[go: up one dir, main page]

CN113848589A - A Specific Target Recognition Method for Passive Magnetic Detection Based on Discrete Meyer Wavelets - Google Patents

A Specific Target Recognition Method for Passive Magnetic Detection Based on Discrete Meyer Wavelets Download PDF

Info

Publication number
CN113848589A
CN113848589A CN202110988556.6A CN202110988556A CN113848589A CN 113848589 A CN113848589 A CN 113848589A CN 202110988556 A CN202110988556 A CN 202110988556A CN 113848589 A CN113848589 A CN 113848589A
Authority
CN
China
Prior art keywords
ferromagnetic
signal
axis
target
magnetic
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.)
Granted
Application number
CN202110988556.6A
Other languages
Chinese (zh)
Other versions
CN113848589B (en
Inventor
查冰婷
黄金波
郑震
张合
周郁
顾钒
李红霞
徐光博
王成君
徐陈又诗
袁海璐
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.)
Nanjing University of Science and Technology
Original Assignee
Nanjing University of Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Science and Technology filed Critical Nanjing University of Science and Technology
Priority to CN202110988556.6A priority Critical patent/CN113848589B/en
Publication of CN113848589A publication Critical patent/CN113848589A/en
Application granted granted Critical
Publication of CN113848589B publication Critical patent/CN113848589B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V3/00—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation
    • G01V3/38—Processing data, e.g. for analysis, for interpretation, for correction
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V3/00—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation
    • G01V3/08—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation operating with magnetic or electric fields produced or modified by objects or geological structures or by detecting devices
    • G01V3/087—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation operating with magnetic or electric fields produced or modified by objects or geological structures or by detecting devices the earth magnetic field being modified by the objects or geological structures
    • 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
    • Y02A90/30—Assessment of water resources

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Geology (AREA)
  • Physics & Mathematics (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Geophysics (AREA)
  • Geochemistry & Mineralogy (AREA)
  • Electromagnetism (AREA)
  • Geophysics And Detection Of Objects (AREA)

Abstract

本发明公开了一种基于离散迈耶小波的被动磁探测特定目标识别方法，通过低通滤波电路处理高频噪声信号，采样获得的数字磁信号；采用PCA算法获得自适应尺度函数，将数据库中铁磁性装甲目标的磁信号数据降维处理，降维到k维数据获得对应尺度函数大小，实现迈耶小波多尺度函数的快速离散小波变换；提取磁信号不同频段信号特征，设计双边滤波判决指数分配算法，获得不同频段的磁信号特征的判决指数；建立多信号融合模型，首先将单轴不同频段磁信号在双边滤波算法下获得判决指数累加获得单轴判决指数，再将三轴磁信号的判决指数累加获得本系统的磁信号的融合判决指数；采用基于固定阈值法的快速判别准则，快速判断探测范围内是否存在铁磁性装甲目标。

Figure 202110988556

The invention discloses a method for identifying a specific target of passive magnetic detection based on discrete Meyer wavelet. A low-pass filter circuit is used to process a high-frequency noise signal, and a digital magnetic signal is obtained by sampling; a PCA algorithm is used to obtain an adaptive scale function, and the iron in a database The dimensionality reduction processing of the magnetic signal data of the magnetic armored target, reducing the dimensionality to k -dimensional data to obtain the corresponding scale function, and realizing the fast discrete wavelet transform of the Meyer wavelet multi-scale function; extracting the signal characteristics of the magnetic signal in different frequency bands, and designing the bilateral filtering decision index allocation algorithm to obtain the judgment index of the magnetic signal characteristics of different frequency bands; to establish a multi-signal fusion model, firstly, the single-axis magnetic signals of different frequency bands are obtained by the bilateral filtering algorithm to obtain the judgment index to accumulate to obtain the single-axis judgment index, and then the judgment index of the three-axis magnetic signal is combined. The fusion judgment index of the magnetic signal of the system is obtained by exponential accumulation; the fast judgment criterion based on the fixed threshold method is used to quickly judge whether there is a ferromagnetic armored target within the detection range.

