CN107942324A - Multi-frame joint Small object double check method based on Doppler's guiding - Google Patents
Multi-frame joint Small object double check method based on Doppler's guiding Download PDFInfo
- Publication number
- CN107942324A CN107942324A CN201711075845.7A CN201711075845A CN107942324A CN 107942324 A CN107942324 A CN 107942324A CN 201711075845 A CN201711075845 A CN 201711075845A CN 107942324 A CN107942324 A CN 107942324A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- detection
- matrix
- msup
- 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
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
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a kind of multi-frame joint Small object double check method based on Doppler's guiding, mainly solves the problems, such as that the prior art is not suitable for detecting sea surface small target under high resolution mode.Implementation step is：1) echo data is obtained, and echo data is tieed up into progress Coherent processing along pulse and obtains data matrix z；2) the detection statistics moment matrix ξ of Generalized Likelihood Ratio threshold detector is calculated using data matrix z1With Doppler's access matrix3) the first re-detection thresholding T is calculated1And reset ξ1With4) Doppler's guidance information is utilizedDesign direction matched filter；5) utilization orientation matched filter and the data matrix ξ reset1', the test statistics ξ of unit d to be detected is calculated, and calculate the second re-detection thresholding T；6) judge that target whether there is by the size of comparing check statistic ξ and the second re-detection thresholding T.The present invention improves Studies of Radar Detection performance, reduces false-alarm probability, available for detecting sea surface small target.
Description
Technical field
The invention belongs to signal processing technology field, and in particular to a kind of small target detecting method, available for the small mesh in sea
Target recognition and tracking.
Background technology
Target detection research under sea clutter background is respectively provided with wide application prospect at military, civilian aspect.Bank base thunder
It is usually operated at up to airborne radar under high-resolution, short scan pattern.Wherein radar is obtained in each ripple position under short scan pattern
Umber of pulse below 10.In the case where distance resolution is higher, umber of pulse is less, the difficulty of sea-surface target detection compared with
It is high.And the extra large spike effect occurred in scanning more increases the difficulty of sea-surface target detection.
Document He, Y., Guan, J.:Meng,X.W.et al.:‘Radar target detection and CFAR
Processing ', (Tsinghua University Press, 2011,2st edn.), pp.30-50 and document Watts, S.:
‘Cell-averaging CFAR gain in spatially correlated K-distributed clutter’,IEEE
Radar Sonar Navig., 1996,143, the pp.321-327 all kinds of CFAR detection methods based on energy proposed are this
Method is widely used in Radar Targets'Detection soon due to easy to implement, calculating speed.CFAR detection side based on energy
Method still can be used in the case where sea clutter statistical property is unknown, umber of pulse is less.But in the case of high resolution range,
The echo of weak target is more easy to be submerged in the signal to noise ratio for causing radar to obtain in strong sea clutter and reduces, and when signal to noise ratio is relatively low
When, the detection performance of the CFAR detection method based on energy is very poor.Moreover, the CFAR detection method based on energy can not
Reduce a large amount of false-alarms that extra large spike effect is brought.Based on these deficiencies, the CFAR detection method based on energy is caused to be difficult to
Complete Detection task.
The content of the invention
It is an object of the invention in view of the above shortcomings of the prior art, propose a kind of multiframe connection based on Doppler's guiding
Small object double check method is closed, to reduce false-alarm probability under high resolution range pattern, improves the detection to sea-surface target
Energy.
To achieve the above object, technical scheme includes as follows：
(1) continuous pulse signal is launched using radar transmitter, radar receiver receives the echo data of M × I × N-dimensional
Matrix X, wherein, M represents that frame number scans number, and I represents range cell number, and N represents accumulation umber of pulse；
(2) the linear thresholding of Generalized Likelihood Ratio under different Doppler's passages is sought along pulse dimension using echo data matrix X
The detection statistic of detector, calculates data matrix Z, wherein, the data of n pages of the m row i row of data matrix Z are Z (m, i, n)：
X (m, i) represents the echo data of the m rows i row of echo data matrix X；λ represents the shape ginseng under inverse Gauss texture
Number；η represents the scale parameter under inverse Gauss texture；M (m, i) represents the echo data of the m rows i row of echo data matrix X
Speckle covariance matrix；p(fn) represent Doppler frequency fnUnder Doppler's steering vector.
(3) data matrix Z is handled as follows along Doppler's passage dimension：
(3a) calculates the maximum detection statistics moment matrix ξ of the linear threshold detector of Generalized Likelihood Ratio1, wherein counting moment matrix
ξ1M rows i row data be ξ1(m,i)：
(3b) calculates the corresponding Doppler's access matrix of maximum test statisticsWherein Doppler's access matrix
M rows i row data be
(4) the false-alarm probability p of the first re-detection is given1With the false-alarm probability p of the second re-detection, and it is real to pass through Monte Carlo
Test and calculate the first re-detection thresholding T1, wherein, p1≥p；
(5) according to the first re-detection thresholding T1Reset maximum detection statistics moment matrix ξ1With Doppler's access matrix
The maximum detection statistics moment matrix ξ that (5a) is reset1' m rows i row data be：
Doppler's access matrix that (5b) is resetM rows i row data be：
(6) the long 2L+1 of window is given, for the unit d to be detected of m rows i row, utilizes its Doppler's channel information
As guiding, Direction matching filtering device of the design corresponding to detection unit d；
(7) the Direction matching filtering device obtained in (6) and the maximum detection statistics moment matrix ξ reset are utilized1', meter
Calculate the test statistics ξ of unit d to be detected；
(8) according to the false-alarm probability p of the second re-detection, the second re-detection thresholding T is calculated by Monte Carlo experiment；
(9) size of comparing check statistic ξ and the second re-detection thresholding T, judges that target whether there is：
If ξ >=T, show that range cell d to be detected has target,
If ξ ＜ T, show that range cell d to be detected does not have target.
The present invention has the following advantages compared with the prior art：
1) present invention, can be in sea due to based on energy measuring, having more popularity compared to existing self-adapting detecting method
Sea-surface target is detected in the case that clutter statistical characteristics are unknown, umber of pulse is less.
2) present invention is using doppler information due to based on accumulation of the energy on direction and guiding search, not only profit
The information of target direction of motion is also used with the energy information of interframe, compared to the existing CFAR detection side by energy
Method, performance are more excellent；Compared to comprehensive scanning class method, Doppler's guiding is more targeted, is ensureing compared with high detection performance
On the premise of, calculation amount is reduced, is particularly suited for being detected sea surface small target under fast scan mode.
3) present invention, compared to existing method, can effectively suppress extra large spike, improve in high-resolution due to the use of double check
Radar reduces false-alarm probability to the detection performance of sea surface small target under pattern.
Brief description of the drawings
Fig. 1 realizes flow chart for the present invention's；
Direction matching filtering device template schematic diagram when Fig. 2 is Direction matching filtering device parameter L=4；
Fig. 3 is the result figure for carrying out target detection under Observed sea clutter with the present invention and existing method；
Fig. 4 is the present invention and existing method detection performance contrast curve.
Embodiment
The invention will be further described below in conjunction with the accompanying drawings：
With reference to Fig. 1, step is as follows for of the invention realizing：
Step 1, echo data is obtained.
Launch continuous pulse signal using radar transmitter, radar receiver receives the echo data square of M × I × N-dimensional
Battle array X, wherein, M represents that frame number scans number, and I represents range cell number, and N represents accumulation umber of pulse.
Step 2, using echo data matrix X, ask the Generalized Likelihood Ratio under different Doppler's passages linear along pulse dimension
The detection statistics moment matrix Z of threshold detector.
(2.1) estimate to obtain from measured data by quantile method of estimation form parameter λ under inverse Gauss texture and
Scale parameter η；
(2.2) for the unit to be estimated of the m rows i row positioned at echo data matrix X, p reference is chosen around it
Unit Zp, calculate the speckle covariance matrix M (m, i) of clutter：
Wherein：For the conjugate transposition of p the reference unit data；
(2.3) Doppler frequency f is calculatednWith Doppler frequency fnCorresponding Doppler's steering vector p (fn)：
Wherein TrFor the pulse repetition period；
(2.4) form parameter λ and scale parameter η, step (2.2) under the inverse Gauss texture obtained according to step (2.1)
The Doppler steering vector p (f that obtained speckle covariance matrix M (m, i) and step (2.3) is obtainedn), calculate Generalized Likelihood Ratio
The detection statistics moment matrix Z of linear threshold detector, wherein, the data of n page of the m row i row of detection statistics moment matrix Z be Z (m,
i,n)：
Wherein:X (m, i) represents the echo data of the m rows i row of echo data matrix X；pH(fn) it is p (fn) conjugation turn
Put.
Step 3, detection statistics moment matrix Z is handled along Doppler's passage dimension.
(3.1) the maximum detection statistics moment matrix ξ of the linear threshold detector of Generalized Likelihood Ratio is calculated1, wherein statistic square
Battle array ξ1M rows i row data be ξ1(m,i)：
(3.2) the corresponding Doppler's access matrix of maximum test statistics is calculatedWherein Doppler's access matrix's
M rows i row data be
Step 4, the first re-detection thresholding T is calculated by Monte Carlo experiment1。
(4.1) the false-alarm probability p of the first re-detection is given1With the false-alarm probability p of the second re-detection wherein, p1≥p；
(4.2) in the maximum detection statistics moment matrix ξ of the linear threshold detector of Generalized Likelihood Ratio1Middle selection V pure clutters
Unit is as training unit, V >=100/p1；
(4.3) each pure clutter unit is taken in maximum detection statistics moment matrix ξ1In it is corresponding value be used as test statistics,
V obtained test statistics is arranged in descending order, takes [the Vp after arrangement1] a test statistics is as the first re-detection door
Limit T1, wherein [Vp1] represent to be no more than real number Vp1Maximum integer.
Step 5, according to the first re-detection thresholding T1Reset maximum detection statistics moment matrix ξ1With Doppler's access matrix
(5.1) the maximum detection statistics moment matrix ξ reset1' m rows i row data be：
(5.2) the Doppler's access matrix resetM rows i row data be：
Step 6, Direction matching filtering device of the design corresponding to detection unit d.
(6.1) the long 2L+1 of window is given, in the case where ensureing that all directions can be scanned, according to the value range of k not
Together, the type of Direction matching filtering device is determined：
Work as k=0,1 ... during 2L-1, to laterally match wave filter；
Work as k=2L, 2L+1 ... it is longitudinally matched wave filter during 4L-1,
The data h of the m rows i row of both matched filters is calculated respectivelyk(m, i) is：
Laterally match wave filter：
Longitudinally matched wave filter：
Wherein：Fix represented to zero rounding, and L is positive integer, L >=4, this example takes L=4, and attached drawing 2 is selected by this example
Direction matching filtering device schematic diagram, the wherein matched filter of serial number 0-7 is laterally matches wave filter；Serial number 8-15's
Matched filter is longitudinally matched wave filter；
(6.2) the corresponding speed v of k-th of different directions matched filter is calculated respectivelyk：
Laterally match wave filter k-th：
K-th of longitudinally matched wave filter：
Wherein, Δ r represents the width of range gate；Δ T represents the sweep time of each frame during more frame scans；
(6.3) the Doppler's passage guidance information corresponding to detection unit d obtained in step (3) is utilizedReally
The fixed Direction matching filtering device k with current unit doppler velocity difference minimum to be detectedc：
Wherein：λ is radar operation wavelength,TrFor the pulse repetition period；
(6.4) with the Direction matching filtering device k of the doppler velocity difference minimum with current unit d to be detectedcCentered on,
Direction matching filtering device group of each N number of Direction matching filtering device in left and right as current unit to be detected is chosen, obtains direction matching
The sequence number matrix k of wave filter groupf：
kf=mod (kc+ j, 4L), j=0, ± 1 ..., ± N,
Wherein：Mod represents remainder operation.
Step 7, the test statistics of unit d to be detected is calculated.
(7.1) the maximum detection statistics moment matrix ξ after resetting1' in, only non-zero unit is detected, to treat
Centered on detection unit d, it is masterplate matrix to choose (2L+1) × (2L+1) dimension datas, by design in masterplate matrix and step (6)
The Direction matching filtering device dot product corresponding to detection unit d, and accumulation summation is carried out to all elements in dot product result, obtained
To unit d to be detected correspond to direction matched filter accumulation and, then calculate masterplate matrix and the side chosen in all (6)
Accumulation to matched filter and, obtain the second re-detection statistics moment matrix ξ2, wherein the second re-detection statistics moment matrix ξ2
The data of k pages of m row i row are ξ2(m,i,k)：
ξ2(m, i, k), k=kf,
Wherein：kfFor the Direction matching filtering device sequence number chosen in step (6)；
(7.2) the second re-detection statistics moment matrix ξ is utilized2, calculate the conclusive judgement detection statistics corresponding to detection unit d
Moment matrix ξ, the m row i column data ξ (m, i) of wherein conclusive judgement detection statistics moment matrix ξ are：
Step 8, the second re-detection thresholding T is calculated by Monte Carlo experiment.
(8.1) the V pure clutter units chosen using in step (4.1) are used as training unit；
(8.2) the range cell d to be detected in replacement step (6) and step (7) is distinguished using V training unit, is laid equal stress on
Multiple step (6) and step (7), obtain the test statistics of each training unit；
(8.3) V obtained test statistics is arranged in descending order, takes [Vp] a test statistics conduct after arrangement
Detection threshold T, p are the false-alarm probability of the second re-detection given in step (4).
Step 9, judge that target whether there is by the size of comparing check statistic ξ and detection threshold T：
If ξ >=T, show that range cell d to be detected has target,
If ξ ＜ T, show that range cell d to be detected does not have target.
The effect of the present invention is described further with reference to emulation experiment.
One, experimental datas
This example uses the Observed sea clutter that land-based radar gathers, and radar pulse repetition frequency is 1800 hertz, away from
It it is 1 meter from resolving power.It is 6 to choose data pulse number, and range cell number is 500, frame number 48, and target is averaged signal to noise ratio as 10 points
Shellfish.
Two, emulation experiments
Emulation 1, using it is of the invention and it is existing be based on the linear threshold detector GLRT-LTD detection methods of Generalized Likelihood Ratio, divide
Other that target is detected, the results are shown in Figure 3, wherein：
Fig. 3 (a) is to use the linear threshold detector detection method of existing Generalized Likelihood Ratio, and takes false-alarm probability p=10-3It is right
The result figure that target is detected；
Fig. 3 (b) is using the method for the present invention, and takes the false-alarm probability p of the first re-detection1=10-2With the second re-detection
False-alarm probability p=10-3The result figure being detected to target；
White area in Fig. 3 represents sea clutter background, and the stain for being linked to be line represents target trajectory, and isolated stain represents
As seen from Figure 3, the present invention effectively suppresses extra large spike compared to the linear threshold detector detection method of Generalized Likelihood Ratio is based on
Extra large spike in scanning, obtained target trajectory become apparent from.The detection probability of the present invention is 0.8125, based on Generalized Likelihood
Detection probability than linear threshold detector detection method is 0.5833.
Emulation 2, with the linear threshold detector GLRT-LTD detection methods of of the invention and existing Generalized Likelihood Ratio in different void
Simulation objectives data are detected under alarm probability, obtained performance comparison curve is as shown in figure 4, wherein：
Detection probability shown in solid for the linear threshold detector detection method of Generalized Likelihood Ratio is with the change of false alarm rate
Curve；Dotted line show the detection probability of the present invention with the change curve of signal to noise ratio.Detection performance of the invention as seen from Figure 4
Curve is always positioned on control methods.
To sum up, the present invention is better than existing method to the detection performance of sea surface small target under high resolution mode.
Claims (8)
1. a kind of multi-frame joint Small object double check method based on Doppler's guiding, including：
(1) continuous pulse signal is launched using radar transmitter, radar receiver receives the echo data matrix of M × I × N-dimensional
X, wherein, M represents that frame number scans number, and I represents range cell number, and N represents accumulation umber of pulse；
(2) the linear Threshold detection of Generalized Likelihood Ratio under different Doppler's passages is sought along pulse dimension using echo data matrix X
The detection statistic of device, calculates data matrix Z, wherein, the data of n pages of the m row i row of data matrix Z are Z (m, i, n)：
<mfenced open = "" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mi>Z</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msup>
<mi>p</mi>
<mi>H</mi>
</msup>
<mrow>
<mo>(</mo>
<msub>
<mi>f</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
</mrow>
<msup>
<mi>M</mi>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mi>X</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
</mrow>
<mrow>
<mo>(</mo>
<msup>
<mi>p</mi>
<mi>H</mi>
</msup>
<mo>(</mo>
<msub>
<mi>f</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
<msup>
<mi>M</mi>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<mo>(</mo>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
<mo>)</mo>
<mi>p</mi>
<mo>(</mo>
<msub>
<mi>f</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
<mo>)</mo>
<mo>(</mo>
<mi>&lambda;</mi>
<mo>/</mo>
<mi>&eta;</mi>
<mo>+</mo>
<mi>X</mi>
<msup>
<mrow>
<mo>(</mo>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
<mo>)</mo>
</mrow>
<mi>H</mi>
</msup>
<msup>
<mi>M</mi>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<mo>(</mo>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
<mo>)</mo>
<mi>X</mi>
<mo>(</mo>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mtable>
<mtr>
<mtd>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>M</mi>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>I</mi>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mn>4</mn>
<mi>N</mi>
<mo>,</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mtd>
</mtr>
</mtable>
</mfenced>
X (m, i) represents the echo data of the m rows i row of echo data matrix X；λ represents the form parameter under inverse Gauss texture；η
Represent the scale parameter under inverse Gauss texture；M (m, i) represents the speckle of the echo data of the m rows i row of echo data matrix X
Covariance matrix；p(fn) represent Doppler frequency fnUnder Doppler's steering vector.
(3) data matrix Z is handled as follows along Doppler's passage dimension：
(3a) calculates the maximum detection statistics moment matrix ξ of the linear threshold detector of Generalized Likelihood Ratio1, wherein statistics moment matrix ξ1's
The data of m rows i row are ξ1(m,i)：
<mrow>
<msub>
<mi>&xi;</mi>
<mn>1</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
<mrow>
<mi>n</mi>
<mo>&Element;</mo>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>...</mo>
<mo>,</mo>
<mn>4</mn>
<mi>N</mi>
<mo>&rsqb;</mo>
</mrow>
</munder>
<mo>{</mo>
<mi>Z</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>}</mo>
<mo>,</mo>
</mrow>
(3b) calculates the corresponding Doppler's access matrix of maximum test statisticsWherein Doppler's access matrixM rows i
The data of row are
<mrow>
<msub>
<mover>
<mi>n</mi>
<mo>^</mo>
</mover>
<mrow>
<mi>f</mi>
<mi>d</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mrow>
<mi>arg</mi>
<mi> </mi>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
<mrow>
<mi>n</mi>
<mo>&Element;</mo>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2...</mn>
<mo>,</mo>
<mn>4</mn>
<mi>N</mi>
<mo>&rsqb;</mo>
</mrow>
</munder>
<mo>{</mo>
<mi>Z</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>}</mo>
<mo>,</mo>
</mrow>
(4) the false-alarm probability p of the first re-detection is given1With the false-alarm probability p of the second re-detection, and pass through Monte Carlo experiment meter
Calculate the first re-detection thresholding T1, wherein, p1≥p；
(5) according to the first re-detection thresholding T1Reset maximum detection statistics moment matrix ξ1With Doppler's access matrix
The maximum detection statistics moment matrix ξ that (5a) is reset1' m rows i row data be：
Doppler's access matrix that (5b) is resetM rows i row data be：
(6) the long 2L+1 of window is given, for the unit d to be detected of m rows i row, utilizes its Doppler's channel informationAs
Guiding, Direction matching filtering device of the design corresponding to detection unit d；
(7) the Direction matching filtering device obtained in (6) and the maximum detection statistics moment matrix ξ reset are utilized1', calculate to be checked
Survey the test statistics ξ of unit d；
(8) according to the false-alarm probability p of the second re-detection, the second re-detection thresholding T is calculated by Monte Carlo experiment；
(9) size of comparing check statistic ξ and the second re-detection thresholding T, judges that target whether there is：
If ξ >=T, show that range cell d to be detected has target,
If ξ ＜ T, show that range cell d to be detected does not have target.
2. the method as described in claim 1, it is characterised in that the test statistics ξ of unit d to be detected is calculated in step (7),
Carry out as follows：
(7.1) the maximum detection statistics moment matrix ξ after resetting1' in, only non-zero unit is detected, with list to be detected
Centered on first d, it is masterplate matrix to choose (2L+1) × (2L+1) dimension datas, and masterplate matrix is corresponding with being designed in step (6)
In the Direction matching filtering device dot product of detection unit d, and accumulation summation is carried out to all elements in dot product result, obtained to be checked
Survey unit d correspond to direction matched filter accumulation and, then calculate masterplate matrix and matched with direction of selection in all (6)
The accumulation of wave filter and, obtain the second re-detection statistics moment matrix ξ2, wherein the second re-detection statistics moment matrix ξ2M rows i row
K pages of data are ξ2(m,i,k)：
ξ2(m, i, k), k=kf,
Wherein：kfFor the Direction matching filtering device sequence number chosen in step (6)；
(7.2) the second re-detection statistics moment matrix ξ is utilized2, calculate the conclusive judgement detection statistic square corresponding to detection unit d
Battle array ξ, the wherein m row i column datas of conclusive judgement detection statistics moment matrix ξ are ξ (m, i)：
<mrow>
<mi>&xi;</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
<mrow>
<mi>k</mi>
<mo>=</mo>
<msub>
<mi>k</mi>
<mi>f</mi>
</msub>
</mrow>
</munder>
<mo>{</mo>
<msub>
<mi>&xi;</mi>
<mn>2</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
<mo>}</mo>
<mo>.</mo>
</mrow>
3. the method as described in claim 1, it is characterised in that the speckle covariance matrix M in step (2) formula Z (m, i, n)
(m, i) is calculated by following formula：
<mrow>
<mi>M</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mi>N</mi>
<mi>p</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>p</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>P</mi>
</munderover>
<mfrac>
<mrow>
<msub>
<mi>Z</mi>
<mi>p</mi>
</msub>
<msubsup>
<mi>Z</mi>
<mi>p</mi>
<mi>H</mi>
</msubsup>
</mrow>
<mrow>
<msubsup>
<mi>Z</mi>
<mi>p</mi>
<mi>H</mi>
</msubsup>
<msub>
<mi>Z</mi>
<mi>p</mi>
</msub>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
Wherein：ZpFor the p reference unit chosen around unit to be estimated.
4. the method as described in claim 1, it is characterised in that the Doppler frequency f in step (2) formula Z (m, i, n)nIt is corresponding
Doppler steering vector p (fn) calculated by following formula：
<mrow>
<mi>p</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>f</mi>
<mi>n</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>,</mo>
<msup>
<mi>e</mi>
<mrow>
<mi>j</mi>
<mn>2</mn>
<msub>
<mi>&pi;f</mi>
<mi>n</mi>
</msub>
<msub>
<mi>T</mi>
<mi>r</mi>
</msub>
</mrow>
</msup>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<msup>
<mi>e</mi>
<mrow>
<mi>j</mi>
<mn>2</mn>
<msub>
<mi>&pi;f</mi>
<mi>n</mi>
</msub>
<msub>
<mi>T</mi>
<mi>r</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>&rsqb;</mo>
<mo>,</mo>
</mrow>
<mrow>
<msub>
<mi>f</mi>
<mi>n</mi>
</msub>
<mo>=</mo>
<mo>-</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mn>2</mn>
<msub>
<mi>T</mi>
<mi>r</mi>
</msub>
</mrow>
</mfrac>
<mo>+</mo>
<mfrac>
<mrow>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
<msub>
<mi>T</mi>
<mi>r</mi>
</msub>
</mrow>
</mfrac>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>...</mo>
<mo>,</mo>
<mn>4</mn>
<mi>N</mi>
<mo>.</mo>
</mrow>
Wherein TrFor the pulse repetition period.
5. the method as described in claim 1, it is characterised in that unknown inverse Gauss texture in step (2) formula Z (m, i, n)
Under form parameter λ and scale parameter η, estimate to obtain from measured data by quantile method.
6. the method as described in claim 1, it is characterised in that in step (4), calculate first by Monte Carlo experiment and examine again
Survey thresholding T1, carry out as follows：
(4.1) in the maximum detection statistics moment matrix ξ of the linear threshold detector of Generalized Likelihood Ratio1Middle selection V pure clutter units are made
For training unit, V >=100/p1, p1For the false-alarm probability of the first given re-detection；
(4.2) each pure clutter unit is taken in data matrix ξ1In it is corresponding value be used as test statistics, by V obtained examine
Statistic arranges in descending order, takes [the Vp after arrangement1] a test statistics is as the first re-detection thresholding T1, wherein [Vp1] table
Show and be no more than real number Vp1Maximum integer.
7. the method as described in claim 1, it is characterised in that design corresponds to the list to be detected of m rows i row in step (6)
The Direction matching filtering device of first d, carries out as follows：
(6.1) the long 2L+1 of window is given, it is different according to the value range of k in the case where ensureing that all directions can be scanned, really
Determine the type of Direction matching filtering device, work as k=0,1 ... during 2L-1, to laterally match wave filter；Work as k=2L, 2L+1,
... it is longitudinal direction matched filter during 4L-1, calculates the data h of the m rows i row of both matched filters respectivelyk(m,
I) it is：
Laterally match wave filter：
Longitudinally matched wave filter：
Wherein：Fix represents that, to zero rounding, L is positive integer L >=4；
(6.2) the corresponding speed v of k-th of different directions matched filter is calculated respectivelykFor：
Laterally match wave filter k-th：
K-th of longitudinally matched wave filter：
Wherein, Δ r represents the width of range gate；Δ T represents the sweep time of each frame during more frame scans；
(6.3) the Doppler's guidance information corresponding to detection unit d obtained in step (5) is utilizedDetermine and current
The Direction matching filtering device k of unit doppler velocity difference minimum to be detectedc：
<mrow>
<msub>
<mi>k</mi>
<mi>c</mi>
</msub>
<mo>=</mo>
<munder>
<mrow>
<mi>arg</mi>
<mi> </mi>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
<mi>k</mi>
</munder>
<mo>{</mo>
<mo>|</mo>
<msub>
<mi>v</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<mfrac>
<mrow>
<msub>
<mi>f</mi>
<mrow>
<msub>
<mover>
<mi>n</mi>
<mo>^</mo>
</mover>
<mrow>
<msup>
<mi>fd</mi>
<mo>&prime;</mo>
</msup>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mi>&lambda;</mi>
</mrow>
<mn>2</mn>
</mfrac>
<mo>|</mo>
<mo>}</mo>
<mo>.</mo>
</mrow>
<mrow>
<msub>
<mi>f</mi>
<mrow>
<msup>
<msub>
<mover>
<mi>n</mi>
<mo>^</mo>
</mover>
<mrow>
<mi>f</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mo>=</mo>
<mo>-</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mn>2</mn>
<msub>
<mi>T</mi>
<mi>r</mi>
</msub>
</mrow>
</mfrac>
<mo>+</mo>
<mfrac>
<mrow>
<msup>
<msub>
<mover>
<mi>n</mi>
<mo>^</mo>
</mover>
<mrow>
<mi>f</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
<msub>
<mi>T</mi>
<mi>r</mi>
</msub>
</mrow>
</mfrac>
<mo>.</mo>
</mrow>
Wherein：λ is radar operation wavelength, TrFor the pulse repetition period；
(6.4) with the Direction matching filtering device k of the doppler velocity difference minimum with current unit d to be detectedcCentered on, choose
Direction matching filtering device group of each N number of Direction matching filtering device in left and right as current unit to be detected, obtains Direction matching filtering
The sequence number matrix k of device groupf：
kf=mod (kc+ j, 4L), j=0, ± 1 ..., ± N,
Wherein：Mod represents remainder operation.
8. the method as described in claim 1 or 6, it is characterised in that the second weight is calculated by Monte Carlo experiment in step (8)
Detection threshold T, carries out as follows：
(8.1) the V pure clutter units chosen using in step (4.1) are used as training unit；
(8.2) the range cell d to be detected in replacement step (6) and step (7) is distinguished using V training unit, and repeats to walk
Suddenly (6) and step (7), obtain the test statistics of each training unit；
(8.3) V obtained test statistics is arranged in descending order, takes [Vp] a test statistics after arrangement as detection
Thresholding T, p are the false-alarm probability of the second re-detection given in step (4).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711075845.7A CN107942324B (en) | 2017-11-06 | 2017-11-06 | Multi-frame combined small target dual detection method based on Doppler guidance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711075845.7A CN107942324B (en) | 2017-11-06 | 2017-11-06 | Multi-frame combined small target dual detection method based on Doppler guidance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107942324A true CN107942324A (en) | 2018-04-20 |
| CN107942324B CN107942324B (en) | 2021-02-26 |
Family
ID=61933344
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711075845.7A Active CN107942324B (en) | 2017-11-06 | 2017-11-06 | Multi-frame combined small target dual detection method based on Doppler guidance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107942324B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2730182C1 (en) * | 2019-08-19 | 2020-08-19 | Иван Васильевич Колбаско | Method of multiple-rundown signal accumulation in radar station when detecting aerial targets in pulse-doppler mode |
| CN112558014A (en) * | 2021-02-23 | 2021-03-26 | 中国人民解放军空军预警学院 | Method and system for detecting adjustable subspace of extended target parameters in clutter |
| CN112558034A (en) * | 2021-02-23 | 2021-03-26 | 中国人民解放军空军预警学院 | Extended target sensitive detector and system during subspace signal mismatch |
| CN112763981A (en) * | 2020-12-15 | 2021-05-07 | 山东农业大学 | Detection method and system for radar pulse signal |
| CN115902881A (en) * | 2022-12-29 | 2023-04-04 | 中国人民解放军空军预警学院 | Method and system for detecting extended target of distributed unmanned airborne radar |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156282A (en) * | 2011-03-25 | 2011-08-17 | 电子科技大学 | Method for detecting radar target based on micro-Doppler effect |
| CN106199537A (en) * | 2016-06-29 | 2016-12-07 | 西安电子科技大学 | Quantile method of estimation based on inverse Gauss texture sea clutter amplitude distribution parameter |
| CN106569193A (en) * | 2016-11-01 | 2017-04-19 | 西安电子科技大学 | Sea-surface small target detection method based on front-back revenue reference particle filter |
| CN107132532A (en) * | 2017-06-13 | 2017-09-05 | 西安电子科技大学 | Small target detecting method based on extra large peak restrained and multi-frame joint |
-
2017
- 2017-11-06 CN CN201711075845.7A patent/CN107942324B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156282A (en) * | 2011-03-25 | 2011-08-17 | 电子科技大学 | Method for detecting radar target based on micro-Doppler effect |
| CN106199537A (en) * | 2016-06-29 | 2016-12-07 | 西安电子科技大学 | Quantile method of estimation based on inverse Gauss texture sea clutter amplitude distribution parameter |
| CN106569193A (en) * | 2016-11-01 | 2017-04-19 | 西安电子科技大学 | Sea-surface small target detection method based on front-back revenue reference particle filter |
| CN107132532A (en) * | 2017-06-13 | 2017-09-05 | 西安电子科技大学 | Small target detecting method based on extra large peak restrained and multi-frame joint |
Non-Patent Citations (1)
| Title |
|---|
| BENZHOU JIN,YONGXUE MA,GANG WU: "Fast scan-to-scan intergration algorithm for small target in sea clutter", 《ELECTRONICS LETTERS》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2730182C1 (en) * | 2019-08-19 | 2020-08-19 | Иван Васильевич Колбаско | Method of multiple-rundown signal accumulation in radar station when detecting aerial targets in pulse-doppler mode |
| CN112763981A (en) * | 2020-12-15 | 2021-05-07 | 山东农业大学 | Detection method and system for radar pulse signal |
| CN112558014A (en) * | 2021-02-23 | 2021-03-26 | 中国人民解放军空军预警学院 | Method and system for detecting adjustable subspace of extended target parameters in clutter |
| CN112558034A (en) * | 2021-02-23 | 2021-03-26 | 中国人民解放军空军预警学院 | Extended target sensitive detector and system during subspace signal mismatch |
| CN115902881A (en) * | 2022-12-29 | 2023-04-04 | 中国人民解放军空军预警学院 | Method and system for detecting extended target of distributed unmanned airborne radar |
| CN115902881B (en) * | 2022-12-29 | 2024-03-29 | 中国人民解放军空军预警学院 | Method and system for detecting extended target of distributed unmanned aerial vehicle radar |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107942324B (en) | 2021-02-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107942324A (en) | Multi-frame joint Small object double check method based on Doppler's guiding | |
| CN106468770B (en) | Nearly optimal radar target detection method under K Distribution Clutter plus noise | |
| CN104569949B (en) | Radar target detection method based on combined adaptive normalized matched filter | |
| CN104316914B (en) | Radar target self-adaptation detection method depending on shape parameters | |
| CN103439697B (en) | Target detection method based on dynamic programming | |
| CN104237879B (en) | A kind of multi-object tracking method in radar system | |
| CN111965632B (en) | Radar target detection method based on Riemann manifold dimensionality reduction | |
| CN106872958B (en) | Radar target self-adapting detecting method based on linear fusion | |
| CN107085205B (en) | Self-adapting detecting method based on clutter covariance matrix structural information | |
| CN105785340B (en) | The angle measurement and recognition methods that monopulse radar interferes target in main lobe and bait | |
| CN106569193B (en) | The small targets detection in sea clutter method filtered based on anterior-posterior to income reference particle | |
| CN107561518A (en) | Three-dimensional radar Plot coherence method based on two-dimentional sliding window local extremum | |
| CN105425230A (en) | Range-extended target adaptive detection method under inverse Gaussian texture complex Gaussian clutter | |
| CN107544061A (en) | Knowledge based engineering is sparse to recover space-time adaptive processing method and system | |
| CN106872969B (en) | Radar target angle estimation method based on MTD pulse accumulation and sliding processing | |
| CN106772299B (en) | One kind is based on apart from matched PD radar weak target Dynamic Programming detection method | |
| CN109143195A (en) | Radar target detection method based on full KL divergence | |
| CN113221631B (en) | Sequence pulse anti-interference target detection method based on convolutional neural network | |
| CN103197297B (en) | Radar moving target detection method based on cognitive framework | |
| CN104714225A (en) | Dynamic programming tracking-before-detection method based on generalized likelihood ratios | |
| CN109934101A (en) | Radar clutter recognition method based on convolutional neural networks | |
| CN108872961B (en) | Radar weak target detection method based on low threshold | |
| CN104680536A (en) | Method for detecting SAR image change by utilizing improved non-local average algorithm | |
| CN106353743B (en) | It is matched with the nearly optimal radar target detection method of equivalent shapes parameter | |
| CN107132532A (en) | Small target detecting method based on extra large peak restrained and multi-frame joint |
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 |