CN102141609A - Method and system for recognizing high-band first-order Bragg peak and split spectrum peak thereof - Google Patents
Method and system for recognizing high-band first-order Bragg peak and split spectrum peak thereof Download PDFInfo
- Publication number
- CN102141609A CN102141609A CN 201010609804 CN201010609804A CN102141609A CN 102141609 A CN102141609 A CN 102141609A CN 201010609804 CN201010609804 CN 201010609804 CN 201010609804 A CN201010609804 A CN 201010609804A CN 102141609 A CN102141609 A CN 102141609A
- Authority
- CN
- China
- Prior art keywords
- peak
- spectrum
- splitting
- splitting spectrum
- bragg peak
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention provides a method and system for recognizing a high-band first-order Bragg peak and a split spectrum peak thereof, and relates to a method and system for recognizing a first-order Bragg peak and a split spectrum peak thereof, which solves the problems that a convention method easily generates bad recognition effect of the Bragg peak, low measurement precision and unavailable recognition of the split spectrum peak under complicated clutter/target background and variable marine environment. The method comprises the following steps of: extracting trend features of the first-order Bragg peak by a method for detecting multi-scale filtering and ridge features, recognizing the Bragg peak by combining overall features of amplitude, continuity, symmetry and the like, and recognizing features of a Bragg split spectrum peak caused by severe-shear ocean currents. The system is used for calculating trend information by a trend information extraction module first, constructing a detection band by combining the ocean currents, extracting peak values and recognizing the split spectrum peak by a split spectrum peak recognition module. The method and the system are suitable for obtaining the ocean information in ocean environment sensing, and the process of rejecting ocean clutters in the target detection.
Description
Technical field
The present invention relates to the recognition methods and the system at a kind of single order Bragg peak and splitting spectrum peak thereof.
Background technology
In extra large attitude environmental remote sensing was used, high frequency over the horizon radar was mainly by obtaining extra large attitude information to the identification of single order, second order marine echo and the parameter estimation of doppler spectral.On the other hand, in testing process to moving-targets such as naval vessel, icebergs, when the speed of target equals or during near the position of single order sea clutter spectrum (Bragg peak), thereby usually owing to can't accurately distinguish target and extra large clutter causes false-alarm and false dismissal.As seen, realizing disturbing the Bragg peak identification under the contaminated environment is the key that addresses the above problem.
Usually, in the water front off sea/in-plant marine site because the Bragg peak is that the part is dominant on amplitude, so the local peak detection method has obtained widespread use.But when existing Ship Target, ionospheric clutter etc. to disturb near the Bragg peak, the performance of this method will descend greatly.Characteristic recognition method has been taken all factors into consideration the multidimensional features such as amplitude/signal to noise ratio (S/N ratio), continuity and symmetry at Bragg peak, has proposed based on Bragg peak Feature Recognition criterion, preferably resolves the identification problem at Bragg peak.Nearest discovers, when existing unusual sudden change (as strong ocean current shear, wind shear, surge, whirlpool etc.) to disturb in the detecting unit, phenomenons such as unusual skew, expansion and division can appear in the spectrum structure at Bragg peak, at this moment, method based on local maximal peak will be lost the Bragg peak energy, and may cause sea-surface target false-alarm to occur in detecting; Weighted-average method is a cost with the loss measuring accuracy then, the ocean current information that can only obtain to be similar to, and may cause and occur false dismissal in the target detection.
Summary of the invention
The present invention is poor for the Bragg peak recognition effect that solves classic method and occur under complicated clutter/target background and the changeable condition of marine environment, measuring accuracy is low, and the problem that can't discern the splitting spectrum peak, thereby provide the recognition methods and the system at a kind of high band single order Bragg peak and splitting spectrum peak thereof.
The recognition methods at a kind of high band single order Bragg peak and splitting spectrum peak thereof, it is realized by following steps:
Step 1, employing high frequency over the horizon radar are measured extra large attitude information, obtain to comprise the echo spectrum of sea area to be detected information;
The method of step 2, employing multi-scale filtering is carried out level and smooth pre-service to the echo spectrum that step 1 obtains, and obtains level and smooth pretreated echo spectrum;
Step 3, the method that adopts ridge to detect obtain level and smooth pretreated echo spectrum to step 2 and detect, and obtain to comprise the testing result at Bragg peak;
Step 4, adopt the jamming target in the testing result that obtains to comprise the Bragg peak in the caustic solution filtering step 3 in the morphological image, and with the tendency information of the result after the filtering interfering target as the Bragg peak;
The tendency information at step 5, the Bragg peak that obtains according to step 4 is determined the initial survey position, and is provided with in the Doppler territory of described initial survey position and detects band, chooses the peak value that is positioned at the Bragg peak that detects band and composes the peak as alternative single order; The width of described detection band is determined according to ocean current speed;
Step 6, the alternative single order spectrum peak that step 5 is chosen carry out the symmetry coupling, by the position at the positive and negative Bragg peak after the coupling is contrasted, select to satisfy the recognition result of the peak value of symmetrical deflection as single order Bragg peak from alternative target;
The single order Bragg peak that obtains in the step 5 continuity in the distance dimension is confirmed as a reference in the forward and backward unit of the recognition result at step 7, the single order Bragg peak that obtains according to step 6, obtains N the alternative point in Bragg peak; Described N is a positive integer;
Step 8, the alternative point in N Bragg peak that step 7 is obtained adopt the method for cubic spline interpolation to carry out smoothing processing, obtain the Bragg peak that the continuously smooth on the distance dimension changes;
Step 9, the local big criterion of employing and offset scope are discerned the splitting spectrum peak position after obtaining to discern to the splitting spectrum peak position in the level and smooth pretreated echo spectrum;
Step 10, adopt the Bragg peak of the continuously smooth variation on the distance dimension that splitting spectrum peak position and step 8 after the identification that the amplitude difference criterion obtains step 9 obtain to compare, obtain the splitting spectrum amplitude difference, with the point beyond the described splitting spectrum amplitude difference as noise spot, and with described noise spot filtering, the splitting spectrum behind the acquisition filtering interfering point;
Splitting spectrum behind step 11, the filtering interfering point that step 10 obtained according to symmetry principle carries out symmetry coupling and confirms that splitting spectrum is blured in acquisition;
Step 12, the fuzzy splitting spectrum that step 11 is obtained carry out the feature affirmation, obtain the final position at the splitting spectrum peak at Bragg peak.
