CN108694714A - Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas - Google Patents
Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas Download PDFInfo
- Publication number
- CN108694714A CN108694714A CN201810458502.7A CN201810458502A CN108694714A CN 108694714 A CN108694714 A CN 108694714A CN 201810458502 A CN201810458502 A CN 201810458502A CN 108694714 A CN108694714 A CN 108694714A
- Authority
- CN
- China
- Prior art keywords
- feature
- indicate
- module
- sar
- adaptive
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10044—Radar image
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses ship seakeeping systems in a kind of adaptive colony intelligence optimization SAR Radar Seas, including SAR radars, database and host computer;SAR radars, database, host computer are sequentially connected, the SAR radars monitor marine site in real time, and in the image data storage to the database for obtaining SAR radars, the host computer includes image pre-processing module, characteristic extracting module, feature selection module, classifier training module, gunz optimization module and result display module.A kind of realization online recognition of present invention offer, ShipTargets identifying system with high accuracy.
Description
Technical field
The present invention relates to radar data process fields, particularly, are related to warship in a kind of adaptive gunz optimization SAR Radar Seas
Ship target identification system.
Background technology
Can highest priority of the ShipTargets as marine monitoring, maritime search and rescue and maritime safety, rapidly and accurately
Ship Target under the complicated sea situation of identification provides support for marine monitoring, safety, search and rescue decision, be greatly related to marine monitoring,
Maritime safety, the quality of maritime search and rescue and effect.The case where judging naval vessel under complicated sea situation, be with its can be accurately identified for
Basis, the emphasis Ship Target of complicated sea situation ShiShimonoseki note is only accurately detected and identifies, it could be to its marine monitoring, sea
Accurately analysis and prediction are made in upper safety, maritime search and rescue etc., to which auxiliary makes correct decision.It is modernIt is marineIt searches and rescues, sea
In the action such as upper monitoring, maritime safety, naval target of the Ship Target as emphasis under complicated sea situation, different types of naval vessel mesh
The maritime affairs behaviors such as progress marine monitoring, maritime safety, maritime search and rescue are marked on to be different.In order to be accurately performed marine prison
The identification of the maritime affairs task such as survey, maritime safety, maritime search and rescue, target is critical issue.Currently based on the Ship Target of SAR image
Detection has had extensive research, and Ship target recognition is identified since the limitation of SAR image resolution ratio just starts to walk,
For some achievements in research having also due to research is not thorough enough, the effect of model is not fine.Therefore, it actively develops based on high score
The Ship target recognition Study of recognition of resolution SAR image has extremely important meaning.
Invention content
In order to overcome the shortcomings of currently based on not high, the of the invention mesh of the ShipTargets recognition accuracy of SAR image
Be provide it is a kind of realize analyze in real time adaptive gunz optimization SAR Radar Seas on ship seakeeping system.
The technical solution adopted by the present invention to solve the technical problems is:A kind of adaptive colony intelligence optimization SAR Radar Seas
Upper ship seakeeping system, including SAR radars, database and host computer, SAR radars, database and host computer phase successively
Even, the SAR radars monitor marine site in real time, and the image data that SAR radars are obtained is stored to the database
In, the host computer includes:
Image pre-processing module is completed to carry out SAR radar image data pretreatments using following process:
1) the SAR image gray level transmitted from database is L, f (x0,y0) it is pixel (x0,y0) at gray value, g
(x0,y0) it is pixel (x0,y0) N × N neighborhoods in pixel average value, wherein x0,y0The abscissa of pixel is indicated respectively
And ordinate;
2) the number of pixels h (m, n) for meeting f=m and g=n by calculating, obtains joint probability density pmn:
pmn=p (m, n)=h (m, n)/M
Wherein, M indicates the total number of image pixel;
3) the mean vector μ of two-dimensional histogram is calculated:
4) probability P that target and background occurs in image is calculated separately0,1With mean vector μ0,1:
Wherein, t, s, subscript 0, subscript 1 indicate f segmentation thresholds, g segmentation thresholds, target area, background area respectively;
5) inter-class variance BCV is calculated:
BCV=P0(μ0-μ)(μ0-μ)′+P1(μ1-μ)(μ1-μ)′;
Wherein, μ indicates mean vector, the transposition of subscript ' representing matrix.
6) optimal threshold is the Two Dimensional Thresholding Xiang Liang [ so that when BCV is maximum value;s0,t0]:
Characteristic extracting module is completed to carry out the extraction of naval vessel characteristic feature using following process:
1) the SAR image slice I (m, n) only comprising a Ship Target transmitted from image pre-processing module, wherein only
Including the binary map of target area is B (m, n), then only include the image T (m, n) of target:
T (m, n)=I (m, n) × B (m, n)
Wherein, × indicate that respective pixel is multiplied;
2) minimum enclosed rectangle of naval vessel body region is acquired according to the major axes orientation of naval vessel individual in B (m, n), then should
The long side length Length of rectangle is the length of naval vessel individual, and the bond length Width of rectangle is the width of naval vessel individual;
3) geometry feature is calculated, including perimeter, area, length-width ratio, shape complexity, target centroid position
It sets and rotary inertia:
PerimeterAreaLength-width ratio R=Length/
Width;Shape complexity C=Length2/4πS;The centroid position of target area
Rotary inertiaIn formula, r represents the distance between target pixel points and barycenter,
4) gray-scale statistical characteristics are calculated, including quality, mean value, coefficient of variation, standard deviation, fractal dimension, add
Weigh packing ratio:
QualityMean valueCoefficient of variationStandard deviationIn formulaIndicate that gray scale logarithm and gray scale logarithm are flat respectively
Fang He;Fractal dimension H=(log10N1-log10N2)/(log10d1-log10d2), the computational methods of this feature are:After segmentation
The binary map B of SAR image slice one a brightest pixel point of the K (taking K=50 here) for remaining target area of structure2(m, n),
It is first d by a size1×d2Window it is continuously slipping in this binary map, write down in window comprising bright spot window it is total
Number scale is N1, with a size it is again then d2×d2Window it is continuously slipping in this binary map, write down and include in the window
The window sum of bright spot is denoted as N2;Weight packing ratio
Feature selection module is completed to select optimal feature subset using following process:
1) inter- object distance is calculatedBetween class distanceAnd class spacing J in classi:
Wherein, i indicates that feature label, ω indicate the label , | of naval vessel classification;|Fi (ω)||2Indicate feature vector Fi (ω)2
Norm,Indicate the population mean of training set sample, NωIndicate that the quantity on ω classes naval vessel, N indicate
Naval vessel sum in training set, E indicate that expectation, subscript W, subscript B are indicated respectively in class, between class.
2) normalization coefficient of variation ρ is calculatedi (ω):
ρi (ω)=E[||Fi (ω)||2 2]-E2[||Fi (ω) 2]/E[||Fi (ω)||2 2]
Wherein, i indicates that feature label, ω indicate the label , | of naval vessel classification;|Fi (ω)||2Indicate feature vector Fi (ω)2
Norm, E[||Fi (ω)||2 2]And E2[||Fi (ω)||2]The mean value of square of feature and square of mean value are indicated respectively.The side of feature
Poor coefficient ρi (ω)It is smaller, show that the stability of the target signature is better;
3) correlation coefficient r is calculatedi,j:
Wherein, i, j indicate feature label , ||Fi||2Indicate feature Fi2 norms,WithF is indicated respectivelyiAnd FjIt is equal
Value, σi,iAnd σj,jF is indicated respectivelyiAnd FjStandard deviation.By the property of related coefficient it is found that 0≤ri,j≤1;If two features
Complete uncorrelated, ri,j=0;If two features are perfectly correlated, ri,j=1;If the correlation between two features is very low, i.e.,
Information redundancy between feature is considerably less, then ri,jIt will be closer to 0;, whereas if the correlation between two features is very
Height, i.e., the information redundancy between feature is very more, then ri,jIt will be closer to 1;
4) optimal feature subset is filtered out by class spacing, normalization coefficient of variation, related coefficient in class obtained above,
Construct optimal input feature value;
Classifier training module is completed to carry out classifier training using following process:
5) N number of SAR radar images x is acquired from feature selection moduleiAs training sample, i=1,2 ..., N;
6) training sample is normalized, obtains normalization sample
7) training sample after normalization is reconstructed, respectively obtains input matrix X and corresponding output matrix Y:
Wherein, D indicates reconstruct dimension, and D is natural number, and D<The value range of N, D are 50-70;
8) X that will be obtained, Y substitute into following linear equation:
WhereinIndicate weight diagonal matrix, K=exp (- ||xi-xj||/θ2) indicate core letter
Number, γ indicate penalty coefficient;Weight factor viIt is calculated by following formula:
WhereinIt is error variance ξiThe estimation of standard deviation, c1, c2For constant;
Solve to obtain function f (x) to be estimated:
Wherein, M is the number of supporting vector, 1v=[1,…,1]′,Subscript ' representing matrix
Transposition,It is Lagrange multiplier, b*It is amount of bias, K=exp (- ||xi-xj||/θ2) it is kernel function, wherein i=1 ..., M, j
=1 ..., M,With exp (- ||x-xi||/θ2) be support vector machines kernel function, θ is nuclear parameter, x
Indicate input variable, γ is penalty coefficient;
Adaptive gunz optimizing module, to the nuclear parameter θ of sorter model and to be punished using APSO algorithm
Penalty factor γ is optimized, and is completed using following process:
1) primary group velocity and position are randomly generated;
2) population diversity index D (t) is calculated:
Wherein, Gbest (t) is the globally optimal solution that entire population was reached in the t times iteration, Fit (Gbest
(t)) the corresponding fitness values of Gbest (t) are indicated, m is population scale, zi(t) it is i-th of particle in the t times iteration
Position, Fit (zi(t)) z is indicatedi(t) corresponding fitness value;
3) renewal learning rate parameter Ψ (t):
4) speed of more new particle and position generate new group;
I=1,2 ..., p;K=1,2
zik(t+1)=zik(t)+uik(t+1)
Wherein, α1It is individual acceleration parameter, α2It is global acceleration parameter,WithIt is the random number between 0-1, t is
Iterations, p are population scale;uik(t+1) be i-th of particle speed of k-th of component in the t+1 times iteration
Degree,
uik(t) it is k-th of component of i-th of particle in the speed of the t times iteration, zik(t+1) be i-th of particle kth
A point
Amount is in the position of the t+1 times iteration, zik(t) be i-th of particle k-th of component in the position of the t times iteration,
Lbestik
It is the optimal solution that k-th of component of i-th of particle reached, k=1,2 correspond respectively to nuclear parameter θ and punishment system
Number γ;
5) judge whether to meet algorithm end condition, if meeting, export the optimal solution of global optimum's particle and its representative, and
Terminate iteration;Otherwise 2) continuation iteration is returned;
Wherein, population scale is 50-100, and individual acceleration parameter is 0.5, and global acceleration parameter is 0.35, individual
The root-mean-square error of ladder-like selection model is adapted to, end condition is that continuous five iteration globally optimal solutions are constant.
Result display module, to be identified the display of result, i.e., by the type for inputting naval vessel in SAR image include
In screen.
The present invention technical concept be:The present invention is directed to SAR radars round-the-clock, all weather operations and the spy penetrated by force
Property, image preprocessing is carried out to the ocean imagery that SAR radars monitor, then carries out the extraction of feature and the selection of feature,
Ocean ship seakeeping model is established finally by the training process of grader, to realize SAR radar ShipTargets
Identification.
Beneficial effects of the present invention are mainly manifested in:1, ShipTargets can be identified in real time;2, recognition methods used
Only need less training sample;3, intelligent, small by interference from human factor.
Description of the drawings
Fig. 1 is the overall structure figure of system proposed by the invention;
Fig. 2 is the functional block diagram of host computer proposed by the invention.
Specific implementation mode
The present invention is illustrated below according to attached drawing.Above-described embodiment is used for illustrating the present invention, rather than to this hair
It is bright to be limited, in the protection domain of spirit and claims of the present invention, to any modifications and changes for making of the present invention,
Both fall within protection scope of the present invention.
Embodiment
Referring to Fig.1, Fig. 2, ship seakeeping system in a kind of adaptive gunz optimization SAR Radar Seas, including SAR radars
1, database 2 and host computer 3, SAR radars 1, database 2 and host computer 3 are sequentially connected, and the SAR radars 1 are to monitored marine site
It is irradiated, and by SAR radar images storage to the database 2, the host computer 3 includes:
Image pre-processing module 4 is completed to carry out SAR radar image data pretreatments using following process:
1) the SAR image gray level transmitted from database is L, f (x0,y0) it is pixel (x0,y0) at gray value, g
(x0,y0) it is pixel (x0,y0) N × N neighborhoods in pixel average value, wherein x0,y0The abscissa of pixel is indicated respectively
And ordinate;
2) the number of pixels h (m, n) for meeting f=m and g=n by calculating, obtains joint probability density pmn:
pmn=p (m, n)=h (m, n)/M
Wherein, M indicates the total number of image pixel;
3) the mean vector μ of two-dimensional histogram is calculated:
4) probability P that target and background occurs in image is calculated separately0,1With mean vector μ0,1:
Wherein, t, s, subscript 0, subscript 1 indicate f segmentation thresholds, g segmentation thresholds, target area, background area respectively;
5) inter-class variance BCV is calculated:
BCV=P0(μ0-μ)(μ0-μ)′+P1(μ1-μ)(μ1-μ)′;
Wherein, μ indicates mean vector, the transposition of subscript ' representing matrix.
6) optimal threshold is the Two Dimensional Thresholding Xiang Liang [ so that when BCV is maximum value;s0,t0]:
Characteristic extracting module 5 is completed to carry out the extraction of naval vessel characteristic feature using following process:
1) the SAR image slice I (m, n) only comprising a Ship Target transmitted from image pre-processing module, wherein only
Including the binary map of target area is B (m, n), then only include the image T (m, n) of target:
T (m, n)=I (m, n) × B (m, n)
Wherein, × indicate that respective pixel is multiplied;
2) minimum enclosed rectangle of naval vessel body region is acquired according to the major axes orientation of naval vessel individual in B (m, n), then should
The long side length Length of rectangle is the length of naval vessel individual, and the bond length Width of rectangle is the width of naval vessel individual;
3) geometry feature is calculated, including perimeter, area, length-width ratio, shape complexity, target centroid position
It sets and rotary inertia:
PerimeterAreaLength-width ratio R=Length/
Width;Shape complexity C=Length2/4πS;The centroid position of target area
Rotary inertiaIn formula, r represents the distance between target pixel points and barycenter,
4) gray-scale statistical characteristics are calculated, including quality, mean value, coefficient of variation, standard deviation, fractal dimension, add
Weigh packing ratio:
QualityMean valueCoefficient of variationStandard deviationIn formulaIndicate that gray scale logarithm and gray scale logarithm are flat respectively
Fang He;Fractal dimension H=(log10N1-log10N2)/(log10d1-log10d2), the computational methods of this feature are:After segmentation
The binary map B of SAR image slice one a brightest pixel point of the K (taking K=50 here) for remaining target area of structure2(m, n),
It is first d by a size1×d2Window it is continuously slipping in this binary map, write down in window comprising bright spot window it is total
Number scale is N1, with a size it is again then d2×d2Window it is continuously slipping in this binary map, write down and include in the window
The window sum of bright spot is denoted as N2;Weight packing ratio
Feature selection module 6 is completed to select optimal feature subset using following process:
1) inter- object distance is calculatedBetween class distanceAnd class spacing J in classi:
Wherein, i indicates that feature label, ω indicate the label , | of naval vessel classification;|Fi (ω)||2Indicate feature vector Fi (ω)2
Norm,Indicate the population mean of training set sample, NωIndicate that the quantity on ω classes naval vessel, N indicate
Naval vessel sum in training set, E indicate that expectation, subscript W, subscript B are indicated respectively in class, between class.
2) normalization coefficient of variation ρ is calculatedi (ω):
ρi (ω)=E[||Fi (ω)||2 2]-E2[||Fi (ω)||2]/E[||Fi (ω)||2 2]
Wherein, i indicates that feature label, ω indicate the label , | of naval vessel classification;|Fi (ω)||2Indicate feature vector Fi (ω)2
Norm, E[||Fi (ω)||2 2]And E2[||Fi (ω)||2]The mean value of square of feature and square of mean value are indicated respectively.The side of feature
Poor coefficient ρi (ω)It is smaller, show that the stability of the target signature is better;
3) correlation coefficient r is calculatedi,j:
Wherein, i, j indicate feature label , ||Fi||2Indicate feature Fi2 norms,WithF is indicated respectivelyiAnd FjIt is equal
Value, σi,iAnd σj,jF is indicated respectivelyiAnd FjStandard deviation.By the property of related coefficient it is found that 0≤ri,j≤1;If two features
Complete uncorrelated, ri,j=0;If two features are perfectly correlated, ri,j=1;If the correlation between two features is very low, i.e.,
Information redundancy between feature is considerably less, then ri,jIt will be closer to 0;, whereas if the correlation between two features is very
Height, i.e., the information redundancy between feature is very more, then ri,jIt will be closer to 1;
4) optimal feature subset is filtered out by class spacing, normalization coefficient of variation, related coefficient in class obtained above,
Construct optimal input feature value;
Classifier training module 7 is completed to carry out classifier training using following process:
1) N number of SAR radar images x is acquired from feature selection moduleiAs training sample, i=1,2 ..., N;
2) training sample is normalized, obtains normalization sample
3) training sample after normalization is reconstructed, respectively obtains input matrix X and corresponding output matrix Y:
Wherein, D indicates reconstruct dimension, and D is natural number, and D<The value range of N, D are 50-70;
4) X that will be obtained, Y substitute into following linear equation:
WhereinIndicate weight diagonal matrix, K=exp (- ||xi-xj||/θ2) indicate core letter
Number, γ indicate penalty coefficient;Weight factor viIt is calculated by following formula:
WhereinIt is error variance ξiThe estimation of standard deviation, c1, c2For constant;
Solve to obtain function f (x) to be estimated:
Wherein, M is the number of supporting vector, 1v=[1,…,1]′,Subscript ' representing matrix
Transposition,It is Lagrange multiplier, b* is amount of bias, K=exp (- ||xi-xj||/θ2) it is kernel function, wherein i=1 ..., M, j
=1 ..., M,With exp (- ||x-xi||/θ2) be support vector machines kernel function, θ is nuclear parameter,
X indicates input variable, and γ is penalty coefficient;
Adaptive gunz optimizing module 9, to the nuclear parameter θ of sorter model and to be punished using APSO algorithm
Penalty factor γ is optimized, and is completed using following process:
1) primary group velocity and position are randomly generated;
2) population diversity index D (t) is calculated:
Wherein, Gbest (t) is the globally optimal solution that entire population was reached in the t times iteration, Fit (Gbest
(t)) the corresponding fitness values of Gbest (t) are indicated, m is population scale, zi(t) it is i-th of particle in the t times iteration
Position, Fit (zi(t)) z is indicatedi(t) corresponding fitness value;
3) renewal learning rate parameter Ψ (t):
4) speed of more new particle and position generate new group;
I=1,2 ..., p;K=1,2
zik(t+1)=zik(t)+uik(t+1)
Wherein, α1It is individual acceleration parameter, α2It is global acceleration parameter,WithIt is the random number between 0-1, t is
Iterations, p are population scale;uik(t+1) be i-th of particle speed of k-th of component in the t+1 times iteration
Degree,
uik(t) it is k-th of component of i-th of particle in the speed of the t times iteration, zik(t+1) be i-th of particle kth
A point
Amount is in the position of the t+1 times iteration, zik(t) be i-th of particle k-th of component in the position of the t times iteration,
Lbestik
It is the optimal solution that k-th of component of i-th of particle reached, k=1,2 correspond respectively to nuclear parameter θ and punishment system
Number γ;
5) judge whether to meet algorithm end condition, if meeting, export the optimal solution of global optimum's particle and its representative, and
Terminate iteration;Otherwise 2) continuation iteration is returned;
Wherein, population scale is 50-100, and individual acceleration parameter is 0.5, and global acceleration parameter is 0.35, individual
The root-mean-square error of ladder-like selection model is adapted to, end condition is that continuous five iteration globally optimal solutions are constant.
Result display module 8 shows the type for inputting naval vessel in SAR image the display of result is identified
In host computer.
The hardware components of the host computer 3 include:I/O elements, the transmission of acquisition and information for data;Data store
Device, the required data sample of storage running and operating parameter etc.;The software journey of function module is realized in program storage, storage
Sequence;Arithmetic unit executes program, realizes specified function;Display module shows the parameter and recognition result of setting.
Above-described embodiment is used for illustrating the present invention, rather than limits the invention, the present invention spirit and
In scope of the claims, to any modifications and changes that the present invention makes, protection scope of the present invention is both fallen within.
Claims (5)
1. ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas, it is characterised in that:Including SAR radars,
Database and host computer, SAR radars, database and host computer are sequentially connected, and the SAR radars supervise marine site in real time
In the image data storage to the database surveyed, and SAR radars are obtained, the host computer includes image preprocessing mould
Block, characteristic extracting module, classifier training module, classifier training module, adaptive gunz optimizing module and result show mould
Block, described image preprocessing module, characteristic extracting module, classifier training module, classifier training module and result show mould
Block is sequentially connected, and classifier training module is connected with adaptive gunz optimizing module.
2. ship seakeeping system in adaptive colony intelligence optimization SAR Radar Seas, feature exist according to claim 1
In:Described image preprocessing module is completed to carry out SAR radar image data pretreatments using following process:
1) the SAR image gray level transmitted from database is L, f (x0,y0) it is pixel (x0,y0) at gray value, g (x0,
y0) it is pixel (x0,y0) N × N neighborhoods in pixel average value, wherein x0,y0The abscissa of pixel is indicated respectively and is indulged
Coordinate;
2) the number of pixels h (m, n) for meeting f=m and g=n by calculating, obtains joint probability density pmn:
pmn=p (m, n)=h (m, n)/M
Wherein, M indicates the total number of image pixel;
3) the mean vector μ of two-dimensional histogram is calculated:
4) probability P that target and background occurs in image is calculated separately0,1With mean vector μ0,1:
Wherein, t, s, subscript 0, subscript 1 indicate f segmentation thresholds, g segmentation thresholds, target area, background area respectively;
5) inter-class variance BCV is calculated:
BCV=P0(μ0-μ)(μ0-μ)′+P1(μ1-μ)(μ1-μ)′;
Wherein, μ indicates mean vector, the transposition of subscript ' representing matrix.
6) optimal threshold is the Two Dimensional Thresholding Xiang Liang [ so that when BCV is maximum value;s0,t0]:
3. ship seakeeping system in adaptive colony intelligence optimization SAR Radar Seas, feature exist according to claim 1
In:The characteristic extracting module is completed to carry out the extraction of naval vessel characteristic feature using following process:
1) the SAR image slice I (m, n) only comprising a Ship Target transmitted from image pre-processing module, wherein including only
The binary map of target area is B (m, n), then only includes the image T (m, n) of target:
T (m, n)=I (m, n) × B (m, n)
Wherein, × indicate that respective pixel is multiplied;
2) minimum enclosed rectangle of naval vessel body region is acquired according to the major axes orientation of naval vessel individual in B (m, n), then the rectangle
Long side length Length be naval vessel individual length, the bond length Width of rectangle is the width of naval vessel individual;
3) geometry feature is calculated, including perimeter, area, length-width ratio, shape complexity, target centroid position with
And rotary inertia:
Perimeter area length-width ratio R=Length/
Width;Shape complexity C=Length2/4πS;The centroid position of target area
In rotational inertia type, r represents the distance between target pixel points and barycenter,
4) gray-scale statistical characteristics are calculated, are filled out including quality, mean value, coefficient of variation, standard deviation, fractal dimension, weighting
Fill ratio:
QualityMean valueCoefficient of variation
Standard deviationIn formula
Gray scale logarithm and gray scale logarithm quadratic sum are indicated respectively;Fractal dimension H=(log10N1-log10N2)/(log10d1-log10d2),
The computational methods of this feature are:With one K for remaining target area of SAR image slice structure after segmentation (K=is taken here
50) the binary map B of a brightest pixel point2One size is first d by (m, n)1×d2Window it is continuous in this binary map
Sliding writes down the window sum comprising bright spot in window and is denoted as N1, with a size it is again then d2×d2Window this two
It is continuously slipping in value figure, it writes down the window sum comprising bright spot in the window and is denoted as N2;Weight packing ratio
4. ship seakeeping system in adaptive colony intelligence optimization SAR Radar Seas, feature exist according to claim 1
In:The feature selection module is completed to select optimal feature subset using following process:
1) inter- object distance is calculatedBetween class distanceAnd class spacing J in classi:
Wherein, i indicates that feature label, ω indicate the label , | of naval vessel classification;|Fi (ω)||2Indicate feature vector Fi (ω)2 norms,Indicate the population mean of training set sample, NωIndicate that the quantity on ω classes naval vessel, N indicate training set
Middle naval vessel sum, E indicate that expectation, subscript W, subscript B are indicated respectively in class, between class.
2) normalization coefficient of variation ρ is calculatedi (ω):
ρi (ω)=E[||Fi (ω)||2 2]-E2[||Fi (ω)||2]/E[||Fi (ω)||2 2]
Wherein, i indicates that feature label, ω indicate the label , | of naval vessel classification;|Fi (ω)||2Indicate feature vector Fi (ω)2 norms, E
[||Fi (ω)||2 2]And E2[||Fi (ω)||2]The mean value of square of feature and square of mean value are indicated respectively.The coefficient of variation of feature
ρi (ω)It is smaller, show that the stability of the target signature is better;
3) correlation coefficient r is calculatedi,j:
Wherein, i, j indicate feature label , ||Fi||2Indicate feature Fi2 norms,WithF is indicated respectivelyiAnd FjMean value,
σi,iAnd σj,jF is indicated respectivelyiAnd FjStandard deviation.By the property of related coefficient it is found that 0≤ri,j≤1;If two features are complete
It is uncorrelated, ri,j=0;If two features are perfectly correlated, ri,j=1;If the correlation between two features is very low, i.e. feature
Between information redundancy it is considerably less, then ri,jIt will be closer to 0;, whereas if the correlation between two features is very high, i.e.,
Information redundancy between feature is very more, then ri,jIt will be closer to 1;
4) optimal feature subset is filtered out by class spacing, normalization coefficient of variation, related coefficient in class obtained above, constructed
Optimal input feature value.
5. ship seakeeping system in adaptive colony intelligence optimization SAR Radar Seas, feature exist according to claim 1
In:The classifier training module is completed to carry out classifier training using following process:
1) N number of SAR radar images x is acquired from feature selection moduleiAs training sample, i=1,2 ..., N;
2) training sample is normalized, obtains normalization sample
3) training sample after normalization is reconstructed, respectively obtains input matrix X and corresponding output matrix Y:
Wherein, D indicates reconstruct dimension, and D is natural number, and D<The value range of N, D are 50-70;
4) X that will be obtained, Y substitute into following linear equation:
WhereinIndicate weight diagonal matrix, K=exp (- ||xi-xj||/θ2) indicate kernel function, γ
Indicate penalty coefficient;Weight factor viIt is calculated by following formula:
WhereinIt is error variance ξiThe estimation of standard deviation, c1, c2For constant;
Solve to obtain function f (x) to be estimated:
Wherein, M is the number of supporting vector, 1v=[1,…,1]′,The transposition of subscript ' representing matrix,It is Lagrange multiplier, b*It is amount of bias, K=exp (- ||xi-xj||/θ2) it is kernel function, wherein i=1 ..., M, j=
1,…,M,With exp (- ||x-xi||/θ2) be support vector machines kernel function, θ is nuclear parameter, x tables
Show that input variable, γ are penalty coefficients;
The adaptive gunz optimizing module, to the nuclear parameter θ of sorter model and to be punished using APSO algorithm
Penalty factor γ is optimized, and is completed using following process:
1) primary group velocity and position are randomly generated;
2) population diversity index D (t) is calculated:
Wherein, Gbest (t) is the globally optimal solution that entire population was reached in the t times iteration, Fit (Gbest (t)) table
Show that the corresponding fitness values of Gbest (t), m are population scale, zi(t) it is position of i-th of particle in the t times iteration,
Fit(zi(t)) z is indicatedi(t) corresponding fitness value;
3) renewal learning rate parameter Ψ (t):
4) speed of more new particle and position generate new group;
zik(t+1)=zik(t)+uik(t+1)
Wherein, α1It is individual acceleration parameter, α2It is global acceleration parameter,WithIt is the random number between 0-1, t is iteration
Number, p are population scale;uik(t+1) it is k-th of component of i-th of particle in the speed of the t+1 times iteration, uik(t) it is
K-th of component of i-th of particle is in the speed of the t times iteration, zik(t+1) it is k-th of component of i-th of particle at the t+1 times
The position of iteration, zik(t) it is k-th of component of i-th of particle in the position of the t times iteration, LbestikIt is i-th of particle
The optimal solution that k-th of component reached, k=1,2 correspond respectively to nuclear parameter θ and penalty coefficient γ;
5) judge whether to meet algorithm end condition, if meeting, export the optimal solution of global optimum's particle and its representative, and terminate
Iteration;Otherwise 2) continuation iteration is returned;
Wherein, population scale is 50-100, and individual acceleration parameter is 0.5, and global acceleration parameter is 0.35, ideal adaptation
The root-mean-square error of ladder-like selection model, end condition are that continuous five iteration globally optimal solutions are constant.
The result display module, to be identified the display of result, i.e., by the type for inputting naval vessel in SAR image include
In screen.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810458502.7A CN108694714A (en) | 2018-05-14 | 2018-05-14 | Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810458502.7A CN108694714A (en) | 2018-05-14 | 2018-05-14 | Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108694714A true CN108694714A (en) | 2018-10-23 |
Family
ID=63847453
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810458502.7A Pending CN108694714A (en) | 2018-05-14 | 2018-05-14 | Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108694714A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102183749A (en) * | 2011-03-03 | 2011-09-14 | 浙江大学 | Sea target detecting system of adaptive radar and method thereof |
CN107274401A (en) * | 2017-06-22 | 2017-10-20 | 中国人民解放军海军航空工程学院 | A kind of High Resolution SAR Images Ship Detection of view-based access control model attention mechanism |
CN107609601A (en) * | 2017-09-28 | 2018-01-19 | 北京计算机技术及应用研究所 | A kind of ship seakeeping method based on multilayer convolutional neural networks |
CN107862271A (en) * | 2017-10-31 | 2018-03-30 | 西安电子科技大学 | A kind of detection method of Ship Target |
-
2018
- 2018-05-14 CN CN201810458502.7A patent/CN108694714A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102183749A (en) * | 2011-03-03 | 2011-09-14 | 浙江大学 | Sea target detecting system of adaptive radar and method thereof |
CN107274401A (en) * | 2017-06-22 | 2017-10-20 | 中国人民解放军海军航空工程学院 | A kind of High Resolution SAR Images Ship Detection of view-based access control model attention mechanism |
CN107609601A (en) * | 2017-09-28 | 2018-01-19 | 北京计算机技术及应用研究所 | A kind of ship seakeeping method based on multilayer convolutional neural networks |
CN107862271A (en) * | 2017-10-31 | 2018-03-30 | 西安电子科技大学 | A kind of detection method of Ship Target |
Non-Patent Citations (1)
Title |
---|
陈文婷: "SAR图像舰船目标特征提取与分类识别方法研究", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Torralba et al. | Statistics of natural image categories | |
CN108490415A (en) | Ship seakeeping system in a kind of intelligence SAR Radar Seas | |
Ji et al. | Seabed sediment classification using multibeam backscatter data based on the selecting optimal random forest model | |
CN110135267A (en) | A kind of subtle object detection method of large scene SAR image | |
CN110472627A (en) | One kind SAR image recognition methods end to end, device and storage medium | |
CN108596156A (en) | A kind of intelligence SAR radar airbound target identifying systems | |
CN105184298A (en) | Image classification method through fast and locality-constrained low-rank coding process | |
CN110287354A (en) | A kind of high score remote sensing images semantic understanding method based on multi-modal neural network | |
CN108629382A (en) | Ship seakeeping system in a kind of gunz optimization SAR Radar Seas | |
Adiwinata et al. | Fish species recognition with faster r-cnn inception-v2 using qut fish dataset | |
CN108596155A (en) | A kind of SAR radars airbound target identifying system | |
CN113436125B (en) | Side-scan sonar simulation image generation method, device and equipment based on style migration | |
CN107766792A (en) | A kind of remote sensing images ship seakeeping method | |
CN109801208A (en) | SAR image change detection based on the optimization of more GPU tasks | |
Wangli et al. | Foxtail Millet ear detection approach based on YOLOv4 and adaptive anchor box adjustment | |
CN108647653A (en) | A kind of intelligence SAR radars land tank target identifying system | |
CN108875578A (en) | A kind of adaptive colony intelligence optimization SAR radar airbound target identifying system | |
Cannaday II et al. | Broad area search and detection of surface-to-air missile sites using spatial fusion of component object detections from deep neural networks | |
CN115205693B (en) | Method for extracting enteromorpha in multi-feature integrated learning dual-polarization SAR image | |
CN108694714A (en) | Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas | |
Wang et al. | CNN based renormalization method for ship detection in VHR remote sensing images | |
CN114663743A (en) | Ship target re-identification method, terminal equipment and storage medium | |
Wang et al. | Coastline detection in SAR images using multi-feature and SVM | |
CN108615029A (en) | A kind of adaptive colony intelligence optimization SAR radars land tank target identifying system | |
CN108932469A (en) | Ship seakeeping system in a kind of SAR Radar Sea |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181023 |