CN117033974A - Method for detecting sea clutter background target based on regenerated kernel Hilbert space - Google Patents
Method for detecting sea clutter background target based on regenerated kernel Hilbert space Download PDFInfo
- Publication number
- CN117033974A CN117033974A CN202310991514.7A CN202310991514A CN117033974A CN 117033974 A CN117033974 A CN 117033974A CN 202310991514 A CN202310991514 A CN 202310991514A CN 117033974 A CN117033974 A CN 117033974A
- Authority
- CN
- China
- Prior art keywords
- vector
- sea clutter
- state vector
- state
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000005259 measurement Methods 0.000 claims abstract description 39
- 230000003190 augmentative effect Effects 0.000 claims abstract description 25
- 230000007704 transition Effects 0.000 claims abstract description 18
- 230000003416 augmentation Effects 0.000 claims abstract description 8
- 238000013507 mapping Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 22
- 230000001133 acceleration Effects 0.000 claims description 3
- 230000007423 decrease Effects 0.000 claims description 3
- 230000004069 differentiation Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 38
- 238000001914 filtration Methods 0.000 abstract description 8
- 230000007246 mechanism Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 230000010287 polarization Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000001172 regenerating effect Effects 0.000 description 2
- 229920000426 Microplastic Polymers 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000739 chaotic effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002592 echocardiography Methods 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005312 nonlinear dynamic Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000008929 regeneration Effects 0.000 description 1
- 238000011069 regeneration method Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000000551 statistical hypothesis test Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/414—Discriminating targets with respect to background clutter
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
- G06F17/13—Differential equations
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Mathematical Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Artificial Intelligence (AREA)
- Radar, Positioning & Navigation (AREA)
- Computer Networks & Wireless Communication (AREA)
- Remote Sensing (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a method for detecting a background target based on regenerated kernel Hilbert space sea clutter, which relates to the technical field of radar target detection and comprises the following steps: collecting signal data containing sea clutter, and combining with system state vector x of radar i And a system measurement vector y i Definition of augmented state vector s i Augmentation of the state vector s by means of a kernel function i Mapping to a high-dimensional feature space; filtering the acquired data containing sea clutter, using a weight vector omega and an augmentation state vector s i Defining new augmented state vectorsn i Then construct the state transition equation F i And measurement equation H i Obtaining and updating system measurement vector y i The method comprises the steps of carrying out a first treatment on the surface of the Calculating updated system measurement vector y i And the original input data u i Inter prediction error e i The method comprises the steps of carrying out a first treatment on the surface of the If the prediction error e i If the target is larger than the preset judgment threshold, the target is considered to exist.
Description
Technical Field
The invention relates to the technical field of radar target detection, in particular to a method for detecting a target based on a regenerated kernel Hilbert space sea clutter background.
Background
Target detection in the sea clutter background is always an important and difficult problem in the fields of radar detection, tracking and identification. Sea clutter has non-gaussian and non-stationary characteristics, which is one of the main constraints affecting small Radar Cross Section (RCS), small target detection performance that floats on the sea surface at low speed. In a complex marine environment, backward electromagnetic scattering of sea clutter is strong, and a target signal in a radar echo is easily covered by sea spike waves caused by sea waves, so that a conventional target power-based detector sea clutter generally has a high False Alarm Rate (FAR).
The main stream sea clutter background target detection method comprises the following steps:
1. the detection method based on sea surface fractal and chaos features comprises the following steps: the fractal-based method has the characteristics of simple calculation, high efficiency and the like, but as the sea clutter time sequence has fractal characteristics only in a non-scale area with a certain time scale, the interval is different along with the change of radar parameters, sea conditions and polarization, and the performance of a detector is also influenced by the deviation of estimation of the non-scale area; when the observation time is longer, the method can obtain better detection performance for the small floating targets on the sea surface, however, if the observation time is shorter, the performance of the detector slides down seriously, a learner improves short-time observation, so that the short-time observation can be improved according to the short-term predictability of sea clutter, a nonlinear prediction model is constructed for the sea clutter echo sequence, statistical hypothesis test is carried out by using a prediction error, the purpose of target detection is achieved, but when the signal-to-clutter ratio is continuously reduced, the detection result is still not ideal, and the detection of the small slow weak targets is difficult to realize.
2. The sea surface target detection method based on micro Doppler characteristics comprises the following steps: in the method based on time-frequency analysis, more details and information which cannot be obtained from the time domain can be obtained by carrying out proper time-frequency transformation, the defect that the time domain and the frequency domain are completely separated in Fourier analysis is overcome, and the time domain and the frequency domain can be considered; however, according to the Hessenberg inaccuracy principle, the longer the length of the time window function in the STFT is, the higher the frequency resolution is, and the worse the time resolution is; if the Wigner-Ville transform in bilinear form is used, the resolution of the time domain and the frequency domain can be improved, but due to the nonlinear transform, serious cross terms occur when a plurality of signal components occur, and the occurrence of the cross terms can reduce the detection performance of the detector on weak targets. While SPWVD can well suppress the cross terms, it sacrifices the operation efficiency of the algorithm and the frequency resolution is also reduced when combined. The time-frequency analysis method adopting FRFT has better detection performance on large ship targets, but still has poor detection performance on slow-speed weak small targets on complex sea surfaces although no cross item exists. The weak and small targets on the sea surface are influenced by sea waves, the movement is complex, and the target signals are usually represented as weak nonlinear frequency modulation signals, so that the detection method faces a plurality of challenges in practical application.
3. The feature-based detection method comprises the following steps: the feature-based detection method is to detect a small sea surface floating target by utilizing multiple types of features with complementary characteristics of distinguishing targets and clutter, convert detection problems into single classification problems in a feature space, and then determine a judgment area by utilizing a machine learning algorithm so as to detect the target. The sea surface target detection is completed by extracting Relative average amplitude (Relative Average Amplitude, RAA), relative Doppler peak height (Relative doppler Peak Height, RPH) and Relative Doppler spectrum Entropy (RVE) based on the original three-feature detection method. The detection method based on the time-frequency domain 3 features adopts smooth pseudo Wigner-Willi distribution to calculate NTFD. Sea clutter and echoes containing targets exhibit different characteristics on NTFD. From these different characteristics, 3 time-frequency features can be extracted: the detection is completed by time-frequency accumulation of NTFD, the number of connected areas in a binary image formed by NTFD bright pixels and the size of the maximum connected area. Based on a polarized target decomposition theoretical basis, according to a sea surface floating small target detection scene, a Freeman-Durden three-component decomposition method is selected to model 3 scattering mechanisms respectively, the energy size under the 3 scattering mechanisms is determined, the relative ratio of the energy of a unit to be detected and the energy of a reference unit is selected so as to extract 3 polarized characteristics, and the energy corresponds to the relative volume scattering mechanism, the relative dihedral angle scattering mechanism and the relative surface scattering mechanism respectively. And carrying out feature detection on the small sea surface floating target by using the 3 types of features through a convex hull learning algorithm. The feature-based detection method is based on a large number of samples (or priori information), a decision area is determined by using a machine learning algorithm to detect the target, the detection mechanism is to distinguish multiple types of features with complementary characteristics of the target and the clutter, engineering application is limited, and application and popularization of the method in actual engineering are prevented.
Disclosure of Invention
The invention aims at: the sea clutter background target detection method capable of achieving weak and small target detection in complex sea surface environments is provided based on the regenerated kernel Hilbert space.
The technical scheme of the invention is as follows: the method for detecting the background target of the sea clutter based on the regenerated kernel Hilbert space comprises the following steps:
s1, collecting signal data containing sea clutter, and combining with a system state vector x of a radar i And a system measurement vector y i Definition of augmented state vector s i Augmentation of the state vector s by means of a kernel function i Mapping to a high-dimensional feature space;
s2, filtering the signal data containing the sea clutter acquired in the step S1, and using a weight vector omega and an augmentation state vector S i Defining a new augmented state vector n i Then construct the state transition equation F i And measurement equation H i Obtaining and updating system measurement vector y i ；
S3, calculating an updated system measurement vector y i And the original input data u i Inter prediction error e i ；
S4, if the prediction error e i If the target is larger than the preset judgment threshold, the target is considered to exist.
In any of the above solutions, further, a system state vector x i Including the position and speed of the radarDegree and acceleration information; system measurement vector y i Including information obtained from the radar's external environment.
In any of the above aspects, further, the state vector s is augmented i Expressed as:
wherein u is i Representing the current input signal of the radar system, including target echo signals and sea clutter signals, x i Representing the system state vector at the current moment, x i-1 Representing the system state vector, y at the last time i Representing a system measurement vector at the current moment; u (u) i Representing the current input signal of the system, wherein the current input signal comprises a target echo and a sea clutter signal; f is the state transition equation and h is the measurement equation.
In any of the above technical solutions, further, the state conversion equation converts the current time state vector into the next time state vector; the metrology equation maps the system state vector to a system metrology vector.
In any of the above solutions, further, the state vector s is augmented with a kernel function i Mapping to a high-dimensional feature space requires the use of gaussian radial basis functions:
at this time, the state vector s is augmented i Can be expressed as:
wherein z is j 、z i For two time-frequency eigenvector samples, beta j Is a weight coefficient, κ (z) j ,z i ) Is z j 、z i The inner product of two time-frequency characteristic vector samples, N is the number of input signal samples, and a is Gaussian radial basisSuper-parameters of the function; II z j -z i ‖ 2 Is the squared Euclidean distance between two time-frequency eigenvector samples, κ a (z j ,z i ) The value of the function decreases with the squared euclidean distance and is between 0 and 1;
augmenting the state vector s using gaussian radial basis functions i The simplification is as follows:
s i ＝Ω T ψ(s i-1 ,u i )，
wherein Ω is a weight vector, ψ (s i-1 ,u i ) To regenerate the kernel hilbert space vector.
In any of the above embodiments, further, a new augmented state vector n i The definition formula is:
construction of the state transition equation F i And measurement equation H i And update the target state y i ：
n i ＝F i-1 (n i-1 +w i-1 )，
y i ＝H i (n i +v i )，
Wherein F is i-1 The state transition equation, w, constructed at the previous time i-1 H is the state noise of the last moment i The measurement equation of the current moment, v i Is the measured noise at the current moment.
In any of the above solutions, further, the state transition equation F i By augmenting the state vector n i The iteration is obtained after differentiation, and finally the measurement state y at the next moment is obtained through a Bayesian recursion method i 。
The beneficial effects of the invention are as follows:
in the scheme of the invention, a regeneration kernel Hilbert space theory is utilized, a kernel method and a representation theorem are applied to carry out linear secondary estimation on a full-state space model in a function space, and the linear estimation of a general nonlinear power system in an original input space is obtained; because the high-dimensional linear representation of the filter is completely learned from observed data, the direct modeling of the target motion state and the measurement equation is avoided, the filter has general approximation property, and the filter can be still used under the condition of no sea clutter priori knowledge;
the method has the advantages that the information quantity is calculated by using a positive definite kernel function meeting the Mercer condition, all information of signals is reserved, high-order moments are extracted, the property of covariance is reserved, and the disadvantage that the traditional Bayesian filtering method can only iterate low-order moments is avoided;
as sea clutter prediction is regarded as a Bayesian tracking filtering problem, a nonlinear model does not need to be established in advance, and online real-time detection can be realized.
Drawings
The advantages of the foregoing and additional aspects of the invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flow chart of a method for regenerating kernel Hilbert space-based sea clutter background target detection according to one embodiment of the invention;
FIG. 2 is a diagram of simulation results of HH polarization conditions based on a regenerative kernel Hilbert spatial sea clutter background target detection method according to one embodiment of the invention;
fig. 3 is a diagram of simulation results of HV polarization conditions based on a regenerated kernel hilbert space sea clutter background target detection method according to an embodiment of the invention.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description. It should be noted that, without conflict, embodiments of the present invention and features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those described herein, and the scope of the invention is therefore not limited to the specific embodiments disclosed below.
As shown in fig. 1, the present embodiment provides a method for detecting a background target based on a regenerated kernel hilbert space sea clutter, which includes:
s1, collecting signal data u containing sea clutter i The augmented state vector is formed by combining the radar system state and radar system measurement:
the augmented state vector provides comprehensive information describing the behavior of the radar system, where s i Augmenting state vector, x for current time i Representing the system state vector at the current moment, x i-1 Representing the system state vector, y at the last time i Representing a system measurement vector at the current moment; u (u) i Representing the current input signal of the system, wherein the current input signal comprises a target echo signal and a sea clutter signal; f is a state transition equation, h is a measurement equation; wherein the system state vector describes internal dynamics of the radar system, including position, velocity, and acceleration information; the system measurement vector represents information obtained by the radar system from the external environment, which in this embodiment is obtained by the sensor.
Mapping the augmented state vector to a high-dimensional feature space using a kernel function, a nonlinear dynamic system with linear weights can be constructed, where the augmented state vector s i Can be expressed as:
wherein beta is j Is a weight coefficient, κ (z) j ,z i ) Is z j 、z i The inner product of two time-frequency eigenvector samples, N, is the number of input signal samples.
The gaussian radial basis function is used in mapping:
a is the super parameter of the radial basis function of Gaussian; II z j -z i ‖ 2 Is z j 、z i The squared euclidean distance between two time-frequency eigenvector samples, κ (z j ,z i ) The value of the function decreases with the square Euclidean distance and is between 0 (limit) and 1 (when z j ＝z i Time) are provided.
Augmenting the current time by a state vector s i The simplification is as follows:
s i ＝Ω T ψ(s i-1 ,u i )，
wherein Ω is a weight vector, ψ (s i-1 ,u i ) To regenerate the kernel hilbert space vector.
Specifically, the regenerated hilbert space has an inner product structure, and the filtering operation can be represented as an inner product calculation in the space. By choosing the appropriate kernel function, the inner product calculation in the regenerated hilbert space corresponds to the nonlinear function operation in the original input space.
The regenerated kernel hilbert space uses time-frequency characteristics to carry out recursive filtering, and the mapped time-frequency characteristic space has higher dimensionality, so that the original input signal is not limited, and the input signal does not need to be subjected to specific pretreatment in advance or is required to have specific statistical properties, which is very important for signals with different characteristics and space-time scales.
S2, filtering the signal data containing the sea clutter acquired in the step S1, and using a weight vector omega and an augmentation state vector S i Constitute a new augmented state vector n at the current instant i Then, a state transition equation and a measurement equation are constructed:
n i ＝F i-1 (n i-1 +w i-1 )，
y i ＝H i (n i +v i )，
wherein F is i-1 The state transition equation, w, constructed at the previous time i -1 is the state noise of the last moment, v i For measuring noise at the current moment, representing errors and uncertainties existing in actual measurement, H i Mapping a system state vector to a system measurement vector to obtain direct observation and measurement values of radar system behaviors; state transition equation F i Converting the current time state vector to the next time state vector, F i By (x, x) for n i The iteration is obtained after differentiation, the state modeling and measurement modeling of the target are not needed, the nonlinear system is not needed to be approximated in a certain way, and finally the system measurement vector y at the next moment can be updated and obtained through a Bayesian recursion method i 。
S3, calculating an updated system measurement vector y i And the original input data u i Inter prediction error e i ：
e i ＝|y i -u i |。
S4, if the prediction error e i If the target is larger than the preset judgment threshold, the target is considered to exist.
In another embodiment of the present invention, ten sets of IPIX published data sets measured in different sea clutter environments on the east coast of canada in 1993 were collected and tested on a plastic pellet with 1.0m diameter anchored, as shown in fig. 2, by the method of the present invention for testing based on the regenerated nuclear hilbert space and three other general methods: the sea surface fractal and chaotic characteristic-based detection method, the micro Doppler characteristic-based sea surface target detection method and the characteristic-based detection method are respectively subjected to simulation analysis under the HH polarization condition, so that the detection accuracy of the method provided by the invention is higher; as shown in FIG. 3, simulation analysis is also performed under the HV polarization condition, and the detection accuracy of the detection method provided by the invention is higher.
