Journal of Spectral Imaging,   Volume 5   Article ID a3   (2016)

Peer reviewed Paper

Model-based co-clustering for hyperspectral images

  • Julien Jacques  
  • Cyril Ruckebusch
Université de Lille, Sciences et technologies, LASIR CNRS, Lille, France
[email protected]
 Search for papers by this author
 Corresponding Author
Université Lyon, Lumière Lyon 2, ERIC, Lyon, France
[email protected]
 Search for papers by this author

A model-based co-clustering algorithm for hyperspectral images is presented. This algorithm, which relies on the probabilistic latent block model for continuous data, aims to cluster both the pixels and the spectral features of the images. This approach has been applied to a benchmark Raman imaging dataset and revealed relevant information for spatial–spectral exploratory investigation of the data.

Keywords: co-clustering, latent block model, hyperspectral images




Abstract Views: