CN107610130A - Extra large land clutter Scene Segmentation based on the ratio between amplitude and phase linearity - Google Patents
Extra large land clutter Scene Segmentation based on the ratio between amplitude and phase linearity Download PDFInfo
- Publication number
- CN107610130A CN107610130A CN201710720482.1A CN201710720482A CN107610130A CN 107610130 A CN107610130 A CN 107610130A CN 201710720482 A CN201710720482 A CN 201710720482A CN 107610130 A CN107610130 A CN 107610130A
- Authority
- CN
- China
- Prior art keywords
- mrow
- distance
- ripple position
- amplitude
- phase
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention discloses a kind of extra large land clutter Scene Segmentation based on the ratio between amplitude and phase linearity, its method and step is：(1) echo sequence is obtained；(2) establish apart from ripple position coordinate system；(3) selected distance ripple position resolution cell；(4) amplitude and phase linearity matrix are calculated；(5) judge whether to have chosen all points in the coordinate system of ripple position, if so, then performing step (6)；Otherwise, step (3) is performed；(6) amplitude and phase linearity matrix are converted into gray level image；(7) bianry image is produced；(8) extra large land clutter scene segmentation figure picture is produced.The present invention is used for estimating with amplitude and phase linearity, has fully demonstrated and has moved or static coherent system platform is plunged into the commercial sea the otherness of land clutter, extra large land clutter scene is split, obtains more accurately result.
Description
Technical field
The invention belongs to communication technical field, the one kind further related in Radar Signal Processing Technology field is based on width
Degree and the extra large land clutter Scene Segmentation of the ratio between phase linearity.The present invention can be used for the echo data obtained to airborne radar
The image of generation, by the extra large land clutter scene cut that the ocean in image, land area are separated.
Background technology
Target detection technique under sea clutter background is a vital research direction in radar application technology, in army
Thing and civil area are widely used.For radar when being worked under to extra large pattern, scanning scene is complicated and scope is larger, and radar returns
Various types of clutters, including sea clutter, land clutter, islands and reefs clutter, coastal waters clutter etc. are usually contained in ripple.Land clutter and island
Reef clutter echo strength is stronger, drastically influence the target detection under sea clutter background, complicated clutter scene and noise performance
Constitute the major obstacle of sea-surface target detection.Therefore, before the target detection under carrying out sea clutter background, it is necessary to Hai Lu
Clutter scene cut is pre-processed.Land clutter and islands and reefs echo are separated from radar clutter by extra large land clutter scene cut
Go out, land clutter and islands and reefs echo are excluded during target detection, reduce land clutter and islands and reefs echo to sea clutter
The influence of target detection under background.The quality of extra large land clutter scene segmentation result will be directly affected under sea clutter background
Target detection performance.
Different from traditional image partition method, the segmentation of radar clutter scene includes changes into gray scale by radar data
Image and gray level image split two parts.The doppler spectral of sea clutter with larger bandwidth and the doppler spectral of land clutter with
Less bandwidth.Due to aircraft motion, echo Doppler is offset with azimuthal variation, by the extra large land clutter of Doppler frequency
Scene cut is often high calculation cost, it is difficult to meets the requirement of extra large land clutter scene cut in real time.If utilize manually
Method directly carries out description segmentation to clutter scene, wastes time and energy, and will bring huge workload, can not equally meet in real time
Scene cut requirement.
Patent " the block adaptive image partition method based on FCM " (number of patent application that Hohai University applies at it
201310726876, publication number CN103761726B) in propose a kind of block adaptive image partition method based on FCM.
This method is respectively adopted neighborhood averaging and treated after segmentation figure picture is handled to be split using based on FCM image partition methods
Split with segmentation figure picture is treated based on standard FCM image partition methods, piecemeal processing is carried out to result figure, to image-region
Block number and the variance for calculating each image-region block, compare the image block variance of same position in two width segmentation figures, selecting party
The less image block of difference is as last segmentation result.This method has taken into full account the half-tone information and space letter of image to be split
Breath, and segmentation figure is treated as block adaptive selection method of partition, the segmentation quality of image is improved, but still have deficiency
Part, due to this method can not the correct unbalanced image of segmenting pixels Density Distribution, cause to split second-rate.
