CN107132518B - A kind of range extension target detection method based on rarefaction representation and time-frequency characteristics - Google Patents
A kind of range extension target detection method based on rarefaction representation and time-frequency characteristics Download PDFInfo
- Publication number
- CN107132518B CN107132518B CN201710422876.9A CN201710422876A CN107132518B CN 107132518 B CN107132518 B CN 107132518B CN 201710422876 A CN201710422876 A CN 201710422876A CN 107132518 B CN107132518 B CN 107132518B
- Authority
- CN
- China
- Prior art keywords
- signal
- echo
- target
- range extension
- extension 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/414—Discriminating targets with respect to background clutter
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
The range extension target detection method based on rarefaction representation and time-frequency characteristics that the present invention provides a kind of, sampling processing is carried out to the echo-signal for removing tiltedly treated wideband radar, one-dimensional range profile is reconstructed by CAMP algorithm using sparse representation theory, obtain the most strong scattering point of range extension target, and the base as corresponding to most strong scattering point and the echo-signal after denoising construct quasi- ambiguity function, the range extension target constant false alarm detector based on quasi- ambiguity function is built, realizes the detection to high-speed maneuver target.The present invention can effectively realize target detection, improve detection performance.
Description
Technical field
The invention belongs to Radar Technology fields, are related to one kind range extension target detection side under white Gaussian noise background
Method.
Background technique
Under white Gaussian noise background, the detection of range extension target is a research hotspot in field of signal processing.
Currently, the range extension target detection device that scholars propose can probably be divided into following a few classes: one, Gerlach in 1997 and
The generalized likelihood test device (SSD-GLRT) based on space density that Steiner is proposed, it is right after being denoised by Nonlinear Mapping
The one-dimensional range profile of range extension target in distance to being integrated, to realize target detection.Although such detector operation
Measure small, but its detection performance is bad；Two, the range extension target detection based on waveform entropy for the propositions such as 2011 Nian Shuipeng are bright is led to
The arithmetic mean of instantaneous value for calculating multiple continuous range extension target one-dimensional range profile waveform entropies is crossed, and then realizes target detection.Though
Such right detector has preferable detection performance, but it has ignored Range cell migration and targeted attitude sensibility；Three, 2013
The maneuvering distance extension target detection device based on dipulse Time-frequency Decomposition characteristic year proposed, by the decomposition and synthesis of signal,
Several maximum singular values are chosen to realize target detection.Although such detector has constant false alarm characteristic and can detecte high speed machine
Moving-target, but its operand is larger, and with the increase for choosing singular value number, detection performance can decline.
Summary of the invention
For overcome the deficiencies in the prior art, the present invention provides a kind of extended distance based on rarefaction representation and time-frequency characteristics
Rarefaction representation is introduced into the range extension target detection field under white Gaussian noise background, using multiple list by object detection method
Frequency signal base carries out rarefaction representation to the wideband radar echo-signal after going tiltedly, by multiple approximate information approximate algorithm (CAMP)
Reconstruct obtains the one-dimensional range profile of range extension target from the echo-signal polluted by white Gaussian noise, meanwhile, by most dissipating by force
Base corresponding to exit point and the echo-signal after denoising obtain target signature by constructing quasi- ambiguity function, and by being based on quasi-mode
The range extension target constant false alarm detector for pasting function realizes the detection of target.
The technical solution adopted by the present invention to solve the technical problems the following steps are included:
(1) treated that wideband radar echo-signal is sampled to going tiltedly, the echo-signal r after being sampled, simultaneously
The corresponding multiple simple signal base of constructionIn formula, fsIndicate sampling frequency
Rate, T0Emit the pulse width of signal, n=0 ..., N-1, N=f for radar systemsT0, K=fs；Then it is asked with CAMP algorithm
Solution obtains the one-dimensional range profile s' of range extension targetHRRP, using the maximum value d of one-dimensional range profile as the most strong scattering of target
Point obtains the base corresponding to itIn formula, γmaxFor ΦDdFrequency；
(2) the one-dimensional range profile s' by reconstructingHRRPIt is calculated with Inverse Fast Fourier TransformsThen
The noise power of r' is estimated with MAD algorithm
(3) false-alarm probability P is determined according to settingF, warpSecondary Monte Carol experimental calculation obtains detection threshold η, is used for
Determine that target whether there is；
(4) by the corresponding base Φ of most strong scattering pointDdWith normalized reconstruct echo-signalCalculate quasi- ambiguity functionIn formula, θ indicates frequency, and ξ indicates time delay,It indicates altogether
Yoke transposition；
(5) selected characteristic TF obtains detection featureFoundationMake
Determine out, determines whether that there are targets, wherein H1Indicate goal hypothesis, H0Indicate no goal hypothesis.
The beneficial effects of the present invention are:
(1) time-frequency characteristics that the present invention chooses can effectively realize target detection；
(2) sparse representation theory is introduced into object detection field by the present invention, goes to improve with the noise removal capability of restructing algorithm
The detection performance of detector；
(3) the multiple simple signal base that the present invention is constructed is unrelated with target velocity, and the detector of proposition can detecte high speed
Maneuvering target and the targe-aspect sensitivity of target are unrelated.
Detailed description of the invention
Fig. 1 is the range extension target constant false alarm detector flow chart in the present invention based on rarefaction representation and time-frequency characteristics；
Fig. 2 is the one-dimensional range profile of echo-signal, and the sparsity for echo-signal is verified；
Fig. 3 is the statistical property schematic diagram of CAMP algorithm reconstructed error；
Fig. 4 is the detection performance schematic diagram of target time-frequency characteristics of the echo-signal based on quasi- ambiguity function.
Specific embodiment
Present invention will be further explained below with reference to the attached drawings and examples, and the present invention includes but are not limited to following implementations
Example.
Sparse representation theory is introduced into object detection field by the present invention, improves detector with the noise removal capability of restructing algorithm
Detection performance, can be used for the detection of high-speed maneuver target.
Realizing the object of the invention key problem in technology is carried out at sampling to the echo-signal for removing tiltedly treated wideband radar
Reason reconstructs one-dimensional range profile by CAMP algorithm using sparse representation theory, obtains the most strong scattering point of range extension target (by institute
Some basis representation in the multiple simple signal base of construction), and the base as corresponding to most strong scattering point and the echo-signal after denoising
Quasi- ambiguity function is constructed, the range extension target constant false alarm detector based on quasi- ambiguity function is built, is realized to high-speed maneuver mesh
Target detection, specific steps include the following:
(1) to go tiltedly treated wideband radar echo-signal carry out sampling processing (sample frequency need to meet Nyquist
Sampling thheorem), the echo-signal r after being sampled, while constructing corresponding multiple simple signal base ΦD(discrete fourier becomes
Change matrix), it is then solved to obtain the one-dimensional range profile s' of range extension target with CAMP algorithmHRRP, obtain the most strong of target and dissipate
Exit point (maximum value of one-dimensional range profile) and corresponding base ΦDd；
(2) the one-dimensional range profile s' by reconstructingHRRPIt is calculated with Inverse Fast Fourier TransformsThen
The noise power of r' is estimated with MAD algorithm
(3) it under the conditions of given false-alarm probability, obtains determining the inspection that target whether there is through Monte Carol experimental calculation
Survey thresholding η；
(4) by the corresponding base Φ of most strong scattering pointDdWith the echo-signal r of normalized reconstructnCalculate quasi- ambiguity function FT
(θ,ξ)；
(5) selected characteristic TF, and determine that target whether there is.
With reference to Fig. 1, the implementation process of the embodiment of the present invention is as follows:
Process 1, to go tiltedly treated wideband radar echo-signal carry out sampling processing, obtain echo-signal r.
(1.1) to obtained echo-signal r, the multiple simple signal base Φ corresponding to it is constructedD, it is indicated by following formula:
In formula, fsIndicate sample frequency, T0Emit the pulse width of signal, n=0 ..., N-1, N=for radar system
fsT0, k=0 ..., K-1, K=fs。
(1.2) rarefaction representation is carried out to echo-signal with multiple simple signal base:
sHRRP=ΦDr
Process 2 is solved to obtain the one-dimensional range profile s' of range extension target with CAMP algorithmHRRP, obtain the most strong of target
Scattering point d=max (s'HRRP), corresponding base ΦDd, by MAD algorithm estimating noise power
(2.1) following formula is solved with CAMP algorithm obtain the one-dimensional range profile s' of range extension targetHRRP:
In formula, w is zero-mean, and noise power isWhite complex gaussian noise, y be white complex gaussian noise pollution echo
Signal,For ΦDInverse matrix (inverse discrete Fourier transform matrix).
(2.2) the one-dimensional range profile s' reconstructedHRRPMaximum value be target most strong scattering point d=max (s'HRRP), institute
Corresponding base is ΦDd, may be expressed as:
In formula, γmaxFor ΦDdFrequency, sdIt (n) is base vector corresponding to most strong scattering point.
(2.3) the one-dimensional range profile s' by reconstructingHRRPIt is calculated with inverse Fourier transformThen it uses
MAD algorithm estimates the noise power of r'Wherein estimated by MAD algorithmProcess, can be obtained by following formula:
In formula, R () indicates to take the real part of vector, and I () indicates to take the imaginary part of vector, and median indicates to calculate vector
Intermediate value.
Process 3, the noise power estimated by MADAfterwards, it tests to obtain detection threshold η by Monte Carol.
(3.1) false-alarm probability P is setF(PF< < 1), obtaining Monte Carol experiment number is
(3.2) by obtaining the corresponding base of most strong scattering point in process 2Pass through following formula
Calculate sd(n) and the quasi- ambiguity function of white complex gaussian noise w (n):
In formula, θ indicates frequency, and ξ indicates time delay,Indicate conjugate transposition.
(3.3) θ=γ is chosenmax, η is obtained by following formula1
η1=∑ | FT(θ,ξ)|2
(3.4) 3.2 and 3.3 are repeated, number isIt obtains
Process 4 calculates the receives echo-signal r' and s of reconstruct by the corresponding base of most strong scattering pointd(n) quasi-mode pastes letter
Number.
(4.1) noise power estimated by process 3R' is normalized:
(4.2) base vector s corresponding to most strong scattering point is obtained as process 2d(n), s is calculated by following formulad(n) and
rnQuasi- ambiguity function:
In formula, θ indicates frequency, and ξ indicates time delay,Indicate conjugate transposition.
Process 5, selected characteristic TF, and determine that target whether there is.
(5.1) selected characteristic TF, the F that 4.2 in process 4 are obtainedT(θ, ξ), selecting frequency point γmaxEnergy and,
Obtain detection feature:
(5.2) the detection feature being calculated by 5.1Detection threshold is obtained with process 3
It compares.FoundationIt decisions making, determines whether that there are targets, wherein H1Indicate goal hypothesis, H0It indicates
There is no goal hypothesis.
Advantages of the present invention can be further illustrated by following emulation experiment.
1, experiment parameter and experiment condition
The one of the measured data to airflight Yark-42 aircraft that the selected data of experiment are enrolled for certain research institute
Part, as shown in Table 1 in relation to wideband radar and aircraft parameter:
The parameter of table 1 wideband radar and aircraft
2, experiment content and interpretation of result
A. echo-signal can be by multiple simple signal base rarefaction representation.Simulation result is as shown in Figure 2, wherein abscissa table
Show that distance unit number, ordinate indicate normalization amplitude and relative error.When Fig. 2 (a) is noiseless, by Fast Fourier Transform (FFT)
The one-dimensional range profile of the 10th echo-signal of the selection obtained with CAMP algorithm, shown in relative error such as Fig. 2 (b), the order of magnitude
It is 10-14；White complex gaussian noise is added in the 100th echo-signal that Fig. 2 (c) show selection, when signal-to-noise ratio is 6dB, by CAMP
The one-dimensional range profile that algorithm obtains, Fig. 2 (d) are to be obtained one-dimensional range profile when signal-to-noise ratio is 6dB by CAMP algorithm and be not present
The relative error of one-dimensional range profile, the order of magnitude 10 are obtained by Fast Fourier Transform (FFT) when noise-2.The experimental results showed that CAMP
Algorithm can from noise restructuring distance extension target one-dimensional range profile.
The reconstructed error approximation Gaussian distributed of B.CAMP algorithm.The 2000th echo-signal is chosen as experiment sample
This, is added white complex gaussian noise, analyzes the statistical property of reconstructed error, when signal-to-noise ratio is 6dB, the reconstruct of CAMP algorithm
The statistical property of error is as shown in Figure 3.Fig. 3 (a) and 3 (b) is respectively that 10000 Monte-Carlo of real and imaginary parts are tested
The statistical property of the reconstructed error arrived, wherein straight line indicates Gaussian probability-density function.It, which is used to approach, measures reconstructed error
Probability density function.It can be seen from the figure that the reconstructed error approximation Gaussian distributed of CAMP algorithm.
C. detector detection performance proposed by the invention is better than traditional range extension target detection device.Choose the 5000th
Secondary echo-signal verifies the target time-frequency characteristics proposed by the present invention based on quasi- ambiguity function as experiment sample, and chooses
Continuous 5000 echo-signals verify the detection performance of constant false alarm detector proposed by the present invention, and experimental result is as shown in Figure 4.
When Fig. 4 (a) is noiseless, the quasi- ambiguity function of multiple simple signal base corresponding to the 5000th echo-signal and most strong scattering point
Energy profile, it is easy to find out, energy focuses primarily upon near most strong scattering point.Fig. 4 (b) be white complex gaussian noise and
The energy profile of the quasi- ambiguity function of simple signal base corresponding to most strong scattering point, it is easy to find out, energy is uniformly to divide
Cloth is in entire fuzzy field.Fig. 4 (c) is when signal-to-noise ratio is 6dB, and CAMP algorithm reconstructs to obtain echo-signal and most strong scattering point
The quasi-mode of corresponding multiple simple signal base pastes function energy distribution map, it is easy to find out as Fig. 4 (a), energy mainly collects
In near most strong scattering point, and its signal-to-noise ratio increases.Fig. 4 (d) also shows CFAR detection proposed by the present invention
Device and SSD-GLRT detector and based on the range extension target detection device of Time-frequency Decomposition characteristic set false-alarm probability as
0.001, detection performance comparable situation when choosing continuous 5000 echo-signals, as shown in Fig. 4 (d), it can be seen that institute of the present invention
The detector detection performance of proposition is better than traditional range extension target detection device.
Claims (1)
1. a kind of range extension target detection method based on rarefaction representation and time-frequency characteristics, it is characterised in that including following steps
It is rapid:
(1) treated that wideband radar echo-signal is sampled to going tiltedly, and the echo-signal r after being sampled is constructed simultaneously
Corresponding multiple simple signal baseIn formula, fsIndicate sample frequency,
T0Emit the pulse width of signal, n=0 ..., N-1, N=f for radar systemsT0, K=fs；Then it is solved with CAMP algorithm
To the one-dimensional range profile s' of range extension targetHRRP, most strong scattering point using the maximum value d of one-dimensional range profile as target obtains
To the base corresponding to itIn formula, γmaxFor ΦDdFrequency；
(2) the one-dimensional range profile s' by reconstructingHRRPIt is calculated with Inverse Fast Fourier TransformsThen MAD is used
Algorithm estimates the noise power of r'
(3) false-alarm probability P is determined according to settingF, warpSecondary Monte Carol experimental calculation obtains detection threshold η, for determining
Target whether there is；
(4) by the corresponding base Φ of most strong scattering pointDdWith normalized reconstruct echo-signalCalculate quasi- ambiguity functionIn formula, θ indicates frequency, and ξ indicates time delay,It indicates altogether
Yoke transposition；
(5) selected characteristic TF obtains detection featureFoundationIt makes and sentencing
It is fixed, determine whether that there are targets, wherein H1Indicate goal hypothesis, H0Indicate no goal hypothesis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710422876.9A CN107132518B (en) | 2017-06-07 | 2017-06-07 | A kind of range extension target detection method based on rarefaction representation and time-frequency characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710422876.9A CN107132518B (en) | 2017-06-07 | 2017-06-07 | A kind of range extension target detection method based on rarefaction representation and time-frequency characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107132518A CN107132518A (en) | 2017-09-05 |
| CN107132518B true CN107132518B (en) | 2019-11-22 |
Family
ID=59733781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710422876.9A Active CN107132518B (en) | 2017-06-07 | 2017-06-07 | A kind of range extension target detection method based on rarefaction representation and time-frequency characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107132518B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108562897B (en) * | 2018-01-26 | 2022-01-11 | 桂林电子科技大学 | Structure sparse imaging method and device of MIMO through-wall radar |
| CN111679264B (en) * | 2020-06-19 | 2023-09-15 | 中国人民解放军空军工程大学 | Intermittent sampling forwarding interference suppression method, device and equipment |
| CN112731388B (en) * | 2020-12-14 | 2023-10-13 | 北京遥感设备研究所 | Target detection method based on effective scattering point energy accumulation |
| CN114492505B (en) * | 2021-12-24 | 2023-05-30 | 西安电子科技大学 | Air group target and extension target identification method based on semi-actual measurement data |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8068385B2 (en) * | 2008-11-24 | 2011-11-29 | Raytheon Company | System and method for enhancing weak target signals for a sensor array |
| CN103885058A (en) * | 2014-02-26 | 2014-06-25 | 西安电子科技大学 | Airborne radar foresight super-resolution imaging method utilizing sparse statistical property |
| CN104122540A (en) * | 2014-07-31 | 2014-10-29 | 电子科技大学 | Matching dictionary and compressive sensing based radar range profile object identification method |
| CN106059971A (en) * | 2016-07-07 | 2016-10-26 | 西北工业大学 | Sparse reconstruction based correlation detection method under signal correlation attenuation condition |
| CN108919263A (en) * | 2018-05-14 | 2018-11-30 | 西安电子科技大学 | ISAR high-resolution imaging method based on maximum mutual information criterion |
-
2017
- 2017-06-07 CN CN201710422876.9A patent/CN107132518B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8068385B2 (en) * | 2008-11-24 | 2011-11-29 | Raytheon Company | System and method for enhancing weak target signals for a sensor array |
| CN103885058A (en) * | 2014-02-26 | 2014-06-25 | 西安电子科技大学 | Airborne radar foresight super-resolution imaging method utilizing sparse statistical property |
| CN104122540A (en) * | 2014-07-31 | 2014-10-29 | 电子科技大学 | Matching dictionary and compressive sensing based radar range profile object identification method |
| CN106059971A (en) * | 2016-07-07 | 2016-10-26 | 西北工业大学 | Sparse reconstruction based correlation detection method under signal correlation attenuation condition |
| CN108919263A (en) * | 2018-05-14 | 2018-11-30 | 西安电子科技大学 | ISAR high-resolution imaging method based on maximum mutual information criterion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107132518A (en) | 2017-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107132518B (en) | A kind of range extension target detection method based on rarefaction representation and time-frequency characteristics | |
| CN105137498B (en) | A kind of the underground objects detection identifying system and method for feature based fusion | |
| CN104569948B (en) | Sub-band adaptive GLRT LTD detection methods under sea clutter background | |
| Guan et al. | Adaptive fractional Fourier transform-based detection algorithm for moving target in heavy sea clutter | |
| CN106569193B (en) | The small targets detection in sea clutter method filtered based on anterior-posterior to income reference particle | |
| Luo et al. | The fractal properties of sea clutter and their applications in maritime target detection | |
| CN104007434B (en) | The detection method of radar moving targets under sea clutter background based on Doppler's over-sampling | |
| Chen et al. | A fast FRFT based detection algorithm of multiple moving targets in sea clutter | |
| CN104155650A (en) | Object tracking method based on trace point quality evaluation by entropy weight method | |
| CN112462355B (en) | Intelligent sea target detection method based on time-frequency three-feature extraction | |
| Li et al. | Radar signal recognition algorithm based on entropy theory | |
| Chen et al. | Robust target identification using a modified generalized likelihood ratio test | |
| Hu et al. | A new way to model nonstationary sea clutter | |
| Hamdollahzadeh et al. | Moving target localization in bistatic forward scatter radars: Performance study and efficient estimators | |
| CN114117912A (en) | Sea clutter modeling and inhibiting method under data model dual drive | |
| CN112731331B (en) | Micro-motion target noise steady identification method based on signal-to-noise ratio adaptive network | |
| Jian et al. | Fractal Poisson model for target detection within spiky sea clutter | |
| Zhang et al. | Time reversal detection in multistatic radar system | |
| Hua et al. | Target detection in sea clutter via weighted averaging filter on the Riemannian manifold | |
| Karabayır et al. | Synthetic-Range-Profile-Based training library construction for ship target recognition purposes of scanning radar systems | |
| Tian et al. | Multiresolution jamming recognition with few-shot learning | |
| CN106019250A (en) | Repeating false target discriminating method based on angular glint | |
| CN105548986B (en) | Object detection method based on prewhitening median ratio detector under sea clutter background | |
| Zhang et al. | A new detecting method for weak targets in sea clutter based on multifractal properties | |
| CN111144316B (en) | Sea surface target detection method based on stack self-encoder |
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 |