplitude deviation information and phase stability, PS points are updated in real time , avoid setting fixed thresholds.

Benefits of technology

It improves the stability of PS points and the real-time selection, reduces errors in deformation inversion, and ensures high-precision deformation monitoring results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113848550B_ABST
    Figure CN113848550B_ABST
Patent Text Reader

Abstract

The present disclosure relates to a method, device and storage medium for identifying permanent scatterers of ground-based radar with an adaptive threshold, which performs grouping operations on acquired time-series SAR images; obtains a set of candidate permanent scatterer points of each group of SAR images by using an amplitude threshold method; generates a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm; obtains a corresponding low amplitude deviation permanent scatterer class and a high phase stability permanent scatterer class through a clustering algorithm based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class; and identifies target permanent scatterer points based on the low amplitude deviation permanent scatterer class and the high phase stability permanent scatterer class. Through the various embodiments of the present disclosure, the stability of the number of PS points can be greatly improved, and the adaptive threshold PS point selection of ground-based radar is realized, which lays a good foundation for high-precision deformation inversion.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the technical field of radar interferometry data processing, and in particular to a ground-based radar adaptive threshold permanent scatterer identification method, a ground-based radar adaptive threshold permanent scatterer identification device and a computer-readable storage medium. Background Art

[0002] In the prior art, the selection method of permanent scatterers (PS) points often requires setting a fixed threshold to achieve implementation, which easily leads to the misselection and omission of PS points. Since the environment of ground-based interferometric synthetic aperture radar (GB-InSAR) is relatively complex during long-term observation, the monitoring results are affected to a certain extent, resulting in the disappearance or emergence of PS points. If the selection of PS points relies solely on a fixed threshold, it cannot achieve the adaptability to cope with different scenarios, which may cause the PS points to fail to be updated in time, resulting in differences in the quantity and quality of the PS points selected during long-term observations, which will introduce large deformation inversion errors, thereby affecting the final deformation inversion results. Summary of the invention

[0003] The present disclosure intends to provide a ground-based radar adaptive threshold permanent scatterer identification method, a ground-based radar adaptive threshold permanent scatterer identification device and a computer-readable storage medium, which can greatly improve the stability of the number of PS points, realize the ground-based radar adaptive threshold PS point selection, and lay a good foundation for high-precision deformation inversion.

[0004] According to one of the solutions of the present disclosure, a method for identifying permanent scatterers using an adaptive threshold of a ground-based radar is provided, comprising:

[0005] Perform grouping operations on the acquired time-series SAR images;

[0006] The set of candidate permanent scatterer points of each group of SAR images is obtained by using the amplitude threshold method;

[0007] Generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm;

[0008] Based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class are obtained through clustering algorithm respectively;

[0009] The target permanent scatterer points are identified based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class.

[0010] In some embodiments, the grouping operation of the acquired time-series SAR images includes:

[0011] In response to the generation of the Nth SAR image, a set of SAR images is formed to perform permanent scatterer point selection processing;

[0012] In response to each newly generated SAR image, it is combined with the previous N-1 SAR images to form a new group for permanent scatterer point selection processing;

[0013] Wherein: N is an integer greater than 20.

[0014] In some embodiments, the step of obtaining a set of candidate permanent scatterer points for each group of SAR images using an amplitude threshold method includes:

[0015] The minimum amplitude time series value of each pixel point extracted from a group of N SAR images is compared, and the pixel point whose minimum amplitude time series value is greater than the amplitude threshold is taken as a candidate permanent scatterer point.

[0016] In some embodiments, the step of generating a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set to obtain a high coherence permanent scatterer class through a clustering algorithm comprises:

[0017] Generate a coherence coefficient sequence according to the coherence coefficient of each pixel in the candidate permanent scatterer point set of each two adjacent registered SAR images in the group;

[0018] The coherence coefficient sequence is used as the input of the clustering algorithm to cluster high coherence permanent scatterer classes from the candidate permanent scatterer points.

[0019] In some embodiments, the method of obtaining a low amplitude dispersion permanent scatterer class includes:

[0020] The amplitude dispersion information of each pixel in the high coherence permanent scatterer class is obtained by using the ratio of the standard deviation to the mean of the amplitude of each pixel in the time-series SAR images within the group.

[0021] The amplitude deviation information of each pixel in the high coherence permanent scatterer class is used as the input of the clustering algorithm, and the low amplitude deviation permanent scatterer class is clustered from the high coherence permanent scatterer class.