The patent that BJ University of Aeronautics ＆ Astronautics applies at it is " a kind of based on the SAR image of wavelet transformation and OTSU threshold values sea
One kind is proposed in land dividing method " (number of patent application 201210536981.2, publication number CN102968798B) and is based on small echo
Conversion and the SAR image sea land dividing method of OTSU threshold values, this method suppress SAR figures using the noise smoothing characteristic of wavelet transformation
Spot as in is made an uproar, and then is partitioned into land area, and base roughly using a kind of non-supervisory, optimal threshold OTSU threshold methods
In the multiscale analysis characteristic of wavelet transformation, the testing result under each yardstick is merged, finally by the follow-up of automation
Processing and Edge track obtain final SAR image sea land segmentation result.The spot that this method fully utilizes wavelet transformation is made an uproar suppression
The adaptive and non-supervisory characteristic of system, multiscale analysis function and OTSU thresholding algorithms, in being applicable for High Resolution SAR Images
Property aspect have significant improvement, but still have weak point, because this method is still more sensitive to noise, cause
The robustness of this method is poor, and segmentation figure is as uneven.
The content of the invention
It is an object of the invention to overcome the shortcomings of above-mentioned prior art, it is proposed that based on the ratio between amplitude and phase linearity
Extra large land clutter Scene Segmentation.Radar is received the echo sequence containing extra large land clutter scene and is converted into gray-scale map by the present invention
Picture, establishes a distance-ripple position coordinate system, in the coordinate system of extraction distance-ripple position a little, distance-ripple that all-pair is answered
Position is as distance-ripple position resolution cell, with the extra large land clutter Scene Segmentation based on the ratio between amplitude and phase linearity, to institute
There is the resolution cell division of distance-ripple position, reach the purpose of extra large land clutter scene cut.
Realizing the basic ideas of the object of the invention is：First, radar is received into the echo sequence containing extra large land clutter scene
Gray level image is converted into, establishes a distance-ripple position coordinate system, point all in the coordinate system of distance-ripple position is extracted, by institute's reconnaissance
Corresponding distance-ripple position is used as distance-ripple position resolution cell；Then, the amplitude and phase of each distance-ripple position resolution cell are calculated
The ratio of the position linearity, obtains amplitude and phase linearity matrix, matrix is converted into gray level image；Finally, maximum kind is utilized
Between variance method obtain the bianry image after Threshold segmentation, using 5*5 structural elements to bianry image carry out morphologic filtering, obtain
Final extra large land clutter scene segmentation figure picture.
To achieve these goals, specific implementation step of the invention is as follows：
(1) echo sequence is obtained：
Radar receives pulse-distance-ripple position three-dimensional echo sequence containing extra large land clutter scene；
(2) distance-ripple position coordinate system is established：
Echo sequence is converted into the gray level image that size is M × L, establishes a distance-ripple position coordinate system, wherein, M tables
Show distance dimension, L represents ripple position dimension；
(3) selected distance-ripple position resolution cell：
From the coordinate system of distance-ripple position optionally a bit, the distance corresponding to the point-ripple position is formed into distance-ripple position to differentiate
Unit；
(4) amplitude and phase linearity matrix are calculated：
(4a) utilizes range value formula, the range value of distance selected by calculating-ripple position resolution cell；
(4b) utilizes phase linearity angle value formula, the phase linearity angle value of distance selected by calculating-ripple position resolution cell；
(4c) using amplitude and phase linearity ratio formula, the range value of distance selected by calculating-ripple position resolution cell with
The ratio of phase linearity angle value；
(4d) calculates amplitude and the element of phase linearity matrix using amplitude and phase linearity matrix element formula；
(5) judge whether to have chosen point all in distance-ripple position coordinate system, if so, then performing step (6)；Otherwise, hold
Row step (3)；
(6) amplitude and phase linearity matrix are converted into gray level image：
Matrix is converted into gray level image using mat2gray sentences in matlab2014a；
(7) bianry image is produced：
(7a) utilizes maximum variance between clusters, obtains the optimal threshold of gray level image；
(7b) judges whether the pixel value of any pixel in gray level image is more than optimal threshold, if so, then performing step
(7c)；Otherwise, step (7d) is performed；
The pixel value that the pixel value of pixel in gray level image is more than the pixel of optimal threshold by (7c) is set to 0；
The pixel value that the pixel value of pixel in gray level image is less than or equal to the pixel of optimal threshold by (7d) is set to 1；
Pixel after the pixel value RS reset-set of all pixels point in image is formed bianry image by (7e)；
(8) extra large land clutter scene segmentation figure picture is produced：
Morphologic filtering is carried out to bianry image using 5*5 structural elements, obtains final extra large land clutter scene segmentation figure
Picture.
