BLIND SOURCE SEPARATION OF ASTRONOMICAL IMAGES
Danielle Nuzillard, Albert Bijaoui
Danielle.Nuzillard@univreims.fr, Bijaoui@obsnice.fr
Celestial sources are studied by means of multispec
tral analysis, for which a pixel value can be consid
ered as a mixture of different sources. We show that
the application of blind source separation (BSS) meth
ods can display interesting features, which could help
the astronomers to improve their description of celes
tial sources. Experiments were done on Hubble Space
Telescope (HST) images of the galaxy 3C120. Different
ways were applied for the separation: tools based on the
correlation between the shifted sources, and tools based
on nonGaussian probability density functions (PDF).
Each obtained source set was quantified by the mutual
information between the sources. Different algorithms
give quite similar decompositions which mainly corre
spond to real phenomena.