AU2020103887A4 - A method for automated endmember identification, selection and extraction from hyperspectral imagery - Google Patents
A method for automated endmember identification, selection and extraction from hyperspectral imagery Download PDFInfo
- Publication number
- AU2020103887A4 AU2020103887A4 AU2020103887A AU2020103887A AU2020103887A4 AU 2020103887 A4 AU2020103887 A4 AU 2020103887A4 AU 2020103887 A AU2020103887 A AU 2020103887A AU 2020103887 A AU2020103887 A AU 2020103887A AU 2020103887 A4 AU2020103887 A4 AU 2020103887A4
- Authority
- AU
- Australia
- Prior art keywords
- endmember
- selection
- identification
- extraction
- automated
- 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.)
- Ceased
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000000605 extraction Methods 0.000 title claims abstract description 30
- 230000003595 spectral effect Effects 0.000 claims abstract description 37
- 238000004458 analytical method Methods 0.000 claims abstract description 20
- 238000011524 similarity measure Methods 0.000 claims abstract description 19
- 239000011159 matrix material Substances 0.000 claims abstract description 12
- 238000010586 diagram Methods 0.000 claims description 13
- 230000000717 retained effect Effects 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 18
- 230000009471 action Effects 0.000 abstract description 2
- 239000000463 material Substances 0.000 description 8
- 230000004048 modification Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000002411 adverse Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000003466 anti-cipated effect Effects 0.000 description 1
- 230000003796 beauty Effects 0.000 description 1
- 238000000701 chemical imaging Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000004611 spectroscopical analysis Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/2823—Imaging spectrometer
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23211—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with adaptive number of clusters
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/2823—Imaging spectrometer
- G01J2003/2826—Multispectral imaging, e.g. filter imaging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/58—Extraction of image or video features relating to hyperspectral data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
Abstract
A METHOD FOR AUTOMATED ENDMEMBER IDENTIFICATION,
SELECTION AND EXTRACTION FROM HYPERSPECTRAL IMAGERY
A method for automated endmember identification, selection and extraction from
5 hyperspectral imagery is disclosed in the present invention. The input 3-dimensional image
cube (10) is encoded into 2-dimensional matrix (20). This encoded sample matrix (20) is
used for further processing and analysis and every single data sample of (20) is evaluated
for Pearson Correlation based purity measure. Data samples from (30) having strong positive
correlation (50) are clustered into automatically into similar groups using distance based
10 similarity measure. Further, all the cluster elements are sorted in descending order based of
MSE values. The first few data samples having minimum MSE value from each of the
cluster are selected as a final set of endmembers (80). The selected set of final endmembers
(80) in decoded into its original input shape (10), so that the final process outcome can be
generated into three different endmember formats including their corresponding spatial
15 coordinates (90), pure spectral signatures (100) and spectral plots (110).
Figure 7 shall be the reference figure.
Application No:
Applicant Name: Total No of Sheets:4
Page4of4
10
Input of 3D image
20
Data Encoding
30
FPositive Correlation Based Purity Measure
50
Strong Positive Correlation
Basd Sample Se action
Distance Metric Based Similarity Measure
80
Spatial Ne ighborhood based
endmember selection
90,100,110
Data D etoddninge
output)
Figure 7 Complete hyperspectral endmember identification, selection and extraction process.
Description
Application No: Applicant Name: Total No of Sheets:4 Page4of4
10
Input of 3D image 20
Data Encoding
30
FPositive Correlation Based Purity Measure
50
Strong Positive Correlation Basd Sample Se action
Distance Metric Based Similarity Measure
80
Spatial Ne ighborhood based endmember selection
90,100,110
Data D etoddninge
output)
Figure 7 Complete hyperspectral endmember identification, selection and extraction process.
Australian Government IP Australia INNOVATION PATENT APPLICATION AUSTRALIA PATENT OFFICE
1. TITLE OF THE INVENTION
2. APPLICANTS (S) NAME NATIONALITY ADDRESS IN DEPARTMENT OF COMPUTER
SCIENCE AND IT, DR BABASAHEB 1. Dr. Karbhari AMBEDKAR MARATHWADA
Vishwanath kale UNIVERSITY, AURANGABAD
431004 (MS)-INDIA. 3. PREAMBLE TO THE DESCRIPTION
The following specification particularly describes the invention and the manner in which it is to be performed
[0001] The present invention generally relates to the Hyperspectral image analysis, and more specifically to the process of automatic endmember identification, selection and extraction from Hyperspectral imagery.
[0002] Background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication.
[0003] The term 'Hyperspectral Imaging' integrates the power of digital imaging and spectroscopy, and it is significantly accepted because of its wider spectral and large geographic coverage. Due to the wider spectral coverage and continuous acquisition nature, the Hyperspectral sensor records the light intensity for a large number of continuous spectral channels. Therefore, every pixel in the image contains a continuous spectrum of radiance or reflectance and can be used to characterize the materials within the scene with great precision and details. Along with the enriched amount of information, it has significantly proven its applicability for material identification, discrimination and mapping in various application domains including Agriculture, Forestry, Soil Sciences, Mineralogy, Hydrology and Oceanography etc. along with the beauty of its note-worthy and multi-domain applications, the Hyperspectral images are having its own data processing challenges, which adversely affect the data analysis outcomes. These challenges includes the adverse atmospheric effects, curse of dimensionality, lack of ground truth data and spectral mixing etc. the ground truth unavailability and spectral mixing problem can be precisely estimated using the image derived endmembers. Where 'endmembers' are the spectral signatures, exclusively characterized by unique surface material. In Hyperspectral image analysis, accurately identified endmembers can be used as reference spectra for supervised classification and spectral unmixing problems.
[00041 The Pixel Purity Index, NFINDR, Automatic Target Generation Process, Vertex Component Analysis, Convex Cone Analysis and Algorithm for Rapid Endmember Determination etc. are the few endmember extraction methods available in the literature. But, these methods are significantly dependent on user defined input parameters (i.e. Number of endmembers to extract, Number of iterations to perform and threshold value etc.) and unfortunately having no specific criteria to define the value of these parameters. Additionally some of these methods get initialized on the basis of randomly generated parameter values. This hypothetically selected parameter values and random initialization nature makes endmember extraction techniques to produce inconsistent set of endmembers during several runs on the same Hyperspectral images.
[0005] To overcome these challenges, an automated Hyperspectral endmember identification, selection and extraction process us designed, which minimizes the number of input parameters and produces the consistent and useful set of endmembers for hyperspectral image classification and spectral unmixing.
[0006] Therefore the present disclosure overcomes the above mentioned problem of lack of consistency in extraction associated with the traditionally available method or system, any of the above mentioned inventions can be used with the presented disclosed technique with or without modification.
[00071 All publications herein are incorporated by reference to the same extent as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. Where a definition or use of a term in an incorporated reference is inconsistent or contrary to the definition of that term provided herein, the definition of that term provided herein applies and the definition of that term in the reference does not apply.
[0008] As used in the description herein and throughout the claims that follow, the meaning of "a "an," and "the" includes plural reference unless the context clearly dictates otherwise. Also, as used in the description herein, the meaning of "in" includes "in" and "on" unless the context clearly dictates otherwise.
[00091 All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g. "such as") provided with respect to certain embodiments herein is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention otherwise claimed. No language in the specification should be construed as indicating any non-claimed element essential to the practice of the invention.
[0010] It is an object of the present disclosure to provide a method to process images obtained from long distance to identify the contents of the image.
[00111 It is an object of the present invention to provide a method for automated end member identification from a hyperspectral image.
[00121 It is yet another object of the present invention to obtain spatial coordinates, spectral plots and pure spectral signatures from a hyperspectral image.
[00131 The present disclosure relates to a traffic surveillance systems, and more particularly to a system and method to identify the vehicles on the road and determine if the vehicle is moving or stationary.
[00141 One should appreciate that although the present disclosure has been explained with respect to a defined set of functional modules, any other module or set of modules can be added/deleted/modified/combined and any such changes in architecture/construction of the proposed system are completely within the scope of the present disclosure. Each module can also be fragmented into one or more functional sub-modules, all of which also completely within the scope of the present disclosure.
[0015] This invention encompasses the six step analysis process, wherein the first and last step includes the input data encoding (DE) and decoding (DD) processes respectively. The seconds step evaluates every data sample using Pearson Correlation based Purity Measure (PCB-PM). Third step uses the Strong Positive Correlation based Sample Selection (SPC-SS) using efficient searching mechanism. The remaining Pearson Correlation Based Eliminated Data Samples (PCB-EDS) are not considered for further analysis. In step four, the only data samples retained in the step three (SPC-SS) are taken for analysis and clustered automatically using Distance Metrics Based Similarity Measure (DMB-SM). Here the cardinality of cluster set gives the Virtual Dimensionality Count (i.e. total number of unique materials) of the image under observation. In step five, using the Spatial Neighborhood Based Automatic Endmember Selection (SNB-AES), the representation data samples having maximum purity are selected from each cluster and marked as a final set of endmembers. The process extracts the final set of endmembers in three different formats including their corresponding spatial coordinates, spectral signatures and spectral plots.
[0016] As per the method of the present invention, the bad band eliminated radiance cube or atmospherically corrected reflectance cube or dimensionality reduced few of the principal components of hyperspectral image is taken as an input for endmember identification, selection and extraction process. Further, the input 3-dimensional image cube is encoded into 2-dimensional matrix, where each row represents a single data sample and each column represents a single variable. This encoded sample matrix is used for further processing and analysis and every single data sample of is evaluated for Pearson Correlation based purity measure. Wherein, sample to sample correlation is computed for all the data samples. Only data samples from having strong positive correlation are retained for further processing and rest of the samples eliminated and not taken in consideration for further analysis. The data samples retained in strong positive correlation based purity measure are clustered into automatically into similar groups using distance based similarity measure. Here the total number of clusters (i.e. cardinality of the cluster set) indicates the Virtual Dimensionality Count (i.e. the total number of unique surface materials within the scene) of the image. The cluster set is used for 8 or 24 spatial neighbourhood based similarity measure which automates the final endmember selection process. Where Mean Squared Error (MSE) is computed between every individual data sample of the image and mean of its 8 or 24 neighbouring samples. Further, all the cluster elements are sorted in descending order based of MSE values. Now the first few data samples having minimum MSE value from each of the cluster are selected as a final set of endmembers. Here, the remaining data samples from all the clusters are eliminated from further analysis. The selected set of final endmembers in decoded into its original input shape, so that the final process outcome can be generated into three different endmember formats including their corresponding spatial coordinates, pure spectral signatures and spectral plots.
[00171 Various objects, features, aspects and advantages of the inventive subject matter will become more apparent from the following detailed description of preferred embodiments, along with the accompanying drawing figures in which like numerals represent like components.
[0018] The accompanying drawings are included to provide a further understanding of the present disclosure, and are incorporated in and constitute a part of this specification. The drawings illustrate exemplary embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.
[0019] Some embodiments of the invention are described herein with reference to the supplementary figures. The description together with the figures, make apparent to the person having ordinary skills in the art how some embodiments may be practiced. The figures provided here are only for the purpose of illustrative description and no attempt is made to show the structural details of an embodiments in more details than is necessary for fundamental understanding of the invention. For the sake of clarity, some of the objects depicted in the figures are not to be scale.
[0020] Figure 1 is block diagram of the input data enrolment and encoding the same for further analysis.
[0021] Figure 2 is a block diagram of the Pearson Correlation based spectral purity measure for all data samples.
[0022] Figure 3 is a block diagram of strong positive correlation based sample searching mechanism.
[0023] Figure 4 is a block diagram of distance metric based spectral similarity measure and clustering of the data samples retained after strong positive correlation based sample searching.
[0024] Figure 5 is a block diagram of an automated selection of pure spectral signatures from each of the clusters generated from the distance metric based spectral similarity measures.
[00251 Figure 6 is a block diagram representing the steps of decoding of the data to identify the input images, end members and the spatial coordinates and spectral plots.
[0026] Figure 7 is a block diagram for complete hyperspectral endmember identification, selection and extraction process.
[00271 In the following description, numerous specific details are set forth in order to provide a thorough understanding of embodiments of the present invention. It will be apparent to one skilled in the art that embodiments of the present invention may be practiced without some of these specific details.
[0028] The following is a detailed description of embodiments of the disclosure depicted in the accompanying drawings. The embodiments are in such detail as to clearly communicate the disclosure. However, the amount of detail offered is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosure as defined by the appended claims.
[0029] Exemplary embodiments will now be described more fully hereinafter with reference to the accompanying drawings, in which exemplary embodiments are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. These embodiments are provided so that this disclosure will be thorough and complete and will fully convey the scope of the invention to those of ordinary skill in the art. Moreover, all statements herein reciting embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future (i.e., any elements developed that perform the same function, regardless of structure).
[0030] Each of the appended claims defines a separate invention, which for infringement purposes is recognized as including equivalents to the various elements or limitations specified in the claims. Depending on the context, all references below to the "invention" may in some cases refer to certain specific embodiments only. In other cases it will be recognized that references to the "invention" will refer to subject matter recited in one or more, but not necessarily all, of the claims.
[00311 Various terms as used herein are shown below. To the extent a term used in a claim is not defined below, it should be given the broadest definition persons in the pertinent art have given that term as reflected in printed publications and issued patents at the time of filing. Although specific terms are used in the following description for sake of clarity, these terms are intended to refer only to particular structure of the invention for illustration in the drawings, and are not intended to define or limit the scope of invention.
[0032] This invention encompasses the six step analysis process, wherein the first and last step includes the input data encoding (DE) and decoding (DD) processes respectively. The seconds step evaluates every data sample using Pearson Correlation based Purity Measure (PCB-PM). Third step uses the Strong Positive Correlation based Sample Selection (SPC-SS) using efficient searching mechanism. The remaining Pearson Correlation Based Eliminated Data Samples (PCB-EDS) are not considered for further analysis. In step four, the only data samples retained in the step three (SPC-SS) are taken for analysis and clustered automatically using Distance Metrics Based Similarity Measure (DMB-SM). Here the cardinality of cluster set gives the Virtual Dimensionality Count (i.e. total number of unique materials) of the image under observation. In step five, using the Spatial Neighbourhoods Based Automatic Endmember Selection (SNB-AES), the representation data samples having maximum purity are selected from each cluster and marked as a final set of endmembers. The process extracts the final set of endmembers in three different formats including their corresponding spatial coordinates, spectral signatures and spectral plots.
[0033] Referring figure 1, the bad band eliminated radiance cube or atmospherically corrected reflectance cube or dimensionality reduced few of the principal components of hyperspectral image is taken as an input (10) for endmember identification, selection and extraction process. Further, the input 3-dimensional image cube (10) is encoded into 2 dimensional matrix (20), where each row represents a single data sample and each column represents a single variable. This encoded sample matrix (20) is used for further processing and analysis.
[0034] In a preferred embodiment, the invention discloses a method for automated endmember identification, selection and extraction from hyperspectral imagery wherein, inputting, an atmospherically corrected reflectance 3 dimensional image (10); encoding, into
Q a two dimensional matrix (20) said input three dimensional image (10); evaluating, Pearson Correlation based purity measure of a one or more said two dimensional matrix (20) to obtain one or more sample correlation (30); retaining, strong positive correlation (50) after searching out of the sample correlations (30); clustering, into similar cluster sets (60), the obtained strong positive correlation samples (50) using distance based similarity measures; identifying a virtual dimensionality count from the total number of clusters; computing mean squared errors between each of the obtained data samples and mean of its 8 or 24 spatial neighboring samples; selecting the final set of endmembers (80) from the data samples (70) having the lowest Mean squared error vales. decoding the final set of endmembers (80) into input shape (10) to obtain corresponding spatial coordinates (90), pre spectral signature (100) and spectral plots (110).
[0035] Referring figure 2, every single data sample of (20) is evaluated for Pearson Correlation based purity measure. Wherein, sample to sample correlation (30) is computed for all the data samples. In an aspect of the present invention, the method for automated endmember identification, selection and extraction from hyperspectral imagery wherein, the cardinality of the total number of clusters formulated using the distance metric based similarity measure are termed as the virtual dimensionality count.
[0036] Referring figure 3, using efficient searching mechanism, only data samples from (30) having strong positive correlation (50) are retained for further processing and rest of the samples (40) eliminated and not taken in consideration for further analysis. In an aspect of the present invention, the method for automated endmember identification, selection and extraction from hyperspectral imagery wherein, the established virtual dimensionality count helps define the number of endmembers to be extracted from the hyperspectral imagery.
[00371 Referring figure 4, the data samples (50) retained in strong positive correlation based purity measure are clustered into automatically into similar groups using distance based similarity measure. Here the total number of clusters (i.e. cardinality of the cluster set) indicates the Virtual Dimensionality Count (i.e. the total number of unique surface materials within the scene) of the image.
[0038] Referring figure 5, the cluster set (60) is used for 8 or 24 spatial neighborhood based similarity measure which automates the final endmember selection process. Where Mean Squared Error (MSE) is computed between every individual data sample of the image and mean of its 8 or 24 neighboring samples. Further, all the cluster elements are sorted in descending order based of MSE values. Now the first few data samples having minimum MSE value from each of the cluster are selected as a final set of endmembers (80). Here, the remaining data samples (70) from all the clusters are eliminated from further analysis.
[00391 Referring figure 6, the automatically selected set of final endmembers (80) in decoded into its original input shape (10), so that the final process outcome can be generated into three different endmember formats including their corresponding spatial coordinates (90), pure spectral signatures (100) and spectral plots (110).
[0040] Referring to figure 7, which is a flowchart illustrating the steps involved in the entire process, the bad band eliminated radiance cube or atmospherically corrected reflectance cube or dimensionality reduced few of the principal components of hyperspectral image is taken as an input (10) for endmember identification, selection and extraction process. Further, the input 3-dimensional image cube (10) is encoded into 2-dimensional matrix (20), where each row represents a single data sample and each column represents a single variable. This encoded sample matrix (20) is used for further processing and analysis and every single data sample of (20) is evaluated for Pearson Correlation based purity measure. Wherein, sample to sample correlation (30) is computed for all the data samples. Only data samples from (30) having strong positive correlation (50) are retained for further processing and rest of the samples (40) eliminated and not taken in consideration for further analysis. The data samples (50) retained in strong positive correlation based purity measure are clustered into automatically into similar groups using distance based similarity measure. Here the total number of clusters (i.e. cardinality of the cluster set) indicates the Virtual Dimensionality Count (i.e. the total number of unique surface materials within the scene) of the image. The cluster set (60) is used for 8 or 24 spatial neighbourhood based similarity measure which automates the final endmember selection process. Where Mean Squared Error (MSE) is computed between every individual data sample of the image and mean of its 8 or 24 neighboring samples. Further, all the cluster elements are sorted in descending order based of MSE values. Now the first few data samples having minimum MSE value from each of the cluster are selected as a final set of endmembers (80). Here, the remaining data samples (70) from all the clusters are eliminated from further analysis. The selected set of final endmembers (80) in decoded into its original input shape (10), so that the final process outcome can be generated into three different endmember formats including their
1i1 corresponding spatial coordinates (90), pure spectral signatures (100) and spectral plots (110).
[0041] This automatically identified, selected and extracted set of endmembers can be used as a reference for supervised classification and spectral unmixing of hyperspectral images. In yet another aspect of the present invention, the method for automated endmember identification, selection and extraction from hyperspectral imagery wherein, the virtual dimensionality count defines the number of principal components to be retained after dimensionality reduction.
[0042] In yet another aspect of the present invention, the method for automated endmember identification, selection and extraction from hyperspectral imagery wherein, the virtual dimensionality count defies the number of clusters in an unsupervised classification. In yet another aspect of the present invention, the method for automated endmember identification, selection and extraction from hyperspectral imagery wherein, the final set of end members are selected automatically.
[0043] Advantages of the invention: This process has application independent usage. No additional input parameters, other than the hyperspectral image itself are required for endmember identification, selection and extraction.
[0044] Any individual user having no domain expertise can use this method for their own application. The automatically estimated Virtual Dimensionality Count will help to define the number of endmembers to extracted from hyperspectral imagery or number of principal components to retain after dimensionality reduction or number of classes for unsupervised classification.
[0045] The foregoing descriptions of the specific embodiments of the present invention have been presented for purpose of illustration and description. They are not intended to be exhaustive or to limit the present invention to the precise forms disclosed, and obviously many modifications and variations are possible in light of the above teaching.
[0046] The embodiments were chosen and described in order to best explain the principles of present invention and its practical applications, to thereby enable others, skilled in the art to best utilize the present invention and various embodiments with various modifications as are suited to the particular use contemplated. It is understood that various omission and substitution of equivalents are contemplated as circumstance may suggest or
1 1 render expedient, but such are intended to cover the application or implementation without departing from the spirit or scope of the present invention.
[00471 While the foregoing describes various embodiments of the invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof. The scope of the invention is determined by the claims that follow. The invention is not limited to the described embodiments, versions or examples, which are included to enable a person having ordinary skill in the art to make and use the invention when combined with information and knowledge available to the person having ordinary skill in the art.
Dated this 01" of December 2020:
1 )
Claims (6)
1. A method for automated endmember identification, selection and extraction from hyperspectral imagery wherein,
inputting, an atmospherically corrected reflectance 3 dimensional image (10);
encoding, into a two dimensional matrix (20) said input three dimensional image (10);
evaluating, Pearson Correlation based purity measure of a one or more said two dimensional matrix (20) to obtain one or more sample correlation (30);
retaining, strong positive correlation (50) after searching out of the sample correlations (30);
identifying a virtual dimensionality count from the total number of clusters;
clustering, into similar cluster sets (60), the obtained strong positive correlation samples (50) using distance based similarity measures;
computing mean squared errors between each of the obtained data samples and mean of its 8 or 24 spatial neighboring samples;
selecting the final set of endmembers (80) from the data samples (70) having the lowest Mean squared error vales.
decoding the final set of endmembers (80) into input shape (10) to obtain corresponding spatial coordinates (90), pre spectral signature (100) and spectral plots (110).
2. The method for automated endmember identification, selection and extraction from hyperspectral imagery as claimed in claim 1 wherein, the cardinality of the total number of clusters formulated using the distance metric based similarity measure are termed as the virtual dimensionality count.
3. The method for automated endmember identification, selection and extraction from hyperspectral imagery as claimed in claim 1 wherein, the established virtual
1'1 dimensionality count helps define the number of endmembers to be extracted from the hyperspectral imagery.
4. The method for automated endmember identification, selection and extraction from hyperspectral imagery as claimed in claim 1 wherein, the virtual dimensionality count defines the number of principal components to be retained after dimensionality reduction.
5. The method for automated endmember identification, selection and extraction from hyperspectral imagery as claimed in claim 1 wherein, the virtual dimensionality count defies the number of clusters in an unsupervised classification.
6. The method for automated endmember identification, selection and extraction from hyperspectral imagery as claimed in claim 1 wherein, the final set of end members are selected automatically.
Date:
1 A
Application No: Applicant Name: Total No of Sheets:4 Page 1 of 4 04 Dec 2020 2020103887
Figure 1. Block diagram of the input data enrolment and encoding the same for further analysis.
Figure 2. Block diagram of the Pearson Correlation based spectral purity measure for all data samples.
Application No: Applicant Name: Total No of Sheets:4 Page 2 of 4 04 Dec 2020 2020103887
Figure 3 Block diagram of strong positive correlation based sample searching mechanism.
Figure 4 Block diagram of distance metric based spectral similarity measure and clustering of the data samples retained after strong positive correlation based sample searching.
Application No: Applicant Name: Total No of Sheets:4 Page 3 of 4 04 Dec 2020 2020103887
Figure 5 Block diagram of an automated selection of pure spectral signatures from each of the clusters generated from the distance metric based spectral similarity measures.
Figure 6 Block diagram representing the steps of decoding of the data to identify the input images, end members and the spatial coordinates and spectral plots
Application No: Applicant Name: Total No of Sheets:4 Page 4 of 4 04 Dec 2020 2020103887
Figure 7 Complete hyperspectral endmember identification, selection and extraction process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2020103887A AU2020103887A4 (en) | 2020-12-04 | 2020-12-04 | A method for automated endmember identification, selection and extraction from hyperspectral imagery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2020103887A AU2020103887A4 (en) | 2020-12-04 | 2020-12-04 | A method for automated endmember identification, selection and extraction from hyperspectral imagery |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| AU2020103887A4 true AU2020103887A4 (en) | 2021-02-11 |
Family
ID=74502199
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| AU2020103887A Ceased AU2020103887A4 (en) | 2020-12-04 | 2020-12-04 | A method for automated endmember identification, selection and extraction from hyperspectral imagery |
Country Status (1)
| Country | Link |
|---|---|
| AU (1) | AU2020103887A4 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113191411A (en) * | 2021-04-22 | 2021-07-30 | 杭州卓智力创信息技术有限公司 | Electronic sound image file management method based on photo group |
| CN113361548A (en) * | 2021-07-05 | 2021-09-07 | 北京理工导航控制科技股份有限公司 | Local feature description and matching method for highlight image |
| CN114120020A (en) * | 2021-11-30 | 2022-03-01 | 哈尔滨工业大学 | Hyperspectral image inter-spectrum sequencing method based on key channel protection and spectral clustering |
| CN116563571A (en) * | 2023-05-16 | 2023-08-08 | 北京师范大学 | Boltzmann entropy similarity-based hyperspectral image band selection method and system |
-
2020
- 2020-12-04 AU AU2020103887A patent/AU2020103887A4/en not_active Ceased
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113191411A (en) * | 2021-04-22 | 2021-07-30 | 杭州卓智力创信息技术有限公司 | Electronic sound image file management method based on photo group |
| CN113191411B (en) * | 2021-04-22 | 2023-02-07 | 杭州卓智力创信息技术有限公司 | Electronic sound image file management method based on photo group |
| CN113361548A (en) * | 2021-07-05 | 2021-09-07 | 北京理工导航控制科技股份有限公司 | Local feature description and matching method for highlight image |
| CN113361548B (en) * | 2021-07-05 | 2023-11-14 | 北京理工导航控制科技股份有限公司 | Local feature description and matching method for highlight image |
| CN114120020A (en) * | 2021-11-30 | 2022-03-01 | 哈尔滨工业大学 | Hyperspectral image inter-spectrum sequencing method based on key channel protection and spectral clustering |
| CN114120020B (en) * | 2021-11-30 | 2024-04-26 | 哈尔滨工业大学 | Method for ordering hyperspectral images among spectrums based on key channel protection and spectrum clustering |
| CN116563571A (en) * | 2023-05-16 | 2023-08-08 | 北京师范大学 | Boltzmann entropy similarity-based hyperspectral image band selection method and system |
| CN116563571B (en) * | 2023-05-16 | 2023-11-21 | 北京师范大学 | Boltzmann entropy similarity-based hyperspectral image band selection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2020103887A4 (en) | A method for automated endmember identification, selection and extraction from hyperspectral imagery | |
| CN110443143B (en) | Multi-branch convolutional neural network fused remote sensing image scene classification method | |
| Marinelli et al. | A novel change detection method for multitemporal hyperspectral images based on binary hyperspectral change vectors | |
| Dev et al. | Color-based segmentation of sky/cloud images from ground-based cameras | |
| Beijbom et al. | Automated annotation of coral reef survey images | |
| Martin et al. | Region-based spatial preprocessing for endmember extraction and spectral unmixing | |
| Charles et al. | Learning sparse codes for hyperspectral imagery | |
| Dópido et al. | A quantitative and comparative assessment of unmixing-based feature extraction techniques for hyperspectral image classification | |
| Mnih | Machine learning for aerial image labeling | |
| Toler-Franklin et al. | Multi-feature matching of fresco fragments | |
| Yang et al. | Superpixel-based unsupervised band selection for classification of hyperspectral images | |
| Zare et al. | Hyperspectral band selection and endmember detection using sparsity promoting priors | |
| CN107992891B (en) | Multispectral remote sensing image change detection method based on spectral vector analysis | |
| Chang et al. | Component analysis-based unsupervised linear spectral mixture analysis for hyperspectral imagery | |
| Marchant et al. | Automated analysis of foraminifera fossil records by image classification using a convolutional neural network | |
| CN111353531B (en) | Hyperspectral image classification method based on singular value decomposition and spatial spectral domain attention mechanism | |
| Deng et al. | Cloud detection in satellite images based on natural scene statistics and gabor features | |
| Wei et al. | Adaboost-based feature relevance assessent in fusing LiDAR and image data for classification of trees and vehicles in urban scenes | |
| Rojas et al. | Comparison of support vector machine-based processing chains for hyperspectral image classification | |
| Merényi et al. | Empowering graph segmentation methods with SOMs and CONN similarity for clustering large and complex data | |
| WO2019129985A1 (en) | Method for forming a neural network for the recognition of a sequence of characters, and associated recognition method | |
| Canham et al. | High spatial resolution hyperspectral spatially adaptive endmember selection and spectral unmixing | |
| Benachir et al. | Modified independent component analysis for initializing non-negative matrix factorization: An approach to hyperspectral image unmixing | |
| CN106033545B (en) | Wave band selection method of determinant point process | |
| Boukir et al. | Classification of remote sensing data using margin-based ensemble methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FGI | Letters patent sealed or granted (innovation patent) | ||
| MK22 | Patent ceased section 143a(d), or expired - non payment of renewal fee or expiry |