[0022] In some embodiments, the method of obtaining a high phase stability permanent scatterer class includes:

[0023] For a set of N time-series SAR images, each adjacent two scenes are processed by interference to generate N-1 interferometric phase images, and the correlation value of the time correlation of each pixel is calculated to obtain the phase stability information of each pixel in the high coherence permanent scatterer class.

[0024] The phase stability information of each pixel in the high coherence permanent scatterer class is used as the input of the clustering algorithm, and the high phase stability permanent scatterer class is clustered from the high coherence permanent scatterer class.

[0025] In some embodiments, the step of identifying target permanent scatterer points based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class includes:

[0026] The intersection of the low amplitude deviation permanent scatterer class set and the high phase stability permanent scatterer class set is taken, and the permanent scatterer points in the intersection are used as the identified target permanent scatterer points.

[0027] In some embodiments, the step of executing a clustering algorithm employs a Gaussian mixture model clustering algorithm.

[0028] According to one of the solutions disclosed in the present invention, a ground-based radar adaptive threshold permanent scatterer identification device is provided.

[0029] A processing module configured to perform grouping operations on the acquired time-series SAR images and obtain a set of candidate permanent scatterer points for each group of SAR images using an amplitude threshold method;

[0030] A clustering module, configured to generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm; based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, obtain the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class through a clustering algorithm respectively;

[0031] The identification module is configured to identify the target permanent scatterer points based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class.

[0032] According to one of the solutions of the present disclosure, a computer-readable storage medium is provided, on which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the following are implemented:

[0033] According to the above ground-based radar adaptive threshold permanent scatterer identification method.

[0034] The ground-based radar adaptive threshold permanent scatterer identification method, ground-based radar adaptive threshold permanent scatterer identification device and computer-readable storage medium of various embodiments of the present disclosure perform grouping operations on acquired time-series SAR images; use the amplitude threshold method to obtain a set of candidate permanent scatterer points for each group of SAR images; generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set to obtain a high-coherence permanent scatterer class through a clustering algorithm; based on the amplitude deviation information and phase stability information of each pixel in the high-coherence permanent scatterer class, obtain the corresponding low-amplitude deviation permanent scatterer class and high-phase stability permanent scatterer class through a clustering algorithm; and identify target permanent scatterer points based on the low-amplitude deviation permanent scatterer class and the high-phase stability permanent scatterer class. By grouping the acquired time-series SAR images, the PS points of each group of SAR images are continuously updated, thus ensuring the real-time performance of the PS point selection algorithm; then the amplitude threshold method is used to roughly select each pixel in each group of SAR images to obtain a set of candidate permanent scatterer points; then the coherence coefficient sequence of each pixel in the candidate permanent scatterer point set is calculated as the input of the clustering algorithm, and the candidate permanent scatterer point set is divided into two categories, namely, high coherence permanent scatterer class and low coherence permanent scatterer class; the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class are calculated as the input of the second and third clustering algorithms, respectively, and the high coherence permanent scatterer class is divided into two categories respectively, and the high-coherence permanent scatterer class is taken as the intersection of the low amplitude deviation permanent scatterer class set and the high phase stability permanent scatterer class set to obtain high-quality pixel points as the target permanent scatterer points.

[0035] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure, as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In the drawings, which are not necessarily drawn to scale, similar reference numerals in different views may represent similar components. Similar reference numerals with letter suffixes or similar reference numerals with different letter suffixes may represent different instances of similar components. The drawings generally illustrate various embodiments by way of example and not limitation, and together with the description and claims, serve to explain the disclosed embodiments.

[0037] Figure 1 A flow chart showing an embodiment of the ground-based radar adaptive threshold permanent scatterer identification method disclosed in the present invention;

[0038] Figure 2 A schematic diagram showing SAR image grouping according to an embodiment of the present disclosure;

[0039] Figure 3 A flow chart of a Gaussian mixture model clustering algorithm according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution of the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings of the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the described embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0041] Unless otherwise defined, the technical terms or scientific terms used in the present disclosure shall have the common meanings understood by persons with ordinary skills in the field to which the present disclosure belongs. "Include" or "comprising" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, but do not exclude other elements or objects.

[0042] In order to keep the following description of the embodiments of the present disclosure clear and concise, the present disclosure omits detailed descriptions of well-known functions and well-known components.

[0043] Ground-based Interferometric Synthetic Aperture Radar (GB-InSAR) is a new type of ground deformation monitoring tool and a derivative of spaceborne synthetic aperture radar on the ground. GB-InSAR has been widely used in ground deformation monitoring due to its advantages of all-day, all-weather, non-contact and high-precision data acquisition. Since the temporal resolution and spatial resolution of the spaceborne SAR system are difficult to meet the monitoring requirements, GB-InSAR has high temporal and spatial resolution and accuracy for deformation monitoring in a small observation area. Therefore, GB-InSAR has unique advantages in monitoring some areas with small observation ranges and fast deformation speeds.

