Mahalanobis kernel for the classification of hyperspectral images

M. Fauvel, A. Villa, J. Chanussot, J. A. Benediktsson

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The definition of the Mahalanobis kernel for the classification of hyperspectral remote sensing images is addressed. Class specific covariance matrices are regularized by a probabilistic model which is based on the data living in a subspace spanned by the p first principal components. The inverse of the covariance matrix is computed in a closed form and is used in the kernel to compute the distance between two spectra. Each principal direction is normalized by a hyperparameter tuned, according to an upper error bound, during the training of an SVM classifier. Results on real data sets empirically demonstrate that the proposed kernel leads to an increase of the classification accuracy by comparison to standard kernels.

Original languageEnglish
Title of host publication2010 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2010
Pages3724-3727
Number of pages4
DOIs
Publication statusPublished - 2010
Event2010 30th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2010 - Honolulu, HI, United States
Duration: 25 Jul 201030 Jul 2010

Publication series

NameInternational Geoscience and Remote Sensing Symposium (IGARSS)

Conference

Conference2010 30th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2010
Country/TerritoryUnited States
CityHonolulu, HI
Period25/07/1030/07/10

Other keywords

  • Classification
  • Hyper-spectral images
  • Mahalanobis kernel
  • Probabilistic principal component analysis
  • Support vector machine

Fingerprint

Dive into the research topics of 'Mahalanobis kernel for the classification of hyperspectral images'. Together they form a unique fingerprint.

Cite this