The echo spectrum that the method for employing multi-scale filtering obtains step 1 in the step 2 carries out level and smooth pretreated method and is:
To three dimensions: f:R
2* → R, its metric space: L:R
2* R
+→ R can be expressed as: L (; T)=g (; T) * f, wherein: * represents convolution, g:R
2* R
+→ R is a gaussian kernel function, and t is a scale parameter, and (x y) is the Cartesian coordinates of certain point in the two-dimensional space;
The derivative of metric space is defined as:
Wherein, α, β are the exponent numbers of differential;
By changing scale parameter t, obtain the different three-dimensional mapping of smothing filtering degree; And then realization is to the level and smooth pre-service of echo spectrum.
Adopt method that ridge detects that step 2 is obtained during level and smooth pretreated echo spectrum detects in the step 3, the preparation method of the ridge point in the ridge testing process is: at first cause the side-play amount at Bragg peak to provide a ridge sensing range according to maximum ocean current in the sea area to be detected; Then in this scope, (v r) asks for the single order of tieing up the direction metric space along Doppler and leads the result to each point in the echo spectrum after step 2 is handled
Lead the result with second order
And when satisfying:
The time, then (v, r) as the ridge point, v and r represent speed and the range coordinate in the RD spectrum respectively with the point in the echo spectrum.
The implication of the big criterion in the part described in the step 9 is: the outer local maximum sharpness of definite division main peak is the splitting spectrum peak.
Offset scope described in the step 9 be: [v
m-Δ
Max, v
m+ Δ
Max], v wherein
mBe the doppler velocity position of main peak, Δ
MaxIt is the radially Bragg peak maximum offset that causes of ocean current of stable state/transient state.
The implication of symmetry principle described in the step 11 is: the skew of splitting spectrum is consistent in the same way.
The implication of the criterion of amplitude difference described in the step 10 is: the amplitude difference at main peak and its splitting spectrum peak exists: between 0～20dB.
The fuzzy splitting spectrum that described in the step 12 step 11 is obtained carries out feature and confirms, the method of the final position at the splitting spectrum peak at acquisition Bragg peak is: investigate the invisible side main peak in splitting spectrum peak spectrum structure on every side, when the spectrum structure shows as the feature that adjacent weak target disturbs, judge then that splitting spectrum exists but invisible; Comprehensive then bilateral segmentation spectrum peak information, thereby the final position at the splitting spectrum peak at acquisition Bragg peak.
Realize a kind of high band single order Bragg peak of said method and the recognition system at splitting spectrum peak thereof, it comprises peak detection block and alternative feature as a result affirmation module, it also comprises tendency information extraction module and splitting spectrum peak identification module, and described tendency information extraction module is made up of multi-scale filtering unit, ridge detecting unit and shape filtering unit; Splitting spectrum peak identification module is by splitting spectrum peak recognition unit, splitting spectrum amplitude difference confirmation unit, splitting spectrum symmetry matching unit and fuzzy splitting spectrum feature confirmation unit, the input end of described multi-scale filtering unit receives sea area to be detected information echo spectrum information, and the output terminal of multi-scale filtering unit is connected with the input end of ridge detecting unit; The output terminal of ridge detecting unit is connected with the input end of shape filtering unit; The output terminal of shape filtering unit is connected with the input end of peak detection block; The output terminal of peak detection block confirms that with alternative feature as a result the input end of module is connected; Alternative feature as a result confirms that the output terminal of module is connected with the input end of splitting spectrum peak recognition unit; The output terminal of splitting spectrum peak recognition unit is connected with the input end of splitting spectrum amplitude difference confirmation unit; The output terminal of splitting spectrum amplitude difference confirmation unit is connected with the input end of splitting spectrum symmetry matching unit; The output terminal of splitting spectrum symmetry matching unit is connected with the input end of fuzzy splitting spectrum feature confirmation unit; Alternative feature is as a result confirmed the recognition result of the recognition result of module output as single order Bragg peak; The recognition result of fuzzy splitting spectrum feature confirmation unit output is as the recognition result at the splitting spectrum peak at Bragg peak.
