A New Subspace Approach for Supervised Hyperspectral Image Classification

1
A New Subspace Approach for Supervised
Hyperspectral Image Classification
Jun Li1,2, José M. Bioucas-Dias2 and Antonio
Plaza1 1Hyperspectral Computing
Laboratory University of Extremadura, Cáceres,
Spain 2Instituto de Telecomunicaçoes, Instituto
Superior Técnico, TULisbon, Portugal Contact
e-mails junli, aplaza_at_unex.es, bioucas_at_lx.it.pt
2
A New Subspace Approach for Hyperspectral
Classification
Talk Outline
1. Challenges in hyperspectral image
classification 2. Subspace projection 2.1.
Subspace projection-based framework 2.2.
Considered subspace projection techniques PCA
versus HySime 2.3. Integration with
different classifiers (LDA, SVM, MLR) 3.
Experimental results 3.1. Experiments with
AVIRIS Indian Pines hyperspectral data 3.2.
Experiments with ROSIS Pavia University
hyperspectral 4. Conclusions and future research
lines
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
3
Challenges in Hyperspectral Image Classification
Concept of hyperspectral imaging using NASA Jet
Propulsion Laboratorys Airborne Visible
Infra-Red Imaging Spectrometer
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
1
4
Challenges in Hyperspectral Image Classification

• Challenges in hyperspectral image classification
• Imbalance between dimensionality and training
  samples, presence of mixed pixels

Ultraspectral (1000s of bands)
Hyperspectral (100s of bands)
Multispectral (10s of bands)
Panchromatic
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
2
5
Challenges in Hyperspectral Image Classification

• Challenges in hyperspectral image classification
• The special characteristics of hyperspectral data
  pose several processing problems
• The high-dimensional nature of hyperspectral data
  introduces important limitations in supervised
  classifiers, such as the limited availability of
  training samples or the inherently complex
  structure of the data
• There is a need to address the presence of mixed
  pixels resulting from insufficient spatial
  resolution and other phenomena in order to
  properly model the hyperspectral data
• There is a need to develop computationally
  efficient algorithms, able to provide a response
  in a reasonable time and thus address the
  computational requirements of time-critical
  remote sensing applications
• In this work, we evaluate the impact of using
  subspace projection techniques prior to
  supervised classification of hyperspectral image
  data while analyzing each of the aforementioned
  items

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
3
6
A New Subspace Approach for Hyperspectral
Classification
Talk Outline
1. Challenges in hyperspectral image
classification 2. Subspace projection 2.1.
Subspace projection-based framework 2.2.
Considered subspace projection techniques PCA
versus HySime 2.3. Integration with
different classifiers (LDA, SVM, MLR) 3.
Experimental results 3.1. Experiments with
AVIRIS Indian Pines hyperspectral data 3.2.
Experiments with ROSIS Pavia University
hyperspectral 4. Conclusions and future research
lines
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
7
Subspace Projection-Based Framework

• Subspace projection-based framework.-
• Hyperspectral image data generally lives in a
  lower-dimensional subspace compared with the
  input feature dimensionality
• This can be exploited to address ill-posed
  problems given by limited training samples
• The projection into such subspaces allows us to
  specifically avoid spectral confusion due to
  mixed pixels, thus reducing their impact in the
  subsequent classification process

J. Li, J. M. Bioucas-Dias and A. Plaza,
Spectral-spatial hyperspectral image
segmentation using sub-space multinomial logistic
regression and Markov random fields, IEEE
Transactions on Geoscience and Remote Sensing, in
press, 2011.
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
4
8
Considered Subspace Projection Techniques PCA
versus HySime

• Principal Component Analysis (PCA).-
• High-dimensional data can be transformed
  effectively according to its distribution in
  feature space (e.g. by finding the most important
  directions or axes, establishing those axes as
  the references of a new coordinate system which
  takes into account data distribution)
• Orders the resulting components in decreasing
  order of variance

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
5
9
Considered Subspace Projection Techniques PCA
versus HySime

• Principal Component Analysis (PCA).-
• High-dimensional data can be transformed
  effectively according to its distribution in
  feature space (e.g. by finding the most important
  directions or axes, establishing those axes as
  the references of a new coordinate system which
  takes into account data distribution)
• Orders the resulting components in decreasing
  order of variance

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
6
10
Considered Subspace Projection Techniques PCA
versus HySime

• Hyperspectral Signal Identification by Minimum
  Error (HySime).-
• A recently developed method for subspace
  identification in remotely sensed hyperspectral
  data, which offers several additional features
  with regards to principal component analysis and
  other subspace projection techniques

J. M. Bioucas-Dias and J. M. P Nascimento,
Hyperspectral subspace identification, IEEE
Transactions on Geoscience and Remote Sensing,
vol. 46, no. 8, pp. 2435-2445, 2008.
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
7
11
Supervised Classification Framework Tested in
this Work

• Supervised Classification Framework.-
• Includes subspace projection and supervised
  classification based on training samples

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
8
12
Integration with different classifiers (LDA, SVM,
MLR)

• Integration of subspace-based framework with
  different classifiers.-
• Three different supervised classifiers tested in
  this work
• Linear discriminant analysis (LDA) find a linear
  combination of features which separate two or
  more classes the resulting combination may be
  used as a linear classifier (only linearly
  separable classes will remain separable after
  applying LDA)
• Support vector machine (SVM) constructs a set of
  hyperplanes in high-dimensional space a good
  separation is achieved by the hyperplane that has
  the largest distance to the nearest training data
  points of any class
• Multinomial logistic regression (MLR) models the
  posterior class distributions in a Bayesian
  framework, thus supplying (in addition to the
  boundaries between the classes) a degree of
  plausibility for such classes

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
9
13
A New Subspace Approach for Hyperspectral
Classification
Talk Outline
1. Challenges in hyperspectral image
classification 2. Subspace projection 2.1.
Classic techniques for subspace projection PCA
versus HySime 2.2. Subspace
projection-based framework 2.3. Integration
with different classifiers (LDA, SVM, MLR) 3.
Experimental results 3.1. Experiments with
AVIRIS Indian Pines hyperspectral data 3.2.
Experiments with ROSIS Pavia University
hyperspectral 4. Conclusions and future research
lines
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
14
Experimental Results Using Real Hyperspectral
Data Sets

• AVIRIS Indian Pines data set.-
• Challenging classification scenario due to
  spectrally similar classes
• Early growth stage of the agricultural features
  (only around 5 coverage of soil)
• 145x145 pixels, 202 spectral bands, 16
  ground-truth classes
• 10366 labeled pixels (random training subsets
  evenly distributed among classes)

False color composition
Ground-truth
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
10
15
Experimental Results Using Real Hyperspectral
Data Sets

• AVIRIS Indian Pines data set.-
• Classification results using 160 training samples
  (10 training samples per class)
• For the SVM classifier we used the Gaussian RBF
  kernel after testing other kernels
• The mean accuracies (after 10 Monte Carlo runs)
  and processing times are reported

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
11
16
Experimental Results Using Real Hyperspectral
Data Sets

• AVIRIS Indian Pines data set.-
• Classification results using 240 training samples
  (15 training samples per class)
• For the SVM classifier we used the Gaussian RBF
  kernel after testing other kernels
• The mean accuracies (after 10 Monte Carlo runs)
  and processing times are reported

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
12
17
Experimental Results Using Real Hyperspectral
Data Sets

• AVIRIS Indian Pines data set.-
• Classification results using 320 training samples
  (20 training samples per class)
• For the SVM classifier we used the Gaussian RBF
  kernel after testing other kernels
• The mean accuracies (after 10 Monte Carlo runs)
  and processing times are reported

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
13
18
Experimental Results Using Real Hyperspectral
Data Sets

• AVIRIS Indian Pines data set.-
• Classification results using 320 training samples
  (20 training samples per class)

SVM (OA65.36)
Subspace SVM (OA70.33)
LDA (OA50.74)
Subspace LDA (OA54.90)
Linear MLR (OA60.38)
Subspace MLR (OA67.53)
Ground-truth
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
14
19
Experimental Results Using Real Hyperspectral
Data Sets

• ROSIS Pavia University data set.-

Overall classification accuracies and kappa
coefficient (in the parentheses) using different
training sets for the ROSIS Pavia University
IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
15
20
Conclusions and Hints at Plausible Future Research

• Conclusions and Future Lines.-
• We have evaluated the impact of subspace
  projection on supervised classification of
  remotely sensed hyperspectral image data sets
• Two dimensionality reduction methods have been
  used PCA and HySime, although many others are
  available and could be used MNF, OSP, VD
• Three different supervised classifiers
  considered LDA, SVM, MLR
• Experimental results indicate that different
  approaches for hyperspectral image classification
  approaches can benefit from subspace projection,
  particularly when very limited training samples
  are available
• Subspace projection can be naturally integrated
  with multinomial logistic regression (MLR)
  classifiers, which greatly benefit from
  dimensionality reduction
• Future work will focus on the evaluation of other
  subspace projection approaches and hyperspectral
  data sets

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
16
21

• IEEE J-STARS Special Issue on Hyperspectral Image
  and Signal Processing

IEEE International Geoscience and Remote Sensing
Symposium (IGARSS 2011), Vancouver, Canada, July
24 29, 2011
17
22
A New Subspace Approach for Supervised
Hyperspectral Image Classification
Jun Li1,2, José M. Bioucas-Dias2 and Antonio
Plaza1 1Hyperspectral Computing
Laboratory University of Extremadura, Cáceres,
Spain 2Instituto de Telecomunicaçoes, Instituto
Superior Técnico, TULisbon, Portugal Contact
e-mails junli, aplaza_at_unex.es, bioucas_at_lx.it.pt