CN106056577B - SAR image change detection based on MDS-SRM Mixed cascading - Google Patents
SAR image change detection based on MDS-SRM Mixed cascading Download PDFInfo
- Publication number
- CN106056577B CN106056577B CN201610333713.9A CN201610333713A CN106056577B CN 106056577 B CN106056577 B CN 106056577B CN 201610333713 A CN201610333713 A CN 201610333713A CN 106056577 B CN106056577 B CN 106056577B
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- region
- merging
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 22
- 230000008859 change Effects 0.000 title claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 17
- 238000012360 testing method Methods 0.000 claims abstract description 13
- 230000001174 ascending effect Effects 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000011218 segmentation Effects 0.000 claims description 7
- 230000001427 coherent effect Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000013316 zoning Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000005267 amalgamation Methods 0.000 abstract description 9
- 238000004458 analytical method Methods 0.000 abstract description 4
- 230000003068 static effect Effects 0.000 abstract description 2
- 238000012545 processing Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 230000002146 bilateral effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10044—Radar image
Abstract
The invention discloses a kind of Mixed cascading change detecting method based on MDS-SRM mainly solves the problems, such as in existing SRM algorithm because the factors such as static ordering cause variation detection inaccurate.Implementation step are as follows: 1. two phase single-channel SAR images of input carry out denoising to it respectively;2. with image configuration log ratio figure and average ratio value figure after denoising, and the two is superimposed as a width binary channels differential image;3. a pair binary channels differential image merges, first time amalgamation result is obtained;4. merging again on the basis of first time amalgamation result, second of amalgamation result is obtained;5. the gray average in each region in pair second of amalgamation result carries out ascending sort, and carries out region merging technique, obtain finally changing testing result.The present invention can effectively increase the detection accuracy of region of variation in SAR image, can be used for the positioning in disaster region and the dilatation analysis in city.
Description
Technical field
The invention belongs to field of image processings, further relate to the detection method in two phase image change regions, can use
In the positioning in disaster region and the dilatation analysis in city.
Background technique
With the fast development of remote sensing technology in recent years, remotely-sensed data quantity is growing, and is widely used in environment prison
The fields such as survey, atmospheric analysis and urban planning.Wherein SAR image variation detection is applied to military and civilian field, mainly relates to
And it is hit in the positioning in such as flood, fire and earthquake natural calamity region, the dilatation analysis in city and Military Application
The assessment of effect, so research SAR image variation detection is of great significance.
SAR image changes the pretreatment of phase images when detection mainly includes, generates disparity map and extraction region of variation three
Step.Wherein, region of variation part is extracted, unsupervised detection can be substantially divided into and has supervisory detection method.In comparison,
Unsupervised approaches can directly obtain variation testing result by cluster or partitioning algorithm processing difference image, compare and agree with
The characteristics of variation detection lacks prior information, therefore be widely used.
Unsupervised approaches based on image segmentation include: maximum between-cluster variance algorithm OTSU, Threshold Segmentation Algorithm K&I, system
It counts partitioning algorithm MRF and statistical regions merges algorithm SRM etc..Wherein SRM algorithm has the probability distribution independent of data false
If this makes it be more suitable for SAR image variation detection with the advantages of possessing preferable anti-noise ability.But SRM algorithm only considers region
Equal value difference carries out static ordering, will lead to the increase of merging error probability, and single use SRM can not obtain final variation inspection
Survey result.
Summary of the invention
It is an object of the invention to be directed to the deficiency of above-mentioned prior art, propose a kind of based on multilayer dynamic order Statistical Area
Domain merges the Mixed cascading change detecting method of algorithm MDS-SRM, to realize the detection of region of variation in SAR image, promotes inspection
The accuracy rate of survey.
It realizes the technical scheme is that being denoised to input picture, using logarithm ratio and the building of mean value ratio method
Binary channels differential image carries out the extraction that region of variation is completed in processing to differential image by Mixed cascading structure, wherein first
Disparity map is converted to super-pixel space from pixel space using SRM algorithm by grade, and difference is completed using MDS-SRM algorithm in the second level
Image small area finally obtains final variation testing result using simplified SRM algorithm to the merging process in big region,
Realize that step includes the following:
(1) the single-channel SAR image X of the identical region of two width different times of input1、X2, and gone using non-local mean algorithm
Two images I except the coherent speckle noise in the two images, after being denoised1And I2;
(2) with the image I after denoising1And I2, construct log ratio figure DI1With average ratio value figure DI2;
(3) by the log ratio disparity map DI of construction1With average ratio value disparity map DI2It is overlapped, obtains into a width bilateral
Road differential image DI;
(4) merge algorithm using statistical regions to merge the pixel in binary channels differential image DI, complete difference
Transformation of the figure from pixel space to regional space obtains merging image DT for the first time1;
(5) merge algorithm to image DT after primary merge using multilayer dynamic order statistical regions1In region carry out two
Secondary merging obtains the secondary merging image DT for possessing large area2:
(5a) traversal is primary to merge image DT1In adjacent region, determine region to number M;
The similarity f (R', R) of the zoning pair (5b):
R' and R indicates two adjacent regions, H in formula1And H2The feature vector in adjacent two region is respectively indicated,WhereinRepresent the average pixel value in the channel k Zone R domain, A (Rk)
Indicate the statistic histogram matrix in Zone R domain, | R | indicate the number of pixels of region R;S-1For the covariance of adjacent area feature vector
The inverse matrix of matrix S (a, b);
(5c) sorts similarity f (R', R) from small to large, obtains merging sequence list Index (i), and i is that traversal merges
The pointer of sequence list selectes the region of similarity f (R, R') minimum value to (R, R') if i=1;
(5d) determines whether the region of minimum value merges (R, R') according to the following formula: if meetingThen merge, calculate adjacent area at this time to number M and updates adjacent area pair
The similarity f of (R, R')n+1(R, R'):
Wherein,Parameter Q represents Statistical Complexity, obtains different degrees of segmentation by adjusting Q
As a result, R|R|Indicate | R | the region of a pixel, and have | | RR|||≤(|R|+1)min(R|,g), g=256, | R | indicate that region contains
Some number of pixels, constant| I | it is the pixel number that image I includes,Represent the average picture in the channel k Zone R domain
Element value;fn(R, R') represent n-th merge before the region R' and Zone R domain similarity, φ andRespectively indicate this similarity
With the weight of history similarity,It updates merging sequence list Index (i), if i=1, selectes fn+1(R, R') is minimum
The region of value judges whether the region of minimum value merges (R, R') to (R, R') again;
Otherwise, i=i+1 judges whether i≤M is true, if so, the corresponding adjacent area pair of Index (i) is then taken, weight
Multiple step (5d);If not, merging terminates, and obtains secondary merging image DT2;
(6) to secondary merging image DT2In each region gray average carry out ascending sort obtain merging sequence arrange
Then table successively carries out region merging technique to the region in merging sequence list using the merging criterion in SRM algorithm, obtain final
Variation testing result.
The present invention has the advantage that
First, the present invention can be made up by constructing binary channels disparity map using the advantage of average ratio and logarithm ratio respectively
Since SAR image contains only single-channel data in experiment, it is unable to fully the shortcomings that utilizing multichannel restriction ability in SRM algorithm;
Second, the present invention by using optimization ranking criteria, joined on the basis of equal value difference region histogram and
To reduce the probability that mistake merges occurs for area discrepancy feature so that sequence is more reasonable;
Third, the present invention, which merges change detecting method structure using Mixed cascading, can effectively improve SAR image variation inspection
The precision of survey.
Detailed description of the invention
Fig. 1 is implementation flow chart of the invention;
Fig. 2 is the result figure for carrying out SAR image variation detection to Bern data set with the present invention;
Fig. 3 is the present invention and existing RFLICM algorithm, SRM algorithm to Bern data set, Ottawa data set and the Yellow River number
According to the variation testing result figure of collection.
Specific embodiment
With reference to the accompanying drawing, realization step of the invention and effect are described in further detail.
Referring to Fig.1, steps are as follows for realization of the invention:
Step 1, the single-channel SAR image X of the identical region of two width different times of input1、X2, and calculated using non-local mean
Method removes the coherent speckle noise in the two images, the image I after being denoised1、I2。
(1a) chooses the noise image X before changing in data set1={ X1(i) | i ∈ I }, I is image pixel fields, to image
In any one pixel i, using non-local mean algorithm denoise, obtain the gray scale estimated value of the point are as follows:
Wherein, ω (i, j) is weight, indicates the similarity degree between ith pixel and j-th of pixel, 0≤ω (i, j)
≤ 1, andTraverse image X1In all pixel, the image I after obtaining the denoising of the first width1;
(1b) chooses the noise image X after changing in data set2={ X2(i) | i ∈ I }, to any one pixel in image
Point i is denoised using non-local mean algorithm, obtains the gray scale estimated value of the point are as follows:
Wherein, weight ω (i, j) indicates the similarity degree between ith pixel and j-th of pixel, 0≤ω (i, j)≤1
AndTraverse image X2In all pixel, the image I after obtaining the denoising of the second width2。
Step 2, with the two images I after denoising1、I2, construct log ratio figure DI1With average ratio value figure DI2。
(2a) uses image I after denoising1、I2Middle coordinate is the pixel value of the pixel of (i, j), calculates log ratio disparity map
DI1Middle pixel coordinate is the pixel value of (i, j):Image I after traversal denoising1、I2In it is all
Pixel obtains log ratio disparity map DI1, wherein I1(i, j), I2(i, j) is respectively image I1, I2Middle coordinate be (i, j) as
The pixel value of vegetarian refreshments;
(2b) uses the image I after denoising1、I2Middle coordinate is the pixel value of the pixel of (i, j), calculates average ratio value difference
Scheme DI2Middle pixel coordinate is the pixel value of (i, j):Scheme after traversal denoising
As I1、I2In all pixel, obtain average ratio value disparity map DI2, wherein μ1(i, j), μ2(i, j) is respectively image I1, I2
In with coordinate be (i, j) pixel centered on the field 2*2 pixel average.
Step 3, the log ratio disparity map DI that will be obtained1With average ratio value disparity map DI2It is superimposed as a width binary channels difference
Image DI.
Step 4, merge algorithm using statistical regions to merge the pixel in binary channels differential image DI, it is poor to complete
Transformation of the different figure from pixel space to regional space obtains merging image DT for the first time1。
(4a) calculates similarity weight to pixel each pair of in binary channels differential image DI:
Wherein pkAnd p'kFor two pixel values adjacent each other;
Similarity weight is carried out ascending sort by (4b) from small to large, and according to collating sequence, and successively selected pixels are to sentencing
Whether the pixel of breaking is to merging: if meetingThen merge, whereinRepresent the channel k
The average pixel value in Zone R domain, Zone R domain are pkPixel affiliated area,It is complicated that parameter Q represents statistics
Degree obtains different degrees of segmentation result, R by adjusting Q|R|Indicate | R | the regional ensemble of a pixel, and have | | RR||≤
(n+1)min(|R|,g), g=256, constant| I | it is the pixel number that image I includes;Otherwise, nonjoinder;When each
Group pixel merges image DT to when all completing this deterministic process to get to first time1。
Step 5, merge algorithm using multilayer dynamic order statistical regions and image DT is merged to first time1In region carry out
It is further to merge, obtain second of amalgamation result DT2。
(5a) traversal merges image DT for the first time1In adjacent region, determine region to number M;
The similarity f (R', R) of the zoning pair (5b):
R' and R indicates two adjacent each other regions, H in formula1And H2The feature vector in adjacent two region is respectively indicated,WhereinRepresent the average pixel value in the channel k Zone R domain, A (Rk)
Indicate the statistic histogram matrix in Zone R domain, | R | indicate the number of pixels of region R;S-1For the covariance of adjacent area feature vector
The inverse matrix of matrix S (a, b);
(5c) sorts similarity f (R', R) from small to large, obtains merging sequence list Index (i), and i is that traversal merges
The pointer of sequence list selectes the region of similarity f (R, R') minimum value to (R, R') if i=1;
(5d) determines whether the region of minimum value merges (R, R') according to the following formula: if meetingThen merge, calculates adjacent area at this time to number M, update adjacent area pair
The similarity f of (R, R')n+1(R, R') and merging sequence list Index (i) select f if i=1n+1The area of (R, R') minimum value
Domain judges whether the region of minimum value merges (R, R') to (R, R') again, fn+1The calculation formula of (R, R') is as follows:
Wherein,Parameter Q represents Statistical Complexity, obtains different degrees of segmentation by adjusting Q
As a result, R|R|Indicate | R | the region of a pixel, and have | | R|R|||≤(|R|+1)min(|R|,g), g=256, | R | indicate region
The number of pixels contained, constant| I | it is the pixel number that image I includes,Represent being averaged for the channel k Zone R domain
Pixel value;fn(R, R') represent n-th merge before the region R' and Zone R domain similarity, φ andRespectively indicate this similarity
With the weight of history similarity,
Otherwise, i=i+1 judges whether i≤M is true, if so, the corresponding adjacent area pair of Index (i) is then taken, weight
Multiple step (5d);If not, merging terminates, and obtains second of merging image DT2。
Step 6, merge image DT to second2In each region gray average carry out ascending sort, then use
Statistical regions merge the merging criterion in algorithm and carry out region merging technique, obtain finally changing testing result figure.
(6a) calculates second and merges image DT2In each region gray average
P in formulaiIndicate the ratio of all pixels point in region shared by pixel of the gray value for i in image, k indicates image
Gray level maximum value, and carried out ascending sort;
(6b) according to collating sequence, successively whether chosen area judges the region to closing to and using following merging criterion
And: if meetingThen merge, whereinRepresent the mean pixel in the channel k Zone R domain
Value,Parameter Q represents Statistical Complexity, obtains different degrees of segmentation result, R by adjusting Q|R|
Indicate | R | the regional ensemble of a pixel, and have | | R|R|||≤(n+1)min(|R|,g), g=256, constant| I | it is figure
The pixel number for including as I;Otherwise, nonjoinder.
Change testing result figure after traversal all areas to get to final.
Effect of the invention is further illustrated in conjunction with following emulation experiment:
1. simulated conditions
It is Intel (R) Core i5-34703.2GHZ, memory 8G, WINDOWS 7 operation that the present invention, which is in central processing unit,
On the PC of system, with the emulation experiment of MATLAB 2013b progress.
2. emulation content
Emulation 1, is changed detection to Bern data set using the method for the present invention, testing result is as shown in Figure 2, in which:
Fig. 2 (a) indicates image before the variation inputted;
Fig. 2 (b) indicates image after the variation inputted;
Fig. 2 (c) is indicated to image carries out the log ratio difference that log ratio is handled before the variation of input, after variation
Figure;
Fig. 2 (d) is indicated to image carries out the average ratio value difference that average ratio value is handled before the variation of input, after variation
Figure;
Fig. 2 (e) indicates to carry out the binary channels disparity map being superimposed as by log ratio disparity map and average ratio value disparity map
Merge obtained first time amalgamation result figure;
Fig. 2 (f) indicates second of the amalgamation result figure merged on the basis of first time amalgamation result figure;
Fig. 2 (g) indicates the final variation testing result further merged on the basis of second of amalgamation result figure
Figure.
As seen from Figure 2, the method for the present invention can effectively come out the Changing Area Detection in SAR image.
Emulation 2, using RFLICM algorithm, SRM algorithm and the method for the present invention to Bern data set, Ottawa data set and Huang
River data set is changed detection, as a result as shown in Figure 3, in which:
Fig. 3 (a) indicates that the input of three group data sets changes preceding image;
Image after the input variation of Fig. 3 (b) three group data sets of expression;
Fig. 3 (c) indicates the testing result of the RFLICM algorithm of three group data sets;
Fig. 3 (d) indicates the testing result of the SRM algorithm of three group data sets;
Fig. 3 (e) indicates the testing result of the MDS-SRM algorithm of three group data sets;
Fig. 3 (f) indicates the canonical reference figure of three group data sets.
As seen from Figure 3, compared to RFLICM algorithm and SRM algorithm, the method for the present invention can effectively improve SAR image
The detection accuracy of middle region of variation.
Above description is only example of the present invention, it is clear that for those skilled in the art, is being understood
After the content of present invention and principle, it may all carry out in form and details without departing substantially from the principle of the invention, structure
Various modifications and variations, but these modifications and variations based on inventive concept still scope of the presently claimed invention it
It is interior.
Claims (4)
1. a kind of SAR image change detection based on MDS-SRM Mixed cascading, comprising:
(1) the single-channel SAR image X of the identical region of two width different times of input1、X2, and should using the removal of non-local mean algorithm
Coherent speckle noise in two images, the two images I after being denoised1And I2;
(2) with the image I after denoising1And I2, construct log ratio disparity map DI1With average ratio value disparity map DI2;
(3) by the log ratio disparity map DI of construction1With average ratio value disparity map DI2It is overlapped, obtains a width binary channels difference
Image DI;
(4) using statistical regions merge algorithm the pixel in binary channels differential image DI is merged, complete disparity map from
Pixel space obtains merging image DT for the first time to the transformation of regional space1;
(5) merge algorithm using multilayer dynamic order statistical regions and image DT is merged to first time1In region carry out secondary conjunction
And obtain the secondary merging image DT for possessing large area2:
(5a) traversal merges image DT for the first time1In adjacent region, determine region to number M;
The similarity f (R', R) of the zoning pair (5b):
R' and R indicates two adjacent regions, H in formula1And H2The feature vector in adjacent two region is respectively indicated,WhereinRepresent the average pixel value in the channel k Zone R domain, A (Rk) table
Show the statistic histogram matrix in Zone R domain, | R | indicate the number of pixels of region R;S-1For the covariance square of adjacent area feature vector
The inverse matrix of battle array S (a, b), k indicate channel;
(5c) sorts similarity f (R', R) from small to large, obtains merging sequence list Index (i), and i is traversal merging sequence
The pointer of list selectes the region of similarity f (R, R') minimum value to (R, R') if i=1;
(5d) determines whether the region of minimum value merges (R, R') according to the following formula: if meetingThen merge, calculate adjacent area at this time to number M and updates adjacent area pair
The similarity f of (R, R')n+1(R, R'):
Wherein,Parameter Q represents Statistical Complexity, obtains different degrees of segmentation result by adjusting Q,
RRIt indicates the region of R pixel, and hasG=256, R indicate the number of pixels that region is contained, constant| I | it is the pixel number that image I includes,Represent the average pixel value in the channel k Zone R domain;fn(R, R') is represented
The similarity in the region R' and Zone R domain before n-th merges, φ andRespectively indicate the power of this similarity and history similarity
Weight,It updates merging sequence list Index (i), if i=1, selectes fn+1The region of (R, R') minimum value to (R, R'),
Judge whether the region of minimum value merges (R, R') again;
Otherwise, i=i+1 judges whether i≤M is true, if so, the corresponding adjacent area pair of Index (i) is then taken, repeats to walk
Suddenly (5d);If not, merging terminates, and obtains secondary merging image DT2;
(6) to secondary merging image DT2In each region gray average carry out ascending sort obtain merging sequence list, so
Region merging technique is successively carried out to the region in merging sequence list using the merging criterion in SRM algorithm afterwards, obtains final change
Change testing result.
2. the SAR image change detection according to claim 1 based on MDS-SRM Mixed cascading, wherein step (1)
It is middle that the coherent speckle noise being originally inputted in two images, the two images I after being denoised are removed using non-local mean algorithm1
And I2, it carries out as follows:
(1a) chooses the noise image X before changing in data set1={ X1(i) | i ∈ I }, I is image pixel fields, is appointed in image
Anticipate a pixel i, is denoised using non-local mean algorithm, obtains the gray scale estimated value of the point are as follows:
Wherein, ω (i, j) is weight, the similarity degree between expression ith pixel and j-th of pixel, 0≤ω (i, j)≤1,
AndTraverse image X1In all pixel, the image I after obtaining the denoising of the first width1;
(1b) chooses the noise image X after changing in data set2={ X2(i) | i ∈ I }, to any one pixel i in image,
It is denoised using non-local mean algorithm, obtains the gray scale estimated value of the point are as follows:
Wherein, weight ω (i, j) indicates the similarity degree between ith pixel and j-th of pixel, 0≤ω (i, j)≤1 andTraverse image X2In all pixel, the image I after obtaining the denoising of the second width2。
3. the SAR image change detection according to claim 1 based on MDS-SRM Mixed cascading, wherein step (2)
The middle image I with after denoising1、I2, construct log ratio disparity map DI1With average ratio value disparity map DI2, in accordance with the following steps into
Row:
(2a) uses image I after denoising1、I2Middle coordinate is the pixel value of the pixel of (i, j), passes through formulaConstruct log ratio disparity map DI1Middle coordinate is the pixel DI of (i, j)1(i, j), traversal denoising
Image I afterwards1、I2In all pixel to get arriving log ratio disparity map DI1;
(2b) uses the image I after denoising1、I2Middle coordinate is the pixel value of the pixel of (i, j), passes through formulaConstruct average ratio value disparity map DI2Middle coordinate is the pixel DI of (i, j)2
(i, j), image I after traversal denoising1、I2In all pixel to get arriving average ratio value disparity map DI2, wherein μ1(i, j), μ2
(i, j) is respectively image I1, I2In with coordinate be (i, j) pixel centered on 2*2 neighborhood territory pixel average value.
4. the SAR image change detection according to claim 1 based on MDS-SRM Mixed cascading, wherein step (4)
It is middle that the pixel in the differential image DI of building is merged using statistical regions merging algorithm, it carries out in accordance with the following steps:
(4a) calculates similarity weight to adjacent pixel each pair of in differential image DI:
Wherein pkAnd p'kFor the pixel value of adjacent pixel pair, k indicates channel;
Similarity weight is carried out ascending sort by (4b) from small to large, and according to collating sequence, successively selected pixels should to judgement
Whether pixel is to merging: if meetingThen merge, whereinRepresent the channel k Zone R
The average pixel value in domain, Zone R domain are pkPixel affiliated area,Parameter Q represents Statistical Complexity,
Different degrees of segmentation result, R are obtained by adjusting Q|R|Indicate | R | the regional ensemble of a pixel, and have | | RR||≤(n+
1)min(R,g), g=256, constant| I | it is the pixel number that image I includes;Otherwise, nonjoinder;When each group of picture
Element merges image DT to when all completing deterministic process to get to first time1。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610333713.9A CN106056577B (en) | 2016-05-19 | 2016-05-19 | SAR image change detection based on MDS-SRM Mixed cascading |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610333713.9A CN106056577B (en) | 2016-05-19 | 2016-05-19 | SAR image change detection based on MDS-SRM Mixed cascading |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106056577A CN106056577A (en) | 2016-10-26 |
CN106056577B true CN106056577B (en) | 2019-02-15 |
Family
ID=57177249
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610333713.9A Active CN106056577B (en) | 2016-05-19 | 2016-05-19 | SAR image change detection based on MDS-SRM Mixed cascading |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106056577B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109446894B (en) * | 2018-09-18 | 2021-10-22 | 西安电子科技大学 | Multispectral image change detection method based on probability segmentation and Gaussian mixture clustering |
CN109919862B (en) * | 2019-02-01 | 2021-08-20 | 北京佳格天地科技有限公司 | Radar image denoising system and method and computer equipment |
CN109738866B (en) * | 2019-02-28 | 2020-09-01 | 电子科技大学 | ALS and MDS-based optical non-ranging positioning method |
CN110163825B (en) * | 2019-05-23 | 2022-11-25 | 大连理工大学 | Human embryo heart ultrasonic image denoising and enhancing method |
CN113425266B (en) * | 2021-07-09 | 2023-02-28 | 上海市第一人民医院 | Skin cancer screening system based on infrared imaging |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101923711A (en) * | 2010-07-16 | 2010-12-22 | 西安电子科技大学 | SAR (Synthetic Aperture Radar) image change detection method based on neighborhood similarity and mask enhancement |
CN104680542A (en) * | 2015-03-15 | 2015-06-03 | 西安电子科技大学 | Online learning based detection method for change of remote-sensing image |
CN104778717A (en) * | 2015-05-05 | 2015-07-15 | 西安电子科技大学 | SAR image change detection method based on oriented difference chart |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6349937B2 (en) * | 2014-05-09 | 2018-07-04 | 日本電気株式会社 | Fluctuation detection apparatus, fluctuation detection method, and fluctuation detection program |
-
2016
- 2016-05-19 CN CN201610333713.9A patent/CN106056577B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101923711A (en) * | 2010-07-16 | 2010-12-22 | 西安电子科技大学 | SAR (Synthetic Aperture Radar) image change detection method based on neighborhood similarity and mask enhancement |
CN104680542A (en) * | 2015-03-15 | 2015-06-03 | 西安电子科技大学 | Online learning based detection method for change of remote-sensing image |
CN104778717A (en) * | 2015-05-05 | 2015-07-15 | 西安电子科技大学 | SAR image change detection method based on oriented difference chart |
Non-Patent Citations (2)
Title |
---|
Fast SAR Image Change Detection Using Bayesian Approach;Han Zhang 等;《The Scientific World Journal》;20140828;第2014卷;1-12 |
在感兴趣的区域层面上进行SAR图像变化检测的方法研究;万红林 等;《测绘学报》;20120430;第41卷(第2期);239-245 |
Also Published As
Publication number | Publication date |
---|---|
CN106056577A (en) | 2016-10-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106056577B (en) | SAR image change detection based on MDS-SRM Mixed cascading | |
US11852778B2 (en) | Static engine and neural network for a cognitive reservoir system | |
CN106228162B (en) | A kind of quick object identification method of mobile robot based on deep learning | |
CN107247938A (en) | A kind of method of high-resolution remote sensing image City Building function classification | |
CN106897681B (en) | Remote sensing image contrast analysis method and system | |
CN103473786B (en) | Gray level image segmentation method based on multi-objective fuzzy clustering | |
CN107784663A (en) | Correlation filtering tracking and device based on depth information | |
Pradhan et al. | Use of different bivariate statistical landslide susceptibility methods: A case study of Kulekhani watershed, Nepal | |
Moua et al. | Correcting the effect of sampling bias in species distribution modeling–A new method in the case of a low number of presence data | |
Erdogan Erten et al. | Combination of machine learning and kriging for spatial estimation of geological attributes | |
Kim et al. | Mapping local vulnerabilities into a 3D city model through social sensing and the CAVE system toward digital twin city | |
CN114998744A (en) | Agricultural machinery track field segmentation method based on motion and vision dual-feature fusion | |
Fouedjio | Clustering of multivariate geostatistical data | |
Eum et al. | Vehicle detection from airborne LiDAR point clouds based on a decision tree algorithm with horizontal and vertical features | |
Wang et al. | Multidimensional particle swarm optimization-based unsupervised planar segmentation algorithm of unorganized point clouds | |
Shafizadeh‐Moghadam et al. | On the spatiotemporal generalization of machine learning and ensemble models for simulating built‐up land expansion | |
CN111986215A (en) | Population calculation method and system based on high-precision building data | |
CN103268586B (en) | A kind of window fusion method based on diffusion theory | |
CN112464997B (en) | Multi-sensor collaborative monitoring method and system based on fuzzy-greedy search decision | |
CN104008376A (en) | Multispectral remote-sensing image mixed pixel decomposition method based on possibility center point clustering | |
CN106844948A (en) | Prediction of Landslide and device | |
CN105787493A (en) | BIM-based method for intelligent extraction of setting-out feature points | |
CN110633890A (en) | Land utilization efficiency judgment method and system | |
CN110763223A (en) | Sliding window based indoor three-dimensional grid map feature point extraction method | |
Tunçer et al. | Informed Design Platform |
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 |