Diferencia entre revisiones de «Endmember Induction Algorithms»

De Grupo de Inteligencia Computacional (GIC)
 
(No se muestran 16 ediciones intermedias de 2 usuarios)
Línea 2: Línea 2:


Download the latest Endmember Induction Algorithms (EIAs) toolbox and the documentation here:
Download the latest Endmember Induction Algorithms (EIAs) toolbox and the documentation here:
* MATLAB: [[Media:EIAs_matlab.zip | toolbox (xxx MB)]] | [[Media:EIAs_matlab_doc.zip | doc (xxx MB)]]
* EIA Toolbox:
* SCILAB: [[Media:EIAs_scilab.zip | toolbox (xxx MB)]] [[Media:EIAs_scilab_doc.zip | doc (xxx MB)]]
** Version 0.4: [[Media:EIA_Toolbox_reduced_v04.zip | download (1.3 MB)]]
** Version 0.3: [[Media:EIA_Toolbox_reduced_v03.zip | download (1.3 MB)]]
* Data:
** For Matlab and Scilab: [[Media:EIA_Toolbox_data.zip | download (259.2 MB)]]
** For Matlab only: [[Media:EIA_toolbox_data_matlab.zip | download (128.3 MB)]]
** For Scilab only: [[Media:EIA_toolbox_data_scilab.zip | download (130.9 MB)]]


This software is distributed under the terms of the [http://www.gnu.org/licenses/gpl.html GNU General Public License] as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This software is distributed under the terms of the [http://www.gnu.org/licenses/gpl.html GNU General Public License] as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.


The here available Endmember Induction Algorithms (EIAs) toolbox has been developed with [http://www.mathworks.com/ MATLAB 7.4] (licensed copy is needed to use it) and [http://www.scilab.org/ SCILAB 5.2] (open source and freely distributed).
The here available Endmember Induction Algorithms (EIAs) toolbox has been developed with [http://www.mathworks.com/ MATLAB 7.4] and [http://www.mathworks.com/ MATLAB 8.1 (R2013a)] (licensed copy is needed to use it) and [http://www.scilab.org/ SCILAB 5.2] (open source and freely distributed).


If you are using the Endmember Induction Algorithms (EIAs) toolbox for your scientific research, please reference it as follows:
If you are using the Endmember Induction Algorithms (EIAs) toolbox for your scientific research, please reference it as follows:
Línea 21: Línea 26:
=== Endmember Induction Algorithms collection ===
=== Endmember Induction Algorithms collection ===


Here you can find separately the EIAs included in the toolbox and their respective bibliographical references:
Here you can find separately the EIAs included in the toolbox's '''version 0.2''' and their respective bibliographical references:


* '''Endmember Induction Heuristic Algorithm (EIHA)'''
* '''Endmember Induction Heuristic Algorithm (EIHA)'''
** MATLAB: [[Media:EIA_EIHA.m | download (4.7 KB)]]
** MATLAB: [[Media:EIA_EIHA.m | download (4.7 KB)]]
** SCILAB: [[Media:EIA_EIHA.sci | download (4.5 KB)]]
** SCILAB: [[Media:EIA_EIHA.sci | download (4.6 KB)]]
     M. Grana, I. Villaverde, J. O. Maldonado, and C. Hernandez
     M. Grana, I. Villaverde, J. O. Maldonado, and C. Hernandez
     ''Two lattice computing approaches for the unsupervised segmentation of hyperspectral images''
     ''Two lattice computing approaches for the unsupervised segmentation of hyperspectral images''
Línea 32: Línea 37:
* '''Incremental Strong Lattice Independent Algorithm (ILSIA)'''
* '''Incremental Strong Lattice Independent Algorithm (ILSIA)'''
** MATLAB: [[Media:EIA_ILSIA.m | download (7.4 KB)]]
** MATLAB: [[Media:EIA_ILSIA.m | download (7.4 KB)]]
** SCILAB: [[Media:EIA_ILSIA.sci | download (7.1 KB)]]
** SCILAB: [[Media:EIA_ILSIA.sci | download (7.3 KB)]]
     M. Grana, D. Chyzhyk, M. García-Sebastián, and C. Hernández
     M. Grana, D. Chyzhyk, M. García-Sebastián, and C. Hernández
     ''Lattice independent component analysis for functional magnetic resonance imaging''
     ''Lattice independent component analysis for functional magnetic resonance imaging''
Línea 39: Línea 44:
* '''Prof. Ritter's WM Algorithm (WM)'''
* '''Prof. Ritter's WM Algorithm (WM)'''
** MATLAB: [[Media:EIA_WM.m | download (2.4 KB)]]
** MATLAB: [[Media:EIA_WM.m | download (2.4 KB)]]
** SCILAB: [[Media:EIA_WM.sci | download (2.3 KB)]]
** SCILAB: [[Media:EIA_WM.sci | download (2.5 KB)]]
     G. X. Ritter and G. Urcid
     G. X. Ritter and G. Urcid
     ''A lattice matrix method for hyperspectral image unmixing''
     ''A lattice matrix method for hyperspectral image unmixing''
Línea 46: Línea 51:
* '''N-FINDR'''
* '''N-FINDR'''
** MATLAB: [[Media:EIA_NFINDR.m | download (3.9 KB)]]
** MATLAB: [[Media:EIA_NFINDR.m | download (3.9 KB)]]
** SCILAB: [[Media:EIA_NFINDR.sci | download (3.8 KB)]]
** SCILAB: [[Media:EIA_NFINDR.sci | download (3.9 KB)]]
     Winter, M. E.
     Winter, M. E.
     ''N-FINDR: an algorithm for fast autonomous spectral endmember determination in hyperspectral data''
     ''N-FINDR: an algorithm for fast autonomous spectral endmember determination in hyperspectral data''
Línea 53: Línea 58:
* '''Fast Iterative PPI (FIPPI)'''
* '''Fast Iterative PPI (FIPPI)'''
** MATLAB: [[Media:EIA_FIPPI.m | download (3.7 KB)]]
** MATLAB: [[Media:EIA_FIPPI.m | download (3.7 KB)]]
** SCILAB: [[Media:EIA_FIPPI.sci | download (3.6 KB)]]
** SCILAB: [[Media:EIA_FIPPI.sci | download (3.7 KB)]]
     Chang, C.-I. and Plaza, A.
     Chang, C.-I. and Plaza, A.
     ''A fast iterative algorithm for implementation of pixel purity index''
     ''A fast iterative algorithm for implementation of pixel purity index''
Línea 60: Línea 65:
* '''Automatic Target Generation Process (ATGP)'''
* '''Automatic Target Generation Process (ATGP)'''
** MATLAB: [[Media:EIA_ATGP.m | download (3.8 KB)]]
** MATLAB: [[Media:EIA_ATGP.m | download (3.8 KB)]]
** SCILAB: [[Media:EIA_ATGP.sci | download (3.8 KB)]]
** SCILAB: [[Media:EIA_ATGP.sci | download (4.0 KB)]]
     A. Plaza and C.-I. Chang
     A. Plaza and C.-I. Chang
     ''Impact of Initialization on Design of Endmember Extraction Algorithms''
     ''Impact of Initialization on Design of Endmember Extraction Algorithms''
Línea 66: Línea 71:


* '''Convex Cone Analysis (CCA)'''
* '''Convex Cone Analysis (CCA)'''
** MATLAB: [[Media:EIA_CCA.m | download (3.5 KB)]]
** MATLAB: [[Media:EIA_CCA.m | download (3.4 KB)]]
** SCILAB: [[Media:EIA_CCA.sci | download (3.4 KB)]]
** SCILAB: [[Media:EIA_CCA.sci | download (3.6 KB)]]
     Ifarraguerri, A. and C.-I. Chang
     Ifarraguerri, A. and C.-I. Chang
     ''Multispectral and hyperspectral image analysis with convex cones''
     ''Multispectral and hyperspectral image analysis with convex cones''
     ''Geoscience and Remote Sensing, IEEE Transactions on'', vol. 37, nº. 2, págs. 756-770, 1999.
     ''Geoscience and Remote Sensing, IEEE Transactions on'', vol. 37, nº. 2, págs. 756-770, 1999.
* '''Vertex Component Analysis (VCA)'''
** MATLAB: [[Media:EIA_VCA.m | download (xxx MB)]]
** SCILAB: [[Media:EIA_VCA.sci | download (xxx MB)]]
    Nascimento, J. M. P. and Dias, J. M. B.
    ''Vertex component analysis: a fast algorithm to unmix hyperspectral data''
    ''Geoscience and Remote Sensing, IEEE Transactions on'', vol. 43, nº. 4, págs. 898-910, 2005.


=== Additional functions ===
=== Additional functions ===
Línea 84: Línea 82:
* '''Lattice Associative Memories (LAMs)'''
* '''Lattice Associative Memories (LAMs)'''
** MATLAB: [[Media:EIA_LAM.m | download (3.1 KB)]]
** MATLAB: [[Media:EIA_LAM.m | download (3.1 KB)]]
** SCILAB: [[Media:EIA_LAM.sci | download (3.0 KB)]]
** SCILAB: [[Media:EIA_LAM.sci | download (3.2 KB)]]


* '''Chebyshev distance'''
* '''Chebyshev distance'''
** MATLAB: [[Media:EIA_CHEBYSHEV.m | download (2.0 KB)]]
** MATLAB: [[Media:EIA_CHEBYSHEV.m | download (2.0 KB)]]
** SCILAB: [[Media:EIA_CHEBYSHEV.sci | download (2.0 KB)]]
** SCILAB: [[Media:EIA_CHEBYSHEV.sci | download (2.1 KB)]]


Some of the algorithms require as input the number of endmembers to search. If unknown, HFC virtual dimensionality algorithm can be used:
Some of the algorithms require as input the number of endmembers to search. If unknown, HFC virtual dimensionality algorithm can be used:
Línea 94: Línea 92:
* '''HFC method'''
* '''HFC method'''
** MATLAB: [[Media:EIA_HFC.m | download (3.2 KB)]]
** MATLAB: [[Media:EIA_HFC.m | download (3.2 KB)]]
** SCILAB: [[Media:EIA_HFC.sci | download (3.3 KB)]]
** SCILAB: [[Media:EIA_HFC.sci | download (3.4 KB)]]
     Chang, C.-I. and Du, Q.
     Chang, C.-I. and Du, Q.
     ''Estimation of number of spectrally distinct signal sources in hyperspectral imagery''
     ''Estimation of number of spectrally distinct signal sources in hyperspectral imagery''
Línea 104: Línea 102:


* '''1D Launcher''': 1-spatial data is defined as a matrix where first dimension represents the spectral information and second dimension the spatial. Examples of 1-spatial data are contingency matrix where each sample (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality).
* '''1D Launcher''': 1-spatial data is defined as a matrix where first dimension represents the spectral information and second dimension the spatial. Examples of 1-spatial data are contingency matrix where each sample (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality).
** MATLAB: [[Media:EIA_1D.m | download (xxx KB)]]
** MATLAB: [[Media:EIA_1D.m | download (7.9 KB)]]
** SCILAB: [[Media:EIA_1D.sci | download (xxx KB)]]
** SCILAB: [[Media:EIA_1D.sci | download (7.4 KB)]]


* '''2D Launcher''': 2-spatial data is defined as a cube where first dimension represents the spectral information and, second and third dimensions are the spatial ones. Examples of 2-spatial data are images where each pixel (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality). For binaries or grey-scale images N==1. For RGB images N==3. For hyperspectral images N is high.
* '''2D Launcher''': 2-spatial data is defined as a cube where first dimension represents the spectral information and, second and third dimensions are the spatial ones. Examples of 2-spatial data are images where each pixel (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality). For binaries or grey-scale images N==1. For RGB images N==3. For hyperspectral images N is high.
** MATLAB: [[Media:EIA_2D.m | download (xxx KB)]]
** MATLAB: [[Media:EIA_2D.m | download (8.1 KB)]]
** SCILAB: [[Media:EIA_2D.sci | download (xxx KB)]]
** SCILAB: [[Media:EIA_2D.sci | download (7.7 KB)]]


* '''3D Launcher''': 3-spatial data is defined as an hypercube where first dimension represents the spectral information and, second, third and fourth dimensions are the spatial ones. Examples of 3-spatial data are hyperspectral MRI images where each voxel (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality).
* '''3D Launcher''': 3-spatial data is defined as an hypercube where first dimension represents the spectral information and, second, third and fourth dimensions are the spatial ones. Examples of 3-spatial data are hyperspectral MRI images where each voxel (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality).
** MATLAB: [[Media:EIA_3D.m | download (xxx KB)]]
** MATLAB: [[Media:EIA_3D.m | download (8.1 KB)]]
** SCILAB: [[Media:EIA_3D.sci | download (xxx KB)]]
** SCILAB: [[Media:EIA_3D.sci | download (7.7 KB)]]


=== Loading hyperspectral data ===
=== Loading hyperspectral data ===


* [[Media:loadHypercubesMatlab.pdf | How to load hyperspectral images with Matlab]]
* [[Media:loadHypercubesMatlab.pdf | How to load hyperspectral images with Matlab]]

Revisión actual - 12:22 6 oct 2014

Endmember Induction Algorithms toolbox

Download the latest Endmember Induction Algorithms (EIAs) toolbox and the documentation here:

This software is distributed under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

The here available Endmember Induction Algorithms (EIAs) toolbox has been developed with MATLAB 7.4 and MATLAB 8.1 (R2013a) (licensed copy is needed to use it) and SCILAB 5.2 (open source and freely distributed).

If you are using the Endmember Induction Algorithms (EIAs) toolbox for your scientific research, please reference it as follows:

   Endmember Induction Algorithms (EIAs) toolbox.
   Grupo de Inteligencia Computacional, Universidad del País Vasco / Euskal Herriko Unibertsitatea (UPV/EHU), Spain. 
   http://www.ehu.es/computationalintelligence/index.php/Endmember_Induction_Algorithms

Copyright 2010 Grupo Inteligencia Computacional, Universidad del País Vasco / Euskal Herriko Unibertsitatea (UPV/EHU).

Acknowledgements to Prof. Gerhard Ritter from the Department of Computer and Information Science and Engineering, University of Florida (USA); Prof. Antonio Plaza from the Department of Technology of Computers and Communications, University of Extremadura (Spain), and coordinator of the Hyper-I-Net project; and to Prof. Chein-I Chang from the Remote Sensing Signal and Image Processing Laboratory, University of Maryland (USA).

Endmember Induction Algorithms collection

Here you can find separately the EIAs included in the toolbox's version 0.2 and their respective bibliographical references:

   M. Grana, I. Villaverde, J. O. Maldonado, and C. Hernandez
   Two lattice computing approaches for the unsupervised segmentation of hyperspectral images
   Neurocomput., vol. 72, nº. 10-12, págs. 2111-2120, 2009.
   M. Grana, D. Chyzhyk, M. García-Sebastián, and C. Hernández
   Lattice independent component analysis for functional magnetic resonance imaging
   Information Sciences, vol. 181, pág. 1910–1928, May. 2011.
   G. X. Ritter and G. Urcid
   A lattice matrix method for hyperspectral image unmixing
   Information Sciences, vol. In Press, Corrected Proof, Oct. 2010.
   Winter, M. E.
   N-FINDR: an algorithm for fast autonomous spectral endmember determination in hyperspectral data
   presented at the Imaging Spectrometry V, Denver, CO, USA, 1999, vol. 3753, págs. 266-275.
   Chang, C.-I. and Plaza, A.
   A fast iterative algorithm for implementation of pixel purity index
   Geoscience and Remote Sensing Letters, IEEE, vol. 3, nº. 1, págs. 63-67, 2006.
   A. Plaza and C.-I. Chang
   Impact of Initialization on Design of Endmember Extraction Algorithms
   Geoscience and Remote Sensing, IEEE Transactions on, vol. 44, nº. 11, págs. 3397-3407, 2006.
   Ifarraguerri, A. and C.-I. Chang
   Multispectral and hyperspectral image analysis with convex cones
   Geoscience and Remote Sensing, IEEE Transactions on, vol. 37, nº. 2, págs. 756-770, 1999.

Additional functions

Some of the algorithms require additional methods:

Some of the algorithms require as input the number of endmembers to search. If unknown, HFC virtual dimensionality algorithm can be used:

   Chang, C.-I. and Du, Q.
   Estimation of number of spectrally distinct signal sources in hyperspectral imagery
   Geoscience and Remote Sensing, IEEE Transactions on, vol. 42, nº. 3, págs. 608-619, 2004.

Data's spatial dimensionality-based launchers

Here you can find launchers for the EIA toolbox which treat data in base to their spatial dimensionality. This launchers are divided into 1D, 2D and 3D.

  • 1D Launcher: 1-spatial data is defined as a matrix where first dimension represents the spectral information and second dimension the spatial. Examples of 1-spatial data are contingency matrix where each sample (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality).
  • 2D Launcher: 2-spatial data is defined as a cube where first dimension represents the spectral information and, second and third dimensions are the spatial ones. Examples of 2-spatial data are images where each pixel (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality). For binaries or grey-scale images N==1. For RGB images N==3. For hyperspectral images N is high.
  • 3D Launcher: 3-spatial data is defined as an hypercube where first dimension represents the spectral information and, second, third and fourth dimensions are the spatial ones. Examples of 3-spatial data are hyperspectral MRI images where each voxel (spatial dimensionality) is an N-dimensional feature vector (spectral dimensionality).

Loading hyperspectral data