Decomposition method of π / 4 reduced polarimetric interferometric synthetic aperture radar

Abstract

The present invention discloses a π / 4 reduced polarimetric interferometric synthetic aperture radar decomposition method, comprising: acquiring polarimetric interferometric SAR image data, converting the full polarimetric SAR data into reduced polarimetric SAR data according to the conversion relationship between the full polarimetric SAR data and the reduced polarimetric SAR data; calculating the optimal parameters of polarimetric interferometry, determining the optimization factor of the volume scattering model; optimizing the volume scattering model according to the volume scattering model optimization factor; and deriving the expression forms of the body scattering energy, the surface scattering energy and the dihedral angle scattering energy according to the optimized volume scattering model. The present invention extends the reduced polarimetric target decomposition method to the reduced polarimetric interferometric target decomposition algorithm, improving the current situation that the existing reduced polarimetric interferometric SAR decomposition method is relatively scarce. Compared with the reduced polarimetric target decomposition algorithm, the problem of over-estimation of the volume scattering energy in the inclined building area is significantly improved, and the decomposition result is more consistent with the actual ground object scattering mechanism.

Description

Technical Field

[0001] The invention relates to a target decomposition technology in the field of reduced polarization synthetic aperture radar, and in particular to a π / 4 reduced polarization interference synthetic aperture radar decomposition method. Background Art

[0002] Polarimetric Synthetic Aperture Radar (PolSAR) is one of the important branches of Synthetic Aperture Radar (SAR), and plays an important role in many fields such as geology, hydrology and environment. As an important branch of polarimetric SAR, reduced polarimetric SAR can avoid the shortcomings of full polarimetric SAR and provide more information than single-polarization and dual-polarization modes. Polarimetric target decomposition technology, as one of the important processing methods of reduced polarimetric SAR images, can directly describe the scattering mechanism of the target, but the problem of over-estimation of volume scattering energy in existing algorithms still limits the effectiveness of the algorithm. Polarimetric interferometry SAR technology adds interference information on the basis of the original polarimetric SAR technology, which can obtain more accurate target decomposition results. However, due to the complexity of the system and the complexity of the algorithm, there are few existing polarimetric interferometry decomposition algorithms, with low computational efficiency and poor robustness. Reduced polarimetric interferometry decomposition technology is in a blank stage, and the application of reduced polarimetric interferometry decomposition methods is still a technical challenge. Summary of the invention

[0003] In view of this, the present invention proposes a π / 4 reduced polarization interferometric synthetic aperture radar decomposition method, the method comprising:

[0004] S1: Acquire polarimetric interferometric SAR image data, and convert the full polarimetric SAR data into reduced polarimetric SAR data according to the conversion relationship between the full polarimetric SAR data and the reduced polarimetric SAR data;

[0005] S2: Calculate the optimal parameters of polarization interferometry and determine the optimization factors of the volume scattering model;

[0006] S3: Optimizing the volume scattering model according to the volume scattering model optimization factor;

[0007] S4: deriving the expressions of body scattering energy, surface scattering energy and dihedral angle scattering energy according to the optimized body scattering model.

[0008] In the above scheme, the method further comprises:

[0009] The two fully polarimetric SAR images acquired in the same area are pre-processed by filtering and multi-viewing, and one of the images is selected as the main image and the other as the secondary image. Subsequently, the fully polarimetric SAR data is converted into π / 4 reduced polarimetric SAR data by using the conversion relationship between the fully polarimetric SAR data and the π / 4 reduced polarimetric SAR data for subsequent processing. The specific implementation method of this step is as follows:

[0010] The ensemble average values ​​of the coherence matrices of the two fully polarized SAR images are obtained as <[T1]> and <[T2]> respectively:

[0011]

[0012] Among them, T ij (i=1,2,3;j=1,2,3) and T′ ij (i=1,2,3; j=1,2,3) represent the i-th row and j-th column elements of the ensemble average values ​​of the primary and secondary full polarization coherence matrices <[T1]> and <[T2]>, respectively; the superscript “H” represents the conjugate transpose of the matrix; k1 and k2 are the Pauli bases of the primary and secondary full polarization images, respectively, and their expressions are and Among them, S HHi (i=1,2) is the backscatter of vertical transmission and vertical reception, S HVi (i=1,2) is the backscatter with vertical transmission and horizontal reception, S VVi (i=1,2) is the backscatter with horizontal transmission and horizontal reception.

[0013] G=(g0,g1,g2,g3) T is the Stokes vector, g k (k=0,1,2,3) represents the element of the kth row of the Stokes vector. The corresponding relationship between the Stokes vector and the fully polarized SAR data is:

[0014]

[0015] The Stokes vectors of the main and secondary images can be expressed as:

[0016]

[0017] Among them, g km (k=0,1,2,3) represents the Stokes vector G m The element of the kth row of g ks (k=0,1,2,3) represents the Stokes vector G s The element of the k-th row of .

[0018] In the above scheme, the method further comprises:

[0019] Polarization interference information can be represented by a 6×6 coherence matrix. The 6×6 semi-positive Hermitian matrix defined by the outer product of k1 and k2 is expressed as:

[0020]

[0021] in,

[0022]

[0023] <·> is the multi-view average. [T m ] and [T s ] is a semi-positive Hermitian matrix, which contains all the polarization information obtained by each group of antennas, [Ω 12 ] is a non-Hermitian matrix, containing polarization information and interference information.

[0024] In polarimetric interferometry SAR, the interference coherence coefficient and phase information are strongly dependent on the polarization state. Similar to the definition of polarimetric similarity parameter, the unit projection vectors ω1 and ω2 are introduced, and the Pauli basis scattering vectors k1 and k2 are projected onto ω1 and ω2 respectively. The polarimetric interferometry similarity parameter (P PISP ) is:

[0025]

[0026] Among them, P PISP ∈[0,1], |·| represents the absolute value operation, [ω1]=[ω 1,opt1 ,ω 1,opt2 ,ω 1,opt3 ] and [ω2] = [ω 2,opt1 ,ω 2,opt2 ,ω 2,opt3 ] are the optimal scattering matrices composed of the optimal scattering vectors of the main and auxiliary images, ω i,optj (i=1,2;j=1,2,3) is the jth element of the optimal scattering matrix of the primary or secondary image. Parameter P PISP The value of is unique and does not change with the change of the target orientation angle. This parameter can be used to ensure the optimal coherence between the two images. Therefore, the optimal parameter P for polarization interference is constructed. PIOP For subsequent decomposition processing, the optimal parameter P of polarization interference PIOP The expression is:

[0027]

[0028] In the above scheme, the method further comprises:

[0029] Using the optimal parameter P of polarization interferometry PIOP , optimize the volume scattering model proposed by Freeman. The expression of the volume scattering model proposed by Freeman is:

[0030]

[0031] The optimal parameter P of polarization interferometry PIOP The expression of the optimized volume scattering model is:

[0032]

[0033] In the above scheme, the method further comprises:

[0034] Stokes Vector G m can be broken down into:

[0035]

[0036] Among them, G v The Stokes vector representing the random volume scattering mechanism, G p The Stokes vector represents the rank-1 scattering mechanism, including surface scattering and dihedral scattering in the π / 4 polarization mode; m v and m p The coefficients of the volume scattering component and the rank-1 scattering component, α and is the scattering variable.

[0037] In order to simplify the subsequent calculation process, c=±cos2α, then the expressions of the elements of the Stokes vector are:

[0038]

[0039] Among them, a, b, c and m v Both can be used p express:

[0040]

[0041] Put a, b, c and m v Substitute a 2 +b 2 +c 2 =1, we can find m p , and then find a, b, c and m v The value of , at the same time, α and The expressions are:

[0042]

[0043] The expressions for volume scattering, dihedral scattering, and surface scattering energy are:

[0044]

[0045] Based on the expressions of volume scattering, dihedral scattering and surface scattering energy, the final image decomposition result can be obtained.

