CN105116410B - The interferometric phase image adaptive filter algorithm matched based on linear model - Google Patents
The interferometric phase image adaptive filter algorithm matched based on linear model Download PDFInfo
- Publication number
- CN105116410B CN105116410B CN201510427496.5A CN201510427496A CN105116410B CN 105116410 B CN105116410 B CN 105116410B CN 201510427496 A CN201510427496 A CN 201510427496A CN 105116410 B CN105116410 B CN 105116410B
- Authority
- CN
- China
- Prior art keywords
- phase
- mrow
- orientation
- image
- estimation
- 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.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
- G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
- G01S13/9021—SAR image post-processing techniques
- G01S13/9023—SAR image post-processing techniques combined with interferometric techniques
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of interferometric phase adaptive filter algorithm matched based on linear model, step of realizing of the invention is：Input main and auxiliary SAR complex image datas；Pass through the multiple interferometric phase image of interference treatment generation；Carry out gradient estimation, estimation orientation to distance to phase gradient；According to the optimum angle gradient of each upward sample point of facet approximation criterion computer azimuth；Carry out orientation and range direction phase compensation；Mean filter processing is carried out to the interferometric phase image after orientation and range direction phase compensation；Export the main and auxiliary SAR image data after phase compensation and mean filter processing；The present invention has in the case of Long baselines, can complete high-quality adaptive-filtering to the main and auxiliary SAR image after registration and handle, and meets the Practical Project performance requirement of high-quality interference synthetic aperture radar InSAR processing, obtains high-quality mapping product.
Description
Technical field
The invention belongs to communication technical field, a kind of interference synthesis hole in the radar exploration technique field is further related to
It is dry in footpath radar (Interferometric Synthetic Aperture Radar, InSAR) topographic survey processing procedure
The processing of phase diagram adaptive-filtering is related to, is a kind of interferometric phase image adaptive filter algorithm matched based on linear model.
Background technology
Interference synthetic aperture radar conventional art synthetic aperture radar (Synthetic Aperture Radar,
SAR) actual three-dimensional scenic is obtained on the basis of peacekeeping azimuth dimension two-dimensional signal, the same field that joint different visual angles are obtained
Two width with certain coherence or several SAR images of scape, pass through interference treatment technology, obtain the dimensional topography of target scene
Information.
When interference synthetic aperture radar InSAR data is obtained, because time dephasing is closed, baseline dephasing is closed and InSAR
The thermal noise of system in itself causes the presence of a large amount of phase speckle noises in interferometric phase image, although having been carried out in interference treatment
Noise in multiple look processing, but interferometric phase image is still than more serious, and the remanent point in phase diagram is very intensive.Dry
The general performance indications counted out with residual as assessment phase diagram noise suppressed in SAR data processing are related to, remanent point is more at most
Illustrate that phase diagram is more serious by noise pollution, i.e. the signal to noise ratio of interference image is lower.Noise drastically influence the image of interference SAR
Quality so that phase unwrapping can not carry out or generate DEM precision reduction.To obtain high-quality INSAR images, it is necessary to right
Noise is effectively suppressed, while the resolution ratio of INSAR images should be kept.Adaptive-filtering processing can effectively reduce interference
Phase speckle noise in phase diagram, lifting phase unwrapping precision and the more accurate absolute phase values of acquisition, and then improve
InSAR system altimetry performances.
Konstantinos P.Papathanassiou, in document " A New Technique for Noise
Filtering of SAR Interferometric Phase Images”(IEEE TRANSACTIONS ON
GEOSCIENCE AND REMOTE SENSING, VOL.36, NO.5, SEPTEMBER1998) in propose a kind of interference synthesis
Interferogram filtering method in aperture radar data processing, the general principle of this method is：Will be each in interferometric phase image
It is Additive noise model that noise like, which is approximately considered, and Directional Windows statistics equal to the same area window size but that direction is different is made an uproar
Sound, the change according to noise level in different directions window selects fixed Directional Windows to be filtered processing along stripe direction, from
And improve the inhibition of noise.
Li Jinwei, Li Zhenfang, paper " a kind of optimal interferometric phase filtering of coherence factor weighting " (《Xi'an electronic section
Skill college journal》, April, volume 42, the 2nd phase in 2014) in propose in a kind of processing of interference synthetic aperture radar InSAR data
Optimal interferometric phase filtering method.This method is to obtain best initial weights by Coherence Estimation, under minimum variance meaning
Processing is weighted to coherence factor, is then weighted according to minimum average B configuration phase difference criterion using one group of fixed-size directionality
Window is filtered processing.
