CN106611423A - SAR image segmentation method based on ridge wave filter and deconvolution structural model - Google Patents
SAR image segmentation method based on ridge wave filter and deconvolution structural model Download PDFInfo
- Publication number
- CN106611423A CN106611423A CN201611269937.4A CN201611269937A CN106611423A CN 106611423 A CN106611423 A CN 106611423A CN 201611269937 A CN201611269937 A CN 201611269937A CN 106611423 A CN106611423 A CN 106611423A
- Authority
- CN
- China
- Prior art keywords
- pixel
- sketch
- sar image
- line
- wave filter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
- G06T2207/20024—Filtering details
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20084—Artificial neural networks [ANN]
Abstract
The invention discloses an SAR image segmentation method based on a ridge wave filter and a deconvolution structural model, mainly used for solving the problem that SAR image segmentation in the prior art is inaccurate. The SAR image segmentation method comprises the following implementation steps of: (1), sketching an SAR image so as to obtain a sketch image; (2), according to an area chart of the SAR image, dividing a pixel subspace of the SAR image; (3), constructing a ridge wave filter set; (4), constructing a deconvolution structural model; (5), segmenting a hybrid aggregation structured surface feature pixel subspace by adopting the SAR image segmentation method based on the ridge wave filter and the deconvolution structural model; (6), performing independent target segmentation based on the sketch line aggregation feature; (7), performing line target segmentation based on a visual semantic rule; (8), performing segmentation of a pixel subspace in a homogeneous area by adopting a polynomial-based logistic regression prior model; and (9), combining segmentation results so as to obtain an SAR image segmentation result. By means of the SAR image segmentation method disclosed by the invention, the good segmentation effect of the SAR image is obtained; and the SAR image segmentation method can be used for semantic segmentation of the SAR image.
Description
Technical field
The invention belongs to technical field of image processing, further relates to the one kind in target identification technology field based on ridge
Synthetic aperture radar SAR (Synthetic Aperture Radar) the image segmentation side of wave filter and deconvolution structural model
Method.The region that the present invention can have different characteristic to synthetic aperture radar SAR image is split exactly, and can be used for
The object detection and recognition of follow-up synthetic aperture radar SAR image.
Background technology
Synthetic aperture radar SAR is the impressive progress in remote sensing technology field, for obtaining the full resolution pricture of earth surface.
Compared with other kinds of imaging technique, SAR has very important advantage, and it is not by air such as cloud layer, rainfall or dense fogs
The impact of condition and intensity of illumination, can round-the-clock, round-the-clock obtain high resolution remote sensing data.SAR technologies for it is military,
Many fields such as agricultural, geography have great importance.Image segmentation is referred to will be schemed according to color, gray scale and Texture eigenvalue
Process as being divided into several mutually disjoint regions.It is for facing at present SAR image to be interpreted by computer
Individual huge challenge, and SAR image segmentation is its steps necessary, it affects very big to further detection, identification.
The conventional method of image segmentation is broadly divided into the method and the method based on statistical model of feature based at present.Due to
SAR unique imaging mechanism, contains many coherent speckle noises, causes the traditional method of many optical imagerys not in SAR image
The segmentation of SAR image can be directly used in.The conventional segmentation methods of SAR image mainly have supervision and semi-supervised method.They
Generally requiring manually experience carries out feature extraction, but the quality of the feature extracted has for the segmentation result of SAR image
Pivotal role.Supervise and semi-supervised method for having, it is desirable to have label data, the label data of SAR image seldom, obtains mark
The cost for signing data is very high.Key technology of the machine learning as unsupervised feature learning, can be used for SAR image segmentation and appoints
Business.However, traditional deep learning method can only often reach feature aspect, SAR image is not preferably excavated in semantic layer
Information on secondary, causes which efficiently accomplish the segmentation to SAR image.
Patent " a kind of specific objective contour images based on depth convolutional neural networks that North China Electric Power University applies at which
Disclose in dividing method " (number of patent application CN201610109536.6, publication number CN105787482A) a kind of based on depth
The image partition method of the specific objective profile of convolutional neural networks.It is big that training image is normalized to same pixel by the method
It is little, the training image for obtaining is input in a convolutional neural networks, through several layers of convolutional layer and full articulamentum, is being connected entirely
Last layer of layer obtains image expression, and is compared with corresponding mark image and obtains forecast error.Using back propagation
Algorithm and stochastic gradient descent method obtain the segmentation of specific objective contour images reducing forecast error to train the neutral net
Training pattern.Although this method has reached the purpose of autonomic learning characteristics of image, but, the deficiency that the method is yet suffered from it
Place is that input picture has been carried out normalized for the convenience on processing by the method, so just destroys the original of image
Structural information.Meanwhile, the method is also marked to image, is divided into training sample and test sample, to reach training volume
The purpose of product network.The processing mode for so having supervision substantially increases the complexity of dividing method.
The patent that Xian Electronics Science and Technology University applies at which is " based on ridge ripple deconvolution network and the SAR image of sparse classification
Disclose in dividing method " (number of patent application CN201510675676.5, publication number CN105374033A) a kind of based on ridge ripple
Deconvolution network and the SAR image segmentation method of sparse classification.Aggregation zone and homogenous region difference of the method to SAR image
Training ridge ripple deconvolution network RDN, obtains the optimal value of ridge ripple deconvolution network median filter group, and using the side of sparse classification
Method completes the segmentation of SAR image.Although the method has reached unsupervised study characteristics of image, but, the method is yet suffered from
Weak point be, when wave filter is initialized, the method for the random initializtion ridge ripple wave filter of employing, and have ignored image
Structural information, so just greatly reduces the accuracy of image segmentation.
The paper that Liu Fang, Duan Yiping, Li Lingling, burnt Li Cheng etc. are delivered at which is " semantic adjacent with self adaptation based on level vision
The SAR image segmentation of the hidden model of domain multinomial " (IEEE Trancactions on Geoscience and Remote
Sensing, 2016,54 (7):Propose in 4287-4301.) a kind of based on level vision semanteme and adaptive neighborhood multinomial
The SAR image segmentation method of hidden model, the method is on the basis of SAR image sketch map, it is proposed that the level vision of SAR image
It is semantic.The level vision semanteme is divided into aggregation zone, structural region and homogenous region SAR image.Based on the division, to not
Different dividing methods are employed with the region of characteristic.For aggregation zone, gray level co-occurrence matrixes feature is extracted, and using local
The method of linear restriction coding obtains the expression of each aggregation zone, and then the method using hierarchical clustering is split.To knot
Structure region, by analyzing side model and line model, devises vision semantic rule positioning border and line target.In addition, border and
Line target contains strong directional information, therefore the hidden model of multinomial devised based on geometry window is split.It is right
Homogenous region, goes to represent center pixel in order to be able to find appropriate neighborhood, devises the hidden mould of multinomial based on self-adapting window
Type is split.The segmentation result in these three regions is integrated into and obtains last segmentation result together.The deficiency of the method it
Place is that the boundary alignment to aggregation zone is not accurate enough, and the determination to homogenous region classification number is not reasonable, the area of segmentation result
Domain concordance is poor, and pinpoint target is not processed in the segmentation of structural region.
The content of the invention
Present invention aims to the deficiency of above-mentioned prior art, proposes that one kind is based on ridge ripple wave filter and deconvolution
The SAR image segmentation method of structural model, with the significantly more efficient segmentation for completing SAR image.
For achieving the above object, technical scheme is as follows:
(1) sketch SAR image:
(1a) to the synthetic aperture radar SAR image being input into, according to the pixel fluctuating characteristic distributions of SAR image, obtain which
Sketch model;
(1b) from sketch model extraction sketch map, obtain the sketch map of synthetic aperture radar SAR image;
(2) divide pixel subspace:
(2a) using sketch line fields method, pixel subspace is divided to synthetic aperture radar SAR image, is synthesized
The administrative division map of aperture radar SAR image;
(2b) administrative division map is mapped in the synthetic aperture radar SAR image of input, obtains synthetic aperture radar SAR image
Middle mixing aggregated structure atural object pixel subspace, homogenous region pixel subspace and structure-pixel subspace;
(3) build ridge ripple filter set:
(3a) extract from the administrative division map of synthetic aperture radar SAR image and mix aggregated structure atural object pixel subspace correspondence
Aggregation zone, it is interval at [0 °, 180 °] in be divided into i.e. 18,18 intervals direction at intervals of 10 °, count each respectively
The line segment bar number of sketch line segment in interval interior aggregation zone;
(3b) to all of sketch line segment in the aggregation zone, according to each interval in line segment bar number number arranged
Sequence, obtains the collating sequence in direction, using the number of degrees in front 6 directions in the collating sequence in direction as ridge ripple wave filter in
Directioin parameter;
(3c) according to the following formula, according to parameter a, θ and b calculates the ridge ripple function of 9 × 9 ridge ripple wave filter:
Y=a × (y1×cosθ+y2×sinθ-b)
Wherein, Y represents the ridge ripple function of ridge ripple wave filter, and a represents the scale parameter of ridge ripple wave filter, the span of a
For [0,3], the discrete interval of a is 0.2, y1Represent the abscissa positions of ridge ripple wave filter pixel, y1Span for [0,
8], y1Discrete interval represent that cosine is operated for 1, cos, θ represents the directioin parameter of ridge ripple wave filter, y2Represent ridge ripple wave filter
The vertical coordinate position of pixel, y2Span be [0,8], y2Discrete interval represent sinusoidal operation for 1, sin;B represents ridge
The displacement parameter of wave filter, when directioin parameter θ for [0 °, 90 °) when, b is in [0,9 × (sin θ+cos θ)] is interval being spaced
Discretization is carried out for 0.2, when directioin parameter θ for [90 °, 180 °) when, b in [9 × cos θ, 9 × sin θ] is interval with intervals of
0.2 carries out discretization;
(3d) according to the following formula, calculate each ridge ripple wave filter:
Wherein, c (Y) represents the ridge ripple wave filter using ridge ripple function Y as parameter, and K represents ridge ripple wave filter frobenius
The inverse of norm, exp represent the index operation with natural constant e as bottom;
(3e) calculated each ridge ripple wave filter group is synthesized into ridge ripple filter set;
(4) construct deconvolution structural model:
(4a) the mixing aggregated structure atural object pixel subspace in synthetic aperture radar SAR image each is not mutually connected
Region, by 31 × 31 window carry out every a sliding window sample, obtain the corresponding multiple images block in each region, by multiple images
Block is sequentially inputted in deconvolution structural model, obtains the input layer of deconvolution structural model;
(4b) carry out convolution operation to reconstruct the image block in input layer using characteristic pattern and ridge ripple wave filter, obtain warp
The warp lamination of product structural model;
(4c) according to the following formula, calculate data fidelity item:
Wherein, E (c) represents data fidelity item, and c represents the ridge ripple wave filter in deconvolution structural model warp lamination, N tables
Show the total number of the image block that each mutual disconnected region to be learnt includes, ∑ represents sum operation, | | | |FExpression is done
Frobenius norms are operated,Represent the square operation of frobenius norms, xiIn representing deconvolution structural model to be constructed
I-th input picture block, MiThe corresponding ridge ripple wave filter of i-th image block being input in representing deconvolution structural model to be constructed
Sum, * represents convolution operation,Represent in deconvolution structural model to be constructed corresponding j-th feature of i-th image block
Figure,Represent in deconvolution structural model to be constructed the corresponding j-th ridge ripple wave filter of i-th image block;
(4d) according to the following formula, computation structure fidelity item:
Wherein, G (c) represents structure fidelity item, and R () represents the operation for seeking all sketch line total lengths in sketch map, SM
() represents the operation extracted with the one-to-one sketch segment of input picture block;
(4e) according to the following formula, calculating target function:
Wherein, L (c) represents object function,Represent when object function L (c) value is minimum, ask for ridge ripple filtering
The operation of device c;
(4f) the ridge ripple filter set obtained by object function guidance learning is exported, the defeated of deconvolution structural model is obtained
Go out layer;
(5) train deconvolution structural model:
(5a) structural failure threshold value is set to into 0.1;
(5b) step (4a) sampling is obtained image block to be sequentially inputted in deconvolution structural model;
(5c) from ridge ripple filter set, six wave filter are randomly selected, its directioin parameter is by statistics in step (3b)
6 directions obtain, its displacement parameter and scale parameter random initializtion, by these initial six wave filter groups into wave filter
Set is used as selected ridge ripple filter set;
(5d) characteristic pattern that 6 sizes for 39 × 39 is initialized for 39 × 39 null matrix with 6 sizes, after initializing
6 sizes for 39 × 39 characteristic pattern as feature set of graphs;
(5e) feature set of graphs and selected ridge ripple filter set are carried out convolution operation to reconstruct input picture block;
(5f) using the structure fidelity item formula in step (4d), calculate the structure fidelity item of reconstruct input picture block;
(5g) judge whether the structure fidelity item of current reconstruct input picture block is less than structural failure threshold value, if so, then hold
Row step (5j), otherwise, execution step (5h);
(5h) using scale parameter more new formula and displacement parameter more new formula, step (4c) data fidelity item is updated respectively
The scale parameter and displacement parameter of formula median ridge wave filter, the ridge ripple filter set after being updated, using characteristic pattern more
New formula, updates characteristic pattern, the feature set of graphs after being updated;
(5i) using the ridge ripple filter set after renewal as selected ridge ripple filter set, by the feature after renewal
Set of graphs re-starts study to input picture block as feature set of graphs, return to step (5e);
(5j) the ridge ripple wave filter that study is obtained is preserved to the reconstruct input picture block ridge ripple wave filter collection for succeeding in school
In conjunction, the study to the input picture block feature is completed, and exports the ridge ripple filter set that the input picture block succeeds in school;
(5k) judge whether all image blocks complete the study of feature by deconvolution structural model, if so, terminate journey
Sequence, otherwise, is input into next image block execution step (5c);
(6) split SAR image mixing aggregated structure atural object pixel subspace:
(6a) by the ridge ripple wave filter collection of all mutually disconnected mixing aggregated structure atural object pixel subspace regional trainings
It is merged and is connected into code book;
(6b) by the mutually ridge ripple filter set of disconnected mixing aggregated structure atural object pixel subspace regional training
All of ridge ripple wave filter, is projected to code book, is obtained projection vector;
(6c) maximum is carried out to the projection vector in each mutual disconnected mixing aggregated structure atural object pixel subspace region
Chi Hua, obtains a structural eigenvector;
(6d) AP clustering algorithms are propagated using neighbour, structural eigenvector is clustered, is obtained and structural eigenvector
The segmentation result of corresponding mixing aggregated structure atural object pixel subspace;
(7) segmenting structure pixel subspace:
(7a) vision semantic rule is used, splits line target;
(7b) feature of gathering based on sketch line, splits pinpoint target;
(7c) result of line target and pinpoint target segmentation is merged, obtains the segmentation knot of structure-pixel subspace
Really.
(8) split homogenous region pixel subspace:
Using the homogenous region dividing method based on multinomial logistic regression prior model, to homogenous region pixel subspace
Split, obtained the segmentation result of homogenous region pixel subspace.
(9) combination and segmentation result:
By the segmentation result of mixing aggregated structure pixel subspace, homogenous region pixel subspace and structure-pixel subspace
Merge, obtain the final segmentation result of synthetic aperture radar SAR image.
The present invention is had the advantage that compared with prior art:
First, the present invention to the mixing aggregated structure atural object pixel subspace in synthetic aperture radar SAR image each
Mutually disconnected region, is carried out sampling every a sliding window by 31 × 31 window, obtains multiple images block, will be the image block for obtaining defeated
Enter in deconvolution structural model, it is not necessary to which the synthetic aperture radar SAR image to being input into is normalized, overcome existing
Have technology need to be normalized synthetic aperture radar SAR image before input and deficiency so that the present invention has
The image block that directly can be obtained by the use of after sampling as input, without destroy input picture block prototype structure information it is excellent
Point.
Second, the present invention utilizes synthetic aperture radar SAR image sketch map, using sketch line fields method, is closed
Into the administrative division map of aperture radar SAR image, administrative division map is mapped to into the synthetic aperture radar SAR image of input, obtains synthesizing hole
Mixing aggregated structure atural object pixel subspace, homogeneous texture pixel subspace and structure-pixel in the radar SAR image of footpath is empty
Between, sampled and feature learning in the pixel subspace of mixing aggregated structure atural object, it is not necessary to which the type of image block is entered
Rower shows, overcomes prior art and input picture is denoted as training image when characteristics of image is learnt and image is marked not
Foot so that the present invention has and need not indicate to input picture, and reduces the advantage of deconvolution structural model complexity.
3rd, in the deconvolution structural model of present invention construction, using the energy error function of input picture block and defeated
Enter the structural failure function of image block to extract the structural information of input picture block, overcome prior art and adopt random initializtion
Ridge ripple wave filter and have ignored the deficiency of the structural information of synthetic aperture radar SAR image so that the present invention have can be more preferable
Utilization SAR image structural information completing the segmentation of synthetic aperture radar SAR image, improve synthetic aperture radar SAR
Image segmentation accuracy.
Description of the drawings
Fig. 1 is the flow chart of the present invention;
Fig. 2 is the analogous diagram of the present invention;
Fig. 3 is extracted in the present invention with the sketch line chart for gathering feature;
Fig. 4 is simulation result schematic diagram of the present invention.
Specific embodiment
The present invention will be further described below in conjunction with the accompanying drawings.
Refer to the attached drawing 1, the present invention's are comprised the following steps that.
Step 1, SAR image sketch.
Input synthetic aperture radar SAR image, by its sketch, obtains the sketch map of synthetic aperture radar SAR image.
1st step, constructs the side being made up of pixel with different directions and yardstick, a template of line, using template
Direction and dimensional information structural anisotropy Gaussian function calculating the weight coefficient of every bit in the template, its mesoscale
Number value is 3~5, and direction number value is 18.
2nd step, according to the following formula, pixel in the calculating synthetic aperture radar SAR image corresponding with template area position
Average and variance yields:
Wherein, μ represents the average of pixel in the synthetic aperture radar SAR image corresponding with template area position, ∑ table
Show sum operation, g represents the position of a pixel in the Ω region of template, and ∈ is represented and belonged to symbol, wgRepresent template the
Weight coefficient of the pixel at g positions, w in Ω regiongSpan be wg∈ [0,1], AgRepresent and template Ω
Pixel value of the pixel at g positions in corresponding synthetic aperture radar SAR image in region, ν are represented and template area position
The variance yields of pixel in corresponding synthetic aperture radar SAR image.
3rd step, according to the following formula, calculates the response value of each pixel comparison value operator in synthetic aperture radar SAR image:
Wherein, R represents the response value of each pixel comparison value operator in synthetic aperture radar SAR image, and min { } is represented
Minimize operation, a and b represents two zoness of different in template, μ respectivelyaAnd μbRepresent respectively and template area a and template
The average of pixel in the corresponding synthetic aperture radar SAR image in region b positions.
4th step, according to the following formula, calculates response value of each pixel to dependency operator in synthetic aperture radar SAR image:
Wherein, C represents response value of each pixel to dependency operator in synthetic aperture radar SAR image, a and b difference
Represent two zoness of different in template, νaAnd νbSynthetic aperture corresponding with template area a and template area b positions is represented respectively
The variance yields of pixel, μ in radar SAR imageaAnd μbSynthesis hole corresponding with template area a and template area b positions is represented respectively
The average of pixel in the radar SAR image of footpath,Represent square root functions.
5th step, according to the following formula, merges response value and the synthesis of pixel comparison value operator in synthetic aperture radar SAR image
Response value of the pixel to dependency operator in aperture radar SAR image, calculates each pixel pair in synthetic aperture radar SAR image
The response value of each template:
Wherein, F represents response value of each pixel to each template in synthetic aperture radar SAR image, R and C difference tables
Show in synthetic aperture radar SAR image that pixel is to dependency operator in pixel comparison value operator and synthetic aperture radar SAR image
Response value,Represent square root functions.
6th step, selects the template with maximum response as the template of pixel in synthetic aperture radar SAR image, and
Using maximum response as the intensity of the pixel, using the direction of the template with maximum response as the direction of the pixel, obtain
Obtain the sideline response diagram and directional diagram of synthetic aperture radar SAR image.
7th step, using the selected template of each pixel in synthetic aperture radar SAR image, obtains synthetic aperture radar
The gradient map of SAR image.
8th step, according to the following formula, the sideline response diagram that will normalize to [0,1] is carried out with the gradient map for normalizing to [0,1]
Fusion, obtains intensity map:
Wherein, I represents the intensity level in intensity map, and x represents the value in the response diagram of sideline, and y represents the value in gradient map.
9th step, using non-maxima suppression method, detects to intensity map, obtains suggestion sketch.
10th step, the pixel in choosing suggestion sketch with maximum intensity, by the picture in suggestion sketch with the maximum intensity
The pixel of element connection connects to form suggestion line segment, obtains suggestion sketch map.
11st step, according to the following formula, calculates the code length gain CLG of sketch line in suggestion sketch map:
Wherein, CLG represents the code length gain of sketch line in suggestion sketch map, and ∑ represents sum operation, and m represents current
The number of pixel in sketch line neighborhood, t represent the numbering of pixel in current sketch line neighborhood, AtIn representing current sketch line neighborhood
The observation of t-th pixel, At,0Expression can not be represented under the hypothesis of structural information in current sketch line, in the sketch line neighborhood
The estimated value of t-th pixel, ln () represent the log operations with e as bottom, At,1Expression can represent structure in current sketch line
Under the hypothesis of information, the estimated value of t-th pixel in the sketch line neighborhood.
12nd step, given threshold T, the span of T is 5~50, selects CLG>The suggestion sketch line of T is used as final sketch
Sketch line in figure, obtains the corresponding sketch map of input synthetic aperture radar SAR image.
The synthetic aperture radar SAR image sketch model that the present invention is used is that Jie-Wu et al. was published in IEEE in 2014
Article on Transactions on Geoscience and Remote Sensing magazines《Local maximal
homogenous region search for SAR speckle reduction with sketch‐based
geometrical kernel function》Proposed in model.
Step 2, divides pixel subspace.
1st step, according to the concentration class of sketch line segment in the sketch map of synthetic aperture radar SAR image, sketch line is divided
To represent the aggregation sketch line of aggregation atural object and representing the sketch line of border, line target and isolated target.
2nd step, according to the statistics with histogram of sketch line segment concentration class, chooses the sketch line that concentration class is equal to optimum concentration class
Duan Zuowei seed line-segment sets { Ek, k=1,2 ..., m }, wherein, EkAny bar sketch line segment in expression seed line-segment sets, k tables
Show the label of any bar sketch line segment in seed line-segment sets, m represents the total number of seed line segment, and { } represents set operation.
3rd step, using the unselected line segment for being added to certain seed line-segment sets sum as basic point, with this basic point recursive resolve
New line segment aggregate.
4th step, one radius of construction are the circular primitive in the optimum concentration class interval upper bound, with the circular primitive to line-segment sets
Line segment in conjunction is expanded, and the line segment aggregate ecto-entad after expansion is corroded, and is obtained with sketch point in sketch map
For the aggregation zone of unit.
5th step, the sketch line to representing border, line target and isolated target, with each sketch point of each sketch line
Centered on construct size for 5 × 5 geometry window, obtain structural region.
6th step, will remove part beyond aggregation zone and structural region as can not sketch region in sketch map.
7th step, by the aggregation zone in sketch map, structural region and can not sketch region, be respectively mapped to synthetic aperture
In radar SAR image, mixing aggregated structure atural object pixel subspace, structure-pixel of synthetic aperture radar SAR image are obtained
Space and homogeneous texture pixel subspace.
Step 3, builds ridge ripple filter set.
1st step, extracts mixing aggregated structure atural object pixel subspace pair from the administrative division map of synthetic aperture radar SAR image
The aggregation zone answered, it is interval interior to be divided into i.e. 18,18 intervals direction at intervals of 10 ° at [0 °, 180 °], count respectively every
The line segment bar number of sketch line segment in individual interval interior aggregation zone.
2nd step, to all of sketch line segment in the aggregation zone, according to the line segment bar number in each interval number carry out
Sequence, obtain the collating sequence in direction, using the number of degrees in front 6 directions in the collating sequence in direction as ridge ripple wave filter in
Directioin parameter.
3rd step, according to the following formula, according to parameter a, θ and b calculates the ridge ripple function of 9 × 9 ridge ripple wave filter:
Y=a × (y1×cosθ+y2×sinθ-b)
Wherein, Y represents the ridge ripple function of ridge ripple wave filter, and a represents the scale parameter of ridge ripple wave filter, the span of a
For [0,3], the discrete interval of a is 0.2, y1Represent the abscissa positions of ridge ripple wave filter pixel, y1Span for [0,
8], y1Discrete interval represent that cosine is operated for 1, cos, θ represents the directioin parameter of ridge ripple wave filter, y2Represent ridge ripple wave filter
The vertical coordinate position of pixel, y2Span be [0,8], y2Discrete interval represent sinusoidal operation for 1, sin;B represents ridge
The displacement parameter of wave filter, when directioin parameter θ for [0 °, 90 °) when, b is in [0,9 × (sin θ+cos θ)] is interval being spaced
Discretization is carried out for 0.2, when directioin parameter θ for [90 °, 180 °) when, b in [9 × cos θ, 9 × sin θ] is interval with intervals of
0.2 carries out discretization.
4th step, according to the following formula, calculates each ridge ripple wave filter:
Wherein, c (Y) represents the ridge ripple wave filter using ridge ripple function Y as parameter, and K represents ridge ripple wave filter frobenius
The inverse of norm, exp represent the index operation with natural constant e as bottom.
Calculated each ridge ripple wave filter group is synthesized ridge ripple filter set by the 5th step.
Step 4, constructs deconvolution structural model.
1st step, does not mutually connect to the mixing aggregated structure atural object pixel subspace in synthetic aperture radar SAR image each
Logical region, is carried out sampling every a sliding window by 31 × 31 window, obtains the corresponding multiple images block in each region, by multiple figures
As block is sequentially inputted to the input layer of deconvolution structural model is obtained in deconvolution structural model.
2nd step, carries out convolution operation to reconstruct the image block in input layer using characteristic pattern and ridge ripple wave filter, obtains anti-
The warp lamination of convolutional coding structure model.
3rd step, according to the following formula, calculates data fidelity item:
Wherein, E (c) represents data fidelity item, and c represents the ridge ripple wave filter in deconvolution structural model warp lamination, N tables
Show the total number of the image block that each mutual disconnected region to be learnt includes, ∑ represents sum operation, | | | |FExpression is done
Frobenius norms are operated,Represent the square operation of frobenius norms, xiIn representing deconvolution structural model to be constructed
I-th input picture block, MiThe corresponding ridge ripple wave filter of i-th image block being input in representing deconvolution structural model to be constructed
Sum, * represents convolution operation,Represent in deconvolution structural model to be constructed corresponding j-th feature of i-th image block
Figure,Represent in deconvolution structural model to be constructed the corresponding j-th ridge ripple wave filter of i-th image block.
4th step, according to the following formula, computation structure fidelity item:
Wherein, G (c) represents structure fidelity item, and R () represents the operation for seeking all sketch line total lengths in sketch map, SM
() represents the operation extracted with the one-to-one sketch segment of input picture block.
5th step, according to the following formula, calculating target function:
Wherein, L (c) represents object function,Represent when object function L (c) value is minimum, ask for ridge ripple filtering
The operation of device c.
6th step, exports the ridge ripple wave filter obtained by object function guidance learning, obtains the output of deconvolution structural model
Layer.
Step 5, trains deconvolution structural model.
Structural failure threshold value is set to 0.1 by the 1st step.
Step 4 sampling is obtained image block and is sequentially inputted in deconvolution structural model by the 2nd step.
3rd step, from ridge ripple filter set, randomly select six wave filter, and its directioin parameter is by counting in step 3
6 directions obtain, its displacement parameter and scale parameter random initializtion, by these initial six wave filter groups into wave filter collection
Cooperate as selected ridge ripple filter set.
4th step, initializes characteristic pattern that 6 sizes for 39 × 39 for 39 × 39 null matrix with 6 sizes, will initialization
6 sizes for 39 × 39 characteristic pattern as feature set of graphs;
5th step, feature set of graphs and ridge ripple filter set are carried out convolution operation to reconstruct input picture block;
6th step, using structure fidelity item formula, calculates the structure fidelity item of reconstruct input picture block.
7th step, judges whether the structure fidelity item of current reconstruct input picture block is less than structural failure threshold value, if so, then
The 10th step is performed, otherwise, the 8th step is performed.
8th step, using scale parameter more new formula and displacement parameter more new formula, is updated the data in fidelity item formula respectively
The scale parameter and displacement parameter of ridge ripple wave filter, the ridge ripple filter set after being updated, using characteristic pattern more new formula,
Update characteristic pattern, the feature set of graphs after being updated.
The more new formula of scale parameter is as follows:
Wherein a represents the scale parameter of ridge ripple wave filter, atFor the yardstick that t steps are tried to achieve, at-1Try to achieve for t-1 steps
Yardstick, δ represent coefficient, and span is [0,1], and ∑ represents addition, xiFor i-th piece 31 × 31 of SAR image sampling block,Table
Show the corresponding j-th ridge ripple wave filter of i-th image block,Represent the corresponding j-th feature segment of i-th image block, γ=
(a, b, θ), K (γ) represent the inverse of ridge ripple wave filter frobenius norms, and e represents natural constant, and Y represents ridge ripple wave filter
Ridge ripple function, y1Represent the abscissa positions of 9 × 9 ridge ripple wave filter pixel, y2Represent 9 × 9 ridge ripple wave filter pixel
The vertical coordinate position of point, θ represent the directioin parameter of ridge ripple wave filter.
The more new formula of displacement parameter is as follows:
Wherein btRepresent the displacement parameter of the ridge ripple wave filter that the t time iteration is tried to achieve, bt-1Join for the displacement that t-1 time is tried to achieve
Number, δ represent coefficient, and span is [0,1], and ∑ represents addition, xiFor i-th piece 31 × 31 of SAR image sampling block,Represent
The corresponding j-th ridge ripple wave filter of i-th image block,Represent the corresponding j-th feature segment of i-th image block, γ=(a,
B, θ), K (γ) represents the inverse of ridge ripple wave filter frobenius norms, and e represents natural constant, and Y represents the ridge of ridge ripple wave filter
Wave function.
Characteristic pattern more new formula is as follows:
WhereinThe characteristic pattern that the t time iteration is tried to achieve is represented,For the characteristic pattern that t-1 iteration is tried to achieve, δ represents step-length,
Span is [0,1], and ∑ represents sum operation, xiI-th input picture block is represented,Represent i-th image block corresponding
J ridge ripple wave filter,Represent corresponding j-th characteristic pattern of i-th image block in t-1 iteration.
9th step, using the ridge ripple filter set after renewal as selected ridge ripple filter set, returns the 4th step, right
Input picture block re-starts study.
10th step, the ridge ripple wave filter that study is obtained is preserved to the input picture block ridge ripple filter set for succeeding in school
In, the study to the input picture block feature is completed, and exports the ridge ripple filter set that the input picture block succeeds in school.
11st step, judges whether all image blocks complete the study of feature by deconvolution structural model, if so, terminates
Program, otherwise, the next image block of input simultaneously performs the 3rd step.
Step 6, segmentation SAR image mixing aggregated structure atural object pixel subspace.
By all mutually not characteristic set splicings of connected region, using spliced characteristic set as code book.
All features to each mutual not connected region, calculate the inner product with each feature in code book respectively, obtain every
Projection vector of all features in individual region on code book.
To each, mutually the projection vector of connected region does not carry out maximum pond, obtains the corresponding structure spy in each region
Levy vector.
AP clustering algorithms are propagated using neighbour, the structural eigenvector of all mutual connected regions is not clustered, is obtained
The segmentation result of mixing aggregated structure atural object pixel subspace.
Step 7, segmenting structure pixel subspace.
Vision semantic rule is used, splits line target.
If i-th sketch line liWith j-th strip sketch line ljThe distance between be Dij, liDirection be Oi, ljDirection be Oj,
The total number of i, j ∈ [1,2 ..., S], S for sketch line.
By width more than 3 pixels line target with two sketch line liAnd ljRepresent, liAnd ljThe distance between DijIt is less than
T1And poor (the O in directioni-Oj) less than 10 degree, wherein T1=5.
If the s article sketch line lsGeometry window wsThe interior average gray per string is AiIf the gray scale difference of adjacent column is
ADi=| Ai-Ai+1|, if zs=[zs1,zs2,...,zs9] for the gray scale difference AD of adjacent columniLabel vector.
By width less than 3 pixels line target with single sketch line lsRepresent, lsGeometry window wsIt is interior, calculate phase
The gray scale difference AD of adjacent columniIf, ADi>T2, then zsi=1;Otherwise zsi=0, zsIn have two elements value be 1, remaining is 0, its
Middle T2=34.
If L1,L2It is the set of the sketch line for representing line target, if Dij<T1And | Oi-Oj|<10, then li,lj∈L1;
If sum is (zs)=2, then ls∈L2, the sum of wherein sum () expression parameter elements.
In structure-pixel subspace, according to the set L of the sketch line of line target1, by liAnd ljBetween region as line mesh
Mark.
In structure-pixel subspace, according to the set L of the sketch line of line target2, l will be coveredsRegion as line target.
Based on the feature of gathering of sketch line, split comprising the following steps that for pinpoint target:
1st step, in the structural region of administrative division map, all sketch wire tags that would not indicate line target are candidate's sketch line
Sketch line in set;
2nd step, randomly selects a sketch line from candidate's sketch line set, with an end points of selected sketch line
Centered on, construct the geometry window that size is for 5 × 5;
3rd step, judges the end points with the presence or absence of other sketch lines in geometry window, if existing, performs the 4th step;Otherwise,
Perform the 6th step;
4th step, judges whether to only exist an end points, if so, carries out the end points place sketch line and current sketch line
Connection;Otherwise, perform the 5th step;
5th step, the sketch line that sketch line selected by connection is located with each end points, chooses wherein angle from all connecting lines
The sketch line that two maximum sketch lines are completed as connection;
6th step, judges the interior end points with the presence or absence of other sketch lines of geometry window of another end points of sketch line, if
Exist, perform the 4th step;Otherwise, perform the 7th step;
7th step, the sketch line to completing attended operation choose the sketch line comprising two and more than two sketch line segments,
Bar number n comprising sketch line segment, wherein n >=2 in sketch line selected by statistics;
8th step, judges that the bar number n of sketch line then performs the 9th step whether equal to 2, if so,;Otherwise, perform the 10th step;
Sketch line of the angle value on sketch line summit in the range of [10 °, 140 °] is gathered spy as having by the 9th step
The sketch line levied;
10th step, selects the sketch line of the angle value on the corresponding n-1 summit of sketch line all in the range of [10 °, 140 °];
11st step, in selected sketch line, is defined as follows two kinds of situations:
Whether the first situation, judge the i-th -1, the adjacent two sketch line segments of i-th sketch line segment, i+1 bar i-th
The same side of bar sketch line segment place straight line, 2≤i≤n-1, if all sketch line segments on sketch line and adjacent segments are all same
Side, then the labelling sketch line is with the sketch line for gathering feature;
Whether second situation, judge the i-th -1, the adjacent two sketch line segments of i-th sketch line segment, i+1 bar i-th
The same side of bar sketch line segment place straight line, 2≤i≤n-1, if there is n-1 bar sketch line segments with adjacent segments same on sketch line
Side, and have a sketch line segment to be adjacent line segment in non-the same side, also the labelling sketch line is with the element for gathering feature
Retouch line;
12nd step, an optional sketch line in the sketch line for gathering feature are held by two of selected sketch line
Point coordinates, determines the distance between two end points, if the end-point distances are in the range of [0,20], then using selected sketch line as table
Show the sketch line of pinpoint target;
13rd step, judge it is untreated whether all selected with the sketch line for gathering feature, if so, perform the 12nd step;
Otherwise, perform the 14th step;
14th step, with the method for super-pixel segmentation, to the sketch line for representing pinpoint target in synthetic aperture radar SAR image
The pixel of surrounding carries out super-pixel segmentation, by super-pixel of the gray value of super-pixel after segmentation in [0,45] or [180,255]
As pinpoint target super-pixel;
15th step, merges pinpoint target super-pixel, using the border of the pinpoint target super-pixel after merging as pinpoint target
Border, obtain the segmentation result of pinpoint target.
The result of line target and pinpoint target segmentation is merged, the segmentation result of structure-pixel subspace is obtained.
Step 8, splits homogenous region pixel subspace.
1st step, arbitrarily chooses a pixel, from the pixel subspace of homogenous region centered on selected pixel
3 × 3 square window is set up, the standard deviation sigma of the window is calculated1。
The length of side of square window is increased by 2 by the 2nd step, obtains new square window, calculates the standard deviation of new square window
σ2。
3rd step, if standard deviation threshold method T3=3, if | σ1-σ2|<T3, then by standard deviation be σ2Square window as final
Square window, perform the 4th step;Otherwise, perform the 2nd step.
4th step, according to the following formula, calculates the prior probability of center pixel in square window:
Wherein, p '1The prior probability of center pixel in square window is represented, exp () represents exponential function operation, η ' tables
Show probabilistic model parameter, η ' values are 1, xk′' represent square window in belong to kth ' class number of pixels, k' ∈ [1 ...,
K'], K' represents the classification number of segmentation, and K' values are 5, xi' represent the pixel for belonging to the i-th ' class in the square window that obtains of the 3rd step
Number.
5th step, the probability density of pixel grey scale is multiplied with the probability density of texture, likelihood probability p' is obtained2, wherein,
The probability density of gray scale is obtained by the distribution of fading channel Nakagami, and the probability density of texture is obtained by t-distribution.
6th step, by prior probability p1' and likelihood probability p2' be multiplied, obtain posterior probability p12'。
Whether the 7th step, also have untreated pixel in judging homogenous region pixel subspace, if having, perform the 1st step;
Otherwise, perform the 9th step.
8th step, according to maximum posteriori criterion, obtains the segmentation result of homogenous region pixel subspace.
Step 9, combination and segmentation result.
By the segmentation of mixing aggregated structure atural object pixel subspace, homogenous region pixel subspace and structure-pixel subspace
As a result merge, obtain the final segmentation result of synthetic aperture radar SAR image.
The effect of the present invention is further described with reference to analogous diagram.
1. simulated conditions:
The present invention emulation hardware condition be:Intellisense and image understanding laboratory graphics workstation;Present invention emulation
The synthetic aperture radar SAR image for being used is:Ku wave band resolution is 1 meter of Piperiver figures.
2. emulation content:
The emulation experiment of the present invention is that the Piperiver figures in SAR image are split, as shown in Fig. 2 (a)
Piperiver schemes.The figure is from the synthetic aperture radar SAR image that Ku wave band resolution is 1 meter.
Using the present invention SAR image sketch step, to Piperiver the retouching of pixel shown in Fig. 2 (a), obtain as
The deeper sketch line of sketch map shown in Fig. 2 (b), wherein color obtains sketch line to represent line target and pinpoint target.
Using the division pixel subspace step of the present invention, to the sketch map compartmentalization shown in Fig. 2 (b), obtain such as Fig. 2
Administrative division map shown in (c).White space in Fig. 2 (c) represents mixing aggregated structure semantic space, and other is homogeneous texture language
Adopted space and structure semanticses space.Administrative division map shown in Fig. 2 (c) is mapped to into former SAR image shown in Fig. 2 (a), is obtained such as Fig. 2 (d)
Shown SAR image mixing aggregated structure atural object pixel subspace figure.
Image Segmentation Methods Based on Features pinpoint target step gathered based on sketch line using the present invention, by administrative division map shown in Fig. 2 (c)
Structural region is mapped to sketch map shown in Fig. 2 (b), obtains the corresponding sketch line of structural region shown in Fig. 3 (a).Fig. 3 (b) institutes
In the corresponding sketch line of structural region for showing, black is the sketch line for representing line target, the structural region correspondence shown in Fig. 3 (c)
Sketch line in, black is the sketch line for representing pinpoint target.Piperiver figures shown in Fig. 2 (a) carry out dividing for pinpoint target
Cut, obtain the segmentation result figure of the pinpoint target shown in Fig. 3 (d), wherein black region represents pinpoint target.
Using the segmentation SAR image mixing aggregated structure atural object pixel subspace step of the present invention, to shown in Fig. 2 (d)
The mixing aggregated structure atural object pixel subspace figure of Piperiver figures is split, and obtains the mixed land cover picture shown in Fig. 4 (a)
Sub-prime space segmentation result figure, its grey area represent untreated ground object space, and the region representation of remaining same color is same
A kind of atural object, the different atural object of the region representation of different colours.
Using the combination and segmentation result step of the present invention, mixing aggregated structure atural object pixel merged shown in Fig. 4 (a) is empty
Between segmentation result and homogenous region pixel subspace segmentation result and structure-pixel subspace segmentation result, obtain Fig. 4 (b), Fig. 4
B () is the final segmentation result of Fig. 2 (a) Piperiver images, Fig. 4 (c) is semantic based on level vision and adaptive neighborhood is more
Final segmentation result figure of the SAR image segmentation method of Xiang Shiyin models to Piperiver images.
3. simulated effect analysis:
Fig. 4 (c) is final segmentation result figure of the inventive method to Piperiver images, and Fig. 4 (d) is regarded based on level
Feel the semantic final segmentation result with the SAR image segmentation method of the hidden model of adaptive neighborhood multinomial to Piperiver images
Figure, by the contrast of segmentation result figure, it could be assumed that, the inventive method is for mixing aggregated structure atural object pixel subspace
Border determines that more accurately, for the segmentation of homogenous region pixel subspace region consistency is substantially preferable, classification number more adduction
Reason, and preferable dividing processing has been carried out to the pinpoint target in structure-pixel subspace.Using the inventive method pairing pore-forming
Footpath radar SAR image is split, and effectively splits SAR image, and improves the accuracy of SAR image segmentation.
Claims (10)
1. a kind of based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, comprise the steps:
(1) sketch SAR image:
(1a) to the synthetic aperture radar SAR image being input into, according to the pixel fluctuating characteristic distributions of SAR image, obtain its sketch
Model;
(1b) from sketch model extraction sketch map, obtain the sketch map of synthetic aperture radar SAR image;
(2) divide pixel subspace:
(2a) using sketch line fields method, pixel subspace is divided to synthetic aperture radar SAR image, synthetic aperture is obtained
The administrative division map of radar SAR image;
(2b) administrative division map is mapped in the synthetic aperture radar SAR image of input, is mixed in obtaining synthetic aperture radar SAR image
Close aggregated structure atural object pixel subspace, homogenous region pixel subspace and structure-pixel subspace;
(3) build ridge ripple filter set:
(3a) from the administrative division map of synthetic aperture radar SAR image, extraction mixing aggregated structure atural object pixel subspace is corresponding poly-
Collection region, it is interval interior to be divided into i.e. 18,18 intervals direction at intervals of 10 ° at [0 °, 180 °], it is interval that each is counted respectively
The line segment bar number of sketch line segment in the interior aggregation zone;
(3b) to all of sketch line segment in the aggregation zone, according to each interval in line segment bar number number be ranked up,
The collating sequence in direction is obtained, the number of degrees using front 6 directions in the collating sequence in direction are as the side in ridge ripple wave filter
To parameter;
(3c) according to the following formula, according to parameter a, θ and b calculates the ridge ripple function of 9 × 9 ridge ripple wave filter:
Y=a × (y1×cosθ+y2×sinθ-b)
Wherein, Y represents the ridge ripple function of ridge ripple wave filter, and a represents the scale parameter of ridge ripple wave filter, the span of a for [0,
3], the discrete interval of a is 0.2, y1Represent the abscissa positions of ridge ripple wave filter pixel, y1Span be [0,8], y1
Discrete interval represent that cosine is operated for 1, cos, θ represents the directioin parameter of ridge ripple wave filter, y2Represent ridge ripple wave filter pixel
The vertical coordinate position of point, y2Span be [0,8], y2Discrete interval represent sinusoidal operation for 1, sin;B represents that ridge ripple is filtered
The displacement parameter of ripple device, when directioin parameter θ for [0 °, 90 °) when, b is in [0,9 × (sin θ+cos θ)] is interval with intervals of 0.2
Carry out discretization, when directioin parameter θ for [90 °, 180 °) when, b is in [9 × cos θ, 9 × sin θ] is interval entering at intervals of 0.2
Row discretization;
(3d) according to the following formula, calculate each ridge ripple wave filter:
Wherein, c (Y) represents the ridge ripple wave filter using ridge ripple function Y as parameter, and K represents ridge ripple wave filter frobenius norms
Inverse, exp represents the index operation with natural constant e as bottom;
(3e) calculated each ridge ripple wave filter group is synthesized into ridge ripple filter set;
(4) construct deconvolution structural model:
(4a) each mutual disconnected area to the mixing aggregated structure atural object pixel subspace in synthetic aperture radar SAR image
Domain, by 31 × 31 window carry out every a sliding window sample, obtain the corresponding multiple images block in each region, by multiple images block according to
It is secondary to be input in deconvolution structural model, obtain the input layer of deconvolution structural model;
(4b) carry out convolution operation to reconstruct the image block in input layer using characteristic pattern and ridge ripple wave filter, obtain deconvolution knot
The warp lamination of structure model;
(4c) according to the following formula, calculate data fidelity item:
Wherein, E (c) represents data fidelity item, and c represents the ridge ripple wave filter in deconvolution structural model warp lamination, and N is represented will
The total number of the image block that the mutual disconnected region of each of study includes, ∑ represent sum operation, | | | |FExpression is done
Frobenius norms are operated,Represent the square operation of frobenius norms, xiIn representing deconvolution structural model to be constructed
I-th input picture block, MiThe corresponding ridge ripple wave filter of i-th image block being input in representing deconvolution structural model to be constructed
Sum, * represents convolution operation,Represent in deconvolution structural model to be constructed corresponding j-th feature of i-th image block
Figure,Represent in deconvolution structural model to be constructed the corresponding j-th ridge ripple wave filter of i-th image block;
(4d) according to the following formula, computation structure fidelity item:
Wherein, G (c) represents structure fidelity item, and R () represents the operation for seeking all sketch line total lengths in sketch map, SM ()
Represent the operation extracted with the one-to-one sketch segment of input picture block;
(4e) according to the following formula, calculating target function:
Wherein, L (c) represents object function,Represent when object function L (c) value is minimum, ask for ridge ripple wave filter c's
Operation;
(4f) the ridge ripple filter set obtained by object function guidance learning is exported, the output of deconvolution structural model is obtained
Layer;
(5) train deconvolution structural model:
(5a) structural failure threshold value is set to into 0.1;
(5b) step (4a) sampling is obtained image block to be sequentially inputted in deconvolution structural model;
(5c) from ridge ripple filter set, six wave filter, its directioin parameter 6 by statistics in step (3b) are randomly selected
Direction obtains, its displacement parameter and scale parameter random initializtion, by these initial six wave filter groups into filter set
As selected ridge ripple filter set;
(5d) characteristic pattern that 6 sizes for 39 × 39 is initialized for 39 × 39 null matrix with 6 sizes, by initialization after 6
Individual size for 39 × 39 characteristic pattern as feature set of graphs;
(5e) feature set of graphs and selected ridge ripple filter set are carried out convolution operation to reconstruct input picture block;
(5f) using the structure fidelity item formula in step (4d), calculate the structure fidelity item of reconstruct input picture block;
(5g) judge whether the structure fidelity item of current reconstruct input picture block is less than structural failure threshold value, if so, then perform step
Suddenly (5j), otherwise, execution step (5h);
(5h) using scale parameter more new formula and displacement parameter more new formula, step (4c) data fidelity item formula is updated respectively
The scale parameter and displacement parameter of median ridge wave filter, the ridge ripple filter set after being updated update public using characteristic pattern
Formula, updates characteristic pattern, the feature set of graphs after being updated;
(5i) using the ridge ripple filter set after renewal as selected ridge ripple filter set, by the feature atlas after renewal
Cooperation is characterized set of graphs, and return to step (5e) re-starts study to input picture block;
(5j) the ridge ripple wave filter that study is obtained is preserved to the reconstruct input picture block in the ridge ripple filter set for succeeding in school,
The study to the input picture block feature is completed, and exports the ridge ripple filter set that the input picture block succeeds in school;
(5k) judge whether all image blocks complete the study of feature by deconvolution structural model, if so, terminate program, it is no
Then, next image block execution step (5c) are input into;
(6) split SAR image mixing aggregated structure atural object pixel subspace:
(6a) the ridge ripple filter set of all mutually disconnected mixing aggregated structure atural object pixel subspace regional trainings is spelled
It is connected into code book;
(6b) will own in the mutually ridge ripple filter set of disconnected mixing aggregated structure atural object pixel subspace regional training
Ridge ripple wave filter, projected to code book, obtained projection vector;
(6c) maximum pond is carried out to the projection vector in each mutual disconnected mixing aggregated structure atural object pixel subspace region,
Obtain a structural eigenvector;
(6d) AP clustering algorithms are propagated using neighbour, structural eigenvector is clustered, obtains relative with structural eigenvector
The segmentation result of the mixing aggregated structure atural object pixel subspace answered;
(7) segmenting structure pixel subspace:
(7a) vision semantic rule is used, splits line target;
(7b) feature of gathering based on sketch line, splits pinpoint target;
(7c) result of line target and pinpoint target segmentation is merged, obtains the segmentation result of structure-pixel subspace.
(8) split homogenous region pixel subspace:
Using the homogenous region dividing method based on multinomial logistic regression prior model, homogenous region pixel subspace is carried out
Segmentation, obtains the segmentation result of homogenous region pixel subspace.
(9) combination and segmentation result:
The segmentation result of mixing aggregated structure pixel subspace, homogenous region pixel subspace and structure-pixel subspace is closed
And, obtain the final segmentation result of synthetic aperture radar SAR image.
2. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, step (1) sketch is comprised the following steps that:
1st step, constructs the side being made up of pixel with different directions and yardstick, a template of line, using the side of template
To with dimensional information structural anisotropy's Gaussian function, count the weight coefficient of every bit in the template, its mesoscale number takes
It is worth for 3~5, direction number value is 18;
2nd step, according to the following formula, calculates the average of pixel in the synthetic aperture radar SAR image corresponding with template area position
And variance yields:
Wherein, μ represents the average of pixel in the synthetic aperture radar SAR image corresponding with template area position, and ∑ is represented to be asked
And operation, g represents the position of a pixel in the Ω region of template, and ∈ is represented and belonged to symbol, wgRepresent template Ω
Weight coefficient of the pixel at g positions, w in regiongSpan be wg∈ [0,1], AgRepresent and the Ω region of template
Pixel value of the middle pixel at g positions in corresponding synthetic aperture radar SAR image, ν represent relative with template area position
The variance yields of pixel in the synthetic aperture radar SAR image answered;
3rd step, according to the following formula, calculates the response value of each pixel comparison value operator in synthetic aperture radar SAR image:
Wherein, R represents the response value of each pixel comparison value operator in synthetic aperture radar SAR image, and min { } is represented and asked most
Little Value Operations, a and b represent two different regions in template, μ respectivelyaAnd μbRepresent respectively and template area a and template region
The average of pixel in the corresponding synthetic aperture radar SAR image in domain b positions;
4th step, according to the following formula, calculates response value of each pixel to dependency operator in synthetic aperture radar SAR image:
Wherein, C represents response value of each pixel to dependency operator in synthetic aperture radar SAR image, and a and b is represented respectively
Two different regions, ν in templateaAnd νbSynthetic aperture thunder corresponding with template area a and template area b positions is represented respectively
The variance yields of pixel, μ up in SAR imageaAnd μbSynthetic aperture corresponding with template area a and template area b positions is represented respectively
The average of pixel in radar SAR image,Represent square root functions;
5th step, according to the following formula, merges the response value and synthetic aperture of pixel comparison value operator in synthetic aperture radar SAR image
Response value of the pixel to dependency operator in radar SAR image, in calculating synthetic aperture radar SAR image, each pixel is to each
The response value of template:
Wherein, F represents response value of each pixel to each template in synthetic aperture radar SAR image, and R and C represents conjunction respectively
Sound of the pixel to dependency operator in pixel comparison value operator and synthetic aperture radar SAR image in aperture radar SAR image
Should be worth,Represent square root functions;
6th step, selects the template with maximum response from the response value of each template, schemes as synthetic aperture radar SAR
The template of pixel as in, and the direction of the template with maximum response is made by maximum response as the intensity of the pixel
For the direction of the pixel, the sideline response diagram and directional diagram of synthetic aperture radar SAR image are obtained;
7th step, using the selected template with maximum response of each pixel in synthetic aperture radar SAR image, obtains
The gradient map of synthetic aperture radar SAR image;
8th step, according to the following formula, merges the value of the response value and gradient map of sideline response diagram, is calculated intensity level, by intensity level
Each pixel constitute the intensity map of synthetic aperture radar SAR image:
Wherein, I represents intensity level, and x represents the value in the response diagram of synthetic aperture radar SAR image sideline, and y represents synthetic aperture thunder
Value up in SAR image gradient map;
9th step, using non-maxima suppression method, detects to intensity map, obtains suggestion sketch;
10th step, chooses the pixel with maximum intensity from suggestion sketch, by the pixel in suggestion sketch with the maximum intensity
The pixel of connection connects to form suggestion line segment, obtains suggestion sketch map;
11st step, according to the following formula, calculates the code length gain of sketch line in suggestion sketch map:
Wherein, CLG represents the code length gain of sketch line in suggestion sketch map, and m represents pixel in current sketch line neighborhood
Number, ∑ represent sum operation, and t represents the numbering of pixel in current sketch line neighborhood, AtRepresent t in current sketch line neighborhood
The observation of individual pixel, At,0Represent on the premise of current sketch line can not represent structural information, t in the sketch line neighborhood
The estimated value of individual pixel, ln () represent the log operations with e as bottom, At,1Represent and can represent that structure is believed in current sketch line
BreathUnder the premise of, the estimated value of t-th pixel in the sketch line neighborhood;
12nd step, given threshold T, the span of T is 5~50, selects CLG>The suggestion sketch line of T is used as in final sketch map
Sketch line, obtain the input corresponding sketch map of synthetic aperture radar SAR image.
3. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, sketch line fields method described in step (2) is comprised the following steps that:
Sketch line, according to the concentration class of sketch line segment in the sketch map of synthetic aperture radar SAR image, is divided into table by the 1st step
The aggregation sketch line for showing aggregation atural object and the sketch line for representing border, line target and isolated target;
2nd step, according to the statistics with histogram of sketch line segment concentration class, chooses the sketch line segment work that concentration class is equal to optimum concentration class
For seed line-segment sets { Ek, k=1,2 ..., m }, wherein, EkAny bar sketch line segment in expression seed line-segment sets, k represent kind
Sub-line section concentrates the label of any bar sketch line segment, m to represent the total number of seed line segment, and { } represents set operation;
3rd step, using the unselected line segment for being added to seed line-segment sets sum as basic point, with the new line segment of this basic point recursive resolve
Set;
4th step, one radius of construction are the circular primitive in the optimum concentration class interval upper bound, with the circular primitive in line segment aggregate
Line segment expanded, the line segment aggregate ecto-entad after expansion is corroded, is obtained in sketch map with sketch point as list
The aggregation zone of position;
5th step, the sketch line to representing border, line target and isolated target, during each the sketch point with each sketch line is
Heart construction size is 5 × 5 geometry window, obtains structural region;
6th step, will remove part beyond aggregation zone and structural region as can not sketch region in sketch map;
7th step, by the aggregation zone in sketch map, structural region and can not sketch region, be respectively mapped to synthetic aperture radar
In SAR image, mixing aggregated structure atural object pixel subspace, the structure-pixel subspace of synthetic aperture radar SAR image are obtained
With homogeneous texture pixel subspace.
4. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, the more new formula of the scale parameter a of the ridge ripple wave filter described in step (5h) is as follows:
Wherein a represents the scale parameter of ridge ripple wave filter, atFor the yardstick that t steps are tried to achieve, at-1For the yardstick that t-1 steps are tried to achieve, δ
Coefficient is represented, span is [0,1], and ∑ represents addition, xiFor i-th piece 31 × 31 of SAR image sampling block,Represent i-th
The corresponding j-th ridge ripple wave filter of individual image block,Represent the corresponding j-th feature segment of i-th image block, γ=(a, b,
θ), K (γ) represents the inverse of ridge ripple wave filter frobenius norms, and e represents natural constant, and Y represents the ridge ripple of ridge ripple wave filter
Function, y1Represent the abscissa positions of 9 × 9 ridge ripple wave filter pixel, y2Represent 9 × 9 ridge ripple wave filter pixel it is vertical
Coordinate position, θ represent the directioin parameter of ridge ripple wave filter.
5. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, the more new formula of the displacement parameter b described in step (5h) is as follows:
Wherein btRepresent the displacement parameter of the ridge ripple wave filter that the t time iteration is tried to achieve, bt-1For the displacement parameter tried to achieve for t-1 time, δ tables
Show coefficient, span is [0,1], ∑ represents addition, xiFor i-th piece 31 × 31 of SAR image sampling block,Represent i-th
The corresponding j-th ridge ripple wave filter of image block,Represent the corresponding j-th feature segment of i-th image block, γ=(a, b, θ),
K (γ) represents the inverse of ridge ripple wave filter frobenius norms, and e represents natural constant, and Y represents the ridge ripple letter of ridge ripple wave filter
Number.
6. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, the characteristic pattern more new formula described in step (5h) is as follows:
WhereinThe characteristic pattern that the t time iteration is tried to achieve is represented,For the characteristic pattern that t-1 iteration is tried to achieve, δ represents step-length, value
Scope is [0,1], and ∑ represents sum operation, xiI-th input picture block is represented,Represent that i-th image block is corresponding j-th
Ridge ripple wave filter,Represent the corresponding j-th feature segment of i-th image block in t-1 iteration.
7. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, the vision semantic rule described in step (7a) is as follows:
If i-th sketch line liWith j-th strip sketch line ljThe distance between be Dij, liDirection be Oi, ljDirection be Oj, i, j
The total number of ∈ [1,2 ..., S], S for sketch line;
By width more than 3 pixels line target with two sketch line liAnd ljRepresent, liAnd ljThe distance between DijLess than T1And
Poor (the O in directioni-Oj) less than 10 degree, wherein T1=5;
If the s article sketch line lsGeometry window wsThe interior average gray per string is AiIf the gray scale difference of adjacent column is ADi=
|Ai-Ai+1|, if zs=[zs1,zs2,...,zs9] for the gray scale difference AD of adjacent columniLabel vector;
By width less than 3 pixels line target with single sketch line lsRepresent, lsGeometry window wsIt is interior, calculate adjacent column
Gray scale difference ADiIf, ADi>T2, then zsi=1;Otherwise zsi=0, zsIn have two elements value be 1, remaining is 0, wherein T2
=34;
If L1,L2It is the set of the sketch line for representing line target, if Dij<T1And | Oi-Oj|<10, then li,lj∈L1;If
sum(zs)=2, then ls∈L2, wherein sum () represent to vector important summation operation.
8. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, the segmentation line target described in step (7a) is comprised the following steps that:
1st step, in structure-pixel subspace, according to the set L of the sketch line of line target1, by liAnd ljBetween region as line
Target;
2nd step, in structure-pixel subspace, according to the set L of the sketch line of line target2, l will be coveredsRegion as line mesh
Mark.
9. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, which is special
Levy and be, the segmentation pinpoint target described in step (7b) is comprised the following steps that:
1st step, in the structural region of administrative division map, all sketch wire tags that would not indicate line target are candidate's sketch line set
In sketch line;
2nd step, randomly selects a sketch line from candidate's sketch line set, during an end points with selected sketch line is
The heart, constructs the geometry window that size is for 5 × 5;
3rd step, judges the end points with the presence or absence of other sketch lines in geometry window, if existing, performs the 4th step;Otherwise, perform
6th step;
4th step, judges whether to only exist an end points, is if so, attached the end points place sketch line and current sketch line;
Otherwise, perform the 5th step;
5th step, the sketch line that sketch line selected by connection is located with each end points choose wherein angle maximum from all connecting lines
Two sketch lines as the sketch line that completes of connection;
6th step, judges the interior end points with the presence or absence of other sketch lines of geometry window of another end points of sketch line, if depositing
Performing the 4th step;Otherwise, perform the 7th step;
7th step, the sketch line to completing attended operation choose the sketch line comprising two and more than two sketch line segments, statistics
Bar number n comprising sketch line segment, wherein n >=2 in selected sketch line;
8th step, judges that the bar number n of sketch line then performs the 9th step whether equal to 2, if so,;Otherwise, perform the 10th step;
Sketch line of the angle value on sketch line summit in the range of [10 °, 140 °] is gathered feature as having by the 9th step
Sketch line;
10th step, selects the sketch line of the angle value on the corresponding n-1 summit of sketch line all in the range of [10 °, 140 °];
11st step, in selected sketch line, is defined as follows two kinds of situations:
The first situation, judges whether the i-th -1, the adjacent two sketch line segments of i-th sketch line segment, i+1 bar are plain at i-th
The same side of line segment place straight line, 2≤i≤n-1 are retouched, if all sketch line segments on sketch line and adjacent segments are all same
Side, then the labelling sketch line is with the sketch line for gathering feature;
Second situation, judges whether the i-th -1, the adjacent two sketch line segments of i-th sketch line segment, i+1 bar are plain at i-th
The same side of line segment place straight line, 2≤i≤n-1 are retouched, if there is n-1 bar sketch line segments with adjacent segments in the same side on sketch line,
And have a sketch line segment to be adjacent line segment in non-the same side, also the labelling sketch line is with the sketch line for gathering feature;
12nd step, an optional sketch line in the sketch line for gathering feature are sat by two end points of selected sketch line
Mark, determines the distance between two end points, if the end-point distances are in the range of [0,20], then only using selected sketch line as expression
The sketch line of vertical target;
13rd step, judge it is untreated whether all selected with the sketch line for gathering feature, if so, perform the 12nd step;Otherwise,
Perform the 14th step;
14th step, with the method for super-pixel segmentation, around the sketch line of expression pinpoint target in synthetic aperture radar SAR image
Pixel carry out super-pixel segmentation, by super-pixel conduct of the gray value of super-pixel after segmentation in [0,45] or [180,255]
Pinpoint target super-pixel;
15th step, merge pinpoint target super-pixel, using the border of the pinpoint target super-pixel after merging as pinpoint target side
Boundary, obtains the segmentation result of pinpoint target.
10. according to claim 1 based on ridge ripple wave filter and the SAR image segmentation method of deconvolution structural model, its
Be characterised by, described in step (8) based on multinomial logistic regression prior model homogenous region dividing method concrete steps
It is as follows:
1st step, is arbitrarily chosen a pixel from the pixel subspace of homogenous region, is set up centered on selected pixel
3 × 3 square window, calculates the standard deviation sigma of the window1;
The length of side of square window is increased by 2 by the 2nd step, obtains new square window, calculates the standard deviation sigma of new square window2;
3rd step, if standard deviation threshold method T3=3, if | σ1-σ2|<T3, then by standard deviation be σ2Square window as final side
Shape window, performs the 4th step;Otherwise, perform the 2nd step;
4th step, according to the following formula, calculates the prior probability of center pixel in square window:
Wherein, p '1The prior probability of center pixel in square window is represented, exp () represents exponential function operation, and η ' represents general
Rate model parameter, η ' values are 1, xk′' represent square window in belong to kth ' class number of pixels, k' ∈ [1 ..., K'], K'
The classification number of segmentation is represented, K' values are 5, xi' represent the number of pixels for belonging to the i-th ' class in the square window that obtains of the 3rd step;
5th step, the probability density of pixel grey scale is multiplied with the probability density of texture, likelihood probability p' is obtained2, wherein, gray scale
Probability density is obtained by the distribution of fading channel Nakagami, and the probability density of texture is obtained by t-distribution;
6th step, by prior probability p1' and likelihood probability p2' be multiplied, obtain posterior probability p12';
Whether the 7th step, also have untreated pixel in judging homogenous region pixel subspace, if having, perform the 1st step;Otherwise,
Perform the 9th step;
8th step, according to maximum posteriori criterion, obtains the segmentation result of homogenous region pixel subspace.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611269937.4A CN106611423B (en) | 2016-12-30 | 2016-12-30 | SAR image segmentation method based on ridge ripple filter and deconvolution structural model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611269937.4A CN106611423B (en) | 2016-12-30 | 2016-12-30 | SAR image segmentation method based on ridge ripple filter and deconvolution structural model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106611423A true CN106611423A (en) | 2017-05-03 |
CN106611423B CN106611423B (en) | 2019-05-21 |
Family
ID=58636801
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611269937.4A Active CN106611423B (en) | 2016-12-30 | 2016-12-30 | SAR image segmentation method based on ridge ripple filter and deconvolution structural model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106611423B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107292268A (en) * | 2017-06-23 | 2017-10-24 | 西安电子科技大学 | The SAR image semantic segmentation method of quick ridge ripple deconvolution Structure learning model |
CN107341813A (en) * | 2017-06-15 | 2017-11-10 | 西安电子科技大学 | SAR image segmentation method based on structure learning and sketch characteristic inference network |
CN107403434A (en) * | 2017-07-28 | 2017-11-28 | 西安电子科技大学 | SAR image semantic segmentation method based on two-phase analyzing method |
CN107480677A (en) * | 2017-08-07 | 2017-12-15 | 北京深睿博联科技有限责任公司 | The method and device of area-of-interest in a kind of identification three-dimensional CT image |
CN107483058A (en) * | 2017-08-17 | 2017-12-15 | 西安电子科技大学 | A kind of match tracing reconstructing method instructed based on structuring observation and direction |
CN108109153A (en) * | 2018-01-12 | 2018-06-01 | 西安电子科技大学 | SAR image segmentation method based on SAR-KAZE feature extractions |
CN109272515A (en) * | 2018-08-17 | 2019-01-25 | 西安电子科技大学 | Based on the SAR image segmentation method that the multiple dimensioned CRF of high-order is unsupervised |
CN109344837A (en) * | 2018-10-22 | 2019-02-15 | 西安电子科技大学 | A kind of SAR image semantic segmentation method based on depth convolutional network and Weakly supervised study |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140043505A1 (en) * | 2011-02-11 | 2014-02-13 | Canadian Space Agency | Method and system of increasing spatial resolution of multi-dimensional optical imagery using sensor's intrinsic keystone |
CN104346814A (en) * | 2014-11-20 | 2015-02-11 | 西安电子科技大学 | SAR (specific absorption rate) image segmentation method based on hierarchy visual semantics |
CN105374033A (en) * | 2015-10-19 | 2016-03-02 | 西安电子科技大学 | SAR image segmentation method based on ridgelet deconvolution network and sparse classification |
CN105389798A (en) * | 2015-10-19 | 2016-03-09 | 西安电子科技大学 | Synthetic aperture radar (SAR) image segmentation method based on deconvolution network and mapping inference network |
-
2016
- 2016-12-30 CN CN201611269937.4A patent/CN106611423B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140043505A1 (en) * | 2011-02-11 | 2014-02-13 | Canadian Space Agency | Method and system of increasing spatial resolution of multi-dimensional optical imagery using sensor's intrinsic keystone |
CN104346814A (en) * | 2014-11-20 | 2015-02-11 | 西安电子科技大学 | SAR (specific absorption rate) image segmentation method based on hierarchy visual semantics |
CN105374033A (en) * | 2015-10-19 | 2016-03-02 | 西安电子科技大学 | SAR image segmentation method based on ridgelet deconvolution network and sparse classification |
CN105389798A (en) * | 2015-10-19 | 2016-03-09 | 西安电子科技大学 | Synthetic aperture radar (SAR) image segmentation method based on deconvolution network and mapping inference network |
Non-Patent Citations (2)
Title |
---|
FANG LIU ET.AL: "SAR Image Segmentation Based on Hierarchical Visual Semantic and Adaptive Neighborhood Multinomial Latent Model", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING》 * |
JIE WU ET.AL: "Local Maximal Homogeneous Region Search for SAR Speckle Reduction With Sketch-Based Geometrical Kernel Function", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107341813A (en) * | 2017-06-15 | 2017-11-10 | 西安电子科技大学 | SAR image segmentation method based on structure learning and sketch characteristic inference network |
CN107341813B (en) * | 2017-06-15 | 2019-10-25 | 西安电子科技大学 | SAR image segmentation method based on Structure learning and sketch characteristic inference network |
CN107292268A (en) * | 2017-06-23 | 2017-10-24 | 西安电子科技大学 | The SAR image semantic segmentation method of quick ridge ripple deconvolution Structure learning model |
CN107403434A (en) * | 2017-07-28 | 2017-11-28 | 西安电子科技大学 | SAR image semantic segmentation method based on two-phase analyzing method |
CN107403434B (en) * | 2017-07-28 | 2019-08-06 | 西安电子科技大学 | SAR image semantic segmentation method based on two-phase analyzing method |
CN107480677A (en) * | 2017-08-07 | 2017-12-15 | 北京深睿博联科技有限责任公司 | The method and device of area-of-interest in a kind of identification three-dimensional CT image |
CN107480677B (en) * | 2017-08-07 | 2020-04-28 | 北京深睿博联科技有限责任公司 | Method and device for identifying interest region in three-dimensional CT image |
CN107483058A (en) * | 2017-08-17 | 2017-12-15 | 西安电子科技大学 | A kind of match tracing reconstructing method instructed based on structuring observation and direction |
CN107483058B (en) * | 2017-08-17 | 2019-10-22 | 西安电子科技大学 | A kind of match tracing reconstructing method instructed based on structuring observation and direction |
CN108109153B (en) * | 2018-01-12 | 2019-10-11 | 西安电子科技大学 | SAR image segmentation method based on SAR-KAZE feature extraction |
CN108109153A (en) * | 2018-01-12 | 2018-06-01 | 西安电子科技大学 | SAR image segmentation method based on SAR-KAZE feature extractions |
CN109272515A (en) * | 2018-08-17 | 2019-01-25 | 西安电子科技大学 | Based on the SAR image segmentation method that the multiple dimensioned CRF of high-order is unsupervised |
CN109272515B (en) * | 2018-08-17 | 2020-10-09 | 西安电子科技大学 | Unsupervised SAR image segmentation method based on high-order multi-scale CRF |
CN109344837A (en) * | 2018-10-22 | 2019-02-15 | 西安电子科技大学 | A kind of SAR image semantic segmentation method based on depth convolutional network and Weakly supervised study |
CN109344837B (en) * | 2018-10-22 | 2022-03-04 | 西安电子科技大学 | SAR image semantic segmentation method based on deep convolutional network and weak supervised learning |
Also Published As
Publication number | Publication date |
---|---|
CN106611423B (en) | 2019-05-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106611423A (en) | SAR image segmentation method based on ridge wave filter and deconvolution structural model | |
CN106611420A (en) | SAR image segmentation method based on deconvolution network and sketch direction constraint | |
CN106683102B (en) | SAR image segmentation method based on ridge ripple filter and convolutional coding structure learning model | |
CN106611421A (en) | SAR image segmentation method based on feature learning and sketch line constraint | |
CN105374033B (en) | SAR image segmentation method based on ridge ripple deconvolution network and sparse classification | |
CN106846322B (en) | The SAR image segmentation method learnt based on curve wave filter and convolutional coding structure | |
CN103049763B (en) | Context-constraint-based target identification method | |
CN109063594A (en) | Remote sensing images fast target detection method based on YOLOv2 | |
CN105608692B (en) | Polarization SAR image segmentation method based on deconvolution network and sparse classification | |
CN107229904A (en) | A kind of object detection and recognition method based on deep learning | |
CN106611422A (en) | Stochastic gradient Bayesian SAR image segmentation method based on sketch structure | |
CN107341813B (en) | SAR image segmentation method based on Structure learning and sketch characteristic inference network | |
CN107123123A (en) | Image segmentation quality evaluating method based on convolutional neural networks | |
CN105930846B (en) | Classification of Polarimetric SAR Image method based on neighborhood information and SVGDL | |
CN106778821A (en) | Classification of Polarimetric SAR Image method based on SLIC and improved CNN | |
CN106203444B (en) | Classification of Polarimetric SAR Image method based on band wave and convolutional neural networks | |
CN106529508A (en) | Local and non-local multi-feature semantics-based hyperspectral image classification method | |
CN106651884B (en) | Mean field variation Bayes's SAR image segmentation method based on sketch structure | |
CN104915676A (en) | Deep-level feature learning and watershed-based synthetic aperture radar (SAR) image classification method | |
CN106909902A (en) | A kind of remote sensing target detection method based on the notable model of improved stratification | |
CN112862792B (en) | Wheat powdery mildew spore segmentation method for small sample image dataset | |
CN107403434B (en) | SAR image semantic segmentation method based on two-phase analyzing method | |
CN109409240A (en) | A kind of SegNet remote sensing images semantic segmentation method of combination random walk | |
CN106600595A (en) | Human body characteristic dimension automatic measuring method based on artificial intelligence algorithm | |
CN105069796B (en) | SAR image segmentation method based on small echo both scatternets |
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 |