CN-122017769-A - SAR soil moisture inversion method based on space neighborhood optimization

Abstract

The invention discloses a SAR soil moisture inversion method based on space neighborhood optimization, which relates to the field of polarized radar remote sensing quantitative inversion, and mainly comprises the following steps: preprocessing the full-polarization synthetic aperture radar image to obtain an observation coherent matrix, constructing a scattering model, carrying out optimization solving by utilizing a minimum residual power criterion to obtain a backscattering coefficient of surface scattering, constructing an improved Alpha approximation model, further obtaining the surface dielectric constant of each pixel, and carrying out moisture conversion and spatial averaging to obtain a final soil moisture estimated value of each pixel. By implementing the SAR soil moisture inversion method based on space neighborhood optimization, provided by the invention, the vegetation coverage soil moisture quantitative inversion supported by single-time-phase and ground-free measured data can be realized.

Inventors

• XIE QINGHUA
• XUE WENXIN
• PENG XING
• CHEN LEI

Assignees

• 中国地质大学（武汉）

Dates

Publication Date

20260512

Application Date

20251229

Claims (10)

1. 1. The SAR soil moisture inversion method based on the space neighborhood optimization is characterized by comprising the following steps of: s1, acquiring a full-polarization synthetic aperture radar image in a crop growth period, and preprocessing the full-polarization synthetic aperture radar image to obtain an observation coherence matrix; s2, constructing a scattering model according to the observation coherent matrix, and carrying out optimization solution according to the scattering model by utilizing a minimum residual power criterion to obtain a backscattering coefficient of surface scattering; S3, constructing an improved Alpha approximation model according to the backward scattering coefficient, and obtaining the surface dielectric constant of each pixel according to the improved Alpha approximation model; And S4, performing moisture conversion and spatial averaging on the surface dielectric constants of the pixels to obtain a final soil moisture estimated value of each pixel.
2. 2. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 1, wherein the preprocessing comprises single view complex number conversion, coherent matrix extraction, polarization filtering, geocoding, and data clipping.
3. 3. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 1, wherein the scattering model is as follows: , Wherein, the Representing an observation coherence matrix; Representing the scattering coefficient of the surface, Representing an X-Bragg surface scattering model coherence matrix; representing the volume scattering coefficient; Representing the coherence matrix of the volume scattering model, Representing the residual matrix.
4. 4. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 1, wherein the minimum residual power criterion is as follows: , Wherein, the Representation minimization; Representing the residual power; Representing the total scattered power; Representing the Bragg scattering coefficients for both H and V polarizations.
5. 5. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 1, wherein step S3 specifically comprises: s31 utilizing Scanning the full-polarization synthetic aperture radar image pixel by a sliding window to obtain the ratio of the backscattering coefficients of two adjacent pixels; s32, constructing an observation equation according to the ratio of the backscattering coefficients; s33, constructing a system of underdetermined equations based on a plurality of observation data of a sliding window according to the observation equations to obtain an improved Alpha approximation model; S34, according to the improved Alpha approximation model, based on the Dubois model, the dielectric constant range is constrained, the least square method is utilized for solving, the Alpha parameter matrix is obtained, and the surface dielectric constant of each pixel is obtained according to the Alpha parameter matrix.
6. 6. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 5, wherein the formula of calculation of the ratio of the backscatter coefficients is: , Wherein, the And Respectively representing two adjacent pixels in the sliding window And Is a backscatter coefficient of (2); And Respectively representing two adjacent pixels in the sliding window And Alpha coefficients of (a); representing radar incident angle; And Respectively representing two adjacent pixels in the sliding window And Is a soil dielectric constant of (a).
7. 7. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 5, wherein the improved Alpha approximation model is: , Wherein, the As a matrix of ratios of the back-scattering coefficients, Is Alpha parameter matrix.
8. 8. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 1, wherein the calculation formula of the moisture conversion is: , Wherein, the Is soil moisture; is the dielectric constant of the earth surface.
9. 9. The SAR soil moisture inversion method based on spatial neighborhood optimization of claim 1, wherein the calculation formula of the spatial average is: , Wherein, the Represent the first Line 1 Final soil moisture estimate for the pixels at the column position; Represent the first Line 1 An accumulation matrix of sums of dielectric constant solutions of pixels at column positions; First, the Line 1 The number of times the pel of the column position is calculated as the center of the sliding window.
10. 10. A computer program product comprising a computer program which, when executed by a processor, implements the steps of the SAR soil moisture inversion method of any one of claims 1-9, based on spatial neighborhood optimization.

Description

SAR soil moisture inversion method based on space neighborhood optimization Technical Field The invention relates to the field of polarized radar remote sensing quantitative inversion, in particular to a SAR soil moisture inversion method based on space neighborhood optimization. Background In agricultural production, soil moisture directly determines the moisture absorption efficiency of crop root systems, is a key factor for crop yield formation, is used as a core index for drought monitoring and a scientific basis for precise irrigation management, and has an irreplaceable effect on guaranteeing grain safety and high-efficiency water resource utilization. Synthetic Aperture Radar (SAR) is used as active microwave remote sensing, has penetration capability, all-weather observation and high resolution characteristics, and realizes quantitative inversion by establishing the relation between the radar backscattering coefficient and the soil dielectric constant. Radar signals are affected by radar system parameters, vegetation parameters, surface parameters and the like in a crop coverage scene. How to separate the vegetation scattering contribution and weaken the influence of surface roughness is a difficult problem of inversion of soil moisture in the crop area at the present stage. As a time sequence inversion method for effectively removing influences of vegetation and surface roughness, the Alpha change detection model has the advantage of being fast and efficient in soil moisture inversion of bare soil and low vegetation coverage. However, in the coverage area of crops, especially in the rapid growth period of crops, the short time sequence of vegetation condition is not easy to satisfy. Disclosure of Invention The invention aims to provide a SAR soil moisture inversion method based on space neighborhood optimization, which can realize quantitative inversion of vegetation coverage soil moisture supported by single-time-phase ground-free measured data. The invention provides a SAR soil moisture inversion method based on space neighborhood optimization, which comprises the following steps: s1, acquiring a full-polarization synthetic aperture radar image in a crop growth period, and preprocessing the full-polarization synthetic aperture radar image to obtain an observation coherence matrix; s2, constructing a scattering model according to the observation coherent matrix, and carrying out optimization solution according to the scattering model by utilizing a minimum residual power criterion to obtain a backscattering coefficient of surface scattering; S3, constructing an improved Alpha approximation model according to the backward scattering coefficient, and obtaining the surface dielectric constant of each pixel according to the improved Alpha approximation model; And S4, performing moisture conversion and spatial averaging on the surface dielectric constants of the pixels to obtain a final soil moisture estimated value of each pixel. The invention also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the SAR soil moisture inversion method based on space neighborhood optimization. The SAR soil moisture inversion method based on space neighborhood optimization has the following beneficial effects: Aiming at the limitation that the coupling effect of vegetation canopy scattering effect and surface roughness parameters is not fully considered by a traditional Alpha approximation model, the invention firstly adopts a generalized polarization SAR binary component decomposition method to separate and correct vegetation effect so as to obtain pure surface scattering components and realize accurate separation of the surface scattering components, then constructs a space neighborhood optimization Alpha algorithm, inputs the components into the Alpha approximation model, traverses the whole scene image pixel by using a 3X 3 sliding window, constructs a linear equation set based on the Alpha approximation model in each window, and establishes a underdetermined equation set with constraint conditions, namely The method comprises the steps of obtaining a space continuous inversion result by adopting a least square optimization method to solve the equation set by introducing a physical constraint condition of dielectric constant, designing an accumulation matrix and a counting matrix by an algorithm, carrying out aggregation treatment on the inversion result of all sliding windows, realizing soil volume water content estimation, and finally outputting a space continuous soil water distribution product. The invention converts the assumption of the traditional change detection method in the time dimension into the assumption of the single-time-phase space dimension, does not depend on time sequence data, converts the time sequence assumption into the space assumption, breaks through the dependence of the traditional method on multi-time-phase da