Beneficial effect: the present invention a kind ofly utilizes characteristics of image in the high frequency over the horizon radar echo spectrum as the indication information of Bragg peak position, and the method for carrying out the identification of Bragg peak in conjunction with the experimental knowledge criterion.This method has not only effectively improved the identification probability at single order Bragg peak under the multiple noise jamming, and splitting spectrum peak, the single order Bragg peak feature that causes at strong shear ocean current, the criterion of identification of splitting spectrum has been proposed, meticulousr, extra large attitude information accurately can be provided, and offer help for the sea-surface target under the extra large clutter background detects.
Description of drawings
Fig. 1 is the schematic flow sheet of method of the present invention, and Fig. 2 is the structural representation of system of the present invention.
Embodiment
Embodiment one, this embodiment is described in conjunction with Fig. 1, the recognition methods at a kind of high band single order Bragg peak and splitting spectrum peak thereof, it is realized by following steps:
Step 1, employing high frequency over the horizon radar are measured extra large attitude information, obtain to comprise the echo spectrum of sea area to be detected information;
The method of step 2, employing multi-scale filtering is carried out level and smooth pre-service to the echo spectrum that step 1 obtains, and obtains level and smooth pretreated echo spectrum;
Step 3, the method that adopts ridge to detect obtain level and smooth pretreated echo spectrum to step 2 and detect, and obtain to comprise the testing result at Bragg peak;
Step 4, adopt the jamming target in the testing result that obtains to comprise the Bragg peak in the caustic solution filtering step 3 in the morphological image, and with the tendency information of the result after the filtering interfering target as the Bragg peak;
The tendency information at step 5, the Bragg peak that obtains according to step 4 is determined the initial survey position, and is provided with in the Doppler territory of described initial survey position and detects band, chooses the peak value that is positioned at the Bragg peak that detects band and composes the peak as alternative single order; The width of described detection band is determined according to ocean current speed;
Step 6, the alternative single order spectrum peak that step 5 is chosen carry out the symmetry coupling, by the position at the positive and negative Bragg peak after the coupling is contrasted, select to satisfy the recognition result of the peak value of symmetrical deflection as single order Bragg peak from alternative target;
The single order Bragg peak that obtains in the step 5 continuity in the distance dimension is confirmed as a reference in the forward and backward unit of the recognition result at step 7, the single order Bragg peak that obtains according to step 6, obtains N the alternative point in Bragg peak; Described N is a positive integer;
Step 8, the alternative point in N Bragg peak that step 7 is obtained adopt the method for cubic spline interpolation to carry out smoothing processing, obtain the Bragg peak that the continuously smooth on the distance dimension changes;
Step 9, the local big criterion of employing and offset scope are discerned the splitting spectrum peak position after obtaining to discern to the splitting spectrum peak position in the level and smooth pretreated echo spectrum;
Step 10, adopt the Bragg peak of the continuously smooth variation on the distance dimension that splitting spectrum peak position and step 8 after the identification that the amplitude difference criterion obtains step 9 obtain to compare, obtain the splitting spectrum amplitude difference, with the point beyond the described splitting spectrum amplitude difference as noise spot, and with described noise spot filtering, the splitting spectrum behind the acquisition filtering interfering point;
Splitting spectrum behind step 11, the filtering interfering point that step 10 obtained according to symmetry principle carries out symmetry coupling and confirms that splitting spectrum is blured in acquisition;
Step 12, the fuzzy splitting spectrum that step 11 is obtained carry out the feature affirmation, obtain the final position at the splitting spectrum peak at Bragg peak.
The echo spectrum that the method for employing multi-scale filtering obtains step 1 in the step 2 carries out level and smooth pretreated method and is:
To three dimensions: f:R
2* → R, its metric space: L:R
2* R
+→ R can be expressed as: L (; T)=g (; T) * f, wherein: * represents convolution, g:R
2* R
+→ R is a gaussian kernel function, and t is a scale parameter, and (x y) is the Cartesian coordinates of certain point in the two-dimensional space;
The derivative of metric space is defined as:
Wherein, α, β are the exponent numbers of differential;
By changing scale parameter t, obtain the different three-dimensional mapping of smothing filtering degree; And then realization is to the level and smooth pre-service of echo spectrum.
Adopt method that ridge detects that step 2 is obtained during level and smooth pretreated echo spectrum detects in the step 3, the preparation method of the ridge point in the ridge testing process is: at first cause the side-play amount at Bragg peak to provide a ridge sensing range according to maximum ocean current in the sea area to be detected; Then in this scope, (v r) asks for the single order of tieing up the direction metric space along Doppler and leads the result to each point in the echo spectrum after step 2 is handled
Lead the result with second order
And when satisfying:
The time, then (v, r) as the ridge point, v and r represent speed and the range coordinate in the RD spectrum respectively with the point in the echo spectrum.The implication of the big criterion in the part described in the step 9 is: the outer local maximum sharpness of definite division main peak is the splitting spectrum peak.