[0044] Permanent Scatterers (PS) interferometry is an important means for GB-InSAR to monitor ground deformation and obtain high-precision deformation information. PS interferometry takes multiple SAR images of a long time series as input, first selects scattering points that can maintain high correlation within the monitoring time range, and defines these scattering points as PS points. At present, the two most classic PS point selection methods are the amplitude deviation method proposed by Italian scholars Ferretti and others and the correlation coefficient method proposed by Italian scholars Berardino and others. By processing the phase information data at the PS point, the deformation information of the observation scene within the monitoring time range is obtained. It can be seen that the accuracy of permanent scatterer extraction will eventually affect the accuracy of deformation information extraction. Therefore, how to extract PS points with strong stability, strong reflectivity and high correlation has become a key issue for GB-InSAR to obtain high-precision deformation information.

[0045] In combination with the above-mentioned background technology, the present disclosure records the corresponding solutions in the form of embodiments to solve the defects in the prior art, but it is not intended to limit the scope of protection of the patent right claimed by the present disclosure.

[0046] As one of the solutions, an embodiment of the present disclosure provides a ground-based radar adaptive threshold permanent scatterer identification method, including:

[0047] Perform grouping operations on the acquired time-series SAR images;

[0048] The set of candidate permanent scatterer points of each group of SAR images is obtained by using the amplitude threshold method;

[0049] Generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm;

[0050] Based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class are obtained through clustering algorithm respectively;

[0051] The target permanent scatterer points are identified based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class.

[0052] In view of the problems raised in the foregoing, the embodiments of the present disclosure aim to propose a method for identifying permanent scatterers using adaptive thresholds for ground-based radars. The method first groups the acquired time-series SAR images, and continuously updates the PS points of each group of SAR images, thereby ensuring the real-time performance of the PS point selection algorithm; then, the amplitude threshold method is used to roughly select each pixel in each group of SAR images to obtain a set of candidate permanent scatterer points; then, the coherence coefficient sequence of each pixel in the set of candidate permanent scatterer points is calculated as the input of the clustering algorithm, and the set of candidate permanent scatterer points is divided into two categories, namely, a high coherence permanent scatterer class and a low coherence permanent scatterer class; the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class are calculated as the input of the second and third clustering algorithms, and the high coherence permanent scatterer class is divided into two categories respectively, to obtain a set of low amplitude deviation permanent scatterer classes U1 and a set of high phase stability permanent scatterer classes U2; finally, the high-quality pixel points obtained by taking the intersection of sets U1 and U2 as the target permanent scatterer points. In each embodiment of the present disclosure, the step of executing the clustering algorithm adopts a Gaussian mixture model (GMM) clustering algorithm. The relevant definition and configuration method of the model will be explained with examples below.

[0053] like Figure 1 As shown, the specific implementation steps may include but are not limited to steps S1 to S6 as examples.

[0054] In some embodiments, the present disclosure may be implemented as follows: the grouping operation of the acquired time-series SAR images includes:

[0055] In response to the generation of the Nth SAR image, a set of SAR images is formed to perform permanent scatterer point selection processing;

[0056] In response to each newly generated SAR image, it is combined with the previous N-1 SAR images to form a new group for permanent scatterer point selection processing;

[0057] Wherein: N is an integer greater than 20.

[0058] Specific implementations may include:

[0059] Step S1: Group the acquired time-series SAR images. The schematic diagram of SAR image grouping is as follows: Figure 2 The specific steps are as follows:

[0060] Step S11: starting from the first image monitored, until the Nth image is generated, a first group of SAR images is formed using the N SAR images from the first to the Nth, and a first permanent scatterer point selection process is performed.

[0061] Step S12: After that, every time a new SAR image is generated, it will be combined with the previous N - 1 SAR images to form a new group for permanent scatterer point selection processing.

[0062] In some embodiments, the present disclosure can be implemented as: The method of obtaining the candidate permanent scatterer point set for each group of SAR images by using the amplitude threshold method includes:

[0063] Comparing the minimum value of the amplitude time series of each pixel point extracted from a group of N SAR images, and taking the pixel points whose minimum value of the amplitude time series is greater than the amplitude threshold as candidate permanent scatterer points.