In summary, the invention provides a method for detecting a background target based on regenerated kernel Hilbert space sea clutter, which comprises the following steps:
s1, collecting signal data containing sea clutter, and combining with a system state vector x of a radar i And a system measurement vector y i Definition of augmented state vector s i Augmentation of the state vector s by means of a kernel function i Mapped to a high-dimensional feature space.
S2, filtering the data containing sea clutter acquired in the step S1, and using a weight vector omega and an augmentation state vector S i Defining a new augmented state vector n i Then construct the state transition equation F i And measurement equation H i Obtaining and updating system measurement vector y i 。
S3, calculating an updated system measurement vector y i And the original input data u i Inter prediction error e i 。
S4, if the prediction error e i If the target is larger than the preset judgment threshold, the target is considered to exist.
The steps in the invention can be sequentially adjusted, combined and deleted according to actual requirements.
Although the invention has been disclosed in detail with reference to the accompanying drawings, it is to be understood that such description is merely illustrative and is not intended to limit the application of the invention. The scope of the invention is defined by the appended claims and may include various modifications, alterations and equivalents of the invention without departing from the scope and spirit of the invention.
Claims (7)
1. The method for detecting the background target of the sea clutter based on the regenerated kernel Hilbert space is characterized by comprising the following steps:
s1, collecting signal data containing sea clutter, and combining with a system state vector x of a radar i And a system measurement vector y i Definition of augmented state vector s i Augmentation of the state vector s by means of a kernel function i Mapping to a high-dimensional feature space;
s2, for the acquired in the step S1Filtering signal data containing sea clutter using a weight vector Ω and an augmented state vector s i Defining a new augmented state vector n i Then construct the state transition equation F i And measurement equation H i Obtaining and updating system measurement vector y i ；
S3, calculating an updated system measurement vector y i And the original input data u i Inter prediction error e i ；
S4, if the prediction error e i If the target is larger than the preset judgment threshold, the target is considered to exist.
2. The method for detecting background targets based on regenerated kernel hilbert space sea clutter according to claim 1, wherein the system state vector x i Including radar position, velocity and acceleration information; the system measures the vector y i Including information obtained from the radar's external environment.
3. The method for detecting a background target based on regenerated kernel hilbert space sea clutter according to claim 1, wherein the augmented state vector s i Expressed as:
wherein u is i Representing the current input signal of the radar system, including target echo signals and sea clutter signals, x i Representing the system state vector at the current moment, x i-1 Representing the system state vector, y at the last time i Representing a system measurement vector at the current moment; u (u) i Representing the current input signal of the system, wherein the current input signal comprises a target echo and a sea clutter signal; f is the state transition equation and h is the measurement equation.
4. The method for detecting a background target based on regenerated kernel hilbert space sea clutter according to claim 3, wherein the state transition equation transitions a current time state vector to a next time state vector; the metrology equation maps a system state vector to a system metrology vector.
5. The method for detecting background targets based on regenerated kernel hilbert space sea clutter according to claim 1, wherein the using kernel function amplifies the state vector s i Mapping to a high-dimensional feature space requires the use of gaussian radial basis functions:
at this time, the state vector s is augmented i Can be expressed as:
wherein z is j 、z i For two time-frequency eigenvector samples, beta j Is a weight coefficient, κ (z) j ,z i ) Is z j 、z i The inner product of two time-frequency characteristic vector samples, N is the number of input signal samples, and a is the super parameter of Gaussian radial basis function; II z j -z i ‖ 2 Is the squared Euclidean distance between two time-frequency eigenvector samples, κ a (z j ,z i ) The value of the function decreases with the squared euclidean distance and is between 0 and 1;
using the gaussian radial basis function to multiply the state vector s i The simplification is as follows:
s i ＝Ω T ψ(s i-1 ,u i )，
wherein Ω is a weight vector, ψ (s i-1 ,u i ) To regenerate the kernel hilbert space vector.
6. The method for detecting background targets based on regenerated kernel Hilbert space sea clutter as claimed in claim 1, wherein the method comprises the steps ofThe new augmented state vector n i The definition formula is:
construction of the state transition equation F i And measurement equation H i And update the target state y i ：
n i ＝F i-1 (n i-1 +w i-1 )，
y i ＝H i (n i +v i )，
Wherein F is i-1 The state transition equation, w, constructed at the previous time i-1 H is the state noise of the last moment i The measurement equation of the current moment, v i Is the measured noise at the current moment.
