CN107527001A - A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information - Google Patents
A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information Download PDFInfo
- Publication number
- CN107527001A CN107527001A CN201710198805.5A CN201710198805A CN107527001A CN 107527001 A CN107527001 A CN 107527001A CN 201710198805 A CN201710198805 A CN 201710198805A CN 107527001 A CN107527001 A CN 107527001A
- Authority
- CN
- China
- Prior art keywords
- space correlation
- correlation information
- linear space
- data collection
- mrow
- 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 OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of hyperspectral image classification method based on Steerable filter and linear space correlation information, and this method includes：Receive hyperspectral image data collection；Spatial texture information is obtained according to the hyperspectral image data collection；Linear space correlation information is obtained according to the hyperspectral image data collection；By the hyperspectral image data collection, spatial texture information and linear space correlation information linear fusion, new data set is obtained；Training set is picked out from the new data set with preset ratio at random, the remainder of the new data set is as test set；The vector machine supported using RBF is trained to the training set, obtains training pattern；The vector machine supported using RBF is classified to the test set, obtains the classification results of the high spectrum image.
Description
Technical field
The present invention relates to remote sensing hyperspectral image process field, more particularly, to one kind based on Steerable filter and linearly
The hyperspectral image classification method of spatial coherence information.
Background technology
High spectrum image spatial information is extracted by wave filter, the classification performance for improving spectrum picture is the previous research of mesh
Focus.Extraction of spatial information method has at present：1) shape filtering feature extraction；2) markov random file feature extraction；3) figure
As segmentation feature extraction；4) with texture blending wave filter extraction spatial information.
Start gradually to increase using filtering method extraction EO-1 hyperion spatial texture information, Shi and Shen et al. utilize multidimensional
Gabor filter carrys out the texture information of multi-angle extraction image, and nicety of grading is improved；Wang et al. employs Gabor
Filtering obtains preferable space characteristics, using Active Learning Method to there is the space neighborhood information of label training sample
Letter, propose the S2ISC semisupervised classification algorithms that a kind of empty spectrum combines；Wang et al. employs Gabor filtering and obtains space characteristics,
It is proposed the SS-LPSVM semisupervised classification algorithms that a kind of empty spectrum label is propagated；Li et al. extracts spatial information with Gabor filter
With arest neighbors information approach (SNR), it is proposed that Gabor-SNR algorithms are classified to high spectrum image；Rajadell et al. is used
Gabor filter extracts space characteristics to the subband of selection, improves classification performance；Wang et al., which is combined, is oriented to bilateral filtering
High-spectrum image space text feature is obtained with form properties feature and can effectively improve nicety of grading；Xia et al. is with improved
Bilateral edge filters algorithm extracts space characteristics, and E-ICA-RGF algorithms pair are proposed with reference to independent component analysis (ICA)
High spectrum image is classified, and classification performance improves much.The identification that many scholars propose filtering and maximum probability combines
Method, Kang et al. proposes the probability optimization method (EPF) based on holding edge filter, passes through supervised classifier first
SVM, classification is optimized using initial classification results edge filter, finally and with maximum probability approach obtains EO-1 hyperion
Classification results；Soomro et al. returns (Elastic NetRegression) with elastomeric network and obtains preliminary classification result, and leads to
Cross two-sided filter and obtain spatial information, finally and with maximum probability approach obtain hyperspectral classification result, all obtain good
Achievement in research.High spectrum image spatial texture information is extracted achieves certain effect for the research of classification, but there is also
Some shortcomings：
1) spatial texture information is excavated not abundant enough；
2) it have ignored booster action of the spatial coherence information to classification hyperspectral imagery.
3) traditional texture blending method easily removes spatial coherence.
The content of the invention
The present invention is to overcome at least one defect described in above-mentioned prior art, there is provided a kind of based on Steerable filter and linear
The hyperspectral image classification method of spatial coherence information.
It is contemplated that at least solves above-mentioned technical problem to a certain extent.
In order to reach above-mentioned technique effect, technical scheme is as follows：
A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information, including：Receive high
Spectral image data collection；Spatial texture information is obtained according to the hyperspectral image data collection；According to the high spectrum image number
Linear space correlation information is obtained according to collection；The hyperspectral image data collection, spatial texture information is related to linear space
Property information linear fusion, obtains new data set；At random training set, the new number are picked out from the new data set with preset ratio
According to the remainder of collection as test set；The vector machine supported using RBF is trained to the training set, is obtained
Training pattern；The vector machine supported using RBF is classified to the test set, obtains the high spectrum image
Classification results.
Preferably, described the step of obtaining spatial texture information according to the hyperspectral image data collection, includes：Pass through spy
Sign dimensionality reduction is handled high-spectral data collection, obtains the hyperspectral image data collection of information content redistribution, and utilize guiding
Filter 20 compositions before the hyperspectral image data collection to information content redistribution to be filtered, obtain spatial texture information.
Preferably, described the step of obtaining linear space correlation information according to the hyperspectral image data collection, includes：
First EO-1 hyperion linear space correlation information matrix is defined according to the hyperspectral image data collection and the second EO-1 hyperion is linear
Spatial coherence information matrix, wherein the first EO-1 hyperion linear space correlation information matrix is the linear space phase of horizontal direction
Closing property information matrix, the second EO-1 hyperion linear space correlation information matrix are the linear space correlation information square of vertical direction
Battle array；First EO-1 hyperion linear space correlation information matrix and the second EO-1 hyperion linear space correlation information matrix are added,
Obtain linear space correlation information.
Preferably, it is described that first EO-1 hyperion linear space correlation information square is defined according to the hyperspectral image data collection
The step of battle array and the second EO-1 hyperion linear space correlation information matrix, includes：
The spatial resolution of the high spectrum image is M × N, then the first EO-1 hyperion linear space correlation information matrix Dl
For：
Second EO-1 hyperion linear space correlation information matrix DvFor：
Wherein, (x, y) be pixel in the position of high spectrum image, M/2 is the center of high spectrum image horizontal direction,
N/2 be vertical direction center, Dl(x, y) is normalizing apart from size of each pixel apart from horizontal direction center
Change, Dv(x, y) be each pixel in abscissa apart from the normalization apart from size of horizontal direction center.
Preferably, it is described to pick out training set from the new data set with preset ratio at random, the new data set its
Remaining part is allocated as also to include after the step of test set：The vector machine method cross validation supported using RBF, find
Optimal parameter combination.
Preferably, the preset ratio is 10% ratio.
Preferably, it is described by the hyperspectral image data collection, spatial texture information and linear space correlation information line
Property fusion, the step of obtaining new data set includes：By the hyperspectral image data collection, spatial texture information and linear space phase
Closing property information three is added, and obtains new data set.
Compared with prior art, the beneficial effect of technical solution of the present invention is：
Spatial texture information and linear space correlation information of the invention by extracting high spectrum image is to high-spectrum
As being classified, it can effectively aid in spectral information to improve nicety of grading, make up the spatial coherence lost in spatial texture information.
Brief description of the drawings
Fig. 1 is a kind of classification hyperspectral imagery side based on Steerable filter and linear space correlation information of an embodiment
The indicative flowchart of method.
Embodiment
Accompanying drawing being given for example only property explanation, it is impossible to be interpreted as the limitation to this patent；
In order to more preferably illustrate the present embodiment, some parts of accompanying drawing have omission, zoomed in or out, and do not represent actual product
Size；
To those skilled in the art, it is to be appreciated that some known features and its explanation, which may be omitted, in accompanying drawing
's.
Technical scheme is described further with reference to the accompanying drawings and examples.
Embodiment 1
Fig. 1 is a kind of classification hyperspectral imagery side based on Steerable filter and linear space correlation information of an embodiment
The indicative flowchart of method.A kind of as shown in figure 1, high spectrum image based on Steerable filter and linear space correlation information point
Class method, including：
S1, receive hyperspectral image data collection D；
S2, spatial texture information D is obtained according to the hyperspectral image data collections；
In the present embodiment, spatial texture information D is obtained according to the hyperspectral image data collectionsThe step of include：Pass through
PCA dimensionality reductions obtain the hyperspectral image data collection D of information content redistribution to high-spectral data collection D processingPCA；Using drawing
Filtering is led to DPCAAbove 20 compositions are filtered, and obtain spatial texture information Ds。
S3, linear space correlation information D is obtained according to the hyperspectral image data collectionG；
In the present embodiment, described the step of obtaining linear space correlation information according to the hyperspectral image data collection, wraps
Include：
It is high according to the hyperspectral image data collection the first EO-1 hyperion linear space correlation information matrix of definition and second
Linear spatial coherence information matrix, wherein the first EO-1 hyperion linear space correlation information matrix is the line of horizontal direction
Property spatial coherence information matrix, the second EO-1 hyperion linear space correlation information matrix is related for the linear space of vertical direction
Property information matrix；By the first EO-1 hyperion linear space correlation information matrix and the second EO-1 hyperion linear space correlation information square
Battle array is added, and obtains linear space correlation information, i.e.,：
The spatial resolution of the high spectrum image is M × N, then the first EO-1 hyperion linear space correlation information matrix Dl
For：
Second EO-1 hyperion linear space correlation information matrix DvFor：
The linear space correlation information D of high spectrum imageGFor：
DG=Dl+DV
Wherein, (x, y) be pixel in the position of high spectrum image, M/2 is the center of high spectrum image horizontal direction,
N/2 be vertical direction center, Dl(x, y) is normalizing apart from size of each pixel apart from horizontal direction center
Change, Dv(x, y) be each pixel in abscissa apart from the normalization apart from size of horizontal direction center.
S4, by the hyperspectral image data collection D, spatial texture information DsWith linear space correlation information DGLinearly melt
Close, obtain new data set R；
It is described by the hyperspectral image data collection D, spatial texture information D in the present embodimentsWith linear space correlation
Information DGLinear fusion, the step of obtaining new data set, include：By the hyperspectral image data collection D, spatial texture information DsWith
Linear space correlation information DGThree is added, and obtains new data set, i.e.,：
R=D+DS+DG。
S5, training set R is picked out with preset ratio from the new data set R at randoms, the remainder of the new data set
As test set Rt；
In the present embodiment, the preset ratio is 10% ratio, i.e., is chosen at random from high-spectral data collection R with 10% ratio
Select training set Rs, remainder is as test set Rt。
S6, the vector machine (SVM) supported using RBF is to the training set RsIt is trained, obtains training mould
Type；
Also include before S6 steps：The SVM method cross validations supported with RBF, find optimal parameter group
Close.
S7, the vector machine supported using RBF is to the test set RtClassified, obtain the high-spectrum
The classification results of picture.
This programme by extract high spectrum image spatial texture information and linear space correlation information to high-spectrum
As being classified, it can effectively aid in spectral information to improve nicety of grading, make up the spatial coherence lost in spatial texture information.
Embodiment 2
Tested using Indian agricultural high-spectral data collection.Wherein, Indian agricultural come from spectrometer (Airborne
Visible Infrared Imaging Spectrometer), it is that in the state of Indiana northwestward, Indian agricultural were received in 1992
The high-spectrum remote sensing collected, there is 20 meters of spatial resolution, it includes 144 × 144 pixels, 220 wave bands, due to
The factor such as noise and water absorption removes 20 wave bands therein, remaining 200 wave bands, includes 16 kinds of vegetation, chooses all 16
In classification, 10% sample composition is randomly selected per class label training set, and remaining 90% be used as test set, specifically species not with
Number of samples is referring to table 1；
The Indian agricultural data images grouped data statistics of table 1
Using overall nicety of grading (Overall accuracy, OA), average nicety of grading (Average accuracy,
AA) and Kappa counts coefficient (Kappa statistic, Kappa) to weigh the precision of sorting algorithm, in order to avoid random
The generation of deviation, each experiment are repeated 10 times and record average result, and verification platform uses Matlab R2012b, E5800,
6GBRAM experiment porch.In order to verify superiority of the SGDS-SVM algorithms in hyperspectral classification, 3 kinds of methods have been used to be compared
Compared with.Method 1：Using SVM, and merge Radial basis kernel function and form RSVM；Method 2：EPF algorithms divide high spectrum image
Class, there are EPF-B-c and EPF-G-c；Method 3：SGDS-SVM.
The classifying quality of the high spectrum image of the present embodiment：
(1) OA of Indian agricultural data set is 96.95%, and 12-13 percentage points is higher by than RSVM entirety niceties of grading,
2-8 percentage points is higher by than EPF algorithm, Kappa coefficients and AA are also greatly improved in addition, fully demonstrate SGDS-SVM
The validity of algorithm, inventive method are substantially better than other two methods.
(2) in order to verify influence of the monitoring data to inventive method, the classification of different training sample testing algorithms is selected
Precision.Indian woods data set overall classification accuracy OA is when training sample is 3% with regard to that can reach 90%；Illustrate that SGDS-SVM is calculated
Method can also obtain preferable nicety of grading in the case of supervision sample on a small quantity.
Embodiment 3
Tested using Salinas mountain valley high-spectral data collection, wherein Salinas mountain valley：From spectrometer
(Airborne Visible Infrared Imaging Spectrometer), is to add within 1992 Li Fuliya states Sa in the U.S.
The image that Li Nasi mountain valleys are collected into, there is 3.7 meters of spatial resolution, it includes 512 × 217 pixels, 224 wave bands,
Because the factor such as noise and water absorption removes 20 wave bands therein, remaining 204 wave bands, comprising 16 kinds of vegetation, choose all
In 16 classifications, 1% sample composition is randomly selected per class label training set, and remaining 99% is used as test set, specifically species
Not and number of samples is referring to table 2；
The Salinas mountain valley data images grouped data statistics of table 2
The present embodiment is using overall nicety of grading (Overall accuracy, OA), average nicety of grading (Average
Accuracy, AA) and Kappa statistics coefficients (Kappa statistic, Kappa) weigh the precision of sorting algorithm, in order to
The generation of random deviation is avoided, each experiment is repeated 10 times and records average result, and verification platform uses Matlab R2012b,
E5800,6GBRAM experiment porch.In order to verify the high spectrum image point for combining Steerable filter and linear space correlation information
Class algorithm (SGDS-SVM algorithms) has used 3 kinds of methods to be compared in the superiority of hyperspectral classification.Method 1：Using SVM,
And merge Radial basis kernel function and form RSVM；Method 2：EPF algorithms are classified to high spectrum image, there is EPF-B-c and EPF-
G-c；Method 3：SGDS-SVM.
The classifying quality of the high spectrum image of the present embodiment：
(1) OA of Salinas mountain valley data set is 98.69%, and 12-13 percentage is higher by than RSVM entirety niceties of grading
Point, 2-8 percentage points is higher by than EPF algorithm, Kappa coefficients and AA are also greatly improved in addition, fully demonstrate SGDS-
The validity of SVM algorithm, inventive method are substantially better than other two methods.
(2) in order to verify influence of the monitoring data to inventive method, the classification of different training sample testing algorithms is selected
Precision, Salinas mountain valley data set overall classification accuracy OA have just exceeded 99% when training sample is 2%, have illustrated SGDS-
SVM algorithm can also obtain preferable nicety of grading in the case of supervision sample on a small quantity.
Same or analogous label corresponds to same or analogous part；
Position relationship is used for being given for example only property explanation described in accompanying drawing, it is impossible to is interpreted as the limitation to this patent；
Obviously, the above embodiment of the present invention is only intended to clearly illustrate example of the present invention, and is not pair
The restriction of embodiments of the present invention.For those of ordinary skill in the field, may be used also on the basis of the above description
To make other changes in different forms.There is no necessity and possibility to exhaust all the enbodiments.It is all this
All any modification, equivalent and improvement made within the spirit and principle of invention etc., should be included in the claims in the present invention
Protection domain within.
Claims (7)
- A kind of 1. hyperspectral image classification method based on Steerable filter and linear space correlation information, it is characterised in that bag Include：Receive hyperspectral image data collection；Spatial texture information is obtained according to the hyperspectral image data collection；Linear space correlation information is obtained according to the hyperspectral image data collection；By the hyperspectral image data collection, spatial texture information and linear space correlation information linear fusion, newly counted According to collection；Training set is picked out from the new data set with preset ratio at random, the remainder of the new data set is as test Collection；The vector machine supported using RBF is trained to the training set, obtains training pattern；The vector machine supported using RBF is classified to the test set, obtains the classification knot of the high spectrum image Fruit.
- 2. the classification hyperspectral imagery side according to claim 1 based on Steerable filter and linear space correlation information Method, it is characterised in that described the step of obtaining spatial texture information according to the hyperspectral image data collection includes：High-spectral data collection is handled by Feature Dimension Reduction, obtains the hyperspectral image data collection of information content redistribution, And 20 compositions are filtered before the hyperspectral image data collection redistributed using guiding filtering to information content, obtain space Texture information.
- 3. the classification hyperspectral imagery side according to claim 1 based on Steerable filter and linear space correlation information Method, it is characterised in that described the step of obtaining linear space correlation information according to the hyperspectral image data collection includes：First EO-1 hyperion linear space correlation information matrix and the second EO-1 hyperion are defined according to the hyperspectral image data collection Linear space correlation information matrix, wherein the first EO-1 hyperion linear space correlation information matrix is the linear sky of horizontal direction Between correlation information matrix, the second EO-1 hyperion linear space correlation information matrix for vertical direction linear space correlation believe Cease matrix；First EO-1 hyperion linear space correlation information matrix and the second EO-1 hyperion linear space correlation information matrix are added, Obtain linear space correlation information.
- 4. the classification hyperspectral imagery side according to claim 3 based on Steerable filter and linear space correlation information Method, it is characterised in that described that first EO-1 hyperion linear space correlation information square is defined according to the hyperspectral image data collection The step of battle array and the second EO-1 hyperion linear space correlation information matrix, includes：The spatial resolution of the high spectrum image is M × N, then the first EO-1 hyperion linear space correlation information matrix DlFor：<mrow> <msub> <mi>D</mi> <mi>l</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <mi>x</mi> <mo>-</mo> <mi>M</mi> <mo>/</mo> <mn>2</mn> </mrow> <mrow> <mi>M</mi> <mo>/</mo> <mn>2</mn> </mrow> </mfrac> <mo>,</mo> <mi>x</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mi>L</mi> <mo>,</mo> <mi>M</mi> <mo>,</mo> <mi>y</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mi>L</mi> <mo>,</mo> <mi>N</mi> </mrow>Second EO-1 hyperion linear space correlation information matrix DvFor：<mrow> <msub> <mi>D</mi> <mi>v</mi> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mi>y</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <mo>(</mo> <mi>y</mi> <mo>-</mo> <mi>N</mi> <mo>/</mo> <mn>2</mn> <mo>)</mo> </mrow> <mrow> <mi>N</mi> <mo>/</mo> <mn>2</mn> </mrow> </mfrac> <mo>,</mo> <mi>x</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mi>L</mi> <mo>,</mo> <mi>M</mi> <mo>,</mo> <mi>y</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mi>L</mi> <mo>,</mo> <mi>N</mi> <mo>;</mo> </mrow>Wherein, (x, y) be pixel in the position of high spectrum image, M/2 is the center of high spectrum image horizontal direction, N/2 For the center of vertical direction, Dl(x, y) is normalization apart from size of each pixel apart from horizontal direction center, Dv(x, y) be each pixel in abscissa apart from the normalization apart from size of horizontal direction center.
- 5. the classification hyperspectral imagery side according to claim 1 based on Steerable filter and linear space correlation information Method, it is characterised in that it is described to pick out training set from the new data set with preset ratio at random, the new data set remaining Also include after the step of part is used as test set：The vector machine method cross validation supported using RBF, finds optimal parameter combination.
- 6. the classification hyperspectral imagery side according to claim 1 based on Steerable filter and linear space correlation information Method, it is characterised in that the preset ratio is 10% ratio.
- 7. the classification hyperspectral imagery side according to claim 1 based on Steerable filter and linear space correlation information Method, it is characterised in that described that the hyperspectral image data collection, spatial texture information and linear space correlation information is linear The step of merging, obtaining new data set includes：The hyperspectral image data collection, spatial texture information are added with linear space correlation information three, newly counted According to collection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710198805.5A CN107527001B (en) | 2017-03-29 | 2017-03-29 | Hyperspectral image classification method based on guided filtering and linear spatial correlation information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710198805.5A CN107527001B (en) | 2017-03-29 | 2017-03-29 | Hyperspectral image classification method based on guided filtering and linear spatial correlation information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107527001A true CN107527001A (en) | 2017-12-29 |
| CN107527001B CN107527001B (en) | 2020-12-22 |
Family
ID=60748629
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710198805.5A Active CN107527001B (en) | 2017-03-29 | 2017-03-29 | Hyperspectral image classification method based on guided filtering and linear spatial correlation information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107527001B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107066959A (en) * | 2017-03-29 | 2017-08-18 | 广东交通职业技术学院 | A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information |
| CN107862330A (en) * | 2017-10-31 | 2018-03-30 | 广东交通职业技术学院 | A kind of hyperspectral image classification method of combination Steerable filter and maximum probability |
| CN108446582A (en) * | 2018-01-25 | 2018-08-24 | 西安电子科技大学 | Hyperspectral image classification method based on textural characteristics and affine propagation clustering algorithm |
| CN111368940A (en) * | 2020-03-25 | 2020-07-03 | 西北工业大学 | A Hyperspectral Image Classification Method Based on Guided Filtering and Kernel Extreme Learning Machine |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101883274A (en) * | 2009-05-08 | 2010-11-10 | 中国科学院沈阳自动化研究所 | A Hyperspectral Image Lossless Compression Method Based on Joint Prediction of Space and Spectrum |
| CN103903007A (en) * | 2014-03-10 | 2014-07-02 | 哈尔滨工程大学 | Hyperspectral semi-supervised classification method based on space-spectral information |
| CN104036289A (en) * | 2014-06-05 | 2014-09-10 | 哈尔滨工程大学 | Hyperspectral image classification method based on spatial and spectral features and sparse representation |
| US20150023553A1 (en) * | 2013-07-18 | 2015-01-22 | U.S. Army Research Laboratory Attn: Rdrl-Loc-I | Image anomaly detection in a target area using polarimetric sensor data polarimetric sensor data |
| CN105654098A (en) * | 2016-03-23 | 2016-06-08 | 深圳大学 | Hyperspectral remote sensing image sorting method and system thereof |
| US9495618B1 (en) * | 2015-08-06 | 2016-11-15 | Digitalglobe, Inc. | Object detection with textural to spectral domain adaptation |
| CN106339674A (en) * | 2016-08-17 | 2017-01-18 | 中国地质大学（武汉） | Hyperspectral image classification method based on edge preservation and graph cut model |
| CN106408034A (en) * | 2016-10-17 | 2017-02-15 | 大连海事大学 | Hyperspectral image classification method based on spatial feature iteration |
| CN106503739A (en) * | 2016-10-31 | 2017-03-15 | 中国地质大学（武汉） | The target in hyperspectral remotely sensed image svm classifier method and system of combined spectral and textural characteristics |
| US20170083747A1 (en) * | 2015-09-21 | 2017-03-23 | The Climate Corporation | Ponding water detection on satellite imagery |
| CN107066959A (en) * | 2017-03-29 | 2017-08-18 | 广东交通职业技术学院 | A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information |
-
2017
- 2017-03-29 CN CN201710198805.5A patent/CN107527001B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101883274A (en) * | 2009-05-08 | 2010-11-10 | 中国科学院沈阳自动化研究所 | A Hyperspectral Image Lossless Compression Method Based on Joint Prediction of Space and Spectrum |
| US20150023553A1 (en) * | 2013-07-18 | 2015-01-22 | U.S. Army Research Laboratory Attn: Rdrl-Loc-I | Image anomaly detection in a target area using polarimetric sensor data polarimetric sensor data |
| CN103903007A (en) * | 2014-03-10 | 2014-07-02 | 哈尔滨工程大学 | Hyperspectral semi-supervised classification method based on space-spectral information |
| CN104036289A (en) * | 2014-06-05 | 2014-09-10 | 哈尔滨工程大学 | Hyperspectral image classification method based on spatial and spectral features and sparse representation |
| US9495618B1 (en) * | 2015-08-06 | 2016-11-15 | Digitalglobe, Inc. | Object detection with textural to spectral domain adaptation |
| US20170083747A1 (en) * | 2015-09-21 | 2017-03-23 | The Climate Corporation | Ponding water detection on satellite imagery |
| CN105654098A (en) * | 2016-03-23 | 2016-06-08 | 深圳大学 | Hyperspectral remote sensing image sorting method and system thereof |
| CN106339674A (en) * | 2016-08-17 | 2017-01-18 | 中国地质大学（武汉） | Hyperspectral image classification method based on edge preservation and graph cut model |
| CN106408034A (en) * | 2016-10-17 | 2017-02-15 | 大连海事大学 | Hyperspectral image classification method based on spatial feature iteration |
| CN106503739A (en) * | 2016-10-31 | 2017-03-15 | 中国地质大学（武汉） | The target in hyperspectral remotely sensed image svm classifier method and system of combined spectral and textural characteristics |
| CN107066959A (en) * | 2017-03-29 | 2017-08-18 | 广东交通职业技术学院 | A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information |
Non-Patent Citations (3)
| Title |
|---|
| LIZHE WANG 等: "Spectral–spatial multi-feature-based deep learning for hyperspectral remote sensing image classification", 《SOFT COMPUTING》 * |
| 倪鼎 等: "基于近邻协同的高光谱图像谱-空联合分类", 《自动化学报》 * |
| 王立国 等: "结合空-谱信息的高光谱图像分类方法", 《黑龙江大学自然科学学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107066959A (en) * | 2017-03-29 | 2017-08-18 | 广东交通职业技术学院 | A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information |
| CN107066959B (en) * | 2017-03-29 | 2020-12-18 | 广东交通职业技术学院 | Hyperspectral image classification method based on guided filtering and linear spatial correlation information |
| CN107862330A (en) * | 2017-10-31 | 2018-03-30 | 广东交通职业技术学院 | A kind of hyperspectral image classification method of combination Steerable filter and maximum probability |
| CN108446582A (en) * | 2018-01-25 | 2018-08-24 | 西安电子科技大学 | Hyperspectral image classification method based on textural characteristics and affine propagation clustering algorithm |
| CN111368940A (en) * | 2020-03-25 | 2020-07-03 | 西北工业大学 | A Hyperspectral Image Classification Method Based on Guided Filtering and Kernel Extreme Learning Machine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107527001B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107066959A (en) | A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information | |
| He et al. | Detection of fake images via the ensemble of deep representations from multi color spaces | |
| WO2021184891A1 (en) | Remotely-sensed image-based terrain classification method, and system | |
| CN108122008B (en) | SAR image recognition method based on sparse representation and multi-feature decision-level fusion | |
| CN107292343A (en) | A kind of Classification of hyperspectral remote sensing image method based on six layers of convolutional neural networks and spectral space information consolidation | |
| CN105046224A (en) | Block self-adaptive weighted histogram of orientation gradient feature based face recognition method | |
| CN105320962B (en) | A kind of road surface breakage kind identification method based on combining classifiers | |
| CN109492593A (en) | Hyperspectral image classification method based on principal component analysis network and space coordinate | |
| CN107871132B (en) | Hyperspectral image classification method for spatial feature adaptive optimization | |
| CN108399355A (en) | A kind of hyperspectral image classification method that spatial information adaptively merges | |
| CN107909109A (en) | SAR image sorting technique based on conspicuousness and multiple dimensioned depth network model | |
| CN107527001A (en) | A kind of hyperspectral image classification method based on Steerable filter and linear space correlation information | |
| CN107358260A (en) | A kind of Classification of Multispectral Images method based on surface wave CNN | |
| CN114781548B (en) | Image scene classification method, device, equipment and storage medium | |
| CN114596274B (en) | Citrus Huanglongbing detection method in natural background based on improved Cascade RCNN network | |
| CN113850284B (en) | A multi-operation detection method based on multi-scale feature fusion and multi-branch prediction | |
| CN113392937A (en) | 3D point cloud data classification method and related device thereof | |
| CN117333713A (en) | Space spectrum combined hyperspectral image classification method based on channel and spatial attention | |
| Pham et al. | Feature profiles from attribute filtering for classification of remote sensing images | |
| CN109598284A (en) | A kind of hyperspectral image classification method based on large-spacing distribution and space characteristics | |
| CN117830787A (en) | Signal Search Method Based on Wavelet Convolution and Former Multi-Scale Feature Fusion | |
| CN110070137A (en) | A kind of hyperspectral image classification method based on adaptive manifold filtering and domain transfer standard convolutional filtering | |
| CN117876760A (en) | A self-supervised hyperspectral image classification method based on redundancy elimination | |
| CN107180419B (en) | A Median Filter Detection Method Based on PCA Network | |
| CN113657199B (en) | Hyperspectral image anomaly detection method based on space-spectrum extraction |
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 |