[0064] The specific implementation may include:

[0065] Step S2: Coarsely select permanent scatterer points for the pixel points of the SAR images within the group by using the amplitude threshold method to obtain a candidate permanent scatterer point set. The specific steps are as follows:

[0066] Step S21: For a group of N SAR images, first perform coarse selection of PS points by using the amplitude threshold method. The definition of the amplitude threshold is:

[0067]

[0068] In the formula: m and n are the total number of rows and the total number of columns in a single image respectively, is the amplitude value after radiometric correction of the pixel point (i, j) in the k - th image, and T is the amplitude threshold, that is, the maximum value of the amplitude means of each image in this group is selected as the amplitude threshold.

[0069] Step S22: Compare the minimum value A of the amplitude time series of each pixel point extracted from the N SAR images in this group min . If A min > T, then this pixel point is considered as a candidate permanent scatterer point; if A min < T, then this pixel point is considered as a non - candidate permanent scatterer point. A total of N1 qualified points are obtained.

[0070] In some embodiments, the present disclosure can be implemented as: Generating a coherence coefficient sequence based on the coherence coefficients of each pixel in the candidate permanent scatterer point set to obtain a high - coherence permanent scatterer class through a clustering algorithm, including:

[0071] Generating a coherence coefficient sequence according to the coherence coefficients of each pixel in the candidate permanent scatterer point sets of every two adjacent registered SAR images within the group;

[0072] Taking the coherence coefficient sequence as the input of the clustering algorithm to cluster out a high - coherence permanent scatterer class from the candidate permanent scatterer points.

[0073] Specific implementations may include:

[0074] Step S3: Calculate the coherence coefficient of each pixel in the candidate permanent scatterer point set of each two adjacent registered SAR images in the group to generate a coherence coefficient sequence, and use it as the input of the Gaussian mixture model (GMM) clustering algorithm to divide the candidate permanent scatterer point set into two categories, namely, high coherence permanent scatterer class and low coherence permanent scatterer class. The specific steps are as follows:

[0075] Step S31: Take the registered SAR images of each two adjacent scenes in the group, first establish a rectangular window with a certain pixel as the center, and the correlation coefficient calculation formula is as follows:

[0076]

[0077] Where: i, j are the positions of pixels, M(i, j) and S(i, j) are the information (complex) blocks of the two adjacent main and auxiliary images, m, n are the sizes of the rectangular sliding windows, and * is the conjugate of the complex number. For the candidate permanent scatterer point set of the SAR image in the group, there are N1 pixel points. For N SAR images, two adjacent SAR images are selected as the main and auxiliary images, and the N-1-dimensional coherence coefficient sequence can be obtained. The coherence coefficient sequence of each pixel is γ l ={γ l1 ,γ l2 ,…,γ lN-1}(where l=1,2,3,…,N1).

[0078] Step S32: Using the coherence coefficient sequence of each pixel in the candidate permanent scatterer point set obtained in step S31 as the input of the GMM clustering algorithm, the candidate permanent scatterer point set in step S22 is divided into two categories. The flowchart is as follows: Figure 3 As shown, the specific steps are as follows:

[0079] Step S321: Use the GMM clustering algorithm to divide the sample space into two categories, using the N-1-dimensional coherence coefficient sequence γ in step S31 l ={γ l1 ,γ l2 ,…,γ lN-1} (where l = 1, 2, 3, ..., N1) as input, its Gaussian mixture model is defined as:

[0080]

[0081] Where: μ i is the N-1 dimensional mean vector of the i-th Gaussian model, Σ i is the N-1×N-1 covariance matrix of the i-th Gaussian model, that is, μ i and Σ iis the parameter of the i-th Gaussian mixture component. i is the weight of the i-th Gaussian model, satisfying α i >0 and

[0082] P(γ|μ i ,Σ i ) satisfies the Gaussian distribution definition:

[0083]

[0084] Step S322: Initialize the parameters {α i ,μ i ,Σ i |i=1,2}.