Figure 202110988556

Description

Passive magnetic detection specific target identification method based on discrete Meyer wavelet
Technical Field
The invention belongs to a target identification technology, and particularly relates to a passive magnetic detection specific target identification method based on discrete Meyer wavelets.
Background
In an actual battlefield, most of attacking objects of weapons and ammunitions are ferromagnetic targets such as helicopters, tanks and armored vehicles, the battlefield environment is complex, and the traditional laser fuze is easily influenced by the ground, trees, cloud mist, smoke dust and the like and cannot accurately identify the ferromagnetic armored targets. Therefore, if the ferromagnetic armor target can be effectively detected and identified, the detection performance of the laser fuse can be greatly improved.
The earth is a huge magnetic field, a ferromagnetic armored target in a battlefield has higher magnetic conductivity, and can cause local geomagnetic field deflection, namely the intensity and the direction of the geomagnetic field near the ferromagnetic armored target can be distorted, and according to the characteristics, the ferromagnetic metal target in a certain range around the fuze can be effectively detected and identified.
Magnetic detection is classified into strong magnetic detection and weak magnetic detection, battlefield ferromagnetic target detection is typical weak magnetic detection, and the current weak magnetic detection methods can be classified into a total magnetic field strength detection method, a magnetic field scalar gradient detection method, a magnetic field vector gradient detection method, a magnetic field gradient tensor detection method and the like.
Sheinker A et al propose a magnetic anomaly signal detection method based on the combination of a fixed AR whitening filter and an orthogonal function OBF, but the method is only suitable for the detection of magnetic anomaly signals with power density of 1/fαThe orthogonal function also needs whitening filtering processing in order to prevent signal distortion. The method is complex and cannot realize quick and efficient judgment.
Ginzburg et al propose a gradient signal processing method for constructing MAD (magnetic analog detector) signal decomposition in an orthogonalized function space based on Gram-Schmidt algorithm, and express magnetic anomaly signals by five mutually independent orthogonal functions. The magnetic signal is directly analyzed, the signal is not preprocessed, the data is complex, the calculation amount is large, and quick and efficient judgment cannot be realized.
A triaxial magnetic difference signal detection processing circuit is adopted by Wangdeli, a Nanjing university of science and technology, and comprises a multistage magnetic signal amplification hardware circuit, a low-pass filtering hardware circuit and a precise full-wave rectification hardware circuit. And extracting three target identification characteristic values of the amplitude, the local change slope and the local delay time of the magnetic signal, and judging that a real target is detected when the characteristic values of the magnetic difference signals of more than two shafts reach a threshold value. The full-wave rectification hardware circuit reversely outputs a negative value signal in the magnetic signal into a positive value signal, so that vector information of the magnetic signal is lost, the extracted magnetic signal characteristics are all characteristics in a time domain, the signal characteristics cannot be analyzed from the angle of a frequency domain, and time domain similar signals cannot be distinguished. The Zhao Cheng lean et al of northwest industry university adopts a magnetic detection fuze target identification algorithm based on target detection of dynamic threshold and fuzzy inference. The algorithm acquires magnetic signals meeting requirements through a dynamic threshold, performs denoising processing on target signals by adopting a wavelet threshold method and a moving average filtering method, extracts peak values, valley values, magnetic field change frequency and magnetic gradients of the target signals to be detected, compares real target characteristic parameters, and discriminates target characteristic quantities by adopting a fuzzy inference algorithm. The dynamic threshold acquisition can cause the loss of signals and the loss of characteristic signals of partial frequency bands due to the sliding average filtering method, wherein the complete magnetic signals cannot be acquired. Under the condition that the signal is incomplete and the characteristic signal of a part of frequency bands is lost, the extracted magnetic signal characteristic has errors, and the identification precision is not high.
Disclosure of Invention
The invention aims to provide a method for identifying a specific target by passive magnetic detection based on Discrete Meyer Wavelet (DMEy), which can accurately detect whether a ferromagnetic armor target exists in a detection range.
The technical solution for realizing the purpose of the invention is as follows: a passive magnetic detection specific target identification method based on discrete Meyer wavelet is characterized in that:
step 1, electrifying a magnetic sensor to work, detecting a ferromagnetic armored target, and outputting magnetic signals of the magnetic field intensity in the directions of three axes x, y and z of a carrier coordinate system by the magnetic sensor on a carrier;
step 2, amplifying the triaxial magnetic signals after respectively performing DC blocking treatment, respectively performing low-pass filtering treatment on the amplified triaxial magnetic signals, filtering high-frequency noise signals, and cutting off frequency fcSetting according to an empirical value; the triaxial magnetic signals after the low-pass filtering processing are subjected to pull-up processing, and triaxial sampling signals x (t), y (t) and z (t) with the reference of 1.65V and the amplitude of 0V-3.3V are output;
step 3, respectively carrying out AD sampling on the three-axis magnetic signals x (t), y (t) and z (t) to obtain digital signals x (n), y (n) and z (n), respectively, carrying out dimensionality reduction on the three-axis characteristic signals of the ferromagnetic armor target by adopting a PCA algorithm according to the three-axis characteristic signals of the ferromagnetic armor target in a typical battlefield ferromagnetic armor target database, and correspondingly obtaining a dimensionality-reduced k-dimensional data matrix
Figure BDA0003231537600000021
And
Figure BDA0003231537600000022
step 4, according to the k-dimension data matrix after dimension reduction optimization
Figure BDA0003231537600000023
And
Figure BDA0003231537600000024
correspondingly acquiring characteristic signals of k frequency bands corresponding to x, y and z axes, and respectively setting k discrete Meyer wavelet transform scale functions corresponding to the x, y and z axes according to the characteristic signals of the k frequency bands of each axis; performing discrete Meyer wavelet transformation of k multi-scale functions on AD sampling signals X (n), y (n) and z (n) of the ferromagnetic armor target to obtain a matrix X after discrete wavelet transformation of the ferromagnetic armor targetMDey、YMDeyAnd ZMDey；
Step 5, according to the matrix X after the discrete wavelet change of the ferromagnetic armor targetMDey、YMDeyAnd ZMDeyPerforming bilateral filtering to obtain weight index, and obtaining decision weight index of each frequency band of x, y and z axes
Figure BDA0003231537600000031
Figure BDA0003231537600000032
And
