CN111025252A - Target feature extraction method and device, computer storage medium and electronic equipment - Google Patents

Target feature extraction method and device, computer storage medium and electronic equipment Download PDF

Info

Publication number
CN111025252A
CN111025252A CN201911254029.1A CN201911254029A CN111025252A CN 111025252 A CN111025252 A CN 111025252A CN 201911254029 A CN201911254029 A CN 201911254029A CN 111025252 A CN111025252 A CN 111025252A
Authority
CN
China
Prior art keywords
target
determining
signal
feature extraction
blur
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201911254029.1A
Other languages
Chinese (zh)
Inventor
刘佳琪
艾夏
刘鑫
刘洪艳
王明亮
王伟东
穆磊
张生俊
张鹏
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.)
China Academy of Launch Vehicle Technology CALT
Beijing Aerospace Changzheng Aircraft Institute
Original Assignee
China Academy of Launch Vehicle Technology CALT
Beijing Aerospace Changzheng Aircraft Institute
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 China Academy of Launch Vehicle Technology CALT, Beijing Aerospace Changzheng Aircraft Institute filed Critical China Academy of Launch Vehicle Technology CALT
Priority to CN201911254029.1A priority Critical patent/CN111025252A/en
Publication of CN111025252A publication Critical patent/CN111025252A/en
Pending legal-status Critical Current

Links

Images

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

Abstract

The target feature extraction method, the target feature extraction device, the computer storage medium and the electronic equipment comprise the following steps: determining a frequency signal S (omega) of the target echo; determining a ambiguity function A (theta, tau) of the frequency signal from the frequency signal; constructing a subdomain A at the origin of a blur plane determined from said blur functionΩ(ii) a According to the sub-field AΩAnd extracting to obtain the characteristic P of the target. By adopting the scheme in the application, the compressed sensing is introduced into the time-frequency joint distribution analysis method, and the calculation efficiency can be greatly improved.

Description

Target feature extraction method and device, computer storage medium and electronic equipment
Technical Field
The present application relates to spatial information countermeasure technology, and in particular, to a target feature extraction method, apparatus, computer storage medium, and electronic device.
Background
The complexity of the spatial multi-target electromagnetic characteristics is embodied in three aspects of the complexity of electromagnetic coupling effect, the complexity of motion form and the complexity of electromagnetic environment. At present, the electromagnetic characteristics of a space single target are deeply known, and a sliding type, distributed type and other scattering center characterization model is provided; the method has the advantages that complex motions such as space target micromotion and the like are preliminarily known, and a single-target micromotion-to-radar wave modulation model is established; the method is mainly used for solving the problems that the electromagnetic environment of a space target and the coupling mechanism of the target are preliminarily researched, and the interaction between the active interference radiation wave and the scattering wave of the target is analyzed. However, theoretical analysis and quantitative description are seriously lacked in the aspects of how to represent the scattering center type of the space multi-target complex electromagnetic coupling effect, the modulation mechanism of the complex motion to radar electromagnetic waves and the evolution rule thereof, and a multi-target scattering + multi-point source radiation electromagnetic representation model.
The key point for solving the problems is to extract various characteristics of the space target, and a traditional time-frequency joint distribution method needs a large amount of calculation, especially when the target bandwidth is wide.
Disclosure of Invention
The embodiment of the application provides a target feature extraction method and device, a computer storage medium and electronic equipment, so as to solve the technical problems.
According to a first aspect of embodiments of the present application, there is provided a target feature extraction method, including the steps of:
determining a frequency signal S (omega) of the target echo;
determining a ambiguity function A (theta, tau) of the frequency signal from the frequency signal;
constructing a subdomain A at the origin of a blur plane determined from said blur functionΩ；
According to the sub-domainAΩAnd extracting to obtain the characteristic P of the target.
According to a second aspect of embodiments of the present application, there is provided a target feature extraction apparatus including:
a signal determination module for determining a frequency signal S (ω) of the target echo;
a function determination module for determining a fuzzy function A (theta, tau) of the frequency signal from the frequency signal;
a subdomain constructing module for constructing a subdomain A at the origin of the blur plane determined according to the blur functionΩ；
A feature extraction module for extracting a feature from the subdomain AΩAnd extracting to obtain the characteristic P of the target.
According to a third aspect of embodiments of the present application, there is provided a computer storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the target feature extraction method as described above.
According to a fourth aspect of embodiments of the present application, there is provided an electronic device comprising a memory for storing one or more programs, and one or more processors; the one or more programs, when executed by the one or more processors, implement the target feature extraction method as described above.
By adopting the target feature extraction method and device, the computer storage medium and the electronic equipment provided by the embodiment of the application, the compressed sensing is introduced into the time-frequency joint distribution analysis method, and the calculation efficiency is greatly improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
FIG. 1 shows sparse reconstruction l in an embodiment of the present applicationpA geometric principle schematic diagram of the algorithm;
FIG. 2 is a schematic diagram showing the composition of observation vector combinations in the embodiment of the present application;
FIG. 3 is a schematic flow chart illustrating an implementation of a target feature extraction method in an embodiment of the present application;
fig. 4 is a schematic structural diagram illustrating a target feature extraction device in the second embodiment of the present application;
fig. 5 is a schematic structural diagram of an electronic device in the fourth embodiment of the present application;
fig. 6 is a schematic diagram illustrating a process of acquiring a reconstructed signal in an embodiment five of the present application;
fig. 7 shows a schematic diagram of suppressing interference caused by cross terms by the CSJTF method in the fifth embodiment of the present application.
Detailed Description
In the process of implementing the present application, the inventors found that:
the dynamic target electromagnetic scattering characteristic information reconstruction algorithm is a core link of dynamic electromagnetic scattering characteristic sparse modeling, and because the dimension M of an observation vector is lower than the dimension N of an original signal, the original signal is recovered by y, so that the problem of underdetermination is solved, namely an equation does not have a definite solution. The compressive sensing theory can be converted into the following minimum l by means of the inherent characteristic of signal sparsity0Solving a norm problem:
Figure BDA0002309803890000031
the solution of the optimization problem is an NP-hard problem, and the combined exhaustive operation in the process of requiring the optimal sparse solution is difficult to realize.
In recent years, a great deal of research is carried out on alternative algorithms with controllable operation complexity by relatives, and a series of reconstruction algorithms with good performance appear, which can be roughly divided into the following categories:
the first possible method is to replace the above-mentioned signal reconstruction problem of equation 0-1 with the following optimization problem:
Figure BDA0002309803890000032
wherein 0 < p < ∞, | | · | | non-volatile gaspIs 1pAnd (4) norm. Different values of p correspond to different algorithms, and the corresponding principle is shown in fig. 1. The larger the value of p is, the more the optimal solution tends to diffuse to each coordinate system of the two-dimensional space, and the smaller the value of p is, the more the optimal solution tends to be sparse. In practice, 0 < p < 1 norm optimization problem is often adopted to reconstruct sparse signals, and different minimum lpThe norm algorithm has different operation complexity and reconstruction performance. The most classical of these algorithms is the minimum l1Norm algorithm, hereinafter referred to as l1The algorithm is as follows:
Figure BDA0002309803890000033
when the sensing matrix Θ satisfies the better RIP property (finite equidistant property), l1The algorithm can obtain the minimum l under relatively more observation matrixes0And (5) reconstructing results equivalent to the norm algorithm. And it belongs to the convex relaxation problem, can be solved by the linear programming method; relative minimum of0In the norm algorithm, the operation degree is greatly simplified.
l1The algorithm is also called as BP (base Pursuit) algorithm, and is a convex optimization problem which can be solved by a Linear programming (Linear Program) method, and related researches have proved that the algorithm is an equivalent form of an equation 0-1 under specific RIP constraints. Meanwhile, the electromagnetic scattering characteristic of the dynamic target is simulated by adopting a mathematical method, and the method is different from the actual measurement with noise, so that the method can be completely used for sparse reconstruction of the dynamic electromagnetic scattering characteristic.
Except that1Besides the algorithm, a greedy iteration algorithm converts the sparse reconstruction problem into an iterative two-times estimation problem. In the greedy iterative algorithm, a non-zero coefficient index set (suppp (f) { i: f (i) ≠ 0}) of a sparse vector of a signal to be reconstructed is called a support set of the signal, and a column vector theta of a sensing matrix isjReferred to as atoms. While the observation vector y is exactly { ΘjJ ∈ supp (f) } linear combinations of atoms within, as shown in FIG. 2.
If the signal can be correctly identified in the real support set Γ ═ supp (f), thenAtom Θ that can be corresponded by a support setjThe sparsity (non-zero coefficient of f) at the position of the sparse signal supporting set is solved by a simple linear algebra method (shown by a black thick line frame in fig. 2), so that the reconstruction of the signal is realized:
Figure BDA0002309803890000041
Figure BDA0002309803890000042
wherein the content of the first and second substances,
Figure BDA0002309803890000043
pseudo-inverse matrices representing sub-matrices whose column indices are composed of elements within ΓcIs the complement of Γ, and f | Γ represents the elements defining the following table as index set Γ. The greedy algorithm is based on this idea, and a support set of signals is identified through an iterative loop to construct a least squares approximation of the signals thereon.
The extraction of the Doppler component mainly comprises short-time Fourier transform, but the method cannot meet the high resolution of two dimensions of time and frequency, and can only meet the high resolution analysis of one component. Therefore, researchers provide a Wegener-Weili distribution (WVD) quadratic time-frequency analysis method, but the method has the problem of serious cross-term interference.
In order to solve the above problems, embodiments of the present application provide a method and an apparatus for extracting target features, a computer storage medium, and an electronic device, where dynamic electromagnetic scattering property sparseness modeling is performed based on compressive sensing, and in order to keep consistent with the modeling method, the embodiments of the present application verify a calculation result of a doppler component by using a joint time-frequency analysis method based on compressive sensing.
The scheme in the embodiment of the application can be implemented by adopting various computer languages, such as object-oriented programming language Java and transliterated scripting language JavaScript.
In order to make the technical solutions and advantages of the embodiments of the present application more apparent, the following further detailed description of the exemplary embodiments of the present application with reference to the accompanying drawings makes it clear that the described embodiments are only a part of the embodiments of the present application, and are not exhaustive of all embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Example one
Fig. 3 shows a schematic flow chart of an implementation of a target feature extraction method in a first embodiment of the present application.
As shown in the figure, the target feature extraction method includes:
step 301, determining a frequency signal S (omega) of a target echo;
step 302, determining a fuzzy function A (theta, tau) of the frequency signal according to the frequency signal;
step 303, constructing a subdomain A at the fuzzy plane origin determined according to the fuzzy functionΩ；
Step 304, according to the sub-domain AΩAnd extracting the time-frequency characteristic P of the target, namely Wegener-Weili distribution (WVD).
By adopting the target feature extraction method provided by the embodiment of the application, the compressed sensing is introduced into a time-frequency joint distribution analysis method, and the calculation efficiency is greatly improved.
In one embodiment, the blur function a (θ, τ) is:
Figure BDA0002309803890000051
where ω denotes frequency, θ denotes frequency offset, and j denotes unit imaginary number
Figure BDA0002309803890000061
τ denotes the time delay.
In one embodiment, the method includes determining the sub-field AΩCalculating the characteristic P of the target according to the following formula:
Figure BDA0002309803890000062
wherein | · | purple sweet1Representing a norm of l1, | · |. non-woven2Is represented by2Norm, F-1Representing a two-dimensional pseudo-fourier transform and epsilon represents the residual.
In one embodiment, the constructing of the subdomain A at the origin of the blur plane determined from the blur functionΩThe method comprises the following steps:
selecting a region with an area smaller than a preset value around the origin of the fuzzy plane as a subfield AΩ(ii) a Signal energy capable of eliminating interference item energy in the region can be obtained;
the sub-field AΩThe quasi-fourier transform of (a) is related to the feature P, satisfying:
AΩ＝F-1{P}。
in one embodiment, the sub-field represents a small portion of the entire blur field, and a sufficiently small region around the center of the blur plane is selected in which signal energy to reject the energy of the interfering item is obtained. Distribution P and subfield A of CSJTFΩBy quasi-Fourier transform, i.e.
AΩ＝F-1{P}
In one embodiment, the fuzzy function a (θ, τ) is transformed with the wigner-willi distribution WVD by:
A(θ,τ)＝F-1{W(t,ω)}；
wherein W (t, ω) represents WVD,
Figure BDA0002309803890000063
F-1representing a two-dimensional pseudo-fourier transform.
In one embodiment, the wigner-willi profile W (t, ω) is obtained by:
1) determining a candidate supporting point Lambda according to the observation vector y and the sensing matrix theta; the candidate support point is a column index which is most relevant to residual errors in the sensing matrix;
2) determining a support set gamma according to the candidate support point lambda;
3) performing a least squares estimation of the signals on the support set Γ;
4) updating the residual error;
5) and outputting the Weigner-Weiley distribution W (t, omega) of the reconstruction signal when the preset termination condition is met, and otherwise, returning to the step 1).
In one embodiment, the support set Γ: performing least square estimation of the signal, specifically using the following formula:
Figure BDA0002309803890000071
wherein f represents the vector coefficient of the sparse signal, y represents the observation vector, and Θ represents the observation matrix.
Example two
Based on the same inventive concept, the embodiment of the present application provides a target feature extraction device, and the principle of the device for solving the technical problem is similar to that of a target feature extraction method, and repeated parts are not repeated.
Fig. 4 shows a schematic structural diagram of a target feature extraction device in the second embodiment of the present application.
As shown in the figure, the target feature extraction device includes:
a signal determining module 401, configured to determine a frequency signal S (ω) of the target echo;
a function determining module 402 for determining a blurring function a (θ, τ) of the frequency signal from the frequency signal;
a sub-field constructing module 403 for constructing a sub-field A at the origin of the blur plane determined according to the blur functionΩ；
A feature extraction module 404 for extracting features from the subdomain AΩAnd extracting to obtain the characteristic P of the target.
By adopting the target feature extraction device provided by the embodiment of the application, the compressed sensing is introduced into a time-frequency joint distribution analysis method, and the calculation efficiency is greatly improved.
In one embodiment, the blur function a (θ, τ) is:
Figure BDA0002309803890000081
where ω denotes frequency, θ denotes frequency offset, and j denotes unit imaginary number
Figure BDA0002309803890000082
τ denotes the time delay.
In one embodiment, the method includes determining the sub-field AΩCalculating the characteristic P of the target according to the following formula:
Figure BDA0002309803890000083
wherein | · | purple sweet1Is represented by1Norm, | · | luminance2Is represented by2Norm, F-1Representing a two-dimensional pseudo-fourier transform and epsilon represents the residual.
In one embodiment, the constructing of the subdomain A at the origin of the blur plane determined from the blur functionΩThe subdomain represents a small portion of the overall fuzzy domain, including:
a sufficiently small region is selected around the center of the blur plane in which signal energy with interference term energy removed can be obtained. Distribution P and subfield A of CSJTFΩBy quasi-Fourier transform, i.e.
AΩ＝F-1{P}
In one embodiment, the fuzzy function a (θ, τ) is transformed with the wigner-willi distribution WVD by:
A(θ,τ)＝F-1{W(t,ω)}；
wherein W (t, ω) represents WVD,
Figure BDA0002309803890000084
F-1representing a two-dimensional pseudo-fourier transform.
In one embodiment, the wigner-willi profile W (t, ω) is obtained by:
1) determining a candidate supporting point Lambda according to the observation vector y and the sensing matrix theta; the candidate support point is a column index which is most relevant to residual errors in the sensing matrix;
2) determining a support set gamma according to the candidate support point lambda;
3) performing a least squares estimation of the signals on the support set Γ;
4) updating the residual error;
5) and outputting the Weigner-Weiley distribution W (t, omega) of the reconstruction signal when the preset termination condition is met, and otherwise, returning to the step 1).
In one embodiment, the least squares estimation of the signals on the support set Γ is performed using the following equation:
Figure BDA0002309803890000091
wherein f represents the vector coefficient of the sparse signal, y represents the observation vector, and Θ represents the observation matrix.
EXAMPLE III
Based on the same inventive concept, embodiments of the present application further provide a computer storage medium, which is described below.
The computer storage medium has a computer program stored thereon, and the computer program, when executed by a processor, implements the steps of the target feature extraction method according to an embodiment.
By adopting the computer storage medium provided by the embodiment of the application, the compressed sensing is introduced into the time-frequency joint distribution analysis method, and the calculation efficiency is greatly improved.
Example four
Based on the same inventive concept, the embodiment of the present application further provides an electronic device, which is described below.
Fig. 5 shows a schematic structural diagram of an electronic device in the fourth embodiment of the present application.
As shown, the electronic device includes memory 501 for storing one or more programs, and one or more processors 502; the one or more programs, when executed by the one or more processors, implement the target feature extraction method as described in embodiment one.
By adopting the electronic equipment provided by the embodiment of the application, the compressed sensing is introduced into the time-frequency joint distribution analysis method, and the calculation efficiency is greatly improved.
EXAMPLE five
In order to facilitate the implementation of the present application, the embodiments of the present application are described with a specific example.
Fig. 6 shows a schematic process diagram for acquiring a reconstructed signal in the fifth embodiment of the present application.
As shown, the process of acquiring the reconstructed signal may include:
1) inputting an observation vector y and a sensing matrix theta;
the observation vector is a known or measured signal, and the sensing matrix is customized by a user.
2) Initialization: vector coefficient f of sparse signal, initial value f (0)0; setting the initial value of residual error r as y, and setting the initial value of iteration count l as 1; the observation vector updates the residual error at each step.
3) Selecting a column index which is most related to the residual error as a candidate support point of the signal by using the following formula;
Figure BDA0002309803890000101
wherein, thetajThe j-th column vector of the sensing matrix is shown, and r is a residual error.
4) Combining the support set with the support set of the previous step to form a new support set; each step forms a new support set based on the support set update of the previous step.
Γ(l)＝Λ∩Γ(l-1)
Wherein l is the number of iterations.
5) Performing a least squares estimation of the signal on the new support set;
Figure BDA0002309803890000102
wherein f represents the vector coefficient of the sparse signal, y represents the observation vector, and Θ represents the observation matrix.
6) Residual error updating;
the residual r may be specifically updated as: r is y- Θ f'.
7) Judging whether a termination condition is met;
the termination conditions may specifically be: the residual error meets the user-defined numerical value or the iteration number reaches the user-set numerical value.
8) If yes, outputting a reconstruction signal;
9) if not, the iteration times are l + +, and the step 3) is returned to.
The Wegener-Weili distribution W (t, omega) of the reconstruction signal is obtained through the process.
Next, the target feature p is extracted by using a joint time-frequency analysis method of compressed sensing, which is specifically described as follows.
A joint time-frequency analysis method (CSJTF) based on compressed sensing is an improvement on a Wegener-Weili distribution WVD method. WVD of the signal is represented as
Figure BDA0002309803890000111
Where W (t, ω) represents WVD, and S (ω) represents a frequency signal. In addition, WVD can also adopt a fuzzy function of the signal to express A (theta, tau)
Figure BDA0002309803890000112
The conversion relation between the fuzzy function and the WVD is
A(θ,τ)＝F-1{W(t,ω)} (0-7)
Wherein, F-1Representing a two-dimensional pseudo-fourier transform.
The CSJTF method for calculating the time-frequency distribution characteristics comprises three steps: firstly, calculating a fuzzy function of a signal by adopting an equation (0-6); second, sub-field A is constructed at the origin of the blur planeΩRepresenting the entire fuzzy domainA part of. By constructing a sub-domain which is small enough at the origin of the fuzzy plane, the cross terms in the time-frequency image can be effectively eliminated; third, the CSJTF profile P is calculated.
Distribution P and subfield A of CSJTFΩBy quasi-Fourier transform, i.e.
AΩ＝F-1{P} (0-8)
Equation (0-8) consists of a series of linear equations of the form y-Tx.
Wherein y is AΩX denotes the vector form of P and T denotes the quasi-fourier transform. However, there is no definite solution for equations (0-8), given subfield AΩIn this case, there are many solutions for P. However, P is known a priori to have sparse properties, e.g. only a very few time-frequency joint curves. From the analysis of the preceding edge based on compressed sensing, the problem can be solved by1The norm gives a reasonable solution, and the calculation formula is as follows
Figure BDA0002309803890000121
Wherein | · | purple sweet1Is represented by1Norm, | · | luminance2Is represented by2Norm, ε represents the residual.
Fig. 7 shows a schematic diagram of suppressing interference caused by cross terms by the CSJTF method in the fifth embodiment of the present application.
As shown in the figure, the time-frequency image obtained by the CSJTF method has almost no cross term interference and is well matched with the analytic solution. Therefore, the embodiment of the application utilizes CSJTF as a time-frequency joint domain high-resolution analysis method for accurately extracting the target characteristic quantity.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the present application.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims (10)

1. A target feature extraction method is characterized by comprising the following steps:
determining a frequency signal S (omega) of the target echo;
determining a ambiguity function A (theta, tau) of the frequency signal from the frequency signal;
constructing a subdomain A at the origin of a blur plane determined from said blur functionΩ；
According to the sub-field AΩAnd extracting to obtain the characteristic P of the target.
2. The method according to claim 1, characterized in that the blurring function a (θ, τ) is:
Figure FDA0002309803880000011
where ω denotes frequency, θ denotes frequency offset, and j denotes unit imaginary number
Figure FDA0002309803880000012
τ denotes the time delay.
3. The method of claim 1, wherein said determining is based on said sub-field aΩCalculating to obtain the characteristic P of the target, specifically calculating according to the following formula:
Figure FDA0002309803880000013
wherein | · | purple sweet1Is represented by1Norm, | · | luminance2Is represented by2Norm, F-1Representing a two-dimensional pseudo-fourier transform and epsilon represents the residual.
4. The method of claim 1, wherein constructing subdomain a at a blur plane origin determined from said blur functionΩThe method comprises the following steps:
selecting a region with an area smaller than a preset value around the origin of the fuzzy plane as a subfield AΩ(ii) a Signal energy capable of eliminating interference item energy can be obtained in the region;
the sub-field AΩThe quasi-fourier transform of (a) is related to the feature P, satisfying:
AΩ＝F-1{P}。
5. the method according to claim 1, wherein the transfer relationship between the blur function a (θ, τ) and the wigner-willi distribution WVD is:
A(θ,τ)＝F-1{W(t,ω)}；
wherein W (t, ω) represents WVD,
Figure FDA0002309803880000021
F-1representing a two-dimensional pseudo-fourier transform.
6. The method according to claim 5, characterized in that said Wegener-Willi distribution W (t, ω) is obtained by:
1) determining a candidate supporting point Lambda according to the observation vector y and the sensing matrix theta; the candidate support point is a column index which is most relevant to residual errors in the sensing matrix;
2) determining a support set gamma according to the candidate support point lambda;
3) performing a least squares estimation of the signals on the support set Γ;
4) updating the residual error;
5) and outputting the Weigner-Weiley distribution W (t, omega) of the reconstruction signal when the preset termination condition is met, and otherwise, returning to the step 1).
7. The method of claim 6, wherein the least squares estimation of the signals over the support set Γ is performed using the following equation:
Figure FDA0002309803880000022
wherein f represents the vector coefficient of the sparse signal, y represents the observation vector, and Θ represents the observation matrix.
8. A target feature extraction device characterized by comprising:
a signal determination module for determining a frequency signal S (ω) of the target echo;
a function determination module for determining a fuzzy function A (theta, tau) of the frequency signal from the frequency signal;
a subdomain constructing module for constructing a subdomain A at the origin of the blur plane determined according to the blur functionΩ；
A feature extraction module for extracting a feature from the subdomain AΩAnd extracting to obtain the characteristic P of the target.
9. A computer storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
10. An electronic device comprising one or more processors, and memory for storing one or more programs; the one or more programs, when executed by the one or more processors, implement the method of any of claims 1 to 7.
CN201911254029.1A 2019-12-10 2019-12-10 Target feature extraction method and device, computer storage medium and electronic equipment Pending CN111025252A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911254029.1A CN111025252A (en) 2019-12-10 2019-12-10 Target feature extraction method and device, computer storage medium and electronic equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911254029.1A CN111025252A (en) 2019-12-10 2019-12-10 Target feature extraction method and device, computer storage medium and electronic equipment

Publications (1)

Publication Number Publication Date
CN111025252A true CN111025252A (en) 2020-04-17

Family

ID=70206260

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911254029.1A Pending CN111025252A (en) 2019-12-10 2019-12-10 Target feature extraction method and device, computer storage medium and electronic equipment

Country Status (1)

Country Link
CN (1) CN111025252A (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102135618A (en) * 2010-12-23 2011-07-27 北京航空航天大学 Frequency domain compressive sensing method aiming at sparse SAR (Synthetic Aperture Radar) images in airspace
CN105891789A (en) * 2016-03-30 2016-08-24 西安电子科技大学 Combined time frequency distribution and compression sensing radar frequency smeared spectrum interference method
US20160259849A1 (en) * 2015-03-05 2016-09-08 Tata Consultancy Services Limited System and method for reducing size of raw data by performing time-frequency data analysis
CN105974412A (en) * 2016-06-07 2016-09-28 电子科技大学 Target feature extraction method used for synthetic aperture radar
CN109101462A (en) * 2018-09-13 2018-12-28 南京邮电大学 The Fourier Transform Algorithm of time loss can be reduced

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102135618A (en) * 2010-12-23 2011-07-27 北京航空航天大学 Frequency domain compressive sensing method aiming at sparse SAR (Synthetic Aperture Radar) images in airspace
US20160259849A1 (en) * 2015-03-05 2016-09-08 Tata Consultancy Services Limited System and method for reducing size of raw data by performing time-frequency data analysis
CN105891789A (en) * 2016-03-30 2016-08-24 西安电子科技大学 Combined time frequency distribution and compression sensing radar frequency smeared spectrum interference method
CN105974412A (en) * 2016-06-07 2016-09-28 电子科技大学 Target feature extraction method used for synthetic aperture radar
CN109101462A (en) * 2018-09-13 2018-12-28 南京邮电大学 The Fourier Transform Algorithm of time loss can be reduced

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
金艳等: "基于α稳定分布噪声稀疏性及最优匹配的跳频信号参数估计", 《电子与信息学报》 *
金艳等: "基于稀疏时频分布的跳频信号参数估计", 《电子与信息学报》 *

Similar Documents

Publication Publication Date Title
Adler et al. Solving ill-posed inverse problems using iterative deep neural networks
Lenthe et al. A spherical harmonic transform approach to the indexing of electron back-scattered diffraction patterns
Van der Tol et al. Image Domain Gridding: a fast method for convolutional resampling of visibilities
Huang et al. Traction force microscopy with optimized regularization and automated Bayesian parameter selection for comparing cells
Johnson et al. Convolution kernel design and efficient algorithm for sampling density correction
Engquist et al. The quadratic Wasserstein metric for inverse data matching
CN109343060B (en) ISAR imaging method and system based on deep learning time-frequency analysis
Li et al. Reconstruction of heterogeneous materials via stochastic optimization of limited-angle X-ray tomographic projections
Guillet et al. Modelling spatially correlated observation errors in variational data assimilation using a diffusion operator on an unstructured mesh
Liu et al. Efficient and qualified mesh generation for Gaussian molecular surface using adaptive partition and piecewise polynomial approximation
Zhongming et al. A monotonic two-step iterative shrinkage/thresholding algorithm for sound source identification based on equivalent source method
CN112257021A (en) Selectable kriging spatial interpolation rainfall estimation method
CN113917540B (en) Method for denoising seismic data by anti-spurious ray beam based on sparse constraint
Gálvez et al. Immunological-based approach for accurate fitting of 3D noisy data points with Bézier surfaces
Metsis et al. A new hierarchical partition of unity formulation of EFG meshless methods
CN109001805B (en) Seismic source inversion uncertainty analysis method, storage medium and server
Boyle et al. Methods for calculating the electrode position Jacobian for impedance imaging
JP4262762B2 (en) Electromagnetic field analysis program
Riis et al. Computed tomography reconstruction with uncertain view angles by iteratively updated model discrepancy
Mou et al. Dictionary learning based low-dose x-ray CT reconstruction using a balancing principle
Li et al. A modified dual-level fast multipole boundary element method for large-scale three-dimensional potential problems
CN111025252A (en) Target feature extraction method and device, computer storage medium and electronic equipment
CN116776695A (en) One-dimensional electromagnetic calculation method, system and equipment based on ultra-high-order finite element technology
Ding et al. Accelerating multi‐dimensional interpolation using moving least‐squares on the GPU
Li et al. Contrast enhancement for cDNA microarray image based on fourth-order moment

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: 20200417