The present invention has advantages below compared with prior art：
First, because the present invention is during extra large land clutter scene image is split, employ amplitude and phase linearity
Be used for estimate, fully demonstrated and plunged into the commercial sea the otherness of land clutter in motion or static coherent system platform, overcome existing
Technology can not correctly split the deficiency of the unbalanced extra large land clutter of Density Distribution so that the present invention improves the segmentation essence of image
Degree.
Second, because the present invention is during extra large land clutter scene image is split, employ maximum between-cluster variance threshold value
Split plot design enters row threshold division to gray level image, and it is adaptive that it, which obtains the process of optimal threshold, it is not necessary to artificial setting
Any parameter, arithmetic speed is fast, overcomes the deficiency that prior art is selected threshold value so that the present invention is in image grey level histogram
During without obvious bimodal or trough, the segmentation result that optimal threshold is more satisfied with can be also obtained.
3rd, because the present invention is during extra large land clutter scene image is split, morphologic filtering is employed to threshold value
Bianry image after segmentation is handled, and ensure that the connectedness of segmentation result Mid-continent domain and sea area, is overcome existing
There is technology to split deficiency at random so that the present invention improves the quality of extra large land clutter scene cut.
Brief description of the drawings
Fig. 1 is the flow chart of the present invention；
Fig. 2 is the analogous diagram of the present invention.
Embodiment
The present invention is described in further detail below in conjunction with the accompanying drawings.
Step 1, radar receives scattering object and returns to the three-dimensional echo sequence containing pulse-distance-ripple position.
Step 2, distance-ripple position coordinate system is established.
Echo sequence is converted into gray level image, establishes distance-ripple position coordinate system.
Step 3, selected distance-ripple position resolution cell.
From the coordinate system of distance-ripple position optionally a bit, the distance corresponding to the point-ripple position is formed into distance-ripple position to differentiate
Unit.
Step 4, amplitude and phase linearity matrix are calculated.
Utilize range value formula, the range value of distance selected by calculating-ripple position resolution cell；
Described range value formula is as follows：
Wherein, PkThe range value of k-th of distance-ripple position resolution cell is represented, i represents the pulse sequence of echo sequence
Number, i=1,2 ... N, N represent the pulse sum of echo sequence, and ∑ represents sum operation, | | represent amplitude behaviour
Make, xi,kRepresent i-th of pulse of echo sequence k-th of distance-ripple position resolution cell, k=1,2 ..., P × Q, P expression away from
From the sum with a distance from-ripple position resolution cell, Q represents the sum of distance-ripple position resolution cell ripple position.
Utilize phase linearity angle value formula, the phase linearity angle value of distance selected by calculating-ripple position resolution cell；
Described phase linearity angle value formula is as follows：
Wherein, θkThe phase linearity angle value of k-th of distance-ripple position resolution cell is represented,Represent evolution operation, n tables
Show the pulse sequence number of echo sequence, n=2,3 ..., N, ()2Represent square operation, Δk(n) echo sequence phase is represented
The winding phase of k-th of distance-ripple position resolution cell and the phase difference of true phase of adjacent pulse,Represent k-th of distance-ripple position point of n-th of pulse of echo sequence
Distinguish the winding phase of unit, φk(n) the true phase of k-th of distance-ripple position resolution cell of n-th of pulse of echo sequence is represented
Position,Represent the distance-winding phase of ripple position resolution cell and the phase difference of true phase of all adjacent pulses of echo sequence
Δk(n) average value,
Utilize amplitude and phase linearity ratio formula, the range value and phase of distance selected by calculating-ripple position resolution cell
The ratio of linear angle value；
Described amplitude and phase linearity ratio formula are as follows：
Wherein, RkRepresent the range value P of k-th of distance-ripple position resolution cellkWith phase linearity angle value θkRatio.