Figure BDA0003231537600000033
step 6, according to the decision weight index of each frequency band of the x, y and z axes
Figure BDA0003231537600000034
And
Figure BDA0003231537600000035
a multi-signal fusion model is adopted to sum the weight indexes of the characteristic signals of different frequency bands to obtain the total decision indexes of the x axis, the y axis and the z axis which are respectively ex、ex、exThen e to ex、ex、exCarrying out summation operation to obtain a total decision index eall；
Step 7, the total judgment index eallAnd an empirical threshold UtvAnd comparing to judge whether a ferromagnetic armor target appears or not, specifically comprising the following steps:
when e isall≥UtvIn order to be in the effective detection rangeA ferromagnetic armor target is present inside; when e isall＜UtvIn time, it is intended that ferromagnetic armor targets do not appear within the effective detection range.
Compared with the prior art, the invention has the remarkable advantages that:
(1) the low-pass filter circuit is adopted to process the high-frequency noise signal, the data volume required to be processed during wavelet decomposition level and discrete wavelet transformation can be reduced, a PCA (Principal Component Analysis) algorithm is adopted to obtain an adaptive scale function according to the signal characteristics of different ferromagnetic armor targets in a typical battlefield ferromagnetic armor target database, and the number of scale functions during discrete Meyer wavelet decomposition is greatly reduced, so that the rapid digital magnetic signal data processing can be realized;
(2) the fast discrete wavelet transform can be realized by adopting a discrete Meyer wavelet function, compared with the traditional wavelet transform operation, the calculation speed is higher, and similar time domain signals can be distinguished from the frequency domain;
(3) the characteristic proportion function and the characteristic distinguishing function are used as weight distribution criteria of the bilateral filtering algorithm, and the characteristics of the ferromagnetic armor target in the time domain and the frequency domain are respectively considered, so that the ferromagnetic armor target is more accurately judged;
(4) a low-pass filter circuit is adopted to process high-frequency noise signals, a PCA algorithm is adopted to perform data dimension reduction filtering processing and a discrete Meyer wavelet is adopted to distinguish ferromagnetic armor target signals and noise signals from time domains and frequencies, so that noise interference can be effectively reduced, and the signal-to-noise ratio and the limit detection distance are improved.
Drawings
Fig. 1 is a flow chart of the method for identifying a specific target by passive magnetic detection based on discrete meier wavelet.
FIG. 2 is a flow chart of the present invention for processing digital magnetic signals in each axis (x-axis is taken as an example, and the processing steps in y-axis, z-axis and x-axis are the same).
Detailed Description
The present invention is described in further detail below with reference to the attached drawing figures.
With reference to fig. 1 and fig. 2, the method for identifying a specific target by passive magnetic detection based on discrete meier wavelet of the present invention includes the following steps:
step 1, electrifying a magnetic sensor to work, detecting a ferromagnetic armored target, and outputting magnetic signals of the magnetic field intensity in the directions of three axes x, y and z of a carrier coordinate system by the magnetic sensor on a carrier.
Step 2, amplifying the triaxial magnetic signals after respectively performing DC blocking treatment, respectively performing low-pass filtering treatment on the amplified triaxial magnetic signals, filtering high-frequency noise signals, and cutting off frequency fcSet according to empirical values. And (3) performing pull-up processing on the triaxial magnetic signal after the low-pass filtering processing, and outputting triaxial sampling signals x (t), y (t) and z (t) with the reference of 1.65V and the amplitude of 0V-3.3V.
The low-pass filter circuit is adopted to process the high-frequency noise signal, so that the data quantity required to be processed during wavelet decomposition level and discrete wavelet transformation can be reduced, and the digital signal processing speed is improved.
Wherein, the calculation formula of the magnification G is as follows:
G＝3.3/Outmax
3.3 supply Voltage, Out, for rail-to-rail amplification circuitsmaxThe maximum amplitude of the magnetic signal is output for the magnetic sensor.
Step 3, respectively carrying out AD sampling on the three-axis magnetic signals x (t), y (t) and z (t) to obtain digital signals x (n), y (n) and z (n), respectively, carrying out dimensionality reduction on the three-axis characteristic signals of the ferromagnetic armor target by adopting a PCA algorithm according to the three-axis characteristic signals of the ferromagnetic armor target in a typical battlefield ferromagnetic armor target database, and correspondingly obtaining a dimensionality-reduced k-dimensional data matrix
Figure BDA0003231537600000041
And
Figure BDA0003231537600000042
the PCA algorithm is adopted to obtain the self-adaptive scale function, so that the number of the scale functions in discrete Meyer wavelet decomposition is greatly reduced, and the rapid digital magnetic signal data processing can be realized.
Typical battlefield ferromagnetic armor targets, exemplified by the x-axisData matrix X of X-axis in databasetargetAs follows:
Figure BDA0003231537600000051
Xtargetthe row elements of (1) represent m data sampling time points, and the column elements represent from 0 to f at one sampling time pointcN frequencies, element xnmRepresenting the magnitude of the amplitude of the nth frequency at the mth time point.
Solving for XtargetCovariance matrix C of (a):
Figure BDA0003231537600000052
diagonalizing the covariance matrix C, and solving the eigenvector xi of CiAnd the feature vector xi is combinediAnd (3) sequencing according to the corresponding eigenvalues from large to small to obtain an eigenvector matrix P:
P＝[ξ1，…，ξn]T
selecting a data matrix X of which the matrix formed by the first k rows of the feature vector matrix P is multiplied by the X axistargetObtaining a k (k < n) dimensional matrix
Figure BDA0003231537600000053
Figure BDA0003231537600000054
Wherein λiFeature vector xi of C rearranged from large to smalliCorresponding characteristic value, λ1≥λ2≥…≥λk…＞＞λn。
The y-axis and the z-axis respectively correspond to the reduced k-dimension data matrix
Figure BDA0003231537600000055
The solving process is the same
Figure BDA0003231537600000056
Step 4, obtaining the k-dimension data matrix after dimension reduction and optimization according to the step 3
Figure BDA0003231537600000057
And
Figure BDA0003231537600000058
acquiring the characteristic signals of k frequency bands corresponding to x, y and z axes, and respectively setting k discrete Meyer wavelet transform scale functions corresponding to the x, y and z axes according to the characteristic signals of the k frequency bands of each axis. Performing discrete Meyer wavelet transformation of k multi-scale functions on AD sampling signals X (n), y (n) and z (n) of the ferromagnetic armor target to obtain a matrix X after discrete wavelet transformation of the ferromagnetic armor targetMDey、YMDeyAnd ZMDey。
The discrete Meyer wavelet function is adopted to realize fast discrete wavelet transform, compared with the traditional wavelet transform operation, the calculation speed is higher, and similar time domain signals can be distinguished from the aspect of frequency domain.
