Morphological Principal Component Analysis for Hyperspectral Image Analysis

Gianni Franchi (1) , Jesus Angulo (1)

1 CMM - Centre de Morphologie Mathématique

This paper deals with a problem of dimensionality reduction for hyperspectral images using the principal component analysis. Hyper-spectral image reduction is improved by adding structural/spatial information to the spectral information, by means of mathematical morphology tools. Then it can be useful in supervised classification for instance. The key element of the approach is the computation of a covariance matrix which integrates simultaneously both spatial and spectral information. Thanks to these new covariance matrices, new features can be extracted. To prove the efficiency of these new features we have conducted an extended study showing the interest of the structural/spatial information.

Mots clés: multivariate image; hyperspectral images; dimensionality reduction; mathematical morphology; nonlinear image processing

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement des images [eess.IV]; Vision par ordinateur et reconnaissance de formes [cs.CV]

Gianni Franchi, Jesus Angulo. Morphological Principal Component Analysis for Hyperspectral Image Analysis. ISPRS International Journal of Geo-Information, 2016, 5 (6), pp.83. ⟨10.3390/ijgi5060083⟩. ⟨hal-01256379⟩