Contribution to Passive Acoustic Ocean Tomography- Part IV: A Data Fusion Strategy For Blind Source Separation and Classification
Résumé
In this paper we applied Fusion algorithm in a passive Ocean Acoustic Tomography context. We applied several types of independent component analysis algorithms an a large number of realistic data and we propose to classify their respective outputs according to the principal properties of the opportunity sources. Finally, we apply a fusions model.. the Belief functions theory to outperform the Blind Source Separation/classification approach.