CN107122764B - ShipTargets detection method based on KpN model - Google Patents

ShipTargets detection method based on KpN model Download PDF

Info

Publication number
CN107122764B
CN107122764B CN201710364490.7A CN201710364490A CN107122764B CN 107122764 B CN107122764 B CN 107122764B CN 201710364490 A CN201710364490 A CN 201710364490A CN 107122764 B CN107122764 B CN 107122764B
Authority
CN
China
Prior art keywords
kpn
model
parameter
shiptargets
detection
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.)
Active
Application number
CN201710364490.7A
Other languages
Chinese (zh)
Other versions
CN107122764A (en
Inventor
高贵
欧阳克威
周石琳
高昇
李高升
何鹃
牟卫华
刘伟
程江华
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.)
National University of Defense Technology
Original Assignee
National University of Defense 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 National University of Defense Technology filed Critical National University of Defense Technology
Priority to CN201710364490.7A priority Critical patent/CN107122764B/en
Publication of CN107122764A publication Critical patent/CN107122764A/en
Application granted granted Critical
Publication of CN107122764B publication Critical patent/CN107122764B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/13—Satellite images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/10—Pre-processing; Data cleansing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/07—Target detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Astronomy & Astrophysics (AREA)
  • Remote Sensing (AREA)
  • Multimedia (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The present invention provides a kind of ShipTargets detection method based on KpN model.Technical solution is: carrying out statistical modeling using KpN distribution to obtained SAR image and is estimated using the logarithm cumulant of SAR image the parameter of KpN model, CFAR detection threshold value is calculated according to the estimated value of KpN model parameter, the detection for ShipTargets is realized using CFAR detection.The present invention can be realized for form parameter in KpN model, scale parameter and the more accurate estimation of noise power, enhance the detection performance to ShipTargets, while the present invention does not need that additional parameter or condition is arranged, succinct easy.

Description

ShipTargets detection method based on KpN model
Technical field
The invention belongs to SAR (synthetic aperture radar, synthetic aperture radar) technical fields, are related to one kind ShipTargets detection method based on KpN model.
Background technique
ShipTargets detection is a key areas of SAR application.Military information monitoring, illegal immigrant supervision and The fields such as a wide range of sea traffic supervision have a wide range of applications.CFAR (constant false alarm rate, constant false alarm Rate) detection be most common ShipTargets detection method.The core of CFAR detection is that sea clutter models, often at present Sea clutter model is mainly K distributed model and G0Distributed model, but both models do not consider channel noise Influence.KpN (K plus Noise, K plus noise) model joined the influence of channel noise on the basis of K distributed model, It can be fitted sea clutter (bibliography: K.D.Ward and R.J.A.Tough, " Radar detection more accurately performance in sea clutter and discrete spikes,"Radar,2002,pp.253-257).But it is existing Have that method is high not enough for the estimated accuracy of KpN model parameter, this limits KpN model naval vessel at sea to a certain extent Effect in target detection.
Summary of the invention
The present invention provides a kind of ShipTargets detection method based on KpN model.This method uses SAR image KpN distribution is carried out statistical modeling and is realized using logarithm cumulant to the parameter Estimation of KpN model, is realized to marine vessel mesh Target detection.
The technical scheme is that
Statistical modeling is carried out and using the logarithm cumulant of SAR image to KpN using KpN distribution to obtained SAR image The parameter of model estimated, according to the estimated value of KpN model parameter calculate CFAR (constant false alarm rate, Constant false alarm rate) detection threshold value, the detection for ShipTargets is realized using CFAR detection.Wherein, it is solved using following formula The estimated value of form parameter v into KpN modelThe estimated value of scale parameter bAnd noise power pnEstimated value
Wherein Ψ () is psi function, and Ψ () is polygamma function, and N is equivalent number, and parameter A's, B, C, D is specific Expression formula is shown below:
Also, the specific formula for calculation of CFAR detection threshold value T is as follows:
WhereinIt is the probability density function of KpN model；PfaFalse alarm rate is indicated, generally according to actual needs It is manually set.
The beneficial effects of the present invention are:
1. carrying out KpN parameter Estimation compared to existing method using logarithm cumulant, can be realized in KpN model Form parameter, scale parameter and the more accurate estimation of noise power, enhance the detection performance to ShipTargets.
2. using logarithm cumulant progress KpN method for parameter estimation not needing that additional ginseng is arranged using proposed by the present invention Several or condition, it is succinct easy.
Detailed description of the invention
Fig. 1 is flow chart of the present invention；
Fig. 2 is experimental data of the invention；
Fig. 3 is experimental result picture of the present invention；
Fig. 4, Fig. 5, Fig. 6 are the results for carrying out theoretical validation.
Specific embodiment
Fig. 1 is flow chart of the present invention, and specific implementation step is as follows:
Statistical modeling is carried out using KpN distribution to obtained SAR image and utilizes logarithm cumulant to the parameter of KpN model Estimated: being first considered that SAR image meets KpN model profile, then estimate the parameter of KpN model, may include following two Step:
The first step calculates image log cumulant, calculation method is as shown in formula one according to original SAR image:
WhereinIndicate single order image log cumulant,Indicate second-order image logarithm cumulant,Indicate three rank images pair Number cumulant, M indicate the pixel total number in image, xiFor the gray value of ith pixel point in image, i ∈ [1, M].
Second step obtains the estimated value of form parameter v in KpN model by carrying out numerical solution to formula twoScale ginseng The estimated value of number bAnd noise power pnEstimated valueIts expression is as follows.
Wherein Ψ () is psi function, and Ψ () is polygamma function, and N is equivalent number, and parameter A's, B, C, D is specific Expression formula is as shown in formula three:
CFAR detection threshold value is calculated according to the estimated value of KpN model parameter, is realized using CFAR detection for marine vessel The detection of target, that is, realize it is following step 3:
Third step utilizes the estimated value of form parameter obtained in second stepThe estimated value of scale parameterAnd noise The estimated value of powerCalculate CFAR detection threshold value T.The specific formula for calculation of CFAR detection threshold value T is as follows:
WhereinIt is the probability density function of KpN model；PfaFalse alarm rate is indicated, generally according to actual needs It is manually set.
Original SAR image is detected, when the gray value of detection pixel point is more than or equal to T, is determined as Ship Target Pixel.Otherwise, it is determined that being background pixel, the detection for ShipTargets is realized.
Experimental data of the invention is original SAR image.Fig. 2 is original SAR image, and wherein abscissa indicates orientation To ordinate indicates distance to the white pixel point in image is the Ship Target for needing to detect.Fig. 3 be using the present invention into Row ShipTargets testing result figure, the abscissa of Fig. 3 indicate orientation, and ordinate indicates distance to white rectangle in figure Frame indicates the Ship Target detected.It is all preferable that comparison diagram 2 and Fig. 3 can be seen that 13 all ShipTargets It detects, and there is no false-alarm, this demonstrates the validity of the method for the present invention.
Further to verify logarithm cumulant for the validity of KpN model parameter estimation, is generated and obeyed using Matlab The random number of KpN model.Fig. 4, Fig. 5, Fig. 6 are to join in invention to the result of KpN model parameter estimation and other two kinds of KpN models The experimental result comparison diagram of number estimation method.Wherein, Fig. 4 is estimated result comparison diagram of three kinds of methods for form parameter v, figure 5 be three kinds of methods for noise parameter pnEstimated result comparison diagram, Fig. 6 is the estimation knot of three kinds of methods for scale parameter b Fruit comparison diagram, Fig. 4, Fig. 5, the abscissa of Fig. 6 all indicate experiment number, and ordinate all indicates the mean square error of estimation, band circle The corresponding Home Parameter Estimation Method of curve, the corresponding curve with rice word is zlog (z) method for parameter estimation, with square Corresponding curve is method for parameter estimation of the invention.It can be found through observation, the mean square error of the method for the present invention estimation is wanted Less than remaining two methods, this illustrates that the accuracy of the method for the present invention estimation is higher.
The probability density function for the KpN model that the present invention utilizes in Fig. 4 experiment is as shown in formula five:
Using formula five, the Mellin of available KpN model converts φZ(s) expression formula is as follows:
Wherein W, what () indicated is Whittaker function.
Second Second Type characteristic function ξ of KpN model can be further obtained according to formula sixZ(s) expression formula It is as follows:
What wherein U () was indicated is Tricomi function.
Expression formula by the theoretical log cumulant of the available KpN model of formula seven is as follows:
WhereinIt indicates i rank theoretical log cumulant, formula one and formula eight is subjected to simultaneous, it is available such as formula two Estimated expression.Further, it is also possible to carry out calculating of the invention using the probability density function of KpN model other forms, no Influence actual effect of the invention.

Claims (1)

1. a kind of ShipTargets detection method based on KpN model, which is characterized in that include the following steps:
Statistical modeling is carried out using KpN distribution to obtained diameter radar image and utilizes pair of diameter radar image Number cumulant estimates the parameter of KpN model, calculates constant false alarm rate detection threshold value according to the estimated value of KpN model parameter, The detection for ShipTargets is realized using constant false alarm rate detection；
Wherein, it solves to obtain the estimated value of form parameter v in KpN model using following formulaThe estimated value of scale parameter bAnd it makes an uproar Acoustical power pnEstimated value
Wherein Ψ () is psi function, and Ψ () is polygamma function, and N is equivalent number, and M indicates that the pixel in image is total The expression of number, parameter A, B, C, D is shown below:
Also, the specific formula for calculation of CFAR detection threshold value T is as follows:
WhereinIt is the probability density function of KpN model；PfaIt indicates false alarm rate, is manually set according to actual needs, KpN model refers to K plus noise model.
CN201710364490.7A 2017-05-22 2017-05-22 ShipTargets detection method based on KpN model Active CN107122764B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710364490.7A CN107122764B (en) 2017-05-22 2017-05-22 ShipTargets detection method based on KpN model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710364490.7A CN107122764B (en) 2017-05-22 2017-05-22 ShipTargets detection method based on KpN model