[0085] Step S323: Let the random variable z j ∈{1,2} represents the generated sample γ l ={γ l1 ,γ l2 ,…,γ lN-1} (where l = 1, 2, 3, ..., N1) Gaussian mixture components, sample γ j The posterior probability P generated by the i-th Gaussian mixture component γ (z j =i|γ j ), and denote it as κ ji ,Right now

[0086]

[0087] According to the parameters in the Gaussian mixture model and formula (5), the coherence coefficient sequence γ of each pixel in the candidate PS point obtained in step S31 is l ={γ l1 ,γ l2 ,…,γ lN-1} (where l = 1, 2, 3, ..., N1) as the input of GMM clustering, GMM clustering will transform the sample set γ k Divided into 2 clusters C = {C1, C2}, each sample γ j The cluster label λ j Determined as follows:

[0088]

[0089] Step S324: Calculate the posterior probability P using the maximum likelihood estimation method γ (γ j ) Maximum log-likelihood function:

[0090]

[0091] Use L(γ) to update the parameters {α i_new ,μ i_new ,Σ i_new |i_new=1,2}, the maximum expectation (EM) algorithm is often used for iterative optimization to continuously update the parameters {α i ,μ i ,Σ i}, the partial derivative of L(γ) with respect to each parameter should be 0, and the parameters of each update iteration are calculated as follows:

[0092]

[0093]

[0094]

[0095] Step S325: Step S323 and Step S324 are looped in sequence until the likelihood function converges and the loop termination condition is reached. 1_new and μ 2_new Take the mean value respectively and get the vector mean μ 1_mean and μ 2_mean By comparing the two vector means, the cluster with the larger vector mean is regarded as a high-coherence permanent scatterer class, and the number of qualified clusters is N2; the cluster with the smaller vector mean is discarded as a low-coherence permanent scatterer class.

[0096] In some embodiments, the present disclosure may be implemented as follows: a method for obtaining a low amplitude dispersion permanent scatterer class includes:

[0097] The amplitude dispersion information of each pixel in the high coherence permanent scatterer class is obtained by using the ratio of the standard deviation to the mean of the amplitude of each pixel in the time-series SAR image.

[0098] The amplitude deviation information of each pixel in the high coherence permanent scatterer class is used as the input of the clustering algorithm, and the low amplitude deviation permanent scatterer class is clustered from the high coherence permanent scatterer class.

[0099] Specific implementations may include:

[0100] Step S4: The amplitude deviation information of each pixel in the high coherence permanent scatterer class is used as the input of the GMM clustering algorithm to divide the high coherence permanent scatterer class into two categories, namely, the low amplitude deviation permanent scatterer class and the high amplitude deviation permanent scatterer class. The specific steps are as follows:

[0101] Step S41: The amplitude deviation information is calculated by using the ratio of the standard deviation to the mean of the amplitude of each pixel in the time series SAR image. The calculation formula is as follows:

[0102]

[0103] In the formula, δ A is the standard deviation of pixel amplitude in the time series SAR image, m A is the mean value of the pixel amplitude in the time series SAR image, D A The amplitude deviation information of each pixel in the high coherence permanent scatterer class set is calculated as D A ={D A_1 ,D A_2 ,…,D A_N2}.

[0104] Step S42: The amplitude deviation information of each pixel in the high coherence permanent scatterer class set obtained in step S41 is used as the input of the GMM clustering algorithm, and the high coherence permanent scatterer class set in step S325 is divided into two categories. Similar operations are performed according to steps S322-S324, the difference being that the input information of the GMM clustering is different, and the amplitude deviation information D obtained in step S41 is used as the input of the GMM clustering algorithm. A ={D A_1 ,D A_2 ,…,D A_N2} as the input for GMM clustering.

[0105] Step S43: Loop through steps S323 and S324 until the likelihood function converges and the loop termination condition is reached.

[0106] Step S44: Compare the two mean values ​​μ after the iteration 1_new and μ 2_new , the clusters with smaller mean are regarded as the low amplitude deviation permanent scatterer class, and the set U1 is obtained, and the number of qualified ones is N3; the clusters with larger mean are discarded as the high amplitude deviation permanent scatterer class.

[0107] In some embodiments, the present disclosure may be implemented as follows: a method for obtaining a permanent scatterer class with high phase stability includes:

[0108] For a set of N time-series SAR images, each two adjacent scenes are processed by interference to generate N-1 interferometric phase images, and the phase stability information of each pixel in the high coherence permanent scatterer class is obtained;

[0109] The phase stability information of each pixel in the high coherence permanent scatterer class is used as the input of the clustering algorithm, and the high phase stability permanent scatterer class is clustered from the high coherence permanent scatterer class.

[0110] Specific implementations may include:

[0111] Step S5: The phase stability information of each pixel in the high coherence permanent scatterer class is used as the input of the GMM clustering algorithm to divide the high coherence permanent scatterer class into two categories, namely, the high phase stability permanent scatterer class and the low phase stability permanent scatterer class. The specific steps are as follows:

[0112] Step S51: For a group of N time-series SAR images, perform interference processing on each two adjacent scenes to generate N-1 interference phase images. The phase stability calculation formula is:

[0113]

[0114] Where: P sta is the temporal correlation of the pixel (phase stability information), is the i-th interference phase value of a certain pixel. The phase stability information sequence of each pixel in the high coherence permanent scatterer class set is calculated as P sta = {P sta_1 ,P sta_2 ,…,P sta_N2}.

[0115] Step S52: The phase stability information sequence of each pixel in the high coherence permanent scatterer class set obtained in step S51 is used as the input of the GMM clustering algorithm, and the high coherence permanent scatterer class set in step S325 is divided into two categories. The same operation is performed according to steps S322-S324, the difference is that the input information of the GMM clustering is different, and the phase stability information sequence P obtained in step S51 is used as the input of the GMM clustering algorithm. sta = {P sta_1 ,P sta_2 ,…,P sta_N2} as the input for GMM clustering.

[0116] Step S53: Loop through steps S323 and S324 until the likelihood function converges and the loop termination condition is reached.

[0117] Step S54: Compare the two mean values ​​μ after the iteration 1_new and μ 2_new , the clusters with larger mean are regarded as permanent scatterers with high phase stability, and the set U2 is obtained, of which the number of qualified ones is N4; the clusters with smaller mean are discarded as permanent scatterers with low phase stability.

[0118] In some embodiments, the present disclosure may be implemented as follows: the step of identifying target permanent scatterer points based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class includes:

[0119] The intersection of the low amplitude deviation permanent scatterer class set and the high phase stability permanent scatterer class set is taken, and the permanent scatterer points in the intersection are used as the identified target permanent scatterer points.

[0120] Specific implementations may include:

[0121] Step S6: Take the intersection of the low amplitude deviation permanent scatterer class set U1 obtained in step S44 and the high phase stability permanent scatterer class set U2 obtained in step S54, that is, U1∩U2, and obtain N after taking the intersection. all points as the final selected target permanent scatterer points.

[0122] As one of the solutions, the embodiment of the present disclosure provides a ground-based radar adaptive threshold permanent scatterer identification device,

[0123] A processing module configured to perform grouping operations on the acquired time-series SAR images and obtain a set of candidate permanent scatterer points for each group of SAR images using an amplitude threshold method;

[0124] A clustering module, configured to generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm; based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, obtain the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class through a clustering algorithm respectively;

[0125] The identification module is configured to identify the target permanent scatterer points based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class.

[0126] As an implementation manner, the processing module of the device of the present disclosure can be further configured as follows in combination with the description of steps S1 to S2 above:

[0127] In response to the generation of the Nth SAR image, a set of SAR images is formed to perform permanent scatterer point selection processing;

[0128] In response to each newly generated SAR image, it is combined with the previous N-1 SAR images to form a new group for permanent scatterer point selection processing;

[0129] Wherein: N is an integer greater than 20;

[0130] The minimum amplitude time series value of each pixel point extracted from a group of N SAR images is compared, and the pixel point whose minimum amplitude time series value is greater than the amplitude threshold is taken as a candidate permanent scatterer point.

[0131] As an implementation manner, the clustering module of the device of the present disclosure can be further configured as follows in combination with the description of steps S3 to S5 above:

[0132] Generate a coherence coefficient sequence according to the coherence coefficient of each pixel of the candidate permanent scatterer point set of each two adjacent registered SAR images in the group;

[0133] The coherence coefficient sequence is used as the input of the clustering algorithm to cluster the high coherence permanent scatterer class from the candidate permanent scatterer points;

[0134] The amplitude dispersion information of each pixel in the high coherence permanent scatterer class is obtained by using the ratio of the standard deviation of the amplitude of each pixel in the time-series SAR image to the mean value.

[0135] The amplitude deviation information of each pixel in the high coherence permanent scatterer class is used as the input of the clustering algorithm to cluster the low amplitude deviation permanent scatterer class from the high coherence permanent scatterer class;

[0136] For a set of N time-series SAR images, each adjacent two scenes are processed by interference to generate N-1 interferometric phase images, and the correlation value of the time correlation of each pixel is calculated to obtain the phase stability information of each pixel in the high coherence permanent scatterer class.

[0137] The phase stability information of each pixel in the high coherence permanent scatterer class is used as the input of the clustering algorithm, and the high phase stability permanent scatterer class is clustered from the high coherence permanent scatterer class.

[0138] As an implementation manner, the identification module of the device of the present disclosure can be further configured as follows in combination with the description of step S6 above:

[0139] The intersection of the low amplitude deviation permanent scatterer class set and the high phase stability permanent scatterer class set is taken, and the permanent scatterer points in the intersection are used as the identified target permanent scatterer points.

[0140] Specifically, one of the inventive concepts of the present disclosure is to be able to perform grouping operations on the acquired time-series SAR images; obtain the candidate permanent scatterer point set of each group of SAR images by using the amplitude threshold method; generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set to obtain a high coherence permanent scatterer class through a clustering algorithm; obtain the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class through a clustering algorithm based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class; identify the target permanent scatterer point based on the low amplitude deviation permanent scatterer class and the high phase stability permanent scatterer class. The acquired time-series SAR images are grouped, and the PS points of each group of SAR images are continuously updated, thus ensuring the real-time performance of the PS point selection algorithm. Secondly, the PS points of the pixels are roughly selected by the amplitude information, and then the GMM clustering algorithm is used to realize the comprehensive consideration of the amplitude and phase of the candidate permanent scatterer points. The entire permanent scatterer point selection process does not need to set any threshold parameters, and the target permanent scatterer point is finally selected through multi-level screening. While ensuring that the number of PS points in each group of images is sufficient, the stability of the number of PS points is also greatly improved, realizing adaptive threshold PS point selection for ground-based radar, laying a good foundation for high-precision deformation inversion.

[0141] The present disclosure also provides a computer-readable storage medium having computer-executable instructions stored thereon. When the computer-executable instructions are executed by a processor, the method for identifying permanent scatterers based on the ground-based radar adaptive threshold is mainly implemented, including:

[0142] Perform grouping operations on the acquired time-series SAR images;

[0143] The set of candidate permanent scatterer points of each group of SAR images is obtained by using the amplitude threshold method;

[0144] Generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm;

[0145] Based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class are obtained through clustering algorithm respectively;

[0146] The target permanent scatterer points are identified based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class.

[0147] In some embodiments, the processor that executes the computer executable instructions may be a processing device including one or more general processing devices, such as a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), etc. More specifically, the processor may be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor that runs other instruction sets, or a processor that runs a combination of instruction sets. The processor may also be one or more special-purpose processing devices, such as an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), a system on a chip (SoC), etc.

[0148] In some embodiments, the computer-readable storage medium may be a memory such as a read-only memory (ROM), a random access memory (RAM), a phase-change random access memory (PRAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), an electrically erasable programmable read-only memory (EEPROM), other types of random access memory (RAM), a flash disk or other form of flash memory, a cache, a register, a static memory, a compact disk read-only memory (CD-ROM), a digital versatile disk (DVD) or other optical storage, a cassette or other magnetic storage device, or any other possible non-temporary medium used to store information or instructions that can be accessed by a computer device.

[0149] In some embodiments, the computer executable instructions may be implemented as a plurality of program modules, and the plurality of program modules together implement the signal processing method for medical images according to any one of the items of the present disclosure.

[0150] The present disclosure describes various operations or functions, which can be implemented as software codes or instructions or defined as software codes or instructions. The display unit can be implemented as a software code or instruction module stored on a memory, which can implement corresponding steps and methods when executed by a processor.

[0151] Such content may be source code or differential code ("delta" or "patch" code) that can be directly executed ("object" or "executable" form). The software implementation of the embodiments described herein may be provided by an article having code or instructions stored thereon, or by a method of operating a communication interface to send data through a communication interface. A machine or computer readable storage medium may enable a machine to perform the functions or operations described, and includes any mechanism for storing information in a form accessible by a machine (e.g., a computing display device, an electronic system, etc.), such as a recordable / unrecordable medium (e.g., a read-only memory (ROM), a random access memory (RAM), a disk storage medium, an optical storage medium, a flash memory display device, etc.). The communication interface includes any mechanism for interfacing with any of hardwired, wireless, optical, etc. media to communicate with other display devices, such as a memory bus interface, a processor bus interface, an Internet connection, a disk controller, etc. The communication interface may be configured by providing configuration parameters and / or sending signals to prepare the communication interface to provide a data signal describing the software content. The communication interface may be accessed by sending one or more commands or signals to the communication interface.

[0152] The computer executable instructions of the embodiments of the present disclosure can be organized into one or more computer executable components or modules. Any number and combination of such components or modules can be used to implement various aspects of the present disclosure. For example, various aspects of the present disclosure are not limited to the specific computer executable instructions or specific components or modules shown in the drawings and described herein. Other embodiments may include different computer executable instructions or components with more or less functions than those shown and described herein.