Using amplitude and phase linearity matrix element formula, amplitude and the element of phase linearity matrix are calculated；
Described amplitude and the formula of phase linearity matrix element are as follows：
Rpq=log10(Rk)；
Wherein, RpqThe element that expression amplitude arranges with pth row q in phase linearity matrix, p represent that distance-ripple position is differentiated
Unit apart from sequence number, p=1,2 ... P, q represents the ripple position sequence number of distance-ripple position resolution cell, q=1,2 ..., Q,
log10Represent denary logarithm operation.
Step 5, judge whether to have chosen point all in distance-ripple position coordinate system, if so, step 6 is then performed, otherwise,
Perform step 3.
Step 6, amplitude and phase linearity matrix are converted into gray level image.
Step 7, bianry image is produced.
(7a) utilizes maximum variance between clusters, obtains the optimal threshold of gray level image.
Described maximum variance between clusters comprise the following steps that：
The first step, gray level image include S gray level, appoint and take thresholding t that gray level image is divided into C1And C2Two classes, C1Table
Show pixel point set of the gray value less than or equal to t, C in gray level image1={ 1,2 ..., t }, C2Represent gray value in gray level image
Pixel point set more than t, C2={ t+1, t+2 ..., S-1 }, 0≤t≤S-1；
Second step, according to the following formula, calculate C1And C2Inter-class variance,
σ2(t)=p1(μ1-μT)2+p2(μ2-μT)2
Wherein, σ2(t) C is represented1And C2Inter-class variance, p1Represent C1The probability of appearance,nλRepresent gray scale
It is worth the pixel sum for λ, W represents the pixel sum of gray level image, p2Represent C2The probability of appearance,The tables of μ 1
Show C1Gray average,μ 2 represents C2Gray average,μ (t) represents gray level image
Gray average,
3rd step, according to the following formula, calculate σ2(t) threshold value during maximum is obtained,
T=Argmax { σ2(t)}
Wherein, T represents inter-class variance σ2(t) threshold value during maximum is obtained, Argmax, which represents to take corresponding to maximum, to be become
Amount operation；
4th step, the optimal threshold using threshold value T as gray level image.
(7b) judges whether the pixel value of the pixel in gray level image is more than optimal threshold, if so, then performing step
(7c)；Otherwise, step (7d) is performed.
The pixel value for the pixel for being more than optimal threshold in the pixel value of pixel in gray level image is set to 0 by (7c).
The pixel value that the pixel value of pixel in gray level image is less than or equal to the pixel of optimal threshold by (7d) is set to 1.
Pixel after the pixel value RS reset-set of all pixels point in image is formed bianry image by (7e)；
Step 8, extra large land clutter scene segmentation figure picture is produced.
Described morphologic filtering refers to, operation and closed operation are carried out out to bianry image using 5*5 structural elements, is filtered out
The isolated group of several pixels is occupied in bianry image, fills the hole in region in flakes, land and the scene areas of ocean two are drawn
Separate, obtain final extra large land clutter scene segmentation figure picture.
The effect of the present invention can be further illustrated by following emulation：
1. simulated conditions：
It in allocation of computer be core i7 3.40GHZ that the emulation experiment of the present invention, which is, internal memory 8G, WINDOWS7 system with
Computer software is configured to what is carried out under Matlab R2014a environment.