Taking the X axis as an example, after discrete Meyer wavelet transformation of k multi-scale functions is carried out on sampling signals X (n) of the ferromagnetic armor target, a matrix X of the ferromagnetic armor target after discrete wavelet transformation is obtainedMDey：
Figure BDA0003231537600000061
Figure BDA0003231537600000062
ωi∈{ω1，…，ωk}
ΨM(ωi) Is a discrete Meyer wavelet function; omega1～ωkFor the frequencies of the characteristic signals of the ferromagnetic armor target corresponding to the k frequency bands after dimensionality reduction, v (x) is an auxiliary function for constructing discrete Meyer wavelets, j isA complex number symbol;
Figure BDA0003231537600000063
matrix X of ferromagnetic armor targets after discrete wavelet transformMDeyThe k-th line of (a) represents the amplitude of the AD sampling signal x (n) on the x-axis at each sampling time corresponding to the discrete meier wavelet transform performed on the k-th scale function.
Matrix Y after discrete wavelet change corresponding to Y-axis and z-axisMDey、ZMDeyAnd XMDeyThe solving method is the same.
A typical battlefield ferromagnetic armor target database includes target magnetic signal data of various ferromagnetic armors (e.g., tanks, infantry combat vehicles, radar vehicles, armor scouts, transport vehicles, etc.), and noise magnetic signal data under environments such as vibration of a detection device carrier (e.g., unmanned aerial vehicles, helicopters, etc.), a detection device carrier interference field, and a geomagnetic background field.
Step 5, according to the matrix X after the discrete wavelet change of the ferromagnetic armor targetMDey、YMDeyAnd ZMDeyPerforming bilateral filtering to obtain weight index to obtain decision weight index of each frequency band of x, y and z axes
Figure BDA0003231537600000071
And
Figure BDA0003231537600000072
the characteristic proportion function and the characteristic distinguishing function are used as weight distribution criteria of the bilateral filtering algorithm, the characteristics of the ferromagnetic armor target in the time domain and the frequency domain are respectively considered, and the ferromagnetic armor target is more accurately judged.
Taking X-axis as an example, matrix X after discrete wavelet transformationMDeyThe weighting index is obtained by bilateral filtering:
Figure BDA0003231537600000073
wherein
Figure BDA0003231537600000074
The weight index of the ith frequency band characteristic signal (i is more than or equal to 0 and less than or equal to k) of the x-axis magnetic signal; i (x)i) The amplitude of the characteristic signal of the ith frequency band of the input magnetic signal; t (x)i) The characteristic signal duration of the ith frequency band of the input magnetic signal; t isi(x) The characteristic signal duration of the ith frequency band corresponding to the specific ferromagnetic armored target and the input magnetic signal in the typical battlefield ferromagnetic armored target database; x is the number ofiIs the characteristic signal of the ith frequency band of the input magnetic signal; wpIs a normalization factor; omega is a characteristic signal domain of the input magnetic signal; x is characteristic information of a particular ferromagnetic armor target in a typical battlefield ferromagnetic armor target database.
fr(||T(xi)-Ti(x) | | l)) is a feature proportion function, weights of proportion values are given according to different feature proportions of feature signals of different frequency bands in the whole ferromagnetic armor target magnetic signal, the larger the feature proportion is, the larger the distribution weight is, and a weight index distribution expression is as follows:
Figure BDA0003231537600000075
x′icharacteristic information of the ith frequency band of a specific ferromagnetic armor target in a typical battlefield ferromagnetic armor target database; sign (x'i) Is x'iThe sign function of (a); sign (x)i) Is x'iThe sign function of (2).
gs(||xi-x | |) is a characteristic distinguishing function, according to the intrinsic magnetic characteristic signal of the ferromagnetic armored target, the intrinsic invariant characteristic can distinguish the frequency band magnetic signal from the noise or time domain similar interference signal in the typical battlefield ferromagnetic armored target database, so that the frequency band signal is distributed with a larger weight, and the weight index distribution expression is as follows:
Figure BDA0003231537600000081
xnoisecharacteristic information of noise in a typical battlefield ferromagnetic armor target database; f (x)i) And (4) a judgment function for distinguishing the graduation of the ith frequency band signal of the ferromagnetic armored target from noise in a typical battlefield ferromagnetic armored target database.
Decision weight index of each frequency band of y and z axes
Figure BDA0003231537600000082
And
Figure BDA0003231537600000083
the solving method is the same.
6, obtaining the decision weight index of each frequency band of the x axis, the y axis and the z axis according to the step 5
Figure BDA0003231537600000084
And
Figure BDA0003231537600000085
a multi-signal fusion model is adopted to sum the weight indexes of the characteristic signals of different frequency bands to obtain the total decision indexes of the x axis, the y axis and the z axis which are respectively ex、ex、exThen e to ex、ex、exCarrying out summation operation to obtain a total decision index eall。
Wherein, the expression of the multi-signal fusion model is as follows,
Figure BDA0003231537600000086
step 7, obtaining the total decision index e in the step 6allAnd an empirical threshold UtvAnd comparing to judge whether a ferromagnetic armor target appears or not, specifically comprising the following steps:
when e isall≥UtvIn order to obtain ferromagnetism in the effective detection rangeAn armor target; when e isall＜UtvIn time, it is intended that ferromagnetic armor targets do not appear within the effective detection range.
Example 1
Taking a detection tank as an example, a three-axis magnetic sensor outputs a magnetic signal, performs amplification, filtering and pull-up processing, performs AD sampling, and converts an analog signal into a digital signal to obtain sampling signals x (n), y (n) and z (n). Performing dimensionality reduction on data by adopting PCA (principal component analysis) self-adaptive algorithm according to magnetic signal data of tanks in typical battlefield ferromagnetic armored target database under interference resistance to obtain k-dimensional data matrix
Figure BDA0003231537600000091
And
Figure BDA0003231537600000092
(account for 80% of the tank magnetic signal data information). k-dimensional data matrix
Figure BDA0003231537600000093
And
Figure BDA0003231537600000094
converting the corresponding frequency band into a scale function of discrete Meyer wavelet transform, and performing multi-window discrete Meyer wavelet transform on the input sampling signals X (n), y (n) and z (n) to obtain a matrix X after the discrete Meyer wavelet transformMDey、YMDeyAnd ZMDeyMatrix X after discrete Meyer wavelet transformMDey、YMDeyAnd ZMDeyAnd comparing the data with the data of the ferromagnetic armor targets and the noise in the typical battlefield ferromagnetic armor target database, and obtaining the weight index through bilateral filtering. Obtaining a total decision index e by the weight indexes of each frequency band and each axis through a multi-signal fusion modelallFinally, the total decision index eallAnd an empirical threshold UtvAnd comparing, and judging whether a tank appears in the effective detection range.

Claims (10)

1.一种基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：1. a passive magnetic detection specific target identification method based on discrete Meyer wavelet, is characterized in that: 步骤1、磁传感器上电工作，探测铁磁性装甲目标，载体上的磁传感器输出载体坐标系x、y、z三轴方向磁场强度的磁信号；Step 1. The magnetic sensor is powered on to detect the ferromagnetic armored target, and the magnetic sensor on the carrier outputs the magnetic signal of the magnetic field strength in the three-axis directions of the carrier coordinate system x, y, and z; 步骤2、对三轴磁信号分别做隔直处理后进行放大，将放大后的三轴磁信号分别进行低通滤波处理，滤除高频噪声信号，截止频率fc根据经验值设定；低通滤波处理后的三轴磁信号进行上拉处理，输出基准为1.65V，幅值为0V～3.3V的三轴采样信号x(t)、y(t)和z(t)；Step 2. Perform DC blocking processing on the three-axis magnetic signals respectively, and then amplify them. The amplified three-axis magnetic signals are respectively subjected to low-pass filtering processing to filter out high-frequency noise signals. The cut-off frequency f c is set according to the empirical value; The three-axis magnetic signal after filtering processing is pulled up, and the output reference is 1.65V, and the three-axis sampling signals x(t), y(t) and z(t) whose amplitude is 0V~3.3V; 步骤3、分别对三轴磁信号x(t)、y(t)和z(t)进行AD采样对应得到数字信号x(n)、y(n)和z(n)，根据典型战场铁磁性装甲目标数据库中铁磁性装甲目标分别对应的x、y、z三轴的特征信号，采用PCA算法对上述铁磁性装甲目标三轴的特征信号进行降维处理，对应获得降维后的k维数据矩阵
Figure FDA0003231537590000011
和
Figure FDA0003231537590000012
Step 3. Perform AD sampling on the three-axis magnetic signals x(t), y(t) and z(t) respectively to obtain digital signals x(n), y(n) and z(n). According to the typical battlefield ferromagnetism The characteristic signals of the ferromagnetic armored targets in the armored target database corresponding to the three axes of x, y, and z, respectively, and the PCA algorithm is used to perform dimension reduction processing on the three-axis characteristic signals of the ferromagnetic armored targets, and the corresponding k-dimensional data matrix after dimensionality reduction is obtained.
Figure FDA0003231537590000011
and
Figure FDA0003231537590000012
 步骤4、根据降维优化后的k维数据矩阵
Figure FDA0003231537590000013
和
Figure FDA0003231537590000014
对应获取x、y、z轴的对应k个频段的特征信号，根据各轴k个频段的特征信号分别设定x、y、z轴对应的k个的离散迈耶小波变换的尺度函数；对铁磁性装甲目标AD采样信号x(n)、y(n)和z(n)进行k个多尺度函数的离散迈耶小波变换，获得铁磁性装甲目标的离散小波变化后的矩阵XMDey、YMDey和ZMDey；Step 4. According to the optimized k-dimensional data matrix after dimensionality reduction
Figure FDA0003231537590000013
and
Figure FDA0003231537590000014
Correspondingly obtain the characteristic signals of the k frequency bands corresponding to the x, y, and z axes, and set the k discrete Meyer wavelet transform scaling functions corresponding to the x, y, and z axes according to the characteristic signals of the k frequency bands of each axis; The AD sampled signals x(n), y(n) and z(n) of the ferromagnetic armored target are subjected to discrete Meyer wavelet transform of k multi-scale functions to obtain the discrete wavelet transformed matrix X MDey , Y of the ferromagnetic armored target MDey and Z MDey ; 步骤5、根据铁磁性装甲目标的离散小波变化后的矩阵XMDey、YMDey和ZMDey进行双边滤波获得权重指数处理，获得x、y、z轴的各频段的判决权重指数
Figure FDA0003231537590000015
Figure FDA0003231537590000016
和
Figure FDA0003231537590000017
Step 5. According to the discrete wavelet-changed matrices X MDey , Y MDey and Z MDey of the ferromagnetic armored target, perform bilateral filtering to obtain the weight index processing, and obtain the decision weight index of each frequency band of the x, y, and z axes
Figure FDA0003231537590000015
Figure FDA0003231537590000016
and
Figure FDA0003231537590000017
 步骤6、根据x、y、z轴的各频段的判决权重指数
Figure FDA0003231537590000018
和
Figure FDA0003231537590000019
采用多信号融合模型，将不同频段特征信号的权重指数求和，获得总的x轴、y轴、z轴的判决指数分别为ex、ex、ex，再ex、ex、ex进行求和运算，获得总判决指数eall；Step 6. According to the decision weight index of each frequency band of the x, y and z axes
Figure FDA0003231537590000018
and
Figure FDA0003231537590000019
Using the multi-signal fusion model, the weight indices of the characteristic signals of different frequency bands are summed, and the total decision indices of the x-axis, y-axis, and z-axis are obtained as e x , e x , e x , and then e x , e x , e Perform a summation operation on x to obtain the total decision index e all ; 步骤7、将总判决指数eall与经验阈值Utv进行对比，判断是否出现铁磁性装甲目标，具体如下：Step 7. Compare the total judgment index e all with the experience threshold U tv to determine whether there is a ferromagnetic armored target, as follows: 当eall≥Utv时，为在有效探测范围内出现铁磁性装甲目标；当eall＜Utv时，为在有效探测范围内未出现铁磁性装甲目标。When e all ≥U tv , it means that there is a ferromagnetic armored target within the effective detection range; when e all <U tv , it means that there is no ferromagnetic armored target within the effective detection range. 2.根据权利要求1所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：步骤2中，对三轴磁信号分别做隔直处理后进行放大，放大倍数G的计算公式为：2. The method for identifying specific targets of passive magnetic detection based on discrete Meyer wavelets according to claim 1, is characterized in that: in step 2, the three-axis magnetic signal is amplified after DC blocking processing, and the calculation of the magnification G is performed. The formula is: G＝3.3/Outmax G=3.3/Out max 3.3为轨对轨放大电路的供电电压，Outmax为磁传感器输出磁信号的最大幅值。3.3 is the power supply voltage of the rail-to-rail amplifying circuit, and Out max is the maximum amplitude of the magnetic signal output by the magnetic sensor. 3.根据权利要求1所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：步骤3中，对x(t)进行AD采样得到数字信号x(n)，根据典型战场铁磁性装甲目标数据库中铁磁性装甲目标对应的x轴的特征信号，采用PCA算法对上述铁磁性装甲目标的特征信号进行降维处理，获得降维后的k维数据矩阵
Figure FDA0003231537590000021
具体如下：3. the passive magnetic detection specific target identification method based on discrete Meyer wavelet according to claim 1, is characterized in that: in step 3, carry out AD sampling to x (t) and obtain digital signal x (n), according to typical battlefield The characteristic signal of the x-axis corresponding to the ferromagnetic armored target in the ferromagnetic armored target database, the PCA algorithm is used to reduce the dimensionality of the characteristic signal of the ferromagnetic armored target, and the k-dimensional data matrix after dimensionality reduction is obtained.
Figure FDA0003231537590000021
details as follows: 典型战场铁磁性装甲目标数据库中x轴的数据矩阵Xtarget如下所示：The data matrix X target of the x-axis in the typical battlefield ferromagnetic armor target database is as follows:
Figure FDA0003231537590000022
Figure FDA0003231537590000022
 Xtarget的行元素代表m个数据采样时间点，列元素代表着一个采样时间点上从0～fc中的n个频率，元素xnm代表着第m个时间点第n个频率的幅值大小；The row elements of X target represent m data sampling time points, the column elements represent n frequencies from 0 to f c at a sampling time point, and the element x nm represents the amplitude of the nth frequency at the mth time point size; 求解Xtarget的协方差矩阵C：Solve the covariance matrix C of the X target :
Figure FDA0003231537590000023
Figure FDA0003231537590000023
 将协方差矩阵C对角化，求解C的特征向量ξi，并将特征向量ξi按照对应的特征值从大到小排序，得到特征向量矩阵P：Diagonalize the covariance matrix C, solve the eigenvector ξ i of C, and sort the eigenvector ξ i according to the corresponding eigenvalues from large to small to obtain the eigenvector matrix P: P＝[ξ1，…，ξn]T P=[ξ 1 ,...,ξ n ] T 选取特征向量矩阵P的前k行组成的矩阵乘以x轴的数据矩阵Xtarget，其中k＜n，得到k维矩阵
Figure FDA0003231537590000024
Select the matrix consisting of the first k rows of the eigenvector matrix P and multiply the x-axis data matrix X target , where k<n, to obtain a k-dimensional matrix
Figure FDA0003231537590000024
Figure FDA0003231537590000031
Figure FDA0003231537590000031
 λi为按照从大到小重新排序后的C的特征向量ξi对应的特征值，λ1≥λ2≥…≥λk…＞＞λn。λ i is the eigenvalue corresponding to the eigenvector ξ i of C reordered in descending order, λ 1 ≥λ 2 ≥...≥λ k ...>>λ n . 4.根据权利要求3所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：步骤3中，y轴、z轴分别对应的降维后的k维数据矩阵
Figure FDA0003231537590000032
求解过程同
Figure FDA0003231537590000033
4. The method for identifying a specific target of passive magnetic detection based on discrete Meyer wavelet according to claim 3, characterized in that: in step 3, the k-dimensional data matrix after dimension reduction corresponding to y-axis and z-axis respectively
Figure FDA0003231537590000032
The solution process is the same
Figure FDA0003231537590000033
 5.根据权利要求4所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：步骤4中，根据降维优化后的k维数据矩阵
Figure FDA0003231537590000034
获取x轴的对应k个频段的特征信号，根据x轴的k个频段的特征信号设定x轴对应的k个的离散迈耶小波变换的尺度函数；对铁磁性装甲目标AD采样信号x(n)进行k个多尺度函数的离散迈耶小波变换，获得铁磁性装甲目标的离散小波变化后的矩阵XMDey，具体如下：5. The method for identifying specific targets of passive magnetic detection based on discrete Meyer wavelets according to claim 4, characterized in that: in step 4, according to the k-dimensional data matrix after dimensionality reduction optimization
Figure FDA0003231537590000034
Obtain the characteristic signals corresponding to the k frequency bands of the x-axis, and set the scaling functions of the k discrete Meyer wavelet transforms corresponding to the x-axis according to the characteristic signals of the k frequency bands of the x-axis; for the ferromagnetic armored target AD sampling signal x( n) Perform the discrete Meyer wavelet transform of k multi-scale functions to obtain the discrete wavelet transformed matrix X MDey of the ferromagnetic armored target, as follows: 铁磁性装甲目标采样信号x(n)经过k个多尺度函数的离散迈耶小波变换后，得到铁磁性装甲目标的离散小波变化后的矩阵XMDey：After the sampling signal x(n) of the ferromagnetic armored target undergoes discrete Meyer wavelet transform of k multi-scale functions, the discrete wavelet transformed matrix X MDey of the ferromagnetic armored target is obtained:
Figure FDA0003231537590000035
Figure FDA0003231537590000035
Figure FDA0003231537590000036
Figure FDA0003231537590000036
 ωi∈{ω1，…，ωk}ω i ∈{ω 1 ,…,ω k } ΨM(ωi)为离散迈耶小波函数；ω1～ωk为铁磁性装甲目标的对应降维后的k个频段的特征信号的频率，v(x)为构造离散迈耶小波的辅助函数，j为复数符号；Ψ M (ω i ) is the discrete Meyer wavelet function; ω 1 ~ω k is the frequency of the characteristic signals of the k frequency bands after the dimension reduction of the ferromagnetic armored target, v(x) is the auxiliary for constructing the discrete Meyer wavelet function, j is a complex number symbol;
Figure FDA0003231537590000041
Figure FDA0003231537590000041
 铁磁性装甲目标的离散小波变化后的矩阵XMDey的第k行代表x轴的AD采样信号x(n)在第k个尺度函数下进行离散迈耶小波变换对应的频率信号在各个采样时间上的幅值。The discrete wavelet-transformed matrix X MDey of the ferromagnetic armored target The kth row of the x-axis represents the AD sampling signal x(n) of the x-axis. The discrete Meyer wavelet transform is performed under the kth scale function. The corresponding frequency signal is at each sampling time the magnitude of . 6.根据权利要求5所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：y轴、z轴对应的离散小波变化后的矩阵YMDey、ZMDey与XMDey的求解方式相同。6. the passive magnetic detection specific target identification method based on discrete Meyer wavelet according to claim 5, is characterized in that: the solution of matrix Y MDey , Z MDey and X MDey after the discrete wavelet change corresponding to y-axis, z-axis the same way. 7.根据权利要求5所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：典型战场铁磁性装甲目标数据库包括各种铁磁性装甲目标磁信号数据，以及探测装置载体的振动、探测装置载体干扰场和地磁背景场环境下的噪声磁信号数据。7. The method for identifying specific targets of passive magnetic detection based on discrete Meyer wavelets according to claim 5, characterized in that: the typical battlefield ferromagnetic armored target database includes various ferromagnetic armored target magnetic signal data, and the detection device carrier Noise magnetic signal data in vibration, detection device carrier interference field and geomagnetic background field environment. 8.根据权利要求6所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：步骤5中，根据铁磁性装甲目标的离散小波变化后的矩阵XMDey进行双边滤波获得权重指数处理获得x轴的各频段的判决权重指数