Offset scope described in the step 9 be: [v
m-Δ
Max, v
m+ Δ
Max], v wherein
mBe the doppler velocity position of main peak, Δ
MaxIt is the radially Bragg peak maximum offset that causes of ocean current of stable state/transient state.
The implication of symmetry principle described in the step 11 is: the skew of splitting spectrum is consistent in the same way.
The implication of the criterion of amplitude difference described in the step 10 is: the amplitude difference at main peak and its splitting spectrum peak exists: between 0～20dB.
The fuzzy splitting spectrum that described in the step 12 step 11 is obtained carries out feature and confirms, the method of the final position at the splitting spectrum peak at acquisition Bragg peak is: investigate the invisible side main peak in splitting spectrum peak spectrum structure on every side, when the spectrum structure shows as the feature that adjacent weak target disturbs, judge then that splitting spectrum exists but invisible; Comprehensive then bilateral segmentation spectrum peak information, thereby the final position at the splitting spectrum peak at acquisition Bragg peak.
Embodiment two, this embodiment is described in conjunction with Fig. 2, based on a kind of high band single order Bragg peak of the recognition methods at embodiment one described a kind of high band single order Bragg peak and splitting spectrum peak thereof and the recognition methods and the system at splitting spectrum peak thereof, it comprises peak detection block 2 and alternative feature as a result affirmation module 3, it also comprises tendency information extraction module 1 and splitting spectrum peak identification module 4, and described tendency information extraction module 1 is made up of multi-scale filtering unit 11, ridge detecting unit 12 and shape filtering unit 13; Splitting spectrum peak identification module 4 is by splitting spectrum peak recognition unit 41, splitting spectrum amplitude difference confirmation unit 42, splitting spectrum symmetry matching unit 43 and fuzzy splitting spectrum feature confirmation unit 44, the input end of described multi-scale filtering unit 11 receives sea area to be detected information echo spectrum information, and the output terminal of multi-scale filtering unit 11 is connected with the input end of ridge detecting unit 12; The output terminal of ridge detecting unit 12 is connected with the input end of shape filtering unit 13; The output terminal of shape filtering unit 13 is connected with the input end of peak detection block 2; The output terminal of peak detection block 2 confirms that with alternative feature as a result the input end of module 3 is connected; Alternative feature as a result confirms that the output terminal of module 3 is connected with the input end of splitting spectrum peak recognition unit 41; The output terminal of splitting spectrum peak recognition unit 41 is connected with the input end of splitting spectrum amplitude difference confirmation unit 42; The output terminal of splitting spectrum amplitude difference confirmation unit 42 is connected with the input end of splitting spectrum symmetry matching unit 43; The output terminal of splitting spectrum symmetry matching unit 43 is connected with the input end of fuzzy splitting spectrum feature confirmation unit 44; Alternative feature is as a result confirmed the recognition result of the recognition result of module 3 outputs as single order Bragg peak; The recognition result of fuzzy splitting spectrum feature confirmation unit 44 outputs is as the recognition result at the splitting spectrum peak at Bragg peak.
Embodiment three, this embodiment and embodiment two described a kind of high band single order Bragg peaks and the recognition methods at splitting spectrum peak thereof and the difference of system are, it comprises that also blind area detection module 5, SNR confirm module 6 and global characteristics affirmation module 7, and the input end of described blind area detection module 5 is connected with the output terminal of fuzzy splitting spectrum feature confirmation unit 44; The output terminal of blind area detection module 5 confirms that with SNR the input end of module 6 is connected, and SNR confirms that the output terminal of module 6 is connected with the input end of global characteristics affirmation module 7.
The present invention is a kind of high band single order Bragg peak and recognition methods of splitting spectrum peak thereof based on knowledge.The identification process of this method is made up of following five modules: tendency information is extracted, and based on the peak value detection of indication information, alternative feature is as a result confirmed, the identification of splitting spectrum peak, subsequent treatment.
One aspect of the present invention has been used multi-scale filtering and ridge characteristic detection method, thereby has extracted the tendency information of Bragg peak in the RD spectrum; Proposed the splitting spectrum recognition methods on the other hand, meticulousr, extra large attitude information accurately can be provided, and offered help for the sea-surface target under the extra large clutter background detects.
Five modules of the present invention have comprised the whole flow processs among Fig. 1, and are specific as follows: the tendency information extraction module comprises that multi-scale filtering, ridge detect and shape filtering three parts; Peak detection block utilizes tendency information as indication, sets up in conjunction with extra large attitude ocean current and detects band, carries out peak value and detects; Alternative feature as a result confirms that module comprises that the symmetry coupling is confirmed, distance dimension continuity is confirmed; Splitting spectrum peak identification module comprises that the identification of splitting spectrum peak, splitting spectrum amplitude difference are confirmed, the splitting spectrum symmetry is mated and fuzzy splitting spectrum feature is confirmed; The subsequent treatment module comprises that the blind area is detected, SNR confirms and global characteristics is confirmed.
Identification process to Fig. 1 specifically is described below:
One, tendency information extraction module
At first, use the method for multi-scale filtering that the RD spectrum is carried out level and smooth pre-service, purpose is that the high frequency noise in the filtering spectrogram disturbs.To any given three dimensions f:R
2* → R, its metric space L:R
2* R
+→ R can be expressed as:
L(·；t)＝g(·；t)*f(1)
Wherein, * represents convolution, g:R
2* R
+→ R is a gaussian kernel function;
In the formula: t is a scale parameter; (x y) is the Cartesian coordinates of certain point in the two-dimensional space.