2. emulation content：
The emulation experiment of the present invention is to use two kinds of prior arts (phase linearity method, standard Fuzzy C Mean Method)
With the present invention, pulse-distance containing the extra large land clutter scene-ripple position three-dimensional echo sequence received respectively to radar is imitated
Very, shown in the result obtained such as Fig. 2 (b), Fig. 2 (c), Fig. 2 (d).
Fig. 2 (a) is to calculate radar to receive pulse-distance-ripple position three-dimensional echo sequence amplitude containing extra large land clutter scene,
The original image obtained after conversion amplitude；
Fig. 2 (b) is the analogous diagram obtained using the phase linearity method of prior art, and wherein Fig. 2 (b) is to use phase
Linearity Method calculates the phase linearity of all distances-ripple position resolution cell in echo sequence, and phase linearity is converted into
Gray level image, threshold process is carried out to gray level image using maximum variance between clusters, bianry image is obtained, using 8*8 structural elements
Element carries out morphologic filtering, obtained extra large land clutter scene segmentation figure picture to bianry image；
Fig. 2 (c) is the analogous diagram obtained using the standard Fuzzy C Mean Method of prior art, and wherein Fig. 2 (c) is use
Standard Fuzzy C Mean Method calculates the cluster centre and subordinated-degree matrix in the image by echo sequence conversion, passes through minimum
Weighted cluster object function produces optimal C sections, obtained extra large land clutter scene segmentation figure picture；
Fig. 2 (d) is the analogous diagram obtained using the present invention, and wherein Fig. 2 (d) is to calculate echo sequence using the inventive method
The amplitude of middle distance-ripple position resolution cell and phase linearity matrix, gray level image is converted into by matrix, utilizes side between maximum kind
Poor method carries out threshold process to gray level image, obtains bianry image, and Mathematical morphology filter is carried out to bianry image using 5*5 structural elements
The extra large land clutter scene segmentation figure picture that ripple obtains；
3. analysis of simulation result：
As can be seen that splitting clutter scene using the extra large land of the phase linearity of prior art from Fig. 2 (b) and Fig. 2 (d)
Dividing method, the hole for causing many needs retained can be connected, and as a result less divided, segmentation are second-rate.And the present invention fills
Divide the otherness for embodying extra large land clutter, more accurate segmentation result can be obtained.
As can be seen that the image partition method clustered using the FCM of prior art from Fig. 2 (c) and Fig. 2 (d), it is impossible to just
The really segmentation unbalanced extra large land clutter of Density Distribution, segmentation are second-rate.And the present invention can preferably split Density Distribution not
Balanced extra large land clutter, can obtain more accurate segmentation result.
Claims (7)
1. a kind of extra large land clutter Scene Segmentation based on the ratio between amplitude and phase linearity, including step are as follows：
(1) echo sequence is obtained：
Radar receives pulse-distance-ripple position three-dimensional echo sequence containing extra large land clutter scene；
(2) distance-ripple position coordinate system is established：
By echo sequence be converted into size be M × L gray level image, establish a distance-ripple position coordinate system, wherein, M represent away from
From dimension, L represents ripple position dimension；
(3) selected distance-ripple position resolution cell：
From the coordinate system of distance-ripple position optionally a bit, the distance corresponding to the point-ripple position is formed into distance-ripple position resolution cell；
(4) amplitude and phase linearity matrix are calculated：
(4a) utilizes range value formula, the range value of distance selected by calculating-ripple position resolution cell；
(4b) utilizes phase linearity angle value formula, the phase linearity angle value of distance selected by calculating-ripple position resolution cell；
(4c) utilizes amplitude and phase linearity ratio formula, the range value and phase of distance selected by calculating-ripple position resolution cell
The ratio of linear angle value；
(4d) calculates amplitude and the element of phase linearity matrix using amplitude and phase linearity matrix element formula；
(5) judge whether to have chosen point all in distance-ripple position coordinate system, if so, then performing step (6)；Otherwise, step is performed
Suddenly (3)；
(6) amplitude and phase linearity matrix are converted into gray level image；
(7) bianry image is produced：
(7a) utilizes maximum variance between clusters, obtains the optimal threshold of gray level image；
(7b) judges whether the pixel value of the pixel in gray level image is more than optimal threshold, if so, then performing step (7c)；It is no
Then, step (7d) is performed；
The pixel value for the pixel for being more than optimal threshold in the pixel value of pixel in gray level image is set to 0 by (7c)；
The pixel value that the pixel value of pixel in gray level image is less than or equal to the pixel of optimal threshold by (7d) is set to 1；
Pixel after the pixel value RS reset-set of all pixels point in image is formed bianry image by (7e)；
(8) extra large land clutter scene segmentation figure picture is produced：
Morphologic filtering is carried out to bianry image using 5*5 structural elements, obtains final extra large land clutter scene segmentation figure picture.
2. the extra large clutter Scene Segmentation in land according to claim 1 based on the ratio between amplitude and phase linearity, it is special
Sign is：Range value formula described in step (4a) is as follows：
<mrow>
<msub>
<mi>P</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>N</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mo>|</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>|</mo>
<mo>;</mo>
</mrow>
Wherein, PkThe range value of k-th of distance-ripple position resolution cell of expression, the pulse sequence number of i expression echo sequences, i=1,
2 ... N, N represent the pulse sum of echo sequence, and Σ represents sum operation, | | represent amplitude operation, xi,kRepresent echo sequence
K-th of distance-ripple position resolution cell of i-th of pulse of row, k=1,2 ..., P × Q, P represent in the resolution cell of distance-ripple position
The sum of distance, Q represent the sum of distance-ripple position resolution cell ripple position.
3. the extra large clutter Scene Segmentation in land according to claim 1 based on the ratio between amplitude and phase linearity, it is special
Sign is：Phase linearity angle value formula described in step (4b) is as follows：
<mrow>
<msub>
<mi>&theta;</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<msqrt>
<mrow>
<mfrac>
<mn>1</mn>
<mrow>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>2</mn>
</mrow>
<mi>N</mi>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<mi>&Delta;</mi>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
<mo>-</mo>
<mover>
<mi>K</mi>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>;</mo>
</mrow>
Wherein, θkThe phase linearity angle value of k-th of distance-ripple position resolution cell is represented,Evolution operation is represented, n is represented back
The pulse sequence number of wave train, n=2,3 ..., N, ()2Represent square operation, Δk(n) echo sequence phase is represented
The winding phase of k-th of distance-ripple position resolution cell and the phase difference of true phase of adjacent pulse, Represent k-th of distance-ripple position point of n-th of pulse of echo sequence
Distinguish the winding phase of unit, φk(n) the true phase of k-th of distance-ripple position resolution cell of n-th of pulse of echo sequence is represented
Position,Represent the distance-winding phase of ripple position resolution cell and the phase difference of true phase of all adjacent pulses of echo sequence
Δk(n) average value,
4. the extra large clutter Scene Segmentation in land according to claim 1 based on the ratio between amplitude and phase linearity, it is special
Sign is：Amplitude and phase linearity ratio formula described in step (4c) is as follows：
<mrow>
<msub>
<mi>R</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>P</mi>
<mi>k</mi>
</msub>
<msub>
<mi>&theta;</mi>
<mi>k</mi>
</msub>
</mfrac>
<mo>=</mo>
<mfrac>
<mrow>
<mfrac>
<mn>1</mn>
<mi>N</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mo>|</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>|</mo>
</mrow>
<msqrt>
<mrow>
<mfrac>
<mn>1</mn>
<mrow>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>n</mi>
<mo>=</mo>
<mn>2</mn>
</mrow>
<mi>N</mi>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<mi>&Delta;</mi>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
<mo>-</mo>
<mover>
<mi>K</mi>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
</mfrac>
<mo>;</mo>
</mrow>
Wherein, RkRepresent the range value P of k-th of distance-ripple position resolution cellkWith phase linearity angle value θkRatio.
5. the extra large clutter Scene Segmentation in land according to claim 1 based on the ratio between amplitude and phase linearity, it is special
Sign is：Amplitude and the formula of phase linearity matrix element described in step (4d) is as follows：
Rpq=log10(Rk)；
Wherein, RpqThe element that expression amplitude arranges with pth row q in phase linearity matrix, p represent distance-ripple position resolution cell
Apart from sequence number, p=1,2 ... P, q represents the ripple position sequence number of distance-ripple position resolution cell, q=1,2 ..., Q, log10Table
Show that denary logarithm operates.
6. the extra large clutter Scene Segmentation in land according to claim 1 based on the ratio between amplitude and phase linearity, it is special
Sign is：Maximum variance between clusters described in step (7a) comprise the following steps that：
The first step, gray level image include S gray level, appoint and take thresholding t that gray level image is divided into C1And C2Two classes, C1Represent ash
Spend the pixel point set that gray value in image is less than or equal to t, C1={ 1,2 ..., t }, C2Represent that gray value is more than in gray level image
T pixel point set, C2={ t+1, t+2 ..., S-1 }, 0≤t≤S-1；
Second step, according to the following formula, calculate C1And C2Inter-class variance,
σ2(t)=p1(μ1-μT)2+p2(μ2-μT)2
Wherein, σ2(t) C is represented1And C2Inter-class variance, p1Represent C1The probability of appearance,nλExpression gray value is λ
Pixel sum, W represent gray level image pixel sum, p2Represent C2The probability of appearance,μ1Represent C1
Gray average,μ2Represent C2Gray average,μ (t) represents the gray scale of gray level image
Average,
3rd step, according to the following formula, calculate σ2(t) threshold value during maximum is obtained,
T=Arg max { σ2(t)}
Wherein, T represents inter-class variance σ2(t) threshold value during maximum is obtained, Arg max represent to take variable corresponding to maximum to grasp
Make；
4th step, the optimal threshold using threshold value T as gray level image.
7. the extra large clutter Scene Segmentation in land according to claim 1 based on the ratio between amplitude and phase linearity, it is special
Sign is：Morphologic filtering described in step (8) refers to, bianry image operate and close using 5*5 structural elements
Operation, the isolated group that several pixels are occupied in bianry image is filtered out, fill the hole in region in flakes, land and the scene of ocean two
Region, which is divided, comes, and obtains final extra large land clutter scene segmentation figure picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710720482.1A CN107610130B (en) | 2017-08-21 | 2017-08-21 | Sea-land clutter scene segmentation method based on amplitude and phase linearity ratio |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710720482.1A CN107610130B (en) | 2017-08-21 | 2017-08-21 | Sea-land clutter scene segmentation method based on amplitude and phase linearity ratio |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107610130A true CN107610130A (en) | 2018-01-19 |
| CN107610130B CN107610130B (en) | 2020-01-31 |
Family
ID=61065625
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710720482.1A Active CN107610130B (en) | 2017-08-21 | 2017-08-21 | Sea-land clutter scene segmentation method based on amplitude and phase linearity ratio |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107610130B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108961255A (en) * | 2018-06-28 | 2018-12-07 | 西安电子科技大学 | Extra large land noise scenarios dividing method based on phase linearity and power |
| CN109543589A (en) * | 2018-11-16 | 2019-03-29 | 西安电子科技大学 | Extra large land Scene Segmentation based on the constant distance of first phase-Doppler and KNN |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110052088A1 (en) * | 2009-08-31 | 2011-03-03 | Yuan Xiaoru | High dynamic range image mapping with empirical mode decomposition |
| CN104239901A (en) * | 2014-09-11 | 2014-12-24 | 西安电子科技大学 | Polarized SAR image classification method based on fuzzy particle swarm and target decomposition |
| CN105427301A (en) * | 2015-11-17 | 2016-03-23 | 西安电子科技大学 | Sea and land clutter scene segmentation method based on direct current component ratio measure |
| CN105844644A (en) * | 2016-03-31 | 2016-08-10 | 西安电子科技大学 | Morphological median derivative-based sea-land clutter scene segmentation method |
-
2017
- 2017-08-21 CN CN201710720482.1A patent/CN107610130B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110052088A1 (en) * | 2009-08-31 | 2011-03-03 | Yuan Xiaoru | High dynamic range image mapping with empirical mode decomposition |
| CN104239901A (en) * | 2014-09-11 | 2014-12-24 | 西安电子科技大学 | Polarized SAR image classification method based on fuzzy particle swarm and target decomposition |
| CN105427301A (en) * | 2015-11-17 | 2016-03-23 | 西安电子科技大学 | Sea and land clutter scene segmentation method based on direct current component ratio measure |
| CN105844644A (en) * | 2016-03-31 | 2016-08-10 | 西安电子科技大学 | Morphological median derivative-based sea-land clutter scene segmentation method |
Non-Patent Citations (1)
| Title |
|---|
| 马晓礼: ""陆海杂波场景快速分割方法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108961255A (en) * | 2018-06-28 | 2018-12-07 | 西安电子科技大学 | Extra large land noise scenarios dividing method based on phase linearity and power |
| CN108961255B (en) * | 2018-06-28 | 2021-09-28 | 西安电子科技大学 | Sea-land noise scene segmentation method based on phase linearity and power |
| CN109543589A (en) * | 2018-11-16 | 2019-03-29 | 西安电子科技大学 | Extra large land Scene Segmentation based on the constant distance of first phase-Doppler and KNN |
| CN109543589B (en) * | 2018-11-16 | 2021-02-02 | 西安电子科技大学 | Sea-land scene segmentation method based on initial phase-Doppler invariant distance and KNN |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107610130B (en) | 2020-01-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108961255B (en) | Sea-land noise scene segmentation method based on phase linearity and power | |
| CN107728115A (en) | Ambient interferences suppressing method based on SVM after a kind of radar target imaging | |
| CN103729846B (en) | LiDAR point cloud data edge detection method based on triangular irregular network | |
| CN106093890A (en) | The method setting up residual clutter suppression radar residual spur | |
| CN111541511B (en) | Communication interference signal identification method based on target detection in complex electromagnetic environment | |
| CN107274401A (en) | A kind of High Resolution SAR Images Ship Detection of view-based access control model attention mechanism | |
| CN101482617A (en) | Synthetic aperture radar image denoising method based on non-down sampling profile wave | |
| CN105427301B (en) | Based on DC component than the extra large land clutter Scene Segmentation estimated | |
| CN104200471A (en) | SAR image change detection method based on adaptive weight image fusion | |
| CN105405132A (en) | SAR image man-made target detection method based on visual contrast and information entropy | |
| CN113673312B (en) | Deep learning-based radar signal intra-pulse modulation identification method | |
| CN102999908A (en) | Synthetic aperture radar (SAR) airport segmentation method based on improved visual attention model | |
| CN110333489A (en) | The processing method to SAR echo data Sidelobe Suppression is combined with RSVA using CNN | |
| CN108985304A (en) | It is a kind of based on the Structure of the deposits extraction method for shallowly cuing open data | |
| CN103983959A (en) | SAR system movement target radial speed estimation method based on data reconstruction | |
| CN106443593A (en) | Self-adaptive oil spill information extraction method based on coherent radar slow-scan enhancement | |
| CN107610130A (en) | Extra large land clutter Scene Segmentation based on the ratio between amplitude and phase linearity | |
| CN109543589B (en) | Sea-land scene segmentation method based on initial phase-Doppler invariant distance and KNN | |
| CN105741262B (en) | The extra large land clutter Scene Segmentation estimated based on energy compaction measure | |
| Yu et al. | Application of a convolutional autoencoder to half space radar hrrp recognition | |
| CN107909595A (en) | Extra large land clutter Scene Segmentation based on amplitude Yu energy compaction measure product | |
| CN104036491A (en) | SAR image segmentation method based on area division and self-adaptive polynomial implicit model | |
| CN105741286B (en) | The SAR image moving-target shadow extraction method being combined based on width | |
| CN104537681A (en) | Method and system for extracting spectrum-separated visual salient region | |
| CN105844644B (en) | Extra large land clutter Scene Segmentation based on morphology intermediate value derivative |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |