CN101976357A - Method and device for classifying fully polarimetric synthetic aperture radar image - Google Patents
Method and device for classifying fully polarimetric synthetic aperture radar image Download PDFInfo
- Publication number
- CN101976357A CN101976357A CN201010508761XA CN201010508761A CN101976357A CN 101976357 A CN101976357 A CN 101976357A CN 201010508761X A CN201010508761X A CN 201010508761XA CN 201010508761 A CN201010508761 A CN 201010508761A CN 101976357 A CN101976357 A CN 101976357A
- Authority
- CN
- China
- Prior art keywords
- msubsup
- mtd
- mtr
- msub
- image data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 239000011159 matrix material Substances 0.000 claims abstract description 126
- 238000012545 processing Methods 0.000 claims abstract description 101
- 230000010287 polarization Effects 0.000 claims description 51
- 238000006243 chemical reaction Methods 0.000 claims description 48
- 238000012800 visualization Methods 0.000 claims description 27
- 230000005540 biological transmission Effects 0.000 claims description 16
- 230000017105 transposition Effects 0.000 claims description 8
- 238000009826 distribution Methods 0.000 abstract description 24
- 238000009827 uniform distribution Methods 0.000 abstract description 22
- 238000003672 processing method Methods 0.000 abstract description 19
- 230000003287 optical effect Effects 0.000 abstract description 14
- 238000011161 development Methods 0.000 abstract description 8
- 238000000638 solvent extraction Methods 0.000 abstract 1
- 230000011218 segmentation Effects 0.000 description 14
- 230000018109 developmental process Effects 0.000 description 7
- 238000003709 image segmentation Methods 0.000 description 3
- 238000007476 Maximum Likelihood Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention provides a method and a device for classifying a fully polarimetric synthetic aperture radar (SAR) image. The method comprises the following steps of: converting a covariance matrix of fully polarimetric SAR image data in a plural form into image data which is fully expressed by an intensive quantity; and classifying the fully polarimetric SAR image based on converted image data expressed by the intensive quantity. Thus, the fully polarimetric SAR image data which obeys different statistical distribution is converted into corresponding backscattering intensity image data which obeys uniform distribution, the conventional remote sensing image processing method based on the remote sensing image data which obeys the uniform distribution, such as an optical remote sensing image development-based partitioning method and an object-oriented classifying method, can be suitable for processing the fully polarimetric SAR image, and the classification accuracy of the fully polarimetric SAR image data is improved.
Description
Technical Field
The invention relates to the technical field of remote sensing image processing, in particular to a method and a device for classifying fully-polarized synthetic aperture radar images.
Background
Many object-oriented remote sensing image processing methods have been developed, and among them, the multi-resolution image segmentation and object-oriented classification techniques developed based on optical remote sensing images are most widely applied. However, the existing remote sensing image processing method is generally applied based on the assumption that the remote sensing image data obeys uniform distribution, such as gaussian distribution.
However, due to different imaging mechanisms, Synthetic Aperture Radar (SAR) image data in a remote sensing image obeys different statistical distributions, which causes the statistical distribution of fully polarized SAR image data to be different from uniform distribution, so that the current remote sensing image processing method based on the unified distribution of the remote sensing image data, such as segmentation and object-oriented classification methods based on optical remote sensing image development, is not suitable for processing the SAR image data, and has a condition of reduced classification accuracy.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a method and a device for classifying images of a fully-polarized synthetic aperture radar, so that the image classification precision of the fully-polarized synthetic aperture radar is improved.
In order to solve the technical problems, the invention provides the following scheme:
the embodiment of the invention provides a classification method of a full-polarization Synthetic Aperture Radar (SAR) image, which comprises the following steps:
converting the fully polarized SAR image data into corresponding backscattering intensity image data;
and carrying out complete polarization SAR image classification processing according to the backscattering intensity image data.
Preferably, the converting the fully polarized SAR image data into corresponding backscatter intensity image data includes:
carrying out radiometric calibration processing on a complex scattering matrix corresponding to the acquired fully-polarized SAR image data;
converting the complex scattering matrix subjected to radiometric calibration processing into a corresponding covariance matrix;
performing multi-visualization processing on the covariance matrix;
and carrying out polarization base conversion processing on the covariance matrix after the multi-visualization processing to obtain backscattering intensity image data corresponding to the fully polarized SAR image data.
Preferably, the converting the complex scattering matrix after the radiometric calibration process into a corresponding covariance matrix includes:
let the complex scattering matrixParameter S ofHV＝SVHWherein S isHHRepresenting the scattering fraction of horizontally polarised transmission, horizontally polarised reception, S, of said complex scattering matrixHVRepresenting the scattering fraction of horizontally polarized transmission, vertically polarized reception in said complex scattering matrix, SVHRepresenting the scattering fraction of the vertically polarized transmission, horizontally polarized reception, S, of said complex scattering matrixVVA scattering portion representing vertical polarization transmission and vertical polarization reception in the complex scattering matrix;
the complex scattering matrix is expressed by a complex vector h with the formula
Wherein, the superscript T represents the matrix transposition;
according to the formula
And calculating to obtain a covariance matrix C corresponding to the complex scattering matrix after radiometric calibration, wherein the superscript x represents complex conjugate, and the superscript T represents matrix transposition.
Preferably, the expression of the covariance matrix C' obtained after performing multi-visualization processing on the covariance matrix is as follows:
preferably, the performing polarization-based conversion processing on the multi-visualization processed covariance matrix to obtain backscatter intensity image data corresponding to the fully-polarized SAR image data includes:
and converting the real number element parameters of 3 real number elements in an upper triangular matrix contained in the covariance matrix after the multi-visualization processing, and the real part parameter Re and the imaginary part parameter IM in the 3 complex number parameters through a polarization basis to obtain a corresponding backscattering strength quantity, wherein the conversion formula is as follows:
wherein,
the parameter σ is the backscattering coefficient, and the subscripts of the parameter σ represent the receive and transmit polarization modes of the polarization base: h represents horizontal, v represents vertical, l represents left circle, r represents right circle, + or +45 represents +45 ° linear, -or-45 represents-45 ° linear.
Preferably, the performing the classification processing of the fully polarimetric SAR image according to the backscatter intensity image data includes:
and performing orthorectification processing on the acquired backscatter intensity image data.
The embodiment of the invention also provides a classification device for the image of the full-polarization Synthetic Aperture Radar (SAR), which comprises the following steps:
the conversion module is used for converting the fully polarized SAR image data into corresponding backscattering intensity image data;
and the classification module is used for carrying out complete polarization SAR image classification processing according to the backscattering intensity image data.
Preferably, the conversion module includes:
the radiometric calibration processing unit is used for carrying out radiometric calibration processing on the complex scattering matrix corresponding to the acquired fully-polarized SAR image data;
the conversion unit is used for converting the complex scattering matrix subjected to radiometric calibration processing into a corresponding covariance matrix;
the multi-visualization processing unit is used for performing multi-visualization processing on the covariance matrix;
and the polarization base conversion processing unit is used for carrying out polarization base conversion processing on the covariance matrix subjected to the multi-visualization processing to obtain backscattering intensity image data corresponding to the fully polarized SAR image data.
Preferably, the classification module includes:
and the orthorectification processing unit is used for performing orthorectification processing on the backscatter intensity image data acquired by the conversion module.
From the above, the method and the device for classifying the fully polarimetric synthetic aperture radar SAR image provided by the invention convert the covariance matrix of the complex fully polarimetric SAR image data into a form completely represented by the intensity, and perform the classification processing of the fully polarimetric 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.
Drawings
Fig. 1 is a first schematic flow chart of a classification method of fully-polarized SAR image data according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a complete polarization SAR image data classification method according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram i of a fully-polarized SAR image data classification device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a fully-polarized SAR image data classification device according to an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a fully polarimetric SAR image classification method, which specifically comprises the following steps as shown in the attached figure 1:
and step 12, carrying out complete polarization SAR image classification processing according to the backscattering intensity image data.
The embodiment of the invention provides a fully-polarized SAR image classification method, which can convert fully-polarized SAR image data subjected to different statistical distributions into corresponding backscattering intensity image data subjected to uniform distribution, so that the existing remote sensing image processing method based on the remote sensing image data subjected to uniform distribution, such as a segmentation and object-oriented classification method based on optical remote sensing image development, is suitable for processing the fully-polarized SAR image data, and further improves the classification precision of the fully-polarized SAR image data.
In an optional embodiment of the present invention, converting the fully polarized SAR image data into corresponding backscatter intensity image data may specifically include the following steps:
carrying out radiometric calibration processing on a complex scattering matrix corresponding to the acquired fully-polarized SAR image data;
converting the complex scattering matrix subjected to radiometric calibration processing into a corresponding covariance matrix;
performing multi-visualization processing on the covariance matrix;
and carrying out polarization base conversion processing on the covariance matrix after the multi-visualization processing to obtain backscattering intensity image data corresponding to the fully polarized SAR image data.
In an optional embodiment of the present invention, the performing the classification processing on the fully polarimetric SAR image according to the backscatter intensity image data may specifically include the following steps:
and performing orthorectification processing on the acquired backscatter intensity image data.
A specific embodiment of the fully polarimetric SAR image classification method provided in the embodiment of the present invention is described in detail below with reference to fig. 2, and specifically includes:
and step 21, acquiring the fully polarized SAR image data.
Specifically, the fully polarized SAR image data can be acquired by a polarized radar.
Since the specific expression form of the fully polarimetric SAR image data may be a complex scattering matrix, in the embodiment of the present invention, the complex scattering matrix may be represented by a horizontal and vertical linear polarization basis, which may be specifically represented by formula (1):
wherein S isHHRepresenting the scattering fraction of horizontally (H) -polarized transmission, horizontally (H) -polarized reception, S, in a complex scattering matrixHVRepresents the scattering portion of the horizontally (H) -polarized transmission and vertically (V) -polarized reception in the complex scattering matrix,SVHrepresenting the scattered part of the vertical (V) -polarized transmission, horizontal (H) -polarized reception, S, of a complex scattering matrixVVRepresents the scattered portion of the vertically (V) polarized transmission, vertically (V) polarized reception in the complex scattering matrix.
The fully polarimetric SAR image data acquired in the embodiment of the present invention can be represented by a complex scattering matrix in a complex form shown in formula (1).
And step 22, carrying out radiometric calibration processing on the acquired full-polarization SAR image data.
In a specific embodiment of the present invention, the obtained fully polarized SAR image data may be subjected to radiometric calibration to obtain a complex scattering matrix after radiometric calibration.
And 23, converting the complex scattering matrix subjected to radiometric calibration processing into a corresponding covariance matrix.
Based on the principle of reciprocity of backscattering, in the embodiment of the invention, the parameter S in the complex scattering matrix can be controlledHV＝SVHThen, the complex scattering matrix corresponding to the fully-polarized SAR image data can be represented by a complex vector h, which can be specifically expressed by formula (2):
where the parameter T represents the matrix transposition and the parameter S represents the matrix transpositionHVFront stage
In order to ensure consistency of the total power calculation. Thereby, a covariance matrix corresponding to the complex scattering matrix can be calculatedC：
Where the superscript denotes complex conjugate and the superscript T denotes matrix transposition.
And 24, performing multi-visualization processing on the covariance matrix.
Since the complex scattering matrix obtained through the radiometric calibration processing 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, and in order to eliminate noise in the SAR image data, in the embodiment of the present invention, the single-view covariance matrix obtained through step 23 may be subjected to multi-view processing, so as to obtain the matrix C:
and 25, performing polarization base conversion processing on the covariance matrix after the multi-visualization processing.
Because the fully-polarized SAR image data obeys different statistical distributions, the remote sensing image processing method based on the remote sensing image data obeying the uniform distribution, such as a segmentation and object-oriented classification method based on the development of an optical remote sensing image, has inconvenience in processing and information extraction of the SAR image data, so in the embodiment of the invention, parameters in the fully-polarized SAR image data can be represented in a parameter form obeying the same distribution, such as Gaussian distribution, and the remote sensing image processing method based on the remote sensing image data obeying the uniform distribution can be used for processing the fully-polarized SAR image data.
In order to unify the statistical distribution form of parameters in the fully-polarized SAR image data, in the embodiment of the invention, a polarization-based conversion processing method which converts the covariance matrix element parameters corresponding to the fully-polarized SAR image data into the parameters only expressed by strength can be specifically adopted.
Since the information of the covariance matrix can be represented by the correlation terms of 3 real element parameters and 3 complex parameter quantities in the upper triangular matrix contained in the covariance matrix, and each complex term can be represented by 1 real part (Re) parameter quantity and 1 imaginary part (IM) parameter quantity, the information of the covariance matrix can be completely represented by the 9 parameter quantities.
In the embodiment of the present invention, the 9 parameter quantities may be converted into 9 backscatter strength quantities through polarization base conversion, and the specific polarization base conversion may be shown as formula (5):
wherein,
in equation (5), the parameter σ is the backscattering coefficient, and the subscript of the parameter σ represents the receive and transmit polarization of the polarization basis: h represents horizontal, v represents vertical, l represents left circle, r represents right circle, + or +45 represents +45 ° linear, -or-45 represents-45 ° linear.
After the processing, 9 backscattering intensity image data corresponding to the fully polarized SAR image data can be obtained, which is equivalent to 9 wave bands of the optical remote sensing image.
Because the covariance matrix corresponding to the fully-polarized SAR image data is subjected to multi-visualization processing, the noise of 9 backscattering intensity images corresponding to the fully-polarized SAR image obtained at the moment is effectively removed, and the statistical distribution of the data approximately conforms to uniform distribution, such as Gaussian distribution.
In a specific embodiment of the present invention, the digital elevation model and the satellite orbit information may also be used to perform an orthorectification process on the nine backscatter intensity image data obtained by the polarization base conversion, so as to eliminate the influence of the terrain on the image quality.
And 27, carrying out complete polarization SAR image classification processing according to the backscattering intensity image data.
After the polarization-based conversion, the covariance matrix corresponding to the fully-polarized SAR image data in the complex form can be changed into image data which is completely represented by the intensity value. Such a group of intensity images conforming to gaussian distribution can be subjected to a remote sensing image processing method based on the remote sensing image data and subject to uniform distribution, such as segmentation based on optical remote sensing image development and object-oriented classification method for classification processing.
In one embodiment of the present invention, the segmentation process may be performed using a multi-resolution segmentation method provided by eCoginization software (eCoginization, 2005). The method is based on a region merging technology from bottom to top from a pixel object, the image segmentation is carried out by comprehensively considering the color information and the shape information of the image, the object-oriented classification is carried out based on the image blocks obtained by segmentation, and the classification method can be unsupervised classification or supervised classification, such as a maximum likelihood method, a minimum distance method and the like.
In one embodiment of the invention, the field measurement and the high-resolution SAR image interpretation can also be utilized to acquire the polygon of the test sample as the reference data for precision evaluation. And establishing a confusion matrix, and generating a precision evaluation report by using the overall precision, the producer precision, the user precision and the Kappa coefficient.
As can be seen from the above description, the basis of the classification method for fully polarimetric SAR image data provided by the embodiment of the present invention is the conversion of the polarization basis and the object-oriented classification. The conversion of the polarization base enables the fully polarized SAR image data which obey different statistical distributions to be represented by a group of strength quantities which accord with uniform distribution, thereby solving the possibility that a large number of mature remote sensing image processing methods which are based on the remote sensing image data and obey uniform distribution, such as optical remote sensing image processing methods, are applied to the fully polarized SAR image data processing. Therefore, the method for classifying the fully-polarized SAR image data provided by the embodiment of the invention has high practical value.
Moreover, the object-oriented classification method is applied to classify the fully-polarized SAR image on the basis of the intensity image obtained after polarization base conversion, so that the influence of speckle noise of the SAR image can be eliminated, a good effect can be achieved on the classification of the high-resolution SAR image in a terrain-broken area, and the classification precision is improved.
An embodiment of the present invention further provides a fully polarized SAR image classification device, as shown in fig. 3, the device may specifically include:
a conversion module 31, configured to convert the fully-polarized SAR image data into corresponding backscatter intensity image data;
and the classification module 32 is used for carrying out complete polarization SAR image classification processing according to the backscatter intensity image data.
The embodiment of the invention provides a fully-polarized SAR image classification device, which can convert fully-polarized SAR image data subjected to different statistical distributions into corresponding backscattering intensity image data subjected to uniform distribution, so that the existing remote sensing image processing method based on the remote sensing image data subjected to uniform distribution, such as a segmentation and object-oriented classification method based on optical remote sensing image development, is suitable for processing the fully-polarized SAR image data, and further improves the classification precision of the fully-polarized SAR image data.
In an alternative embodiment of the present invention, the conversion module 31 may specifically include a radiometric scaling processing unit 311, a conversion unit 312, a multi-visualization processing unit 313, and a polarization-based conversion processing unit 314. Wherein:
and the radiometric calibration processing unit 311 is configured to perform radiometric calibration processing on the complex scattering matrix corresponding to the acquired fully-polarized SAR image data.
Specifically, the radiometric calibration processing unit 311 may apply to a complex scattering matrix corresponding to the fully-polarized SAR image data acquired by the apparatus
And carrying out radiometric calibration processing to obtain the complex scattering matrix after radiometric calibration processing.
In complex scattering matrix, SHHRepresenting the scattering fraction of horizontally (H) -polarized transmission, horizontally (H) -polarized reception, S, in a complex scattering matrixHVRepresenting the scattering fraction of horizontally (H) -polarized transmission and vertically (V) -polarized reception in a complex scattering matrix, SVHRepresenting the scattered part of the vertical (V) -polarized transmission, horizontal (H) -polarized reception, S, of a complex scattering matrixVVRepresents the scattered portion of the vertically (V) polarized transmission, vertically (V) polarized reception in the complex scattering matrix.
A conversion unit 312, configured to convert the complex scattering matrix subjected to the radiometric calibration processing by the radiometric calibration processing unit 311 into a corresponding covariance matrix.
Specifically, the conversion unit 312 may be based on the principle of reciprocal backscattering, and may enable the parameter S in the complex scattering matrixHV＝SVHThen, the complex scattering matrix corresponding to the fully-polarized SAR image data can be represented by a complex vector h, which can be specifically expressed by formula (2):
where the parameter T represents the matrix transposition and the parameter S represents the matrix transpositionHVFront stage
In order to ensure consistency of the total power calculation. Thus, the covariance matrix C corresponding to the complex scattering matrix can be calculated:
where the superscript denotes complex conjugate and the superscript T denotes matrix transposition.
A multi-visualization processing unit 313, configured to perform multi-visualization processing on the covariance matrix obtained through the conversion by the conversion unit 312.
Since the complex scattering matrix obtained by the radiometric calibration processing unit 311 is a single-view complex scattering matrix, the covariance matrix C obtained by the single-view complex scattering matrix by the conversion unit 312 is also single-view, and in order to eliminate noise in the SAR image data, in the embodiment of the present invention, the multi-view processing unit 313 may perform multi-view processing on the single-view covariance matrix obtained by the conversion processing unit 312, so as to obtain the matrix C:
and the polarization base conversion processing unit 314 is configured to perform polarization base conversion processing on the covariance matrix C' subjected to the multi-visualization processing by the multi-visualization processing unit 313, and acquire backscatter intensity image data corresponding to the fully-polarized SAR image data.
Specifically, because the fully-polarized SAR image data obeys different statistical distributions, the remote sensing image processing method based on the remote sensing image data obeys uniform distribution, such as a segmentation and object-oriented classification method based on the development of an optical remote sensing image, has inconvenience in processing and information extraction of the SAR image data, so in the embodiment of the present invention, parameters in the fully-polarized SAR image data can be represented in a parameter form obeying the same distribution, such as gaussian distribution, so that the remote sensing image processing method based on the remote sensing image data obeys uniform distribution can be used for processing the fully-polarized SAR image data.
In order to unify the statistical distribution form of parameters in the fully-polarized SAR image data, in the embodiment of the invention, a polarization-based conversion processing method which converts the covariance matrix element parameters corresponding to the fully-polarized SAR image data into the parameters only expressed by strength can be specifically adopted.
Since the information of the covariance matrix can be represented by the correlation terms of 3 real element parameters and 3 complex parameter quantities in the upper triangular matrix contained in the covariance matrix, and each complex term can be represented by 1 real part (Re) parameter quantity and 1 imaginary part (IM) parameter quantity, the information of the covariance matrix can be completely represented by the 9 parameter quantities.
In the embodiment of the present invention, the 9 parameter quantities may be converted into 9 backscatter strength quantities through polarization base conversion, and the specific polarization base conversion may be shown as formula (5):
wherein,
in equation (5), the parameter σ is the backscattering coefficient, and the subscript of the parameter σ represents the receive and transmit polarization of the polarization basis: h represents horizontal, v represents vertical, l represents left circle, r represents right circle, + or +45 represents +45 ° linear, -or-45 represents-45 ° linear.
After the processing, 9 backscattering intensity image data corresponding to the fully polarized SAR image data can be obtained, which is equivalent to 9 wave bands of the optical remote sensing image.
Because the covariance matrix corresponding to the fully-polarized SAR image data is subjected to multi-visualization processing, the noise of 9 backscattering intensity images corresponding to the fully-polarized SAR image obtained at the moment is effectively removed, and the statistical distribution of the data approximately conforms to uniform distribution, such as Gaussian distribution.
In an optional embodiment of the present invention, the classification module 32 may specifically include an orthorectification processing unit 321, configured to perform an orthorectification process on the backscatter intensity image data acquired by the conversion module 31.
Specifically, the orthorectification processing unit 321 may perform orthorectification processing on the nine backscatter intensity image data obtained by the conversion module 31 by using the digital elevation model and the satellite orbit information, so as to eliminate the influence of the terrain on the image quality.
After the conversion based on the polarization basis, the covariance matrix corresponding to the fully polarized SAR image data in the complex form can be changed into the image data which is completely represented by the intensity. Such a group of intensity images conforming to gaussian distribution can be subjected to a remote sensing image processing method based on the remote sensing image data and subject to uniform distribution, such as segmentation based on optical remote sensing image development and object-oriented classification method for classification processing. Therefore, in an alternative embodiment of the present invention, the classification module 32 may perform the segmentation process using a multi-resolution segmentation method provided by the eCognition software (eCognition, 2005). The method is based on a region merging technology from bottom to top from a pixel object, the image segmentation is carried out by comprehensively considering the color information and the shape information of the image, the object-oriented classification is carried out based on the image blocks obtained by segmentation, and the classification method can be unsupervised classification or supervised classification, such as a maximum likelihood method, a minimum distance method and the like.
As can be seen from the above description, the classification apparatus for fully polarized SAR image data according to the embodiment of the present invention enables fully polarized SAR image data that obey different statistical distributions to be represented by a set of strength amounts that conform to uniform distribution through the conversion of the polarization basis, thereby solving the possibility that a large number of mature remote sensing image processing methods based on the remote sensing image data that obey uniform distribution, such as an optical remote sensing image processing method, are applied to the processing of the fully polarized SAR image data. Therefore, the method for classifying the fully-polarized SAR image data provided by the embodiment of the invention has high practical value.
Moreover, the object-oriented classification method is applied to classify the fully-polarized SAR image on the basis of the intensity image obtained after polarization base conversion, so that the influence of speckle noise of the SAR image can be eliminated, a good effect can be achieved on the classification of the high-resolution SAR image in a terrain-broken area, and the classification precision is improved.
The foregoing is merely an embodiment of the present invention, and it should be noted that those skilled in the art can make various modifications and improvements without departing from the principle of the present invention, and such modifications and improvements should be considered as the protection scope of the present invention.
Claims (9)
1. A full-polarization Synthetic Aperture Radar (SAR) image classification method is characterized by comprising the following steps:
converting the fully polarized SAR image data into corresponding backscattering intensity image data;
and carrying out complete polarization SAR image classification processing according to the backscattering intensity image data.
2. The method of claim 1, wherein converting the fully polarized SAR image data into corresponding backscatter intensity image data comprises:
carrying out radiometric calibration processing on a complex scattering matrix corresponding to the acquired fully-polarized SAR image data;
converting the complex scattering matrix subjected to radiometric calibration processing into a corresponding covariance matrix;
performing multi-visualization processing on the covariance matrix;
and carrying out polarization base conversion processing on the covariance matrix after the multi-visualization processing to obtain backscattering intensity image data corresponding to the fully polarized SAR image data.
3. The method of claim 2, wherein converting the radiometric calibration-processed complex scattering matrix into a corresponding covariance matrix comprises:
let the complex scattering matrix
Parameter S inHV＝SVHWherein S isHHRepresenting the scattering fraction of horizontally polarised transmission, horizontally polarised reception, S, of said complex scattering matrixHVRepresenting the scattering fraction of horizontally polarized transmission, vertically polarized reception in said complex scattering matrix, SVHRepresenting the scattering fraction of the vertically polarized transmission, horizontally polarized reception, S, of said complex scattering matrixVVA scattering portion representing vertical polarization transmission and vertical polarization reception in the complex scattering matrix;
the complex scattering matrix is expressed by a complex vector h with the formulaWherein, the superscript T represents the matrix transposition;
according to the formula
And calculating to obtain a covariance matrix C corresponding to the complex scattering matrix after radiometric calibration, wherein the superscript x represents complex conjugate, and the superscript T represents matrix transposition.
4. The method according to claim 3, wherein the covariance matrix C' obtained after performing multi-visualization on the covariance matrix is expressed as:
5. the method of claim 2, wherein the performing polarization-based conversion processing on the multi-visualization processed covariance matrix to obtain backscatter intensity image data corresponding to the fully-polarized SAR image data comprises:
and converting the real number element parameters of 3 real number elements in an upper triangular matrix contained in the covariance matrix after the multi-visualization processing, and the real part parameter Re and the imaginary part parameter IM in the 3 complex number parameters through a polarization basis to obtain a corresponding backscattering strength quantity, wherein the conversion formula is as follows:
wherein,
the parameter σ is the backscattering coefficient, and the subscripts of the parameter σ represent the receive and transmit polarization modes of the polarization base: h represents horizontal, v represents vertical, l represents left circle, r represents right circle, + or +45 represents +45 ° linear, -or-45 represents 45 ° linear.
6. The method of claim 1, wherein the performing a fully polarimetric SAR image classification process based on the backscatter intensity image data comprises:
and performing orthorectification processing on the acquired backscatter intensity image data.
7. The utility model provides a full polarization synthetic aperture radar SAR image classification device which characterized in that includes:
the conversion module is used for converting the fully polarized SAR image data into corresponding backscattering intensity image data;
and the classification module is used for carrying out complete polarization SAR image classification processing according to the backscattering intensity image data.
8. The apparatus of claim 7, wherein the conversion module comprises:
the radiometric calibration processing unit is used for carrying out radiometric calibration processing on the complex scattering matrix corresponding to the acquired fully-polarized SAR image data;
the conversion unit is used for converting the complex scattering matrix subjected to radiometric calibration processing into a corresponding covariance matrix;
the multi-visualization processing unit is used for performing multi-visualization processing on the covariance matrix;
and the polarization base conversion processing unit is used for carrying out polarization base conversion processing on the covariance matrix subjected to the multi-visualization processing to obtain backscattering intensity image data corresponding to the fully polarized SAR image data.
9. The apparatus of claim 7, wherein the classification module comprises:
and the orthorectification processing unit is used for performing orthorectification processing on the backscatter intensity image data acquired by the conversion module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010508761XA CN101976357A (en) | 2010-10-18 | 2010-10-18 | Method and device for classifying fully polarimetric synthetic aperture radar image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010508761XA CN101976357A (en) | 2010-10-18 | 2010-10-18 | Method and device for classifying fully polarimetric synthetic aperture radar image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101976357A true CN101976357A (en) | 2011-02-16 |
Family
ID=43576241
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010508761XA Pending CN101976357A (en) | 2010-10-18 | 2010-10-18 | Method and device for classifying fully polarimetric synthetic aperture radar image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101976357A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103617427A (en) * | 2013-12-13 | 2014-03-05 | 首都师范大学 | Polarized SAR (Synthetic Aperture Radar) image classifying method |
| CN106815559A (en) * | 2016-12-21 | 2017-06-09 | 中国科学院深圳先进技术研究院 | A kind of utilization SAR data monitoring oyster arranges method and device, the user equipment in region |
| CN108037504A (en) * | 2017-11-28 | 2018-05-15 | 中国科学院遥感与数字地球研究所 | A kind of method of the polarimetric synthetic aperture radar rapid polarization correction based on three point targets |
| CN110023780A (en) * | 2016-11-25 | 2019-07-16 | Iee国际电子工程股份公司 | Changeable Polarization Radar System and method for detect and classify vehicle interior vehicle occupant and other objects |
| CN110516552A (en) * | 2019-07-29 | 2019-11-29 | 南京航空航天大学 | A kind of multipolarization radar image classification method and system based on timing curve |
| CN115236655A (en) * | 2022-09-01 | 2022-10-25 | 成都理工大学 | Landslide identification method, system, equipment and medium based on fully-polarized SAR |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101344587A (en) * | 2008-08-15 | 2009-01-14 | 哈尔滨工业大学 | Multi-component decomposition method used for polarization synthetic aperture radar image |
| CN101464956A (en) * | 2009-01-21 | 2009-06-24 | 电子科技大学 | Method for classifying polarization synthetic aperture radar image based on sub-aperture analysis |
| CN101571915A (en) * | 2009-06-16 | 2009-11-04 | 大连海事大学 | Method for identifying oil spill of SAR image based on characteristic value |
-
2010
- 2010-10-18 CN CN201010508761XA patent/CN101976357A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101344587A (en) * | 2008-08-15 | 2009-01-14 | 哈尔滨工业大学 | Multi-component decomposition method used for polarization synthetic aperture radar image |
| CN101464956A (en) * | 2009-01-21 | 2009-06-24 | 电子科技大学 | Method for classifying polarization synthetic aperture radar image based on sub-aperture analysis |
| CN101571915A (en) * | 2009-06-16 | 2009-11-04 | 大连海事大学 | Method for identifying oil spill of SAR image based on characteristic value |
Non-Patent Citations (2)
| Title |
|---|
| 《2010 IEEE International Geoscience and Remote Sensing Symposium(IGARSS)》 20100730 Xin Tian等 Comparison of crop classification capabilities of spaceborne multi-parameter SAR data 359-362 1-9 , * |
| 《IEEE Transactions on Geoscience and Remote Sensing》 20031231 Hoekman,D.H.等 A new polarimetric classification approach evaluated for agricultural crops 2881-2889 1-9 , * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103617427A (en) * | 2013-12-13 | 2014-03-05 | 首都师范大学 | Polarized SAR (Synthetic Aperture Radar) image classifying method |
| CN103617427B (en) * | 2013-12-13 | 2016-06-29 | 首都师范大学 | Classification of Polarimetric SAR Image method |
| CN110023780A (en) * | 2016-11-25 | 2019-07-16 | Iee国际电子工程股份公司 | Changeable Polarization Radar System and method for detect and classify vehicle interior vehicle occupant and other objects |
| CN110023780B (en) * | 2016-11-25 | 2023-08-15 | Iee国际电子工程股份公司 | Polarized radar system and method for detecting and classifying vehicle occupants and other objects within a vehicle |
| CN106815559A (en) * | 2016-12-21 | 2017-06-09 | 中国科学院深圳先进技术研究院 | A kind of utilization SAR data monitoring oyster arranges method and device, the user equipment in region |
| CN106815559B (en) * | 2016-12-21 | 2020-04-14 | 中国科学院深圳先进技术研究院 | Method and device for monitoring oyster arrangement area by utilizing SAR data, and user equipment |
| CN108037504A (en) * | 2017-11-28 | 2018-05-15 | 中国科学院遥感与数字地球研究所 | A kind of method of the polarimetric synthetic aperture radar rapid polarization correction based on three point targets |
| CN110516552A (en) * | 2019-07-29 | 2019-11-29 | 南京航空航天大学 | A kind of multipolarization radar image classification method and system based on timing curve |
| CN115236655A (en) * | 2022-09-01 | 2022-10-25 | 成都理工大学 | Landslide identification method, system, equipment and medium based on fully-polarized SAR |
| CN115236655B (en) * | 2022-09-01 | 2022-12-20 | 成都理工大学 | Landslide identification method, system, equipment and medium based on fully-polarized SAR |
| US11747498B1 (en) | 2022-09-01 | 2023-09-05 | Chengdu University Of Technology | Method, system, device and medium for landslide identification based on full polarimetric SAR |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Dabboor et al. | Towards sea ice classification using simulated RADARSAT Constellation Mission compact polarimetric SAR imagery | |
| CN101976357A (en) | Method and device for classifying fully polarimetric synthetic aperture radar image | |
| CN103761752B (en) | A kind of processing method and processing device of polarization synthetic aperture radar image | |
| CN109375189A (en) | Polarimetric radar remote sensing images city goal decomposition method based on cross scatter model | |
| CN102999761B (en) | Based on the Classification of Polarimetric SAR Image method that Cloude decomposes and K-wishart distributes | |
| CN103839077B (en) | A kind of Classification of Polarimetric SAR Image method of the low-rank representation based on super-pixel feature | |
| CN102540157A (en) | Ground feature classifying method based on simplified polarization synthetic aperture radar data | |
| Romeiser et al. | A new approach to ocean wave parameter estimates from C-band ScanSAR images | |
| Henry et al. | Proximal radar sensors for precision viticulture | |
| Shirvany et al. | Estimation of the degree of polarization for hybrid/compact and linear dual-pol SAR intensity images: Principles and applications | |
| Tsai et al. | 3D computation of gray level co-occurrence in hyperspectral image cubes | |
| CN105005767A (en) | Microwave remote sensing image based forest type identification method | |
| CN105956622B (en) | Polarization SAR image classification method based on multiple features combining modeling | |
| CN108764326B (en) | City impervious stratum extraction method based on deep belief network | |
| CN104361346A (en) | K-SVD and sparse representation based polarization SAR (synthetic aperture radar) image classification method | |
| CN107239757B (en) | Polarized SAR image target detection method based on deep ladder network | |
| CN104751183A (en) | Polarimetric SAR image classification method based on tensor MPCA | |
| CN112147591A (en) | Polarized radar sea surface ship detection method and device | |
| CN106405545A (en) | Double-time different-mode dual-polarization SAR-type Pauli false color image synthesis method | |
| Perko et al. | The capabilities of TerraSAR-X imagery for retrieval of forest parameters | |
| Vasile et al. | Normalized coherency matrix estimation under the SIRV model. Alpine glacier POLSAR data analysis | |
| Chau et al. | High altitude large-scale plasma waves in the equatorial electrojet at twilight | |
| CN114545410B (en) | Crop lodging monitoring method based on synthetic aperture radar dual-polarized data coherence | |
| CN114926696A (en) | Remote sensing image classification method based on complex matrix and multi-feature collaborative learning | |
| Kuang et al. | Agricultural field boundary delineation using a cascaded deep network model from polarized SAR and multispectral images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110216 |