7. The method for detecting background targets based on regenerated kernel Hilbert space sea clutter according to claim 6, wherein said state transition equation F i By amplifying the state vector n i The iteration is obtained after differentiation, and finally the measurement state y at the next moment is obtained through a Bayesian recursion method i 。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310991514.7A CN117033974A (en) | 2023-08-08 | 2023-08-08 | Method for detecting sea clutter background target based on regenerated kernel Hilbert space |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310991514.7A CN117033974A (en) | 2023-08-08 | 2023-08-08 | Method for detecting sea clutter background target based on regenerated kernel Hilbert space |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117033974A true CN117033974A (en) | 2023-11-10 |
Family
ID=88644165
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310991514.7A Pending CN117033974A (en) | 2023-08-08 | 2023-08-08 | Method for detecting sea clutter background target based on regenerated kernel Hilbert space |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117033974A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118551312A (en) * | 2024-07-29 | 2024-08-27 | 中国人民解放军海军航空大学 | Target detection method based on Spearman correlation characteristic re-expression |
-
2023
- 2023-08-08 CN CN202310991514.7A patent/CN117033974A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118551312A (en) * | 2024-07-29 | 2024-08-27 | 中国人民解放军海军航空大学 | Target detection method based on Spearman correlation characteristic re-expression |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102721952B (en) | Method for steadily classifying ground moving target based on super-resolution Doppler spectrum | |
| CN111707999B (en) | Sea surface floating small target detection method based on combination of multiple features and ensemble learning | |
| CN106569193B (en) | The small targets detection in sea clutter method filtered based on anterior-posterior to income reference particle | |
| CN104360336B (en) | Novel method for extracting radar target micro-motion cycle in self-adaptive mode | |
| CN113534065B (en) | Radar target micro-motion feature extraction and intelligent classification method and system | |
| CN101984360B (en) | Normalized leakage LMS self-adaptive mobile target detector based on FRFT | |
| CN103064063B (en) | Poly-phase code radar signal waveform automatic identification method based on continuous wave Doppler (CWD) feature | |
| CN108226928B (en) | Inverse Synthetic Aperture Radar Imaging Method Based on Expectation Propagation Algorithm | |
| CN108919254A (en) | The CS-PHD method of the motor-driven small and weak multi-target detection tracking of Gao Zhongying radar | |
| CN101872484B (en) | Method for detecting self-adaptive growth of a plurality of dim target tracks in image domain | |
| CN102914769A (en) | Joint fractal-based method for detecting small target under sea clutter background | |
| CN107861109A (en) | A kind of target micro-doppler curve extracting method based on order particles filtering | |
| CN106772352A (en) | A kind of PD radars extension Weak target detecting method based on Hough and particle filter | |
| CN114117912A (en) | Sea clutter modeling and inhibiting method under data model dual drive | |
| Chen et al. | LFM signal detection and estimation based on deep convolutional neural network | |
| CN110501683B (en) | Sea-land clutter classification method based on four-dimensional data characteristics | |
| CN117033974A (en) | Method for detecting sea clutter background target based on regenerated kernel Hilbert space | |
| CN115774248B (en) | Small target detection method and system on sea surface based on multiple image features | |
| Hongzhi et al. | Particle swarm optimization for rigid body reconstruction after micro-Doppler removal in radar analysis | |
| Hou et al. | Sparse coding-inspired high-resolution ISAR imaging using multistage compressive sensing | |
| CN115097408A (en) | Sea surface small target detection method based on improved 1D-Lenet network | |
| CN105334507B (en) | The detection method to offshore floating radar target based on polarization multiple features | |
| CN105353353B (en) | Multiple Search Particle Probability Hypothesis Density Filtering Method for Multiple Target Tracking | |
| Zhang et al. | A novel method for sea clutter suppression and target detection via deep convolutional autoencoder | |
| CN106772357A (en) | AI PHD wave filters under signal to noise ratio unknown condition |
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 |