CN114066816B - SAR image unsupervised change detection method based on static wavelet transformation extraction - Google Patents
SAR image unsupervised change detection method based on static wavelet transformation extraction Download PDFInfo
- Publication number
- CN114066816B CN114066816B CN202111227373.9A CN202111227373A CN114066816B CN 114066816 B CN114066816 B CN 114066816B CN 202111227373 A CN202111227373 A CN 202111227373A CN 114066816 B CN114066816 B CN 114066816B
- Authority
- CN
- China
- Prior art keywords
- image
- sar
- images
- difference
- change detection
- 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
- 230000008859 change Effects 0.000 title claims abstract description 32
- 238000001514 detection method Methods 0.000 title claims abstract description 30
- 230000003068 static effect Effects 0.000 title claims abstract description 25
- 230000009466 transformation Effects 0.000 title claims abstract description 17
- 238000000605 extraction Methods 0.000 title claims abstract description 11
- 238000001914 filtration Methods 0.000 claims abstract description 9
- 238000009499 grossing Methods 0.000 claims abstract description 5
- 238000000034 method Methods 0.000 claims description 34
- 238000010586 diagram Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 10
- 238000003064 k means clustering Methods 0.000 claims description 5
- 239000002356 single layer Substances 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 12
- 230000006870 function Effects 0.000 description 12
- 238000000354 decomposition reaction Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 230000004927 fusion Effects 0.000 description 4
- 239000000284 extract Substances 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008602 contraction Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 239000010410 layer Substances 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 238000012800 visualization 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/10—Image enhancement or restoration using non-spatial domain filtering
-
- 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 using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20064—Wavelet transform [DWT]
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20224—Image subtraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an SAR image unsupervised change detection method based on static wavelet transformation extraction, which comprises the following implementation steps: first is the pretreatment phase. Filtering and smoothing the two SAR images by using a Lee filter; the next is the phase of generating the disparity map. Generating a difference map using a logarithmic ratio operator; finally, the analysis of the difference map stage. The difference map is first decomposed using SWT2 (db 2), (two-dimensional static wavelet transform, db2 being the wavelet basis function) to obtain an approximate image, a horizontal detail image, a vertical detail image, and a diagonal detail image. And then the EM_GGM is used for respectively decomposing the four groups of images, finally, a second-order neighborhood window probability is selected to form a feature vector, and a result is obtained through K-means.
Description
Technical Field
The invention belongs to the technical field of image change detection, and relates to an SAR image unsupervised change detection method based on static wavelet transformation extraction.
Background
SAR is an active side-looking radar system and the imaging geometry is of the oblique projection type. Therefore, SAR images and optical images have large differences in imaging mechanism, geometric characteristics, radiation characteristics and the like. SAR realizes high-resolution microwave imaging by utilizing the synthetic aperture principle, and has the characteristics of all-weather, high resolution, large breadth and the like. Therefore, the SAR system has unique advantages in the aspects of disaster monitoring, environment monitoring, ocean monitoring, resource exploration, crop estimation, mapping, military and the like, can play a role which is difficult to play by other remote sensing means, and is therefore more and more valued in all countries of the world.
The basic flow paradigm of single polarization SAR image change detection is as follows: 1) preprocessing 2) generating a difference map 3) analyzing the difference map. The two steps of generating a difference map and analyzing the difference map are important research directions of SAR image change detection in recent years, and the purpose of the two steps is mainly to reduce the influence of speckle noise on SAR images.
1) Is to prepare the basis for generating the difference map.
2) The purpose of the generation of the difference map is to primarily distinguish unchanged classes from changed classes in the two SAR images, the generation of the difference map is to actually find a matrix capable of representing the distance between the two SAR images, and the matrix is the difference map after the visualization processing. We can construct a difference map of the same size as both by some sort of difference operation. The improved difference diagram generation algorithm is as follows, 1. Log-ratio (LR) operator operation is performed, and one-step Log operation is further performed on the basis of the ratio difference diagram. According to the method, speckle noise in the SAR image is converted into additive noise, and the difference image is subjected to logarithmic conversion to obtain nonlinear contraction, so that the contrast of the variable type and the non-variable type is enhanced. The logarithmic operation itself can reduce a large difference due to the ratio operation, so that the effect of the wild point on the background portion of the unchanged class can be further reduced, and the effect is effective when the changed region is smaller than the unchanged region, but the pixel value of the edge region is easily blurred because the logarithmic operation has strong contractility. 2. The Mean-ratio (MR) operator operation, the object to be compared is not the corresponding isolated pixel point, but the average value of the neighborhood where the pixel point is located, so that the method has a certain degree of inhibition effect on the independently-appearing wild point, but is not easy to effectively inhibit the influence of noise if the noise does not appear in a punctiform form but appears in a flaking form due to the lack of telescopic transformation. 3. And combining a difference map method (Combined Difference Image, CDI), wherein the method carries out parameter weighting on the difference map and the LR difference map to obtain a new difference map. The CDI method carries out mean value filtering and median value filtering on the difference value difference image and the LR difference image respectively, noise interference and wild points are preliminarily removed, and then a final fusion difference image is obtained by using manually weighted parameters. The method is simple and feasible, is suitable for parallel processing and has high speed; however, the method comprises manual parameters, and the optimal parameter values can be obtained through multiple tests, so that automatic selection is not easy to carry out according to the self-properties of the images. ) And fourthly, a Neighborhood-based Ratio difference map algorithm (NR), wherein the NR operator is a weighted average of the contrast value difference map and the MR difference map. This weight may characterize whether the location of the center pixel is in a homogeneous region or a heterogeneous region, with a low value corresponding to the homogeneous region and a high value corresponding to the heterogeneous region. The method fully combines the gray information and the space information of the pixel points, the weighting parameters are completely determined by the self-properties of the images, and the robustness of the difference image construction is improved. 5. The Wavelet Fusion (WF) method firstly performs Wavelet transformation on the generated LR and MR difference maps respectively, and extracts the low frequency band of the MR difference map and the high frequency band of the LR difference map respectively, that is, extracts the whole information of the MR difference map and the detailed information of the LR difference map. And then fusing the LL, the LH, the HL and the HH according to a fusion rule based on the neighborhood to generate a new wavelet transformation diagram. And finally, carrying out wavelet inverse transformation to obtain a WF fusion difference graph. This approach combines the properties of wavelet transforms, combining the advantages of both LR and MR disparity maps. 6. Constructing a difference map (Intensity and Texture, IT) by combining the texture and intensity characteristics of the SAR images, carrying out sparse and low-rank-coefficient decomposition on the two input SAR images to respectively obtain corresponding intensity and texture information, respectively constructing the difference map on the two information, and then fusing the two information. The method not only extracts the main change area in the SAR image, but also can prevent the speckle noise from influencing the performance of the difference map, and especially has stronger robustness in maintaining the performance.
3) The analysis of the difference map is carried out, and after the difference map is generated, the analysis is needed to be carried out, so that a black-white binary image is finally generated. There are four commonly used analytical methods: threshold analysis, cluster analysis, cut-of-graph analysis, and level set analysis. The threshold analysis method is to divide the differential image into 2 classes by taking the threshold pixel value as a boundary after finding out an optimal threshold value through a certain threshold selection method; the clustering analysis method is to obtain 2 clustering centers of unchanged classes and changed classes by applying a clustering algorithm to the difference graph, and then to divide the 2 classes by a neighbor method; the graph cut analysis method is another classification method of images, which essentially classifies labels of unchanged classes and changed classes into pixel points, and the method is characterized in that energy optimization is continuously carried out on given constraint functions, and when the energy reaches the minimum, the image pixels can correspond to the optimal labels; the level set analysis method converts the evolution problem of the two-dimensional closed curve into an implicit mode of the level set function curved surface evolution in the three-dimensional space to solve, namely, a three-dimensional level set function is constructed, and then a curve set formed by a solution with a value of zero is solved, so that an image segmentation result is obtained.
Disclosure of Invention
The invention aims to provide an SAR image unsupervised change detection method based on static wavelet transformation extraction, and the SAR image can be accurately separated by adopting the method.
The technical scheme adopted by the invention is that the SAR image unsupervised change detection method based on static wavelet transformation extraction specifically comprises the following steps:
step 1, preprocessing two SAR images;
step 2, generating a difference graph by adopting a logarithmic ratio operator based on the result obtained in the step 1;
and 3, analyzing the difference graph generated in the step 2.
The invention is also characterized in that:
the specific process of the step 1 is as follows: and filtering and smoothing the two SAR images by using a Lee filter.
The specific process of the step 2 is as follows:
let two SAR images be I respectively 1 、I 2 Image I 1 、I 2 The difference diagram D of (2) is:
D＝|log(I 1 )-log(I 2 )| (1)。
the specific process of the step 3 is as follows:
step 3.1, performing single-layer static wavelet transformation on the difference map D obtained in the step 2 by using db2 wavelet basis function;
step 3.2, classifying the images obtained in the step 3.1 by adopting an EM-GGM algorithm;
and 3.3, carrying out K-means clustering on the images classified in the step 3.2.
The specific process of the step 3.3 is as follows: and carrying out vhad feature selection on the image pixels in each second-order neighborhood, and finally clustering by using a K-means algorithm.
The beneficial effects of the invention are as follows: the invention firstly uses the Lee filter to carry out filtering treatment, eliminates noise and reduces the influence of noise points on the change detection result as much as possible. A difference map is then generated using the log ratio operator. Finally, the difference graph is decomposed through SWT2 (db 2), (two-dimensional static wavelet transformation, db2 is a wavelet basis function), then EM_GGM is used for decomposition, window probability is selected to form a feature vector, and a result is obtained through K-means. The detection method provided by the invention has the following advantages: (a) minimizing the effect of speckle noise in the SAR image; (B) When analyzing the difference map, SWT2 (db 2) algorithm is introduced, and low-pass filtering can be used in the frequency domain to reduce the influence of noise. The em_ggm algorithm is introduced, and GGM is a more robust and flexible model for describing the changed and unchanged pixels in the disparity map.
Drawings
FIG. 1 is a flow chart of an SAR image unsupervised change detection method based on static wavelet transform extraction of the present invention;
FIG. 2 is a change detection reference map of the Bern dataset;
FIG. 3 is a change detection diagram of Bern dataset obtained by adopting the SAR image unsupervised change detection method extracted based on static wavelet transformation;
FIG. 4 is a change detection reference diagram of the Ottawa dataset;
FIG. 5 is a change detection diagram of an Ottawa dataset obtained by adopting the SAR image unsupervised change detection method based on static wavelet transform extraction of the invention;
FIG. 6 is a change detection reference diagram of the Shihmen Reservoir dataset;
fig. 7 is a change detection diagram of Shihmen Reservoir data set obtained by adopting the method for detecting the unsupervised change of the SAR image extracted based on static wavelet transformation.
Detailed Description
The invention will be described in detail below with reference to the drawings and the detailed description.
The invention discloses an SAR image unsupervised change detection method based on static wavelet transform extraction, which comprises the steps of firstly decomposing a static wavelet transform (Stationary wavelet transform, SWT) of a difference image, wherein the static wavelet transform is a wavelet transform algorithm designed for overcoming the defect of translation invariance of discrete wavelet transform (discrete wavelet transform, DWT). The static wavelet transform differs from the discrete wavelet transform in part, principally by passing through each orderHigh pass filterAndlow pass filterThereafter, the filter is upsampled (Upsampling) instead of downsampling the discrete wavelet transform after it has passed through the filter.
And secondly, respectively carrying out expectation maximization algorithm (EM-GGM) decomposition of a generalized Gaussian model on the four groups of images obtained by SWT decomposition. The four groups of images are an approximate image, a horizontal detail image, a vertical detail image and a diagonal detail image respectively, the EM algorithm based on the generalized Gaussian model is utilized to carry out histogram fitting on the class condition distribution of the changed class and the unchanged class, initial parameters are given to generate a lookup table at the beginning, the step E is to calculate posterior estimation of each pixel belonging to two opposite classes by using Bayesian rules according to the parameters in the iteration t times, the step M is to provide new parameter estimation values of the generalized Gaussian model for the opposite classes, iteration is carried out until the difference value of the posterior probabilities of the last two times is smaller than a set minimum value, iteration is stopped, and otherwise the step E is returned.
And finally, classifying by a K-Means clustering algorithm, wherein the feature selection is vhad second-order neighborhood. v is the vertical window, h is the horizontal window, a is the full window, and d is the diagonal window. The K-Means clustering algorithm utilizes two points with the maximum inter-class distance and the minimum intra-class distance to find a proper clustering center through iteration. The squared euclidean distance is used as a basis for sample clustering.
As shown in fig. 1, the method specifically comprises the following steps:
and step 1, a pretreatment stage. Filtering and smoothing the two SAR images by using a Lee filter;
and 2, generating a difference map stage. Generating a difference map using a logarithmic ratio operator;
the specific process of the step 2 is as follows:
let two SAR images be I respectively 1 、I 2 Image I 1 、I 2 The difference diagram D of (2) is:
D＝|log(I 1 )-log(I 2 )| (1)。
and 3, analyzing a difference graph stage. The difference map is first decomposed using SWT2 (db 2), (two-dimensional static wavelet transform, db2 being the wavelet basis function) to obtain an approximate image, a horizontal detail image, a vertical detail image, and a diagonal detail image. And then the EM_GGM is used for respectively decomposing the four groups of images, finally, a second-order neighborhood window probability is selected to form a feature vector, and a result is obtained through K-means.
The specific process of the step 3 is as follows:
step 3.1, swt2 (two-dimensional static wavelet transform):
the image is subjected to single-layer static wavelet transformation by using db2 wavelet basis function. MATLAB provides a function swt of the two-dimensional static wavelet transform, because it does not downsample the decomposition coefficients, the single-layer decomposition is formally the same as the multi-layer decomposition results.
Step 3.2, classifying the images obtained in the step 3.1 by adopting an EM-GGM algorithm;
selecting GGM to conditional probability Density function P (x i |c)，P(x i Modeling of u), c is a label of a changed class, and u is a label of an unchanged class; the probability density function is:
wherein mu m 、σ m 、γ m The mean value, standard deviation and shape parameters of the two classes are respectively; for parameter setsθ＝{P(c)，P(u)，μ c ，μ u ，σ c ，σ u ，γ c ，γ u Reliable estimates of P (c), P (u) are prior probabilities that can be learned by EM algorithms; the algorithm is a general method for learning maximum likelihood estimates of parameters in incomplete data problems. It includes a desired step (E step) and a maximization step (M step) in the iteration and iterates until convergence is reached. The EM-GGM algorithm is described as follows:
starting: initializing parameters:
θ 0 ＝{P(c)，P(u)，μ c ，μ u ，σ c ，σ u ，γ c ，γ u } (3)
generating a parameter lookup table, and initializing a threshold T to be 100:
e, step E: from the parameters of the t-th iteration, the posterior estimate for each pixel belonging to two opposite classes is calculated using bayesian rules as:
P t (m|x i )＝P t (m)P t (x i |m)/P t (x i ) (3)；
P t (x i )＝P t (c)P t (x i |c)+P t (u)P t (x i |u) (4)；
m step: updating model parameters:
where |x| represents the number of pixels, here n, representing n pixels. P (x) i ) For pixel x at position i in the differential image i Probability density function of (a). X represents the set of all pixels.
And (3) convergence: setting the iteration number 200, and stopping if the iteration number is satisfied. Otherwise, continuing to execute the step E.
And 3.3, carrying out K-means clustering on the images classified in the step 3.2.
And (3) carrying out vhad feature selection on pixels in each second-order neighborhood, and finally clustering by using a K-means algorithm.
Examples
The effect of the invention can be specifically illustrated by simulation experiments:
1. experimental conditions
The CPU of the microcomputer used in the experiment is Intel Pentium43.0GHz memory 1GB, and the programming platform is Matlab 7.0.1. The SAR images used for the experiments were the Bern dataset, ottawa dataset, shihmen Reservoir dataset.
2. Experimental details
First is the pretreatment phase. Filtering and smoothing the two SAR images by using a Lee filter; the next is the phase of generating the disparity map. Generating a difference map using a logarithmic ratio operator; finally, the analysis of the difference map stage. The difference map is first decomposed using SWT2 (db 2), (two-dimensional static wavelet transform, db2 being the wavelet basis function) to obtain an approximate image, a horizontal detail image, a vertical detail image, and a diagonal detail image. And then the EM_GGM is used for respectively decomposing the four groups of images, finally, a second-order neighborhood window probability is selected to form a feature vector, and a result is obtained through K-means.
3. Experimental results
Table 1 shows the change detection evaluation index of the method proposed by the present invention and the other three methods on three data sets;
TABLE 1
FIG. 2 is a change detection reference map of the Bern dataset; FIG. 3 is a graph of the change detection of Bern dataset obtained using the method of the present invention; FIG. 4 is a change detection reference diagram of the Ottawa dataset; FIG. 5 is a graph of the change detection of an Ottawa dataset using the method of the present invention; FIG. 6 is a change detection reference diagram of the Shihmen Reservoir dataset; FIG. 7 is a graph of the change detection of Shihmen Reservoir dataset obtained using the method of the present invention; experimental results show that the method provided by the invention can obtain a good change detection result.
Claims (2)
1. The SAR image unsupervised change detection method based on static wavelet transformation extraction is characterized by comprising the following steps of: the method specifically comprises the following steps:
step 1, preprocessing two SAR images;
the specific process of the step 1 is as follows: filtering and smoothing the two SAR images by using a Lee filter;
step 2, generating a difference graph by adopting a logarithmic ratio operator based on the result obtained in the step 1;
the specific process of the step 2 is as follows:
let two SAR images be I respectively 1 、I 2 Image I 1 、I 2 The difference diagram D of (2) is:
D＝|log(I 1 )-log(I 2 )| (1)；
step 3, analyzing the difference graph generated in the step 2;
the specific process of the step 3 is as follows:
step 3.1, performing single-layer static wavelet transformation on the difference map D obtained in the step 2 by using db2 wavelet basis function;
step 3.2, classifying the images obtained in the step 3.1 by adopting an EM-GGM algorithm;
and 3.3, carrying out K-means clustering on the images classified in the step 3.2.
2. The method for detecting the unsupervised change of the SAR image based on the static wavelet transform extraction as set forth in claim 1, wherein: the specific process of the step 3.3 is as follows: and carrying out vhad feature selection on the image pixels in each second-order neighborhood, and finally clustering by using a K-means algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111227373.9A CN114066816B (en) | 2021-10-21 | 2021-10-21 | SAR image unsupervised change detection method based on static wavelet transformation extraction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111227373.9A CN114066816B (en) | 2021-10-21 | 2021-10-21 | SAR image unsupervised change detection method based on static wavelet transformation extraction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114066816A CN114066816A (en) | 2022-02-18 |
| CN114066816B true CN114066816B (en) | 2023-08-15 |
Family
ID=80235107
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111227373.9A Active CN114066816B (en) | 2021-10-21 | 2021-10-21 | SAR image unsupervised change detection method based on static wavelet transformation extraction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114066816B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114814993B (en) * | 2022-03-25 | 2023-04-18 | 河海大学 | Microwave attenuation snowfall intensity monitoring method based on DCGAN and 2D-CNN |
| CN117173587B (en) * | 2023-08-23 | 2024-07-12 | 哈尔滨工程大学 | Feature refinement fusion change detection method based on heterogeneous image depth conversion |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1148332A2 (en) * | 2000-04-18 | 2001-10-24 | The University of Hong Kong | Method of and device for inspecting images to detect defects |
| CN103955943A (en) * | 2014-05-21 | 2014-07-30 | 西安电子科技大学 | Non-supervision change detection method based on fuse change detection operators and dimension driving |
| CN110210407A (en) * | 2019-06-04 | 2019-09-06 | 武汉科技大学 | A kind of Misty Image well-marked target detection method |
| CN111681197A (en) * | 2020-06-12 | 2020-09-18 | 陕西科技大学 | Remote sensing image unsupervised change detection method based on Siamese network structure |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7039239B2 (en) * | 2002-02-07 | 2006-05-02 | Eastman Kodak Company | Method for image region classification using unsupervised and supervised learning |
-
2021
- 2021-10-21 CN CN202111227373.9A patent/CN114066816B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1148332A2 (en) * | 2000-04-18 | 2001-10-24 | The University of Hong Kong | Method of and device for inspecting images to detect defects |
| CN103955943A (en) * | 2014-05-21 | 2014-07-30 | 西安电子科技大学 | Non-supervision change detection method based on fuse change detection operators and dimension driving |
| CN110210407A (en) * | 2019-06-04 | 2019-09-06 | 武汉科技大学 | A kind of Misty Image well-marked target detection method |
| CN111681197A (en) * | 2020-06-12 | 2020-09-18 | 陕西科技大学 | Remote sensing image unsupervised change detection method based on Siamese network structure |
Non-Patent Citations (1)
| Title |
|---|
| 融合多尺度深度特征的SAR图像变化检测;刘梦岚;杨学志;贾璐;汪骏;;合肥工业大学学报(自然科学版)(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114066816A (en) | 2022-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tirandaz et al. | A two-phase algorithm based on kurtosis curvelet energy and unsupervised spectral regression for segmentation of SAR images | |
| Venugopal | Automatic semantic segmentation with DeepLab dilated learning network for change detection in remote sensing images | |
| Celik | A Bayesian approach to unsupervised multiscale change detection in synthetic aperture radar images | |
| Kayabol et al. | Unsupervised amplitude and texture classification of SAR images with multinomial latent model | |
| Khmag et al. | Natural image noise removal using nonlocal means and hidden Markov models in transform domain | |
| Helmy et al. | Image segmentation scheme based on SOM–PCNN in frequency domain | |
| CN114066816B (en) | SAR image unsupervised change detection method based on static wavelet transformation extraction | |
| CN111046772A (en) | Multi-temporal satellite remote sensing island shore line and development and utilization information extraction method | |
| CN104732244A (en) | Wavelet transform, multi-strategy PSO (particle swarm optimization) and SVM (support vector machine) integrated based remote sensing image classification method | |
| Hidane et al. | Nonlinear multilayered representation of graph-signals | |
| Gu et al. | A two-component deep learning network for SAR image denoising | |
| Molini et al. | Towards deep unsupervised SAR despeckling with blind-spot convolutional neural networks | |
| Li et al. | Gamma correction-based automatic unsupervised change detection in SAR images via FLICM model | |
| Wang et al. | Color image segmentation using PDTDFB domain hidden Markov tree model | |
| Annavarapu et al. | An adaptive watershed segmentation based medical image denoising using deep convolutional neural networks | |
| Xiaolan et al. | Texture Feature Extraction Method Combining Nonsubsampled Contour Transformation with Gray Level Co-occurrence Matrix. | |
| Katartzis et al. | A hierarchical Markovian model for multiscale region-based classification of vector-valued images | |
| Lu et al. | SAR image despeckling via structural sparse representation | |
| Wu et al. | Object-Enhanced YOLO Networks for Synthetic Aperture Radar Ship Detection | |
| CN114240990A (en) | SAR image point target segmentation method | |
| CN114066815B (en) | SAR image change detection method based on fuzzy c-means clustering | |
| CN115410093B (en) | Remote sensing image classification method based on dual-channel coding network and conditional random field | |
| An et al. | Unsupervised segmentation of hidden Markov fields corrupted by correlated non-Gaussian noise | |
| CN113486967B (en) | SAR image classification algorithm combining graph convolution network and Markov random field | |
| Khmag et al. | Natural image noise removal using non local means and hidden Markov models in stationary wavelet transform domain |
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 |