[0046] The advantages and beneficial effects of the present invention are as follows: the present invention extends the reduced polarization target decomposition method to the reduced polarization interferometric target decomposition algorithm, improving the current situation that the reduced polarization interferometric SAR decomposition method is relatively scarce. In the decomposition results obtained by the present invention, compared with the reduced polarization target decomposition algorithm, the problem of overestimation of the volume scattering energy in the inclined building area is significantly improved, and the decomposition results are more consistent with the actual ground object scattering mechanism. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 The present invention is a flowchart of a π / 4 reduced polarization interferometric synthetic aperture radar decomposition method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to enable a more detailed understanding of the features and technical contents of the embodiments of the present invention, the implementation of the embodiments of the present invention is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not intended to limit the embodiments of the present invention.

[0049] Figure 1 FIG. 4 is a flow chart of a π / 4 reduced polarization interferometric synthetic aperture radar decomposition method according to an embodiment of the present invention. Figure 1 As shown, the π / 4 reduced polarization interferometric synthetic aperture radar decomposition method according to an embodiment of the present invention comprises the following steps:

[0050] S1: Acquire polarimetric interferometric SAR image data, and convert the full polarimetric SAR data into reduced polarimetric SAR data according to the conversion relationship between the full polarimetric SAR data and the reduced polarimetric SAR data;

[0051] S2: Calculate the optimal parameters of polarization interferometry and determine the optimization factors of the volume scattering model;

[0052] S3: Optimizing the volume scattering model according to the volume scattering model optimization factor;

[0053] S4: deriving the expressions of body scattering energy, surface scattering energy and dihedral angle scattering energy according to the optimized body scattering model.

[0054] Specifically, the step S1 further includes:

[0055] The two fully polarimetric SAR images acquired in the same area are pre-processed by filtering and multi-viewing, and one of the images is selected as the main image and the other as the secondary image. Subsequently, the fully polarimetric SAR data is converted into π / 4 reduced polarimetric SAR data by using the conversion relationship between the fully polarimetric SAR data and the π / 4 reduced polarimetric SAR data for subsequent processing. The specific implementation method of this step is as follows:

[0056] The ensemble average values ​​of the coherence matrices of the two fully polarized SAR images are obtained as <[T1]> and <[T2]> respectively:

[0057]

[0058] Among them, T ij (i=1,2,3;j=1,2,3) and T ij ′(i=1,2,3;j=1,2,3)represents the i-th row and j-th column elements of the ensemble average values ​​of the primary and secondary full polarization coherence matrices <[T1]> and <[T2]>, respectively; the superscript “ H " represents the conjugate transpose of the matrix; k1 and k2 are the Pauli basis of the primary and secondary full polarization images, respectively.

[0059] Among them, S HHi (i=1,2) is the backscatter of vertical transmission and vertical reception, S HVi (i=1,2) is the backscatter with vertical transmission and horizontal reception, S VVi (i=1,2) is the backscatter with horizontal transmission and horizontal reception.

[0060] G=(g0,g1,g2,g3) T is the Stokes vector, g k (k=0,1,2,3) represents the element of the kth row of the Stokes vector. The corresponding relationship between the Stokes vector and the fully polarized SAR data is:

[0061]

[0062] The Stokes vectors of the main and secondary images can be expressed as:

[0063]

[0064] Among them, g km (k=0,1,2,3) represents the Stokes vector G m The element of the kth row of g ks (k=0,1,2,3) represents the Stokes vector G s The element of the k-th row of .

[0065] Specifically, the step S2 further includes the following sub-steps:

[0066] Polarization interference information can be represented by a 6×6 coherence matrix. The 6×6 semi-positive Hermitian matrix defined by the outer product of k1 and k2 is expressed as:

[0067]

[0068] in,

[0069]

[0070] ·> is the multi-view average. [T m ] and [T s ] is a semi-positive Hermitian matrix, which contains all the polarization information obtained by each group of antennas, [Ω 12 ] is a non-Hermitian matrix, containing polarization information and interference information.

[0071] In polarimetric interferometry SAR, the interference coherence coefficient and phase information are strongly dependent on the polarization state. Similar to the definition of polarimetric similarity parameter, the unit projection vectors ω1 and ω2 are introduced, and the Pauli basis scattering vectors k1 and k2 are projected onto ω1 and ω2 respectively. The polarimetric interferometry similarity parameter (P PISP ) is:

[0072]

[0073] Among them, P PISP ∈[0,1], |·| represents the absolute value operation, [ω1]=[ω 1,opt1 ,ω 1,opt2 ,ω 1,opt3 ] and [ω2] = [ω 2,opt1 ,ω 2,opt2 ,ω 2,opt3 ] are the optimal scattering matrices composed of the optimal scattering vectors of the main and auxiliary images respectively. Parameter P PISP The value of is unique and does not change with the change of the target orientation angle. This parameter can be used to ensure the optimal coherence between the two images. Therefore, the optimal parameter P for polarization interference is constructed. PIOP For subsequent decomposition processing, the optimal parameter P of polarization interference PIOP The expression is:

[0074]

[0075] Specifically, the step S3 further includes:

[0076] Using the optimal parameter P of polarization interferometry PIOP, optimize the volume scattering model proposed by Freeman. The expression of the volume scattering model proposed by Freeman is:

[0077]

[0078] The optimal parameter P of polarization interferometry PIOP The expression of the optimized volume scattering model is:

[0079]

[0080] Specifically, the step S4 further includes:

[0081] Stokes Vector G m can be broken down into:

[0082]

[0083] Among them, G v The Stokes vector representing the random volume scattering mechanism, G p The Stokes vector represents the rank-1 scattering mechanism, including surface scattering and dihedral scattering in the π / 4 polarization mode; m v and m p The coefficients of the volume scattering component and the rank-1 scattering component, α and is the scattering variable.

[0084] In order to simplify the subsequent calculation process, c=±cos2α, then the expressions of the elements of the Stokes vector are:

[0085]

[0086] Among them, a, b, c and m v Both can be used p express:

[0087]

[0088] Put a, b, c and m v Substitute a 2 +b 2 +c 2 =1, we can find m p , and then find a, b, c and m v The value of , at the same time, α and The expressions are:

[0089]

[0090] The expressions for volume scattering, dihedral scattering, and surface scattering energy are:

[0091]

[0092] Based on the expressions of volume scattering, dihedral scattering and surface scattering energy, the final image decomposition result can be obtained.

[0093] So far, the present embodiment has been described in detail with reference to the accompanying drawings. Based on the above description, those skilled in the art should have a clear understanding of the π / 4 reduced polarization interferometric synthetic aperture radar decomposition method of the present invention.

[0094] The methods provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0095] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A π / 4 reduced polarization interferometric synthetic aperture radar decomposition method, characterized in that: The method comprises: S1: Acquire polarimetric interferometric SAR image data, and convert the full polarimetric SAR data into reduced polarimetric SAR data according to the conversion relationship between the full polarimetric SAR data and the reduced polarimetric SAR data; S2: Calculate the optimal parameters of polarization interferometry and determine the optimization factors of the volume scattering model; S3: Optimizing the volume scattering model according to the volume scattering model optimization factor; S4: deriving the expressions of body scattering energy, surface scattering energy and dihedral angle scattering energy according to the optimized body scattering model; Wherein, step S2 also includes: Polarization interference information is represented by a 6×6 coherence matrix; the 6×6 semi-positive Hermitian matrix defined by the outer product of k1 and k2 is expressed as: in, <·> is the multi-view average; [T m ] and [T s ] is a semi-positive Hermitian matrix, which contains all the polarization information obtained by each group of antennas, [Ω 12 ] is a non-Hermitian matrix, containing polarization information and interference information; Step S2 also includes: In polarimetric interferometry SAR, the interference coherence coefficient and phase information are dependent on the polarization state. Similar to the definition of polarimetric similarity parameter, unit projection vectors ω1 and ω2 are introduced, and the Pauli basis scattering vectors k1 and k2 are projected onto ω1 and ω2 respectively. The polarimetric interferometry similarity parameter P PISP The expression is: Among them, P PISP ∈[0,1], |·| represents the absolute value operation, [ω1]=[ω 1,opt1 ,ω 1,opt2 ,ω 1,opt3 ] and [ω2] = [ω 2,opt1 ,ω 2,opt2 ,ω 2,opt3 ] are the optimal scattering matrices composed of the optimal scattering vectors of the main and auxiliary images, ω i,optj The jth element of the optimal scattering matrix of the primary or secondary image, where i = 1, 2; j = 1, 2, 3; parameter P PISP The value of is unique and does not change with the change of the target orientation angle, which is used to ensure the optimization of the coherence between the two images; therefore, the optimal parameter P of polarization interference is constructed. PIOP For subsequent decomposition processing, the optimal parameter P of polarization interference PIOP The expression is: Step S3 also includes: Using the optimal parameter P of polarization interferometry PIOP , optimize the volume scattering model proposed by Freeman. The expression of the volume scattering model proposed by Freeman is: The optimal parameter P of polarization interferometry PIOP The expression of the optimized volume scattering model is: Step S4 also includes: Stokes Vector G m is broken down into: Among them, G v The Stokes vector representing the random volume scattering mechanism, G p The Stokes vector represents the rank-1 scattering mechanism, including surface scattering and dihedral scattering in the π / 4 polarization mode; m v and m p The coefficients of the volume scattering component and the rank-1 scattering component, α and is a scattering variable; Step S4 also includes: In order to simplify the subsequent calculation process, c=±cos2α, then the expressions of the elements of the Stokes vector are: Among them, a, b, c and m v All in m p express: Put a, b, c and m v Substitute a 2 +b 2 +c 2 =1, find m p , and then find a, b, c and m v The value of , at the same time, α and The expressions are: The expressions for volume scattering, dihedral scattering, and surface scattering energy are: According to the expressions of volume scattering, dihedral scattering and surface scattering energy, the final image decomposition result is obtained.

2. The π / 4 reduced polarization interferometric synthetic aperture radar decomposition method according to claim 1, characterized in that: Features: Step S1 also includes: Filtering and multi-look preprocessing operations are completed on two fully polarimetric SAR images acquired in the same area, and one of the images is selected as the main image and the other as the secondary image. Subsequently, the conversion relationship between fully polarimetric SAR data and π / 4 reduced polarimetric SAR data is used to convert the fully polarimetric SAR data into π / 4 reduced polarimetric SAR data for subsequent processing.

3. The π / 4 reduced polarization interferometric synthetic aperture radar decomposition method according to claim 2, characterized in that: In step S1, the specific implementation is as follows: The ensemble average values ​​of the coherence matrices of the two fully polarized SAR images are obtained as <[T1]> and [T2]> respectively: Among them, T ij and T ij ′ represents the i-th row and j-th column elements of the ensemble average values ​​of the primary and secondary full polarization coherence matrices <[T1]> and <[T2]>, respectively, where i=1,2; j=1,2,3; the superscript "H" represents the conjugate transpose of the matrix; k1 and k2 are the Pauli bases of the primary and secondary full polarization images, respectively, and their expressions are and Among them, S HHi is the backscatter of vertical transmission and vertical reception, S HVi is the backscatter with vertical transmission and horizontal reception, S VVi is the backscatter of horizontal transmission and horizontal reception, where i=1,2.

4. The π / 4 reduced polarization interferometric synthetic aperture radar decomposition method according to claim 3, characterized in that: Features: Step S1 also includes: G=(g0,g1,g2,g3) Τ is the Stokes vector, gk represents the element of the kth row of the Stokes vector, where k = 0, 1, 2, 3; the corresponding relationship between the Stokes vector and the fully polarized SAR data is: The Stokes vectors of the main and secondary images are expressed as: Among them, g km represents the Stokes vector G m The element of the kth row of g ks represents the Stokes vector G s The elements of the k-th row of , where k = 0, 1, 2, 3.

Images