[0153] The above description is intended to be illustrative rather than restrictive. For example, the above examples (or one or more of them) can be used in combination with each other. For example, a person of ordinary skill in the art can use other embodiments when reading the above description. In addition, in the above-mentioned specific embodiments, various features can be grouped together to simplify the present disclosure. This should not be interpreted as an intention that a disclosed feature that is not required to be protected is necessary for any claim. On the contrary, the subject matter of the present disclosure may be less than all the features of a specific disclosed embodiment. Thus, the following claims are incorporated into the specific embodiments as examples or embodiments, wherein each claim is independently used as a separate embodiment, and it is considered that these embodiments can be combined with each other in various combinations or arrangements. The scope of the present disclosure should be determined with reference to the attached claims and the full scope of equivalent forms granted by these claims.

[0154] The above embodiments are only exemplary embodiments of the present disclosure and are not intended to limit the present disclosure. The protection scope of the present disclosure is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present disclosure within the essence and protection scope of the present disclosure, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present disclosure.

Claims

1. A method for identifying permanent scatterers using an adaptive threshold of a ground-based radar, comprising: Perform grouping operations on the acquired time-series SAR images; The set of candidate permanent scatterer points of each group of SAR images is obtained by using the amplitude threshold method; Generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm; Based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class are obtained through clustering algorithm respectively; The target permanent scatterer points are identified based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class; wherein: The grouping operation of the acquired time-series SAR images includes: In response to N SAR images are generated to form a group of SAR images for permanent scatterer point selection processing; in response to each newly generated SAR image, N -1 SAR images are combined to form a new group for permanent scatterer point selection processing; among them: N is an integer greater than 20; The method of using the amplitude threshold method to obtain a set of candidate permanent scatterer points for each group of SAR images includes: comparing a group N The minimum value of the amplitude time series of each pixel point extracted from the SAR image is used, and the pixel point whose minimum value of the amplitude time series is greater than the amplitude threshold is taken as the candidate permanent scatterer point; The method of generating a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set to obtain a high coherence permanent scatterer class through a clustering algorithm includes: generating a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set of each two adjacent registered SAR images in the group; using the coherence coefficient sequence as an input of the clustering algorithm to cluster a high coherence permanent scatterer class from the candidate permanent scatterer points; The method of obtaining the low amplitude deviation permanent scatterer class includes: obtaining the amplitude deviation information of each pixel in the high coherence permanent scatterer class by using the ratio of the standard deviation of the amplitude of each pixel in the time-series SAR image within the group to the mean value; using the amplitude deviation information of each pixel in the high coherence permanent scatterer class as the input of the clustering algorithm to cluster the low amplitude deviation permanent scatterer class from the high coherence permanent scatterer class; The method of obtaining a permanent scatterer with high phase stability includes: for a set of N Time-series SAR images, each of which is generated by interferometric processing N -1 interferometric phase image, calculate the correlation value of the time correlation of each pixel to obtain the phase stability information of each pixel in the high coherence permanent scatterer class; use the phase stability information of each pixel in the high coherence permanent scatterer class as the input of the clustering algorithm, and cluster the high phase stability permanent scatterer class from the high coherence permanent scatterer class; The method of identifying the target permanent scatterer point based on the low amplitude deviation permanent scatterer class and the high phase stability permanent scatterer class includes: taking the intersection of the low amplitude deviation permanent scatterer class set and the high phase stability permanent scatterer class set, and taking the permanent scatterer point in the intersection as the identified target permanent scatterer point.

2. The method according to claim 1, wherein: The step of executing the clustering algorithm adopts the Gaussian mixture model clustering algorithm.

3. A ground-based radar adaptive threshold permanent scatterer identification device using the method of claim 1, comprising: A processing module configured to perform grouping operations on the acquired time-series SAR images and obtain a set of candidate permanent scatterer points for each group of SAR images using an amplitude threshold method; A clustering module, configured to generate a coherence coefficient sequence based on the coherence coefficient of each pixel in the candidate permanent scatterer point set, so as to obtain a high coherence permanent scatterer class through a clustering algorithm; based on the amplitude deviation information and phase stability information of each pixel in the high coherence permanent scatterer class, obtain the corresponding low amplitude deviation permanent scatterer class and high phase stability permanent scatterer class through a clustering algorithm respectively; The identification module is configured to identify the target permanent scatterer points based on the low amplitude dispersion permanent scatterer class and the high phase stability permanent scatterer class.

4. A computer-readable storage medium having computer-executable instructions stored thereon, wherein when the computer-executable instructions are executed by a processor, the computer-executable instructions implement: A method for identifying permanent scatterers by adaptive threshold of a ground-based radar according to any one of claims 1 to 2.