CN113281714A - Bird target detection method based on radar micro Doppler feature enhancement - Google Patents

Bird target detection method based on radar micro Doppler feature enhancement Download PDF

Info

Publication number
CN113281714A
CN113281714A CN202110488406.9A CN202110488406A CN113281714A CN 113281714 A CN113281714 A CN 113281714A CN 202110488406 A CN202110488406 A CN 202110488406A CN 113281714 A CN113281714 A CN 113281714A
Authority
CN
China
Prior art keywords
bird
micro
doppler
target
bird 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
Application number
CN202110488406.9A
Other languages
Chinese (zh)
Inventor
陈宇晨
程巧灵
董辰初
韩旻
邢嘉悦
杨磊
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.)
Civil Aviation University of China
Original Assignee
Civil Aviation University of China
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 Civil Aviation University of China filed Critical Civil Aviation University of China
Priority to CN202110488406.9A priority Critical patent/CN113281714A/en
Publication of CN113281714A publication Critical patent/CN113281714A/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
    • G01S7/415—Identification of targets based on measurements of movement associated with the target

Landscapes

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

Abstract

A bird target detection method based on radar micro Doppler feature enhancement. The method comprises the steps of establishing a bird target radar echo model to obtain bird target radar echo signals; extracting bird micro Doppler characteristics from bird target radar echo signals; enhancing the bird micro Doppler characteristics by using an alternating direction multiplier method to obtain an enhanced bird micro Doppler characteristic spectrogram; obtaining partial flight parameters of the birds from the enhanced bird micro Doppler characteristic spectrogram, obtaining other parameters through the relation between the flight parameters and the quality, identifying the bird target, and finally obtaining the type of the bird target. The method can effectively solve the problem that the micro Doppler features are difficult to extract due to low signal-to-noise ratio and low sampling rate in actual life, and can obtain a high-resolution micro Doppler spectrogram through a small amount of observation data so as to identify birds.

Description

Bird target detection method based on radar micro Doppler feature enhancement
Technical Field
The invention belongs to the technical field of radar imaging, and particularly relates to a bird target detection Method based on radar micro-Doppler feature enhancement, which is used for realizing micro-Doppler feature enhancement by firstly establishing a micro-Doppler feature optimization model by using Short-Time Fourier Transform (STFT) and Compressed sensing theory (CS), and then solving the optimization model by using an Alternating Direction Multiplier Method (ADMM).
Background
The existing air target monitoring mainly comprises infrared detection, optical detection and radar detection. Compared with the prior art, the radar has strong anti-interference performance, has strong penetrating power, is not blocked by wind, fog, cloud and rain, and can remotely detect the target in all weather and for a long time. In the technical field of radar imaging, unlike the traditional radar, the non-coherent processing is adopted, only the position and the speed of a point target can be obtained, the inverse synthetic aperture radar obtains the azimuth high resolution by adopting a coherent processing method, and obtains the range high resolution by transmitting a large-bandwidth signal. Therefore, radar is a relatively ideal detection means for target tracking imaging.
When a bird target is detected and identified by using a radar, the echo of the target not only contains the radial motion information of the bird body, but also contains periodic micro-motion information of bird wing flapping and the like, the micro-motion information can also cause periodic modulation of the radar echo phase, namely, extra frequency modulation is generated near the Doppler frequency to form a sideband, the phenomenon is called micro Doppler effect, and the sideband modulated signal is called micro Doppler signal of the target. Such as flapping of wings of birds, is a unique feature different from other targets in the air, and generates unique micro-doppler signals, which are critical factors for target identification. For the purpose of identifying birds, how to extract micro doppler characteristic information (micro doppler frequency) of birds is important.
Because the echo signals modulated by birds received by the radar belong to time-varying non-stationary signals, the most classical micro-Doppler frequency extraction method for the signals is a joint time-frequency analysis method, the method mainly converts time-domain echo signals to a time-frequency spectrum, describes the time-varying signals from a two-dimensional time-frequency domain, better reveals the instantaneous frequency variation characteristic of the signals, and simultaneously separates the bird main body movement and the micro-movement thereof according to different performances on the time-frequency spectrum. In many joint time-frequency analysis methods, considering the simplicity of operation and the requirement of low cross term interference, the most common method is STFT in linear time-frequency analysis, but the time resolution and the frequency resolution cannot meet the requirement at the same time. And radar echo signal in reality can receive many-sided influences such as environmental noise, and the data of flight birds that can gather are few, and this all causes little Doppler characteristic to be difficult to extract.
In order to solve the above problems, people introduce a Compressed Sensing theory (CS) based on the natural sparseness or compressibility of echo signals generated by the flapping of birds, so as to compress and process the signals of bird targets. The compressed sensing theory indicates that a signal can be multiplied by an original signal by using a matrix, dimensionality reduction is performed on the signal to realize compression of the signal, reconstruction of the compressed signal can be realized by using a sparse representation algorithm, and the original signal is finally restored. The current algorithms for feature extraction based on compressed sensing can be divided into 3 categories: the problem that how to select a relatively ideal optimization algorithm to realize micro-Doppler feature enhancement is needed to be solved at present is a Greedy algorithm of matching pursuit, a Bayesian method and a convex optimization algorithm.
In summary, it is a current challenge how to enhance and extract the micro-doppler feature of birds by using the existing joint time-frequency analysis method and the algorithm of feature extraction based on compressed sensing, so as to identify birds.
Disclosure of Invention
In order to solve the above problems, the present invention aims to provide a bird target detection method based on radar micro-doppler feature enhancement.
In order to achieve the above object, the bird target detection method based on radar micro-doppler feature enhancement provided by the invention comprises the following steps in sequence:
step 1, establishing a bird target radar echo model by using a bird target dynamic flapping wing model flying at low altitude to obtain a bird target radar echo signal;
step 2, aiming at the periodicity of a bird target dynamic flapping wing model, extracting bird micro Doppler characteristics from the bird target radar echo signals by using short-time Fourier transform;
introducing sparse prior in a time-frequency domain, and enhancing the bird micro-Doppler characteristics by using an alternating direction multiplier method to obtain an enhanced bird micro-Doppler characteristic spectrogram;
and 4, obtaining partial flight parameters of the birds from the enhanced bird micro Doppler characteristic spectrogram, and obtaining other parameters according to the relation between the flight parameters and the quality, so as to identify the bird target and finally obtain the type of the bird target.
In step 1, the method for establishing a bird target radar echo model by using the bird target dynamic flapping wing model flying at low altitude to obtain a bird target radar echo signal comprises the following steps:
constructing a bird target dynamic flapping wing model for flapping wing motion of a bird target by using a method for simplifying the body structure of the bird target under the condition of only considering the distribution of upper arm and forearm bones; on the basis of the bird target dynamic flapping wing model, a radar scattering area superposition method based on geometric theory is used for obtaining a bird target radar echo signal Sbird(t)。
In step 2, aiming at the periodicity of the bird target dynamic flapping wing model, the method for extracting bird micro Doppler characteristics from the bird target radar echo signals by using short-time Fourier transform comprises the following steps:
processing the bird target radar echo signals by using short-time Fourier transform of a Gaussian window, wherein the time domain signals are the bird target radar echo signals S obtained in the step 1bird(t) denotes complex conjugation, g (t) denotes a gaussian window:
Figure BDA0003051416810000031
obtaining a bird micro Doppler characteristic frequency spectrum diagram in an ideal state, wherein two sin-like functions are generated by micro Doppler modulation of radar echo due to micro wing motion of birds when the birds perform flapping wing motion, and the amplitude value of the sin-like functions is the maximum micro Doppler frequency offset so as to extract bird micro Doppler characteristics.
In step 3, introducing sparse prior in the time-frequency domain, and enhancing the bird micro-doppler characteristic by using an alternating direction multiplier method to obtain an enhanced bird micro-doppler characteristic spectrogram, the method comprises the following steps:
by utilizing the signal sparsity concept in the compressive sensing theory, the bird target radar echo signal S scattered by a plurality of birds isbird(t) constructing a sparse model by using a bird target radar echo data matrix Y, and reconstructing a bird target radar echo data matrix X to be recovered after feature enhancement from a conventional azimuth Fourier dictionary matrix AfExtracting a plurality of bird target radar echo data matrixes X to be recovered after the optimal sparse representation characteristics of atoms are enhanced, and finally solving the minimization l1Norm problem, will solve the minimum l1The norm problem is converted by a Lagrange method, so that a final micro Doppler characteristic optimization model can be obtained;
Figure BDA0003051416810000032
Figure BDA0003051416810000033
wherein Y isn∈CM×N，Xn∈RK×NIs a bird target radar echo data matrix and a corresponding bird target radar echo data matrix to be recovered in N time windows with the length of M, namely the frequency coefficient of a micro Doppler signal, K represents the frequency coefficient X of the micro Doppler signal under different time windowsnThe number of (2);
then introduces an alternative direction multiplier method in a convex optimization algorithm to solve the final micro Doppler characteristic optimization modelSolution, which updates each optimization variable by parallel iteration, to the frequency coefficient X of the micro-Doppler signalnEstimating, finally obtaining a bird target radar echo data matrix X to be recovered after characteristic enhancement through parallel operation of the N submodules, and obtaining higher signal resolution under a relatively smaller iteration number m so as to complete micro Doppler characteristic enhancement of bird target radar echo signals and obtain an enhanced bird micro Doppler characteristic spectrogram.
In step 4, the method for obtaining part of flight parameters of birds from the enhanced bird micro doppler characteristic spectrogram and obtaining other parameters through the relation between the flight parameters and the quality to identify bird targets and finally obtain the types of the bird targets comprises the following steps:
reading out the micro Doppler frequency offset f from the enhanced bird micro Doppler characteristic spectrogrammicroTime T of bird observation by radarbirdAnd the number of cycle repetitions N during this timebirdIn time, wing length L and flapping frequency f of bird target can be approximately calculatedflapThen, obtaining other flight parameters of the birds including wing areas and wing loads through a functional relation between bird flight parameters and bird mass, which is described by a micro Doppler characteristic characterization function of the bird parameters; finally, the flight parameters are combined with bird motion characteristics, flight time and space characteristics and group habits in real life, so that the type of the bird target can be identified.
The bird target detection method based on radar micro Doppler characteristic enhancement provided by the invention has the following beneficial effects: the Method firstly utilizes Short-Time Fourier Transform (STFT) and compressive sensing theory (CS) to establish a micro Doppler feature optimization model, and then utilizes an Alternating Direction Multiplier Method (ADMM) to solve the optimization model to realize the radar-based micro Doppler feature enhancement of micro Doppler feature enhancement. The method can effectively solve the problem that the micro Doppler features are difficult to extract due to low signal-to-noise ratio and low sampling rate in actual life, and can obtain a high-resolution micro Doppler spectrogram through a small amount of observation data so as to identify birds.
Drawings
FIG. 1 is a flow chart of a bird target detection method based on radar micro-Doppler feature enhancement provided by the invention;
FIG. 2 is a radar tracking bird identification simulation diagram;
FIG. 3 is a diagram of bird micro-Doppler signature spectra in an ideal state;
FIG. 4 is a diagram illustrating the effect of using the alternative direction multiplier method to process and image the data after down-sampling and noise adding; wherein, the image is obtained after down sampling and noise adding, (b) is an enhanced bird micro Doppler characteristic spectrogram obtained after processing by using an alternative direction multiplier method;
FIG. 5 is a graph of ideal micro-Doppler characteristic spectrum under different parameters; wherein (a) is to reduce the flapping frequency to fflap0Hz, changing the bird's flying posture from flapping wing state to gliding state, (b) increasing flapping wing frequency to fflap2.0Hz, (c) to increase the bird translation speed to V2 m/s, and (d) to increase the forearm length to L21m images, (e) increase flapping amplitude to Aflap165 ° images.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings.
As shown in fig. 1, the bird target detection method based on radar micro-doppler feature enhancement provided by the invention comprises the following steps in sequence:
step 1, establishing a bird target radar echo model by using a bird target dynamic flapping wing model flying at low altitude to obtain a bird target radar echo signal;
when constructing a bird target dynamic flapping wing model flying at low altitude, firstly simplifying a complex geometric model, omitting feathers and muscles of birds, omitting hands, and only considering the distribution of upper arm and forearm bones, wherein the length of the upper arm is assumed to be L1The length of the forearm is L2Bird translation speed is V, flapping wing frequency is fflapThe wing-twisting frequency is ftwistSweeping wingFrequency fsweepWhile A isflap、AtwistAnd AsweepThe flapping wing amplitude, the twisting wing amplitude and the sweeping wing amplitude are respectively, the corresponding flapping wing angle is alpha (t), the corresponding twisting wing angle is beta (t) and the corresponding sweeping wing angle is gamma (t), and when the parameters are changed, other different birds are represented. Constructing a bird target dynamic flapping wing model by using the parameters;
secondly, because the flying motion of birds is complex, when a bird target dynamic flapping wing model is used for establishing a corresponding bird target Radar echo model, Radar scattering area (RCS for short) superposition method based on geometric theory can be used for processing, namely, the structures of all parts of the birds are simplified into ellipsoids, then Radar echo signals of the single ellipsoid are modeled based on the Radar scattering area, and then the Radar echo models are added to obtain an approximate and simplified bird target Radar echo model, and the bird target Radar echo model is used for solving bird target Radar echo signals sbird(ii) a Wherein it is assumed that the radar transmits a very narrow rectangular pulse sequence at a frequency fcPulse width is Δ, pulse interval is Δ T:
Figure BDA0003051416810000051
wherein s isbody、sright_upperarm、sright_forearm、sleft_upperarm、sleft_forearmAnd σbody、σright_upperarm、σright_forearm、σleft_upperarm、σleft_forearmRespectively representing radar echo signals and radar scattering areas corresponding to ellipsoids simplified from bird bodies, right upper arms, right forearms, left upper arms and left forearms, c representing wave velocity, k representing total number of pulses in an observation period, R (t) representing the distance between the radar and the ellipsoids at the time t, and rect (t) representing a rectangular function.
Figure 2 pictorially illustrates the situation where radar tracking identifies birds, where the circles represent radar.
Step 2, aiming at the periodicity of a bird target dynamic flapping wing model, extracting bird micro Doppler characteristics from the bird target radar echo signals by using short-time Fourier transform;
processing the bird target radar echo signals by using short-time Fourier transform of a Gaussian window, wherein the time domain signals are the bird target radar echo signals S obtained in the step 1bird(t) denotes complex conjugation, g (t) denotes a gaussian window:
Figure BDA0003051416810000061
the bird micro-Doppler characteristic frequency spectrum diagram in an ideal state as shown in fig. 3 can be obtained, wherein the two sin-like functions are generated by micro-Doppler modulation of radar echo due to micro-wing motion of birds during flapping, and the amplitude of the sin-like functions is the maximum micro-Doppler frequency offset, so that bird micro-Doppler characteristics can be extracted.
When the length L of the upper arm of the bird is changed1Length of forearm L2Frequency of flapping wing fflapIn the process of internal parameters, different bird target radar echo signals can be obtained, and then time-frequency analysis is carried out on the obtained signals, so that bird micro Doppler characteristic spectrograms of different bird parameters in an ideal state can be obtained as shown in fig. 5.
Introducing sparse prior in a time-frequency domain, and enhancing the bird micro-Doppler characteristics by using an alternating direction multiplier method to obtain an enhanced bird micro-Doppler characteristic spectrogram;
firstly, a signal sparsity concept in a compressed sensing theory is utilized to build the bird target radar echo signal obtained in the step 1 into the following sparse model (sparse representation is essentially a linear approximation process):
Y＝AfX+N
Af＝[A(Fbird(1)),A(Fbird(2)),...,A(Fbird(n))]
Figure BDA0003051416810000062
wherein Y represents bird target radar return signal s scattered by multiplebirdX represents a bird target radar echo data matrix to be recovered after characteristic enhancement, and N represents additive noise; fbird(n) azimuthal Doppler frequency due to bird flight, AfRepresenting a conventional orientation fourier dictionary matrix.
When reconstructing the bird target radar data matrix X to be recovered after feature enhancement, the conventional azimuth Fourier dictionary matrix A is required to be usedfExtracting a plurality of atoms to optimally express an avian target radar data matrix X to be recovered after the characteristic enhancement, wherein the method with the least non-zero elements in various sparse expression methods is the optimal sparse expression method which is equivalent to solving the method l0Norm minimization problem, so a micro Doppler feature optimization model can be initially established:
Figure BDA0003051416810000063
Figure BDA0003051416810000064
wherein | · | purple sweet0Is represented by0Norm, | X | luminance0And representing the number of nonzero items in the bird target radar data matrix X to be recovered after the characteristics are enhanced. But due to l0The norm minimization problem, which is a discontinuous, NP-hard problem, is usually reduced to solving the minimization l1Problem of norm:
Figure BDA0003051416810000071
Figure BDA0003051416810000072
wherein by solving for bird targets within N time windows of length MRadar echo data matrix Yn∈CM×NCorresponding bird target radar echo data matrix (namely frequency coefficient of micro Doppler signal) X to be recoveredn∈RK×NThe purpose of time and frequency analysis can be achieved, and K represents the frequency coefficient X of the micro Doppler signal under different time windowsnThe number of the bird target radar echo signals in the nth time window is as follows:
Yn＝AXn+Nn
minimize the solution above by l1The norm problem is converted by a Lagrange method, so that a final micro Doppler characteristic optimization model can be obtained:
Figure BDA0003051416810000073
Figure BDA0003051416810000074
and the micro Doppler characteristic enhancement of the bird target radar echo signal can be completed by solving the final micro Doppler characteristic optimization model.
And then, an alternating direction multiplier method in a convex optimization algorithm is introduced to solve the final micro Doppler feature optimization model, so that higher signal resolution is obtained under the condition of reducing iteration times. Introducing an auxiliary variable znBy construction of
Figure BDA0003051416810000075
The micro Doppler feature optimization model is simpler and more visual:
Figure BDA0003051416810000076
s.t.Xn＝zn
further simplify and define it
Figure BDA0003051416810000077
Then the constraint of the target problem is sorted to d n0, while introducing an enhanced lagrange function:
Figure BDA0003051416810000078
where ρ represents a penalty parameter, λnRepresenting the Lagrange multiplier matrix lambda ∈ RN×KColumn n. On the basis, the frequency coefficient X of the micro Doppler signal can be estimated by iteratively updating each optimization variable in paralleln：
Figure BDA0003051416810000081
Where m represents the number of iterations.
The alternative direction multiplier method is to use the frequency coefficient X of the micro Doppler signalnThe estimation is divided into five steps to be solved, and finally, a bird target radar data matrix X to be recovered after the characteristics are enhanced is obtained through the parallel operation of the N submodules. The alternative direction multiplier method needs multiple iterations to converge to an accurate solution, but the obtained spectral analysis diagram can be popularized to appropriate accuracy by selecting appropriate punishment parameters rho and iteration times m under relatively small iteration times.
The conditions of the signal-to-noise ratio SNR of 0dB and the sampling rate n of 4096 are selected to simulate the situation of low signal-to-noise ratio and low sampling rate in actual life, and the conditions are processed according to the above alternative direction multiplier method to obtain the enhanced bird micro Doppler characteristic spectrogram shown in fig. 4.
Step 4, obtaining part of flight parameters of the birds from the enhanced bird micro Doppler characteristic spectrogram, and obtaining other parameters according to the relation between the flight parameters and the quality, so as to identify the bird target and finally obtain the type of the bird target;
reading radar observation from the enhanced bird micro Doppler characteristic frequency spectrogram obtained by processing the bird target radar echo signal in the step 3Time T of birdsbirdNumber of cycle repetitions N within this timebirdAnd a micro Doppler frequency offset fmicroThen observing the time T of the birds by using a radarbirdAnd the number of cycle repetitions N during this timebirdApproximate calculation of flapping frequency f of bird targetflap：
Figure BDA0003051416810000082
Micro Doppler frequency offset fmicroThe expression of (a) is:
Figure BDA0003051416810000083
wherein L represents the wing length of the bird target, which is equal to the length of the upper arm as L1And forearm length L2Summing; thetamaxThe maximum included angle between the airfoil surface and the horizontal plane is represented, alpha and beta respectively represent the azimuth angle and the pitch angle of the bird target relative to the radar and are related to the coordinate positions of the bird and the radar; w 2 pi fflapλ represents radar wavelength, both known data:
in practical application, through observation of birds and collection of corresponding data, a micro Doppler characteristic characterization function of bird parameters can be obtained, which explains the functional relation between the flight parameters of the birds and the quality of the birds, as shown in the following table, so that the flapping frequency f of the bird target is utilizedflapAnd a micro Doppler frequency offset fmicroObtaining other flight parameters of the birds, such as wing area, wing load and the like, through the functional relation between the flight parameters of the bird target and the mass of the bird target; and finally, combining the flight parameters with bird motion characteristics, flight time and space characteristics and group habits in real life to identify the type of the bird target.
Figure BDA0003051416810000091
The effect of the present invention will be further described with reference to the simulation data experiment.
1. Simulation conditions are as follows:
all simulation parameters are shown in the following table:
Figure BDA0003051416810000092
2. simulation data packet experimental analysis:
by comparing the micro-doppler feature spectrogram in the ideal state shown in fig. 3 with the effect chart of data processing imaging after downsampling and noise addition by using the alternating direction multiplier method shown in fig. 4, it can be observed that the time-frequency analysis chart at this time is different from the ideal situation, but the micro-doppler feature can be extracted by reading some parameters through the image to obtain the corresponding micro-motion feature of the bird target, so as to identify the bird target, which indicates that the micro-doppler feature enhancement by using the alternating direction multiplier method based on STFT not only basically realizes the function of denoising, but also reconstructs the high-resolution micro-doppler feature from a small number of radar observation signals.
Through the micro doppler characteristic spectrogram in fig. 3, it can be read that the amplitude of the sin-like function with a larger amplitude is about 100Hz, and the amplitude of the sin-like function with a smaller amplitude is about 50Hz, in this figure, the sin-like function is equivalent to micro doppler frequency offset, and the amplitude is the maximum micro doppler frequency offset, and the current wing length of the bird can be calculated through step 4: upper arm is L10.4761m with a half span length L1+L20.9523m, and a parameter L in simulation1＝0.5m，L1+L2It can be known that when the micro doppler frequency offset is read out from the spectrogram, the wing length of the bird target can be estimated, and then the bird can be identified by the functional relationship between the flight parameters and the mass of the bird target.

Claims (5)

1. A bird target detection method based on radar micro Doppler feature enhancement is characterized by comprising the following steps: the bird target detection method based on radar micro Doppler feature enhancement comprises the following steps in sequence:
step 1, establishing a bird target radar echo model by using a bird target dynamic flapping wing model flying at low altitude to obtain a bird target radar echo signal;
step 2, aiming at the periodicity of a bird target dynamic flapping wing model, extracting bird micro Doppler characteristics from the bird target radar echo signals by using short-time Fourier transform;
introducing sparse prior in a time-frequency domain, and enhancing the bird micro-Doppler characteristics by using an alternating direction multiplier method to obtain an enhanced bird micro-Doppler characteristic spectrogram;
and 4, obtaining partial flight parameters of the birds from the enhanced bird micro Doppler characteristic spectrogram, and obtaining other parameters according to the relation between the flight parameters and the quality, so as to identify the bird target and finally obtain the type of the bird target.
2. The method for bird target detection based on radar micro-doppler signature enhancement as claimed in claim 1 wherein: in step 1, the method for establishing a bird target radar echo model by using the bird target dynamic flapping wing model flying at low altitude to obtain a bird target radar echo signal comprises the following steps:
constructing a bird target dynamic flapping wing model for flapping wing motion of a bird target by using a method for simplifying the body structure of the bird target under the condition of only considering the distribution of upper arm and forearm bones; on the basis of the bird target dynamic flapping wing model, a radar scattering area superposition method based on geometric theory is used for obtaining a bird target radar echo signal Sbird(t)。
3. The method for bird target detection based on radar micro-doppler signature enhancement as claimed in claim 1 wherein: in step 2, aiming at the periodicity of the bird target dynamic flapping wing model, the method for extracting bird micro Doppler characteristics from the bird target radar echo signals by using short-time Fourier transform comprises the following steps:
processing the bird target radar echo signals by using short-time Fourier transform of a Gaussian window, wherein the time domain signals are the bird target radar echo signals S obtained in the step 1bird(t) denotes complex conjugation, g (t) denotes a gaussian window:
Figure FDA0003051416800000011
obtaining a bird micro Doppler characteristic frequency spectrum diagram in an ideal state, wherein two sin-like functions are generated by micro Doppler modulation of radar echo due to micro wing motion of birds when the birds perform flapping wing motion, and the amplitude value of the sin-like functions is the maximum micro Doppler frequency offset so as to extract bird micro Doppler characteristics.
4. The method for bird target detection based on radar micro-doppler signature enhancement as claimed in claim 1 wherein: in step 3, introducing sparse prior in the time-frequency domain, and enhancing the bird micro-doppler characteristic by using an alternating direction multiplier method to obtain an enhanced bird micro-doppler characteristic spectrogram, the method comprises the following steps:
by utilizing the signal sparsity concept in the compressive sensing theory, the bird target radar echo signal S scattered by a plurality of birds isbird(t) constructing a sparse model by using a bird target radar echo data matrix Y, and reconstructing a bird target radar echo data matrix X to be recovered after feature enhancement from a conventional azimuth Fourier dictionary matrix AfExtracting a plurality of bird target radar echo data matrixes X to be recovered after the optimal sparse representation characteristics of atoms are enhanced, and finally solving the minimization l1Norm problem, will solve the minimum l1The norm problem is converted by a Lagrange method, so that a final micro Doppler characteristic optimization model can be obtained;
Figure FDA0003051416800000021
Figure FDA0003051416800000022
wherein Y isn∈CM×N，Xn∈RK×NIs a bird target radar echo data matrix and a corresponding bird target radar echo data matrix to be recovered in N time windows with the length of M, namely the frequency coefficient of a micro Doppler signal, K represents the frequency coefficient X of the micro Doppler signal under different time windowsnThe number of (2);
then introduces the alternative direction multiplier method in the convex optimization algorithm to solve the final micro Doppler characteristic optimization model, and updates each optimization variable through parallel iteration to carry out frequency coefficient X of the micro Doppler signalnEstimating, finally obtaining a bird target radar echo data matrix X to be recovered after characteristic enhancement through parallel operation of the N submodules, and obtaining higher signal resolution under a relatively smaller iteration number m so as to complete micro Doppler characteristic enhancement of bird target radar echo signals and obtain an enhanced bird micro Doppler characteristic spectrogram.
5. The method for bird target detection based on radar micro-doppler signature enhancement as claimed in claim 1 wherein: in step 4, the method for obtaining part of flight parameters of birds from the enhanced bird micro doppler characteristic spectrogram and obtaining other parameters through the relation between the flight parameters and the quality to identify bird targets and finally obtain the types of the bird targets comprises the following steps:
reading out the micro Doppler frequency offset f from the enhanced bird micro Doppler characteristic spectrogrammicroTime T of bird observation by radarbirdAnd the number of cycle repetitions N during this timebirdIn time, wing length L and flapping frequency f of bird target can be approximately calculatedflapAnd obtaining other flight parameters including wing area and wing load of the birds through the functional relation between the bird flight parameters and the bird mass described by the micro Doppler characteristic characterization function of the bird parametersCounting; finally, the flight parameters are combined with bird motion characteristics, flight time and space characteristics and group habits in real life, so that the type of the bird target can be identified.
CN202110488406.9A 2021-05-06 2021-05-06 Bird target detection method based on radar micro Doppler feature enhancement Pending CN113281714A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110488406.9A CN113281714A (en) 2021-05-06 2021-05-06 Bird target detection method based on radar micro Doppler feature enhancement

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110488406.9A CN113281714A (en) 2021-05-06 2021-05-06 Bird target detection method based on radar micro Doppler feature enhancement

Publications (1)

Publication Number Publication Date
CN113281714A true CN113281714A (en) 2021-08-20

Family

ID=77277973

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110488406.9A Pending CN113281714A (en) 2021-05-06 2021-05-06 Bird target detection method based on radar micro Doppler feature enhancement

Country Status (1)

Country Link
CN (1) CN113281714A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116449368A (en) * 2023-06-14 2023-07-18 中国人民解放军国防科技大学 Imaging method, device and equipment of short-distance millimeter wave MIMO-SAR

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109190149A (en) * 2018-07-20 2019-01-11 北京理工大学 A kind of emulation verification method extracting frequency of fluttering its wings up and down based on birds EM scatter model
CN109409225A (en) * 2018-09-21 2019-03-01 清华大学 Unmanned plane classification method and device based on the fusion of radar multipath signal time-frequency characteristics
CN111708011A (en) * 2020-07-10 2020-09-25 南京天朗防务科技有限公司 Micro Doppler velocity measurement method based on compressed sensing

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109190149A (en) * 2018-07-20 2019-01-11 北京理工大学 A kind of emulation verification method extracting frequency of fluttering its wings up and down based on birds EM scatter model
CN109409225A (en) * 2018-09-21 2019-03-01 清华大学 Unmanned plane classification method and device based on the fusion of radar multipath signal time-frequency characteristics
CN111708011A (en) * 2020-07-10 2020-09-25 南京天朗防务科技有限公司 Micro Doppler velocity measurement method based on compressed sensing

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
刘振: "基于压缩感知的随机调制雷达信号处理方法与应用研究", 《中国优秀博硕士学位论文全文数据库（博士） 信息科技辑》 *
林益耳: "基于微多普勒特征的人类行为分类方法研究", 《中国优秀博硕士学位论文全文数据库(博士)信息科技辑》 *
陈宏昆 等: "鸟类目标微多普勒分析及参数估计", 《电讯技术》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116449368A (en) * 2023-06-14 2023-07-18 中国人民解放军国防科技大学 Imaging method, device and equipment of short-distance millimeter wave MIMO-SAR
CN116449368B (en) * 2023-06-14 2023-08-25 中国人民解放军国防科技大学 Imaging method, device and equipment of short-distance millimeter wave MIMO-SAR

Similar Documents

Publication Publication Date Title
Huizing et al. Deep learning for classification of mini-UAVs using micro-Doppler spectrograms in cognitive radar
Liao et al. Identification of ground targets from sequential high-range-resolution radar signatures
Yuan et al. Micro-Doppler analysis and separation based on complex local mean decomposition for aircraft with fast-rotating parts in ISAR imaging
CN110244303B (en) SBL-ADMM-based sparse aperture ISAR imaging method
CN108226927B (en) SAR imaging method based on weighted iteration minimum sparse Bayesian reconstruction algorithm
CN112099008B (en) SA-ISAR imaging and self-focusing method based on CV-ADMMN
CN104076360B (en) The sparse target imaging method of two-dimensional SAR based on compressed sensing
CN108562884A (en) A kind of Air-borne Forward-looking sea-surface target angle ultra-resolution method based on maximum a posteriori probability
CN111025257A (en) Micro-motion target high-resolution time-frequency diagram reconstruction method based on sparse Bayesian learning
Zhu et al. Classification of UAV-to-ground targets based on enhanced micro-Doppler features extracted via PCA and compressed sensing
CN109507666A (en) The sparse frequency band imaging method of ISAR based on off-network variation bayesian algorithm
Abdulatif et al. Towards adversarial denoising of radar micro-Doppler signatures
CN113281714A (en) Bird target detection method based on radar micro Doppler feature enhancement
CN105759264B (en) Fine motion target defect echo high-resolution imaging method based on time-frequency dictionary
CN108919263B (en) ISAR high-resolution imaging method based on maximum mutual information criterion
Wang et al. A novel STAP method based on structured sparse recovery of clutter spectrum
CN113608218A (en) Frequency domain interference phase sparse reconstruction method based on back projection principle
CN113376606A (en) Clutter suppression method for fast convergence sparse Bayes along clutter ridge
CN110046326B (en) Time-frequency DOA estimation method
CN112784916B (en) Air target micro-motion parameter real-time extraction method based on multitask convolutional network
CN106405509B (en) The piecemeal processing method of space-time adaptive signal
CN116027293A (en) Rapid sparse angle super-resolution method for scanning radar
Cifola et al. Target/clutter disentanglement using deep adversarial training on micro-Doppler signatures
CN112099010B (en) ISAR (inverse synthetic aperture radar) imaging method for target with micro-motion component based on structured non-convex low-rank representation
CN113421281A (en) Pedestrian micromotion part separation method based on segmentation theory

Legal Events

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

Application publication date: 20210820

WD01 Invention patent application deemed withdrawn after publication