CN-122017755-A - GB-SAR atmospheric delay phase correction method based on cluster analysis

CN122017755ACN 122017755 ACN122017755 ACN 122017755ACN-122017755-A

Abstract

The invention provides a GB-SAR atmospheric delay phase correction method based on cluster analysis, which comprises the steps of collecting single vision complex images of a plurality of time phases by using GB-SAR equipment, generating a short base line interferogram pair, screening to obtain permanent scatterer points, carrying out model fitting on the permanent scatterer points based on preset large-scale atmospheric phase correction to obtain large-scale estimated values, carrying out clustering on the permanent scatterer points through space dimension clustering and time dimension clustering based on preset small-scale atmospheric phase correction to obtain space-time sample clusters, calculating or complementing the space-time sample clusters to obtain small-scale estimated values, and adding the large-scale estimated values and the small-scale estimated values to obtain total atmospheric phase estimated values. According to the invention, by combining the large-scale fitting advantage of the PSM method and the phase stacking thought of the CSS method, a homogeneous sample is obtained through space and time two-dimensional cluster analysis, and the accurate removal of the small-scale turbulent atmosphere is realized.

Inventors

ZHAO MINGQIN
LIAO MINGSHENG
LI HAOCHENG
Fan Shaobiao
CHENG DESHENG
LEI JINFENG
WEI HONGYI
WANG RUIQING
HAN QUANBIN
DONG JIE

Assignees

中国南水北调集团中线有限公司河南分公司

Dates

Publication Date: 20260512
Application Date: 20260122

Claims (10)

1. The GB-SAR atmospheric delay phase correction method based on cluster analysis is characterized by comprising the following steps of: acquiring single-view complex images of a plurality of time phases by using GB-SAR equipment, generating a short baseline interferogram pair, and screening to obtain permanent scatterer points; Performing model fitting on the permanent scatterer points based on preset large-scale atmospheric phase correction to obtain a large-scale estimated value; Based on the preset small-scale atmospheric phase correction, clustering the permanent scatterer points through space dimension clustering and time dimension clustering to obtain space-time sample clusters, and calculating or complementing the space-time sample clusters to obtain small-scale estimated values; and adding the large-scale estimated value and the small-scale estimated value to obtain a total atmospheric phase estimated value.
2. The method for correcting the atmospheric delay phase of the GB-SAR based on the cluster analysis of claim 1, wherein the steps of acquiring single vision complex images of a plurality of time phases by using the GB-SAR equipment, generating a pair of short base line interferograms, and screening to obtain permanent scatterer points comprise the following steps: Generating a short baseline interferogram pair from single-view complex images of a plurality of time phases according to a preset scanning mode of equipment; screening by adopting an amplitude dispersion index and a coherence threshold value to obtain a permanent scatterer point; And carrying out phase unwrapping on the short baseline interferogram pair to determine the whole cycle number of the winding phase, and outputting an unwrapped interference phase sequence.
3. The method for correcting the GB-SAR atmospheric delay phase based on cluster analysis according to claim 1, wherein model fitting the permanent scatterer points based on a preset large-scale atmospheric phase correction to obtain a large-scale estimated value comprises: Obtaining the distance between any permanent scatterer point and the radar, and obtaining a four-dimensional modeling polynomial function of the permanent scatterer point from the three-dimensional coordinates of any permanent scatterer point, the distance between any permanent scatterer point and the radar and the atmospheric property function at any moment; and solving the optimal weight parameter of the four-dimensional modeling polynomial function by using the coordinates of all the permanent scatterer points and the unwrapped interference phase sequence, and obtaining the atmospheric phase estimated value at any two moments.
4. The method for correcting the atmospheric delay phase of GB-SAR based on cluster analysis according to claim 1, wherein said permanent scatterer points are clustered by spatial dimension clustering, comprising: Taking any central moment as a reference, determining 2N symmetrical short baseline phases as first clustering samples; determining a clustering range, and searching a plurality of nearest permanent scatterer points around the current permanent scatterer point in the clustering range by using a KD tree method to form a first time-space sample cluster with similar phase sequences in a short time sequence range; if the number of samples of the first time-space sample cluster is greater than a first preset condition, determining that the space contains similar phase signals.
5. The method for clustering analysis-based GB-SAR atmospheric delay phase correction according to claim 4, wherein clustering the permanent scatterer points by time-dimensional clustering comprises: Taking the phases of the permanent scatterer points of the first time-space sample cluster in the number of the samples in the space dimension as clustering data to form 2N samples, and forming a second aggregate sample; taking the atmospheric phase obtained by stacking the first space-time sample cluster through the conventional phase as a clustering center of a second space-time sample cluster to obtain a second space-time sample cluster which is formed at the same time as the atmospheric component in the first cluster with the same current space, and determining that an average required stable condition is met if the number of samples of the second space-time sample cluster is not empty; Calculating phase stacking by adopting samples of the first time-space sample cluster and the second time-space sample cluster to obtain an atmospheric component average estimated value at any central moment; and obtaining the small-scale atmospheric estimated value in the interference diagram by using the difference between the average estimated values of the atmospheric components at any two central moments.
6. The method for correcting the atmospheric delay phase of the GB-SAR based on the cluster analysis according to claim 5, wherein calculating or complementing the space-time sample cluster to obtain a small-scale estimated value comprises: Classifying the space-time sample clusters according to a given morphological coefficient threshold by adopting an iterative K-Means classification method; if the number is smaller than the given morphological coefficient threshold, determining an optimal cluster number by adopting phase stacking calculation; if the first space-time sample cluster or the second space-time sample cluster is insufficient or the calculation condition is not met, performing tertiary interpolation complementation by using the calculated permanent scatterer points around, and outputting a small-scale estimated value.
7. The method for correcting the GB-SAR atmospheric delay phase based on cluster analysis according to claim 1, wherein adding the large-scale estimated value and the small-scale estimated value to obtain a total atmospheric phase estimated value comprises: and adding the large-scale estimated value and the small-scale estimated value, and then performing phase winding operation to obtain a total atmospheric phase estimated value.
8. The GB-SAR atmospheric delay phase correction system based on cluster analysis is characterized by comprising: The screening module is used for acquiring single vision complex images of a plurality of time phases by using the GB-SAR equipment, generating a short baseline interferogram pair, and screening to obtain permanent scatterer points; the first correction module is used for carrying out model fitting on the permanent scatterer points based on preset large-scale atmospheric phase correction to obtain a large-scale estimated value; The second correction module is used for carrying out clustering on the permanent scatterer points through space dimension clustering and time dimension clustering based on preset small-scale atmospheric phase correction to obtain space-time sample clusters, and calculating or completing the space-time sample clusters to obtain small-scale estimated values; and the synthesis module is used for adding the large-scale estimated value and the small-scale estimated value to obtain a total atmospheric phase estimated value.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements a cluster analysis based GB-SAR atmospheric delay phase correction method according to any one of claims 1 to 7 when executing the program.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements a cluster analysis based GB-SAR atmospheric delay phase correction method according to any one of claims 1 to 7.

Description

GB-SAR atmospheric delay phase correction method based on cluster analysis Technical Field The invention relates to the technical field of radar measurement, in particular to a GB-SAR atmospheric delay phase correction method based on cluster analysis. Background Foundation synthetic aperture radar has become one of the important tools in the field of surface deformation monitoring since the later development of the 20 th century. The technology realizes high-precision measurement of the ground surface micro displacement by comparing the phase differences of radar images in different time phases. Compared with the spaceborne InSAR, the Ground-based synthetic aperture radar (GB-Based Synthetic Aperture Radar SAR) has the remarkable advantages of flexible deployment, short image acquisition period (usually several minutes to tens of minutes) and high measurement accuracy (up to the sub-millimeter level), and has been widely applied to the monitoring of the slope stability of an open mine, the early warning of landslide and the deformation detection of artificial buildings such as bridges and dams. For example, in a surface mine scene, GB-SAR can monitor small displacement of a side slope in real time to prevent collapse accidents, in landslide monitoring, the GB-SAR can provide high-frequency monitoring data to support emergency response decisions, and in artificial building monitoring, small changes caused by bridge vibration or dam leakage can be detected. However, the sub-millimeter measurement accuracy of GB-SAR is deeply affected by atmospheric disturbances. Atmospheric delay phase effects are major sources of error, including large scale slowly varying atmospheric components (e.g., uniform changes due to global temperature or pressure gradients) and small scale turbulent atmospheric components (e.g., rapid fluctuations due to local vapor vortices or wind field disturbances). These atmospheric errors result from spatial and temporal inhomogeneities in the atmospheric refractive index, causing phase delays in the radar signal propagation path, superimposing spurious deformation amounts in the interference phase. Especially in the scene of high humidity area with a monitoring distance of several kilometers and changeable environment near river or with complex weather conditions, the error can reach several millimeters or more, which severely restricts the precision of practical engineering application. Domestic and foreign studies have verified the severity of atmospheric effects through manual deformation accuracy verification experiments, for example, some tests have shown that the atmosphere can introduce a deformation error of about 2 mm or a phase deviation of 1 rad. In the prior art, the atmospheric correction method is mainly divided into an external data method and an internal data method. The external data method relies on parameters such as temperature, pressure, humidity and the like acquired by a weather station, and simulates the change of the atmospheric refractive index through a linear or exponential model, but has the characteristics of low data space and time sampling rate and difficulty in meeting the small-range high-precision monitoring of GB-SAR. The internal data method is based on the information of the GB-SAR interference data itself, such as using the phase information of a permanent scatterer (PERMANENT SCATTERER, PS) or an artificial corner reflector, and estimating the atmosphere by using polynomial model fitting and spatial interpolation (hereinafter referred to as PSM method). The PSM method works well when removing large scale slowly varying atmospheres, but each interferogram is processed separately, time-dimensional information is not considered, and it is difficult to remove locally residual small scale turbulent atmospheres. The common co-date phase stacking method (hereinafter CSS method) in on-board InSAR removes turbulent components in the time dimension, but assuming that the deformation becomes linearly stable, the method is not suitable for the deformation experimental scenes of sudden deformation or manual control of the side slope common to GB-SAR, and the mutation information is lost. In summary, the defects of the prior art are mainly characterized in that (1) the sampling rate of an external data method is low, the dependence is strong, the real-time performance of GB-SAR cannot be adapted, (2) the PSM method is incomplete in removing small-scale atmosphere and affects the precision of residual errors, (3) the CSS method considers time dimension, but ignores deformation nonlinearity and easily loses high-frequency mutation information, and (4) the method for refining the fusion space-time dimension is lacking, so that the atmosphere removal and deformation protection cannot be simultaneously realized. These problems often lead to false alarms or reduced accuracy in practical engineering, and new methods are needed to solve the problems. Disclosure of Invention