The derivative of metric space can be defined as:
Wherein, (α β) is the exponent number of differential.Can obtain the different three-dimensional mapping of smothing filtering degree by changing scale parameter t.
To filtered result, the method that adopts ridge to detect is obtained the Bragg peak: at first cause the side-play amount at Bragg peak to provide a comparatively wide in range ridge sensing range according to the maximum ocean current of search coverage; Then in this scope, (v r) asks for single order and the second order of tieing up the direction metric space along Doppler and leads the result to each point in the RD spectrum after treatment
If:
Judge that then (v r) is the ridge point.
Owing to have interference such as clutter and naval vessel, usually both comprised among the result that said method obtains that the Bragg peak also comprised part jamming target (as ionospheric clutter, Ship Target etc.).And interference sections is relatively independent in the RD chromatic graph, and they are not possessing continuity on distance dimension, therefore can use the in addition filtering of caustic solution in the morphological image.
Two, based on the peak detection block of indication information
Get the Bragg peak skew that ocean current speed maximal value radially causes and be Δ
Max, at initial survey position (v
0, r
i) the Doppler territory be provided with and detect band [v
0-Δ
Max, v
0+ Δ
Max].Choose and detect the interior local peaking of band as alternative target.In the method, its initial survey position (v
0, r
i) be to determine with reference to the tendency information that last joint obtains.
Three, alternative feature is as a result confirmed module
Symmetry coupling: contrast by positive and negative Bragg peak position, from alternative target, select to satisfy the peak value of symmetrical deflection as recognition result.
Distance dimension continuity is confirmed: with the front and back reference unit is basis, investigates in the continuity of distance dimension for single order Bragg to be confirmed peak;
Cubic spline interpolation:, therefore, obtain the secondary indication information of Bragg peak position by the cubic spline function interpolation to a plurality of alternative some situation after confirming through the symmetry coupling because cubic spline function is the simplest smooth curve function.Can obtain the distance dimension by this method and go up the Bragg peak indication information that continuously smooth changes.
Four, splitting spectrum peak identification module
When satisfying certain limit the integration time of high frequency over the horizon radar,, will cause the division at Bragg peak if transient changing appears in object element yardstick inside surface ocean current field structure.The present invention claims that the big peak of amplitude is a main peak in the Bragg peak of the zero any side of Doppler, and another closes on and amplitude is time big is the splitting spectrum peak.Along with the variation of ocean current shear and target resolution element relative position, the relative amplitude of Relative Zero Doppler's splitting spectrum in the same way will have bigger variation, and amplitude differs about about 0～20dB.Find in the actual observation, when the signal to noise ratio (S/N ratio) at splitting spectrum peak hour, the influence that its identification is easy to be interfered.Show as, another survey is invisible in that zero Doppler's a side is visible, claims this phenomenon fuzzy for the division of Bragg peak among the present invention.
According to splitting spectrum mechanism of production and observation experience, can provide the criterion of identification at splitting spectrum peak:
It is local that criterion--the splitting spectrum peak generally still shows as the bigger spike in part except that main peak greatly.
The doppler position at offset scope--main peak and splitting spectrum peak is no more than the radially Bragg peak maximum offset Δ that causes of ocean current of stable state/transient state at interval
MaxTherefore the splitting spectrum hunting zone can be set be [v
m-Δ
Max, v
m+ Δ
Max], v wherein
mDoppler velocity position for main peak.
Criterion--factors such as ocean current shear are symmetrical to the effect at positive and negative Bragg peak to symmetry, so the skew of splitting spectrum also is consistent in the same way.Consider factors such as resolving power, systematic error disturbance, under medium resolution (100～200 seconds integration times) condition, can increase the permissible error of 1 resolution element.
The amplitude difference criterion--can find that by measured data observation the amplitude difference at main peak and splitting spectrum peak is about 0～20dB, the decidable that surpasses this scope is noise jamming.
Fuzzy splitting spectrum peak criterion--because the splitting spectrum peak that noise jamming causes zero Doppler's one side only is as seen.Need to investigate the invisible side main peak in splitting spectrum peak spectrum structure on every side this moment.When the spectrum structure showed as the feature of adjacent weak target interference, the decidable splitting spectrum existed but is invisible.Can judge according to main peak broadening degree in the implementation procedure.
The SNR criterion--reject under the SNR algorithm for estimating of criterion supports having comprised unruly-value rejecting, the main clutter discerned, the SNR at splitting spectrum peak also needs to satisfy certain SNR.Because the splitting spectrum peak that is lower than this value can be ignored for the value and the influence of the detection of the estimation of ocean current and sea-surface target.
In processing, take all factors into consideration above criterion, especially may there be the situation that can not satisfy above criterion simultaneously at the splitting spectrum peak, just more need overall consideration.
Five, subsequent treatment module
Be subjected to clutter and noise effect, the whole Bragg peak information in the expression RD spectrum that tendency information sometimes can not be complete; And the appearance at splitting spectrum peak has also increased the risk that occurs false-alarm in the detection to a certain extent.In detection architecture, increased " blind area detection " and " global characteristics affirmation " two processing procedures for this reason, the former task be investigate do not provide tendency information just/negative zone, Bragg peak, confirm alternative single order spectrum by symmetry then; The latter then uses symmetry and smoothness criterion, and the false-alarm that causes owing to splitting spectrum peak, high clutter and ship target is controlled.Both combine and can address the above problem preferably with other process in the system.
Method of the present invention is single order sea clutter/Bragg peak recognition methods in a kind of effective high-frequency ground wave over-the-horizon radar echo spectrum, this method is by the indication of tendency information, and the identification at the splitting spectrum peak that strong ocean current shear is caused, single order Bragg peak has been obtained higher identification probability.
Claims (10)
1. the recognition methods at high band single order Bragg peak and splitting spectrum peak thereof, it is characterized in that: it is realized by following steps:
Step 1, employing high frequency over the horizon radar are measured extra large attitude information, obtain to comprise the echo spectrum of sea area to be detected information;
The method of step 2, employing multi-scale filtering is carried out level and smooth pre-service to the echo spectrum that step 1 obtains, and obtains level and smooth pretreated echo spectrum;
Step 3, the method that adopts ridge to detect obtain level and smooth pretreated echo spectrum to step 2 and detect, and obtain to comprise the testing result at Bragg peak;
Step 4, adopt the jamming target in the testing result that obtains to comprise the Bragg peak in the caustic solution filtering step 3 in the morphological image, and with the tendency information of the result after the filtering interfering target as the Bragg peak;
The tendency information at step 5, the Bragg peak that obtains according to step 4 is determined the initial survey position, and is provided with in the Doppler territory of described initial survey position and detects band, chooses the peak value that is positioned at the Bragg peak that detects band and composes the peak as alternative single order; The width of described detection band is determined according to ocean current speed;
Step 6, the alternative single order spectrum peak that step 5 is chosen carry out the symmetry coupling, by the position at the positive and negative Bragg peak after the coupling is contrasted, select to satisfy the recognition result of the peak value of symmetrical deflection as single order Bragg peak from alternative target;
The single order Bragg peak that obtains in the step 5 continuity in the distance dimension is confirmed as a reference in the forward and backward unit of the recognition result at step 7, the single order Bragg peak that obtains according to step 6, obtains N the alternative point in Bragg peak; Described N is a positive integer;
Step 8, the alternative point in N Bragg peak that step 7 is obtained adopt the method for cubic spline interpolation to carry out smoothing processing, obtain the Bragg peak that the continuously smooth on the distance dimension changes;
Step 9, the local big criterion of employing and offset scope are discerned the splitting spectrum peak position after obtaining to discern to the splitting spectrum peak position in the level and smooth pretreated echo spectrum;
Step 10, adopt the Bragg peak of the continuously smooth variation on the distance dimension that splitting spectrum peak position and step 8 after the identification that the amplitude difference criterion obtains step 9 obtain to compare, obtain the splitting spectrum amplitude difference, with the point beyond the described splitting spectrum amplitude difference as noise spot, and with described noise spot filtering, the splitting spectrum behind the acquisition filtering interfering point;
Splitting spectrum behind step 11, the filtering interfering point that step 10 obtained according to symmetry principle carries out symmetry coupling and confirms that splitting spectrum is blured in acquisition;
Step 12, the fuzzy splitting spectrum that step 11 is obtained carry out the feature affirmation, obtain the final position at the splitting spectrum peak at Bragg peak.
2. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof, the echo spectrum that it is characterized in that in the step 2 adopting the method for multi-scale filtering that step 1 is obtained carries out level and smooth pretreated method and is:
To three dimensions: f:R
2* → R, its metric space: L:R
2* R
+→ R can be expressed as: L (; T)=g (; T) * f, wherein: * represents convolution, g:R
2* R
+→ R is a gaussian kernel function, and t is a scale parameter, and (x y) is the Cartesian coordinates of certain point in the two-dimensional space;
The derivative of metric space is defined as:
Wherein, α, β are the exponent numbers of differential;
By changing scale parameter t, obtain the different three-dimensional mapping of smothing filtering degree; And then realization is to the level and smooth pre-service of echo spectrum.
3. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof, it is characterized in that in the step 3 adopting method that ridge detects that step 2 is obtained during level and smooth pretreated echo spectrum detects, the preparation method of the ridge point in the ridge testing process is: at first cause the side-play amount at Bragg peak to provide a ridge sensing range according to maximum ocean current in the sea area to be detected; Then in this scope, (v r) asks for the single order of tieing up the direction metric space along Doppler and leads the result to each point in the echo spectrum after step 2 is handled
Lead the result with second order
And when satisfying:
The time, then (v, r) as the ridge point, v and r represent speed and the range coordinate in the RD spectrum respectively with the point in the echo spectrum.
4. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof, it is characterized in that the implication of the big criterion in the part described in the step 9 is: the outer local maximum sharpness of definite division main peak is the splitting spectrum peak.
5. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof, what it is characterized in that the offset scope described in the step 9 is: [v
m-Δ
Max, v
m+ Δ
Max], v wherein
mBe the doppler velocity position of main peak, Δ
MaxIt is the radially Bragg peak maximum offset that causes of ocean current of stable state/transient state.
6. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof is characterized in that the implication of symmetry principle described in the step 11 is: the skew of splitting spectrum is consistent in the same way.
7. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof, it is characterized in that the implication of the criterion of amplitude difference described in the step 10 is: the amplitude difference at main peak and its splitting spectrum peak exists: between 0～20dB.
8. the recognition methods at a kind of high band single order Bragg peak according to claim 1 and splitting spectrum peak thereof, it is characterized in that the fuzzy splitting spectrum that described in the step 12 step 11 is obtained carries out feature and confirms, the method of the final position at the splitting spectrum peak at acquisition Bragg peak is: investigate the invisible side main peak in splitting spectrum peak spectrum structure on every side, when the spectrum structure shows as the feature that adjacent weak target disturbs, judge then that splitting spectrum exists but invisible; Comprehensive then bilateral segmentation spectrum peak information, thereby the final position at the splitting spectrum peak at acquisition Bragg peak.
9. realize a kind of high band single order Bragg peak of recognition methods at claim 1 described a kind of high band single order Bragg peak and splitting spectrum peak thereof and the recognition system at splitting spectrum peak thereof, it comprises peak detection block (2) and alternative feature as a result affirmation module (3), it is characterized in that: further comprising tendency information extraction module (1) and splitting spectrum peak identification module (4), described tendency information extraction module (1) is made up of multi-scale filtering unit (11), ridge detecting unit (12) and shape filtering unit (13); Splitting spectrum peak identification module (4) is by splitting spectrum peak recognition unit (41), splitting spectrum amplitude difference confirmation unit (42), splitting spectrum symmetry matching unit (43) and fuzzy splitting spectrum feature confirmation unit (44), the input end of described multi-scale filtering unit (11) receives sea area to be detected information echo spectrum information, and the output terminal of multi-scale filtering unit (11) is connected with the input end of ridge detecting unit (12); The output terminal of ridge detecting unit (12) is connected with the input end of shape filtering unit (13); The output terminal of shape filtering unit (13) is connected with the input end of peak detection block (2); The output terminal of peak detection block (2) confirms that with alternative feature as a result the input end of module (3) is connected; Alternative feature as a result confirms that the output terminal of module (3) is connected with the input end of splitting spectrum peak recognition unit (41); The output terminal of splitting spectrum peak recognition unit (41) is connected with the input end of splitting spectrum amplitude difference confirmation unit (42); The output terminal of splitting spectrum amplitude difference confirmation unit (42) is connected with the input end of splitting spectrum symmetry matching unit (43); The output terminal of splitting spectrum symmetry matching unit (43) is connected with the input end of fuzzy splitting spectrum feature confirmation unit (44); Alternative feature is as a result confirmed the recognition result of the recognition result of module (3) output as single order Bragg peak; The recognition result of fuzzy splitting spectrum feature confirmation unit (44) output is as the recognition result at the splitting spectrum peak at Bragg peak.
10. a kind of high band single order Bragg peak according to Claim 8 and the recognition methods and the system at splitting spectrum peak thereof, it is characterized in that it comprises that also blind area detection module (5), SNR confirm module (6) and global characteristics affirmation module (7), the input end of described blind area detection module (5) is connected with the output terminal of fuzzy splitting spectrum feature confirmation unit (44); The output terminal of blind area detection module (5) confirms that with SNR the input end of module (6) is connected, and SNR confirms that the output terminal of module (6) is connected with the input end of global characteristics affirmation module (7).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010609804 CN102141609A (en) | 2010-12-28 | 2010-12-28 | Method and system for recognizing high-band first-order Bragg peak and split spectrum peak thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010609804 CN102141609A (en) | 2010-12-28 | 2010-12-28 | Method and system for recognizing high-band first-order Bragg peak and split spectrum peak thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102141609A true CN102141609A (en) | 2011-08-03 |
Family
ID=44409269
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010609804 Pending CN102141609A (en) | 2010-12-28 | 2010-12-28 | Method and system for recognizing high-band first-order Bragg peak and split spectrum peak thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102141609A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102680954A (en) * | 2012-05-28 | 2012-09-19 | 哈尔滨工业大学 | Quantitative detection method of color noise correlation property based on autocorrelation function difference |
| CN104180800A (en) * | 2014-08-20 | 2014-12-03 | 广州海格通信集团股份有限公司 | Correction method and system based on track points of ADS-B (Automatic Dependent Surveillance Broadcast) system |
| CN105718723A (en) * | 2016-01-18 | 2016-06-29 | 吉林大学 | Spectrum peak position detection method in mass spectrum data processing |
| CN106959442A (en) * | 2016-12-21 | 2017-07-18 | 国家海洋局第海洋研究所 | Ground wave radar first-order sea echo composes extracting method under strong interference environment based on many domain informations |
| CN107369297A (en) * | 2016-04-27 | 2017-11-21 | 科达海洋传感器有限公司 | For tsunami detection and pre-warning system |
| CN110954886A (en) * | 2019-11-26 | 2020-04-03 | 南昌大学 | High-frequency ground wave radar first-order echo spectrum region detection method taking second-order spectrum intensity as reference |
| CN115965825A (en) * | 2023-03-17 | 2023-04-14 | 南京信息工程大学 | Multi-spectral satellite image ship iceberg classification method based on lightweight network |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101806887A (en) * | 2010-03-19 | 2010-08-18 | 清华大学 | Space tracking filter-based sea clutter suppression and target detection method |
-
2010
- 2010-12-28 CN CN 201010609804 patent/CN102141609A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101806887A (en) * | 2010-03-19 | 2010-08-18 | 清华大学 | Space tracking filter-based sea clutter suppression and target detection method |
Non-Patent Citations (1)
| Title |
|---|
| 《Journal of Systems Engineering and Electronics》 20090831 Li Yang, Zhang Ning, Yang Qiang Characteristic-Knowledge-Aided Spectral Detection of High Frequency First-Order Sea Echo 第20卷, 第4期 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102680954A (en) * | 2012-05-28 | 2012-09-19 | 哈尔滨工业大学 | Quantitative detection method of color noise correlation property based on autocorrelation function difference |
| CN102680954B (en) * | 2012-05-28 | 2013-12-11 | 哈尔滨工业大学 | Quantitative detection method of color noise correlation property based on autocorrelation function difference |
| CN104180800A (en) * | 2014-08-20 | 2014-12-03 | 广州海格通信集团股份有限公司 | Correction method and system based on track points of ADS-B (Automatic Dependent Surveillance Broadcast) system |
| CN105718723A (en) * | 2016-01-18 | 2016-06-29 | 吉林大学 | Spectrum peak position detection method in mass spectrum data processing |
| CN107369297A (en) * | 2016-04-27 | 2017-11-21 | 科达海洋传感器有限公司 | For tsunami detection and pre-warning system |
| CN106959442A (en) * | 2016-12-21 | 2017-07-18 | 国家海洋局第海洋研究所 | Ground wave radar first-order sea echo composes extracting method under strong interference environment based on many domain informations |
| CN106959442B (en) * | 2016-12-21 | 2019-06-25 | 自然资源部第一海洋研究所 | Ground wave radar first-order sea echo composes extracting method under strong interference environment based on multiple domain information |
| CN110954886A (en) * | 2019-11-26 | 2020-04-03 | 南昌大学 | High-frequency ground wave radar first-order echo spectrum region detection method taking second-order spectrum intensity as reference |
| CN110954886B (en) * | 2019-11-26 | 2023-03-24 | 南昌大学 | High-frequency ground wave radar first-order echo spectrum region detection method taking second-order spectrum intensity as reference |
| CN115965825A (en) * | 2023-03-17 | 2023-04-14 | 南京信息工程大学 | Multi-spectral satellite image ship iceberg classification method based on lightweight network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102141609A (en) | Method and system for recognizing high-band first-order Bragg peak and split spectrum peak thereof | |
| CN106569196B (en) | A kind of compressed sensing based ground radar multi-target detection method | |
| CN107678003A (en) | Object detection method and device under a kind of ground wave radar sea clutter background | |
| Teague et al. | A comparison of multifrequency HF radar and ADCP measurements of near-surface currents during COPE-3 | |
| JP6415288B2 (en) | Radar equipment | |
| CN110488301B (en) | Multi-source information fusion sonar comprehensive target identification method | |
| CN103197297B (en) | Radar moving target detection method based on cognitive framework | |
| CN106990403B (en) | Low-altitude target tracking method based on the fusion of multiband two-stage information | |
| CN113109837B (en) | Data processing method of laser radar system | |
| CN104007424B (en) | Maneuvering target detection method based on time-frequency analysis | |
| CN110954886B (en) | High-frequency ground wave radar first-order echo spectrum region detection method taking second-order spectrum intensity as reference | |
| CN113608195B (en) | Laser radar full waveform data decomposition sounding method and device and electronic equipment | |
| CN102721966B (en) | Below high precision depth-sounding method and system by coherent depth-sounding sonar | |
| CN109116326A (en) | A kind of adaption radar ocean clutter cancellation method based on medion estimator | |
| CN111751797A (en) | High-frequency ground wave radar first-order and second-order echo spectrum boundary determining method based on azimuth angle | |
| CN108414992A (en) | A kind of object detection method based on phase information clutter map | |
| CN103308910A (en) | Method for detecting offshore non-navigational state ship target by using high-frequency ground wave radar | |
| CN106959442A (en) | Ground wave radar first-order sea echo composes extracting method under strong interference environment based on many domain informations | |
| CN108490425B (en) | Angle measuring method of bistatic MIMO radar | |
| CN103344950A (en) | Sea-state radar internal wave parameter retrieval method | |
| CN105353362B (en) | A kind of captain's method of estimation for merging the two kinds of measurements of energy and rate of change | |
| CN117148358A (en) | Barnes spatial filtering-based Doppler weather radar speed deblurring method | |
| Lopera et al. | Time-frequency domain signature analysis of GPR data for landmine identification | |
| CN109581350B (en) | Radar distance and speed measurement method and device based on time-frequency integral interpolation | |
| Evans et al. | Evaluation of three antenna pattern measurements for a 25 MHz seasonde |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110803 |