Sparse reconstruction techniques for near-field underwater acoustic imaging

Nikolaos Stefanakis 1, * Jacques Marchal 2 Valentin Emiya 3 Nancy Bertin 1 Rémi Gribonval 4 Pierre Cervenka 2
* Corresponding author
1 METISS - Speech and sound data modeling and processing
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
4 PANAMA - Parcimonie et Nouveaux Algorithmes pour le Signal et la Modélisation Audio
Inria Rennes – Bretagne Atlantique , IRISA-D5 - SIGNAUX ET IMAGES NUMÉRIQUES, ROBOTIQUE
Abstract : The use of sparse priors has shown interesting potential in various acoustic or radar imaging applications. In this paper, sparse reconstruction is applied for underwater acoustic imaging using a newly built flexible sonar device. We investigate several models concerning the linear mapping between the image domain and the observation domain. In particular, we define a point-scatterer model in which the apparent back-scatter coefficient of a given reflector varies with respect to the specific emitter and receiver locations. To handle this problem, we adapt a multi-channel version of the orthogonal matching pursuit and we apply it on real data in order to obtain images of an underwater target placed at a small distance from the sonar. The techniques are shown to overcome bottlenecks that are apparent with more standard approaches that assume far-field conditions when building the image.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-00810911
Contributor : Hal System <>
Submitted on : Monday, April 23, 2012 - 10:00:00 AM
Last modification on : Tuesday, April 2, 2019 - 2:34:20 AM
Document(s) archivé(s) le : Sunday, December 18, 2016 - 4:03:40 PM

File

hal-00810911.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-00810911, version 1

Citation

Nikolaos Stefanakis, Jacques Marchal, Valentin Emiya, Nancy Bertin, Rémi Gribonval, et al.. Sparse reconstruction techniques for near-field underwater acoustic imaging. Acoustics 2012, Apr 2012, Nantes, France. pp.1617. ⟨hal-00810911⟩

Share

Metrics

Record views

1526

Files downloads

188