CN-121995341-A - SAR water depth inversion method based on self-adaptive elliptic window function

CN121995341ACN 121995341 ACN121995341 ACN 121995341ACN-121995341-A

Abstract

The invention discloses a SAR water depth inversion method based on a self-adaptive elliptic window function, which is characterized in that by analyzing the energy distribution characteristics of a main peak of a two-dimensional spectrum of a SAR subgraph, the self-adaptive adjustment is realized in the directions of the elliptic window, the ratio of the long axis to the short axis and the window function type, so that the directivity expression of the main peak of the frequency spectrum and the accuracy of the wavelength extraction are obviously improved. The method can remarkably improve the definition of the main peak of the frequency spectrum and the inversion precision of the water depth, and provides a more stable and accurate technical path for SAR water depth inversion in a complex near-shore environment.

Inventors

JIA DONGZHEN
XU ZONGHAO
Guo Jiongfu
HE XIUFENG
SONG MINFENG
ZHANG HUI

Assignees

河海大学

Dates

Publication Date: 20260508
Application Date: 20260311

Claims (10)

1. The SAR water depth inversion method based on the self-adaptive elliptic window function is characterized by comprising the following specific steps: Step 1, acquiring SAR images, priori water depth information and measured water depth data, and carrying out unified projection, spatial registration and vertical reference integrated treatment; Step 2, sliding clipping is carried out on the SAR image obtained in the step 1 according to a set step length, so as to form a sub-image sequence for spectrum analysis; Step 3, performing fast Fourier transform FFT calculation on each sub-graph obtained in the step 2, identifying a main energy connected domain and representing the main wave direction and spectrum shape characteristics; Step 4, for each sub-graph obtained in the step 2, anisotropic elliptical windowing aligned with the corresponding main wave direction obtained in the step 3 is performed; And 5, performing FFT (fast Fourier transform) calculation on the windowed map obtained in the step 4 to obtain a spectrogram, extracting the position of the main peak and the corresponding wavelength and direction information thereof, and completing water depth inversion by utilizing a linear dispersion relation.
2. The method according to claim 1, wherein in the step of performing anisotropic elliptical windowing aligned with the corresponding main wave direction obtained in step 3, the main direction angle of the elliptical window to be added is obtained by: Calculating the centroid position and the second-order center matrix of the main energy connected domain obtained in the step 3, and quantifying the distribution characteristics of the spectrum energy in different directions; And extracting the characteristic vector direction corresponding to the maximum characteristic value of the second-order center matrix through characteristic decomposition, and taking the characteristic vector direction as the main axis direction of a main peak of the frequency spectrum, namely the main direction angle of the elliptical window.
3. The method of claim 2, wherein the step of obtaining the long and short axes of the elliptical window comprises: , , In the formula, a and b respectively represent a major axis and a minor axis, area represents a pixel area of a main energy communication domain, H and W respectively represent a height and a width of a corresponding sub-graph, and e represents an eccentricity of a main peak of a frequency spectrum.
4. The method of claim 1, wherein the added elliptical window is shaped according to a peak-to-bandwidth ratio PBR: When PBR < alpha, a Blackman window is selected as a basis function to construct an elliptical window; when alpha is less than or equal to PBR < beta, a Hamming window is selected as a basis function, and an elliptical window is constructed; when PBR is more than or equal to beta, a Hanning window is adopted as a basis function, and an elliptical window is constructed; Wherein, alpha and beta are both preset PBR threshold values.
5. The method of claim 4, wherein the peak-to-bandwidth ratio PBR is defined as the ratio of spectral dominant peak energy to global average energy: , , , Wherein, the As the energy of the main peak of the wave, As a background energy of the light, And (3) with Respectively positive and negative frequency neighborhoods of the main peak, Representing the frequency coordinates, Ω being the spectrally efficient domain, The power spectral density of the spectrum is represented, Indicating removal And (3) with The remaining regions in the post spectrum.
6. The method of claim 5, wherein the step of determining the position of the probe is performed, , wherein, Is centered on the origin and has a radius Is a small disc low frequency exclusion zone.
7. The method according to claim 1, wherein the step of extracting the position of the main peak and the corresponding wavelength and direction information in the step 5 includes: The main energy connected domain with the largest area is extracted by utilizing binary segmentation and area connectivity analysis of the energy position of the main peak in the spectrogram, and the geometric centroid coordinates of the main energy connected domain are calculated ; Extracting a second maximum peak from the residual spectrum after removing the main peak, and calculating the aggregate centroid coordinates of the corresponding region ; Calculating the wavelength corresponding to the main peak Wave direction angle : , , In the formula, , M and N represent the length and width of the spectrogram.
8. The method according to claim 7, wherein the water depth inversion is performed in step 5 by using a linear dispersion relation, and the following inversion model is constructed: , in the formula, Expressed in terms of wavelength Sum wave period The water depth function value is an independent variable, and g is the gravity acceleration.
9. A computer readable storage medium having stored thereon a computer program, characterized in that the computer program when executed by a processor implements the steps of a SAR water depth inversion method based on an adaptive elliptical window function as described above.
10. An electronic device, comprising: A memory for storing a computer program; A processor for implementing the steps of the SAR water depth inversion method based on the adaptive elliptic window function as described above when executing the computer program.

Description

SAR water depth inversion method based on self-adaptive elliptic window function Technical Field The invention relates to an SAR water depth inversion method based on a self-adaptive elliptic window function. Background The shallow sea high-precision water depth information is basic supporting data for ocean development and treatment, and has important significance for port channel planning and maintenance, coast disaster toughness assessment, coastal zone refined management and the like. The traditional acoustic sounding with single beam or multiple beams has advantages in point location precision, but is limited by cost, efficiency and operation conditions in large-scale, rapid updating and near-shore complex terrain scenes, and is difficult to meet the requirements of high space-time resolution and high updating frequency. The optical remote sensing water depth inversion method can realize effective inversion in clear water, but is highly dependent on water transparency, is easily influenced by factors such as suspended matters, substrate change, water disturbance, cloud layers and the like, and has limited applicability under turbid or changeable sea conditions. When the swell propagates to the shore, the wavelength can be shortened under the action of shallowing and refraction, and the information of the swell wavelength can be mapped into the information of local water depth according to a linear dispersion equation. The synthetic aperture radar (SYNTHETIC APERTURE RADAR, SAR) has the capabilities of day and night, imaging under the cloud and wide revisiting, and can provide information such as surge wavelength, wave direction and the like under moderate wind fields and imaging geometric conditions, so as to provide observation input for offshore sounding. The current mainstream method is mostly based on Two-dimensional fast fourier transform (Two-Dimensional Fast Fourier Transform, 2D FFT) to locate main peaks from SAR image spectrum, extract main wavelengths, and back-push water depth. However, the spectrum is susceptible to noise, edge effects, and the like, resulting in unstable spectrum quality and blurred main peak positioning. Meanwhile, the wave period is difficult to directly acquire by an image, is often replaced by a fixed or empirical value, and is easy to introduce systematic deviation under different water depths and topography conditions, so that inversion accuracy is reduced. The above problems are particularly pronounced in the shallow offshore areas, breaker zones and areas where the surge-shore interactions are significant. Therefore, reducing the uncertainty of the spectrum, reducing the dependence on fixed period assumptions, and realizing robust estimation of wavelength and wave period is one of the key challenges of current SAR water depth inversion. In order to inhibit spectrum leakage and improve main peak resolvability, partial research is performed on applying weighted windows to SAR images in a spectrum calculation stage, namely window functions with different weights are applied to a receiving end, so that spectrum performance is balanced, and spectrum performance is improved. The traditional window functions, such as rectangular windows and various cosine windows, are widely applied to analysis of sea wave spectrums, but have limited capability of inhibiting leakage and side lobes in a scene with obvious near-shore boundary effect and nonuniform wave field, and main peaks are still easy to diffuse and unstable in positioning, so that water depth inversion deviation is introduced. Therefore, it is difficult to guarantee the spectrum quality under complex conditions by relying on a fixed window alone, and it is necessary to introduce more adaptive window function design and parameter adjustment. On the other hand, to reduce the reliance on fixed cycle assumptions, it is desirable to accurately reflect the offshore dynamics process in the physical modeling. The mapping relationship between period and wavelength is no longer stable under the influence of near-shore swell-shore flow interaction and a broken band, so that systematic deviation is easy to occur based on inversion of a fixed wave period. Zhang et al review shows that the coupling of waves to the flow field can significantly change the wave velocity and propagation characteristics, thereby affecting the wavelength extraction and sounding accuracy in SAR images. The higher-order dispersion relation proposed by Ge and the like further indicates that, especially under medium water depth conditions, the conventional linear dispersion relation has difficulty in accurately describing the wave propagation variation, thereby limiting inversion accuracy. Such physical-based improvements promote reliability of period and phase velocity estimation, but cannot replace quality control at the spectrum end, and under complex hydrodynamic conditions, the adaptability of fixed period assumptions is still insufficient. Therefore, ho