The core of interferogram filtering method in above-mentioned two kinds of interference synthetic aperture radar InSAR datas processing,
All be from interferometric phase in itself with its present in noise, processing is filtered to interferometric phase image using stationary window.
But, the deficiency that both processing modes are present is in interference of data of synthetic aperture radar processing procedure, to have ignored major-minor
The influence for the phase gradient that SAR image orientation is produced by topography variation.Filter window size is not accounted for not same district simultaneously
Otherness between the filter effect of domain noise level.Due to the difference of different zones noise level in interferometric phase image, if with
Stationary window, which is filtered processing to interferometric phase image, can cause high s/n ratio region transitions to filter and lose detailed information, low letter
Make an uproar than region and filtering deficiency then occur and remain much noise influence subsequent phase solution and twine operation.To above two interference synthesis hole
Interferogram filtering processing method in the radar InSAR data processing of footpath is slow to local phase change in interferometric phase image
Influence can be neglected, reliable data product can be obtained.And the region differed greatly for local phase change is above-mentioned
Interferogram filtering processing method is in the engineer applied that interference synthetic aperture radar InSAR data is handled by the essence of product
Degree is affected greatly.
The content of the invention
In order to overcome the shortcoming of above-mentioned prior art, it is an object of the invention to provide it is a kind of based on linear model match
Interferometric phase image adaptive filter algorithm, available for existing, a large amount of phase speckle noises, local phase stripe order recognition frequency are very fast
And the changeable interferometric phase image of frequency, needed for existing interferogram filtering method can not meet processing during performance requirement, dry
Relate in data of synthetic aperture radar processing procedure, to the interferometric phase image of major-minor SAR image to carrying out efficient adaptive filtering
Processing；By estimation orientation to, distance to phase gradient, calculate different estimation beneath windows positions to optimal estimation phase, then
Adaptive-filtering processing is carried out to interferometric phase image by orientation and range direction phase compensation.
To achieve these goals, the technical solution adopted by the present invention is：
The interferometric phase image adaptive filter algorithm matched based on linear model, is comprised the following steps：
(1) main and auxiliary SAR complex image datas are inputted；
The main image data that (1a) input interference synthetic aperture radar InSAR primary antennas are obtained；
What the auxiliary antennas of (1b) input interference synthetic aperture radar InSAR were obtained has completed and the complete registering auxiliary figure of master image
As data；
(1c) inputs the processing parameter and systematic parameter of interference synthetic aperture radar InSAR imagings；
(2) multiple interferometric phase image is generated by interference treatment：
(2a) carries out interference treatment to the main and auxiliary SAR image of input, obtains multiple interferometric phase image；
(3) carry out gradient estimation, estimation orientation to distance to phase gradient：
(3a) uses phase gradient method of estimation, and orientation is carried out to multiple interferometric phase diagram data under different estimation windows
Interferometric phase gradient is estimated, obtains the orientation phase gradient estimate of each sample data in multiple interferometric phase image；
(3b) use phase gradient method of estimation, it is different estimation windows under to multiple interferometric phase diagram data carry out distance to
Interferometric phase gradient is estimated, obtains the distance of each sample data in multiple interferometric phase image to phase gradient estimate；
(4) according to the optimum angle gradient of each upward sample point of facet approximation criterion computer azimuth：
The phase gradient of each sample point is estimated in multiple interferometric phase image under the different windows that (4a) is obtained using step (3)
Value；
(4b) estimates the cost function of phase according to facet approximation criterion and one-dimensional linear model definition on orientation,
(4c) determines the minimum deflection estimation of orientation to determine the optimum angle Grad of orientation by cost function；
(5) orientation and range direction phase compensation are carried out：
The orientation estimation phase that (5a) is obtained using step (4c), orientation is carried out to sample point in multiple interferometric phase image
Phase compensation；
(5b) using step (3b) obtain distance to estimation phase, sample point in multiple interferometric phase image is entered row distance to
Phase compensation；
(6) mean filter processing is carried out to the interferometric phase image after orientation and range direction phase compensation；
(7) main and auxiliary SAR image data of the output after phase compensation and mean filter processing.
Compared with prior art, the beneficial effects of the invention are as follows：
First, the present invention is using estimation orientation to obtaining main and auxiliary SAR image area data respectively to phase gradient with distance
Between phase pushing figure, the filtering parameter of different zones each sample point under different windows is calculated, according to orientation phase
Minimum deflection estimation criterion carries out self-adapting window phase estimation to interferometric phase image.Overcome prior art interference synthetic aperture
Interferogram filtering approach application stationary window in radar data InSAR processing methods is filtered operation, in Local Phase
Position stripe order recognition differs greatly the region intensive with fringe distribution, and process performance can not meet the defect of engineer applied requirement, subtract
It is small that the difficulty of phase unwrapping is carried out to filtered interferometric phase image and more real absolute phase values can be obtained.
Second, one-dimensional linear model and facet approximation in orientation of the present invention, estimation orientation is to optimal
The method of phase gradient, then compensates filtering process to orientation phase.With prior art interference synthetic aperture radar
Interferogram filtering method in InSAR data processing is compared, and the present invention has preferably adapted to local phase striped along orientation
To SPA sudden phase anomalies, and be adaptation to the ground change, be effectively kept coherency information, improved major-minor SAR image pair
Between coherence.
Brief description of the drawings
Fig. 1 is flow chart of the invention.
Fig. 2 emulates data integral experiment result figure, and wherein Fig. 2 (a) is fixed filter window (3x3), and Fig. 2 (b) is fixation
Filter window (5x5), Fig. 2 (c) is fixed filter window (7x7), and Fig. 2 (d) is adaptive-filtering window, and Fig. 2 (e) misses for phase
Difference figure (3x3), Fig. 2 (f) is phase error figure (5x5), and Fig. 2 (g) is phase error figure (7x7), and Fig. 2 (h) is phase error figure
(adaptive-filtering window).
The local experimental result picture of Fig. 3 emulation data, wherein Fig. 3 (a) is interferometric phase image before filtering, and Fig. 3 (b) filters to be fixed
Ripple window (3x3), Fig. 3 (c) is fixed filter window (9x9), and Fig. 3 (d) is adaptive-filtering window, and Fig. 3 (e) is local phase
Error Graph (13x13), Fig. 3 (f) is local phase errors figure (15x15), and Fig. 3 (g) is local phase errors figure (17x17), Fig. 3
(h) it is local phase errors figure (adaptive-filtering window).
Embodiment
Describe embodiments of the present invention in detail with reference to the accompanying drawings and examples.
Referring to the drawings 1, specific implementation step of the invention is as follows：
Step 1, input SAR image data and auxiliary parameter.
The main image data and interference synthetic aperture radar InSAR that interference synthetic aperture radar InSAR primary antennas are obtained
Auxiliary antenna obtain with being used in master image completely registering auxiliary view data and processing procedure and systematic parameter and imaging
The related auxiliary parameter of processing is input in system, the major-minor SAR image of input to meet interference treatment coherence and into
As the quality requirement in terms of quality.
Step 2, multiple interferometric phase image.
Major-minor SAR image to input carries out interference treatment, obtains multiple interferometric phase image；
Step 3, phase gradient is estimated
Concrete methods of realizing is illustrated to exemplified by phase gradient by estimation orientation below.Using the phase of second order auto-correlation function
Gradient estimation algorithm, quickly steadily estimates each data of interference pattern piecemeal again, the orientation phase ladder in each range cell
Degree.The estimation of orientation interferometric phase gradient is carried out using this method, the range cell that multiple interferometric phase image coordinate is n is estimated first
Orientation single order covariance and second order covariance：
Wherein, C1,nIt is the orientation single order covariance of n range cell, C to represent multiple interferometric phase image coordinate2,nRepresent multiple
Interferometric phase image coordinate is the orientation second order covariance of n range cell, and I (m, n) represents multiple interferometric phase image, and m represents multiple
Interference pattern orientation coordinate, n represents multiple interference pattern range cell coordinate, and M represents multiple interference pattern Data in Azimuth Direction length, I*(·)
Expression takes conjugate operation, and Σ represents sum operation.
Then according still further to following formula, estimation orientation is to interferometric phase gradient：
Wherein, knDenotation coordination is the orientation interferometric phase gradient of n range cell, C1,nRepresent multiple interferometric phase image
Coordinate is the orientation single order covariance of n range cell, C2,nRepresent side of the multiple interferometric phase image coordinate for n range cell
Position is to second order covariance, and exp () represents constant e index operation, and arg () represents to take phase angle to operate.
Step 4, the optimum angle gradient of each upward sample point of computer azimuth.
First, the phase gradient of each sample point is estimated in multiple interferometric phase image under the different windows obtained using step (3)
Evaluation；Then the cost function of phase is estimated on orientation according to facet approximation criterion and one-dimensional linear model definition；Most
Determine the minimum deflection estimation of orientation to determine the optimum angle Grad of orientation by cost function afterwards.
, can be with one-dimensional come approximate for orientation because SAR complex images have two dimensions.Estimate in fixed
Count in window, carried out curve fitting by the sample data in window using least square method, these sample points can be approx
Straight line is fitted, the gradient (slope) value of this straight line is equal to the estimation phase of orientation.Join when window size changes
Cause the straight incline degree fitted also different with the data difference of fitting, therefore, the different window for same sample point
The different estimation phase of mouth size correspondence.
It is located at for anisotropic each sample point on Two-dimensional Surfaces, but can will be individual in part filter window
Sample point is approximately considered at grade.That is the phase estimation Grad each put is equal.The intraoral sample of estimating window
The phase estimation Grad of point can be represented by the formula：
Im=sqrt (- 1)
Wherein im is imaginary unit, sa(i, j) is the estimation phase of estimation window central pixel point (i, j), win and window
Relation between size is：Window size=2 × win+1,It is then phase estimation Grad, exp () represents constant e
Index operation.
Step 5, phase compensation filter.
First, the orientation phase parameter obtained using step (4), orientation is carried out to sample point in multiple interferometric phase image
Phase compensation；Then using step (3) obtain distance to phase parameter, sample point in multiple interferometric phase image is entered row distance to
Phase compensation；An additive phase is multiplied by the basis of original phase.
Step 6, mean filter processing is carried out to the interferometric phase image after orientation, range direction phase compensation.
Step 7, interferometric phase image of the output after phase compensation and mean filter.
2nd, emulation data processing experiment：
Emulating data experiment, satellite is platform in a distributed manner, and the simulation parameter of InSAR systems is as shown in the table：
| Parameter | Value | Parameter | Value |
| Satellite altitude (km) | 600 | Baseline length (m) | 813.5 |
| Radar downwards angle of visibility (°) | 45 | Distance is to sample rate (MHz) | 160 |
| Azimuth resolution (m) | 3 | Range resolution ratio (m) | 3 |
| Master image Doppler center (Hz) | 0 | Auxiliary image Doppler center (Hz) | 0 |
Fig. 2 (a) (e) is respectively stationary window (3x3) filtered interferometric phase image and phase error figure, Fig. 2 (b) (f)
Respectively stationary window (5x5) filtered interferometric phase image and phase error figure, Fig. 2 (c) (g) is respectively stationary window
(7x7) filtered interferometric phase image and phase error figure.Fig. 2 (d) (h) is that the method for the invention handles obtained interference phase
Bitmap and phase error figure.As shown in Figure 2, noise level is with showing in the interferometric phase image obtained through the method for the invention processing
There is method compared to lower.
Fig. 3 (a) is the interferometric phase image before regional area is filtered, and Fig. 3 (b) filters for regional area stationary window (3x3)
Interferometric phase image afterwards, Fig. 3 (c) is regional area stationary window (9x9) filtered interferometric phase image, and Fig. 3 (d) is partial zones
The filtered interferometric phase image of domain self-adapting window, Fig. 3 (e) is that the filtered phase of regional area stationary window (13x13) is missed
Difference figure, Fig. 3 (f) is the filtered phase error figure of regional area stationary window (15x15), and Fig. 3 (g) is regional area fixed window
Mouth (17x17) filtered phase error figure, Fig. 3 (h) is the filtered phase error figure of self-adapting window.From the figure 3, it may be seen that right
In interferometric phase image for striped close quarters, phase error in obtained interferometric phase image is handled through the method for the invention
It is smaller compared with the conventional method.
Table 1 is the filter result contrast of adaptive-filtering window and stationary window, and comparison other includes the big of phase error
The small and residual number counted out.
The self-adapting window of table 1 is compared with stationary window filter result
| Window size | 3x3 | 5x5 | 7x7 | 9x9 | 11x11 | 13x13 | Self-adapting window |
| It is residual to count out (number) | 6997 | 414 | 49 | 31 | 26 | 29 | 25 |
| Phase error (rad) | 0.577 | 0.226 | 0.206 | 0.210 | 0.214 | 0.225 | 0.205 |
It is of the present invention it was found from the statistical result of Fig. 2, Fig. 3 result and table 1 to residual number and phase error
Method can not only preferably filter out the phase noise in interferometric phase image, and can preferably keep the continuous of interferometric phase
Property.
Claims (7)
1. a kind of interferometric phase image adaptive filter algorithm matched based on linear model, it is characterised in that comprise the following steps：
Step (1), inputs main and auxiliary SAR complex image datas；
Step (2), multiple interferometric phase image is generated by interference treatment；
Step (3), carry out gradient estimation, estimation orientation to distance to phase gradient：
Using phase gradient method of estimation, orientation interferometric phase is carried out to multiple interferometric phase diagram data under different estimation windows
Gradient is estimated, obtains the orientation phase gradient estimate of each sample data in multiple interferometric phase image；
Using phase gradient method of estimation, distance is carried out to multiple interferometric phase diagram data under different estimation windows to interferometric phase
Gradient is estimated, obtains the distance of each sample data in multiple interferometric phase image to phase gradient estimate；
Step (4), according to the optimum angle gradient of each upward sample point of facet approximation criterion computer azimuth, step is as follows：
The phase gradient of each sample point is estimated in multiple interferometric phase image under step (4a), the different windows obtained using step (3)
Evaluation；
Step (4b), the cost letter of phase is estimated according to facet approximation criterion and one-dimensional linear model definition on orientation
Number；
Step (4c), determines the minimum deflection estimation of orientation to determine the optimum angle gradient of orientation by cost function
Value；
Step (5), carries out orientation and range direction phase compensation；
Step (6), mean filter processing is carried out to the interferometric phase image after orientation and range direction phase compensation；
Step (7), exports the main and auxiliary SAR image data after phase compensation and mean filter processing.
2. the interferometric phase image adaptive filter algorithm matched according to claim 1 based on linear model, it is characterised in that
In the step (1), input step is as follows：
Step (1a), the main image data that input interference synthetic aperture radar InSAR primary antennas are obtained；
Step (1b), it is completely registering auxiliary with master image that what the auxiliary antennas of input interference synthetic aperture radar InSAR were obtained completed
View data；
Step (1c), inputs the processing parameter and systematic parameter of interference synthetic aperture radar InSAR imagings.
3. the interferometric phase image adaptive filter algorithm matched according to claim 1 based on linear model, it is characterised in that
In the step (2), the interference treatment includes the interference treatment to main and auxiliary SAR complex image datas, and processing is real by following formula
It is existing：
I=Rconj (S)
Wherein, I represents the interference pattern of main and auxiliary SAR complex images, i.e. interferometric phase image again, and R represents main SAR complex image datas,
S represents auxiliary SAR complex image datas, and conj () represents to take conjugate operation.
4. the interferometric phase image adaptive filter algorithm matched according to claim 1 based on linear model, it is characterised in that
The phase gradient method of estimation is comprised the following steps that：
The first step：According to the following formula, estimate respectively multiple interferometric phase image coordinate for the orientation single order covariance of n range cell and
Second order covariance：
<mrow>
<msub>
<mi>C</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>n</mi>
</mrow>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mi>M</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</munderover>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<msup>
<mi>I</mi>
<mo>*</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>+</mo>
<mn>1</mn>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>C</mi>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>n</mi>
</mrow>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mi>M</mi>
<mo>-</mo>
<mn>2</mn>
</mrow>
</munderover>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<msup>
<mi>I</mi>
<mo>*</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mo>+</mo>
<mn>2</mn>
<mo>,</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
Wherein, C1,nIt is the orientation single order covariance of n range cell, C to represent multiple interferometric phase image coordinate2,nRepresent multiple interference
Phase diagram coordinate is the orientation second order covariance of n range cell, and I (m, n) represents multiple interferometric phase image, and m represents multiple interference
Figure orientation unit coordinate, n represents multiple interference pattern distance to unit coordinate, and M represents multiple interference pattern Data in Azimuth Direction length, I*
() represents to take conjugate operation, and ∑ represents sum operation；
Second step：According to the following formula, estimation orientation is to interferometric phase gradient：
<mrow>
<msub>
<mi>k</mi>
<mi>n</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mi>arg</mi>
<mo>{</mo>
<msub>
<mi>C</mi>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>n</mi>
</mrow>
</msub>
<mi>exp</mi>
<mo>&lsqb;</mo>
<mo>-</mo>
<mi>j</mi>
<mn>2</mn>
<mi>arg</mi>
<mo>{</mo>
<msub>
<mi>C</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>n</mi>
</mrow>
</msub>
<mo>}</mo>
<mo>&rsqb;</mo>
<mo>}</mo>
</mrow>
<mn>2</mn>
</mfrac>
<mo>+</mo>
<mi>arg</mi>
<mo>{</mo>
<msub>
<mi>C</mi>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>n</mi>
</mrow>
</msub>
<mo>}</mo>
</mrow>
Wherein, knDenotation coordination is the orientation interferometric phase gradient of n multiple interference pattern range cell, and n represents multiple interference pattern distance
Unit coordinate, C1,nIt is the orientation single order covariance of n range cell, C to represent multiple interferometric phase image coordinate2,nRepresent multiple interference
Phase diagram coordinate is the orientation second order covariance of n range cell, and j is imaginary unit, and exp () represents constant e index
Operation, arg () represents to take phase angle to operate.
5. the interferometric phase image adaptive filter algorithm matched according to claim 1 based on linear model, it is characterised in that
One-dimensional linear model and facet approximation criterion in the step (4b)：
Because SAR complex images have two dimensions, with it is one-dimensional come apparent azimuth to, it is intraoral in fixed estimating window, pass through window
Intraoral sample data is carried out curve fitting using least square method, and sample point approx fits straight line, this straight line
The gradient be slope value be equal to orientation estimation phase；The data difference of fitting is participated in when window size changes causes fitting
The straight incline degree gone out is also different, therefore, the different estimation phase of different window sizes correspondence for same sample point；
It is located at for anisotropic each sample point on a Two-dimensional Surfaces, but can be by individual sample in part filter window
Point is approximately considered at grade, i.e., the phase estimation Grad each put is equal, and the phase of the intraoral sample point of estimating window is estimated
Meter Grad is represented with following formula：
Wherein im is imaginary unit, sa(i, j) is the estimation phase of estimation window central pixel point (i, j), win and window size
Between relation be：Window size=2win+1, thenFor phase estimation Grad, exp () represents constant e finger
Number operation；
Optimal estimation phase is determined by minimum deflection estimation in the step (4c)：
I=Rconj (S)
Error_min (i, j)=min (error { i, j, win })
Sa=sa (i, j)
Wherein, I (i, j) represents multiple interferometric phase image, and i represents multiple interferometric phase image orientation unit coordinate, and j represents multiple interference phase
Bitmap distance to unit coordinate,For orientation phase estimation Grad, M represents that multiple interferometric phase image Data in Azimuth Direction is long
Degree, N represents multiple interferometric phase map distance to data length, and error (i, j) represents estimated bias, and error_min (i, j) is represented
Position is minimum deflection estimation of (i, j) place sample point under different windows, and sa represents that minimum deflection estimates corresponding orientation
Optimal estimation phase.
6. the interferometric phase image adaptive filter algorithm matched according to claim 1 based on linear model, it is characterised in that
The phase compensation includes：
Step (5a), the orientation phase gradient estimate obtained using step (4b), is clicked through to sample in multiple interferometric phase image
Row orientation phase compensation；
Step (5b), the distance obtained using step (3b) is clicked through to phase gradient estimate to sample in multiple interferometric phase image
The phase compensation of line-spacing descriscent.
7. the interferometric phase image adaptive filter algorithm matched according to claim 1 based on linear model, it is characterised in that
The phase compensation is realized using following formula：
Win_slope (i, j)=exp (imsa (i, j) [- win:win])*exp(ima·sr(i,j)·[-win:
win])
Corr (i, j)=mean (mean (I (i, j) * win_slope (i, j)))
Phase (i, j)=angle (corr (i, j))
Wherein, angle () represents to take plural number phase angle computing, and win_slope (i, j) represents orientation and distance to compensation phase
The size of position, corr (i, j) represents the coefficient correlation of the multiple interferometric phase image after phase compensation and mean filter, phase
(i, j) represents the phase value of multiple interferometric phase image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510427496.5A CN105116410B (en) | 2015-07-20 | 2015-07-20 | The interferometric phase image adaptive filter algorithm matched based on linear model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510427496.5A CN105116410B (en) | 2015-07-20 | 2015-07-20 | The interferometric phase image adaptive filter algorithm matched based on linear model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105116410A CN105116410A (en) | 2015-12-02 |
| CN105116410B true CN105116410B (en) | 2017-09-05 |
Family
ID=54664443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510427496.5A Expired - Fee Related CN105116410B (en) | 2015-07-20 | 2015-07-20 | The interferometric phase image adaptive filter algorithm matched based on linear model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105116410B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108663678B (en) * | 2018-01-29 | 2022-01-18 | 西北农林科技大学 | Multi-baseline InSAR phase unwrapping algorithm based on mixed integer optimization model |
| CN109143234B (en) * | 2018-07-16 | 2021-03-26 | 西安空间无线电技术研究所 | InSAR filtering method and system for estimating interference phase gradient based on FFT high precision |
| CN116224332B (en) * | 2023-01-17 | 2023-10-20 | 中国矿业大学 | Radar interference phase quality estimation method for coordinating multiple indexes |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5463397A (en) * | 1993-10-25 | 1995-10-31 | Hughes Aircraft Company | Hyper-precision SAR interferometry using a dual-antenna multi-pass SAR system |
| CN103197316B (en) * | 2013-03-26 | 2015-03-18 | 中国科学院电子学研究所 | Protective phase airborne interference synthetic aperture radar (SAR) terrain altitude self-adaption motion compensation method |
| CN103454633A (en) * | 2013-07-12 | 2013-12-18 | 电子科技大学 | Interference SAR movable base line processing method based on back-projection algorithm |
| CN103543453B (en) * | 2013-10-28 | 2017-01-18 | 北京理工大学 | Elevation inversion method for geosynchronous orbit synthetic aperture radar interference |
| CN103809180B (en) * | 2014-03-12 | 2016-08-17 | 西安电子科技大学 | For InSAR topographic Pre-Filter processing method |
-
2015
- 2015-07-20 CN CN201510427496.5A patent/CN105116410B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN105116410A (en) | 2015-12-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107301661B (en) | High-resolution remote sensing image registration method based on edge point features | |
| CN111142105B (en) | ISAR imaging method for complex moving target | |
| WO2016086699A1 (en) | Wavelet domain insar interferometric phase filtering method in combination with local frequency estimation | |
| CN106249236B (en) | A kind of spaceborne InSAR long-short baselines image joint method for registering | |
| Gallant | Adaptive smoothing for noisy DEMs | |
| CN110764087B (en) | Sea surface wind direction inverse weighting inversion method based on interference imaging altimeter | |
| CN104200471B (en) | SAR image change detection based on adaptive weight image co-registration | |
| CN102402685A (en) | Method for segmenting three Markov field SAR image based on Gabor characteristic | |
| CN103473755B (en) | Based on the sparse denoising method of SAR image that change detects | |
| CN101685158B (en) | Hidden Markov tree model based method for de-noising SAR image | |
| CN103809180B (en) | For InSAR topographic Pre-Filter processing method | |
| CN105869146A (en) | Saliency fusion-based SAR image change detection method | |
| Zhu et al. | Improving TanDEM-X DEMs by non-local InSAR filtering | |
| CN105116410B (en) | The interferometric phase image adaptive filter algorithm matched based on linear model | |
| Du et al. | A novel multifractal estimation method and its application to remote image segmentation | |
| Thomas et al. | A robust motion estimation algorithm for PIV | |
| Yeh et al. | Applying adaptive LS-PIV with dynamically adjusting detection region approach on the surface velocity measurement of river flow | |
| CN111681272B (en) | SAR image processing method based on singular power spectrum | |
| Li et al. | An interferometric phase noise reduction method based on modified denoising convolutional neural network | |
| CN104680536A (en) | Method for detecting SAR image change by utilizing improved non-local average algorithm | |
| CN104200434A (en) | Non-local mean image denoising method based on noise variance estimation | |
| CN115951350A (en) | Permanent scatterer point extraction method, device, equipment and medium | |
| CN103871031A (en) | Kernel regression-based SAR image coherent speckle restraining method | |
| CN113205564B (en) | SAR intelligent target edge reconstruction method | |
| CN111220980A (en) | Ground-based SAR nonlinear atmospheric phase compensation method |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170905 Termination date: 20180720 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |