CN101976357A

CN101976357A - A full polarization synthetic aperture radar image classification method and device

Info

Publication number: CN101976357A
Application number: CN201010508761XA
Authority: CN
Inventors: 田昕; 凌飞龙; 陈尔学; 白黎娜
Original assignee: Research Institute Of Forest Resource Information Techniques Chinese Academy Of Forestry
Current assignee: Research Institute Of Forest Resource Information Techniques Chinese Academy Of Forestry
Priority date: 2010-10-18
Filing date: 2010-10-18
Publication date: 2011-02-16

Abstract

The invention provides a method and a device for classifying fully polarimetric Synthetic Aperture Radar (SAR) images. The method comprises the steps of converting a covariance matrix of the complete polarization SAR image data in a complex form into a form completely represented by the intensity, and carrying out classification processing on the complete polarization SAR image based on the converted image data represented by the intensity. Therefore, the fully-polarized SAR image data which obey different statistical distributions are converted into corresponding backscattering intensity image data which obey uniform distribution, so that the existing remote sensing image processing method which is based on the remote sensing image data and obeys uniform distribution, such as a segmentation and object-oriented classification method developed based on an optical remote sensing image, can be suitable for processing the fully-polarized SAR image data, and further improves the classification precision of the fully-polarized SAR image data.

Description

A full polarization synthetic aperture radar image classification method and device

技术领域technical field

本发明涉及遥感图像处理技术领域，具体涉及一种全极化合成孔径雷达图像分类方法及装置。The invention relates to the technical field of remote sensing image processing, in particular to a method and device for classifying images of full polarization synthetic aperture radar.

背景技术Background technique

目前已经发展了很多面向对象的遥感图像处理方法，其中，基于光学遥感图像发展的多分辨图像分割和面向对象的分类技术应用最广。而现有遥感图像处理方法一般都基于遥感图像数据服从统一分布，例如高斯分布的假设而应用的。At present, many object-oriented remote sensing image processing methods have been developed. Among them, the multi-resolution image segmentation and object-oriented classification technology based on the development of optical remote sensing images are the most widely used. However, existing remote sensing image processing methods are generally applied based on the assumption that remote sensing image data obeys a uniform distribution, such as a Gaussian distribution.

但是，由于成像机理不同，遥感图像中的合成孔径雷达(SAR：SyntheticAperture Radar)图像数据服从不同的统计分布，这导致全极化SAR图像数据的统计分布异于统一分布，因此，目前基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法，并不适于对SAR图像数据进行处理，存在分类精度降低的情况。However, due to different imaging mechanisms, the synthetic aperture radar (SAR: Synthetic Aperture Radar) image data in remote sensing images obeys different statistical distributions, which causes the statistical distribution of full-polarization SAR image data to be different from the uniform distribution. Remote sensing image processing methods with uniform distribution of data, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images, are not suitable for processing SAR image data, and the classification accuracy is reduced.

发明内容Contents of the invention

本发明所要解决的技术问题是提供一种全极化合成孔径雷达图像分类方法及装置，从而提高全极化合成孔径雷达图像分类精度。The technical problem to be solved by the present invention is to provide a full polarization synthetic aperture radar image classification method and device, thereby improving the classification accuracy of the full polarization synthetic aperture radar image.

为解决上述技术问题，本发明提供方案如下：In order to solve the problems of the technologies described above, the present invention provides the following solutions:

本发明实施例提供了一种全极化合成孔径雷达SAR图像分类方法，包括：An embodiment of the present invention provides a method for classifying an all-polarization synthetic aperture radar SAR image, including:

将全极化SAR图像数据转换为对应的后向散射强度图像数据；Convert the full polarization SAR image data into corresponding backscatter intensity image data;

根据所述后向散射强度图像数据，进行全极化SAR图像分类处理。According to the backscattering intensity image data, full polarization SAR image classification processing is performed.

优选的，所述将全极化SAR图像数据转换为对应的后向散射强度图像数据包括：Preferably, said converting full polarization SAR image data into corresponding backscattering intensity image data includes:

将获取的全极化SAR图像数据对应的复散射矩阵进行辐射定标处理；Perform radiation calibration processing on the complex scattering matrix corresponding to the acquired full-polarization SAR image data;

将经过辐射定标处理后的复散射矩阵转换为对应的协方差矩阵；Convert the complex scattering matrix processed by radiation calibration into the corresponding covariance matrix;

将所述协方差矩阵进行多视化处理；Carrying out multi-visualization processing to the covariance matrix;

将多视化处理后的协方差矩阵进行极化基转换处理，获取全极化SAR图像数据对应的后向散射强度图像数据。The covariance matrix after multi-visualization processing is subjected to polarization base conversion processing to obtain the backscattering intensity image data corresponding to the full polarization SAR image data.

优选的，所述将经过辐射定标处理后的复散射矩阵转换为对应的协方差矩阵包括：Preferably, said converting the complex scattering matrix processed by radiation calibration into a corresponding covariance matrix includes:

令所述复散射矩阵的参数S_HV＝S_VH，其中，S_HH表示所述复散射矩阵中水平极化发射、水平极化接收的散射部分，S_HV表示所述复散射矩阵中水平极化发射、垂直极化接收的散射部分，S_VH表示所述复散射矩阵中垂直极化发射、水平极化接收的散射部分，S_VV表示所述复散射矩阵中垂直极化发射、垂直极化接收的散射部分；Let the complex scattering matrix The parameter S _HV =S _VH , wherein, _SHH represents the scattering part of the horizontal polarization emission and horizontal polarization reception in the complex scattering matrix, and _SHV represents the horizontal polarization emission and vertical polarization reception in the complex scattering matrix The scattering part of S _VH represents the scattering part of vertical polarization transmission and horizontal polarization reception in the complex scattering matrix, and S _VV represents the scattering part of vertical polarization transmission and vertical polarization reception in the complex scattering matrix;

用复矢量h表示所述复散射矩阵，表示公式为

其中，上角标T表示矩阵转置；Represent described complex scattering matrix with complex vector h, expression formula is

Among them, the superscript T indicates matrix transposition;

根据公式

计算得到经过辐射定标处理后的复散射矩阵对应的协方差矩阵C，其中，上角标*表示复数共轭，上角标T表示矩阵转置。According to the formula

The covariance matrix C corresponding to the complex scattering matrix after radiation calibration processing is calculated, where the superscript * indicates complex conjugation, and the superscript T indicates matrix transposition.

优选的，将所述协方差矩阵进行多视化处理后获取的协方差矩阵C’表达式为：Preferably, the expression of the covariance matrix C ' obtained after the covariance matrix is multi-visualized is:

${C C}^{' '} = = [\begin{matrix} < < {S S}_{hh hh} {S S}_{hh hh}^{* *} > > & < < {S S}_{hh hh} {S S}_{hv hv}^{* *} > > & < < {S S}_{hh hh} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{hv hv} {S S}_{hh hh}^{* *} > > & < < {S S}_{hv hv} {S S}_{hv hv}^{* *} > > & < < {S S}_{hv hv} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{vv vv} {S S}_{hh hh}^{* *} > > & < < {S S}_{vv vv} {S S}_{hv hv}^{* *} > > & < < {S S}_{vv vv} {S S}_{vv vv}^{* *} > > \end{matrix}]$

优选的，所述将多视化处理后的协方差矩阵进行极化基转换处理，获取全极化SAR图像数据对应的后向散射强度图像数据包括：Preferably, performing polarization base conversion processing on the covariance matrix after multi-visibility processing, and obtaining the backscattering intensity image data corresponding to the full polarization SAR image data includes:

将所述多视化处理后的协方差矩阵所包含的上三角矩阵中3个实数元素参数量，以及3个复数参数量中的实部Re参数量和虚部IM参数量通过极化基转换，获取对应的后向散射强度量，转换公式为：The three real number element parameters in the upper triangular matrix included in the multi-visualization processed covariance matrix, and the real part Re parameter and the imaginary part IM parameter of the three complex number parameters are converted by polarization basis , to obtain the corresponding backscattering intensity, the conversion formula is:

$A A (\begin{matrix} < < {S S}_{hh hh} {S S}_{hh hh}^{* *} > > \\ < < {S S}_{vv vv} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{hv hv} {S S}_{hv hv}^{* *} > > \\ Re Re [[< < {S S}_{hh hh} {S S}_{vv vv}^{* *} > >]] \\ IM IM [[< < {S S}_{hh hh} {S S}_{vv vv}^{* *} > >]] \\ Re Re [[< < {S S}_{hh hh} {S S}_{hv hv}^{* *} > >]] \\ IM IM [[< < {S S}_{hh hh} {S S}_{hv hv}^{* *} > >]] \\ Re Re [[< < {S S}_{hv hv} {S S}_{vv vv}^{* *} > >]] \\ IM IM [[< < {S S}_{hv hv} {S S}_{vv vv}^{* *} > >]] \end{matrix}) = = (\begin{matrix} {σ σ}_{hh hh}^{00} \\ {σ σ}_{vv vv}^{00} \\ {σ σ}_{+ + + + 4545}^{00} \\ {σ σ}_{- - - - 4545}^{00} \\ {σ σ}_{ll ll}^{00} \\ {σ σ}_{rr rr}^{00} \\ {σ σ}_{h h + + 4545}^{00} \\ {σ σ}_{hl hl}^{00} \\ {σ σ}_{+ + 4545 l l}^{00} \end{matrix})$

其中，in,

$A A = = [\begin{matrix} 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & 6.28319 6.28319 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & 6.28319 6.28319 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & - - 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & - - 12.5664 12.5664 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & - - 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 12.5664 12.5664 \\ 6.28319 6.28319 & 0.0000 0.0000 & 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 6.28319 6.28319 & 0.0000 0.0000 & 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 \\ 6.28319 6.28319 & 3.14159 3.14159 & 6.28319 6.28319 & 0.0000 0.0000 & - - 6.28319 6.28319 & 6.28319 6.28319 & - - 6.28319 6.28319 & 6.28319 6.28319 & 6.28319 6.28319 \end{matrix}]$

参数σ为后向散射系数，参数σ的下标代表极化基的接收和发射极化方式：h代表水平、v代表垂直、l代表左圆、r代表右圆、+或+45代表+45°线性、-或-45代表-45°线性。The parameter σ is the backscattering coefficient, and the subscript of the parameter σ represents the receiving and transmitting polarization modes of the polarized base: h stands for horizontal, v stands for vertical, l stands for left circle, r stands for right circle, + or +45 stands for +45 °Linear, - or -45 stands for -45° linear.

优选的，所述根据所述后向散射强度图像数据，进行全极化SAR图像分类处理包括：Preferably, said performing full-polarization SAR image classification processing according to said backscattering intensity image data includes:

对获取的后向散射强度图像数据进行正射校正处理。Perform orthorectification processing on the acquired backscattering intensity image data.

本发明实施例还提供了一种全极化合成孔径雷达SAR图像分类装置，包括：The embodiment of the present invention also provides a full polarization synthetic aperture radar SAR image classification device, including:

转换模块，用于将全极化SAR图像数据转换为对应的后向散射强度图像数据；A conversion module, configured to convert full-polarization SAR image data into corresponding backscattering intensity image data;

分类模块，用于根据所述后向散射强度图像数据，进行全极化SAR图像分类处理。A classification module, configured to perform full-polarization SAR image classification processing according to the backscattering intensity image data.

优选的，所述转换模块包括：Preferably, the conversion module includes:

辐射定标处理单元，用于将获取的全极化SAR图像数据对应的复散射矩阵进行辐射定标处理；a radiation calibration processing unit, configured to perform radiation calibration processing on the complex scattering matrix corresponding to the acquired full-polarization SAR image data;

转换单元，用于将经过辐射定标处理后的复散射矩阵转换为对应的协方差矩阵；A conversion unit, configured to convert the complex scattering matrix processed by radiation calibration into a corresponding covariance matrix;

多视化处理单元，用于将所述协方差矩阵进行多视化处理；A multi-visualization processing unit, configured to perform multi-visualization processing on the covariance matrix;

极化基转换处理单元，用于将多视化处理后的协方差矩阵进行极化基转换处理，获取全极化SAR图像数据对应的后向散射强度图像数据。The polarization base conversion processing unit is configured to perform polarization base conversion processing on the multi-view processed covariance matrix to obtain backscattering intensity image data corresponding to the full polarization SAR image data.

优选的，所述分类模块包括：Preferably, the classification module includes:

正射校正处理单元，用于对转换模块获取的后向散射强度图像数据进行正射校正处理。The orthorectification processing unit is configured to perform orthorectification processing on the backscattering intensity image data acquired by the conversion module.

从以上所述可以看出，本发明提供的全极化合成孔径雷达SAR图像分类方法及装置，通过将复数形式的全极化SAR图像数据的协方差矩阵转换为完全由强度量表示的形式，并基于转换后的由强度量表示的图像数据，进行全极化SAR图像分类处理。从而将服从不同统计分布的全极化SAR图像数据转换为对应的服从统一分布的后向散射强度图像数据，以便于现有基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法，可以适于对全极化SAR图像数据进行处理，进而提高了全极化SAR图像数据的分类精度。As can be seen from the above, the full polarization SAR image classification method and device provided by the present invention, by converting the covariance matrix of the full polarization SAR image data in complex form into a form completely represented by the intensity quantity, And based on the converted image data represented by the intensity quantity, the full polarization SAR image classification process is performed. Therefore, the fully polarized SAR image data subject to different statistical distributions is converted into the corresponding backscattering intensity image data subject to a uniform distribution, so as to facilitate the existing remote sensing image processing methods based on remote sensing image data subject to a uniform distribution, such as based on optical remote sensing images The developed segmentation and object-oriented classification methods are suitable for processing full-polarization SAR image data, thereby improving the classification accuracy of full-polarization SAR image data.

附图说明Description of drawings

图1为本发明实施例提供的全极化SAR图像数据分类方法流程示意图一；FIG. 1 is a schematic flow diagram of a method for classifying full-polarization SAR image data provided by an embodiment of the present invention;

图2为本发明实施例提供的全极化SAR图像数据分类方法流程示意图二；FIG. 2 is a second schematic flow diagram of a method for classifying full-polarization SAR image data provided by an embodiment of the present invention;

图3为本发明实施例提供的全极化SAR图像数据分类装置结构示意图一；Fig. 3 is a schematic structural diagram of a device for classifying full-polarization SAR image data provided by an embodiment of the present invention;

图4为本发明实施例提供的全极化SAR图像数据分类装置结构示意图二。Fig. 4 is a second structural schematic diagram of a device for classifying full-polarization SAR image data provided by an embodiment of the present invention.

具体实施方式Detailed ways

本发明实施例提供了一种全极化SAR图像分类方法，如附图1所示，具体可以包括以下步骤：An embodiment of the present invention provides a method for classifying full-polarization SAR images, as shown in Figure 1, which may specifically include the following steps:

步骤11，将全极化SAR图像数据转换为对应的后向散射强度图像数据；Step 11, converting the full polarization SAR image data into corresponding backscattering intensity image data;

步骤12，根据所述后向散射强度图像数据，进行全极化SAR图像分类处理。Step 12, performing full polarization SAR image classification processing according to the backscattering intensity image data.

本发明实施例提供了全极化SAR图像分类方法，通过将服从不同统计分布的全极化SAR图像数据转换为对应的服从统一分布的后向散射强度图像数据，从而可以使现有基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法，适于对全极化SAR图像数据进行处理，进而提高了全极化SAR图像数据的分类精度。The embodiment of the present invention provides a method for classifying fully polarimetric SAR images. By converting fully polarimetric SAR image data subject to different statistical distributions into corresponding backscattering intensity image data subject to uniform distribution, the existing remote sensing image-based Remote sensing image processing methods with data subject to uniform distribution, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images, are suitable for processing full-polarization SAR image data, thereby improving the classification accuracy of full-polarization SAR image data.

在本发明一个可选的实施例中，将全极化SAR图像数据转换为对应的后向散射强度图像数据具体可以包括以下步骤：In an optional embodiment of the present invention, converting the full-polarization SAR image data into corresponding backscattering intensity image data may specifically include the following steps:

在本发明一个可选的实施例中，根据所述后向散射强度图像数据，进行全极化SAR图像分类处理具体可以包括以下步骤：In an optional embodiment of the present invention, performing full-polarization SAR image classification processing according to the backscattering intensity image data may specifically include the following steps:

下面结合附图2，对本发明实施例提供的全极化SAR图像分类方法的一个具体实施例进行详细的描述，具体包括：A specific embodiment of the full-polarization SAR image classification method provided by the embodiment of the present invention is described in detail below in conjunction with accompanying drawing 2, specifically including:

步骤21，获取全极化SAR图像数据。Step 21, acquiring full-polarization SAR image data.

具体的，可通过极化雷达获取全极化SAR图像数据。Specifically, full-polarization SAR image data may be acquired through a polarization radar.

由于全极化SAR图像数据的具体表现形式可为复散射矩阵，因此，本发明实施例中，可以以水平和垂直线性极化基表示复散射矩阵，具体可如公式(1)所示：Since the specific form of full-polarization SAR image data can be a complex scattering matrix, therefore, in the embodiment of the present invention, the complex scattering matrix can be represented by horizontal and vertical linear polarization bases, specifically as shown in formula (1):

$S S = = [\begin{matrix} {S S}_{HH HH} & {S S}_{HV HV} \\ {S S}_{VH VH} & {S S}_{VV VV} \end{matrix}] - - - - - - ((11))$

其中，S_HH表示复散射矩阵中水平(H)极化发射、水平(H)极化接收的散射部分，S_HV表示复散射矩阵中水平(H)极化发射、垂直(V)极化接收的散射部分，S_VH表示复散射矩阵中垂直(V)极化发射、水平(H)极化接收的散射部分，S_VV表示复散射矩阵中垂直(V)极化发射、垂直(V)极化接收的散射部分。Among them, S _HH represents the scattering part of the horizontal (H) polarized emission and horizontal (H) polarized reception in the complex scattering matrix, and _SHV represents the horizontal (H) polarized emission and vertical (V) polarized reception in the complex scattering matrix S _VH represents the scattering part of vertical (V) polarized emission and horizontal (H) polarized reception in the complex scattering matrix, S _VV represents the vertical (V) polarized emission, vertical (V) polarized scatter part of the reception.

本发明实施例中获取的全极化SAR图像数据即可以由公式(1)所示的复数形式的复散射矩阵表示。The fully polarized SAR image data acquired in the embodiment of the present invention can be represented by a complex scattering matrix in complex form as shown in formula (1).

步骤22，对获取的全极化SAR图像数据进行辐射定标处理。Step 22, performing radiometric calibration processing on the acquired full-polarization SAR image data.

在本发明的一个具体实施例中，还可以对获取的全极化SAR图像数据进行辐射定标处理，得到辐射定标后的复散射矩阵。In a specific embodiment of the present invention, radiation calibration processing may also be performed on the acquired full-polarization SAR image data to obtain a complex scattering matrix after radiation calibration.

步骤23，将经过辐射定标处理后的复散射矩阵转换为对应的协方差矩阵。Step 23, converting the complex scattering matrix after the radiation calibration processing into a corresponding covariance matrix.

基于后向散射互易性原理，本发明实施例中，可令复散射矩阵中的参数S_HV＝S_VH，则全极化SAR图像数据对应的复散射矩阵可以用于一个复矢量h表示，具体可如公式(2)所示：Based on the principle of backscattering reciprocity, in the embodiment of the present invention, the parameter S _HV =S _VH in the complex scattering matrix can be set, then the complex scattering matrix corresponding to the fully polarized SAR image data can be represented by a complex vector h, Specifically, it can be shown as formula (2):

$h h = = {[\begin{matrix} {S S}_{HH HH} & \sqrt{22} {S S}_{HV HV} & {S S}_{VV VV} \end{matrix}]}^{T T} - - - - - - ((22))$

其中，参数T表示矩阵转置，参数S_HV前的

是为了确保总功率计算的一致性。由此，可以计算出复散射矩阵对应的协方差矩阵C：Among them, the parameter T represents the matrix transpose, and the parameter S before _HV

is to ensure consistency in total power calculations. From this, the covariance matrix C corresponding to the complex scattering matrix can be calculated:

$C C = = {hh hh}^{* * T T} = = [\begin{matrix} {| | {S S}_{HH HH} | |}^{22} & \sqrt{22} {S S}_{HH HH} {S S}_{HV HV}^{* *} & {S S}_{HH HH} {S S}_{VV VV}^{* *} \\ \sqrt{22} {S S}_{HV HV} {S S}_{HH HH}^{* *} & 22 {| | {S S}_{HV HV} | |}^{22} & \sqrt{22} {S S}_{HV HV} {S S}_{VV VV}^{* *} \\ {S S}_{VV VV} {S S}_{HH HH}^{* *} & \sqrt{22} {S S}_{VV VV} {S S}_{HV HV}^{* *} & {| | {S S}_{VV VV} | |}^{22} \end{matrix}] - - - - - - ((33))$

其中，上角标*表示复数共轭，上角标T表示矩阵转置。Among them, superscript * means complex conjugate, and superscript T means matrix transpose.

步骤24，对协方差矩阵进行多视化处理。Step 24, perform multi-visualization processing on the covariance matrix.

由于在经过步骤22辐射定标处理得到的复散射矩阵为单视复散射矩阵，因此，步骤23由单视复散射矩阵得到的协方差矩阵C同样为单视的，为了消除SAR图像数据中的噪声，本发明实施例中，可对经过步骤23所获取的单视协方差矩阵进行多视化处理，从而得到矩阵C：Since the complex scattering matrix obtained through radiation calibration in step 22 is a single-view complex scattering matrix, the covariance matrix C obtained from the single-view complex scattering matrix in step 23 is also single-view. In order to eliminate the SAR image data Noise, in the embodiment of the present invention, multi-visual processing can be performed on the single-view covariance matrix obtained in step 23, so as to obtain the matrix C:

${C C}^{' '} = = [\begin{matrix} < < {S S}_{hh hh} {S S}_{hh hh}^{* *} > > & < < {S S}_{hh hh} {S S}_{hv hv}^{* *} > > & < < {S S}_{hh hh} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{hv hv} {S S}_{hh hh}^{* *} > > & < < {S S}_{hv hv} {S S}_{hv hv}^{* *} > > & < < {S S}_{hv hv} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{vv vv} {S S}_{hh hh}^{* *} > > & < < {S S}_{vv vv} {S S}_{hv hv}^{* *} > > & < < {S S}_{vv vv} {S S}_{vv vv}^{* *} > > \end{matrix}] - - - - - - ((44))$

步骤25，对多视化处理后的协方差矩阵进行极化基转换处理。Step 25, performing polarization base conversion processing on the covariance matrix after the multi-view processing.

由于全极化SAR图像数据服从不同的统计分布，导致基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法，在对SAR图像数据进行处理和信息提取时存在不便，因此本发明实施例中，可将全极化SAR图像数据中的参数用服从同一分布，例如高斯分布的参数形式表示，使得基于遥感图像数据服从统一分布的遥感图像处理方法可以被用于全极化SAR图像数据的处理。Since the full-polarization SAR image data obeys different statistical distributions, remote sensing image processing methods based on remote sensing image data subject to uniform distribution, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images, are used to process SAR image data and information It is inconvenient to extract, so in the embodiment of the present invention, the parameters in the fully polarized SAR image data can be expressed in the form of parameters that obey the same distribution, such as Gaussian distribution, so that the remote sensing image processing method based on the remote sensing image data that obeys the uniform distribution can be It is used for the processing of full polarization SAR image data.

为了统一全极化SAR图像数据中参数的统计分布形式，本发明实施例中，具体可以采用一种把全极化SAR图像数据对应的协方差矩阵元素参数量转换成只用强度量表示的极化基转化处理方法。In order to unify the statistical distribution form of the parameters in the full-polarization SAR image data, in the embodiment of the present invention, a method of converting the covariance matrix element parameters corresponding to the full-polarization SAR image data into polar Chemical base conversion treatment method.

由于协方差矩阵的信息可以用其包含的上三角矩阵中3个实数元素参数量和3个复数参数量的相关项表示，而且，每个复数项可各用1个实部(Re)参数量和1个虚部(IM)参数量表述，因此，利用上述9个参数量就可以完全表示协方差矩阵的信息。Since the information of the covariance matrix can be represented by the relevant items of the 3 real element parameters and 3 complex parameter quantities in the upper triangular matrix contained in it, and each complex number item can use 1 real part (Re) parameter quantity and one imaginary part (IM) parameter, therefore, the information of the covariance matrix can be fully expressed by using the above nine parameters.

本发明实施例中，上述9个参数量可以通过极化基转换，转换成9个后向散射强度量，具体极化基转换可如公式(5)所示：In the embodiment of the present invention, the above-mentioned 9 parameter quantities can be converted into 9 backscattering intensity quantities through polarization base conversion, and the specific polarization base conversion can be shown in formula (5):

其中，in,

$A A = = [\begin{matrix} 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & 6.28319 6.28319 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & 6.28319 6.28319 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & - - 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & - - 12.5664 12.5664 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & - - 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 12.5664 12.5664 \\ 6.28319 6.28319 & 0.0000 0.0000 & 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 6.28319 6.28319 & 0.0000 0.0000 & 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 \\ 6.28319 6.28319 & 3.14159 3.14159 & 6.28319 6.28319 & 0.0000 0.0000 & - - 6.28319 6.28319 & 6.28319 6.28319 & - - 6.28319 6.28319 & 6.28319 6.28319 & 6.28319 6.28319 \end{matrix}] - - - - - - ((55))$

公式(5)中，参数σ为后向散射系数，参数σ的下标代表极化基的接收和发射极化方式：h代表水平、v代表垂直、l代表左圆、r代表右圆、+或+45代表+45°线性、-或-45代表-45°线性。In formula (5), the parameter σ is the backscattering coefficient, and the subscript of the parameter σ represents the receiving and transmitting polarization modes of the polarized base: h stands for horizontal, v stands for vertical, l stands for left circle, r stands for right circle, + Or +45 represents +45° linearity, - or -45 represents -45° linearity.

经过以上处理后，既可以得到全极化SAR图像数据对应的9个后向散射强度图像数据，相当于光学遥感图像的9个波段。After the above processing, 9 backscattering intensity image data corresponding to the full-polarization SAR image data can be obtained, which is equivalent to 9 bands of the optical remote sensing image.

由于对全极化SAR图像数据对应的协方差矩阵进行了多视化处理，因此，此时得到的全极化SAR图像对应的9个后向散射强度图像噪声得到有效去除，数据统计分布近似符合统一分布，例如高斯分布，为本发明实施例后续采用基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法奠定了基础。Since the covariance matrix corresponding to the full-polarization SAR image data has been multi-visualized, the noise of the nine backscattering intensity images corresponding to the full-polarization SAR image obtained at this time has been effectively removed, and the statistical distribution of the data approximately conforms to Uniform distribution, such as Gaussian distribution, lays the foundation for the subsequent adoption of remote sensing image processing methods based on remote sensing image data subject to uniform distribution in the embodiment of the present invention, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images.

步骤26，对获取的后向散射强度图像数据进行正射校正处理。Step 26, performing orthorectification processing on the acquired backscattering intensity image data.

在本发明的一个具体实施例中，还可以利用数字高程模型和卫星轨道信息对极化基转换得到的九个后向散射强度图像数据进行正射校正处理，消除地形对图像质量的影响。In a specific embodiment of the present invention, the digital elevation model and satellite orbit information can also be used to perform orthorectification processing on the nine backscattering intensity image data obtained through polarization base conversion, so as to eliminate the influence of terrain on image quality.

步骤27，根据后向散射强度图像数据，进行全极化SAR图像分类处理。Step 27, according to the backscattering intensity image data, perform full-polarization SAR image classification processing.

基于极化基转换后，复数形式的全极化SAR图像数据对应的协方差矩阵就可以变成全部由强度量表示的图像数据。这样一组符合高斯分布的强度量图像即可以应用基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法进行分类处理。After conversion based on the polarization basis, the covariance matrix corresponding to the fully polarimetric SAR image data in the form of complex numbers can become image data represented by intensity quantities. Such a group of intensity images conforming to the Gaussian distribution can be applied to remote sensing image processing methods based on remote sensing image data subject to uniform distribution, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images for classification.

在本发明的一个具体实施例中，可以采用eCognition软件(eCognition，2005)提供的多分辨率分割方法进行分割处理。该方法是以从一个像素对象开始自下而上的区域合并技术为基础，综合考虑图像的颜色信息和形状信息进行图像分割，基于分割得到的图像块进行面向对象分类，分类方法可以是非监督分类，也可以是监督分类，如最大似然法和最小距离法等。In a specific embodiment of the present invention, the multi-resolution segmentation method provided by eCognition software (eCognition, 2005) can be used for segmentation processing. This method is based on a bottom-up region merging technology starting from a pixel object, and comprehensively considers the color information and shape information of the image for image segmentation, and performs object-oriented classification based on the segmented image blocks. The classification method can be unsupervised classification , and can also be supervised classification, such as maximum likelihood method and minimum distance method.

在本发明的一个具体实施例中，还可以利用野外测量和高分辨率SAR图像解译获取检验样本多边形作为精度评价的参考数据。建立混淆矩阵，利用总体精度、生产者精度、用户精度和Kappa系数生成精度评价报告。In a specific embodiment of the present invention, the inspection sample polygon can also be obtained by using field measurement and high-resolution SAR image interpretation as reference data for accuracy evaluation. Establish a confusion matrix, and use the overall accuracy, producer accuracy, user accuracy and Kappa coefficient to generate an accuracy evaluation report.

通过上述描述可以看出，本发明实施例提供的全极化SAR图像数据分类方法的基础为极化基的转换和面向对象的分类。其中极化基的转换使得服从不同统计分布的全极化SAR图像数据由一组符合统一分布的强度量来表示，从而解决了大量成熟的基于遥感图像数据服从统一分布的遥感图像处理方法，例如光学遥感图像处理方法应用于全极化SAR图像数据处理的可能性。因此，本发明实施例提供的全极化SAR图像数据分类方法具有很强的实用价值。It can be seen from the above description that the basis of the classification method for full-polarization SAR image data provided by the embodiment of the present invention is the conversion of the polarization basis and the object-oriented classification. Among them, the conversion of the polarization basis makes the fully polarized SAR image data subject to different statistical distributions represented by a set of intensity quantities conforming to the uniform distribution, thus solving a large number of mature remote sensing image processing methods based on remote sensing image data subject to the uniform distribution, such as The possibility of optical remote sensing image processing method applied to full polarization SAR image data processing. Therefore, the full-polarization SAR image data classification method provided by the embodiment of the present invention has strong practical value.

而且，在极化基转换后得到的强度图像的基础上应用面向对象分类方法对全极化SAR图像进行分类处理，可以消除SAR图像斑点噪声的影响，对地形破碎地区的高分辨率SAR图像的分类也能达到好的效果，提高分类的精度。Moreover, on the basis of the intensity image obtained after polarization base conversion, the object-oriented classification method is applied to classify the full-polarization SAR image, which can eliminate the influence of speckle noise in the SAR image and improve the accuracy of the high-resolution SAR image in the terrain fragmented area. Classification can also achieve good results and improve the accuracy of classification.

本发明实施例还提供了一种全极化SAR图像分类装置，如附图3所示，该装置具体可以包括：The embodiment of the present invention also provides a full-polarization SAR image classification device, as shown in Figure 3, the device may specifically include:

转换模块31，用于将全极化SAR图像数据转换为对应的后向散射强度图像数据；A conversion module 31, configured to convert the full-polarization SAR image data into corresponding backscattering intensity image data;

分类模块32，用于根据所述后向散射强度图像数据，进行全极化SAR图像分类处理。The classification module 32 is configured to perform full-polarization SAR image classification processing according to the backscattering intensity image data.

本发明实施例提供了全极化SAR图像分类装置，通过将服从不同统计分布的全极化SAR图像数据转换为对应的服从统一分布的后向散射强度图像数据，从而可以使现有基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法，适于对全极化SAR图像数据进行处理，进而提高了全极化SAR图像数据的分类精度。An embodiment of the present invention provides a device for classifying fully polarimetric SAR images. By converting fully polarimetric SAR image data subject to different statistical distributions into corresponding backscattering intensity image data subject to uniform distribution, the existing remote sensing image-based Remote sensing image processing methods with data subject to uniform distribution, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images, are suitable for processing full-polarization SAR image data, thereby improving the classification accuracy of full-polarization SAR image data.

在本发明一个可选实施例中，转换模块31具体可以包括辐射定标处理单元311，转换单元312，多视化处理单元313，极化基转换处理单元314。其中：In an optional embodiment of the present invention, the conversion module 31 may specifically include a radiation calibration processing unit 311 , a conversion unit 312 , a multi-visualization processing unit 313 , and a polarization base conversion processing unit 314 . in:

辐射定标处理单元311，用于将获取的全极化SAR图像数据对应的复散射矩阵进行辐射定标处理。The radiation calibration processing unit 311 is configured to perform radiation calibration processing on the complex scattering matrix corresponding to the acquired full-polarization SAR image data.

具体的，辐射定标处理单元311可以对装置获取的全极化SAR图像数据对应的复散射矩阵

进行辐射定标处理，得到辐射定标处理后的复散射矩阵。Specifically, the radiation calibration processing unit 311 can calculate the complex scattering matrix corresponding to the fully polarized SAR image data acquired by the device

Radiation calibration processing is performed to obtain a complex scattering matrix after radiation calibration processing.

复散射矩阵中，S_HH表示复散射矩阵中水平(H)极化发射、水平(H)极化接收的散射部分，S_HV表示复散射矩阵中水平(H)极化发射、垂直(V)极化接收的散射部分，S_VH表示复散射矩阵中垂直(V)极化发射、水平(H)极化接收的散射部分，S_VV表示复散射矩阵中垂直(V)极化发射、垂直(V)极化接收的散射部分。In the complex scattering matrix, _SHH represents the scattering part of horizontal (H) polarized emission and horizontal (H) polarized reception in the complex scattering matrix, and _SHV represents the horizontal (H) polarized emission, vertical (V) polarized part of the complex scattering matrix Scattering part of polarization reception, S _VH represents the scattering part of vertical (V) polarization emission and horizontal (H) polarization reception in the complex scattering matrix, S _VV represents the vertical (V) polarization emission, vertical ( V) Scattered part of polarized reception.

转换单元312，用于将经过辐射定标处理单元311辐射定标处理后的复散射矩阵转换为对应的协方差矩阵。The conversion unit 312 is configured to convert the complex scattering matrix processed by the radiation calibration processing unit 311 into a corresponding covariance matrix.

具体的，转换单元312可以基于后向散射互易性原理，可令复散射矩阵中的参数S_HV＝S_VH，则全极化SAR图像数据对应的复散射矩阵可以用于一个复矢量h表示，具体可如公式(2)所示：Specifically, the conversion unit 312 can be based on the principle of backscattering reciprocity, and the parameter S _HV =S _VH in the complex scattering matrix can be set, then the complex scattering matrix corresponding to the fully polarized SAR image data can be used for a complex vector h to represent , as shown in formula (2):

其中，参数T表示矩阵转置，参数S_HV前的

多视化处理单元313，用于将转换单元312经过转换得到的协方差矩阵进行多视化处理。The multi-visualization processing unit 313 is configured to perform multi-visualization processing on the covariance matrix converted by the conversion unit 312 .

由于经过辐射定标处理单元311辐射定标处理得到的复散射矩阵为单视复散射矩阵，因此，转换单元312由单视复散射矩阵得到的协方差矩阵C同样为单视的，为了消除SAR图像数据中的噪声，本发明实施例中，多视化处理单元313可对经过转换单元312转换处理得到的单视协方差矩阵进行多视化处理，从而得到矩阵C：Since the complex scattering matrix obtained by the radiation calibration processing unit 311 is a single-view complex scattering matrix, the covariance matrix C obtained by the conversion unit 312 from the single-view complex scattering matrix is also single-view. In order to eliminate SAR Noise in the image data, in the embodiment of the present invention, the multi-view processing unit 313 can perform multi-view processing on the single-view covariance matrix converted and processed by the conversion unit 312, thereby obtaining the matrix C:

极化基转换处理单元314，用于将多视化处理单元313经过多视化处理后的协方差矩阵C’进行极化基转换处理，获取全极化SAR图像数据对应的后向散射强度图像数据。The polarization-based conversion processing unit 314 is configured to perform polarization-based conversion processing on the covariance matrix C' after multi-visualization processing by the multi-visualization processing unit 313, and obtain the backscattering intensity image corresponding to the full-polarization SAR image data data.

具体的，由于全极化SAR图像数据服从不同的统计分布，导致基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法，在对SAR图像数据进行处理和信息提取时存在不便，因此本发明实施例中，可将全极化SAR图像数据中的参数用服从同一分布，例如高斯分布的参数形式表示，使得基于遥感图像数据服从统一分布的遥感图像处理方法可以被用于全极化SAR图像数据的处理。Specifically, since the full-polarization SAR image data obeys different statistical distributions, the remote sensing image processing methods based on the uniform distribution of remote sensing image data, such as the segmentation and object-oriented classification methods based on the development of optical remote sensing images, are used in the SAR image data There is inconvenience in processing and information extraction, so in the embodiment of the present invention, the parameters in the fully polarized SAR image data can be expressed in the form of parameters that obey the same distribution, such as Gaussian distribution, so that the remote sensing image based on the remote sensing image data obeys the uniform distribution The processing method can be used for the processing of full polarization SAR image data.

其中，in,

在本发明一个可选实施例中，分类模块32具体可以包括正射校正处理单元321，用于对转换模块31获取的后向散射强度图像数据进行正射校正处理。In an optional embodiment of the present invention, the classification module 32 may specifically include an orthorectification processing unit 321 configured to perform orthorectification processing on the backscattering intensity image data acquired by the conversion module 31 .

具体的，正射校正处理单元321可以利用数字高程模型和卫星轨道信息对转换模块31得到的九个后向散射强度图像数据进行正射校正处理，消除地形对图像质量的影响。Specifically, the orthorectification processing unit 321 can use the digital elevation model and satellite orbit information to perform orthorectification processing on the nine backscatter intensity image data obtained by the conversion module 31, so as to eliminate the influence of terrain on image quality.

由于基于极化基转换后，复数形式的全极化SAR图像数据对应的协方差矩阵就可以变成全部由强度量表示的图像数据。这样一组符合高斯分布的强度量图像即可以应用基于遥感图像数据服从统一分布的遥感图像处理方法，例如基于光学遥感图像发展的分割与面向对象分类方法进行分类处理。因此，在本发明的一个可选实施例中，分类模块32可以采用eCognition软件(eCognition，2005)提供的多分辨率分割方法进行分割处理。该方法是以从一个像素对象开始自下而上的区域合并技术为基础，综合考虑图像的颜色信息和形状信息进行图像分割，基于分割得到的图像块进行面向对象分类，分类方法可以是非监督分类，也可以是监督分类，如最大似然法和最小距离法等。After conversion based on the polarization basis, the covariance matrix corresponding to the complex number form of fully polarimetric SAR image data can become image data represented by intensity quantities. Such a group of intensity images conforming to the Gaussian distribution can be applied to remote sensing image processing methods based on remote sensing image data subject to uniform distribution, such as segmentation and object-oriented classification methods based on the development of optical remote sensing images for classification. Therefore, in an optional embodiment of the present invention, the classification module 32 can use the multi-resolution segmentation method provided by eCognition software (eCognition, 2005) to perform segmentation processing. This method is based on a bottom-up region merging technology starting from a pixel object, and comprehensively considers the color information and shape information of the image for image segmentation, and performs object-oriented classification based on the segmented image blocks. The classification method can be unsupervised classification , and can also be supervised classification, such as maximum likelihood method and minimum distance method.

通过上述描述可以看出，本发明实施例提供的全极化SAR图像数据分类装置，通过极化基的转换使得服从不同统计分布的全极化SAR图像数据由一组符合统一分布的强度量来表示，从而解决了大量成熟的基于遥感图像数据服从统一分布的遥感图像处理方法，例如光学遥感图像处理方法应用于全极化SAR图像数据处理的可能性。因此，本发明实施例提供的全极化SAR图像数据分类方法具有很强的实用价值。It can be seen from the above description that the fully polarimetric SAR image data classification device provided by the embodiment of the present invention makes the fully polarimetric SAR image data subject to different statistical distributions be classified by a set of intensity quantities conforming to a uniform distribution through the transformation of the polarization basis. In this way, a large number of mature remote sensing image processing methods based on the uniform distribution of remote sensing image data are solved, such as the possibility of optical remote sensing image processing methods being applied to full polarization SAR image data processing. Therefore, the full-polarization SAR image data classification method provided by the embodiment of the present invention has strong practical value.

以上所述仅是本发明的实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以作出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。The above is only the embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications should also be regarded as Be the protection scope of the present invention.

Claims

1. A kind of all-polarization synthetic aperture radar SAR image classification method, it is characterized in that, comprising:

Convert the full polarization SAR image data into corresponding backscatter intensity image data;

According to the backscattering intensity image data, full polarization SAR image classification processing is performed.

2. The method according to claim 1, wherein said converting the full polarization SAR image data into corresponding backscattering intensity image data comprises:

Perform radiation calibration processing on the complex scattering matrix corresponding to the acquired full-polarization SAR image data;

Convert the complex scattering matrix processed by radiation calibration into the corresponding covariance matrix;

Carrying out multi-visualization processing to the covariance matrix;

The covariance matrix after multi-visualization processing is subjected to polarization base conversion processing to obtain the backscattering intensity image data corresponding to the full polarization SAR image data.

3. The method according to claim 2, wherein said converting the complex scattering matrix after radiation calibration processing into a corresponding covariance matrix comprises:

Let the complex scattering matrix

The parameter S _HV =S _VH in , wherein, _SHH represents the scattering part of horizontally polarized transmission and horizontally polarized reception in the complex scattering matrix, and _SHV represents the horizontally polarized transmission and vertical polarization in the complex scattering matrix The received scattering part, S _VH represents the scattering part of vertical polarization transmission and horizontal polarization reception in the complex scattering matrix, and S _VV represents the scattering part of vertical polarization transmission and vertical polarization reception in the complex scattering matrix;

Represent described complex scattering matrix with complex vector h, expression formula is Among them, the superscript T indicates matrix transposition;

According to the formula

4. method according to claim 3, is characterized in that, described covariance matrix is carried out the covariance matrix C ' expression that obtains after multi-visualization processing is:

{C C}^{' '} = = [\begin{matrix} < < {S S}_{hh hh} {S S}_{hh hh}^{* *} > > & < < {S S}_{hh hh} {S S}_{hv hv}^{* *} > > & < < {S S}_{hh hh} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{hv hv} {S S}_{hh hh}^{* *} > > & < < {S S}_{hv hv} {S S}_{hv hv}^{* *} > > & < < {S S}_{hv hv} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{vv vv} {S S}_{hh hh}^{* *} > > & < < {S S}_{vv vv} {S S}_{hv hv}^{* *} > > & < < {S S}_{vv vv} {S S}_{vv vv}^{* *} > > \end{matrix}]

5. The method according to claim 2, wherein the covariance matrix after multi-visualization processing is carried out into a polarization base conversion process, and obtaining the corresponding backscattering intensity image data of the full polarization SAR image data comprises :

The three real number element parameters in the upper triangular matrix included in the multi-visualization processed covariance matrix, and the real part Re parameter and the imaginary part IM parameter of the three complex number parameters are converted by polarization basis , to obtain the corresponding backscattering intensity, the conversion formula is:

A A (\begin{matrix} < < {S S}_{hh hh} {S S}_{hh hh}^{* *} > > \\ < < {S S}_{vv vv} {S S}_{vv vv}^{* *} > > \\ < < {S S}_{hv hv} {S S}_{hv hv}^{* *} > > \\ Re Re [[< < {S S}_{hh hh} {S S}_{vv vv}^{* *} > >]] \\ IM IM [[< < {S S}_{hh hh} {S S}_{vv vv}^{* *} > >]] \\ Re Re [[< < {S S}_{hh hh} {S S}_{hv hv}^{* *} > >]] \\ IM IM [[< < {S S}_{hh hh} {S S}_{hv hv}^{* *} > >]] \\ Re Re [[< < {S S}_{hv hv} {S S}_{vv vv}^{* *} > >]] \\ IM IM [[< < {S S}_{hv hv} {S S}_{vv vv}^{* *} > >]] \end{matrix}) = = (\begin{matrix} {σ σ}_{hh hh}^{00} \\ {σ σ}_{vv vv}^{00} \\ {σ σ}_{+ + + + 4545}^{00} \\ {σ σ}_{- - - - 4545}^{00} \\ {σ σ}_{ll ll}^{00} \\ {σ σ}_{rr rr}^{00} \\ {σ σ}_{h h + + 4545}^{00} \\ {σ σ}_{hl hl}^{00} \\ {σ σ}_{+ + 4545 l l}^{00} \end{matrix})

in,

A A = = [\begin{matrix} 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & 6.28319 6.28319 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & 6.28319 6.28319 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & - - 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & - - 12.5664 12.5664 \\ 3.14159 3.14159 & 3.14159 3.14159 & 12.5664 12.5664 & - - 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 12.5664 12.5664 \\ 6.28319 6.28319 & 0.0000 0.0000 & 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 \\ 6.28319 6.28319 & 0.0000 0.0000 & 6.28319 6.28319 & 0.0000 0.0000 & 0.0000 0.0000 & 0.0000 0.0000 & - - 12.5664 12.5664 & 0.0000 0.0000 & 0.0000 0.0000 \\ 6.28319 6.28319 & 3.14159 3.14159 & 6.28319 6.28319 & 0.0000 0.0000 & - - 6.28319 6.28319 & 6.28319 6.28319 & - - 6.28319 6.28319 & 6.28319 6.28319 & 6.28319 6.28319 \end{matrix}]

The parameter σ is the backscattering coefficient, and the subscript of the parameter σ represents the receiving and transmitting polarization modes of the polarized base: h stands for horizontal, v stands for vertical, l stands for left circle, r stands for right circle, + or +45 stands for +45 °Linear, - or -45 stands for 45° linear.

6. The method according to claim 1, wherein, according to the backscattering intensity image data, performing full polarization SAR image classification processing comprises:

Perform orthorectification processing on the acquired backscattering intensity image data.

7. A fully polarized synthetic aperture radar SAR image classification device, characterized in that it comprises:

A conversion module, configured to convert full-polarization SAR image data into corresponding backscattering intensity image data;

A classification module, configured to perform full-polarization SAR image classification processing according to the backscattering intensity image data.

8. The device according to claim 7, wherein the conversion module comprises:

a radiation calibration processing unit, configured to perform radiation calibration processing on the complex scattering matrix corresponding to the acquired full-polarization SAR image data;

A conversion unit, configured to convert the complex scattering matrix processed by radiation calibration into a corresponding covariance matrix;

A multi-visualization processing unit, configured to perform multi-visualization processing on the covariance matrix;

The polarization base conversion processing unit is configured to perform polarization base conversion processing on the multi-view processed covariance matrix to obtain backscattering intensity image data corresponding to the full polarization SAR image data.

9. The device according to claim 7, wherein the classification module comprises:

The orthorectification processing unit is configured to perform orthorectification processing on the backscattering intensity image data acquired by the conversion module.