Publications (2)

Publication Number Publication Date
CN107122764A CN107122764A (en) 2017-09-01
CN107122764B true CN107122764B (en) 2019-08-06

Family

ID=59727560

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710364490.7A Active CN107122764B (en) 2017-05-22 2017-05-22 ShipTargets detection method based on KpN model

Country Status (1)

Country Link
CN (1) CN107122764B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108985292A (en) * 2018-05-23 2018-12-11 中国地质大学（武汉） A kind of SAR image CFAR object detection method and system based on multi-scale division
CN112560897B (en) * 2020-11-24 2024-02-09 西安电子科技大学 Radar ship classification capability evaluation method based on captain statistical model

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102176000A (en) * 2011-01-27 2011-09-07 电子科技大学 Sea clutter suppression method for marine radar
CN104504711A (en) * 2014-12-29 2015-04-08 西安交通大学 Vascular image processing method based on circular contour polarity
CN105005983A (en) * 2015-04-13 2015-10-28 西南科技大学 SAR image background clutter modeling and target detection method
CN106468770A (en) * 2016-09-23 2017-03-01 西安电子科技大学 Closely optimum radar target detection method under K Distribution Clutter plus noise
CN106646403A (en) * 2016-11-16 2017-05-10 电子科技大学 K distributed radar clutter real-time simulation method and system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102176000A (en) * 2011-01-27 2011-09-07 电子科技大学 Sea clutter suppression method for marine radar
CN104504711A (en) * 2014-12-29 2015-04-08 西安交通大学 Vascular image processing method based on circular contour polarity
CN105005983A (en) * 2015-04-13 2015-10-28 西南科技大学 SAR image background clutter modeling and target detection method
CN106468770A (en) * 2016-09-23 2017-03-01 西安电子科技大学 Closely optimum radar target detection method under K Distribution Clutter plus noise
CN106646403A (en) * 2016-11-16 2017-05-10 电子科技大学 K distributed radar clutter real-time simulation method and system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"Parameter estimation for Pareto and K distributed clutter with noise";Stephen Bocquet;《IET Radar, Sonar and Navigation》;20151231;第1-10页
"Sea Clutter Modeling in Presence of Thennal Noise Using Beta-Prime Texture Distribution";Izzeddine Chalabi，Amar Mezache;《International Multi-Conference on Systems, Signals &Devices》;20151231;第1-5页

Also Published As

Publication number Publication date
CN107122764A (en) 2017-09-01

Similar Documents

Publication Publication Date Title
CN103761748B (en) Anomaly detection method and device
CN106872958B (en) Radar target self-adapting detecting method based on linear fusion
CN111965632B (en) Radar target detection method based on Riemann manifold dimensionality reduction
CN105699952B (en) Double quantile methods of estimation of sea clutter K profile shape parameters
CN107102302B (en) Radar target fusion detection method based on non-Gaussian parameters
CN109143195B (en) Radar target detection method based on full KL divergence
CN102928822B (en) Radar target length calculation method based on high-resolution one-dimensional range profiles
CN103792522B (en) Multi-radar marine target robust association algorithm based on credible association pair
CN107390194B (en) A kind of radar target detection method based on the graceful divergence of full Donald Bragg
CN107122764B (en) ShipTargets detection method based on KpN model
CN104076350A (en) Method for detecting floating targets on sea based on normalized Doppler spectrum
CN114114192A (en) Cluster target detection method
KR101817011B1 (en) Clutter reduction Method and Apparatus for active sonar based on clustering characteristics
CN105158734A (en) Single-vector hydrophone passive positioning method based on array invariants
CN106353743A (en) Asymptotically optimal radar target detection method matched to equivalent shape parameter
CN113608193A (en) Radar multi-target distance and speed estimation method based on UNet
CN106019250B (en) Based on angle scintillations relay type decoy discrimination method
CN110988856B (en) Target detection trace agglomeration algorithm based on density clustering
CN117368877A (en) Radar image clutter suppression and target detection method based on generation countermeasure learning
CN107479052A (en) Ground concealed target detection method based on Generalized Gaussian Distribution Model
CN107315169B (en) Clutter covariance matrix estimation method based on second-order statistic similarity
CN113093135A (en) Target detection method and device based on F norm normalized distance
CN112098952B (en) Radar reconnaissance clutter suppression method based on time domain statistical processing
CN111123233B (en) Sea clutter texture estimation method based on correlation
Hongwei et al. Remote passive sonar detection by relative multiscale change entropy

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
GR01 Patent grant
GR01 Patent grant