CN109490886A - A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region - Google Patents

A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region Download PDF

Info

Publication number
CN109490886A
CN109490886A CN201811434914.3A CN201811434914A CN109490886A CN 109490886 A CN109490886 A CN 109490886A CN 201811434914 A CN201811434914 A CN 201811434914A CN 109490886 A CN109490886 A CN 109490886A
Authority
CN
China
Prior art keywords
polarization
oil
matrix
sar
sea
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
Application number
CN201811434914.3A
Other languages
Chinese (zh)
Inventor
李煜
张渊智
袁子峰
郭菁菁
郭华秋
潘洪沅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Technology
Original Assignee
Beijing University of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing University of Technology filed Critical Beijing University of Technology
Priority to CN201811434914.3A priority Critical patent/CN109490886A/en
Publication of CN109490886A publication Critical patent/CN109490886A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications
    • G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G01S13/904—SAR modes
    • G01S13/9076—Polarimetric features in SAR
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications
    • G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications
    • G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G01S13/9021—SAR image post-processing techniques
    • G01S13/9029—SAR image post-processing techniques specially adapted for moving target detection within a single SAR image or within multiple SAR images taken at the same time

Landscapes

  • Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • Geophysics And Detection Of Objects (AREA)

Abstract

The invention discloses a kind of accurate extracting methods in polarimetric synthetic aperture radar remote sensing offshore spilled oil region, comprising: carries out polarization filtering to SAR image, reduces influence of the noise to sea information extraction；Effective polarization characteristic is extracted using multiple means；Classified using storehouse self-encoding encoder (SAE) to sea mineral oil and non-oil diaphragm area；Morphological scale-space is carried out to classification results, obtains mineral oil region mask；Space density Threshold segmentation is carried out to backscattering cross intensity image, obtains high-precision Oil Boundary；Mask is merged with Oil Boundary, obtains accurate offshore spilled oil extracted region result.Method of the invention solves the problems, such as that the differentiation of prior art mineral oil in fluid and biological oil film likelihood object and Oil Boundary are precisely extracted.

Description

A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region
Technical field
The invention belongs to Remote Sensing Image Processing Technology more particularly to it is a kind of utilize SAR remote sensing image carry out sea The method that face oil spilling region is accurately extracted.
Background technique
Marine oil spill is one of chief threat of Coastal Zone Environment, and timely and effectively oil spilling monitoring is to protection marine ecology peace It has great significance entirely.Synthetic aperture radar (SAR) has the advantages of not influenced by sexual intercourse, all weather operations, supervises in remote sensing There is unique advantage in survey.The sea Bragg diffraction because the presence of oil spilling can decay, forms dark space in SAR image, because This traditional SAR offshore spilled oil detection is mainly realized by detection sea blackspot.However, studies have shown that some likelihood objects are (such as Biological oil film) similar blackspot can be also formed in SAR image, it brings challenges for the accurate detection of offshore spilled oil.
Compared with single polarization SAR, full-polarization SAR system can not only obtain the backscatter intensity information of atural object, moreover it is possible to Enough obtain polarization scattering characteristics of the atural object under any different polarized electromagnetic wave irradiations.Studies have shown that can be with using polarization information Effectively distinguish mineral oil and biological oil film likelihood object.Since the back scattering noise in cross polarization channel is relatively low, merge from same Polarize image zooming-out Oil Boundary and from polarization characteristic obtain oil film classification information can be realized more accurate sea overflow Oil extract.
Summary of the invention
For the problems of the prior art, the present invention provides a kind of polarimetric synthetic aperture radar remote sensing offshore spilled oil region essence True extracting method, this method solve the limited problems of slick extracted region precision in the prior art.
The technical solution adopted by the present invention is a kind of polarimetric synthetic aperture radar remote sensing offshore spilled oil region accurately side of extraction Method, steps are as follows for the realization of this method:
S1 pretreatment
To SAR image geocoding, sea mineral oil and non-mineral oil (including sea and biological oil film likelihood object) are extracted Training sample region.Its covariance matrix C (if data do not provide) is obtained by the back scattering matrix S of polarimetric SAR image, And Stokes matrix.
S2 polarization filtering
Polarization filtering is carried out to polarimetric SAR image, selecting filtering algorithm includes the filtering of Boxcar and Lopez based on model Device etc. calculates the polarization covariance and coherence matrix of SAR image after filtering.
S3 feature extraction
A series of polarization characteristics are extracted from the covariance matrix and coherence matrix of polarimetric SAR image, later Selection utilization ReliefF algorithm carries out selection to feature or carries out dimensionality reduction to feature using principal component analysis (PCA), obtains effectively distinguishing The feature set of mineral oil and biological oil film likelihood object.
The classification of S4 oil spilling
Storehouse self-encoding encoder (SAE) is trained based on the polarization characteristic of training sample, distinguishes mineral oil and non-mineral Oil covering water area, classifies to polarimetric SAR image using trained classifier.Oil spilling classification can also select support Other classifiers such as vector machine or random forest.
The fusion of S5 information
1) process of convolution is carried out to the channel VV backscatter signal using Gauss convolution kernel, space density is carried out to result Threshold segmentation；2) Morphological scale-space is carried out to oil spilling classification results, comprising: corrosion, expansion；Union 1) will be sought with result 2), Obtain the intact slick region of high accuracy and boundary.
Compared with prior art, the present invention by the oil spilling classification information based on polarization SAR feature and is based on same polarization The Oil Boundary information of backscattering cross, in conjunction with polarization filtering, storehouse self-encoding encoder, the technologies such as Morphological scale-space are realized high The offshore spilled oil extracted region of precision provides important technology branch for the protection of marine environment and relevant research and operational use It holds.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this Some embodiments of invention without any creative labor, may be used also for those of ordinary skill in the art To obtain other attached drawings according to these attached drawings.
Fig. 1 is the flow diagram of oil spilling extracting method proposed by the present invention；
Fig. 2 is polarimetric SAR image schematic diagram provided in an embodiment of the present invention；
Fig. 3 is SAR image schematic diagram after polarization filtering provided in an embodiment of the present invention；
Fig. 4 is mineral oil classification results schematic diagram provided in an embodiment of the present invention；
Fig. 5 is gaussian kernel function filtering and capacity-threshold segmentation result schematic diagram, and (a) and (b) is respectively that the present invention is implemented The gaussian kernel function filtering and capacity-threshold segmentation result schematic diagram that example provides.
Fig. 6 is corrosion and expansion results schematic diagram, and (a) and (b) is respectively corrosion and expansion provided in an embodiment of the present invention Result schematic diagram.
Fig. 7 is that final oil spilling extracts result schematic diagram.
Specific embodiment
In order to preferably explain the present invention, in order to understand, with reference to the accompanying drawing, by specific embodiment, to this hair It is bright to be described in detail.
In the following description, multiple and different aspects of the invention will be described, however, for common skill in the art For art personnel, the present invention can be implemented just with some or all structures or process of the invention.In order to explain Definition for, specific number, configuration and sequence are elaborated, however, it will be apparent that these specific details the case where Under the present invention also can be implemented.It in other cases, will no longer for some well-known features in order not to obscure the present invention It is described in detail.
Aiming at the problem that existing offshore spilled oil detection technique is difficult to differentiate between mineral oil film and biological oil film, the present invention proposes one The method that kind classifies to offshore spilled oil using polarization SAR feature, while being obtained using the high s/n ratio of same polarization channel image More complete Oil Boundary is taken, the high-precision offshore spilled oil extracted region of two kinds of information realizations is merged.
S1 full-polarization SAR data prediction
Pretreatment (such as needs) is formatted, calibrated etc. to full-polarization SAR data to be treated, obtains complete polarization SAR covariance matrix C.If that obtain is repolarization collision matrix S, it is translated into covariance matrix C by the following method:
Back scattering when wherein each element in repolarization collision matrix S respectively represents certain polar transmitter and receives Coefficient, subscript h indicate the polarization mode of horizontal polarization, and v indicates the polarization mode of vertical polarization, ShvFor horizontal polarization transmitting, hang down Straight polarization reception.
Covariance matrix reflection is that the second-order statistics of radar raster-displaying signal are needed pair by its physical significance The result repeatedly observed is counted, and using spacial average (more views) substitution time domain average in practical application, that is, counts certain pixel Point value in point surrounding neighbors.
Therefore covariance matrix:
Wherein<>represents spacial average, and n takes 3 × 3,5 × 5,7 × 7 neighborhoods, and i indicates the serial number of neighborhood territory pixel.
The polarization coherence matrix T of multiple view picture:
That is:
Wherein * is conjugation.
The Stokes vector that echo is received under any transmitting signal is obtained by Complete polarimetry matrix, that is, collision matrix:
Wherein g0, g1,g2,g3For each component of Stokes vector, EthAnd EtvRespectively a certain polar transmitter state is lauched The back scattering matrix of gentle polarization reception, Re and Im respectively represent real and imaginary parts.Emitting polarization mode t is vertical (V) hair It penetrates, therefore EthAnd EtvJust correspond to SvhAnd Svv。
S2 polarization filtering
To reduce influence of the speckle noise to polarimetric SAR image target decomposition, covariance matrix is filtered, filter Have:
S2.1Boxcar filtering
Using each point in sliding window traversal image, the mean value of image in window is assigned to window center point:
Wherein X is filtered image, and Y is image to be filtered, and N is window size, m, n be respectively pixel to be filtered cross, Ordinate, p and q are horizontal, ordinate offset when traversing pixel in sliding window.
Filtering of the S2.2Lopez based on model
S2.2.1 carries out more view filtering to the diagonal entry in polarization SAR covariance matrix；
S2.2.2 is filtered the off diagonal element in polarization SAR covariance matrix based on polarization model, the model Speckle noise is considered as to the combination of multiplicative noise and additive noise.
The polarization Lee filtering of S2.3 exquisiteness:
S2.3.1 carries out border template matching, selected directions window on SPAN image；
S2.3.2 is filtered covariance matrix using local statistics filter in direction window.
S3 feature extraction
Based on filtered covariance matrix and coherence matrix, or by its calculated Stokes vector, extract as follows The combination of polarization characteristic constitutive characteristic, constitutive characteristic vector inputs classifier together with optical signature.
S3.1Cloude-Pottier characteristics of decomposition
Eigenvalues Decomposition is implemented to more view polarization SAR coherence matrixes, calculates polarization entropy H, average polarization angleAnd normalization Polarize saddle point height NPH.
Wherein λiFor the characteristic value for the coherence matrix that polarizes.Feature H and NPH is related with the complexity of surface scattering mechanism, former The H and NPH on oil covering sea are larger, and the sea H and NPH of the covering of oil-free or biological oil film likelihood object are lower.Average polarization AngleWith area scattering, volume scattering or dihedral angle in back scattering scatter which kind of scattering ingredient occupy an leading position it is related.
S3.2 polarizability, ellipticity, same polarization phase difference standard deviation
It is pure that Stokes matrix calculating by horizontally or vertically emitting polarized wave signal can measure polarization of ele The polarizability of degree:
Wherein, gt1, gt2, gt3, gt0Element respectively in Stokes vector.
Transmitting polarization t is horizontal (H) polarization, calculates polarized signal polarization ellipse using Stokes matrix component and calculating degree Ellipticity:
gt3, gt0Element respectively in Stokes vector, P are polarizability.
Calculate same polarization phase differenceAnd its standard deviation is calculated in n × n window.
The standard deviation that crude oil covers the same polarization phase difference in sea area is larger, and it is covered on sea or biological oil film likelihood object It is smaller in sea area.
S3.3 related coefficient, coherence factor, consistency coefficient
The channel HH and VV related coefficient is calculated by covariance matrix:
Coherence factor is calculated by correlation matrix:
Wherein, T12, T11And T22Element respectively in coherence matrix.
Calculate the consistency coefficient μ of full polarimetric SAR；
The coherence of element in related coefficient and coherence factor reflection polarization SAR signal back scattering or covariance matrix, And consistency coefficient reflects the difference of area scattering and volume scattering component in polarization SAR scattering mechanism.Sea area is covered for crude oil, with Upper three characteristic values are all higher, and cover sea area for oil-free or biological oil film, and three is lower, therefore these three spies can be used Sign is to detect offshore spilled oil.
The classification of S4 oil spilling
By above-mentioned polarization characteristic constitutive characteristic vector, as the input of classifier, using storehouse self-encoding encoder to feature into Row is extracted and is abstracted.The basic unit of storehouse self-encoding encoder is self-encoding encoder, is made of a three-layer neural network.From input layer To the output of hidden layer, input signal is encoded, and from hidden layer to output, the output of hidden layer is decoded.Given input to Amount, autocoder are intended to minimize input x and export the difference between y, i.e. minimum reconstruction error.
Stacking self-encoding encoder (Stacked Auto-Encoder, SAE) is to cascade multiple self-encoding encoders, so that they will Input of the output of one hidden layer of previous autocoder as its subsequent autocoder.Given training sample, often The unsupervised level Training strategy that one layer of self-encoding encoder all passes through greed is trained, and upper layer is the abstract table of related higher-level time Show.Stack autocoder can more effectively establish depth mind by initializing its weight in the region close to optimal value Through network.Finally based on the label of training sample, network parameter is finely adjusted using direction of error propagation algorithm.
Classifier makes full use of polarization SAR feature, effectively carries out to mineral oil and non-mineral oil (biological oil film and sea) It distinguishes, exports the bianry image of mineral oil overlay area.
The fusion of S5 information
S5.1 carries out process of convolution to the channel VV backscatter signal using Gauss convolution kernel, and it is close to carry out space to result Spend Threshold segmentation:
Obtain the more complete oil film overlay area in boundary.
S5.2 Morphological scale-space: corroding oil spilling classification results, removes small false target.Then it is carried out with circle Expansion will belong to the regional connectivity of same oil film, remove the hole due to caused by noise.
S5.3 merges classification results with Morphological scale-space result, obtain high accuracy and boundary it is intact sea oil Diaphragm area.
Ifinal=ISDT.*IDilation
Wherein Ifinal, ISDT, IDilationRespectively final oil film extracts as a result, space density Threshold segmentation result and expansion Processing result.
Embodiment
As shown in Figure 1, to show the offshore spilled oil based on polarimetric synthetic aperture radar remote sensing in the present embodiment accurate by Fig. 1 The flow diagram of extracting method.
Embodiment is using the polarimetric SAR image obtained in the detection experiment of Northern Europe offshore spilled oil, as shown in a in Fig. 2, wherein The blackspot region in the upper left corner and the lower right corner is respectively biological oil film and mineral oil, and intermediate blackspot region is oil emulsion, characteristic Between mineral oil and seawater.The purpose of offshore spilled oil detection is effective extraction mineral oil overlay area, by its same bio oil Film distinguishes.
First according to step 1, to SAR image geocoding, extract sea mineral oil and non-mineral oil (including sea and Biological oil film likelihood object) training sample region (as shown in the b in Fig. 2).Pass through the back scattering matrix S of polarimetric SAR image Obtain its covariance matrix C (if data do not provide) and Stokes matrix.
According to step 2, polarimetric SAR image is filtered based on the method for model using Lopez, as a result such as Fig. 3 institute Show.It can be seen that speckle noise is weakened in filtered image significantly.
According to step 3, polarization entropy H, average polarization angle are extractedWith polarization saddle point height NPH, polarizability, ellipticity, together Polarization phases difference standard deviation, related coefficient, coherence factor, the polarization characteristics such as consistency coefficient.
Offshore spilled oil is realized using the label information of feature vector and sample training storehouse self-encoding encoder according to step 4 Classification, as a result as shown in figure 4, white is oil spilling overlay area.
According to step 5,1) convolution is carried out to the channel VV backscattering cross image using gaussian kernel function, then utilized Capacity-threshold divides (using threshold value for 22dB in embodiment), obtains the fine boundary information of oil film, as shown in Figure 5.2) to excessive Oil classification result is corroded and is expanded (as shown in fig. 6, using radius first for 11 circular configuration in embodiment).It will 1) and 2) Result take union, obtain the fused oil spilling region of information accurately extract as a result, as shown in Figure 7.
Finally, it should be noted that the needs of various parameters designed by this method are adjusted according to the specific interest of practical application It is whole.Above-described embodiments are merely to illustrate the technical scheme, rather than its limitations；Although referring to aforementioned implementation Invention is explained in detail for example, those skilled in the art should understand that: it still can be to aforementioned implementation Technical solution documented by example is modified, or is equivalently replaced to part of or all technical features；And these are repaired Change or replaces, the range for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.

Claims (6)

1. a kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region, it is characterised in that: the reality of this method It is existing that steps are as follows,
S1 pretreatment
To SAR image geocoding, extracts sea mineral oil and non-mineral oil includes the training on sea and biological oil film likelihood object Sample areas；Its covariance matrix C and Stokes matrix are obtained by the back scattering matrix S of polarimetric SAR image；
S2 polarization filtering
Polarization filtering is carried out to polarimetric SAR image, selecting filtering algorithm includes the filter of Boxcar and Lopez based on model, Calculate the polarization covariance and coherence matrix of SAR image after filtering；
S3 feature extraction
A series of polarization characteristics are extracted from the covariance matrix and coherence matrix of polarimetric SAR image, later Selection utilization ReliefF Algorithm select to feature or carries out dimensionality reduction to feature using principal component analysis PCA, obtain effectively distinguishing mineral oil with The feature set of biological oil film likelihood object；
The classification of S4 oil spilling
Storehouse self-encoding encoder SAE is trained based on the polarization characteristic of training sample, distinguishes mineral oil and non-mineral oil covering Water area classifies to polarimetric SAR image using trained classifier；Oil spilling classification can also select support vector machines Or other classifiers such as random forest；
The fusion of S5 information
1) process of convolution is carried out to the channel VV backscatter signal using Gauss convolution kernel, space density threshold value is carried out to result Segmentation；2) Morphological scale-space is carried out to oil spilling classification results, comprising: corrosion, expansion；Union 1) will be sought with result 2), obtained The intact slick region of high accuracy and boundary.
2. a kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region according to claim 1, Be characterized in that: full-polarization SAR data prediction treatment process is as follows,
Full-polarization SAR data to be treated are formatted, calibrate pretreatment, obtain full-polarization SAR covariance matrix C；If that obtain is repolarization collision matrix S, it is translated into covariance matrix C by the following method:
Back scattering system when wherein each element in repolarization collision matrix S respectively represents certain polar transmitter and receives Number, subscript h indicate the polarization mode of horizontal polarization, and v indicates the polarization mode of vertical polarization, ShvEmit for horizontal polarization, vertically Polarization reception；
Time domain average is substituted using spacial average, that is, counts the point value in certain pixel surrounding neighbors；
Therefore covariance matrix:
Wherein<>represents spacial average, and n takes 3 × 3,5 × 5,7 × 7 neighborhoods, and i indicates the serial number of neighborhood territory pixel；
The polarization coherence matrix T of multiple view picture:
That is:
Wherein * is conjugation；
The Stokes vector that echo is received under any transmitting signal is obtained by Complete polarimetry matrix, that is, collision matrix:
Wherein g0, g1,g2,g3For each component of Stokes vector, EthAnd EtvRespectively a certain polar transmitter state is lauched gentle The back scattering matrix of polarization reception, Re and Im respectively represent real and imaginary parts；Transmitting polarization mode t is Vertical Launch, therefore EthAnd EtvJust correspond to SvhAnd Svv。
3. a kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region according to claim 1, Be characterized in that: the treatment process of polarization filtering is as follows,
Influence for reduction speckle noise to polarimetric SAR image target decomposition, is filtered covariance matrix, filter has:
S2.1 Boxcar filtering
Using each point in sliding window traversal image, the mean value of image in window is assigned to window center point:
Wherein X is filtered image, and Y is image to be filtered, and N is window size, and m, n are respectively that the horizontal, vertical of pixel to be filtered sits Mark, p and q are horizontal when being pixel in traversal sliding window, the offset of ordinate；
Filtering of the S2.2 Lopez based on model
S2.2.1 carries out more view filtering to the diagonal entry in polarization SAR covariance matrix；
S2.2.2 is filtered the off diagonal element in polarization SAR covariance matrix based on polarization model, and the model is by spot Spot noise is considered as the combination of multiplicative noise and additive noise；
The polarization Lee filtering of S2.3 exquisiteness:
S2.3.1 carries out border template matching, selected directions window on SPAN image；
S2.3.2 is filtered covariance matrix using local statistics filter in direction window.
4. a kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region according to claim 1, Be characterized in that: the step of feature extraction, is as follows,
Based on filtered covariance matrix and coherence matrix, or by its calculated Stokes vector, extract following polarization The combination of feature constitutive characteristic, constitutive characteristic vector inputs classifier together with optical signature；
S3.1 Cloude-Pottier characteristics of decomposition
Eigenvalues Decomposition is implemented to more view polarization SAR coherence matrixes, calculates polarization entropy H, average polarization angleWith normalization polarization saddle Point height；
α=P1α1+P2α2+P3α3
Wherein λiFor the characteristic value for the coherence matrix that polarizes；Feature H and NPH is related with the complexity of surface scattering mechanism, crude oil covering The H and NPH on sea are big, and the sea H and NPH of the covering of oil-free or biological oil film likelihood object are lower；Average polarization angleWith it is rear Into scattering area scattering, volume scattering or dihedral angle scatter which kind of scattering ingredient occupy an leading position it is related；
S3.2 polarizability, ellipticity, same polarization phase difference standard deviation
Stokes matrix by horizontally or vertically emitting polarized wave signal, which calculates, can measure polarization of ele ingredient purity Polarizability:
Wherein, gt1, gt2, gt3, gt0Element respectively in Stokes vector；
Transmitting polarization t is horizontal (H) polarization, calculates the ellipse of polarized signal polarization ellipse using Stokes matrix component and calculating degree Circle rate:
gt3, gt0Element respectively in Stokes vector, P are polarizability；
Calculate same polarization phase differenceAnd its standard deviation is calculated in n × n window；
The standard deviation that crude oil covers the same polarization phase difference in sea area is big, and it is covered in sea area on sea or biological oil film likelihood object It is small；
S3.3 related coefficient, coherence factor, consistency coefficient
The channel HH and VV related coefficient is calculated by covariance matrix:
Coherence factor is calculated by correlation matrix:
Wherein, T12, T11And T22Element respectively in coherence matrix；
Calculate the consistency coefficient μ of full polarimetric SAR；
The coherence of element in related coefficient and coherence factor reflection polarization SAR signal back scattering or covariance matrix, and one Cause the difference of area scattering and volume scattering component in property coefficient reflection polarization SAR scattering mechanism；For crude oil cover sea area, above three A characteristic value is all high, and covers sea area for oil-free or biological oil film, and three is low, therefore with these three features come to sea Oil spilling is detected.
5. a kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region according to claim 1, Be characterized in that: the implementation process of oil spilling classification is as follows,
By polarization characteristic constitutive characteristic vector, as the input of classifier, feature is extracted using storehouse self-encoding encoder and It is abstract；The basic unit of storehouse self-encoding encoder is self-encoding encoder, is made of a three-layer neural network；From input layer to hidden layer Output, input signal is encoded, and from hidden layer to output, the output of hidden layer is decoded；Given input vector, it is automatic to compile Code device is intended to minimize input x and exports the difference between y, i.e. minimum reconstruction error；
Stacking self-encoding encoder SAE is to cascade multiple self-encoding encoders, so that they hide one of previous autocoder Input of the output of layer as its subsequent autocoder；Given training sample, each layer of self-encoding encoder all pass through the nothing of greed Supervision level Training strategy is trained, and upper layer is the abstract expression of related higher-level time；Stack autocoder is by leaning on The region of nearly optimal value initializes its weight more effectively to establish deep neural network；Finally based on the label of training sample, Network parameter is finely adjusted using direction of error propagation algorithm；
Classifier makes full use of polarization SAR feature, is that biological oil film and sea distinguish to mineral oil and non-mineral oil, exports The bianry image of mineral oil overlay area.
6. a kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region according to claim 1, Be characterized in that: the implementation process of information fusion is as follows,
S5.1 carries out process of convolution to the channel VV backscatter signal using Gauss convolution kernel, carries out space density threshold to result Value segmentation:
Obtain the complete oil film overlay area in boundary；
S5.2 Morphological scale-space: corroding oil spilling classification results, removes small false target；Then it is carried out with circle swollen It is swollen, the regional connectivity of same oil film will be belonged to, remove the hole due to caused by noise；
S5.3 merges classification results with Morphological scale-space result, obtains the intact slick area of high accuracy and boundary Domain；
Ifinal=ISDT.*IDilation
Wherein Ifinal, ISDT, IDilationRespectively final oil film extracts as a result, space density Threshold segmentation result and expansion process As a result.
CN201811434914.3A 2018-11-28 2018-11-28 A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region Pending CN109490886A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811434914.3A CN109490886A (en) 2018-11-28 2018-11-28 A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811434914.3A CN109490886A (en) 2018-11-28 2018-11-28 A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region

Publications (1)

Publication Number Publication Date
CN109490886A true CN109490886A (en) 2019-03-19

Family

ID=65698072

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811434914.3A Pending CN109490886A (en) 2018-11-28 2018-11-28 A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region

Country Status (1)

Country Link
CN (1) CN109490886A (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110221300A (en) * 2019-07-17 2019-09-10 北京无线电测量研究所 The intersection backup switching network of Planar integration Ka wave band for remote sensing SAR radar
CN110346795A (en) * 2019-06-24 2019-10-18 中国地质大学深圳研究院 Marine oil overflow drift dynamic prediction method and system based on carried SAR monitoring
CN110646795A (en) * 2019-09-16 2020-01-03 武汉大学 Ocean oil spill detection method and system of simple polarization SAR
CN110991257A (en) * 2019-11-11 2020-04-10 中国石油大学(华东) Polarization SAR oil spill detection method based on feature fusion and SVM
CN111025291A (en) * 2019-11-06 2020-04-17 中国石油大学(华东) Ocean oil spill detection method based on new characteristics of fully-polarized SAR
CN112014808A (en) * 2020-09-07 2020-12-01 航天科工微电子系统研究院有限公司 Missile-borne double-base SAR anti-interference system and working method thereof
CN112989940A (en) * 2021-02-08 2021-06-18 国家海洋环境监测中心 Raft culture area extraction method based on high-resolution three-satellite SAR image
CN113516657A (en) * 2021-09-14 2021-10-19 中国石油大学(华东) Self-adaptive weight-based fully-polarized SAR image sea surface oil spill detection method
CN114511790A (en) * 2022-02-11 2022-05-17 广东海启星海洋科技有限公司 Method and device for realizing marine oil spill event monitoring based on optical satellite data
CN115236598A (en) * 2022-05-11 2022-10-25 西安电子科技大学 Subspace distance extension target detection method based on polarized radar
CN115546199A (en) * 2022-11-09 2022-12-30 烟台大学 Method for acquiring offshore oil spill area by utilizing self-attention analysis SAR image
CN116299305A (en) * 2023-05-22 2023-06-23 南京信息工程大学 Multi-feature SAR oil spill detection method
CN116912699A (en) * 2023-09-06 2023-10-20 交通运输部天津水运工程科学研究所 Port oil spill diffusion trend prediction method and system based on image processing
CN117095308A (en) * 2023-10-19 2023-11-21 中国石油大学(华东) Genetic algorithm-based full-polarization SAR image sea surface oil spill detection method
CN117911253A (en) * 2024-03-20 2024-04-19 福建省计量科学研究院（福建省眼镜质量检验站） Polarized SAR image processing method and device

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102608604A (en) * 2012-03-14 2012-07-25 清华大学 Method for detecting sea surface oil spillage on basis of common polarimetric characteristic power diagram
EP2848962A1 (en) * 2012-05-04 2015-03-18 Oil Finder Servicos de Sensoriamento Remoto e Modelagem Computacional Ltda. Method for identifying and classifying petroleum exuding areas at the bottom of the sea by reverse modelling
CN105866775A (en) * 2016-03-28 2016-08-17 南京信息工程大学 Method for detecting oil spill at sea based on full-polarized synthetic aperture radar image
CN106872976A (en) * 2016-12-28 2017-06-20 山东科技大学 A kind of marine oil spill information extracting system based on radar data

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102608604A (en) * 2012-03-14 2012-07-25 清华大学 Method for detecting sea surface oil spillage on basis of common polarimetric characteristic power diagram
EP2848962A1 (en) * 2012-05-04 2015-03-18 Oil Finder Servicos de Sensoriamento Remoto e Modelagem Computacional Ltda. Method for identifying and classifying petroleum exuding areas at the bottom of the sea by reverse modelling
CN105866775A (en) * 2016-03-28 2016-08-17 南京信息工程大学 Method for detecting oil spill at sea based on full-polarized synthetic aperture radar image
CN106872976A (en) * 2016-12-28 2017-06-20 山东科技大学 A kind of marine oil spill information extracting system based on radar data

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
YU LI等: ""Marine Oil Spill Detection Based on the Comprehensive Use of Polarimetric SAR Data"", 《SUSTAINABILITY 2018》 *

Cited By (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110346795B (en) * 2019-06-24 2021-06-01 中国地质大学深圳研究院 Marine oil spill drift dynamic prediction method and system based on airborne SAR monitoring
CN110346795A (en) * 2019-06-24 2019-10-18 中国地质大学深圳研究院 Marine oil overflow drift dynamic prediction method and system based on carried SAR monitoring
CN110221300A (en) * 2019-07-17 2019-09-10 北京无线电测量研究所 The intersection backup switching network of Planar integration Ka wave band for remote sensing SAR radar
CN110221300B (en) * 2019-07-17 2021-09-10 北京无线电测量研究所 Cross backup switch network of planar integrated Ka waveband for remote sensing SAR radar
CN110646795A (en) * 2019-09-16 2020-01-03 武汉大学 Ocean oil spill detection method and system of simple polarization SAR
CN111025291A (en) * 2019-11-06 2020-04-17 中国石油大学(华东) Ocean oil spill detection method based on new characteristics of fully-polarized SAR
CN110991257A (en) * 2019-11-11 2020-04-10 中国石油大学(华东) Polarization SAR oil spill detection method based on feature fusion and SVM
CN110991257B (en) * 2019-11-11 2023-04-18 中国石油大学(华东) Polarized SAR oil spill detection method based on feature fusion and SVM
CN112014808A (en) * 2020-09-07 2020-12-01 航天科工微电子系统研究院有限公司 Missile-borne double-base SAR anti-interference system and working method thereof
CN112014808B (en) * 2020-09-07 2024-04-30 航天科工微电子系统研究院有限公司 Missile-borne double-base SAR anti-interference system and working method thereof
CN112989940A (en) * 2021-02-08 2021-06-18 国家海洋环境监测中心 Raft culture area extraction method based on high-resolution three-satellite SAR image
CN112989940B (en) * 2021-02-08 2023-08-01 国家海洋环境监测中心 Raft culture area extraction method based on high-resolution third satellite SAR image
CN113516657A (en) * 2021-09-14 2021-10-19 中国石油大学(华东) Self-adaptive weight-based fully-polarized SAR image sea surface oil spill detection method
CN113516657B (en) * 2021-09-14 2021-12-10 中国石油大学(华东) Self-adaptive weight-based fully-polarized SAR image sea surface oil spill detection method
CN114511790A (en) * 2022-02-11 2022-05-17 广东海启星海洋科技有限公司 Method and device for realizing marine oil spill event monitoring based on optical satellite data
CN115236598A (en) * 2022-05-11 2022-10-25 西安电子科技大学 Subspace distance extension target detection method based on polarized radar
CN115546199B (en) * 2022-11-09 2024-03-15 烟台大学 Method for acquiring offshore oil spill area by analyzing SAR image through self-attention
CN115546199A (en) * 2022-11-09 2022-12-30 烟台大学 Method for acquiring offshore oil spill area by utilizing self-attention analysis SAR image
CN116299305A (en) * 2023-05-22 2023-06-23 南京信息工程大学 Multi-feature SAR oil spill detection method
CN116299305B (en) * 2023-05-22 2023-08-15 南京信息工程大学 Multi-feature SAR oil spill detection method
CN116912699A (en) * 2023-09-06 2023-10-20 交通运输部天津水运工程科学研究所 Port oil spill diffusion trend prediction method and system based on image processing
CN116912699B (en) * 2023-09-06 2023-12-05 交通运输部天津水运工程科学研究所 Port oil spill diffusion trend prediction method and system based on image processing
CN117095308A (en) * 2023-10-19 2023-11-21 中国石油大学(华东) Genetic algorithm-based full-polarization SAR image sea surface oil spill detection method
CN117095308B (en) * 2023-10-19 2024-01-05 中国石油大学(华东) Genetic algorithm-based full-polarization SAR image sea surface oil spill detection method
CN117911253A (en) * 2024-03-20 2024-04-19 福建省计量科学研究院（福建省眼镜质量检验站） Polarized SAR image processing method and device
CN117911253B (en) * 2024-03-20 2024-07-02 福建省计量科学研究院（福建省眼镜质量检验站） Polarized SAR image processing method and device

Similar Documents

Publication Publication Date Title
CN109490886A (en) A kind of accurate extracting method in polarimetric synthetic aperture radar remote sensing offshore spilled oil region
Chen et al. Ship detection from coastal surveillance videos via an ensemble Canny-Gaussian-morphology framework
Zhang Applying data fusion techniques for benthic habitat mapping and monitoring in a coral reef ecosystem
CN104408482B (en) A kind of High Resolution SAR Images object detection method
CN109447073B (en) Infrared weak and small target detection method based on tensor robust principal component analysis
CN109427055B (en) Remote sensing image sea surface ship detection method based on visual attention mechanism and information entropy
CN110378308A (en) The improved harbour SAR image offshore Ship Detection based on Faster R-CNN
CN109117802A (en) Ship Detection towards large scene high score remote sensing image
CN106557740B (en) The recognition methods of oil depot target in a kind of remote sensing images
CN110348356A (en) A kind of successive frame RD images steganalysis method based on depth light stream network
CN105512622B (en) A kind of visible remote sensing image sea land dividing method based on figure segmentation and supervised learning
CN106529472B (en) Object detection method and device based on large scale high-resolution high spectrum image
Corbane et al. Fully automated procedure for ship detection using optical satellite imagery
CN105139034B (en) A kind of ship detecting method that combination spectrum filters out
Li et al. Detection and monitoring of oil spills using moderate/high-resolution remote sensing images
Zhang et al. Evaluation and improvement of generalization performance of SAR ship recognition algorithms
Sun et al. ABLAL: Adaptive background latent space adversarial learning algorithm for hyperspectral target detection
Deng et al. Ship detection from optical satellite image using optical flow and saliency
CN109800637A (en) A kind of remote sensing image small target detecting method
Tello et al. Automatic detection of spots and extraction of frontiers in SAR images by means of the wavelet transform: Application to ship and coastline detection
Baruah et al. Analysis of Traffic Sign Recognition for Automated Transportation Systems Using Neural Networks
Selvi et al. A novel approach for ship recognition using shape and texture
Lei et al. Ship extraction using post CNN from high resolution optical remotely sensed images
Graziano SAR-based ship route estimation by wake components detection and classification
Shihab et al. A comparison of segmentation techniques for target extraction in ground‐penetrating radar data

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20190319

WD01 Invention patent application deemed withdrawn after publication