Figure FDA0003231537590000042
具体如下：8. the passive magnetic detection specific target identification method based on discrete Meyer wavelet according to claim 6, is characterized in that: in step 5, carry out bilateral filtering to obtain weight according to the matrix X MDey after the discrete wavelet change of ferromagnetic armored target Exponential processing to obtain the decision weight index of each frequency band of the x-axis
Figure FDA0003231537590000042
details as follows: 将离散小波变化后的矩阵XMDey通过双边滤波获得权重指数：The weight index is obtained by bilateral filtering of the matrix X MDey after the discrete wavelet change:
Figure FDA0003231537590000043
Figure FDA0003231537590000043
 其中
Figure FDA0003231537590000044
为x轴磁信号的第i个频段特征信号的权重指数，0≤i≤k；I(xi)为输入磁信号的第i个频段的特征信号的幅值；T(xi)为输入磁信号的第i个频段的特征信号持续时间；Ti(x)为典型战场铁磁性装甲目标数据库中特定铁磁性装甲目标与输入磁信号对应的第i个频段的特征信号持续时间；xi为输入磁信号的第i个频段的特征信号；Wp为正规化因子；Ω为输入磁信号的特征信号域；x为典型战场铁磁性装甲目标数据库中特定铁磁性装甲目标的特征信息；in
Figure FDA0003231537590000044
is the weight index of the characteristic signal of the ith frequency band of the x-axis magnetic signal, 0≤i≤k; I(x i ) is the amplitude of the characteristic signal of the ith frequency band of the input magnetic signal; T( xi ) is the input The characteristic signal duration of the ith frequency band of the magnetic signal; T i (x) is the characteristic signal duration of the ith frequency band corresponding to the specific ferromagnetic armored target in the typical battlefield ferromagnetic armored target database and the input magnetic signal; xi is the characteristic signal of the ith frequency band of the input magnetic signal; W p is the normalization factor; Ω is the characteristic signal domain of the input magnetic signal; x is the characteristic information of the specific ferromagnetic armored target in the typical battlefield ferromagnetic armored target database; fr(||T(xi)-Ti(x)||)为特征占比函数，根据不同频段的特征信号在整个铁磁性装甲目标磁信号中的特征占比的不同，赋予所占比值的权重，特征占比越大，则分配权重越大，权重指数分配表达式如下：f r (||T(x i )-T i (x)||) is the feature ratio function. According to the different feature ratios of the feature signals of different frequency bands in the magnetic signal of the entire ferromagnetic armored target, the proportion of The weight of the ratio, the larger the feature proportion, the greater the distribution weight, the weight index distribution expression is as follows:
Figure FDA0003231537590000051
Figure FDA0003231537590000051
 x′i为典型战场铁磁性装甲目标数据库中特定铁磁性装甲目标的第i个频段的特征信息；sign(x′i)为x′i的符号函数；sign(xi)为x′i的符号函数；x′ i is the characteristic information of the i-th frequency band of a specific ferromagnetic armor target in the typical battlefield ferromagnetic armor target database; sign(x′ i ) is the sign function of x′ i ; sign(x i ) is the x′ i symbolic function; gs(||xi-x||)为特征区分函数，根据铁磁性装甲目标固有磁特征信号，此固有不变的特征可将该频段磁信号与典型战场铁磁性装甲目标数据库中所述噪声或者时域的相似干扰信号进行区分，因此该频段信号分配权重更大，权重指数分配表达式如下：g s (||x i -x||) is the feature discrimination function. According to the inherent magnetic characteristic signal of the ferromagnetic armored target, this inherently invariable characteristic can compare the magnetic signal of this frequency band with that described in the typical battlefield ferromagnetic armored target database. Noise or similar interference signals in the time domain are distinguished, so the signal in this frequency band is assigned a greater weight, and the weight index assignment expression is as follows:
Figure FDA0003231537590000052
Figure FDA0003231537590000052
 xnoise为典型战场铁磁性装甲目标数据库中噪声的特征信息；f(xi)为铁磁性装甲目标第i个频段信号与典型战场铁磁性装甲目标数据库中噪声区分度的判断函数。x noise is the characteristic information of the noise in the typical battlefield ferromagnetic armored target database; f(x i ) is the judgment function of the ith frequency band signal of the ferromagnetic armored target and the noise in the typical battlefield ferromagnetic armored target database. 9.根据权利要求8所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于：y、z轴的各频段的判决权重指数
Figure FDA0003231537590000061
与
Figure FDA0003231537590000062
的求解方式相同。9. The method for identifying specific targets of passive magnetic detection based on discrete Meyer wavelets according to claim 8, wherein: the decision weight index of each frequency band of the y and z axes
Figure FDA0003231537590000061
and
Figure FDA0003231537590000062
is solved in the same way. 10.根据权利要求8所述的基于离散迈耶小波的被动磁探测特定目标识别方法，其特征在于，步骤6中，多信号融合模型表达式如下：10. The method for identifying specific targets of passive magnetic detection based on discrete Meyer wavelets according to claim 8, characterized in that, in step 6, the multi-signal fusion model expression is as follows:
Figure FDA0003231537590000063
Figure FDA0003231537590000063
CN202110988556.6A 2021-08-26 2021-08-26 Passive magnetic detection specific target identification method based on discrete Meyer wavelet Active CN113848589B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110988556.6A CN113848589B (en) 2021-08-26 2021-08-26 Passive magnetic detection specific target identification method based on discrete Meyer wavelet

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110988556.6A CN113848589B (en) 2021-08-26 2021-08-26 Passive magnetic detection specific target identification method based on discrete Meyer wavelet

Publications (2)

Publication Number Publication Date
CN113848589A true CN113848589A (en) 2021-12-28
CN113848589B CN113848589B (en) 2024-12-17

Family

ID=78976135

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110988556.6A Active CN113848589B (en) 2021-08-26 2021-08-26 Passive magnetic detection specific target identification method based on discrete Meyer wavelet

Country Status (1)

Country Link
CN (1) CN113848589B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114858004A (en) * 2022-04-27 2022-08-05 西安工业大学 Detonation control method for anti-unmanned aerial vehicle based on weak magnetic detection technology

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6208951B1 (en) * 1998-05-15 2001-03-27 Council Of Scientific & Industrial Research Method and an apparatus for the identification and/or separation of complex composite signals into its deterministic and noisy components
CN102679980A (en) * 2011-11-15 2012-09-19 哈尔滨工程大学 Target tracking method based on multi-scale dimensional decomposition
CN103837889A (en) * 2013-11-29 2014-06-04 成都理工大学 Seismic attribute multi-source information fusion method
CN108256488A (en) * 2018-01-19 2018-07-06 中国人民解放军陆军装甲兵学院 A kind of radar target identification method based on micro-Doppler feature extraction and deep learning
CN109100809A (en) * 2018-06-29 2018-12-28 中国科学院上海光学精密机械研究所 Weak magnetic signal noise suppressed and signal extracting device and method under earth magnetism background

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6208951B1 (en) * 1998-05-15 2001-03-27 Council Of Scientific & Industrial Research Method and an apparatus for the identification and/or separation of complex composite signals into its deterministic and noisy components
CN102679980A (en) * 2011-11-15 2012-09-19 哈尔滨工程大学 Target tracking method based on multi-scale dimensional decomposition
CN103837889A (en) * 2013-11-29 2014-06-04 成都理工大学 Seismic attribute multi-source information fusion method
CN108256488A (en) * 2018-01-19 2018-07-06 中国人民解放军陆军装甲兵学院 A kind of radar target identification method based on micro-Doppler feature extraction and deep learning
CN109100809A (en) * 2018-06-29 2018-12-28 中国科学院上海光学精密机械研究所 Weak magnetic signal noise suppressed and signal extracting device and method under earth magnetism background

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陈德宏;徐舒;沈浩;: "一种战场声目标识别的多特征提取算法", 安徽工业大学学报(自然科学版), no. 02, 15 April 2017 (2017-04-15) *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114858004A (en) * 2022-04-27 2022-08-05 西安工业大学 Detonation control method for anti-unmanned aerial vehicle based on weak magnetic detection technology
CN114858004B (en) * 2022-04-27 2023-07-28 西安工业大学 An anti-drone detonation control method based on weak magnetic detection technology

Also Published As

Publication number Publication date
CN113848589B (en) 2024-12-17

Similar Documents

Publication Publication Date Title
Fan et al. Magnetic anomaly detection using one-dimensional convolutional neural network with multi-feature fusion
CN102437984A (en) Modulation Signal Recognition Method Based on Complexity Features under Low SNR
CN113791449B (en) Aviation magnetic anomaly OBF decomposition detection method based on support vector machine
CN105652325A (en) Exponential fit-adaptive Kalman-based ground-air electromagnetic data de-noising method
Qiao et al. Multi-scale eigenvalues Empirical Mode Decomposition for geomagnetic signal filtering
CN110443139B (en) Hyperspectral remote sensing image noise band detection method for classification
Wang et al. Frequency characteristics analysis for magnetic anomaly detection
Riggs et al. Identification of metallic mine-like objects using low frequency magnetic fields
Xu et al. Magnetic anomaly detection using multifeature fusion-based neural network
CN112364768A (en) Distributed optical fiber intrusion identification method based on airspace features and machine learning
CN113848589B (en) Passive magnetic detection specific target identification method based on discrete Meyer wavelet
CN112633147A (en) Magnetic anomaly detection method based on multi-feature fusion and intelligent iterative optimization algorithm
CN116304561B (en) Ocean target magnetic anomaly signal denoising and detecting method based on deep migration learning
Zhou et al. Aeromagnetic anomaly detection under low SNR conditions using multiscale wavelet energy accumulation
CN112069987B (en) Interference type automatic identification method based on statistical manifold optimization dimension reduction
CN114371474A (en) Intelligent radar signal sorting method and system based on convolution noise reduction self-encoder
CN116894223B (en) Airborne underwater anomaly detection method based on self-adaptive cancellation and ResNet neural network
Dula et al. Landmine classification using possibilistic k-nearest neighbors with wideband electromagnetic induction data
Ren et al. Active jamming signal recognition based on residual neural network
CN117579216B (en) UAV communication jamming task generation method based on deep learning
Kim et al. CNN-based UAV detection with short time fourier transformed acoustic features
Adler-Golden Improved hyperspectral anomaly detection in heavy-tailed backgrounds
CN116192292B (en) A UAV rapid identification system and method based on two-dimensional code mapping
CN115545080B (en) Method and device for on-line detection of lubricating oil debris
Du et al. Magnetic anomaly detection based on singular spectrum analysis and orthonormal basis functions

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant