CN106291492B - A kind of adaptive targets detection method based on fine clutter map - Google Patents
A kind of adaptive targets detection method based on fine clutter map Download PDFInfo
- Publication number
- CN106291492B CN106291492B CN201610570313.XA CN201610570313A CN106291492B CN 106291492 B CN106291492 B CN 106291492B CN 201610570313 A CN201610570313 A CN 201610570313A CN 106291492 B CN106291492 B CN 106291492B
- Authority
- CN
- China
- Prior art keywords
- thresholding
- valuation
- unit
- false alarm
- clutter
- 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
Abstract
The invention belongs to Radar Signal Processing Technology fields, disclose a kind of adaptive targets detection method based on fine clutter map.Using improved fining clutter map, different update modes can be carried out to land clutter and sea clutter according to the difference of clutter attribute.It is different from traditional clutter map update method for only updating amplitude information, counts and updates two kinds of parameters of amplitude and variance simultaneously in the present invention, covariance information is used in subsequent signal weighting, extra large spike bring false-alarm can be effectively inhibited.Meanwhile traditional constant false alarm gate method is distinguished, the present invention uses while seeking the technical solution of fast, slow thresholding constant false alarm thresholding, and the automatic selection of fast, slow constant false alarm thresholding is carried out according to certain criterion, is effectively improved constant false alarm effect.
Description
Technical field
The present invention relates to Radar Signal Processing Technology field more particularly to a kind of adaptive targets based on fine clutter map
Detection method.
Background technique
The military change to increasingly sharpen proposes increasingly higher demands to the detection performance of radar.It is often deposited around target
In complicated strong land clutter, sea clutter and other a variety of clutters etc..Noise intensity is generally all very big, often compares target echo
Signal is much better than, causes severe jamming to target detection.So radar must try to clutter reduction before carrying out target detection.
Detection for slower-velocity target is handed over since the Doppler frequency spectrum and slower-velocity target signal Doppler frequency spectrum of clutter exist
Folded, either traditional kalmus filter or MTI/MTD technology are all the stop-band recesses by the way that depth is presented at zero-frequency,
Essence is that clutter reduction is by way of frequency domain filtering to attempt to be promoted signal to noise ratio, while the problem of bringing is filtering clutter pair
The slower-velocity target tangentially to fly also has certain loss, so that slower-velocity target signal to noise ratio is difficult to be promoted, which results in low speed
Target is difficult to the problem of detecting.
Traditional super clutter detection technique based on clutter map, signal pulse improve unobvious without accumulation, signal-to-noise ratio；
The update for carrying out different modes for different clutter types is not considered when clutter map is updated simultaneously, leads to clutter map effect
It is not ideal enough.Conventional method divides clutter region using clutter sketch map, only counts amplitude when establishing clutter sketch map
Information, way are simple.Meanwhile target detection is carried out under different noises, clutter and jamming pattern, it should use different perseverances
False-alarm detector obtains high detection probability while to guarantee that false-alarm probability is constant.Traditional super clutter detection method is difficult to
Effectively realize adaptive approach selection and processing, target detection performance is difficult to effectively be promoted.
Summary of the invention
For the deficiency of above-mentioned prior art, the purpose of the present invention is to provide a kind of based on the adaptive of fine clutter map
Object detection method, the radar self-adaption object detection method based on fine clutter map can be realized in clutter, noise background
Lower radar self-adaption target detection.
The technical thought that technical solution of the present invention is realized are as follows: present invention employs improved fining clutter maps, can
According to the difference of clutter attribute, different update modes is carried out to land clutter and sea clutter.It is different from traditional update amplitude
The clutter map update method of information, the present invention in simultaneously count and update two kinds of parameters of amplitude and variance, covariance information is used for
In subsequent signal weighting, it can effectively inhibit extra large spike bring false-alarm.Meanwhile traditional constant false alarm gate method is distinguished, this
Invention uses while seeking the technical solution of fast, slow constant false alarm thresholding, and carries out fast, slow constant false alarm thresholding according to certain criterion
It is automatic to choose, it is effectively improved constant false alarm effect.
In order to achieve the above objectives, the embodiment of the present invention, which adopts the following technical scheme that, is achieved.
A kind of adaptive targets detection method based on fine clutter map, described method includes following steps:
Step 1, the continuous L of radar emission postimpulse target echo signals are obtained, to L target echo signal successively into
Row process of pulse-compression, the signal after obtaining L process of pulse-compression；
Step 2, modulus operation is carried out to the signal after the L process of pulse-compression, modulus value is believed after obtaining L pulse pressure
Number；
Step 3, the clutter map for obtaining target echo signal, according to L pulse pressure rear mold value signal to the parameter of clutter map into
Row updates, and the parameter of the clutter map includes clutter map valuation thresholding and estimate of variance；
Step 4, slow thresholding constant false alarm valuation is carried out according to first of pulse pressure rear mold value signal, it is permanent empty obtains slow thresholding
Alert valuation thresholding；Fast thresholding constant false alarm valuation is carried out according to first of pulse pressure rear mold value signal, obtains fast thresholding constant false alarm valuation door
Limit；And according to the slow thresholding constant false alarm valuation thresholding and the fast thresholding constant false alarm valuation thresholding, constant false alarm thresholding is determined；
Step 5, using the clutter map valuation thresholding and the constant false alarm thresholding to first of pulse pressure rear mold value signal
It is normalized and weights, obtain normalization weighting modulus value signal；
Step 6, it enables the value of l add 1, and is repeated in and executes step 3 to step 5, to obtain L normalization weighting modulus value
Signal, the signal after described L normalization weighting modulus value signal to be overlapped to obtain pulse accumulation, l=1 ..., L；
Step 7, detection threshold is set, element each in the signal after the pulse accumulation and the detection threshold are carried out
Compare, if some element is greater than the detection threshold in the signal after the pulse accumulation, then it is assumed that distance belonging to the element
Element memory is in target.
Beneficial effects of the present invention: first, the method for the present invention is clear in structure, be easy to hardware realization, fining clutter map and
The application of constant false alarm effectively can detect radar target in noise range and clutter area, promote radar detection ability；Second, the present invention
Method can be realized the selection of adaptive threshold, return automatically according to the suitable valuation thresholding of environmental selection locating for target to signal
One changes, and effectively promotes the reliability of detection；Third, the method for the present invention can be realized the non-coherent product of the weighting to normalized signal
It is tired, it can effectively inhibit extra large spike bring false-alarm；4th, changeable parameters are arranged in the method for the present invention, can be according to use process
In external environment variation carry out flexible modulation, flexibility, reusability and the portability that the method for effectively improving uses.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is that the process of the adaptive targets detection method provided in an embodiment of the present invention based on fine clutter map is illustrated
Figure；
Fig. 2 is the implementation process schematic diagram of clutter map valuation thresholding；
Fig. 3 is the implementation process schematic diagram of constant false alarm valuation thresholding；
Fig. 4 is the implementation process schematic diagram that thresholding normalizes selection；
Fig. 5 a, 5b, 5c be respectively land clutter area, sea clutter area, noise range pulse compression result modulus value；
Fig. 6 a, 6b, 6c are respectively the modulus value after land clutter area, sea clutter area, 10 circle of noise range clutter map update；
Fig. 7 a, 7b, 7c are respectively the slow thresholding constant false alarm valuation in land clutter area, fast thresholding constant false alarm valuation and adaptively selected
Threshold value；
Fig. 8 a, 8b, 8c are respectively the slow thresholding constant false alarm valuation in sea clutter area, fast thresholding constant false alarm valuation and adaptively selected
Threshold value；
Fig. 9 a, 9b, 9c are respectively the slow thresholding constant false alarm valuation in noise range, fast thresholding constant false alarm valuation and adaptively selected door
Limit value；
Figure 10 a, 10b, 10c are respectively land clutter area, sea clutter area, noise range clutter map normalization level；
Figure 11 a, 11b, 11c are respectively land clutter area, sea clutter area, noise range constant false alarm normalization level；
Figure 12 a, 12b, 12c are respectively land clutter area, sea clutter area, the adaptively selected rear weight level in noise range；
Figure 13 a, 13b, 13c are respectively land clutter area, sea clutter area, the non-coherent result in noise range.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The embodiment of the present invention provides a kind of adaptive targets detection method based on fine clutter map, referring to Fig.1, the side
Method includes the following steps:
Step 1, the continuous L of radar emission postimpulse target echo signals are obtained, to L target echo signal successively into
Row process of pulse-compression, the signal after obtaining L process of pulse-compression.
Specifically, it is exactly matched filtering process that pulse compression process is practical, realizes that pulse compression can pass through volume in time domain
Product carries out, and can also realize in frequency domain, and when pulse compressed coefficient sampling number m is very big, the operand of convolution is also very big,
It at this moment is suitable for using frequency-domain impulse compression method, to reduce operand.
Step 2, modulus operation is carried out to the signal after the L process of pulse-compression, modulus value is believed after obtaining L pulse pressure
Number.
Specifically, the signal after process of pulse-compression will first carry out envelope detection, and signal loses phase after envelope detection
Position information and only remain amplitude information.
Step 3, the clutter map for obtaining target echo signal, according to L pulse pressure rear mold value signal to the parameter of clutter map into
Row updates, and the parameter of the clutter map includes clutter map valuation thresholding and estimate of variance；
It should be noted that first radar scanning period, clutter map initial value is updated for 0, and second period starts,
The clutter map for obtaining target echo signal was the value of clutter map of corresponding of the upper radar scanning period apart from orientation
Step 3 specifically includes following sub-step:
The scanning range of radar is carried out the equal sector division of azimuth-range by (3a), obtains M × N number of clutter map unit,
In, M indicates the 360 degree of scannings range of radar being divided into M part in orientation, N expression by the scanning range of radar distance to
On be divided into N parts；Wherein M > 1, N > 1；Set the initial value u of the clutter map valuation thresholding of each clutter map unit0, each clutter map
The initial value σ of the estimate of variance of unit0 2；
You need to add is that first radar scanning period, clutter map is just begun setting up, so first lap does not do clutter map
Normalization weighting, only carries out signal normalization weighting with constant false alarm thresholding.
(3b) obtains the clutter attribute of i-th of clutter map unit, and the clutter attribute is divided into two kinds, one kind be land clutter and
Noise, another kind are sea clutter；The initial value of i is 1, i=1 ..., M × N；
(3c) obtains i-th of clutter if the clutter attribute of i-th of clutter map unit is land clutter and noise
The maximum value X1 for the first pulse pressure rear mold value signal for including in figure unit, after the first pulse pressure rear mold value signal is L pulse pressure
The signal in i-th of clutter map unit is fallen into modulus value signal, and using following more new formula to i-th of clutter map unit
Clutter map valuation thresholding and estimate of variance be updated:
uj=k1 × uj-1+(1-k1)×X1
σj 2=k2 × σj-1 2+(1-k2)×[uj-X1]2
Wherein, j indicates that the current scan period number of radar, 1, the k1 and k2 of j are preset iteration factor, X1=max
(x1..., xE), E is the data amount check for the pulse pressure rear mold value signal for including in i-th of clutter map unit；
(3d) is obtained in i-th of clutter map unit if the clutter attribute of i-th of clutter map unit is sea clutter
The average value X2 for the second pulse pressure rear mold value signal for including, the second pulse pressure rear mold value signal are L pulse pressure rear mold value signal
In fall into the signal in i-th of clutter map unit, and using following more new formula to the clutter map of i-th of clutter map unit
Valuation thresholding and estimate of variance are updated:
uj=k3 × uj-1+(1-k3)×X2
Wherein, j indicates that the current scan period number of radar, j >=1, k3 and k4 are preset iteration factor,E is the pulse pressure rear mold value signal for including in i-th of clutter map unit
Data amount check；
(3e) successively carries out sub-step (3b) to sub-step to each clutter map unit in M × N number of clutter map unit
(3d), to obtain the clutter map valuation thresholding and estimate of variance of each clutter map unit.
Different update modes is carried out according to clutter attribute, the characteristic for being conducive to more preferably distinguish different clutters carries out accordingly
Processing, undated parameter includes covariance information, can reflect the smoothness of clutter, non-to be weighted to subsequent normalized signal
Correlative accumulation.
Step 4, slow thresholding constant false alarm valuation is carried out according to first of pulse pressure rear mold value signal, it is permanent empty obtains slow thresholding
Alert valuation thresholding；Fast thresholding constant false alarm valuation is carried out according to first of pulse pressure rear mold value signal, obtains fast thresholding constant false alarm valuation door
Limit；And according to the slow thresholding constant false alarm valuation thresholding and the fast thresholding constant false alarm valuation thresholding, constant false alarm thresholding is determined.
The purpose that two constant false alarm valuations are arranged is for noise jamming using slow thresholding constant false alarm valuation thresholding, for miscellaneous
Wave area edge is using fast thresholding constant false alarm valuation thresholding to reduce false-alarm.And by setting certain criterion, can choose automatically
Constant false alarm thresholding valuation result.
Step 4 specifically includes following sub-step:
It include P distance unit in (4a) first of pulse pressure rear mold value signal, after xp indicates the pulse pressure of p-th of distance unit
Modulus value signal；P=1 ..., P；L ∈ { 1,2 ..., L }；
(4b) using p-th of distance unit as detected unit, the initial value of p is 1；
T1 distance unit on the left of detected unit is set as protection location by (4c), on the left of the protection location successively
Continuous C1 reference unit is chosen, using the average value of the pulse pressure rear mold value signal of C1 reference unit as the detected list
The slow thresholding constant false alarm valuation thresholding of member；If the number of reference unit is less than C1, the slow thresholding constant false alarm valuation is set
Thresholding is zero；T1 >=1, C1 >=1；
T2 distance unit on the left of detected unit is set as left protection location by (4d), by the T2 on the right side of detected unit
A distance unit is set as right protection location, C2 left reference units is chosen on the left of the left protection location, from the right protection
C2 right reference units are chosen on the right side of unit, and the pulse pressure rear mold value signal of C2 left reference units is averaged, is denoted as first
The pulse pressure rear mold value signal of C2 right reference units is averaged, is denoted as the second average value by average value；And more described first
The size of average value and second average value is estimated biggish average value as the fast thresholding constant false alarm of the detected unit
It is worth thresholding；If the number of left reference unit is less than C2, using second average value as the fast thresholding of the detected unit
Constant false alarm valuation thresholding, if the number of right reference unit is less than C2, using first average value as the detected unit
Fast thresholding constant false alarm valuation thresholding；
(4e) if be detected unit on the left of distance unit number be less than C1+T1, by fast thresholding constant false alarm valuation door
Limit the constant false alarm thresholding y that unit is detected as thisp, otherwise calculate the slow thresholding constant false alarm valuation thresholding of the detected unit
With the absolute value of the difference of fast thresholding constant false alarm valuation thresholding, and difference threshold is preset, if the detected unit is slow
The absolute value of the difference of thresholding constant false alarm valuation thresholding and fast thresholding constant false alarm valuation thresholding is greater than the difference threshold, then will be fast
Thresholding constant false alarm valuation thresholding is detected the constant false alarm thresholding y of unit as thisp, otherwise, by slow thresholding constant false alarm valuation thresholding
The constant false alarm thresholding y of unit is detected as thisp；
(4f) enables the value of p add 1, and is repeated in sub-step (4b) to sub-step (4e), to obtain first of pulse pressure rear mold
The corresponding constant false alarm thresholding of P distance unit for including in value signal.
Step 5, using the clutter map valuation thresholding and the constant false alarm thresholding to first of pulse pressure rear mold value signal
It is normalized and weights, obtain normalization weighting modulus value signal.
Step 5 specifically includes following sub-step:
It include P distance unit, x in (5a) first of pulse pressure rear mold value signalpAfter the pulse pressure for indicating p-th of distance unit
Modulus value signal；P=1 ..., P；L ∈ { 1,2 ..., L }；
(5b) is normalized first of pulse pressure rear mold value signal according to clutter map valuation thresholding, obtains clutter map
Valuation thresholding normalizes result na={ x1/u(j-1)1, x2/u(j-1)2..., xp/u(j-1)p... xP/u(j-1)P, wherein u(j-1)pFor thunder
Up to clutter map unit belonging to p-th of distance unit is corresponding miscellaneous in first of pulse pressure rear mold value signal in -1 scan period of jth
Wave figure valuation thresholding；
(5c) is normalized first of pulse pressure rear mold value signal according to constant false alarm thresholding, obtains constant false alarm thresholding
Normalize result nc={ x1/y1, x2/y2..., xp/yp... xP/yP, wherein ypFor pth in first of pulse pressure rear mold value signal
The corresponding constant false alarm thresholding of a distance unit；
(5d) normalizes result n to the clutter map valuation thresholding respectivelyaResult n is normalized with the constant false alarm thresholdingcInto
Row weighting processing obtains clutter map valuation normalization weighted signal ACP and constant false alarm valuation normalization weighted signal CFAR:
ACP=(a/ (20 × lg σ)) × na
CFAR=b × nc
Wherein, a and b is the constant coefficient of setting, and σ is and clutter map valuation thresholding u(j-1)pCorresponding standard deviation estimate value；
(5e) preset signals amplitude threshold T makees clutter map valuation normalization weighted signal ACP if CFAR-ACP > T
Modulus value signal is weighted for normalization, otherwise, using constant false alarm valuation normalization weighted signal CFAR as normalization weighting modulus value letter
Number.
Step 6, it enables the value of l add 1, and is repeated in and executes step 3 to step 5, to obtain L normalization weighting modulus value
Signal, the signal after described L normalization weighting modulus value signal to be overlapped to obtain pulse accumulation, l=1 ..., L.
Specifically: if the signal of a certain distance unit is relatively steady, amplitude should be accumulated by, and otherwise just having can
It can be extra large spike, the amplitude of extra large spike should be suppressed.Non-inherent accumulation side is directly carried out relative to traditional normalized signal
Method carries out non-inherent accumulation again after being weighted to normalized signal, can effectively inhibit extra large spike bring false-alarm, and accumulate
Signal energy.
What non-inherent accumulation utilized is target echo correlation this weak characteristic of ambient noise correlation by force, can be effective
Improve echo-signal signal-to-noise ratio.
Step 7, detection threshold is set, element each in the signal after the pulse accumulation and the detection threshold are carried out
Compare, if some element is greater than the detection threshold in the signal after the pulse accumulation, then it is assumed that distance belonging to the element
Element memory is in target.
Effect of the present invention is described further and is analyzed below with reference to MATLAB emulation experiment
(1), simulated conditions:
For convenience, the present invention is emulated using the data after AD sampling, this batch data is adopted by exclusive data
Storage collects, and is copied to PC machine hard disk and saves, and emulation input data length is 11 radar scanning circles, and every circle includes
4000 PRT, each PRT data are 5000 sampled points, and PRT beginning includes 40 points of frame head+pattern-word, data 4960
Point.
(2), emulation content:
According to above-mentioned steps and simulated conditions, emulation experiment, letter are carried out according to the present invention in software MATLAB 2015b
It is described as follows:
1., to length be 4960 points data carry out process of pulse-compression.It is quick to 8192 points of input data progress first
Fourier transformation FFT obtains the radar echo signal of frequency domain, is that 600 pulse compression coefficients carry out 8192 point FFT by length；It will input
The FFT result of the data and FFT result of pulse compression coefficient is multiplied again and Fast Fourier Transform Inverse is believed after obtaining pulse compression
Number.Result is truncated with 4400 points of length.As a result modulus processing is carried out.
2., determine that clutter map unit size, the size of clutter map unit azimuth-range are radar resolution cell side respectively
Position and the integral multiple apart from size.Radar coverage orientation is bisected into 500 equal portions for 360 degree, the distance unit of sampling is put down
It is divided into 2200 equal portions, i.e., the number of samples that each clutter unit includes is 8*2=16 point.According to clutter area attribute, to 10 circle data
It is updated according to the clutter map that updating factor 1/8 carries out corresponding criterion, calculates clutter map valuation thresholding and variance.
3., the fast thresholding constant false alarm thresholding estimator of design, 1 protection location of unit to be detected or so Ge Liu, then left and right to
It is outer to select 16 reference units, it carries out cell-average and selects macrooperation, obtain fast thresholding valuation；
Slow thresholding constant false alarm thresholding estimator is designed, stays 1 protection location on the left of unit to be detected, then select 128 to the left
Reference unit carries out cell-average operation, obtains slow thresholding valuation；
Thresholding S is set as 10000, by the difference absolute value of fast, slow thresholding valuation compared with thresholding, final thresholding is chosen and estimates
Value.
4., carry out signal normalization operation, set simulation parameter a=48, b=1, thresholding T=1, carry out signal weighting；
5., carry out 28 pulses non-inherent accumulation；
6., setting detection threshold F=60 non-coherent result size compared with thresholding thought that there are targets greater than thresholding.
(3), analysis of simulation result:
It is the value of land clutter area pulse compression result modulus referring to Fig. 5 a.Referring to Fig. 5 b, for sea clutter area pulse compression knot
The value of fruit modulus.It is the value of noise range pulse compression result modulus referring to Fig. 5 c.In Fig. 5 a, 5b, 5c, horizontal axis indicates points, indulges
The logarithm of axis expression range value.
As can be seen that possible very close or even clutter the Time Domain Amplitude of amplitude is wanted in echo signal and the time domain of clutter
More than signal amplitude.If directly differentiated, it be easy to cause a large amount of false-alarms.
Modulus value referring to Fig. 6 a, after 10 circles are updated for land clutter area clutter map；Referring to Fig. 6 b, more for sea clutter area clutter map
Modulus value after new 10 circle；Modulus value referring to Fig. 6 c, after 10 circles are updated for noise range clutter map；In Fig. 6 a, 6b, 6c, horizontal axis is indicated
Points, the longitudinal axis indicate the logarithm of range value.
Can be seen that land clutter from Fig. 6 a, 6b, 6c, the amplitude of sea clutter is generally relatively strong, it rises and falls also more acutely, noise range
Amplitude it is weaker, amplitude also has certain fluctuating, but fluctuating value is relatively small；
It is the slow thresholding constant false alarm valuation in land clutter area referring to Fig. 7 a；Referring to Fig. 7 b, estimate for the fast thresholding constant false alarm in land clutter area
Value；Referring to Fig. 7 c, constant false alarm valuation after adaptively being chosen for land clutter area.In Fig. 7 a, 7b, 7c, horizontal axis indicates points, longitudinal axis table
Show the logarithm of range value.
As can be seen from Fig., the thresholding that land clutter area finally adaptively chooses largely is the fast thresholding valuation of constant false alarm, is said
It is reasonable that bright thresholding, which is set as 10000,.
It is the slow thresholding constant false alarm valuation in sea clutter area referring to Fig. 8 a；Referring to Fig. 8 b, estimate for the fast thresholding constant false alarm in sea clutter area
Value；Referring to Fig. 8 c, constant false alarm valuation after adaptively being chosen for sea clutter area.In Fig. 8 a, 8b, 8c, horizontal axis indicates points, longitudinal axis table
Show the logarithm of range value.
As can be seen from Fig., the thresholding that noise range is finally adaptively chosen largely is the fast thresholding valuation of constant false alarm, explanation
It is reasonable that thresholding, which is set as 10000,.
It is the slow thresholding constant false alarm valuation in noise range referring to Fig. 9 a；It is the fast thresholding constant false alarm valuation in noise range referring to Fig. 9 b；
Referring to Fig. 9 c, constant false alarm valuation after adaptively being chosen for noise range.In Fig. 9 a, 9b, 9c, horizontal axis indicates points, and the longitudinal axis indicates width
The logarithm of angle value.
As can be seen from Fig., the thresholding that noise range is finally adaptively chosen largely is the slow thresholding valuation of constant false alarm, explanation
It is reasonable that thresholding, which is set as 10000,.
0a referring to Fig.1 normalizes level for land clutter area clutter map；0b referring to Fig.1 normalizes for sea clutter area clutter map
Level；0c referring to Fig.1 is that the trivial clutter map of noise normalizes level；In Figure 10 a, 10b, 10c, horizontal axis indicates points, longitudinal axis table
Show range value.
1a referring to Fig.1 normalizes level for land clutter area constant false alarm；1b referring to Fig.1 normalizes for sea clutter area constant false alarm
Level；1c referring to Fig.1 is that the trivial constant false alarm of noise normalizes level；In Figure 11 a, 11b, 11c, horizontal axis indicates points, longitudinal axis table
Show range value.
2a referring to Fig.1 is the adaptively selected rear weight level in land clutter area；2b referring to Fig.1 adaptively selects for sea clutter area
Select rear weight level；2c referring to Fig.1 is the adaptively selected rear weight level in noise range；In Figure 12 a, 12b, 12c, horizontal axis is indicated
Points, the longitudinal axis indicate range value.
As can be seen from Fig.: extra large spike is effectively suppressed, and is conducive to subsequent non-inherent accumulation.
3a referring to Fig.1 is the non-coherent result in land clutter area；3b referring to Fig.1 is the non-coherent result in sea clutter area；Referring to figure
13c is the non-coherent result in noise range；In Figure 13 a, 13b, 13c, horizontal axis indicates points, and the longitudinal axis indicates range value.
As can be seen that after treatment, signal is detected, and the application and adaptive threshold selection of fine clutter map,
Effectively control false alarm rate.
In conclusion emulation experiment demonstrates correctness of the invention, validity and reliability.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range；In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to encompass within these modification and variations.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.
Claims (3)
1. a kind of adaptive targets detection method based on fine clutter map, which is characterized in that described method includes following steps:
Step 1, the continuous L of radar emission postimpulse target echo signals are obtained, arteries and veins is successively carried out to L target echo signal
Compression processing is rushed, the signal after obtaining L process of pulse-compression；
Step 2, modulus operation is carried out to the signal after the L process of pulse-compression, obtains L pulse pressure rear mold value signal；
Step 3, the clutter map for obtaining target echo signal carries out more according to parameter of the L pulse pressure rear mold value signal to clutter map
Newly, the parameter of the clutter map includes clutter map valuation thresholding and estimate of variance；
Step 4, slow thresholding constant false alarm valuation is carried out according to first of pulse pressure rear mold value signal, obtains slow thresholding constant false alarm valuation door
Limit；Fast thresholding constant false alarm valuation is carried out according to first of pulse pressure rear mold value signal, obtains fast thresholding constant false alarm valuation thresholding；And root
According to the slow thresholding constant false alarm valuation thresholding and the fast thresholding constant false alarm valuation thresholding, constant false alarm thresholding is determined；
The step 4 specifically includes following sub-step:
It include P distance unit, x in (4a) first of pulse pressure rear mold value signalpModulus value is believed after indicating the pulse pressure of p-th of distance unit
Number；P=1 ..., P；l∈{1,2,...,L}；
(4b) using p-th of distance unit as detected unit, the initial value of p is 1；
T1 distance unit on the left of detected unit is set as protection location by (4c), is successively chosen on the left of the protection location
Continuous C1 reference unit, using the average value of the pulse pressure rear mold value signal of C1 reference unit as the detected unit
Slow thresholding constant false alarm valuation thresholding；If the number of reference unit is less than C1, the slow thresholding constant false alarm valuation thresholding is set
It is zero；T1≥1,C1≥1；
T2 distance unit on the left of detected unit is set as left protection location by (4d), by the T2 on the right side of detected unit away from
It is set as right protection location from unit, C2 left reference units are chosen on the left of the left protection location, from the right protection location
C2 right reference units are chosen on right side, and the pulse pressure rear mold value signal of C2 left reference units is averaged, and it is average to be denoted as first
The pulse pressure rear mold value signal of C2 right reference units is averaged, is denoted as the second average value by value；And more described first is average
The size of value and second average value, using biggish average value as the fast thresholding constant false alarm valuation door of the detected unit
Limit；It is permanent empty using second average value as the fast thresholding of the detected unit if the number of left reference unit is less than C2
Alert valuation thresholding, if the number of right reference unit is less than C2, using first average value as the fast of the detected unit
Thresholding constant false alarm valuation thresholding；
(4e) makees fast thresholding constant false alarm valuation thresholding if the distance unit number being detected on the left of unit is less than C1+T1
The constant false alarm thresholding y of unit is detected for thisp, otherwise calculate the slow thresholding constant false alarm valuation thresholding of the detected unit and fast
The absolute value of the difference of thresholding constant false alarm valuation thresholding, and difference threshold is preset, if the slow thresholding of the detected unit
The absolute value of the difference of constant false alarm valuation thresholding and fast thresholding constant false alarm valuation thresholding is greater than the difference threshold, then by fast thresholding
Constant false alarm valuation thresholding is detected the constant false alarm thresholding y of unit as thisp, otherwise, using slow thresholding constant false alarm valuation thresholding as
This is detected the constant false alarm thresholding y of unitp；
(4f) enables the value of p add 1, and is repeated in sub-step (4b) to sub-step (4e), believes to obtain modulus value after first of pulse pressure
The corresponding constant false alarm thresholding of P distance unit for including in number；
Step 5, first of pulse pressure rear mold value signal is carried out using the clutter map valuation thresholding and the constant false alarm thresholding
It normalizes and weights, obtain normalization weighting modulus value signal；
Step 6, it enables the value of l add 1, and is repeated in and executes step 3 to step 5, to obtain L normalization weighting modulus value letter
Number, the signal after described L normalization weighting modulus value signal to be overlapped to obtain pulse accumulation, l=1 ..., L；
Step 7, detection threshold is set, element each in the signal after the pulse accumulation is compared with the detection threshold
Compared with if some element is greater than the detection threshold in signal after the pulse accumulation, then it is assumed that distance list belonging to the element
There are targets in member.
2. a kind of adaptive targets detection method based on fine clutter map according to claim 1, which is characterized in that step
Rapid 3 specifically include following sub-step:
The scanning range of radar is carried out the equal sector division of azimuth-range by (3a), obtains M × N number of clutter map unit, wherein M
It indicates for the 360 degree of scannings range of radar to be divided into M parts in orientation, N indicates in distance to divide the scanning range of radar upwards
It is N parts；Wherein M > 1, N > 1；Set the initial value u of the clutter map valuation thresholding of each clutter map unit0, each clutter map unit
The initial value σ of estimate of variance0 2；
(3b) obtains the clutter attribute of i-th of clutter map unit, and the clutter attribute is divided into two kinds, and one kind is land clutter and makes an uproar
Sound, another kind are sea clutter；The initial value of i is 1, i=1 ..., M × N；
(3c) obtains i-th of clutter map list if the clutter attribute of i-th of clutter map unit is land clutter and noise
The maximum value X1 for the first pulse pressure rear mold value signal for including in member, the first pulse pressure rear mold value signal are modulus value after L pulse pressure
The signal in i-th of clutter map unit is fallen into signal, and using following more new formula to the miscellaneous of i-th of clutter map unit
Wave figure valuation thresholding and estimate of variance are updated:
uj=k1 × uj-1+(1-k1)×X1
σj 2=k2 × σj-1 2+(1-k2)×[uj-X1]2
Wherein, j indicates that the current scan period number of radar, j >=1, k1 and k2 are preset iteration factor, X1=max (x1,…,
xE), E is the data amount check for the pulse pressure rear mold value signal for including in i-th of clutter map unit；
(3d) includes in acquisition i-th of clutter map unit if the clutter attribute of i-th of clutter map unit is sea clutter
The second pulse pressure rear mold value signal average value X2, the second pulse pressure rear mold value signal is the decline of L pulse pressure rear mold value signal
Enter the signal in i-th of clutter map unit, and the clutter map valuation using following more new formula to i-th of clutter map unit
Thresholding and estimate of variance are updated:
uj=k3 × uj-1+(1-k3)×X2
Wherein, j indicates that the current scan period number of radar, j >=1, k3 and k4 are preset iteration factor,E is the pulse pressure rear mold value signal for including in i-th of clutter map unit
Data amount check；
(3e) successively carries out sub-step (3b) to sub-step (3d) to each clutter map unit in M × N number of clutter map unit, from
And obtain the clutter map valuation thresholding and estimate of variance of each clutter map unit.
3. a kind of adaptive targets detection method based on fine clutter map according to claim 1, which is characterized in that step
Rapid 5 specifically include following sub-step:
It include P distance unit, x in (5a) first of pulse pressure rear mold value signalpModulus value is believed after indicating the pulse pressure of p-th of distance unit
Number；P=1 ..., P；l∈{1,2,...,L}；
(5b) is normalized first of pulse pressure rear mold value signal according to clutter map valuation thresholding, obtains clutter map valuation
Thresholding normalizes result na={ x1/u(j-1)1,x2/u(j-1)2,...,xp/u(j-1)p,...xP/u(j-1)P, wherein u(j-1)pFor radar
The corresponding clutter of clutter map unit belonging to p-th of distance unit in first of pulse pressure rear mold value signal in -1 scan period of jth
Figure valuation thresholding；
(5c) is normalized first of pulse pressure rear mold value signal according to constant false alarm thresholding, obtains constant false alarm thresholding normalizing
Change result nc={ x1/y1,x2/y2,...,xp/yp,...xP/yP, wherein ypFor p-th in first of pulse pressure rear mold value signal away from
Constant false alarm thresholding corresponding from unit；
(5d) normalizes result n to the clutter map valuation thresholding respectivelyaResult n is normalized with the constant false alarm thresholdingcAdded
Power processing obtains clutter map valuation normalization weighted signal ACP and constant false alarm valuation normalization weighted signal CFAR:
ACP=(a/ (20 × lg σ)) × na
CFAR=b × nc
Wherein, a and b is the constant coefficient of setting, and σ is and clutter map valuation thresholding u(j-1)pCorresponding standard deviation estimate value；
(5e) preset signals amplitude threshold T, if CFAR-ACP > T, using clutter map valuation normalization weighted signal ACP as returning
One changes weighting modulus value signal, otherwise, weights modulus value signal using constant false alarm valuation normalization weighted signal CFAR as normalization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610570313.XA CN106291492B (en) | 2016-07-19 | 2016-07-19 | A kind of adaptive targets detection method based on fine clutter map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610570313.XA CN106291492B (en) | 2016-07-19 | 2016-07-19 | A kind of adaptive targets detection method based on fine clutter map |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106291492A CN106291492A (en) | 2017-01-04 |
| CN106291492B true CN106291492B (en) | 2019-07-09 |
Family
ID=57651710
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610570313.XA Active CN106291492B (en) | 2016-07-19 | 2016-07-19 | A kind of adaptive targets detection method based on fine clutter map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106291492B (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107132532B (en) * | 2017-06-13 | 2019-12-10 | 西安电子科技大学 | Small target detection method based on sea peak inhibition and multi-frame combination |
| CN107255804B (en) * | 2017-07-31 | 2020-04-10 | 西安电子科技大学 | Clutter contour map-based moving target display filter filtering method |
| CN107748357B (en) * | 2017-09-01 | 2019-09-10 | 中国电子科技集团公司第二十八研究所 | A kind of radar fine day G- Design method of FIR filter |
| CN107678008A (en) * | 2017-09-07 | 2018-02-09 | 西安电子工程研究所 | A kind of plural clutter map CFAR detection method |
| CN108562879B (en) * | 2018-04-18 | 2022-03-22 | 南京理工大学 | Constant false alarm detection method for ship-borne radar based on FPGA |
| CN109782248B (en) * | 2019-02-18 | 2022-09-13 | 中国电子科技集团公司第二十研究所 | Radar clutter processing method |
| CN109901145A (en) * | 2019-03-19 | 2019-06-18 | 安徽雷炎电子科技有限公司 | Sea clutter elimination method based on fine hodograph |
| CN109814071B (en) * | 2019-03-19 | 2023-08-29 | 安徽雷炎电子科技有限公司 | Regional threshold rapid convergence method based on complex environment and application thereof |
| CN110082754B (en) * | 2019-04-19 | 2021-09-17 | 北京海兰信数据科技股份有限公司 | Weak target self-adaptive detection system and detection method based on tracking feedback |
| CN111123212B (en) * | 2019-12-11 | 2023-11-17 | 安徽四创电子股份有限公司 | Signal processing method of scene surveillance radar based on complex clutter background |
| CN111175706B (en) * | 2020-01-08 | 2022-05-27 | 中国船舶重工集团公司第七二四研究所 | Interpolation clutter map method based on local threshold judgment |
| CN111398911B (en) * | 2020-03-24 | 2022-03-08 | 中国人民解放军海军航空大学 | MIMO radar target detection method and device |
| CN112098952B (en) * | 2020-08-19 | 2022-11-08 | 中国电子科技集团公司第二十九研究所 | Radar reconnaissance clutter suppression method based on time domain statistical processing |
| CN112180341B (en) * | 2020-09-29 | 2022-05-17 | 中国船舶重工集团公司第七二四研究所 | Method for realizing selection of background self-adaptive CFAR algorithm |
| CN113093180B (en) * | 2021-03-20 | 2023-10-13 | 中国人民解放军空军工程大学 | Clustering extrapolation-based slow weak target detection method |
| CN113253231B (en) * | 2021-05-17 | 2023-10-13 | 成都西科微波通讯有限公司 | Clutter map detection and update method based on one-dimensional range profile features |
| CN114415123B (en) * | 2022-04-01 | 2022-07-19 | 北京海兰信数据科技股份有限公司 | Non-coherent neighborhood based weighting pulse accumulation processing method and system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103246556A (en) * | 2013-05-21 | 2013-08-14 | 中国人民解放军防空兵学院 | General performance simulation method and framework of air surveillance radar |
-
2016
- 2016-07-19 CN CN201610570313.XA patent/CN106291492B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103246556A (en) * | 2013-05-21 | 2013-08-14 | 中国人民解放军防空兵学院 | General performance simulation method and framework of air surveillance radar |
Non-Patent Citations (6)
| Title |
|---|
| Design of adaptive broadband beamformer with fixed feedback weights;WU Kai等;《Journal of Jilin University(Engineering and Technology Edition)》;20160131;第26卷(第1期);第303-310页 |
| 一种双模式双参数杂波图恒虚警检测器的实现;林彦 等;《系统工程与电子技术》;20031231;第25卷(第2期);第133-136页 |
| 一种新的高速多目标检测及参数估计方法;郑纪彬 等;《西安电子科技大学学报（自然科学版）》;20130430;第40卷(第2期);第82-88页 |
| 一种机载雷达信号处理系统的实现;柳兵 等;《火控雷达》;20070331;第36卷(第1期);第66-78页 |
| 恒虚警检测器中的双门限方法;方云;《现代电子》;20001231(第3期);第30-39页 |
| 毫米波汽车防撞雷达恒虚警门限设定方法;蒋特珍 等;《红外与毫米波学报》;20050630;第24卷(第3期);第217-220页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106291492A (en) | 2017-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106291492B (en) | A kind of adaptive targets detection method based on fine clutter map | |
| CN107861107B (en) | Double-threshold CFAR (computational fluid dynamics) and trace point agglomeration method suitable for continuous wave radar | |
| CN111352102B (en) | Multi-target number detection method and device based on frequency modulation continuous wave radar | |
| CN204495996U (en) | broadband sonar receiver | |
| US8816899B2 (en) | Enhanced target detection using dispersive vs non-dispersive scatterer signal processing | |
| CN106127110B (en) | A kind of human body fine granularity motion recognition method based on UWB radar and optimal SVM | |
| CN104569948B (en) | Sub-band adaptive GLRT LTD detection methods under sea clutter background | |
| CN109031287A (en) | ULTRA-WIDEBAND RADAR human body respiration signal detecting method through walls based on Faster-RCNN network | |
| CN110286368A (en) | A kind of Falls Among Old People detection method based on ULTRA-WIDEBAND RADAR | |
| CN104267388B (en) | One is target clutter map detection method at a slow speed | |
| CN109633629A (en) | Terahertz frequency range single rotor unmanned plane target characteristic micro-Doppler feature extracting method | |
| IL111543A (en) | Radar return signal processing method | |
| KR102275960B1 (en) | System and method for searching radar targets based on deep learning | |
| CN109188385A (en) | High speed weak object detection method under clutter background | |
| Wang et al. | Hand gesture recognition using radial and transversal dual micromotion features | |
| CN109116326A (en) | A kind of adaption radar ocean clutter cancellation method based on medion estimator | |
| CN108414992A (en) | A kind of object detection method based on phase information clutter map | |
| CN109765529A (en) | A kind of millimetre-wave radar anti-interference method and system based on digital beam froming | |
| CN112327286B (en) | Method, device, equipment and storage medium for classifying daily activities under low complexity | |
| CN106950546B (en) | The non-homogeneous clutter suppression method weighted again based on mahalanobis distance | |
| CN106199539A (en) | Ground bounce removal method based on prewhitening filter | |
| CN109507654A (en) | Phase information calculation method under a kind of complex environment based on LS | |
| CN106324576B (en) | Radar target adaptive matched filter method based on autoregression model | |
| CN112255607A (en) | Sea clutter suppression method | |
| CN111044996A (en) | LFMCW radar target detection method based on dimension reduction approximate message transfer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220628 Address after: 215300 floor 3, building 7, No. 826 Xintang Road, Yushan Town, Suzhou City, Jiangsu Province Patentee after: Liguo Intelligent Technology (Kunshan) Co.,Ltd. Address before: 710071 No. 2 Taibai South Road, Shaanxi, Xi'an Patentee before: XIDIAN University |