Volume scattering model determination device and volume scattering model determination method

Abstract

This volume scattering model determination device is configured so as to comprise: a covariance matrix acquiring unit (1) that acquires a covariance matrix corresponding to a scattering matrix of an observation target obtained by alternately radiating two mutually orthogonal polarized waves toward the observation target; and a parameter setting unit (2) that sets, as model parameters of a generalized volume scattering model having the scattering matrices of a plurality of scattered bodies mapped thereto, a plurality of values of model parameters including parameters respectively indicating the scattering matrix elements of each scattered body, the distribution of the plurality of scattered bodies, and the average tilt angle of the plurality of scattered bodies, and outputs a plurality of generalized volume scattering models including the model parameters that were set. The volume scattering model determination device further comprises: a coefficient adjusting unit (3) that multiplies each of the plurality of generalized volume scattering models output by the parameter setting unit (2) by a fitting coefficient and adjusts each fitting coefficient such that the difference between the covariance matrix and each of the generalized volume scattering models after being multiplied by the fitting coefficient decreases; and a model selecting unit (4) that compares the plurality of differences after the fitting coefficient adjustment by the coefficient adjusting unit (3) to one another, and selects a generalized volume scattering model from among the plurality of generalized volume scattering models output by the parameter setting unit (2) on the basis of the results of comparing the differences.

Description

Volume scattering model determination device and volume scattering model determination method

[0001] The present disclosure relates to a volume scattering model determination device and a volume scattering model determination method.

[0002] There is a volume scattering model determination device that determines a volume scattering model from scattering data obtained by irradiating a polarized wave toward an observation target. The volume scattering model is a model that represents radio wave scattering from multiple scatterers dispersed in space. As such a volume scattering model determination device, for example, Patent Document 1 discloses a device that determines an ungeneralized volume scattering model using scattering components included in the scattering data. The ungeneralized volume scattering model is a model in which scatterers are fixed to dipoles, and the distribution of the multiple scatterers and the average tilt angle of the multiple scatterers are both fixed.

[0003] Japanese Patent Application Laid-Open No. 2020-180865

[0004] The device disclosed in Patent Document 1 has the problem that if the scatterer is not a dipole, if the distribution of multiple scatterers dispersed in space differs from a fixed distribution, or if the average tilt angle of the multiple scatterers differs from a fixed average tilt angle, it may not be possible to correctly obtain a volume scattering model that simulates the features of the scatterers.

[0005] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a volume scattering model determination device that can obtain a generalized volume scattering model that simulates the features of a scatterer.

[0006] a parameter setting unit that sets a plurality of model parameter values ​​as model parameters of a generalized volume scattering model onto which the scattering matrices of a plurality of scatterers are mapped, the model parameter values ​​including parameters indicating elements of the scattering matrices of each of the scatterers, the distribution of the plurality of scatterers, and the average tilt angle of the plurality of scatterers, and outputs a plurality of generalized volume scattering models including the set model parameters. The volume scattering model determination device also includes a coefficient adjustment unit that multiplies each of the plurality of generalized volume scattering models output from the parameter setting unit by a fitting coefficient and adjusts each fitting coefficient so that a difference between each generalized volume scattering model and the covariance matrix after multiplication by the fitting coefficient is small, and a model selection unit that compares the plurality of differences after adjustment of the fitting coefficients by the coefficient adjustment unit and selects one of the plurality of generalized volume scattering models output from the parameter setting unit based on the comparison result of the differences.

[0007] According to the present disclosure, a generalized volume scattering model that simulates the feature quantities of a scatterer can be obtained.

[0008] FIG. 1 is a configuration diagram showing a volume scattering model determination device according to embodiment 1. FIG. 2 is a hardware configuration diagram showing hardware of the volume scattering model determination device according to embodiment 1. FIG. 3 is a hardware configuration diagram of a computer when the volume scattering model determination device is realized by software, firmware, or the like. FIG. 4 is a flowchart showing a volume scattering model determination method, which is a processing procedure of the volume scattering model determination device. FIG. 5 is a flowchart showing a volume scattering model determination method according to embodiment 1. ｇ and [C ０ ] eigenvalue λ １ , λ ２ , λ ３ 8A is an explanatory diagram showing the relationship between <[Cm]> and <[Cv ｇ (α, β, σ, ψ ０) ]> and the fitting coefficient x ｇ and [C ０ ] eigenvalue λ １ , λ ２ , λ ３ FIG. 8B is an explanatory diagram showing the relationship between <[Cm]> and <[Cv ｇ (α, β, σ, ψ ０ ) ]> and the fitting coefficient x ｇ and [C ０ ] eigenvalue λ １ , λ ２ , λ ３ 1 is an explanatory diagram showing the relationship between ０ = (A + B + C) / (a ​​+ b + c) is an explanatory diagram showing that the sum of the eigenvalues ​​becomes zero. ｇ Fig. 1 is an explanatory diagram showing an example of a PS (Polarmetric Signature) shape of a feature simulated by Fig. 2. Fig. 2 is a configuration diagram showing a volume scattering model determination device according to a second embodiment. Fig. 3 is a hardware configuration diagram showing hardware of a volume scattering model determination device according to a second embodiment.

[0009] In order to explain the present disclosure in more detail, embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0010] Embodiment 1. Fig. 1 is a configuration diagram showing a volume scattering model determination device according to embodiment 1. Fig. 2 is a hardware configuration diagram showing the hardware of the volume scattering model determination device according to embodiment 1. The volume scattering model determination device shown in Fig. 1 comprises a covariance matrix acquisition unit 1, a parameter setting unit 2, a coefficient adjustment unit 3, and a model selection unit 4.

[0011] The covariance matrix acquisition unit 1 is realized, for example, by a covariance matrix acquisition circuit 11 shown in FIG. 2 . The covariance matrix acquisition unit 1 acquires a covariance matrix Cm corresponding to a scattering matrix [S] of an observation target obtained by alternately irradiating the observation target with two mutually orthogonal polarized waves. The observation target contains one or more scatterers dispersed in space. The covariance matrix acquisition unit 1 outputs the covariance matrix Cm to the coefficient adjustment unit 3.

[0012] The parameter setting unit 2 is realized by, for example, a parameter setting circuit 12 shown in Fig. 2. The parameter setting unit 2 sets parameters α and β (see formula (1) below) indicating elements of the scattering matrix of each scatterer, a parameter σ indicating the distribution of the plurality of scatterers, and a parameter ψ indicating the average tilt angle of the plurality of scatterers as model parameters of the generalized volume scattering model Cv onto which the scattering matrix [S] of the plurality of scatterers is mapped. ０ The model parameters (α, β, σ, ψ ０ ) are set in multiple ways. The generalized volume scattering model Cv is a model that allows the distribution of multiple scatterers or the average tilt angle of multiple scatterers to be set arbitrarily. The model parameter α is an element of the scattering matrix [S], and the model parameter α = |α|exp(jψ α The range of |α| is 0≦|α|≦+1. α ≦+180. The model parameter β is an element of the scattering matrix [S], and the model parameter β=|β|exp(jψ β The range of |β| is 0≦|β|≦+1. β The model parameter σ is a parameter that indicates the distribution of multiple scatterers, and the value range of the model parameter σ is 0≦σ≦(π ２ / 12) １／２ The model parameter ψ ０ is a parameter indicating the average tilt angle of multiple scatterers, and the model parameter ψ ０ The value range of is -90≦ψ ０ ≦+90. The parameter setting unit 2 sets the model parameters (α, β, σ, ψ ０ ) to the coefficient adjustment unit 3 and the model selection unit 4. Here, the model parameters included in the generalized volume scattering model Cv are (α, β, σ, ψ ０ However, this is only an example, and the generalized volume scattering model Cv is (α, β, σ, ψ ０ ) may contain different model parameters.

[0013] The coefficient adjustment unit 3 is realized by, for example, the coefficient adjustment circuit 13 shown in FIG. 2 . The coefficient adjustment unit 3 acquires the covariance matrix Cm from the covariance matrix acquisition unit 1, and acquires multiple generalized volume scattering models Cv from the parameter setting unit 2. The coefficient adjustment unit 3 multiplies each of the multiple generalized volume scattering models by a fitting coefficient x. The coefficient adjustment unit 3 adjusts each fitting coefficient x so that the difference between each generalized volume scattering model Cv and the covariance matrix Cm after multiplication by the fitting coefficient becomes small. The coefficient adjustment unit 3 outputs multiple differences after adjusting the fitting coefficients to the model selection unit 4 as the differences between each generalized volume scattering model Cv and the covariance matrix Cm.

[0014] The model selection unit 4 is realized by, for example, the model selection circuit 14 shown in Fig. 2. The model selection unit 4 acquires a plurality of generalized volume scattering models Cv from the parameter setting unit 2, and acquires a plurality of differences after fitting coefficient adjustment from the coefficient adjustment unit 3. The model selection unit 4 compares the plurality of differences with each other, and selects one of the generalized volume scattering models Cv based on the comparison result of the plurality of differences.

[0015] In Fig. 1, it is assumed that each of the components of the volume scattering model determination device, namely, a covariance matrix acquisition unit 1, a parameter setting unit 2, a coefficient adjustment unit 3, and a model selection unit 4, is realized by dedicated hardware such as that shown in Fig. 2. That is, it is assumed that the volume scattering model determination device is realized by a covariance matrix acquisition circuit 11, a parameter setting circuit 12, a coefficient adjustment circuit 13, and a model selection circuit 14. Each of the covariance matrix acquisition circuit 11, the parameter setting circuit 12, the coefficient adjustment circuit 13, and the model selection circuit 14 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof.

[0016] The components of the volume scattering model determination device are not limited to those realized by dedicated hardware, and the volume scattering model determination device may be realized by software, firmware, or a combination of software and firmware. The software or firmware is stored in the memory of a computer as a program. The computer refers to hardware that executes a program, and includes, for example, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, a processor, or a DSP (Digital Signal Processor).

[0017] 3 is a hardware configuration diagram of a computer when the volume scattering model determination device is realized by software, firmware, etc. When the volume scattering model determination device is realized by software, firmware, etc., a program for causing a computer to execute the respective processing procedures of the covariance matrix acquisition unit 1, parameter setting unit 2, coefficient adjustment unit 3, and model selection unit 4 is stored in memory 21. Then, a processor 22 of the computer executes the program stored in memory 21.

[0018] 2 shows an example in which each of the components of the volume scattering model determination device is realized by dedicated hardware, and Fig. 3 shows an example in which the volume scattering model determination device is realized by software, firmware, etc. However, this is merely an example, and some of the components in the volume scattering model determination device may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, etc.

[0019] Next, the operation of the volume scattering model determination device shown in Fig. 1 will be described. Fig. 4 is a flowchart showing a volume scattering model determination method, which is a processing procedure of the volume scattering model determination device. Fig. 5 is an explanatory diagram showing an example of a generalized volume scattering model Cv on which the scattering matrices [S] of five scatterers are mapped. The generalized volume scattering model Cv shown in Fig. 5 is mapped with the scattering matrix [S] of a left-handed scatterer that rotates to the left, the scattering matrix [S] of a dextrorotatory scatterer that rotates to the right, the scattering matrix [S] of a dihedral CR scatterer that is a dihedral corner reflector, the scattering matrix [S] of either a dipole scatterer or a wire scatterer, and the scattering matrix [S] of a spherical scatterer, a planar scatterer, or a trihedral CR scatterer. The scattering matrix [S] of a left-handed scatterer, the scattering matrix [S] of a right-handed scatterer, and the scattering matrix [S] of a scatterer such as a trihedral CR are isotropic scatterers, while the scattering matrix [S] of a bihedral CR scatterer and the scattering matrix [S] of a scatterer such as a dipole are anisotropic scatterers. The scattering matrix [S] of each scatterer contains elements represented by α and β, as shown in the following formula (1). In particular, the scattering matrix [S] of a left-handed scatterer is expressed as shown in the following formula (2), and the scattering matrix [S] of a dextrorotatory scatterer is expressed as shown in the following formula (3). Furthermore, the scattering matrix [S] of a bihedral CR scatterer is expressed as shown in the following formula (4), the scattering matrix [S] of a scatterer such as a dipole is expressed as shown in the following formula (5), and the scattering matrix [S] of a scatterer such as a trihedral CR is expressed as shown in the following formula (6).

[0020]

[0021] For example, the scattering matrix [S ｄｉｐｏｌｅ ] is expressed by the following equation (7): As shown in FIG. 7, the scattering matrix [S ｄｉｐｏｌｅ[(ψ)] is expressed using a rotation matrix [R(ψ)] as shown in the following equation (8). Fig. 6 is an explanatory diagram showing a dipole tilted vertically in the polarization plane. Fig. 7 is an explanatory diagram showing a dipole tilted by ψ in the polarization plane. In Figs. 6 and 7, the horizontal axis indicates the direction of H polarization, and the vertical axis indicates the direction of V polarization.

[0022]

[0023] The scattering matrix of the scatterer [S ｄｉｐｏｌｅ (ψ)], the covariance matrix [C ｄｉｐｏｌｅ (ψ)] is expressed as the following equation (9).

[0024]

[0025] Such scatterers are distributed as ｖ Covariance matrix C when there are an infinite number of covariances according to (ψ) ｖ is expressed as the following equation (10): ｖ By varying (ψ), it is possible to represent a large number of scatterer groups. ｖ By replacing with the simpler index ρ and solving the integral in equation (10), a generalized volume scattering model Cv can be obtained.

[0026]

[0027] The covariance matrix acquisition unit 1 acquires a covariance matrix Cm corresponding to a scattering matrix [S] of an observation target obtained by alternately irradiating the observation target with two mutually orthogonal polarized waves. Specifically, when the observation target is, for example, a plurality of observation areas AR １ ~ A.R. Ｎ When the covariance matrix acquisition unit 1 includes a plurality of observation areas AR １ ~ A.R. Ｎ In order, one of the observation areas AR ｎ The covariance matrix acquisition unit 1 acquires a covariance matrix Cm corresponding to the scattering matrix [S] of the observation area AR (step ST1 in FIG. 4 ). n=1, . . . , N, where N is an integer equal to or greater than 1. ｎThe covariance matrix Cm corresponding to the scattering matrix [S] is output to the coefficient adjustment unit 3. Here, one covariance matrix C ０ is expressed as the following equation (11).

[0028]

[0029] covariance matrix C ０ can be calculated from one pixel and corresponds to one scatterer. If a group of surrounding pixels is used to average multiple scatterers, the covariance matrix C corresponding to the scattering matrix [S] can be calculated as shown in Equation (12) below.

[0030] In equation (12), for example, M is 100 (10 pixels vertically by 10 pixels horizontally).

[0031] The parameter setting unit 2 sets model parameters (α, β, σ, ψ) included in the generalized volume scattering model Cv onto which the scattering matrices [S] of a plurality of scatterers are mapped. ０ Specifically, the parameter setting unit 2 sets a plurality of values ​​for the model parameters (α, β, σ, ψ) included in the generalized volume scattering model Cv. ０ ) within the range of possible values ​​of the model parameters (α, β, σ, ψ ０ ) are set in G ways (step ST2 in FIG. 4), where G is an integer equal to or greater than 2. The parameter setting unit 2 sets G values ​​of the set model parameters (α, β, σ, ψ ０ ) G generalized volume scattering models Cv １ ~CV Ｇ are output to the coefficient adjustment unit 3 and the model selection unit 4, where g=1, . . . , G.

[0032] The coefficient adjustment unit 3 acquires the covariance matrix Cm from the covariance matrix acquisition unit 1, and calculates G generalized volume scattering models Cv １ ~CV Ｇ The coefficient adjustment unit 3 obtains the generalized volume scattering model Cv ｇ (g = 1, ..., G) with fitting coefficient x ｇ(g = 1, ..., G) Generalized volume scattering model Cv ｇ is expressed as a matrix.

[0033] In formula (13), C ０ is the excess component during fitting, and C ０ The smaller is the fitting accuracy.

[0034] The coefficient adjustment unit 3 calculates the generalized volume scattering model Cv ｇ and the difference ΔC between the covariance matrix Cm ｇ [C ０ ] is small, the fitting coefficient x ｇ (Step ST3 in FIG. 4). In the volume scattering model determination device shown in FIG. 1, the coefficient adjustment unit 3 adjusts [C ０ ] is the fitting coefficient x ｇ The following is required.

[0035]

[0036] Generally, <[Cm]> and <[Cv ｇ (α, β, σ, ψ ０ )]> and often do not match. Fitting coefficient x ｇ As the increase in [C ０ ] monotonically decreases and becomes negative as shown in Figure 8A. ｇ (α, β, σ, ψ ０ ) ]> and the fitting coefficient x ｇ and [C ０ ] eigenvalue λ １ , λ ２ , λ ３ In the example of FIG. 8A, three eigenvalues ​​λ １ , λ ２ , λ ３ Among them, the eigenvalue λ ３ becomes negative at x' at first, and the sum of all eigenvalues ​​at that time is λ １ +λ ２ On the other hand, <[Cm]> and <[Cv ｇ (α, β, σ, ψ０ ) ]> is the case where [C ０ ] becomes a zero matrix, then [C ０ ] eigenvalue λ １ , λ ２ , λ ３ are simultaneously zero at x=1, as shown in Figure 8B. Figure 8B shows that <[Cm]> and <[Cv ｇ (α, β, σ, ψ ０ ) ]> and the fitting coefficient x ｇ and [C ０ ] eigenvalue λ １ , λ ２ , λ ３ FIG.

[0037] From the above, we can see that various model parameters ｇ (α, β, σ, ψ ０ ) ]> and create a ｇ (α, β, σ, ψ ０ ) ]> and <[Cm]>, the <[Cv ｇ (α, β, σ, ψ ０ )】>, then [C ０ ] eigenvalue λ １ , λ ２ , λ ３ The sum of [C ０ ] eigenvalue λ １ , λ ２ , λ ３ In order to analytically examine, equation (13) is expressed as equation (15) below.

[0038]

[0039] The determinant representing the eigenvalues ​​of this matrix is ​​expressed as in the following equation (16).

[0040]

[0041] Due to the relationship between the solution of the cubic equation and the coefficients, the sum of the three eigenvalues ​​can be calculated as shown in the following equation (17).

[0042]

[0043] The covariance matrix is ​​<[Cv ｇ (α, β, σ, ψ ０ The diagonal elements of [Cm] and [Cm] are always positive real numbers. Therefore, as shown in FIG. 9, equation (17) becomes a linear function with a negative slope, and x ０ = (A + B + C) / (a ​​+ b + c), the sum of the eigenvalues ​​becomes zero. ０ = (A + B + C) / (a ​​+ b + c), the sum of the eigenvalues ​​is zero. ｇ (α, β, σ, ψ ０ ) ]> is the same, then x = x ０ = 1, and [C ０ ] is zero, and each eigenvalue is zero. ｇ (α, β, σ, ψ ０ )]>, considering that the eigenvalues ​​move away from zero as the difference between x and x increases, ０ By evaluating the index that accumulates the absolute values ​​of each eigenvalue in ｇ (α, β, σ, ψ ０ )】>, or [Cv] similar to <[Cm]> ｇ (α, β, σ, ψ ０ ) ]> can be obtained by applying the square root to the sum of the eigenvalues ​​shown in equation (17), as shown in the following equation (18), the evaluation function G ｅ (α, β, σ, ψ ０ ) is defined, the evaluation function G ｅ (α, β, σ, ψ ０ ) is minimized.

[0044]

[0045] This index can also be calculated using the relationship between the solution of the cubic equation and the coefficients, as shown in the following equation (19).

[0046]

[0047] Using equation (17), equation (19) is rewritten as follows: Note that the eigenvalues ​​of a Hermitian matrix are all real numbers, and therefore their simple squares are always positive real numbers.

[0048]

[0049] This gives x = x ０ In this case, the evaluation function G ｅ (α, β, σ, ψ ０ ) can be obtained.

[0050]

[0051] The coefficient adjustment unit 3 adjusts the evaluation function G ｅ (α, β, σ, ψ ０ ) is calculated. ｅ (α, β, σ, ψ ０ ) is the generalized volume scattering model Cv ｇ (g = 1, ..., G) and the covariance matrix Cm. ｇ The fitting coefficient x ｇ and the covariance matrix Cm is the generalized volume scattering model Cv ｇ If it matches, the evaluation function G ｅ (α, β, σ, ψ ０ ) = 0. The fitting coefficient x ｇ and the covariance matrix Cm, and the generalized volume scattering model Cv ｇ The larger the difference between ｅ (α, β, σ, ψ ０ The coefficient adjustment unit 3 adjusts the fitting coefficient x ｇ G differences ΔC after adjusting １ ~ΔC Ｇ All the evaluation functions G calculated according to the formula (21) are ｅ (α, β, σ, ψ ０ ) is output to the model selection unit 4.

[0052] The model selection unit 4 receives G differences ΔC from the coefficient adjustment unit 3. １ ~ΔC Ｇ As a result, all evaluation functions G ｅ (α, β, σ, ψ０ ) and obtains G generalized volume scattering models Cv １ ~CV Ｇ The model selection unit 4 obtains all the evaluation functions G ｅ (α, β, σ, ψ ０ ) are compared with each other. ｅ (α, β, σ, ψ ０ ) with each other, G differences ΔC １ ~ΔC Ｇ The model selection unit 4 compares the G differences ΔC １ ~ΔC Ｇ Based on the comparison results, G generalized volume scattering models Cv １ ~CV Ｇ (Step ST4 in FIG. 4). Specifically, the model selection unit 4 selects one of the generalized volume scattering models from G differences ΔC １ ~ΔC Ｇ Among them, for example, the minimum difference ΔC ｍｉｎ Then, the model selection unit 4 selects G generalized volume scattering models Cv １ ~CV Ｇ The smallest difference ΔC identified from ｍｉｎ Generalized volume scattering model Cv ｇ Select .

[0053] In the volume scattering model determination device shown in FIG. 1, the model selection unit 4 selects G generalized volume scattering models Cv １ ~CV Ｇ Among these, the smallest difference ΔC ｍｉｎ Generalized volume scattering model Cv ｇ However, this is merely an example, and the model selection unit 4 may select a difference ΔC other than the minimum within a range that does not cause practical problems. ｍｉｎ Generalized volume scattering model Cv ｇ The model selection unit 4 may select, for example, the second smallest difference ΔC ｍｉｎ Generalized volume scattering model Cv ｇ may be selected.

[0054] The generalized volume scattering model Cv selected by the model selection unit 4ｇ is a set of multiple observation areas AR １ ~ A.R. Ｎ Any of the observation areas AR ｎ This is a generalized volume scattering model for all observation areas AR １ ~ A.R. Ｎ The processing of steps ST1 to ST4 in FIG. 4 is repeated until a generalized volume scattering model for is obtained.

[0055] The generalized volume scattering model Cv selected by the model selection unit 4 ｇ is the observation area AR ｎ The generalized volume scattering model Cv ｇ 10 is an explanatory diagram showing an example of a PS shape of a feature simulated by the generalized volume scattering model Cv. In FIG. 10, ψ is the orientation angle, and χ is the elliptically angle. FIG. 10 illustrates the PS shape when the feature is an urban area, the PS shape when the feature is vegetation, and the PS shape when the feature is the sea surface. FIG. 10 also illustrates the PS shape when the transmission polarization and the reception polarization are parallel polarization (Co-Pol), and the PS shape when the transmission polarization and the reception polarization are orthogonal polarization (X-Pol). Therefore, for example, an image processing device (not shown) may simulate the PS shape of a feature when the generalized volume scattering model Cv ｇ By analyzing the PS shape of the feature simulated by the , the type of feature can be identified.

[0056] In the first embodiment described above, the volume scattering model determination device is configured to include: a covariance matrix acquisition unit 1 that acquires a covariance matrix corresponding to the scattering matrix of an observation target obtained by alternately irradiating the observation target with two mutually orthogonal polarized waves; and a parameter setting unit 2 that sets a plurality of values ​​of model parameters as model parameters of a generalized volume scattering model onto which the scattering matrices of a plurality of scatterers are mapped, the model parameters including parameters indicating the elements of the scattering matrix of each scatterer, the distribution of the plurality of scatterers, and the average tilt angle of the plurality of scatterers, and outputs a plurality of generalized volume scattering models including the set model parameters. The volume scattering model determination device also includes: a coefficient adjustment unit 3 that multiplies each of the plurality of generalized volume scattering models output from the parameter setting unit 2 by a fitting coefficient and adjusts each fitting coefficient so that a difference between each generalized volume scattering model and the covariance matrix after multiplication by the fitting coefficient is small; and a model selection unit 4 that compares the plurality of differences after adjustment of the fitting coefficients by the coefficient adjustment unit 3 and selects one of the plurality of generalized volume scattering models output from the parameter setting unit 2 based on the comparison result of the differences. Therefore, the volume scattering model determination device can obtain a generalized volume scattering model that simulates the characteristics of the scatterer.

[0057] Second Embodiment In a second embodiment, a volume scattering model determination device including a parameter setting unit 5 will be described.

[0058] Fig. 11 is a configuration diagram showing a volume scattering model determination device according to embodiment 2. In Fig. 11, the same reference numerals as in Fig. 1 indicate the same or corresponding parts, and detailed description thereof will be omitted. Fig. 12 is a hardware configuration diagram showing the hardware of the volume scattering model determination device according to embodiment 2. In Fig. 12, the same reference numerals as in Fig. 2 indicate the same or corresponding parts, and detailed description thereof will be omitted. The volume scattering model determination device shown in Fig. 11 comprises a covariance matrix acquisition unit 1, a parameter setting unit 5, a coefficient adjustment unit 3, and a model selection unit 4.

[0059] The parameter setting unit 5 is realized by, for example, a parameter setting circuit 15 shown in Fig. 12. The parameter setting unit 5 sets, as model parameters of the generalized volume scattering model Cv onto which the scattering matrix [S] of a plurality of scatterers is mapped, parameters α and β indicating elements of the scattering matrix of each scatterer, a parameter σ indicating the distribution of the plurality of scatterers, and a parameter ψ indicating the average tilt angle of the plurality of scatterers. ０ The model parameters (α, β, σ, ψ ０ Specifically, the parameter setting unit 5 sets a plurality of values ​​of the model parameters (α, β, σ, ψ) related to the anisotropic scatterer. ０ The anisotropy parameters (σ, ψ) included in ０ ) compared to the step size of the model parameters (α, β, σ, ψ) for an isotropic scatterer. ０ The anisotropy parameters (σ, ψ) included in ０ The model parameters (α, β, σ, ψ) are adjusted so that the step size becomes coarse. ０ The parameter setting unit 5 sets a plurality of values ​​for the set model parameters (α, β, σ, ψ ０ ) to the coefficient adjustment unit 3 and the model selection unit 4, respectively.

[0060] In Fig. 11, it is assumed that each of the components of the volume scattering model determination device, namely, the covariance matrix acquisition unit 1, the parameter setting unit 5, the coefficient adjustment unit 3, and the model selection unit 4, is realized by dedicated hardware such as that shown in Fig. 12. In other words, it is assumed that the volume scattering model determination device is realized by a covariance matrix acquisition circuit 11, a parameter setting circuit 15, a coefficient adjustment circuit 13, and a model selection circuit 14. Each of the covariance matrix acquisition circuit 11, the parameter setting circuit 15, the coefficient adjustment circuit 13, and the model selection circuit 14 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination of these.

[0061] The components of the volume scattering model determination device are not limited to those realized by dedicated hardware, and the volume scattering model determination device may be realized by software, firmware, or a combination of software and firmware. When the volume scattering model determination device is realized by software, firmware, or the like, programs for causing a computer to execute the respective processing procedures of the covariance matrix acquisition unit 1, parameter setting unit 5, coefficient adjustment unit 3, and model selection unit 4 are stored in memory 21 shown in Fig. 3. Then, processor 22 shown in Fig. 3 executes the programs stored in memory 21.

[0062] 12 shows an example in which each of the components of the volume scattering model determination device is realized by dedicated hardware, while Fig. 3 shows an example in which the volume scattering model determination device is realized by software, firmware, etc. However, this is merely an example, and some of the components in the volume scattering model determination device may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, etc.

[0063] Next, the operation of the volume scattering model determination device shown in Fig. 11 will be described. However, apart from the parameter setting unit 5, the volume scattering model determination device is the same as that shown in Fig. 1. Therefore, only the operation of the parameter setting unit 5 will be described here.

[0064] In the example of FIG. 5 , the scattering matrix [S] of a left-handed scatterer, the scattering matrix [S] of a right-handed scatterer, and the scattering matrix [S] of a scatterer such as a three-sided CR are isotropic scatterers, while the scattering matrix [S] of a two-sided CR scatterer and the scattering matrix [S] of a scatterer such as a dipole are anisotropic scatterers. Isotropic scatterers have lower sensitivity to rotation than anisotropic scatterers. For this reason, the model parameters (α, β, σ, ψ) related to the isotropic scatterer are ０ ) in the anisotropic model parameters (σ, ψ ０ ) even if the step size is set roughly, there is little problem in obtaining appropriate model parameters.

[0065] Therefore, the parameter setting unit 5 sets the model parameters (α, β, σ, ψ) related to the anisotropic scatterer. ０ The anisotropy parameters (σ, ψ) included in ０ ) compared to the step size of the model parameters (α, β, σ, ψ) for an isotropic scatterer. ０ The anisotropy parameters (σ, ψ) included in ０ The model parameters (α, β, σ, ψ) are adjusted so that the step size becomes coarse. ０ The parameter setting unit 5 sets a plurality of values ​​of the model parameters (α, β, σ, ψ) that are different from each other. ０ ) G generalized volume scattering models Cv １ ~CV Ｇ are output to the coefficient adjustment unit 3 and the model selection unit 4, respectively.

[0066] In the above-described second embodiment, the volume scattering model determination device shown in Fig. 11 is configured so that the parameter setting unit 5 sets a plurality of model parameter values ​​so that the step size of the anisotropic parameter values ​​included in the model parameters related to isotropic scatterers is coarser than the step size of the anisotropic parameter values ​​included in the model parameters related to anisotropic scatterers. Therefore, the volume scattering model determination device shown in Fig. 11 can obtain a generalized volume scattering model that simulates the characteristics of scatterers, just like the volume scattering model determination device shown in Fig. 1, and can also reduce the processing load on the coefficient adjustment unit 3 more than the volume scattering model determination device shown in Fig. 1.

[0067] 1 and 11 , the covariance matrix acquisition unit 1 acquires a covariance matrix Cm corresponding to the scattering matrix [S] of the object to be observed, and outputs the covariance matrix Cm to the coefficient adjustment unit 3. However, this is merely an example, and the covariance matrix acquisition unit 1 may, for example, convert the covariance matrix Cm into a coherency matrix Tm using a transformation matrix L as shown in the following equation (22), and output the coherency matrix Tm as the covariance matrix Cm to the coefficient adjustment unit 3. In other words, the matrix output from the covariance matrix acquisition unit 1 to the coefficient adjustment unit 3 is not limited to the covariance matrix Cm, and may be any matrix that can be converted from the covariance matrix Cm using a transformation matrix.

[0068] In equation (22), T is a mathematical symbol indicating transposition.

[0069] In addition, the present disclosure allows for free combination of the respective embodiments, modification of any of the components of the respective embodiments, or omission of any of the components of the respective embodiments.

[0070] The present disclosure is suitable for a volume scattering model determination device and a volume scattering model determination method.

[0071] REFERENCE SIGNS LIST 1 Covariance matrix acquisition unit, 2 Parameter setting unit, 3 Coefficient adjustment unit, 4 Model selection unit, 5 Parameter setting unit, 11 Covariance matrix acquisition circuit, 12 Parameter setting circuit, 13 Coefficient adjustment circuit, 14 Model selection circuit, 15 Parameter setting circuit, 21 Memory, 22 Processor.

Claims

a parameter setting unit that sets a plurality of model parameter values ​​as model parameters of a generalized volume scattering model on which the scattering matrices of a plurality of scatterers are mapped, the model parameter values ​​including parameters indicating elements of the scattering matrix of each scatterer, a distribution of the plurality of scatterers, and an average tilt angle of the plurality of scatterers, and outputs a plurality of generalized volume scattering models including the set model parameters; a coefficient adjustment unit that multiplies each of the plurality of generalized volume scattering models output from the parameter setting unit by a fitting coefficient, and adjusts each fitting coefficient so that a difference between each generalized volume scattering model after multiplication by the fitting coefficient and the covariance matrix becomes small; and a model selection unit that compares the plurality of differences after adjustment of the fitting coefficient by the coefficient adjustment unit, and selects one of the generalized volume scattering models from the plurality of generalized volume scattering models output from the parameter setting unit based on a comparison result of the differences.

2. The volume scattering model determination device according to claim 1, characterized in that the parameter setting unit sets multiple values ​​of the model parameters within a range of possible values ​​of the model parameters included in the generalized volume scattering model to which the scattering matrices of the multiple scatterers are mapped.

3. The volume scattering model determination device according to claim 1, characterized in that the generalized volume scattering model in which the model parameter values ​​are set by the parameter setting unit has a scattering matrix of an isotropic scatterer and a scattering matrix of an anisotropic scatterer mapped thereto, the generalized volume scattering model includes, as model parameters, parameters indicating elements of the scattering matrix and anisotropic parameters, and the parameter setting unit sets a plurality of values ​​of the model parameters so that the increments of the values ​​of the anisotropic parameters included in the model parameters related to the isotropic scatterer are coarser than the increments of the values ​​of the anisotropic parameters included in the model parameters related to the anisotropic scatterer.

4. A volume scattering model determination method, in which a covariance matrix acquisition unit acquires a covariance matrix corresponding to a scattering matrix of an observation target obtained by alternately irradiating the observation target with two mutually orthogonal polarized waves; a parameter setting unit sets a plurality of values ​​of model parameters as model parameters of a generalized volume scattering model onto which the scattering matrices of a plurality of scatterers are mapped, the model parameters including parameters indicating elements of the scattering matrix of each scatterer, a distribution of the plurality of scatterers, and an average tilt angle of the plurality of scatterers, and outputs a plurality of generalized volume scattering models including the set model parameters; a coefficient adjustment unit multiplies each of the plurality of generalized volume scattering models output from the parameter setting unit by a fitting coefficient and adjusts each fitting coefficient so that a difference between each generalized volume scattering model after multiplication by the fitting coefficient and the covariance matrix becomes small; and a model selection unit compares the plurality of differences after the fitting coefficient adjustment by the coefficient adjustment unit with each other, and selects one of the generalized volume scattering models from the plurality of generalized volume scattering models output from the parameter setting unit based on a comparison result of the differences.

Images