Method and system for simulating radar backscattering coefficient of rice field in heading period

Abstract

The invention relates to the technical field of agricultural remote sensing and information, and particularly discloses a method and system for simulating a radar backscattering coefficient of a rice field in a heading period, and the method comprises the steps: obtaining geometric and physiological parameters of the rice field, and synchronously and actually measuring radar data; dividing the canopy into a spike leaf mixing layer, a stem leaf mixing layer and an underlying surface layer; five scattering mechanisms of ears, leaves and stems are decomposed; calculating component complex permittivity based on the model; constructing a fine three-dimensional geometric model of the ears, the leaves and the stems; obtaining a scattering matrix of each component through electromagnetic simulation; calculating an interlayer attenuation coefficient and a propagation constant by using the forward scattering matrix; calculating phase deviations under different scattering paths; and finally, performing coherence superposition on the backscattering contributions of the components in combination with phase deviation to obtain a total backscattering electric field and a coefficient. Through combination of component-level fine electromagnetic simulation and a scene-level radiation transfer theory, high-precision and high-efficiency scattering simulation is realized, and remote sensing monitoring of rice is effectively supported.

Description

Technical Field

[0002] This invention relates to the fields of agricultural remote sensing and information technology, and more specifically, to a method and system for simulating the radar backscattering coefficient of rice fields during the heading stage. Background Technology

[0004] Rice is one of the world's three major food crops and plays a vital role in my country's food structure. Rice primarily grows in cloudy and rainy areas, and synthetic aperture radar (SAR) offers advantages such as all-weather, all-day coverage and high penetration, providing an important technical means for large-scale rice monitoring. Radar backscattering coefficient is one of the most commonly used parameters in rice monitoring. Accurate simulation of this parameter is a prerequisite for correctly understanding the microwave scattering mechanism of the paddy field canopy and is the foundation for quantitative remote sensing applications such as rice mapping, parameter inversion (e.g., leaf area index, biomass, canopy height), and yield estimation. As the basic scattering unit of paddy fields, the panicle, leaves, and stems have complex geometric structures. Traditional radar backscattering coefficient simulation methods mostly model rice components using simple geometric shapes (e.g., cylinders, elliptical disks, rectangular thin sheets) and then approximate the scattering field using electromagnetic scattering theory. This simplification in modeling causes the rice components, especially the panicle, to deviate from their actual shape. As the most important component of rice, the parameters of the rice panicle are strongly correlated with the radar backscattering coefficient of the paddy field scene. However, few existing methods consider modeling the rice panicle, and the modeling of the rice panicle is often oversimplified (e.g., cylindrical), resulting in low simulation accuracy of the electromagnetic scattering characteristics of the rice panicle, which in turn leads to low simulation accuracy of the radar backscattering coefficient of paddy fields during the heading stage. Therefore, by establishing a three-dimensional geometric model of key rice components (such as panicle, leaf, and stem), we can perform detailed modeling of the complex geometric morphology of rice components, especially the rice panicle, and solve its electromagnetic scattering characteristics under different parameter combinations (such as waveband, polarization, incident angle, and azimuth angle). By integrating the electromagnetic simulation results of different components and the radiative transfer equation, we can simulate the propagation and scattering process of electromagnetic waves in the rice canopy, as well as the attenuation characteristics, and finally achieve high-precision simulation of the radar backscattering coefficient of paddy fields during the heading stage.

[0005] The backscattering coefficient of paddy field radar is mainly obtained through three methods: microwave scattering model, measurement experiment, and electromagnetic simulation.

[0006] (1) The microwave scattering model starts from the electromagnetic scattering theory and uses rigorous mathematical physics methods to establish a functional relationship between the radar backscattering coefficient and the geometric and physiological parameters of rice. Although this method has good universality, the simplification of rice component modeling leads to the distortion of its geometric structure, and therefore the simulation results of the electromagnetic scattering characteristics of the components are not accurate, which is the key factor limiting the simulation accuracy of the radar backscattering coefficient of rice fields.

[0007] (2) Measurement experiments can extract the backscattering coefficient of paddy field radar from scatterometer measurement or SAR imaging results, which can yield relatively reliable results. However, most measurement sensors operate under specific bands, polarizations, and incident angles, and can only obtain observation data under limited parameter combinations. Moreover, they focus on large-scale scene observations. Existing measurement methods and resolutions are insufficient for accurate measurement of rice plants and components.

[0008] (3) Electromagnetic simulation starts from the three-dimensional modeling of rice components and directly performs electromagnetic solutions on the three-dimensional models of rice components, plants, and scenes. It can simulate and calculate targets of any complex shape, and the calculation results can reach industrial-grade accuracy. The three-dimensional model of the rice scene can be regarded as an electrically large target (relative to wavelength). The computational load and memory requirements for electromagnetic simulation of it are large, which is difficult to meet the needs of practical applications. Summary of the Invention

[0010] To address the aforementioned technical problems in related technologies, this invention provides a method and system for simulating the radar backscattering coefficient of rice paddies during the heading stage, which can solve the above problems.

[0011] To achieve the above-mentioned technical objectives, the technical solution of the present invention is implemented as follows:

[0012] A method for simulating the radar backscattering coefficient of rice fields during the heading stage includes the following steps:

[0013] Data acquisition and processing: Select rice fields with uniform growth and heading stage, and conduct on-site measurements to obtain the geometric morphological structure and physiological characteristic parameters of rice components, plants, and scenes. The geometric morphological structure includes the geometric dimensions of panicles, stems, and leaves, and the plant growth spacing. The physiological characteristics include water content. Obtain the average value of multiple measurements of each parameter; simultaneously obtain the measurement results of the radar backscattering coefficient of the rice field scene.

[0014] Paddy field canopy layering: The paddy field scene is divided into three layers from top to bottom: the first layer L1 is the mixed layer of panicles and leaves, with a height of L1, including panicles and leaves; the second layer L2 is the mixed layer of stems and leaves, with a height of L2, located below L1, including stems and leaves; the third layer L3 is the underlying surface layer, located below L2, which is water or soil.

[0015] Scattering Mechanism Decomposition: Under the backscattering regime, corresponding to the first-order solution of the radiative transfer equation, the main scattering mechanism of the rice canopy is decomposed, yielding five different scattering mechanisms for the panicle, leaf, and stem, respectively. Figure 1As shown, scattering mechanism 1 represents direct backscattering, scattering mechanism 2 represents the incident wave being reflected by the underlying surface and then bistatically scattered by the scatterer, scattering mechanism 3 is the reverse of 2, scattering mechanism 4 represents the incident wave being reflected by the underlying surface, then backscattered by the scatterer, and then reflected a second time by the underlying surface, and scattering mechanism 5 represents forward scattering.

[0016] Simulation of dielectric properties of rice components: Using an empirical model (Debye-Cole bicolor dielectric model) relating dielectric constant to vegetation water content, the dielectric properties of rice panicles, leaves, and stems were simulated to obtain the complex dielectric constant of the rice components at different frequencies, expressed as:

[0017]

[0018] In the above formula, This indicates the non-dispersive residue. This indicates that under room temperature conditions ( The dielectric constant of free water, This represents the volume fraction of free water. This represents the dielectric constant of bound water. The volume fraction of bound water is expressed as follows:

[0019]

[0020]

[0021]

[0022]

[0023]

[0024] in, This indicates the weight and moisture content of the rice component. Indicates the frequency of electromagnetic waves. Represents the imaginary unit. It is the ionic conductivity. The value is 8.5‰, representing salinity.

[0025] Three-dimensional geometric modeling of rice components: The measured geometric morphological parameters of rice are used as input. Gaussian curves are used to simulate the central axis of rice panicle and leaves, ellipsoids are used to simulate grains, Hermite curves are used to simulate leaf outlines, and slender cylinders are used to simulate stems, so as to establish a detailed three-dimensional geometric model of rice panicle, leaves and stems.

[0026] Simulation of electromagnetic scattering characteristics of three-dimensional model of rice component: Using the three-dimensional geometric model of rice component and complex permittivity as input, the scattering matrix corresponding to five scattering mechanisms of each component is obtained by simulation using electromagnetic simulation software (such as FEKO, HFSS) under radar wave illumination of different frequencies, polarizations, incident angles and azimuth angles.

[0027] Electromagnetic wave attenuation characteristics simulation: The forward scattering matrix of different components in the rice canopy (corresponding to scattering mechanism 5) was obtained through simulation. Then, the forward scattering matrix of the scatterers in different layers was statistically averaged to obtain the attenuation coefficient of electromagnetic waves propagating in layers L1 and L2, which are expressed as follows:

[0028]

[0029]

[0030] in, Indicates the polarization mode for transmitting and receiving electromagnetic waves. Represents the wave number in free space. Indicates the radar wave illumination area. and These represent the attenuation coefficients of layer L1 and layer L2, respectively. , , These represent the number of spikes, leaves, and stems within the corresponding layer, respectively. , , These represent the spike, leaf, and stem along the direction of electromagnetic wave propagation, respectively. The scattering matrix elements, where < > denote the statistical averaging operation. The imaginary unit is used to calculate the propagation constants of electromagnetic waves in layers L1 and L2. , is represented as:

[0031] Spatial location simulation of rice components: Rice plants are evenly distributed along rows and columns, with individual plants randomly distributed around the center of each clump. First, the number of clumps and the coordinates of the clump centers are determined based on the paddy field scene parameters; then, the spatial coordinates of an individual plant within that clump are simulated; for an individual rice plant, with the panicle apex as the origin, the spatial coordinates of the panicle, leaves, and stems are simulated separately; finally, a coordinate system transformation is performed to unify the spatial positions of the panicle, leaves, and stems to the global coordinate system.

[0032] Phase shift calculation at different component locations: Since scattering mechanism 5 does not contribute to backscattering, the backscattering process of the paddy canopy includes a total of 12 scattering mechanisms. The propagation paths of these 12 scattering mechanisms within the paddy canopy can be obtained from... Figure 2It is composed of four basic propagation paths, and the lengths of these four basic propagation paths are as follows:

[0033]

[0034]

[0035]

[0036]

[0037] in, The imaginary unit, and These represent the attenuation coefficients of layer L1 and layer L2, respectively. This represents the spatial location vector of components within the paddy field canopy. This represents the direction vector of the incident wave in free space. Represents the unit direction vector of the incident wave along Components of the axis, This represents the angle of incidence. Based on the simulated spatial positions of each component in the canopy, the phase shift at different component positions in the canopy is calculated.

[0038] Scattering Mechanism Synthesis: Considering the phase shift at different component locations within the paddy field canopy, the scattered electric fields from different components and scattering mechanisms are coherently superimposed using the Coherent Superposition Approximation (CAA) to synthesize the total radar backscattered electric field of the paddy field, expressed as:

[0039]

[0040] in, Represents the natural constant. This indicates the distance from the radar antenna phase center to the rice paddy scene. Indicates the components of spike, leaf, and stem. Represents the scattering matrices of different components. Indicates the direction of the incident electromagnetic wave. Indicates the direction of the scattered electromagnetic wave. Indicates the corresponding angle of incidence and Fresnel reflection coefficient under polarization.

[0041] Simulation of radar backscattering coefficient in paddy fields: The final simulation yielded the following radar backscattering coefficient for paddy fields at the heading stage:

[0042]

[0043] Simulation accuracy verification and evaluation: The simulation results of the above model are verified and the accuracy is evaluated by measuring the radar backscattering coefficient of the rice paddy scene.

[0044] A radar backscattering coefficient system for simulating rice paddies in the heading stage includes:

[0045] The data acquisition and processing module is used to acquire and process measured parameters of paddy fields;

[0046] The stratification and decomposition module is used to perform canopy stratification and scattering mechanism decomposition;

[0047] The dielectric and geometry modeling module is used to simulate the dielectric properties of components and build three-dimensional geometric models;

[0048] The electromagnetic simulation module is used to simulate and calculate the scattering matrix of each component. The electromagnetic simulation module integrates the FEKO or HFSS electromagnetic simulation software interface.

[0049] The attenuation calculation module is used to calculate the attenuation coefficient and propagation constant of electromagnetic waves in the canopy;

[0050] The spatial location simulation module is used to simulate the spatial distribution of each component in the paddy field;

[0051] The phase shift calculation module is used to calculate the phase shift under each scattering path;

[0052] The scattering synthesis and coefficient calculation module is used to synthesize the total scattering electric field and calculate the backscattering coefficient.

[0053] The verification and evaluation module is used to compare simulation results with measured data to evaluate accuracy.

[0054] Beneficial effects of the present invention: The method of this application

[0055] Significantly improved simulation accuracy: By constructing a detailed three-dimensional geometric model of key components such as rice panicle, leaves, and stems, and utilizing this model and its precise dielectric properties in electromagnetic simulation, the complex spatial structure and electromagnetic properties of rice during the panicle stage can be realistically depicted, thus solving the problem of large errors in solving the scattering field caused by the simplification or distortion of geometric shapes in traditional methods.

[0056] A hybrid simulation method is proposed that effectively balances computational efficiency and simulation accuracy. This method combines component scattering data based on rigorous electromagnetic simulation with radiative transfer theory. First, the scattering matrix of key components is obtained through high-precision electromagnetic simulation. Then, it is organically embedded into the solution framework of the radiative transfer equation. This strategy retains the high-precision advantage of electromagnetic simulation in depicting complex and detailed scattering, while efficiently handling macroscopic rice paddy scenes containing a large number of scatterers through radiative transfer theory. Thus, while ensuring simulation accuracy, it effectively overcomes the bottleneck of huge computational load and memory requirements faced by traditional full-wave electromagnetic simulation methods when dealing with large-scale agricultural scenes, making the method both high-precision and practical.

[0057] A fine characterization of complex coherent scattering effects was achieved: by simulating the attenuation of electromagnetic waves in layers and accurately deriving the phase shift generated when electromagnetic waves propagate to panicle, stem, and leaf components at different spatial locations within the canopy, a coherent superposition method considering this phase shift was finally adopted when synthesizing the total scattered electric field. This fully takes into account the interference effect caused by the spatial distribution of components within the rice canopy, and can more realistically reflect the physical process of the interaction between radar waves and rice fields, thereby further improving the accuracy and reliability of the simulation results. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] The present invention will now be described in further detail with reference to the accompanying drawings.

[0061] Figure 1 This is a schematic diagram of the five main scattering mechanisms of the rice canopy in this invention;

[0062] Figure 2 This is a schematic diagram of the four basic propagation paths of electromagnetic waves in the rice canopy according to the present invention;

[0063] Figure 3 This is a flowchart of the method for simulating the radar backscattering coefficient of rice fields during the heading stage according to the present invention. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0066] Example 1:

[0067] This embodiment provides a method for simulating the radar backscattering coefficient of rice fields during the heading stage, such as... Figures 1-3 As shown, the specific steps include:

[0068] S1. Data Acquisition and Processing: A typical paddy field with uniform growth and in the heading stage (e.g., 15-20 days after heading) was selected as the research object. Using tools such as measuring tapes and calipers, at least 30 sample points were measured in the field to obtain the geometric morphological structural parameters of key rice components (panicle, stem, and leaves), including but not limited to: panicle length, panicle width, grain size and distribution density, stem height and diameter, leaf length, width, thickness, and curvature. Simultaneously, the weight moisture content of each component was measured using the drying method. Measure the row spacing and plant spacing (or clump spacing) of the plants. Take the average of multiple measurements for all parameters and use it as the input parameters for this scene. Simultaneously, use a ground-based or airborne radar system to acquire measured backscattering coefficient data for this paddy field scene at a specific frequency (e.g., L, C, X bands), polarization mode (HH, VV, HV), and incident angle range (e.g., 20°-50°) for subsequent model validation.

[0069] S2. Paddy Field Canopy Layering: Based on the measured rice plant height (H) and panicle position, the vertical structure of the paddy field is divided into three layers. The first layer, L1, is a mixed panicle-leaf layer, with a height d1 equal to the distance from the bottom to the top of the panicle. This layer mainly contains the rice panicles and a small number of upward-curving leaves located in the panicle area. The second layer, L2, is a mixed stem-leaf layer, with a height d2 = H - d1, located below L1. It mainly contains the stems and most of the drooping or spreading leaves. The third layer, L3, is the underlying surface layer, located below L2. Under water conditions, it is a water layer; after drainage, it becomes a moist or dry soil layer.

[0070] S3. Scattering Mechanism Decomposition: Based on the first-order solution of radiative transfer theory (i.e., considering only first-order scattering), the physical process of radar wave interaction with paddy field is analyzed. For each scatterer (ear, leaf, stem) within the canopy, its possible backscattering contribution mainly comes from four basic paths and their combinations, totaling 12 scattering mechanisms, such as... Figure 2As shown. Furthermore, a fifth scattering mechanism is defined as forward scattering. This mechanism does not directly contribute to the backscattered signal, but its scattering intensity is used to calculate the attenuation of electromagnetic waves during propagation through the canopy. The four basic propagation paths include: ① directly from the top of the canopy to the scatterer and back; ② from the top of the canopy to the scatterer, reflected by the underlying surface, and back; ③ reflected by the underlying surface to the scatterer, and then directly back; ④ reflected by the underlying surface to the scatterer, and then reflected by the underlying surface and back.

[0071] S4. Simulation of dielectric properties of rice components: The weight and water content of rice panicles, leaves, and stems measured in step S1 ( The input is substituted into the Debye-Körle two-color dispersion dielectric model for calculation. This model considers the vegetation dielectric constant as a mixture of free water, bound water, and dry matter. The specific calculation formulas are shown in formulas (1) to (6) in the invention content. Through this model, the target radar electromagnetic wave frequency ( Under these conditions, the complex complex permittivity of each component of rice ( ).

[0072] S5. 3D geometric modeling of rice components: Using the geometric parameters obtained in step S1, construct detailed models of each component using 3D modeling software (such as CAD) or through programming (such as MATLAB, Python):

[0073] Rice ear: A Gaussian curve is used to simulate the main axis of the ear, reflecting its natural curvature; each grain is simplified as an ellipsoid, randomly distributed along the main axis at a certain density and angle.

[0074] Blade: The midrib profile of the blade is defined by a cubic Hermite curve to control its bending and twisting; the blade surface is generated by a specific width curve (such as a rectangle or more complex shape) swept along the midrib, which can be given thickness.

[0075] Stem: Simplified as a straight or slightly curved slender cylinder.

[0076] These models are exported in standard formats (such as .stl, .step) for subsequent electromagnetic simulations.

[0077] S6. Simulation of Electromagnetic Scattering Characteristics of Rice Components: Import the 3D component model generated in step S5 and the complex permittivity calculated in step S4 into full-wave electromagnetic simulation software (such as FEKO, HFSS). Set a plane wave incident, covering the frequency and incident angle measured by the radar in step S1. ) and azimuth ( The simulation is performed individually for each component (ear, leaf, stem). By setting appropriate boundary conditions and solvers (such as the method of moments, physical optics, or high-frequency asymptotic method), the scattering field of the component in the directions corresponding to the five scattering mechanisms is calculated, and its scattering matrix elements are extracted.

[0078] S7. Simulation of Electromagnetic Wave Attenuation Characteristics: From the simulation results of step S6, extract the scattering matrix elements of all components in the forward scattering direction (i.e., the scattering angle equals the incident angle, and the scattering azimuth angle differs from the incident azimuth angle by 180 degrees). , , According to the Foldy approximation theory, the effective wavenumber (propagation constant) of electromagnetic waves in a random discrete medium is related to the average forward scattering amplitude of the scatterer. Statistical averaging was performed on the scatterers (ear / leaf, leaf / stem) in layers L1 and L2, and the attenuation coefficients of each layer for p-polarized waves were calculated using formulas (7) and (8). and Furthermore, the complex propagation constant of electromagnetic waves in each layer is obtained according to formula (9). .

[0079] S8. Spatial Position Simulation of Rice Components. Based on the measured row and column (clump) spacing, a two-dimensional grid is generated in the computer, with each grid point representing the center position of a clump of rice. For each clump, multiple plant positions are randomly generated within a certain radius around the center of the clump, based on the number of plants per clump. For a single plant, a local coordinate system is established with the panicle tip as the origin. Based on the relative positional relationships between components (such as leaf node position, leaf attachment angle, etc., which can be obtained statistically from measured data), the three-dimensional coordinates of each panicle (usually one), each leaf, and each stem node on that plant are determined in the local coordinate system. Finally, through translation and rotation, the local coordinates of all components are uniformly transformed into a global coordinate system with the scene center as the origin. This generates a precise set of spatial positions for all scatterers in the entire simulation scene.

[0080] S9. Phase shift calculation. For the 12 mechanisms contributing to backscattering, their propagation paths can all be decomposed into combinations of the aforementioned four basic paths. Specific calculation formulas are shown in formulas (10) to (13) in the invention description. For example, This represents the complex phase shift along the "direct-to-direct" path for components within layer L1, and it includes the free-space wavenumber. The determined geometric phase, and the attenuation portion of the electromagnetic wave as it passes through the L1 layer. The resulting phase shift and amplitude attenuation are calculated for each scatterer and each polarization to obtain the phase shift matrix.

[0081] S10. Scattering Mechanism Synthesis. The coherent superposition approximation (CAA) method is used to synthesize the total backscattered electric field of the entire paddy field scene. For each scatterer in the global coordinate system, according to its type (ear, leaf, stem) and level (L1 or L2), its corresponding backscattering matrix element (corresponding to the four contribution mechanisms) is retrieved from the scattering matrix database in step S6. These scattering matrix elements are multiplied by the phase offset factor of the corresponding path calculated in step S9 to obtain the contribution of the scatterer to the total scattered field. Simultaneously, the Fresnel reflection coefficient of the underlying surface is considered. Finally, the contributions of all scatterers under all effective scattering mechanisms are vector-superimposed (coherent superposition), as shown in Equation (14), to obtain the total scattered electric field. The core of this step is to take into account the relative phase interference effect caused by the spatial positional differences between different scatterers.

[0082] S11. Calculation of backscattering coefficient. Based on the definition of radar backscattering coefficient, the total scattered electric field calculated in step S10 is... The square of the amplitude is normalized to the incident power and the irradiated area A. The specific calculation formula is shown in (15), and the simulated radar backscattering coefficient is finally obtained. This coefficient is a function of frequency, polarization, and angle of incidence.

[0083] S12. Simulation Accuracy Verification and Evaluation. The simulation results obtained in step S11 are then used to verify and evaluate the accuracy. Compare the simulated and measured backscattering coefficients obtained in step S1 with the actual backscattering coefficients. Plot the curves of the simulated and measured values ​​as a function of the incident angle, as well as a scatter plot. Calculate evaluation indicators such as root mean square error (RMSE), coefficient of determination (R²), and mean bias. By analyzing these indicators, verify the accuracy and reliability of this method in simulating radar backscattering in rice paddies during the heading stage.

[0084] Example 2:

[0085] A system for simulating radar backscattering coefficients in rice paddies during the heading stage is provided to implement the method described in Example 1. This system can be integrated into a server, workstation, or high-performance computing cluster with computing capabilities and includes the following functional modules:

[0086] Data acquisition and processing module: configured to receive or input geometric, moisture content and measured radar data from the field, and perform preprocessing and averaging.

[0087] Layering and Decomposition Module: Configured to automatically divide the canopy into layers (L1, L2, L3) based on input parameters and determine various scattering mechanisms.

[0088] Dielectric and Geometric Modeling Module: Includes a built-in Debye-Kör dielectric model calculation submodule and a parametric 3D geometric modeling submodule for generating geometric model files and dielectric parameters for components.

[0089] Electromagnetic Simulation Module: This module is one of the core computing modules, with built-in interfaces to commercial electromagnetic simulation software (such as FEKO and HFSS) or integrated open-source electromagnetic solvers. It receives the geometric model and dielectric parameters, automatically sets the simulation task, and batch calculates and extracts the scattering matrices of each component.

[0090] Attenuation calculation module: configured to read forward scattering data from electromagnetic simulation results, perform statistical averaging according to formulas (7) and (8), and calculate the attenuation coefficient and propagation constant of each layer.

[0091] Spatial location simulation module: Configured to run a random distribution algorithm based on scene layout parameters to generate a list of precise locations of all scatterers in the global coordinate system.

[0092] Phase offset calculation module: configured to calculate the complex phase offset of all scatterer-path combinations according to formulas (10)-(13) based on the scatterer position, propagation constant and incident wave parameters.

[0093] Scattering synthesis and coefficient calculation module: This module is another core calculation module. It is configured to call the scattering matrix database and phase offset data, perform large-scale complex vector superposition operation according to the CAA principle and formula (14), and finally output the backscattering coefficient according to formula (15).

[0094] Verification and Evaluation Module: Configured to visualize and compare simulation results with measured data and evaluate quantitative accuracy, and generate a verification report.

[0095] Central control and database module (not shown in the diagram, but actually exists): used to coordinate the sequential execution of each module, and manage input parameters, intermediate data (such as scattering matrix library, position coordinates) and final results.

[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for simulating the radar backscattering coefficient of rice paddies during the heading stage, characterized in that, Includes the following steps: S1 Data Acquisition and Processing: Acquire the geometric morphological structure parameters, physiological characteristic parameters, and synchronously measured radar backscattering coefficients of the rice field in the heading stage to be simulated; S2 Paddy Field Canopy Layering: The paddy field canopy is divided from top to bottom into panicle and leaf mixed layer L1, stem and leaf mixed layer L2, and underlying surface layer L3; S3 Scattering Mechanism Decomposition: Based on the first-order solution of the radiative transfer equation, five scattering mechanisms corresponding to the spike, leaf, and stem components are obtained respectively, including direct backscattering, underlying surface reflection-scatterer bistatic scattering, scatterer bistatic scattering-underlying surface reflection, underlying surface reflection-scatterer backscattering-underlying surface reflection, and forward scattering. Among them, the forward scattering mechanism is used to characterize the electromagnetic wave propagation attenuation, and the other scattering mechanisms are used to characterize the backscattering contribution. Simulation of dielectric properties of S4 rice components: Based on the obtained water content parameters of rice components, the complex dielectric constants of panicles, leaves and stems at different frequencies were calculated using an empirical model of dielectric constant and water content. 3D geometric modeling of S5 rice components: Using the acquired geometric morphology parameters, detailed 3D geometric models of rice panicles, leaves, and stems are established respectively. Simulation of electromagnetic scattering characteristics of S6 rice component: Based on the three-dimensional geometric model and complex permittivity, the scattering matrix of each component under different radar system parameters corresponding to the five scattering mechanisms is calculated using electromagnetic simulation software. S7 Electromagnetic wave attenuation characteristic simulation: Based on the forward scattering matrix, the attenuation coefficient of electromagnetic wave propagation in L1 and L2 layers is obtained by statistical averaging, and then the propagation constant is calculated. S8 Rice Component Spatial Position Simulation: Based on the parameters of the rice paddy scene, the spatial position of all panicle, leaf, and stem components in the rice paddy is simulated and determined in the global coordinate system; S9 Phase shift calculation: Based on the propagation path of each scattering mechanism, the spatial position of the components, and the attenuation coefficient obtained in step S7, calculate the phase shift of different components in the canopy under each scattering path. S10 Scattering Mechanism Synthesis: Considering the phase shift, the scattering electric fields of different components and the scattering mechanism corresponding to the backscattering contribution are coherently superimposed to synthesize the total radar backscattering electric field of the paddy field. S11 Backscattering Coefficient Calculation: Based on the total radar backscattering electric field, the simulated radar backscattering coefficient is calculated.

2. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, In step S2, the spike-leaf mixture layer L1 contains spikes and leaves, and the height of layer L1 is d1. The stem-leaf mixture layer L2 contains stems and leaves, and the height of layer L2 is d2. It is located below layer L1. The underlying surface layer L3 is a water layer or a soil layer, and it is located below layer L2.

3. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, In step S4, the empirical model for the dielectric constant and water content is the Debye-Cole two-color dispersion dielectric model. Using this model, the dielectric properties of rice panicles, leaves, and stems are simulated to obtain the complex dielectric constant of rice components at different frequencies, expressed as: In the above formula, Indicates the non-dispersive residue. This indicates that under room temperature conditions ( The dielectric constant of free water, This represents the volume fraction of free water. This represents the dielectric constant of bound water. The volume fraction of bound water is expressed as follows: in, This indicates the weight and moisture content of the rice component. Indicates the frequency of electromagnetic waves. Represents the imaginary unit. It is the ionic conductivity. The value is 8.5‰, representing salinity.

4. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, In step S5, the three-dimensional geometric modeling includes: simulating the central axis of the rice panicle with a Gaussian curve, simulating the grains with an ellipsoid, simulating the leaf outline with a Hermite curve, and simulating the stem with a slender cylinder.

5. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, The formula for calculating the attenuation coefficient in step S7 is as follows: in, Indicates the polarization mode for transmitting and receiving electromagnetic waves. Represents the wave number in free space. Indicates the area illuminated by radar waves. and These represent the attenuation coefficients of layer L1 and layer L2, respectively. , , These represent the number of spikes, leaves, and stems within the corresponding layer, respectively. , , These represent the spike, leaf, and stem along the direction of electromagnetic wave propagation, respectively. The scattering matrix elements, where < > denote the statistical averaging operation. The imaginary unit is used to calculate the propagation constants of electromagnetic waves in layers L1 and L2. , represented as: 。 6. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, The spatial location simulation in step S8 is as follows: the uniform row and column distribution of rice plants is determined by taking the clump as the unit, a single plant is randomly distributed in each clump, and a local coordinate system of the plant is established with the panicle apex as the origin, and finally transformed to the global coordinate system.

7. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, In step S9, the phase offset is calculated based on the following four basic propagation path length combinations: in, The imaginary unit, and These represent the attenuation coefficients of layer L1 and layer L2, respectively. This represents the spatial location vector of components within the paddy field canopy. This represents the direction vector of the incident wave in free space. Represents the unit direction vector of the incident wave along Components of the axis, The incident angle is represented by the phase shift at different component positions in the canopy, calculated based on the spatial positions of each component obtained from the simulation.

8. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, In step S10, the coherent superposition approximation is used to coherently superimpose the scattered electric fields, and the total radar backscattered electric field is expressed as: in, Represents the natural constant. This indicates the distance from the radar antenna phase center to the rice paddy scene. Indicates the components of spike, leaf, and stem. Represents the scattering matrices of different components. Indicates the direction of the incident electromagnetic wave. Indicates the direction of the scattered electromagnetic wave. Indicates the corresponding angle of incidence and Fresnel reflection coefficient under polarization.

9. The method for simulating the radar backscattering coefficient of rice paddies during the heading stage according to claim 1, characterized in that, In step S11, the radar backscattering coefficient of the rice field at the heading stage is finally obtained from the simulation as follows: 。 10. A system for implementing the method according to any one of claims 1-8, characterized in that, include: The data acquisition and processing module is used to acquire and process measured parameters of paddy fields; The stratification and decomposition module is used to perform canopy stratification and scattering mechanism decomposition; The dielectric and geometry modeling module is used to simulate the dielectric properties of components and build three-dimensional geometric models; The electromagnetic simulation module is used to simulate and calculate the scattering matrix of each component; The attenuation calculation module is used to calculate the attenuation coefficient and propagation constant of electromagnetic waves in the canopy; The spatial location simulation module is used to simulate the spatial distribution of each component in the paddy field; The phase shift calculation module is used to calculate the phase shift under each scattering path; The scattering synthesis and coefficient calculation module is used to synthesize the total scattering electric field and calculate the backscattering coefficient. The verification and evaluation module is used to compare simulation results with measured data to evaluate accuracy. The electromagnetic simulation module integrates FEKO or HFSS electromagnetic simulation software interfaces.

Images